Academia.eduAcademia.edu
I THE CAMBRIDGE HISTORY OF CHINA General Editors D E N I S TWITCHETT and J O H N K. FAIRBANK Volume 6 Alien regimes and border states, 907—1368 Cambridge Histories Online © Cambridge University Press, 2008 Work on this volume was partially supported by the National Endowment for the Humanities, Grants RO-21512-87 and RO-22077-90. Cambridge Histories Online © Cambridge University Press, 2008 THE CAMBRIDGE HISTORY OF CHINA Volume 6 Alien regimes and border states, 907—1368 edited by HERBERT FRANKE and DENIS TWITCHETT CAMBRIDGE UNIVERSITY PRESS Cambridge Histories Online © Cambridge University Press, 2008 CAMBRIDGE UNIVERSITY PRESS Cambridge, New York, Melbourne, Madrid, Cape Town, Singapore, Sao Paulo Cambridge University Press 40 West 20th Street, New York, NY 10011-4211, USA www.cambridge.org Information on this title:www.cambridge.org/9780521243315 © Cambridge University Press 1994 This publication is in copyright. Subject to statutory exception and to the provisions of relevant collective licensing agreements, no reproduction of any part may take place without the written permission of Cambridge University Press. First published 1994 Reprinted 2002, 2006 Printed in the United States of America A catalogue recordfor this book is availablefromthe British Library. Library of Congress Cataloguing in Publication Data (Revised for vol. 6) The Cambridge history of China. Vol 1. edited by Denis Twitchett and Michael Loewe; v. 6 edited by Herbert Franke and Denis Twitchett; v. 7 edited by Frederick W. Mote and Denis Twitchett; v. 11 edited by John K. Fairbank and Kwang-Ching Liu; v. 13 edited by John K. Fairbank and Albert Feuerwerker; v. 14 edited by Roderick MacFarquhar and John K. Fairbank; v. 15 edited by Roderick MacFarquhar and John K. Fairbank. Includes bibliographies and indexes. Contents: v 1. The Ch'in and Han Empires, 221 B.C.—AD. 220 — v. 3. Sui and T'ang China, 589—906, pt. 1 — —v. 6 Alien regimes and border states, 710—1368. 1. China — History. I. Twitchett, Denis Crispin. II. Fairbank, John King, 1907—1991. DS735.C3145 951' .03 76—29852 ISBN-13 978-0-521-24331-5 hardback ISBN-10 0-521-24331-9 hardback Cambridge University Press has no responsibility for the persistence or accuracy of URLs for external or diird-party Internet Web sites referred to in this publication and does not guarantee that any content on such Web sites is, or will remain, accurate or appropriate. Cambridge Histories Online © Cambridge University Press, 2008 GENERAL EDITORS' PREFACE When The Cambridge History of China was first planned, more than two decades ago, it was naturally intended that it should begin with the very earliest periods of Chinese history. However, the production of the series has taken place over a period of years when our knowledge both of Chinese prehistory and of much of the first millennium B.C. has been transformed by the spate of archaeological discoveries that began in the 1920s and has been gathering increasing momentum since the early 1970s. This flood of new information has changed our view of early history repeatedly, and there is not yet any generally accepted synthesis of this new evidence and the traditional written record. In spite of repeated efforts to plan and produce a volume or volumes that would summarize the present state of our knowledge of early China, it has so far proved impossible to do so. It may well be another decade before it will prove practical to undertake a synthesis of all these new discoveries that is likely to have some enduring value. Reluctantly, therefore, we begin the coverage of The Cambridge History of China with the establishment of the first imperial regimes, those of Ch'in and Han. We are conscious that this leaves a millennium or more of the recorded past to be dealt with elsewhere, and at another time. We are equally conscious of the fact that the events and developments of the first millennium B.C. laid the foundations for the Chinese society and its ideas and institutions that we are about to describe. The institutions, the literary and artistic culture, the social forms, and the systems of ideas and beliefs of Ch'in and Han were firmly rooted in the past and cannot be understood without some knowledge of this earlier history. As the modern world grows more interconnected, historical understanding of it becomes ever more necessary and the historian's task ever more complex. Fact and theory affect each other even as sources proliferate and knowledge increases. Merely to summarize what is known becomes an awesome task, yet a factual basis of knowledge is increasingly essential for historical thinking. Since the beginning of the century, the Cambridge histories have set a pattern in the English-reading world for multivolume series containing chapters written by specialists under the guidance of volume editors. The Cambridge Histories Online © Cambridge University Press, 2008 vi GENERAL EDITORS' PREFACE Cambridge Modern History, planned by Lord Acton, appeared in sixteen volumes between 1902 and 1912. It was followed by The Cambridge Ancient History, The Cambridge Medieval History, The Cambridge History of English Literature, and Cambridge histories of India, of Poland, and of the British Empire. The original Modern History has now been replaced by The New Cambridge Modern History in twelve volumes, and The Cambridge Economic History of Europe is now being completed. Other Cambridge histories include histories of Islam, Arabic literature, Iran, Judaism, Korea, South East Asia, Central Asia, Africa, Japan, and Latin America. In the case of China, Western historians face a special problem. The history of Chinese civilization is more extensive and complex than that of any single Western nation, and only slightly less ramified than the history of European civilization as a whole. The Chinese historical record is immensely detailed and extensive, and Chinese historical scholarship has been highly developed and sophisticated for many centuries. Yet until recent decades the study of China in the West, despite the important pioneer work of European sinologists, had hardly progressed beyond the translation of some few classical historical texts, and the outline history of the major dynasties and their institutions. Recently Western scholars have drawn more fully upon the rich traditions of historical scholarship in China and also in Japan, and greatly advanced both our detailed knowledge of past events and institutions, and also our critical understanding of traditional historiography. In addition, the present generation of Western historians of China can also draw on the new outlooks and techniques of modern Western historical scholarship and on recent developments in the social sciences, while continuing to build on the solid foundations of rapidly progressing European, Japanese, and Chinese studies. Recent historical events, too, have given prominence to new problems, while throwing into question many older conceptions. Under these multiple impacts the Western revolution in Chinese studies is steadily gathering momentum. When The Cambridge History of China was first planned in 1966, the aim was to provide a substantial account of the history of China as a benchmark for the Western history-reading public: an account of the current state of knowledge in six volumes. Since then the outpouring of current research, the application of new methods, and the extension of scholarship into new fields have further stimulated Chinese historical studies. This growth is indicated by the fact that the history has now become a planned fifteen volumes but will still leave out such topics as the history of art and of literature, many aspects of economics and technology, and all the riches of local history. The striking advances in our knowledge of China's past over the last decade will continue and accelerate. Western historians of this great and Cambridge Histories Online © Cambridge University Press, 2008 GENERAL EDITORS PREFACE VII complex subject are justified in their efforts by the needs of their own peoples for greater and deeper understanding of China. Chinese history belongs to the world not only as a right and necessity, but also as a subject of compelling interest. JOHN K. FAIRBANK DENIS TWITCHETT Cambridge Histories Online © Cambridge University Press, 2008 Cambridge Histories Online © Cambridge University Press, 2008 CONTENTS General editors' preface List of maps, tables, and Preface to Volume 6 List of abbreviations figures Introduction page v xiv xvii xix i by H E R B E R T F R A N K E , Universitat Muncben, Emeritus, and D E N I S T W I T C H E T T , Princeton University, Emeritus The Late T'ang balance of power The frontier Foreigners Vassals and overlords Multistate system Treaty relations Modes of government Multilinguality The Han Chinese under alien domination i The Liao 3 7 11 14 16 18 21 30 36 43 by D E N I S T W I T C H E T T and KLAUS-PETER TIETZE Introduction The predynastic Khitan The background of A-pao-chi's rise to power The rise of A-pao-chi A-pao-chi becomes the new khaghan and ascends the throne The succession crisis and the reign of T'ai-tsung The succession of Shih-tsung The development of government institutions Relation with regimes in China The reign of Mu-tsung, 951—969 IX Cambridge Histories Online © Cambridge University Press, 2008 43 44 53 56 60 68 75 76 80 81 CONTENTS The reign of Ching-tsung, 969—982: confrontation with Sung The regency of Empress Dowager Ch'eng-t'ien Changes in internal government Foreign relations The reign of Hsing-tsung The reign of Tao-tsung The reign of T'ien-tso and the collapse of the Liao Final disaster The Hsi Hsia by R U T H D U N N E L , Kenyon College Introduction The ethnogenesis of the Tangut people The surrender to T'ang and settlement in the Ordos The end of the T'ang The Tanguts during the Wu-tai period The Tangut move toward independence, 982—1002 Liang-chou and Tangut expansion into Ho-hsi The rise of the Tibetan Tsung-ko dynasty Li Te-ming, 1004-1032 Ho-hsi after the Tangut conquest Li Yiian-hao (Wei-ming Nang-hsiao, Ching-tsung), 10321048 The succession to Wei-ming Yiian-hao A state in peril: the reigns of I-tsung (1048-1068), Huitsung (1068-1086), and Ch'ung-tsung to 1100 The reign of Hui-tsung (1068-1086) Hsia comes of age: Ch'ung-tsung (1086—1139) and Jentsung (1140-1193) The last years of the Hsia state and the Mongolian conquest The Chin dynasty by HERBERT 84 87 91 98 114 123 139 149 154 154 155 158 161 164 168 172 173 176 179 180 189 191 193 197 205 215 FRANKE General remarks The Jurchen people and their predynastic history The reign of A-ku-ta and the founding of the Chin dynasty From war to coexistence: Chin—Sung relations before the treaty of 1142 The political history of Chin after 1142 The annihilation of Chin, 1215-1234 Cambridge Histories Online © Cambridge University Press, 2008 215 216 220 226 235 259 CONTENTS Government structure Social structure Ethnic groups Economic conditions Scholarship, literature, and the arts Religious life Conclusion 4 The rise of the Mongolian empire and Mongolian rule in north China by THOMAS A L L S E N , Trenton State College Mongolia and Temiijin, ca. 1150—1206 Chinggis khan and the early Mongolian state, 1206-1227 The organization of the empire: the reigns of Ogodei and Giiyiig The empire at its apogee: the reign of Mongke, 1251-1259 Epilogue: The empire on the eve of civil war XI 265 277 279 291 304 313 319 321 321 342 365 390 411 5 The reign of Khubilai khan 414 by M O R R I S ROSSABI, City University of New York, Columbia University The early years Khubilai and China, 1253—1259 Khubilai versus Arigh Boke Foreign expansion Social and economic policies Khubilai as emperor of China Khubilai and religion Khubilai and Chinese culture Preservation of the Mongolian heritage Economic problems in later years The regime of Sangha and economic and religious abuses Disastrous foreign expeditions Khubilai's last years 6 Mid-Yuan politics by H S I A O C H ' I - C H ' I N G , University of Singapore Introduction The reign of Temiir khaghan (Emperor Ch'eng-tsung), 1294-1307 Cambridge Histories Online © Cambridge University Press, 2008 414 418 422 429 445 454 457 465 471 473 478 482 488 490 490 492 Xii CONTENTS Temiir's succession The conservator of Khubilai's achievements Signs of decline The transition to peace The dominance of Empress Bulukhan The reign of Khaishan (Emperor Wu-tsung), 1307-1311 Administrative anomalies The "new deals" The reign of Ayurbarwada khaghan (Emperor Jen-tsung), 494 496 498 501 504 505 507 510 1311—1320 513 The reign of Shidebala khaghan (Emperor Ying-tsung), 1320-1323 The coup d'etat at Nan-p'o The reign of Yesiin Temiir khaghan (Emperor T'ai-ting), 1323-1328 The reign of Tugh Temur (Emperor Wen-tsung), 13281332 The failure of the succession arrangements The period in retrospect 7 Shun-ti and the end of Yuan rule in China • by J O H N D A R D E S S , University of Kansas Yuan China at the accession of Toghon Temur (Shun-ti) ToghSn Temiir's enthronement and Bayan's chancellorship, 1333-134° Toghto and his opposition, 1340-1355 The disintegration of the Yuan Conclusion: Why did the Yuan dynasty fall? 8 The Yuan government and society by E L I Z A B E T H E N D I C O T T - W E S T , Harvard University Government Society 9 Chinese society under Mongol rule, 1215-1368 by F R E D E R I C K W. M O T E , Princeton University, Emeritus The Mongolian period in Chinese history The population of Yiian China Social—psychological factors Social classes: traditional and new elites Cambridge Histories Online © Cambridge University Press, 2008 527 532 535 541 556 557 561 561 566 572 580 584 587 587 608 616 616 618 622 627 CONTENTS Confucian households Diffusion of elite roles The meaning of the Yiian drama in Yiian period social history Cultural diversity The Western Asians' changing relationship to the Chinese elite Social classes: sub-elite and commoner Other aspects of Yiian dynasty social history Households in bondage Xlll 635 638 640 643 644 648 657 661 Bibliographical essays Bibliography 665 727 Glossary-Index 777 Cambridge Histories Online © Cambridge University Press, 2008 MAPS, TABLES, AND FIGURES Maps 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 The Khitan and north China, A.D. 908 The Khitan and north China, A.D. 924 The Khitan and north China, A.D. 943 The Khitan and north China, A.D. 959 The Liao campaigns against Koryo, 993-1019 The Liao invasion of Sung, 1004 The Liao empire, ca. 1045 The fall of Liao and the Chin invasions, 1017-24 The Hsi Liao (Khara Khitan) empire, 1131 —1213 The growth of the Hsi Hsia state The Tsung-ko region The Hsi Hsia state, 1111 The first Mongolian invasion of Hsi Hsia, 1209 Chinggis's invasion of Hsi Hsia, 1226-7 The Chin invasions of Northern Sung Chin raids south of the Yangtze, 1129-30 The Chin empire The change of the Huang-ho's course, 1194 Chinggis's campaigns against Chin The rebellions in Manchuria and Shantung Mukhali's campaigns against Chin The destruction of Chin, 1234 Population distribution in Chin, ca. 1211 The steppe world, 1190 The campaigns in Manchuria, 1211 —16 Chinggis's western campaigns xiv Cambridge Histories Online © Cambridge University Press, 2008 page 55 65 71 83 101 106 118 145 152 171 174 184 209 2I2 228 231 236 246 253 255 260 262 280 322 353 355 MAPS, TABLES, A N D FIGURES XV 27 28 29 Batu's invasion of Russia and Europe The Mongolian invasions of Koryo The conquest of Ta-li 369 371 406 30 31 The Mongolian khanates The conquest of Southern Sung 425 432 32 33 34 The Yuan empire The Yuan capital, Ta-tu The Yuan grain transport system 438 456 477 35 36 Khubilai's invasions of Japan Khubilai's invasions of Southeast Asia 483 486 37 Registered population distribution of Yuan China 619 Tables 1 Liao emperors and their regnal titles 2 Hsia emperors and regnal titles 3 Chin emperors and regnal titles 4 Regnal titles of Mongolian rulers 5 Capital cities 6 The Yao-lieh khaghans 7 Total population of Chin 8 Numbers of households registered in Hopei and Shantung xxii xxiv xxvi xxviii xxix 51 278 621 Figures 1 2 3 A-pao-chi's descent and the structure of the Yeh-lii clan Outline genealogy of Liao Genealogy of the Hsia ruling house xx xxi xxiii 4 5 6 Genealogy of the early Jurchen rulers Genealogy of the Chin emperors Genealogy of Mongolian rulers xxv xxv xxvii Cambridge Histories Online © Cambridge University Press, 2008 Cambridge Histories Online © Cambridge University Press, 2008 PREFACE TO VOLUME 6 Names and terms in Volume 6 ofThe Cambridge History of China are transliterated as follows: Chinese is romanized according to the Wade-Giles system, which, for all its imperfections, is employed in most of the serious literature on China written in English. There are a few exceptions, which are noted below. Japanese is romanized according to the Hepburn system. Mongolian is transliterated according to the system found in A. Mostaert, Dictionnaire ordos, vol. 3, "Index des mots du mongol ecrit et du mongol ancien" (Peiping, 1944). The only deviations from the Mostaert system are q, which becomes kh; y, which becomes gh; Z, which becomes ch; /, which becomes sh; and / , which becomes /'. Tibetan is transliterated according to the system used in Sarat Chandra Das, A Tibetan—English Dictionary (Calcutta, 1902). Persian is transliterated according to the Library of Congress system. Turkic is transliterated according to the system found in V. M. Nadeliaev et al., Drevnetiurkskii slovar' (Leningrad, 1969), with the following modifications: y becomes gh; if becomes ch; and / becomes sh. Chinese and Japanese personal names follow their native form, that is, with the surname preceding the given name. In the case of Chinese and Japanese authors of Western-language works, the names are given in the published form, in which the given name may sometimes precede the surname (e.g., Hok-lam Chan), and the orthography may employ a system other than Wade-Giles. Chinese place names are romanized according to the Wade—Giles system, with the exception of those places familiar in the English-language literature in nonstandard postal spellings. For a list of these, see G. William Skinner, Modern Chinese Society: A Critical Bibliography (Stanford: Stanford University Press, 1973), vol. 1, Introduction, p. xii. Modern place names are generally not hyphenated (e.g., Hopei is the modern province of that name), whereas contemporary place names are hyphenated (Ho-pei, Hsi-lu, the Chin period circuit). The maps are based on the standard historical atlas of China: T'an Ch'ihsiang, ed., Chung-kuo li-shih ti-t'u chi, 2nd rev. ed. (Shanghai: Chung-hua xvn Cambridge Histories Online © Cambridge University Press, 2008 xviii PREFACE TO VOLUME 6 ti-t'u hsiieh she, 1974—6; Peking: Ti-t'u ch'u-pan she, 1980—1), vol. 6, covering Sung, Liao, Hsi-Hsia, and Chin, and vol. 7, covering the Yuan period. Maps 23 and 37 are adapted from the Bulletin of the Museum of Far Eastern Antiquities, 59 (1987), pp. 214, 215. The Chinese official titles generally follow Charles O. Hucker's A Dictionary of Official Titles in Imperial China (Stanford: Stanford University Press, 1985). However, for our period, this work is incomplete in some respects. The usage of titles under all the regimes covered in this volume was constantly changing, and the reader should remember that the same office title was often used by the Sung, Liao, Hsia, Chin, and Yuan with somewhat different connotations, sometimes necessitating differing English translations. Emperors are referred to by their temple names during their reign and by their personal names before their succession to the throne. Lists of emperors, giving the various regnal titles used during their reigns, are provided in Tables 1-4. Dates normally follow the Chinese form rather than the Western calendar. The reader should be aware that the Chinese year is normally converted to a corresponding Western year but that these do not correspond exactly. Thus, for example, the Treaty of Shan-yuan was concluded in the Chinese year equivalent to 1004, though this was actually 24 January 1005. ACKNOWLEDGMENTS The editors of this volume have faced great problems arising from the complexity of the period and its sources, and the variety of languages and cultures that impinge on its story. We wish to express out gratitude to all those members of the international scholarly community whose advice we have requested on points of detail and to thank them for their careful responses. In particular, we wish to acknowledge the assistance of Professor Frederick W. Mote, who read and commented in detail on all the contributions to this volume, and the editorial work of Dr. James Geiss, who devoted a decade of meticulous scholarly attention to this and other volumes of The Cambridge History of China. We also wish to thank Professor Elizabeth Endicott-West, who not only contributed her own chapter to this volume but also assisted the editors in ensuring that the usage of Mongolian, Turkic, Tibetan, and Persian terms is uniform throughout, and Mrs. Soo-won Kim, who advised us on Korean problems. Preparing this volume took several years, and our work was made possible by the generous support of the National Endowment for the Humanities and by Princeton University. DCT HF Cambridge Histories Online © Cambridge University Press, 2008 ABBREVIATIONS BIHP CS CTKC CTS HCP HJAS HTS HYS JAS KHCPTS LS SKCS SPPY SPTK SS THY TSCC YS YTC Bulletin of the Institute of History and Philology (Academia Sinica) (Chung yang yen chiu yiian, Li shih yii yen yen chiu so chi K'an) Chin shih Ch'i-tan kuo chih Chiu Tang shu Hsu Tzu-chih t'ung-chien ch'ang-pien HarvardJournal of Asiatic Studies Hsin Tang shu K'o Shao-min, Hsin Yuan shih Journal of Asian Studies Kuo hsiieh chi pen ts'ung shu Liao shih Ssu k'u ch'iian shu chen pen Ssupupeiyao Ssu pu ts'ung k'an Sung shih Tang hui yao Ts'ung shu chi ch'eng Yuan shih Ta Yiian sheng cheng kuo ch'ao tien chang xix Cambridge Histories Online © Cambridge University Press, 2008 Nou-li-ssu* (Su-tsu) I I Chih-shen* Ti-lu-ku* I I (died young) Shu-la Yun-tt-shib* (Hsiian-tsu) (murdered by T'ieh-la) Tiieh-la* /I (died young) I Tu-lu-chiin* / Yen-ku-chih* / (driven from / office by / Hsia-li Ko-la Sa-la-U" (l-tsu) I1 Hsia-ti# y I S X y I T A-pao-chi* ,-~~ Yen-mu # (three times) Ma-lu ^ ,T-ieh-li-te» Niao-ku-chih i i La r i _ Shih-lu i Sa-la-ti* (Te-tsu) Yin-ti-shih Tieh-la An-tuan The Five Divisions The Six Divisions The First Patriarchal Household The Second Patriarchal Household The Horizontal Tents The Third Patriarchal Household Wu yuan ssu Liu yiian ssu Mmg-fu fang Cbung-fu fang Heng cbang Cbi-fu fang Note: * designates members serving as chieftain (i-li-chin). Some, as shown, served more than one term. Names in italics are A-pao-chi's paternal ancestors, subsequently canonized with imperial titles. No dates can be assigned; Nou-li-ssu was contemporary with the An-Lu-shan uprising (755). F I G U R E i . A-pao-chi's descent and the structure of the Yeh-lii clan Cambridge Histories Online © Cambridge University Press, 2008 Su Sa-la-ti (Te-tsu) Lou-kuo YUAN (Shih-tsung) (947-51) | CHING (Mu-tsung) (951-69) HSIEN (Ching-tsung) J969-82) | 6. LUNG-HSU (Sheng-tsung) roo^-tnln i (982-1051) Lung-ch'ing Li-ko 1 T'ien-te P'en-tu Ti-lieh Cha-ko Pi-she Hsi-yin Yao-shih-nu (died young) Pieh-ku-te A-lien I Yiian Chiin C. YA-LI 9. YEN-HSI (Tien-iso) (1101-1125)1 Yen-sa-ko Lung-yu ' Shih-tu 1 Ya-li-kuo Su An-tuan Chih-mo Chung-yuan 7. TSUNG-CHEN (Hsing-tsung)l /1/12 1 <<\ (1031-55) fr J Ho-lu-wo HUNG-CHI (Tao-tsung) I (1055-1101)1 ' | Yin-ti-shih Li-hu TE-KUANG (Tai-tsung) (927-47) |— Pei h.a. 917 Hou-a-pu (died young) Tieh-la La-ko A-PAO-CHI (Tai-tsu) (907-26) [- (Prince of Liang, enthroned 1123) Wu'-ke Kou-erh Hou-ka 1 Nieh-lu-ku I A. Ch'un (Hsiian-tsung) "Emperor" of Pei-Liao 1122 Ta-lu Hsi-ni-lieh (died young) Ao*lu-wo (Prince of Chin) Note: There is great confusion about the sons of T'ien-tso. The table shows six sons as listed in the royal genealogy, LS. 64, pp. 994-7. 1-9 A,B,C, Successors to the Liao throne Were all briefly enthroned and given regnal titles during the last years of T'ien-tso. FIGURE 2. Outline genealogy of Liao Cambridge Histories Online © Cambridge University Press, 2008 B. TING (Prince of Ch'in, enthroned 1122) Ning Wan TABLE I Liao emperors and their regnal titles" T'ai-tsu (r. 9O7-26) b Shen-ts'e1 T'ien-tsan T'ien-hsien T'ien-hsien Hui-t'ung Ta-t'ung T'ien-lu Ying-li Pao-ning Ch'ien-heng T'ung-ho K'ai-t'ai T'ai-p'ing Ching-fu Ch'ung-hsi Ch'ing-ning Hsien-yung T'ai-k'ang' Ta-an Shou-lung' Ch'ien-t'ung T'ien-ch'ing Tai-tsung (r. 927-47) Shih-tsung (r. 947—51) Mu-tsung (r. 951—69) Ching-tsung (r. 969—82) Sheng-tsung (r. 982-1031) Sheng-tsung cont'd Hsing-tsung (r. 1031-55) Tao-tsung (r. 1055-1101) T'ien-tso (r. 1101— 25) 916 922—26 926 926-38 938-47 947d 947-51 951-69 969-79 979-83 983-1012 1012—21 1021-31 1031—2 1032-55 1055-65 1065-75 1075-85 1085-95 Pao-ta 1095-1101 IIOI—II 1111—21 1121—5 Chien-fu 1122 Hsuan-tsung (reigned in Southern Capital, 1122) "This table lists information from Liao shih, chaps. 1-30. Ch'i-tan kuo-chih has various differences. See the note in Arthur C. Moule, The rulers of China (London, 1957), pp. 91—3 and table, p. 97. T'ai-tsu's accession appears twice in Liao shih, in 907 and 916. Probably 907 is the year when he became paramount leader of the Ch'i-tan, and 916 is the date when he became ruler of a Chinese-style Ch'i-tan state. T'ai-tsu's accession and the dynastic founding are dated 916 by Ch'i-tan kuo-chih. Before that date, Liao shih simply numbers the years. There is some doubt whether Shen-ts'e and T'ien-tsan ever existed: They may have been invented later to push back the date of the independent Ch'i-tan state to 916. Ch'i-tan kuochih dates T'ien-hsien as 927 to 937. ^h'i-tan kuo-chih omits Ta-t'ung and dates T'ien-lu as 948 to 951. 'Ch'i-tan kuo-chih omits T'ai-k'ang and Ta-an. 'Shou-ch'ang in Ch'i-tan kuo-chih. XX11 Cambridge Histories Online © Cambridge University Press, 2008 #1—10: Holders of the Ting-nan military governorship at Hsia-chou. Ssu-chung ^—^^^^—^— ? ^ — ^ — ^ — T'o-pa Ssu-kung ^ — — Ssu-chien I I Jen-yen 1 I (d. 895) Li Jen-fu 4 I-ching I-yin I I-ch'ao 6| (C'o (Kuang) -yen 2 Jen-yu I-ch'ang 5 3 K'o-jui I I Chi-p'eng 9 Chi-yun 8 Li Chi-chien (963-1004): Tai-tsu 10 Li Te-ming (983-1032): Tai-tsung I Wei-ming Nang-hsiao (Li Yuan-hao): Ching-tsung, r. 1032-48 I Liang-tso: 1-tsung, r. 1048-67 I I Ch'ien-shun: Ch'ung-tsung, r. Ping-ch'ang: Hui-tsung, r. 1068-86 1086-1139 1 > 1 Jen-hsiao : Jen-tsung, r. 1139-93 1 j 1 1 Yen-tsung Huan-tsung Ch'un-yu: r. 1193-1206 Jcn-yu l 1 An-ch'iian : Hsiang-tsung r. 1206-11 Tsun-hsiang: Shen-tsung, r. 1211-23 Te-wang: Hsien-tsung, r. 1223-6 Ch'ing-p'ing chun-wang 1 1 Hsien, r . 1226-7 Modified from Wu Tien-ch'ih, Hsi-hsia shih kao (1983), p. 292. FIGURE 3 . Genealogy of the Hsia ruling house xxm Cambridge Histories Online © Cambridge University Press, 2008 TABLE 2 Hsia emperors and regnal titles Ching-tsung (r. 1032-48) Hsien-tao (1032) Kuang-yiin (103;) (originally K'ai-yiin) Ta-ch'ing (1036-8) T'ien-shou li-fe yen-tso (1038-48) I-csung (r. 1048—67) Yen-ssu ning-kuo (1049) T'ien-yu ch'ui-sheng (1050—2) Fu-shcng ch'eng-tao (1053—6) Ch'an-tu (1057—62) Kung-hua (1063-7) Hui-tsung (r. 1068-86) Ch'ien-tao (1068—9) T'ien-ssu li-sheng kuo-ch'ing (io69~?iO74) Ta-an (1074—84?) T'ien-an li-ting (? 1085-6) Ch'ung-csung (r. 1086— 1139) T'ien-i chih-p'ing (?io86—9) T'ien-yu min-an ( 1 0 9 0 - 7 ) Yung-an (1098— 1100) Chen-kuan (1101 —13) Yung-ning (1114-18) Yuan-te (1119—26) Cheng-te (1127-34) Ta-te (1135-9) Jen-tsung (r. 1139-93) Ta-ch'ing (1140—3) Jen-ch'ing (1144-8) T'ien-sheng (1149—?! 169) Ch'ien-yu (1170—93) Huan-tsung (r. 1193-1206) T'ien-ch'ing (1194-1206) Hsiang-tsung (r. 1206—11) Ying-t'ien (1206-9) Huang-chien (1210—11) Shen-csung (r. 1211 - 2 3) Kuang-ting (1211-23) Hsien-csung (r. 1223-6) Ch'ien-ting (1223—6) Hsien (r. 1226—7) ?Pao-i (1226-7) XXIV Cambridge Histories Online © Cambridge University Press, 2008 (l)Han-p'u I (2) Wu-lu O)Pa-hai I (4) Sui-ko (5) Shih-lu (6) Wu-ku-nai (b. 1021, d. 1074) (7) Ho-li-po (b. 1038, t. 1074-92) (10) Wu-ya-shu (b. 1061. r. 1103-13) (8) Po-la-shu (b. 1042, r. 1092-4) (II) A-ku-ta (Chin Tai-tsu) (b. 1068, r. 1113 (1115 emperor) - 1 1 2 3 ) (9) Ying-ko (b. 1053, r. 1094-1103) (12) Wu-chi-mai (Chin: T'ai-tsung) (b. 1075, r. 1123-35) Noli: The names of the early Jurchen rulers are sometimes written differently in the Chinese sources. Our genealogy follows the Chin thih. In 1135-6, all the former rulers of the Wan-yen clan were given imperial designations posthumously. FIGURE 4 . Genealogy of the early Jurchen rulers (1) T'ai-tsu (b. 1068, r. 111.3 (1115 emperor) -1123) Tsung-kan ung (4) Hai-ling wang (b. 1122, r. 1149-61) (2) T'ai-tsung (b. 1075, r. 1123-35) brother of T'ai-tsu Tsung-hsiin (3) Hsi-tsung (b. 1119, r. 1135-49) (5) Shih-tsung (b. 1123, r. 1161-89) (7) Wei-shao wang (r. 1208-13) Yiin-kung (8) Hsiian-tsung (b. 1163, r. 1213-23) I Tsung-fu (6) Chang-tsung (b. 1168, r. 1189-1208) (9) Ai-tsung (b. 1198, r. 1223-34) Note: This genealogy of Chin emperors indicates the patrilineal descendance. For a full genealogy that lists also the other members of the imperial Wan-yen clan, see Toyama Gunji, Kimkiibi knkyu (Kyoto, 1964), the end of the volume. FIGURE 5. Genealogy of the Chin emperors XXV Cambridge Histories Online © Cambridge University Press, 2008 TABLE 3 Chin emperors and regnal titles T'ai-tsu T'ai-tsung Hsi-tsung Hai-ling Wang Shih-tsung Chang-tsung Wei Shao Wang Hsiian-tsung Ai-tsung Shou-kuo T'ien-fu T'ien-hui T'ien-chiian Huang-t'ung T'ien-te Chen-yuan Cheng-lung Ta-ting Ming-ch'ang Ch'eng-an T'ai-ho Ta-an Ch'ung-ch'ing Chih-ning Chen-yu Hsing-cing Yiian-kuang Cheng-ta K'ai-hsing T'ien-hsing Mo-ti 1115-17 1117-23 1123-38 1138—41 1141-50 1150-3 1153-6 1156—61 1161-90 1190—6 1196—1201 1201—9 1209—12 1212—13 1213 1213-17 1217-22 1222—4 1224-32 1232 1232-4 1234 XXVI Cambridge Histories Online © Cambridge University Press, 2008 Yesiigei Lieh-tsu Temijjin Tai-tsu 1167-1227 r. 1206-1227 Jochi d. 1127 Golden Horde Khans I Khans in Central Asia regency of Oghul Khaimish widow of Ting-tsung 1246-1251 Kammala Hsien-tsung 1263-1302 Yesiin Temiir T'ai-ting huang-ti 1293-1328 r. 1323-1328 Aragibag r. 10.-11.1328 Tugh Temiir Wen-tsung 1304-1332 r. 1328-2.1329 reascended, r. 1.329-1.332 Tolui Jui-tsung d. 1232 regency of Toregene, widow of T'ai-tsung 1241-1246 Giiyiig Ting-tsung 1206-1248 r. 1246-1248 I I Ogddei T'ai-tsung 1186-1248 r. 1229-1241 Chaghadai Mb'ngke Hsien-tsung 1209-1259 r. 1251-1259 Chen-chin Yu-tsung 1243-1285 Khubilai Shih-rsu 1215-1294 r. 1260-1294 J Temiir Ch'eng-tsung 1265-1307 r. 1294-1307 Khaishan Wu-tsung 1281-1311 r. 1308-1311 Ayurbarwada Jen-tsung 1285-1320 r. 1311-1320 Irinjibal Ning-tsung 1326-1332 r. 10.-12.13.32 I Arigh Boke Il-Khans in Persia _L Darmabala Shun-tsung 1264-1292 Khoshila Ming-tsung 1.300-1329 r. 1329 T Hiilegu Shidebala Ying-tsung 1303-1.323 r. 1321-1323 Toghon Temiir Shun-ti 1320-1.370 r. 1333-1368(1370) Ayushiridara Chao-tsung 1339-1378 FIGURE 6. Genealogy of Mongolian rulers xxvn Cambridge Histories Online © Cambridge University Press, 2008 Toghus Temiir 1.342-1388 r. 1379-1388 TABLE 4 Regnal titles of Mongolian rulers Mongolian name Regnal titles Temiijin Mongolian temple name Chinese temple name Chinggis khan T'ai-tsu (1206-27) Ogodei T'ai-tsung (1229-41) T'ing-tsung Giiyiik (1246-58) Mongke Khubilai Temiir Khaishan Ayurbarwada Shidebala Yesiin Temiir Arigibag Tugh Temiir (abdicated 1329) Khoshila Tugh Temiir (restored 1329) Irinjibal Toghon Temiir Ayushiridara Togus Temiir Hsien-tsung (1251-9) Chung-t'ung (1260-3) Chih-yuan (1264-94) Yuan-chen (1295-6) Ta-te (1297-1307) Chih-ta (1308-11) Huang-ch'ing (1312-13) Yen-yu (1314-20) Chih-chih (1321-3) T'ai-ting (1324-7) Chih-hod328) T'ien-shun (1328) Tien-li Sechen khaghan Shih-tsu Oljeitii khaghan Cheng-tsung Kiiliig khaghan Wu-tsung Buyantu khaghan Jen-tsung Gegen khaghan Ying-tsung T'ai-ting huang-ci Jayaghatu khaghan Wen-tsung Khutughtu khaghan Ming-tsung (1328-9) (1329) Chih-shun (I33O-3) Ning-tsung (1332) Yiian-t'ung (1333-4) Chih-yuan (1335-40) Chih-cheng (1341-68) Hsiian-kuang (1371-8) T'ien-yiian (1379-88) Ukhaghacu khaghan Shun-ti Hui-tsung Biliktii khan Chao-tsung Usakhal khan XXVH1 Cambridge Histories Online © Cambridge University Press, 2008 TABLE 5 Capital cities Capitals Liao Supreme Capital Modern location Dates City plan in Steinhardt, Chinese imperial city planning 918—1120 Bolon Khoton; Inner Mongolian A.R. p. 124 929-38 938-1122 938-1118 Liao-yang; Liaoning Peking Liao-yang; Liaoning p. 126 1006-1121 Ning-ch'eng; Hopei p. 127 1044—1122 Ta-t'ung; Shansi Yin-ch'uan; Ninghsia Hui A.R. Chung-hsing fu name changed, early twelfth century date unclear Chin Supreme Capital (Shang-ching) nth c-1115 1119-38 pp. 128-9 1173-1215 1138—50 1153-1215 1120-53 1153-1215 1215-33 1122-53 1153-1232 1132-53? 1117-32 1153-1212 1122—1212 A-ch'eng; Heilungkiang Lin-tung (Bolon Left Banner) Inner Mongolian A.R. A-ch'eng A-ch'eng Lin-tung; Inner Mongolia A.R. Ning-ch'eng; Hopei Ning-ch'eng; HopeiPeking Lo-yang; Honan Peking K'ai-feng; Honan Liao-yang; Liaoning Liao-yang; Liaoning Liao-yang; Liaoning Ta-t'ung; Shansi Mongolian People's Republic Nr. To-lun Inner Mongolian A.R. Nr. To-lun Inner Mongolian A.R. p. 149 1256—64 1264—1370 1267-136? Peking pp. 157-8 (Shang-ching) Southern Capital (Nan-ching) Eastern Capital (Tung-ching) Central Capital (Chung-ching) Western Capital (Hsi-ching) Hit Hsia Hsing-ch'ing fu 1138-53 Northern Capital (Pei-ching) Central Capital (Chung-tu) Southern Capital (Nan-ching) Eastern Capital (Tung-ching) Western Capital (Hsi-ching) Yuan Khara khorum K'ai-p'ing fu Supreme Capital (Shang-ching) The Great Capital (Ta-tu) Yin-ch'uan; Ninghsia Hui A.R. XXIX Cambridge Histories Online © Cambridge University Press, 2008 pp. 151-2 Cambridge Histories Online © Cambridge University Press, 2008 INTRODUCTION The four regimes that form the subject matter of this volume have generally received negative treatment from traditional Chinese historians and have been viewed as an interruption in the grand sweep of Chinese history. Each was established by a non-Chinese ruling group, who maintained their own cultural identity while ruling over a multiethnic state including large Han Chinese populations, and each controlled large territories that had long been ruled by Chinese. Each regime presented a challenge to the integrity of Chinese culture and to China's deeply rooted assumptions about its cultural supremacy and international order. Yet these regimes were remarkably successful. The Khitan Liao dynasty lasted longer than had any previous Chinese dynasty except for the Han and the T'ang. For more than a century after the fall of the T'ang dynasty in 907, the Tangut in northern Shensi and Kansu clung tenaciously to the regional authority that they had acquired under the late T'ang and then formed their own empire of Hsia, which as an independent state survived for two centuries more. More than two centuries after the collapse of Khitan power, the official historians of the Mongolian Yiian court grudgingly conceded to the Liao the title of a legitimate dynasty but denied that status to Hsia, who would seem to us to have had almost as good a claim. Both were long-lasting, stable regimes, firmly rooted in territories that had been in part settled by Han Chinese for a millennium, and both regimes survived in the face of a hostile Chinese regime, the Sung, whose population outnumbered them twenty to one and whose economic resources were even more overpoweringly superior. Both fought the Sung to a standstill, forced the Sung to recognize their existence and parity as sovereign regimes, and extracted large subsidies from them in exchange for peace. The Jurchen Chin were still more successful, emerging from what had always been a remote borderland of the Chinese world first to conquer the Liao empire and then to wrest control over all of north China from the Sung. Whereas the Khitan and Tangut had been multiracial border regimes in which the Han Chinese, though probably the majority of the population and certainly the producers of much of the country's wealth, had not been over- Cambridge Histories Online © Cambridge University Press, 2008 2 INTRODUCTION whelming in numbers, the Chin empire had a burgeoning Chinese population of between thirty million and forty million. The Jurchen regime in China was necessarily far more of an accommodation to traditional methods of government than had been those of the Khitan or Tangut. It was also a state that could claim far more realistically to have been an equal rival to the Sung, "another China." The Mongolian Yuan dynasty was quite distinct from its predecessors. For the Liao, its Chinese domains had been, at least in the beginning, an almost accidental extension of a regime whose political heart always remained in its home grasslands. The Jurchen had kept their tribal home territories in the northeast but soon became essentially a Chinese dynasty with their capital in China. The Mongols overwhelmed the Hsia and Chin and took possession of northern China as merely one part of an unparalleled explosion of military power aimed at world conquest, which made them masters of northern Eurasia from the borders of Hungary and Poland to the Sea of Japan. The Mongols' conception of "All under Heaven" was far grander than anything dreamed of by the Han, T'ang, or Sung, or indeed by any other Chinese regime. China became for a while only part of a much larger political order. By the time Khubilai conquered the Southern Sung and brought the entire country under Mongolian rule, the Mongolian empire had been fractured into rival khanates, but China — now the whole of China, with well over ioo million people - remained just one part of a larger empire. Unlike its predecessors, when the Yuan regime in China collapsed the Mongols simply retreated to their homeland in the steppe and, there, continued to be a major power for several centuries. Each of these regimes was in its own way successful, and each of their dominant peoples proved adaptable and yet maintained their own identity. They controlled large areas of north China for centuries. The region around Peking, for example, remained in alien hands for well over four centuries, and the western part of Kansu Province was restored to Chinese control under the first Ming emperor after six centuries of foreign domination. Moreover, the whole of north China was under non-Chinese rule for well over two hundred years. Although an ever-increasing part of China thus fell into foreign hands, these regimes may also be looked at in a different light, as a part of a much lengthier reverse process by which Chinese-style bureaucratic governance became the political norm in East Asia and was adopted and adapted by regimes outside Chinese control and beyond what had been traditionally Chinese territory. This development can be traced back to KoguryS, to the other Korean kingdoms of Silla and Paekche, to the unified Silla from the seventh century onward, and to seventh- and eighth-century Japan. The Cambridge Histories Online © Cambridge University Press, 2008 THE LATE T'ANG BALANCE OF POWER 3 immediate predecessor of the Liao as a stable northeastern state was Po-hai (Parhae) in southeastern Manchuria (719—926), the conquest of which was perhaps a more important stage in the establishment of Khitan power than was the acquisition of the sixteen Chinese prefectures in 937. Po-hai was an independent state, with five capitals, a sophisticated Chinese-style bureaucratic government, an elite with a mastery of literary Chinese, and a developed culture that had emerged and flourished for two centuries on the borders of the T'ang empire. 1 A somewhat similar regime had been in existence since the early eighth century in Nan-chao, occupying modern Yunnan.2 These two fiercely independent "satellite regimes" of T'ang represented the same general political development - the adaptation of Chinese institutions by non-Chinese populations - one aspect of the larger process by which surrounding peoples were incorporated into the Chinese institutional and cultural system. This could happen in a variety of ways: In the case of the Hsia, Koguryd, and also Vietnam — which finally threw off Chinese rule in the early tenth century and became an independent state on the Chinese model* — some of their peoples had lived under Chinese provincial administration for centuries and, when they finally broke free to form their own states, continued to employ familiar methods of government. At the other extreme, Japan, Pohai, and Nan-chao occupied areas that had never been effectively ruled by a Chinese dynasty, but their peoples were familiar with China and its institutions and imitated familiar Chinese models when they themselves formed independent states. And these various adaptations took place in a world that was fundamentally changing. THE LATE T'ANG BALANCE OF POWER One of the problems of traditional Chinese historiography in its dealings with foreign peoples has been its failure to match unchanging theory with constantly evolving actuality. The ancient ideal of the "five zones of submission" envisioned a world in which China, or rather its ruling dynasty, the bearers of the Mandate of Heaven to control mankind, were the sole legitimate possessors of unquestioned authority - authority that was at once political, cultural, and moral. The surrounding peoples, the "barbarians" who did not fully participate in Chinese culture, should in this ideal model voluntar1 On the emergence of Po-hai, see Denis C. Twitchett and John K. Fairbank, eds., Sui and T'ang China, 389—906, pt. 1, vol. 3 of The Cambridge history ofChina (Cambridge, 1979), pp. 440—3. 2 On Nan-chao, see Charles Backus, The Nan-chao kingdom and T'ang China's southwestern frontier (Cambridge, 1981). 3 See Keith W. Taylor, The birth of Vietnam (Berkeley and Los Angeles, 1983). Cambridge Histories Online © Cambridge University Press, 2008 4 INTRODUCTION ily submit to the emperor and become his vassals, and their countries would become a sort of outer perimeter under the emperor's moral authority but beyond his real control and outside the realm of civilian direct administration. Despite centuries of contacts with the real outside world, this model persisted as a symbol of the world as the Chinese elite imagined it should be. Remnants of these attitudes, based on an imaginary world sharply divided between Chinese and barbarians, continued to undermine China's relations with other peoples for centuries after the period covered by this volume. This was a theory that may have had some validity in the distant past, when China was surrounded on all sides by peoples with a somewhat lower level of cultural development and with loose and ill-articulated forms of political organization. But it had changed forever during the T'ang. In the Sui period, only one of China's neighbors, KoguryS in northern Korea and southeastern Manchuria, had any claim to be a "state" with a mainly sedentary population and stable institutions. All the other border peoples, from Yunnan to the border of Hopei, were tribal peoples, many of them living a seminomadic life without any permanent large-scale structures of government, although they might join together to become a potent threat in times of crisis. Perhaps more important, none of them had a written language except Kogury6, which employed Chinese. By 750 that situation had completely changed: Late T'ang China was ringed with stable states - Nan-chao in Yunnan; the great aggressive Tibetan kingdom along its long border with Szechwan, Kansu, and what is now Sinkiang; the 'Abbassid caliphate in the far west; the Turkish (T'u-chiieh) and later the Uighur "empires" in the Mongolian steppe; Po-hai in Manchuria; Silla in Korea; and, far away, Japan. All these states had literate elites, some employing Chinese as a written language and others using writing systems of their own. Events in the 750s and 760s had further stabilized this situation. After the disasters of An Lu-shan's rebellion, T'ang forces abandoned their far western protectorates in Sinkiang and also the extensive areas under regular Chinese civil administration in Turfan, Hami, and the Kansu corridor. All Kansu was occupied by the Tibetans. From the 730s to the 750s T'ang armies had intervened in the Pamir region, fought the armies of the 'Abbassid caliph on the Talas near Ferghana, invaded Nan-chao, and attempted to conquer the Khitan in the northeast. After 763 the T'ang took up a totally defensive posture. No T'ang expeditionary force was ever again sent against a neighboring state with an eye to conquest. Even when the chance arose in the 850s, the T'ang court deliberately decided not to reoccupy the lost prefectures of the northwest. During the late eighth and early ninth centuries there had gradually emerged a novel and stable international situation in which the T'ang used Cambridge Histories Online © Cambridge University Press, 2008 THE LATE T'ANG BALANCE OF POWER 5 diplomacy as much as armed force and in which the other players on the international scene also gradually evolved stable relationships among themselves that were underpinned by treaties. In 822 the T'ang entered this interstate system when they finally concluded a treaty with Tibet on the basis of equality. And China was no longer the center around which international relations revolved, although ambassadors and embassies continued regularly to visit Ch'ang-an. In the west the Uighurs, Tibetans, Nan-chao, and the Arabs were involved in conflicts among themselves and developed their own network of alliances and treaties, and in the northeast Silla, Po-hai, and Japan formed another diplomatic network employing Chinese as their common language and formalities derived from the T'ang system. In neither of these networks were the T'ang direct participants. In 840, the only immediate neighbors of China who were not an organized state were the tribal nations of the Khitan and Hsi, living to the north of modern Hopei and in western Liao-ning. They were, for the moment, vassals of the Uighur khaghans, though they still also retained close and regular relations with the Chinese court. Then around 840 the stability of northern Asia began to unravel. First, the Tibetan kingdom suddenly collapsed, owing to internal causes that remain to be satisfactorily explained. Almost immediately thereafter the Uighur empire disintegrated, and the Uighurs abandoned their capital city, Karabalghasun, and their homeland in Mongolia to settle in eastern Sinkiang, Turfan, Hami, and the Kansu corridor. Their Khitan and Hsi vassals transferred their loyalty to the T'ang court. Toward the end of the century came a collapse of central authority that spread through East Asia like an epidemic: The T'ang empire was destroyed by Huang Ch'ao's rebellion and was an empire only in name from 880 onward. Long before its formal end in 907 in reality it had disintegrated into numerous independent local regimes, competing for hegemony and constantly at war with one another. For more than half a century after 907, China was divided among as many as ten regional states. In the last years of the century, central authority began to break down in Japan; in Korea, the Silla kingdom broke up into three regional warlord states; in Manchuria, Pohai went into a terminal decline; and in the far southwest Nan-chao too fell apart. The fragmentation of China itself during the Five Dynasties was paralleled everywhere in East Asia. It was against this background that the Khitan state of Liao emerged. There was no sudden breakdown of an international order imposed by the T'ang, as is sometimes suggested. That order had disappeared forever in the late eighth century, modified into something quite new and replaced by a novel framework of international relations. But this, too, had collapsed in its Cambridge Histories Online © Cambridge University Press, 2008 6 INTRODUCTION turn, and in A.D. 900 the international situation had been fluid for some sixty years, and governments were collapsing everywhere. In this situation of near anarchy, the Khitan gradually, almost accidentally, became players in the game of warlord politics in north China, as well as in the world of the steppe, the control of which was their first objective. Moreover, the situation was to remain fragmented for years to come. In China itself, the political fragmentation lasted for almost a century, from the fall of Ch'ang-an to Huang Ch'ao in 880 to the final conquest of the Northern Han by the Sung armies in 979. For much of that period China was divided into as many as nine or ten regional states, and until 960 the north was ruled by a succession of unstable and short-lived military regimes. During that period, military force determined the shape of politics and continued to be a major factor in the first decades of the Sung. Not only was the early tenth century dominated by the generals in north China; it also was a period during which many of the regional warlords were non-Han generals, particularly Sha-t'o Turks. Li K'o-yung, the greatest of these, had been instrumental in putting down Huang Ch'ao and was the de facto ruler of northern Shansi long before the formal end of the T'ang in 907, and one of the contenders for power in the constant civil wars that blighted the twilight of the T'ang in the 880s and 890s. After the fall of T'ang his state (known as Chin) became an independent entity. In 921 Li K'o-yung's successor reunified north China by destroying the Liang and then set himself up as emperor of the restored Later T'ang dynasty (923—37). For a quarter of a century, all of north China was under Sha-t'o rule, first under the Later T'ang and then under their successors the Chin (937—46). Li K'o-yung had already involved the Khitan in a short-lived alliance as early as 905. By the 920s they were drawn in as full-fledged participants in north Chinese politics: The Chin were their puppets and ceded to them sixteen border prefectures with Chinese populations. They were also wooed as allies by the independent courts of southern China. Between 944 and 947 the Khitans attempted to invade China, took the Chin capital, destroyed the dynasty and briefly set up one of their own, but wisely decided that the risks were too great and withdrew, leaving north China in the hands of yet another Sha-t'o military dynasty, the Han (947-50). Although this regime soon collapsed as a dynasty, its successors retained their independence in the Sha-t'o homeland in Shansi until 979. Large parts of northern China were thus under the rule of the Sha-t'o for many years: in the case of northern Shansi, for a century. But the Sha-t'o were not the only foreign leaders of local regimes on Chinese soil at this time. The northwest, which was Tibetan-occupied territory until the 840s, was split among a variety of local warlords: Chinese in Tunhuang; Uighur in Turfan, Cambridge Histories Online © Cambridge University Press, 2008 THE FRONTIER 7 Kan-chou, and Su-chou; Tibetan in Liang-chou; and Tangut on the southern borders of the Ordos. The Tanguts, like the Sha-t'o, had been settled in the region by the T'ang as a frontier garrison force and had clung to their local power through the disturbed times of the early tenth century. They later emerged as founders of a powerful multiethnic empire, the Hsia or Hsi Hsia, which in the early eleventh century assimilated all the fragmented local regimes of the northwest under its own control. Like the Sha-t'o, the Tanguts were not alien invaders of Chinese territory but were non-Han people who had been settled inside the T'ang frontiers by deliberate government policy and had long been a part of the T'ang provincial and military organization. The roots of two of our major regimes, the Liao and the Hsi Hsia, can thus be traced into the political and military turmoil of the late ninth and early tenth centuries. It is also important to see how they were rooted in the frontier order established by the T'ang. THE FRONTIER Traditional histories of China depict the Khitan, Jurchen, and Mongols as "outsiders," inrupting into "Chinese" territory. But this is a misleading oversimplification that needs to be laid to rest forever.4 In spite of what is shown in modern historical atlases, the T'ang, like its predecessors, never had any clearly defined and demarcated northern frontier. People still sometimes spoke of the "Great Wall," and some remnants of earlier fortifications still existed, but this term was used only to express a vague idea of China's limits.5 There was never a continuous defensive line or a defined frontier. There was a line of fortified border prefectures and counties, a few fortresses in strategic places, and a scattering of military colonies, military stud farms, beacon signal towers, and military picket-outposts. It was a defense in depth, with its backbone formed by powerful provincial armies at Ling-chou, T'aiyiian, Ta-t'ung, and Peking. Only on the Tibetan frontier did the T'ang maintain a massive static defense, and only the Tibetan frontier was from time to time demarcated in disputed areas by mutual negotiation. But in the north, T'ang control was defined by the extent of the border prefects' authority, and this constantly changed. This "frontier" was further obscured by another aspect of early T'ang military policy. The tribal peoples of the border districts were brought 4 For an interesting new analysis of the relations between nomadic peoples and their sedentary neighbors, see Anatoli M. Khazanov, Nomads andthe outside world, trans. Julia Crookenden (Cambridge, 1984). For a new interpretation of China's relations with its steppe neighbors since the Han period, see Thomas J. Barfield, The perilous frontier: Nomadic empires and China, ed. Charles Tilly (Oxford, 1989). 5 On this problem, see Arthur N. Waldron, The Great Wall: From history to myth (Cambridge, 1990). Cambridge Histories Online © Cambridge University Press, 2008 8 INTRODUCTION partially within China's political orbit by a complex system of agreements and understandings through which they were incorporated into a system of indirect rule, under which their leadership was co-opted into the T'ang government system "on a loose rein" and given titles, offices, ranks, and emoluments. The tribal groups were organized in a loose hierarchy of prefectures (chou) and governments-general (tu-tu-fu) overseen by T'ang border commanders. They did not in fact participate in the T'ang system of government; rather, their chieftains ruled their peoples according to native custom. Their hierarchy of Chinese titles was largely a formalization in Chinese terms of the existing tribal structures, in the stable continuance of which both their own chieftains and the T'ang government alike had a vested interest. The tribal leaders were periodically invested with new titles and were given gifts and subsidies to ensure their continuing loyalty. The larger and more powerful border peoples and their paramount leaders were inducted even further into the T'ang order by the bestowal of the T'ang imperial surname, with the implication of their incorporation into the imperial family's system of kinship; by marriage of their rulers with T'ang princesses, to establish a dynastic marriage relationship; and by the education of their future rulers in China as "hostage princes," usually serving long terms as officers in the imperial guards. The embassies that attended the Chinese court often included large numbers of prominent tribesmen in addition to the ambassador, and these people, too, became familiar to some extent with things Chinese. None of these measures, of course, went very far in making Han Chinese out of the tribal leaders. But they were successful in that they gave tribal leaders a firsthand knowledge of the capital and court and of Chinese institutions and methods of government, and they produced an influential minority among those familiar with the Chinese language and customs. None of the immediate neighbors of T'ang China was really ignorant of China, whether or not they chose to imitate Chinese models. Thus the idea of a "frontier" in T'ang China was a multilayered one. There was the outer zone of people who were part of the "Chinese world" by virtue of their participation in the tributary system; there was an inner zone of tribal people under very loose indirect rule; and there was the outer limit of the T'ang military defense system and the outer limit of effective civil government. There was also, of course, the abiding "ecological frontier" between those regions suitable for permanent agriculture and those that would support only a pastoral economy, which limited the possible expansion of a Han Chinese farming population. And last, there was a frontier sometimes too casually assumed to be the same as this ecological boundary, between those regions with a dense more or less homogenous Han Chinese population and those Cambridge Histories Online © Cambridge University Press, 2008 THE FRONTIER 9 inhabited by other peoples. In the T'ang neither the military defense system nor the limits of civil administration represented either an ethnic or a cultural frontier. Many non-Han people had lived for centuries inside these frontiers, intermingled and intermarried with Han Chinese and other ethnic groups, some of them partly or wholly assimilated. This had long been so, at least since the massive settlements of hundreds of thousands of Ch'iang, Hsiung-nu, Hsien-pei, and other border peoples in Later Han times. There was a further huge infusion of non-Han invaders during the fourth and fifth centuries. It is customary to think of these peoples as having been assimilated fairly rapidly into the Chinese way of life, but in the sixth century many of them still retained a powerful ethnic and cultural identity of their own, and some Han Chinese in the border zones adopted some aspects of their non-Han life-style. The early T'ang government settled further some large groups of non-Han people - Ch'iang, Tangut, T'u-yii-hun, Tibetans, Turks, Uighurs, Khitan, and even Sogdians from Central Asia - in their northern frontier prefectures. These peoples numbered in the hundreds of thousands. Some willingly assimilated and adopted a settled life as farmers, though Chinese attempts to force this on other groups were bitterly resisted. Others remained primarily herdsmen. But they were important to the T'ang government, as they oversaw the huge government pastures that produced horses for the cavalry and other livestock. Many of these minority groups retained their tribal structure and their tribal leaders. Many of their menfolk were enlisted in the T'ang armies and provided a great part of its cavalry, their own chieftains serving as their commanders. The modern province of Kansu, the border areas of eastern Tsinghai, and the northern parts of Shensi, Shansi, and Hopei formed a broad zone where a minority of Han settlers and much larger non-Han populations co-existed, for the most part peacefully, and where garrison troops, Han and non-Han, with their military settlements and colonies, lived side by side with both local farmers and semi-nomadic herdsmen. There was some degree of intermarriage, and the identity of these peoples was far from uniform, with some thinking of themselves primarily as T'ang subjects and others clinging fiercely to their tribal legacy. It is thus totally misleading to imagine the T'ang's northern frontier, even at the height of its dynastic power, as a sharply delineated national boundary in the modern sense, clearly defining areas of sovereignty and separating different peoples. Instead, it was a broad transitional zone in which identities, loyalties, and authority were constantly changing and striking new balances. Cambridge Histories Online © Cambridge University Press, 2008 IO INTRODUCTION The collapse of the T'ang empire between 880 and 907 and the emergence of various new independent states both inside and outside what had been T'ang territory brought about sweeping political changes but left the confused multiethnic borderland of northern China much as before. There was no longer a Chinese central power, but its local military leadership remained intact and even more aggressive, as the central restraints had been removed. The main difference was that the borderland now became the region from which new military and political forces arose to influence the rest of northern China. The Sha-t'o Turkish power based in Shansi and the Tangut regime in the borders of the Ordos had been developed from parts of the T'ang border military establishment. The Sha-t'o attempt to renew the T'ang was largely based on a border general's conception of what T'ang power had been. It was also a continuation of an ancient trend: Ever since the rise of Toba Wei, one centralizing regime after another had been founded by border generals based in the no-man's-land of the northern frontier — the Toba themselves, who rose to power from the Ta-t'ung region; their successors the Western Wei and Northern Chou, whose rulers were originally commanders of northwestern garrisons; the Sui who came from the same group; and the T'ang, who were closely related to both the Northern Chou and the Sui and had their original power base in T'ai-yiian. All were originally military regimes, and all were able to mobilize support from both the Han Chinese and the non-Han worlds. The same trend continued, but with important differences: The Liao arose on the northern borders but, after acquiring an important foothold in China, decided against serious conquest. The Chin arose from being a dependent people on the northern border of the Liao's Chinese-style empire. The Mongols, who are often treated as sui generis, had been as we shall see later frontier dependents of the Chin for many years before Temiijin's rise to power. The difference in the case of the Jurchen and Mongols was that after the tenth century the frontier itself had moved: The outer frontier of empire to which the Chin and the Mongols related was not the traditional frontier of the Chinese world, or the frontier between the Liao and Chin and the Sung, but that of the "extended Chinese world" brought into being by the Khitan conquests in Mongolia, Liaoning, Kirin, and Heilungkiang and their establishment there of a state and a system of border relationships based on the T'ang Chinese model. All these peoples came to power not as strangers and complete outsiders in the Chinese system, but after long having been a part of it. Living on the fringes they were perhaps better acquainted with the provincial border than with the real centers of dynastic power and culture, but to some degree they were participants nevertheless. Cambridge Histories Online © Cambridge University Press, 2008 FOREIGNERS II FOREIGNERS Just how alien were these peoples to the societies that they conquered by force of arms? And what did they have in common? In a larger historical perspective it is possible to view the dynasties of conquest as one more phase in the age-old confrontation between the Chinese and their northern neighbors that has existed since the Western Chou period, when the heartland of Chou in the Wei River valley was invaded by seminomadic peoples.6 Under the Ch'in and Han, the Hsiung-nu confederation was the chief adversary of the Chinese, followed in the third century A.D. by the Hsien-pi and other tribes, who were successful in dominating China's northern provinces and establishing states of their own on Chinese soil. After the T'ang dynasty had lost its hegemony and finally disintegrated into several states from the early tenth century onward, the border conflict took on a new form. The reunification of China under the Sung after 960 was paralleled by a steadily growing consolidation of polities founded by peoples north of the borders of China proper. It would, however, be wrong to see the polities founded by the northerners as nomad empires in contrast with the stable empire of the sedentary Chinese. The peoples who founded the states of Liao, Hsia, Chin, and Yiian were not, by any means, all nomads. Both the Khitans and the Mongols in the beginning had an economy based on pastoralism, and their wealth was predicated on their vast herds of horses, sheep, and camels. Nevertheless, none of the "nomadic" neighbors of China was exclusively pastoral. There was always some marginal agriculture and also extensive trade that supplied them with those goods that could not be produced from cattle raising. Before their conquests in China, the Khitan had some farming and permanent settlements, with Chinese artisans and Chinese and Uighur traders. The Jurchens were not nomads at all in the strict sense. Even those "wild" Jurchen living in the forested mountains of eastern Manchuria were a sedentary people, living by hunting, fishing, and some agriculture. They did not live in tents, but in villages consisting of wooden cabins. The Jurchen in the plains of Manchuria, which had been a part of the Po-hai state until its annexation by the Khitans, were not nomads either, although they had huge herds of horses. These differences in life-style and economy among the Jurchen are reflected in the terms "civilized" (shu) and "wild" {sheng) Jurchen, which were already current under the Liao. The Tanguts, too, had a complex mixed economy long before their rise to independence. Thus to lump to6 See Herbert Franke, "The role of"the state as a structural element in polyethnic societies," in Foundations and limits of state power in China, ed. Stuart R. Schram (London and Hong Kong, 1987), pp. 87-112. Cambridge Histories Online © Cambridge University Press, 2008 12 INTRODUCTION gether the founders of all these regimes and characterize them as "nomad invaders" would certainly be a naive oversimplification. Another simplification of which historians must be aware is terminology. When we use the terms Khitan, Jurchen, Tangut, or Mongol, we should remember that each of these terms referred not to a single homogenous people but a complex reality. The names of Khitan, Jurchen, or Tangut are in reality those of federations under the leadership of Khitans and so forth. The names are those of the linguistically dominant groups within these federations. The federations themselves were each polyethnic and multilingual, as the various contributions to this volume will make abundantly clear. The Khitan federation, for example, included tribes and ethnic groups with Turkic affiliations such as the Hsi or the Uighurs; others were undoubtedly Mongols like the Shih-wei or of Tungus stock like the civilized Jurchen, but the common language within the federation must have been Khitan. Later the federation was expanded to include Po-hai and Han Chinese. The same is true for the Jurchens, among whose confederates we find other Tungus tribes and, in addition, Mongols. The Mongols themselves had absorbed Turkicspeaking tribes like the Onggiid, not to mention Central Asians who had come within the Mongolian orbit after the great expeditions against Central and Western Asia. The ethnic and linguistic composition of the peoples bordering on China in the north and west was always fluid: Whole tribes either voluntarily joined the dominant tribe or were placed under their leadership by force or persuasion. It was a common policy of all these peoples to incorporate the warriors of a subdued or allied tribe into their armies, generally under their previous commanders. All conquerors followed this same integrative policy in regard to the Chinese population after the annexation during their conquests of regions with preponderantly Chinese sedentary settlements. The technical skills of the Chinese were welcome for the manufacture of weapons and the operation of siege machinery for use against walled towns. Other Chinese recruits served as infantrymen, whereas the cavalry remained in most cases the prerogative of the non-Chinese contingents. The "Khitan," "Jurchen," and "Mongolian" armies were always multinational and included a large number of Chinese soldiers. It is therefore rather doubtful that we should view the wars of the Sung against their adversaries as purely national or racial wars of resistance against foreigners, and we might even regard the wars between Sung on the one hand and Liao, Hsia, Chin, or the Mongols on the other as almost a special form of Chinese civil war in which one side fought under foreign command and disposed of more or less strong contingents of non-Chinese troops. Such suggestions are, of course, tentative, and a more definite interpretaCambridge Histories Online © Cambridge University Press, 2008 FOREIGNERS 13 tion would require more research into warfare in the period from the tenth to the thirteenth centuries and, above all, a numerical breakdown of the Liao, Chin, and Mongolian armies according to nationalities. But to cite one example, when the Chin ruler, the prince of Hai-ling (r. 1150-61), mobilized his state for a campaign against the Sung in 1159 and 1160, the predominantly Jurchen meng'-an mou-k'o army numbered 120,000 men, whereas not fewer than 150,000 Han Chinese were conscripted for the campaign, plus 30,000 sailors for warfare on the waterways and lakes of central China. The majority of his military rank and file was therefore not "Jurchen" but Chinese. Finally, it must be remembered that the antagonism between Chinese and non-Chinese cannot be construed in the traditional Chinese way as a confrontation between high civilization and barbarism. It cannot, by any means, be assumed that the conquerors who founded their states on Chinese soil from the tenth century onward came from nowhere and started from a very low level of political organization and cultural achievement. The Hsi Hsia state of the Tanguts was a special case insofar as they were not conquerors or invaders but had been living for centuries in the same region, which became the nucleus of their state. The Tanguts' ethnogenesis was less one of conquest but, rather, of a gradual absorption of other tribal elements into a federation that included also Chinese, Tibetans, and smaller ethnic groups in the Ordos region and what is now Kansu Province. They, too, could not by the wildest stretch of the imagination be described as primitive barbarians when they achieved formal independence in the middle of the eleventh century. Although it would be an exaggeration to regard the polities of the various federations as fully sinicized, it is a historical fact that the complex interaction between the Chinese empires and their so-called barbarian neighbors had been going on for centuries. One indicator of Chinese influence on the institutional framework of the neighboring peoples is the many loan words of official titles that were borrowed from Chinese. By the early T'ang period the Turks (T'u-chiieh) had already adopted several titles from Chinese. Many native Khitan official titles were also borrowed from Chinese, for example, hsin-kun, which in Chinese is chiang-chun "general," or the title of hsiangwen, which occurs in several transcriptions and is derived from Chinese hsiang-kung, an honorific designation for a chancellor or minister. The Mongols too, even before the proclamation of Chinggis khan as supreme ruler in 1206, had adopted into their language the Chinese word ivang, "king," as ong, and t'ai-tzu, "prince," via the Turkic tays'i, as taisi in Mongolian. Both words are used in the Secret history of the Mongols. Such loan words show the prestige and influence of Chinese institutions and terminology even if they Cambridge Histories Online © Cambridge University Press, 2008 14 INTRODUCTION were now employed in a social or political setting different from that of their Chinese models. VASSALS AND OVERLORDS Another aspect of the interaction between Chinese and foreigners that had begun long before the dynasties of conquest were founded are the tributary relations and the role of non-Chinese polities in the organization of the frontier areas. In Chinese political terminology, they were regarded as outer vassals {wai-ch'en). These vassals were only loosely integrated into the Chinese orbit; they were expected to appear with local products as tribute at the Chinese imperial court. The Khitans had done this more or less regularly since the seventh century A.D., and many Khitan grandees were rewarded with honorary ranks and noble titles by the T'ang emperors. Some of them were awarded the surname Li, which shows that they had been given the privilege of bearing the family name of the ruling T'ang house. Such adoptions were by no means rare and served to draw foreign leaders closer to the imperial court. The conferment of offices also played a role in the loose integration of non-Han leaders into the Chinese institutional hierarchy. We find, for example, that in 649 the Khitans were organized into the government-general (tu tu-fu) of the Sung-mo region in Manchuria. It is not always easy to determine to what extent this inclusion in the Chinese outer hierarchies was purely nominal. Such tribes or nations certainly remained fully independent in managing their internal affairs. But the resonant Chinese titles granted to their leaders conferred prestige. The situation of the Jurchens in the Liao state was similar to that of the Khitan in the T'ang world. For generations the ancestors of the dynastic founder A-ku-ta had held the title of commanding prefect (chieh-tu shih) in the Liao border hierarchy, so that he too was not an absolute newcomer on the political stage when he assumed the title of emperor in 1115. In addition to the national Chinese state of Sung, during the Liao dynasty a new center of political gravity had evolved in the north where the modes of dealing with frontier populations followed the precedents of T'ang and Sung. It has been suggested that the case was different for the Mongols and that they rose to power completely outside the Chinese framework of frontier organization. This would imply that the Mongolian tribes under the leadership of Temiijin, who was proclaimed khan in 1206, had not previously been subject to interaction with an imperial court and that their consolidation as the dominant power in the steppe region was not affected by their involvement in tributary relations and the concomitant formalities such as the conferment of titles by a central power. This interpretation, however, has to Cambridge Histories Online © Cambridge University Press, 2008 VASSALS AND OVERLORDS 15 be abandoned. It was based on the picture of the early history of the Mongols before 1206 given in the Secret history of the Mongols. This source represents Temiijin and the Mongols, both his allies and adversaries, as steppe nomads completely untouched by Chinese or any other high civilization, without statehood and without close relations with a Chinese empire. In the second half of the twelfth century this empire would have been, for the Mongols, the Chin state, and the Chin dynastic history {Chin shih) is completely silent on the subject of Mongols as vassals. The Sung sources, however, give a completely different picture and show that the ancestors of Chinggis Khan were not only adversaries of the Chin but also vassals with a state {kuo) of their own.7 It can be easily explained why both the Secret history and the Chin dynastic history have omitted this episode. The Secret history is a romanticized epic in which the career of Temiijin appears as a rise from the humblest of beginnings to supreme power; there was no room in this national epic for any mention of the fact that Temiijin's ancestors had been vassals of another state. The Chin dynastic history, too, was compiled under Mongolian rule, and its authors somewhat understandably omitted all references to the vassal status of Chinggis khan or his predecessors. The information from which we may conclude that the early Mongols had a kuo of their own all comes from Sung sources. It is reported that in 1147 after an outbreak of warfare, the Mongols were pacified by Chin presents and that the ruler of their state of Meng-ku took an imperial title and adopted his own reign title (nien-hao), T'ien-hsing (Heavenly Rising). 8 The Mongolian leader under whom this occurred has been identified by some scholars as Khabul khan, Temiijin's grandfather who, even according to the Secret history, "ruled over all the Mongols." Sung sources also mention that Temiijin himself had attended the Chin court in his capacity as one of its outer vassals.' It is therefore clear that Temiijin was not, as the Secret history wishes to make us believe, an adventurer with an obscure tribal background but the descendant of a princely family that had received recognition, investiture, and gifts from the Chin court. These examples show how deeply and how long the leaders of the rising powers of Khitans, Jurchens, and Mongols had already been exposed to contacts with a central imperial power and that by the time they were about to found an imperial state they had achieved a considerable level of political and cultural sophistication. They knew and understood the "Chinese" sys7 See Charles A. Peterson, "First Sung reactions to the Mongol invasion of the north, 1211—1217," in Crisis and prosperity in Sung China, ed. John W. Haeger (Tucson, 1975), p. 248. 8 Yu-wen Mou-chao, Ta Chin kuo chih (KHCPTS ed.), 12, pp. 9 9 - 1 0 0 . 9 See Li Hsin-ch'uan, Chien-yen i lai cb'ao yeh tsa chi (KHCHTS ed.), 19, p. 585. Cambridge Histories Online © Cambridge University Press, 2008 16 INTRODUCTION tern. Their rise to imperial status took place within the institutional framework that had dominated Sino-foreign relations, and so it is not difficult to imagine how foreign rulers themselves might have aspired to become Sons of Heaven and to hand out investitures or receive tribute instead of being enfeofred by, or sending delegations bearing tributes to, their overlords. The image of the universal emperors of early T'ang who had ruled over a much larger territory than the Sung had also certainly influenced the leaders of the neighboring tribal federations, who gradually asserted themselves and eventually succeeded in becoming emperors and Sons of Heaven in their own right. MULTISTATE SYSTEM The fragmentation of the geographical area of China into several states, each under its own Son of Heaven, was certainly nothing new for the Chinese. Such a division had already existed for more than three centuries between Han and Sui and again came into being during the period of the Five Dynasties in the early tenth century. Insofar as the simultaneous existence of a plurality of states is concerned, the period of the dynasties of conquest was not different from earlier ages of political disunity. But there were some new elements in the situation. One of these was a major shift in political centers. The region around Peking (Yen) had been for centuries a comparatively unimportant border area in the far north, essentially a border garrison city that did not play a considerable political, cultural, or economic role in China as a whole. When the Liao made Yen-ching (their Southern Capital, Nan-ching) one of their five capitals and the principal administrative center for the entire settled region of the empire, this changed the situation drastically. The Liao were followed by the Chin who also ruled from Yen-ching, now called their Central Capital, Chung-tu, and the Mongols whose emperors resided with a large part of their metropolitan offices in the city, now called the Great Metropolis, Ta-tu. Both the Chin and the Mongolian Yiian built an imposing imperial city in what is now Peking, with sumptuous palaces and parks. Much that impresses the visitor even today goes partly back to the Chin and Yiian periods when Peking for the first time in history became a national metropolis. During the Five Dynasties the two cities that had throughout recorded history been the rival loci of supreme political authority, Ch'ang-an and Loyang, finally lost their preeminence. After the 88os Ch'ang-an was in ruins and would never again be more than a provincial capital, and the whole northwest declined into the backwater. After the Liang set up its capital in K'ai-feng in Ho-nan, the center of the transport network of the eastern plain, Loyang went into decline as well. K'ai-feng was again adopted as the capital Cambridge Histories Online © Cambridge University Press, 2008 MULTISTATE SYSTEM 17 by the Sung who reunified the empire. The first step in the transfer of the center of Chinese political gravity to the northeast took place after 1127 when the Sung lost all of north China and their capital at K'ai-feng. The Southern Sung regime subsequently established its "temporary capital" at Hang-chou, which developed into a second metropolis, even more splendid than K'ai-feng. Meanwhile the Chin, masters of north China, established their Central Capital at Peking. With the destruction of the Southern Sung state in the years following 1276, Hang-chou too, lost forever its role as a national political center, and for almost a century the whole of China was governed from Peking until the Mongolian court was expelled in 1368 and withdrew into the Mongolian steppe country. The Ming at first established their capital in Nanking, but after 1420 the court was moved back to Peking, where it remained until the end of the empire in this century. This shift of the political center to the northeast changed Peking, previously a provincial backwater with no cultural identity, into the capital of a unified China, a populous metropolitan city producing a wide range of manufactures and the scene of manifold cultural activities. Economically, however, the northeast had been an impoverished and unproductive area even before it bore the full weight of the first Mongolian onslaughts, and Peking remained dependent on imports of grain from the rich provinces in the Yangtze region, which in turn necessitated a complete reorganization of the domestic transport system by both land and water. As we have seen, China had repeatedly been reunified by regimes originating on the northern border. This had happened under the Sui and the Sung and was reenacted under the Yuan. The idea that China could be unified only from the north had almost become a stereotype by the thirteenth century, and it was used as a political argument by Khubilai when after his accession he proposed that the Sung conclude a peace with the Mongols. Of course both emperors in Peking and Hang-chou considered themselves as legitimate rulers over the Chinese world. The aspirations of the Mongols, however, exceeded by far those of the Chinese states of the past, for the Mongols regarded as their legitimate domain not only China but the entire world. This idea was unmistakably expressed in the messages that they sent to rulers in Western and Central Asia and even in Europe demanding surrender to the great khan and that show that the term "All under Heaven" (t'ien-hsia) that had originated in China had for the Mongols a much wider and more comprehensive meaning and included virtually the whole known world as parts of their "world empire in waiting." This ideology also lay behind the farreaching expeditions that Khubilai dispatched at enormous cost as far away as Burma, Champa, Java, and Japan. All those countries had, in a loose way, found a marginal place in the tributary system of the Sung, sending missions Cambridge Histories Online © Cambridge University Press, 2008 18 INTRODUCTION when it suited their rulers. But it had never occurred to the Sung court to try to enforce tribute by sending military expeditions overseas, despite their considerable naval power. The imposition by the Mongols of indirect rule over Tibet can also be seen as an expression of their worldwide ambitions. No Chinese dynasty had ever exercised any degree of authority over Tibet; the Sung had been content to stabilize their Tibetan frontiers and left it at that. The huge economic, technical, and military potential of China was thus used by the Mongols in the service of an expansionist ideology whose ambitious aims surpassed anything to which purely Chinese states had ever aspired in the past. TREATY RELATIONS The political organization of the East Asian world before the reunification of China under the Yuan can be described as an age of treaties. Although in principle a treaty recognizing other regimes as either legitimate or equal might seem incompatible with the ideology of an imperial state claiming to rule over All under Heaven, the conclusion of treaties with other powers had a long history in China. As early as the second century B.C. the Han had practiced a policy of appeasement toward the Hsiung-nu and tried to keep this unruly tribal federation away from China's borders through gifts, mostly of silk, and marital alliances, and the making of concessions to neighboring political entities if these were required by circumstances became a recurrent element in Chinese international relations. Such agreements were, however, always considered to be only a temporary and second-best solution in dealing with dangerous enemies, a solution to which a dynasty might resort when all-out victory and submission could not be achieved. The period of the dynasties of conquest in the eleventh and twelfth centuries was, for the Sung, characterized by a high degree of realpolitik. Neither the Khitan nor the Jurchen state could be defeated by military means, and the history of Sung-Liao and Sung-Chin relations became an alternation between relatively short wars and long periods of peace bought chiefly by large payments in textiles and silver. The treaty of Shan-yuan concluded early in 1005 between Sung and Liao became a model for handling later conflicts, for after the Liao state had succumbed to the Chin the latter regarded themselves as the legal successors of Liao and therefore entitled to receive the same gifts from the Sung court as they had done. In addition to the promise of gifts (which the recipients preferred to call "tribute" (kung), a term meticulously avoided by the Sung because of its implications of subjection), the contents of the treaty agreed on the correct delimitation of the borders, together with rules prescribing how to deal with disturbances in the border Cambridge Histories Online © Cambridge University Press, 2008 TREATY RELATIONS 19 region and unauthorized crossings of the frontier. The treaties or corollary agreements also established licensed border markets for state-supervised international trade. But perhaps the most important inherent content of a treaty was the mutual recognition of the two states, and the establishment of regular diplomatic intercourse between them. Routine embassies went both ways, for felicitation at the New Year and on the birthday of the other ruler. Other occasions for sending embassies were death of an emperor or a close relative. These diplomatic exchanges — which were entirely Chinese in concept, ritual, and rhetoric - were regulated in great detail on both sides, and much attention was paid to the ceremonial rules for the travel of the ambassadors and their reception in audience. Apart from such regular embassies it was always the practice to send envoys if problems arose or if a treaty was under negotiation. This intensive diplomatic intercourse required a great deal of paperwork, and the contemporary sources have preserved a huge amount of diplomatic correspondence, so that modern historians find themselves almost overwhelmed by the sheer mass of materials. There is hardly a period in Chinese premodern history for which diplomacy can be studied in such detail as for the eleventh and twelfth centuries. The Sung, and probably also the Liao and Chin envoys, had the statutory duty to write a detailed report of the mission after their return to the home capital. Some of these reports have survived and provide fascinating information on conditions in the northern states as viewed by the Sung envoys, and it must be regretted that no similar accounts from the other side survive. The recognition of the other imperial states did, however, not imply a true equality of status. Their ruler's rank as emperor (buang-tt) was mutually conceded, but with a fine distinction added: Pseudofamilial terminology provided a differentiation in status. The two emperors addressed themselves as older or younger brother or as uncle and nephew so that at least a pretense of nonequality could be maintained. China was, as a result of this development, ruled in theory by one single fictitious "family." This implied hierarchical differences between the rival emperors, for in the Chinese family system no equal status was imaginable. A painful exception from this pattern was, for the Sung, the treaty of 1141, which stipulated that the Sung were subjects (ch'eri) of the Chin, so that the Sung emperor could be addressed in Chin correspondence by his family and personal name, a deep humiliation that was redressed only more than two decades later in 1164 after the Chin invasion of Sung under the prince of Hai-ling had failed and both sides were anxious to resume a policy of peaceful coexistence. From a formal point of view the treaties regulating the bilateral relations Cambridge Histories Online © Cambridge University Press, 2008 2O INTRODUCTION differed from Western concepts insofar as no common paper was drafted to which both sides added their signatures and seals. Instead, each party issued an oath-letter with identical contents in which adherence to the stipulations of the agreement was solemnly promised, invoking Heaven and the gods of the soil. The contents, of course, had to be negotiated beforehand so that the treaty's solemnization by oath became a formality. This procedure took for granted that both sides were acting according to the same standards. In other words, only those states with similar or nearly similarly organized governments and chanceries could be treaty partners. The treaty partners of the Sung — the Liao, Hsi Hsia, and Chin - all had adopted Chinese institutions that could guarantee procedural equality. This kind of equality extended also to the reception of embassies. The rituals surrounding them went back to ancient Chinese concepts. These rites and ceremonies can already be found described in the ritual compilations Li-chi and /-// (the Book of rites and the Book of ceremonials) and reflect the multistate system of the late Chou period. There was thus a continuity linking preimperial China with the sophisticated diplomatic procedures of a later age. There can be no doubt that diplomatic intercourse through embassies and correspondence strengthened and accelerated the absorption of the conquest states into the Chinese world. Not only was the procedure modeled on Chinese precedents; the diplomatic language was Chinese. It does not seem that any letter from Liao or Chin or Hsi Hsia to the Sung was ever written in the native language and script or that a parallel text to the Chinese version was drafted in their own language. We also do not know whether the oathletters of the northern states that had to be ritually deposited in their own ancestral temples were written in the native language. It seems safe to assume that however multilingual the states might have been, the diplomatic lingua franca throughout continental East Asia was Chinese. This had not always been so. When the T'ang emperor Mu-tsung concluded a treaty with Tibet in 821-2, two versions were drawn up, one in Chinese and one in Tibetan, and on the occasion of the treaty the Tibetans erected a stele in Lhasa on which both the Tibetan and the Chinese texts were inscribed. There is no evidence for such bilinguality in treaties from the eleventh to the thirteenth century. Even the letters written by the Mongols to the Sung in the thirteenth century were, it seems, exclusively in Chinese and not accompanied by Mongolian versions. The Mongols used either Chinese personnel or non-Chinese who were sufficiently sinicized to conduct their diplomacy in East Asia, just as the Liao and Chin had done previously. The national Chinese state of Sung was therefore more or less dispensed from studying Khitan, Jurchen, Tangut, or Mongolian. We know that some Sung Cambridge Histories Online © Cambridge University Press, 2008 MODES OF GOVERNMENT 21 envoys had learned to speak a few words of Khitan or Jurchen, but nobody in Sung China could read their script. The rather detailed regulations for Sung embassy personnel never included any sort of language training, nor would any Sung official have been able to read a text in a non-Chinese script. This isolationist and sinocentric attitude underwent a profound change during the early reigns of Ming when a Bureau of Interpreters was established that provided a basic knowledge of foreign languages and scripts for intercourse with their authorities. If China went through a period of disunity until 1276, the political fragmentation was in any case overarched by a common Chinese civilization in many fields including, as we showed, the technicalities of diplomatic intercourse. The regionalism inherent in the political division of China was to some extent balanced by other factors tending to include the "barbarian" states in a larger cultural unity that was Chinese. The borders of the Sung state were never hermetically sealed, even though private travel was not possible for ordinary individuals living on either side of the border. Trade, diplomacy, and, above all, the memory of a common cultural heritage did much to alleviate the political division of China, at least for the intellectuals on both sides of the border. The multistate system that evolved after the Five Dynasties period remained in its fundamental elements very much Chinese despite the many foreign influences in the northern states. It is nevertheless remarkable that outside observers - Marco Polo is an example - did not realize the basic unity of China. For the fourteenthcentury Europeans, Cathay — a name derived from the Khitan ethnic designation and meaning "northern China" — was a country different from Manzi (southern China), and only in the "age of discoveries" in the sixteenth century did Europeans become aware that Cathay and Manzi were in reality parts of that greater unity that we now call China. MODES OF GOVERNMENT It is self-evident that each of the states of conquest in China differed from one another and from the Chinese Sung state in many respects and that their institutions, even if modeled on those of the Chinese, were by no means simple copies of that model. At a higher level of abstraction it is, however, possible to suggest some generalizations that apply in varying degrees to all these states. All their polities suffered from a deep-rooted conflict between nativism and acculturation in ruling the multiethnic territories under their jurisdiction. All were under pressure from the basic antagonism between autocratic and bureaucratic elements on the one hand and feudal—patrimonial Cambridge Histories Online © Cambridge University Press, 2008 22 INTRODUCTION institutions on the other. These conflicting tendencies affected each of those states at all governmental and administrative levels, as the following chapters of this volume will amply illustrate. Personalized authority A common element that emerges most clearly with the Mongols is the strong personalization of imperial power. In a warlike tribal society with few or no government institutions, the personal relations between leader and follower were of primary importance. The chieftain or ruler would select his close companions (the Mongolian term is ndkdr) from among experienced and loyal fighters, whereas outsiders would be looking for a charismatic leader to whom they could offer their services. The Chinese political tradition had no close parallels for such personalized relationships. Even if in later phases the ndkdr tie between ruler and companion was formalized through conferment on the "companion" of a Chinese official title, the social reality still conformed to the inherited usages. Yeh-lii Ch'u-ts'ai (i 190-1244), for example, the famous adviser of Chinggis Khan, appears in Chinese sources as the holder of a Chinese title, but in fact he must be regarded more as a ndkdr of the khan (who honored him with the nickname urtu saqal, "long beard") than as a chancellor in the Chinese sense of the word. An important characteristic of the ruler—companion relationship was that it transcended tribal and national barriers. Whoever was thought able to contribute to the ruler's prestige and power was welcome without regard to his ethnic or social background. In the late Yuan period the personal friends of the emperor were called i-tut, a rendering of the Turkic word inaq, "friend, favorite." This type of relationship was, of course, much more feudal than bureaucratic. Another element that played a great role under the Liao, Chin, and Yuan was the household of the emperor and, in particular, his bodyguards. The exercise of imperial power rested less with the institutions copied from the Chinese than with the households of the ruler and his close relatives. The Altaic term for the household or camp of a ruler is ordo (in Mongolian) or ordu (in Turkic and Mongolian), a word from which eventually the English word horde and its cognates in other European languages were derived. Under the Liao the emperor's ordo was the backbone of his military power and his household organization, which included all sorts of servants, retainers, and higher and lower dignitaries. Many of the general rules for state offices did not apply, it seems, to the management of the ordo, which therefore formed a sort of personal state within the state. Not only the emperor but also the Liao empresses and princes of the blood had their own ordo. The Chinese term used for ordo was wang-fu, "princely administration." Such administrations Cambridge Histories Online © Cambridge University Press, 2008 MODES OF G O V E R N M E N T 23 also existed in purely Chinese states such as the T'ang and Sung but there they were essentially a royal prince's household service and formed a regular part of the bureaucracy. The ordo had far broader functions, was more loosely organized, and was quite different from the parallel Chinese institution. 10 Under the Chin the imperial guards and, above all, the personal regiments {mou-k'o) of the emperor and of the princes functioned to some extent like the Liao ordo.11 The military units of the Chin imperial guards were mostly Jurchens, but the mou-k'o families attached to the imperial clan also included a great number of slaves. Formal princely administrations (wang-fu) were introduced in 1191, but some such administrations must have existed at least twenty years earlier than this. The bodyguards of the Mongols called kesig go back to the times of the dynastic founder and consisted of elite soldiers and trusted followers selected regardless of tribal affiliation.12 No clear difference between the bodyguard and the general household of the emperor can be found. The duties of the members of the guard were not limited to protecting the emperor's person; they also were responsible for the emperor's domestic services, which accounts for some of them holding such titles as ba'urchi, "cook," or ayagchi, "cupbearer. " It apparently was customary for the higher dignitaries of the guards and the household to hold Chinese titles in addition to their Mongolian titles. Under the early Mongolian rulers the kesig organization also assumed the functions of an administrative body. From the time of Khubilai the kesig lost some of its previous political importance following the introduction of more and more Chinese style administrative bodies, but the structural dualism of coexisting bodyguard—household and formal bureaucratic administration continued until the end of Yuan. An important function of the guards was always that of a reservoir of reliable manpower at the emperor's disposal for employment as his personal agents. Even as late as 1346, guard members (kesigden) were dispatched to the provinces as supervising governors (darughachi). Autocrats and joint deliberation Recent research has done much to modify earlier theories according to which the dynasties of conquest were governed by despotic autocrats. The personal power and prestige of their rulers were sometimes extremely great, but it can 10 SeeJing-shenTao, Thejunhen in twelfth-century China: A study ofj/mriztf/ion (Seattle, 1977), pp. 46—51. 11 SeeMikami Tsugio, Kindai Joshin shakai no kenkyii vol. 1 of his Kimhi kenkyii (Tokyo, 1972), pp. 109— 418. This work, first published in 1937 under the title Kindai Joshin no Kenkyii, was extensively revised and supplemented by the author and reprinted in his collected studies on Chin history under the title Kindai Joshin shakai no kenkyii. 12 See Hsiao Ch'i-ching, The military establishment of the Yuan dynasty (Cambridge, Mass., 1978), pp. 37-48. Cambridge Histories Online © Cambridge University Press, 2008 24 INTRODUCTION be shown that all peoples who conquered parts or the whole of China had as a part of the heritage of their tribal past a strong tradition of deliberative and joint decision making. The early Khitans selected their leaders at a council of tribal chieftains and frequently held such a council when planning a campaign. The Jurchens, too, had the custom before a campaign of convoking a military assembly at which the action to be taken was discussed by all those present, including common soldiers. This and other similar customs have induced some modern scholars to describe early Jurchen society, perhaps somewhat overenthusiastically, as a "military democracy." Such traditions continued even after the founding of imperial states modeled on Chinese dynasties. We know, for example, that at the Chin court in 1197, votes were taken among higher-ranking officials in order to determine what course should be taken against the Mongolian menace. The organization of such a ballot to decide military issues was a survival of native traditions of joint decision making and at the same time served as a check on the emperor's autocratic power. Similar war councils at which policies and tactics were discussed also existed among the Tanguts. By far the most illuminating and best-explored examples of joint decision making were the Mongols' diets or tribal convocations, the kburiltai. At these meetings a new ruler would be elected or proclaimed, a process that can be sufficiently explained only if one takes into account that the Mongolian empire was somehow regarded as the family patrimony of the Chinggisid family. As there was no other formal regulation for the succession, every male member of the ruler's clan believed himself a potential successor with a claim to the throne. The khuriltai assemblies were not electoral in the strict sense of the word and did not involve the casting of votes. The claimants came to the khuriltai accompanied by their armed followers, and considerations of power, prestige, and charisma influenced the ultimate proclamation. A consensus did not always emerge as the result of the assembly. Sometimes dissenting pretenders would convoke their own khuriltai, and for this reason even the reign of Khubilai was more than once threatened by other rival claimants to the supreme dignity of khaghan. Such pretenders might come from his own lineage, like Khubilai's younger brother Arigh Boke, or from another Chinggiskhanid line, like his rival Khaidu. All this points to a precariousness of imperial power and its transmission. The fact that so many rulers of the Liao, Chin, and Yuan eliminated their predecessors or rivals by murder or banishment may be seen as a direct result of the absence of fixed rules for inheritance and also of the unpredictable consensual element in proclaiming a successor. It cannot be attributed to a typically "barbarian" savagery. The consensual element was also present in the Yuan government system at lower levels. Collective and deliberative decision making was a standard Cambridge Histories Online © Cambridge University Press, 2008 MODES OF GOVERNMENT 25 practice in most administrations. The officials had to appear at daily conferences, and failure to attend was a punishable offense. Such sessions took place every morning in the offices of the capital and the regional administrative agencies, down to the counties. The attending officials had to sign their names or, because many were illiterate, to append their name-seal in the register to prove that they were present. Among the Chinese terms for these sessions were yuan-tso, "to sit in a circle," and yiian-i, "to deliberate in a circle," which evoke for us the image of a roundtable conference. Such procedures differed from the normal Chinese government practice, in which responsibility for decisions was individual rather than collective, and they may often have resulted in delay, indecision, and a lack of common purpose in the bureaucracy. Central authority A study of the governmental systems of the dynasties of conquest also reveals a relatively low degree of centralization, contrary to earlier assumptions that tended to depict their governments as highly autocratic and centralized. The Liao dynasty had from early times a dual system of administration, one for the Khitans and other tribes and one for governing the Chinese population, called, respectively, the Northern and Southern administrations. Their system of government was a chaotic jumble of offices, and their areas of responsibility were ill defined. The personal power of great officials far outweighed the prestige or authority conferred on them by office. A similar system can be observed for the early reigns of the Chin dynasty. The Jurchen political system was in the early twelfth century based on the institution of po-cbi-lieh, a Jurchen word that may be loosely rendered as "appointed chieftain." Under the dynastic founder T'ai-tsu the po-chi-lieh were primarily concerned with governing the Jurchen population. In 1126 a Chinese bureaucratic system was formally introduced, but only for the subject Chinese population. The early Chin system of government seems therefore to have been a conscious imitation of Liao dualism with, however, the important difference that the territorial spheres of jurisdiction and administration were less clearly defined than under the Liao. The interplay between the two types of administration was quite convoluted, and the history of early Chin governmental structures is therefore singularly complicated. Ruthless action to restructure the Chin government was taken by the prince of Hai-ling. He tried to eliminate the influence of the Jurchen aristocracy and to restructure his state according to Chinese models and to introduce stronger centralization by any means, including bloody purges. He also abolished the hereditary offices of most of the meng-an mou-k'o commanders and tried to transform their posi- Cambridge Histories Online © Cambridge University Press, 2008 26 INTRODUCTION tions into those of regular officials who did not inherit their office automatically but had to be appointed and could be dismissed, if necessary, by the imperial government. The introduction of a more Chinese bureaucratic system in imitation of T'ang and Sung models resulted, however, in despotism under a Chinese disguise, for the prince of Hai-ling was skillful in manipulating Chinese political traditions so as to augment his own personal power. He also seems, on the other hand, to have realized that although Chinese institutions might centralize authority, a full and thorough adoption of Chinese governmental values might have diminished or checked his own personal power. Up to the end of the dynasty the Chin governmental system thus remained a mixture of native traditions with Chinese bureaucratic practice. Centralism under the Mongols was, despite the popular belief that they exercised strong centralized power, very limited. The fact that they reunified China has sometimes obscured the other fact that their government was characterized by a striking lack of system and a sometimes chaotic fragmentation of authority. Tribal units under recognized hereditary leaders continued to play a prominent role, and their leaders exercised strict personal control over their own subjects. A special feature of Mongolian rule in China was the numerous appanages granted to members of the imperial clan, imperial relatives, and meritorious generals. Such fief holders frequently had their own armies, and their territories were fiscally more or less exempt from the control of the financial administrations responsible for taxing the empire as a whole. A functional bureaucracy of the Chinese type was first (and imperfectly) introduced after the Mongols had annexed the northern half of the Chin state in 1214—15, and further effective steps to create an efficient bureaucracy were taken only much later, chiefly under Khubilai. But even then the sinicization of the state structure was anything but uniform. It has, for example, been said with good reason that the provinces (bsing-sbeng) of Yuan China had quite a different character from the Sung provinces and were more like governments of external territories or separate vassal states surrounding the metropolitan domain. They were internally centralized to some extent but were rather loosely linked with the imperial domain around the capital of Ta-tu (Peking).'3 Viewed from this angle, Yuan China appears almost as a conglomeration of regions under strong regional governments. This relative lack of powerful central control certainly contributed to the gradual disintegration of the state after 1340 when local rebellions and secessionist warlords threatened the empire's unity. 13 See David M. Farquhar, "Structure and function in the Yuan imperial government," in China under Mongol rule, ed. John D . Langlois, Jr. (Princeton, 1981), pp. 52—53. Cambridge Histories Online © Cambridge University Press, 2008 MODES OF GOVERNMENT 27 Even the military system of the Mongols was not strongly centralized. There was a Bureau of Military Affairs (Shu-mi yuan) in the capital that established norms for military units throughout China and beyond, but a system of direct subordination and a strong line of command involved only the imperial guards and a few other contingents in north China. The guards themselves were a mixed body of regiments recruited from many nationalities, ranging from the Ossetes of the Caucasus region to the Jurchens from Manchuria. Another unusual feature of Yuan government must also be mentioned in this context. One of the most important ministries was the Court for Buddhist and Tibetan AfFairs (Hsiian-cheng yiian).'4 It had quite diverse and seemingly incompatible duties: On the one hand it supervised the Buddhist clergy in the Yiian state, and on the other it acted like a provincial government for Tibet and the adjacent regions, with great powers that even included mobilizing military expeditions in times of unrest. This, however, did not imply the subordination of authority in local government in Tibet to the Mongol central government. The heads of this agency were in most cases Lamaist Buddhist clerics. All this was not only alien to the Chinese political tradition but also provides another example of the unsystematic structures of the Yiian government. China under the Yiian was anything but a monolithic centralized state, despite the fiction offered in the Yiian shih that a Chinese centralized civil administration prevailed. Fragmented legal systems Fragmentation rather than homogeneity was also characteristic of the legal systems in the states of conquest. Chinese traditional law did not leave much room for the recognition of ethnic differences, and once a non-Chinese ethnic group had been absorbed into the state territory, their legal treatment followed Chinese statutory law. The only exception from this rule can be found in the T'ang code, which ruled that offenses committed between nonacculturated foreigners {hua-wai jen, "people outside civilization") should be judged according to their native customs. If such persons committed offenses against Chinese, they were prosecuted and punished according to the articles of the code.'' The basic principle according to which the application of law follows territorial criteria is in legal theory called ius soli. It is opposed to the personality principle {ius sanguinis), which allows for different legal treatment 14 See Herbert Franke, "Tibetans in Yiian China," in China under Mongol rule, ed. John D. Langlois, Jr. (Princeton, 1981), pp. 2 9 6 - 3 2 8 . 15 See Tangliisbui (Peking, 1983 ed.), {article 48], 6, art. 4 , p. 133; Wallace Johnson, ThtTangcodi. Vol. 1: General principles (Princeton, 1979), p. 252. Cambridge Histories Online © Cambridge University Press, 2008 28 INTRODUCTION of different ethnic groups. All states of conquest were multinational and included a large Chinese population; their legal systems generally applied the principle of ius sanguinis. Under the Liao, Chinese law — that is, codified T'ang law - was used for the Chinese and the Po-hai population, with, however, some modifications that consisted chiefly of harsher punishments than those provided by T'ang law. Tribal customary law was applied to Khitans and other non-Chinese ethnic groups. The Liao made no attempts to create a comprehensive and systematic code, although on several occasions, existing rules and ordinances were collected together and promulgated.'6 The Tangut, by contrast, created very complex codified laws, written in Tangut, that are an amalgam of T'ang law and Tangut custom. This code has survived in large part, and a translation is now available.'7 Throughout the twelfth century the laws of Chin remained a mixture of Chinese law and customary law of the Jurchens and other ethnic groups. Chinese (T'ang) law was adopted only gradually, a process that culminated in the T'ai-ho code (T'ai-ho lu), which was promulgated in 1201. The T'ai-ho code was to a large extent based on the T'ang code and survived the fall of Chin in 1234, for even after the Mongolian conquest of northern China, it was still applied to the Chinese population.'8 It was abrogated only in 1271 after the Mongolian regime had adopted the dynastic name of Yuan. But in spite of their code, Chin law included not a few principles that were alien to Chinese legal theory and practice, chiefly in family and inheritance law. Among these we might mention the toleration of levirate and the permission for sons to set up their own households during the lifetime of their parents. Compared with T'ang law, the stipulations of the T'ai-ho code, of which fragments have survived, were sometimes more draconian and tended to strengthen the authority of the heads of the family over wives and junior relatives. The differentiation of law and legal procedure under the Mongols was even greater than under the preceding dynasties. Jurisdiction was fragmented according to nationalities.'9 For example, the Bureau of Affairs of the Imperial Clan (Ta tsung-cheng-fii), which had the functions of a court of appeal, 16 Herbert Franke, "Chinese law in a multinational society: The case of the Liao (907-1125)," paper presented to the History of Chinese Medieval Law Conference, Bellagio, Italy, August 1981; and "The 'Treatise on punishments' in the Liao history," Central Asiatic Journal, 27 (1983), pp. 9—38. 17 See Evgenii I. Kychanov, Izmennyi i zanovo utverzhdennyi kodeks daiisa tsarstvovaniia nebanoeprotsvetanie ( / / 4 9 - 1 1 6 9 ) , vol. 1 (Moscow, 1988); vol. 2 (Moscow, 1987); vols. 3 and 4 (Moscow, 1989). 18 See Herbert Franke, "Jurchen customary law and the Chinese law of the Chin dynasty," in State and law in East Asia: Festschrift Karl Biinger, ed. Dieter Eikemeier and Herbert Franke (Wiesbaden, 1981), pp. 215—33; also see Franke, "The legal system of the Chin dynasty," in CollectedStudies on Sung history dedicated to Professor Janus T. C. Liu in celebration of his seventieth birthday, ed. Tsuyoshi Kinugawa (Kyoto, 1989), pp. 3 8 7 - 4 0 9 . 19 See the introduction of Paul Ratchnevsky, Un code da Yuan (Paris, 1937), vol. 1, pp. v-xcix. Cambridge Histories Online © Cambridge University Press, 2008 MODES OF GOVERNMENT 29 had only jurisdiccional powers over Mongols. Cases involving Central Asians were decided at the appeal level by the Bureau of Guardianship (Tu-hu fu). There were a few exceptions to the principle of ethnicity. One of these concerned mixed marriages. Even if only one of the marriage partners was Mongolian - husband or wife - Mongolian law had to be applied. The introduction of mixed courts must also be seen as an expression of the ius sanguinis principle. For example, all cases between Uighurs, who were subjects of their holy king, iduq qut of Khara Khocho in Turfan, and Chinese had to be tried by a mixed court. Special mixed courts were also introduced for certain professional groups, including military personnel. Serious offenses committed by members of the Buddhist or Taoist clergy came under the jurisdiction of the normal civilian courts, but less serious disputes between laymen and clergy were jointly decided by the chief monk and the local civilian official. Cases between the physician and a patient or the patient's family were to be decided by a spokesman for the medical profession and the local official. The same procedure was followed for cases between members of the musicians' guild and outsiders. In this way, the principle of personality, ethnic, and professional affiliation pervaded the whole legal system under the Yuan. Law and the judicature were fragmented to a high degree. Moreover, the Mongolian regime did not have a comprehensive and systematic code like those of the T'ang, Hsia, Chin, and Sung. Judicial practice followed the multitude of individual regulations and ordinances collected in several handbooks. Some of these have wholly or partly survived, and so it is possible to study the Yuan legal system in far more detail than is possible for the Liao and Chin. The standing of officials A semi-legal practice that had a deep influence on the atmosphere at court was that of "court beatings." Under all the conquest dynasties, officials of all ranks could be punished with beatings, on the order and in the presence of the emperor himself. At lower administrative levels, too, officials were not exempt from corporal punishment. Such corporal punishment of officials had been common under Sui Wen-ti. 2 ° Court beatings had sometimes been carried out under the T'ang, but only in exceptional cases.21 By contrast the Sung adhered not only in theory but also in practice to the age-old principle 20 On Sui Wen-ti's capricious and atrocious cruelty to his officials, see T'ang Ch'eng-ye, Sui Wen-li cheng chihshih kungcbihyin chiu (Taipei, 1967), pp. 81—3. 21 Some cases occurred during the "reign of terror" under Empress Wu. The most notorious case, that of Chiang Chiao, an imperial favorite of high rank who was flogged at court under Hsiian-tsung and subsequently died in 722, was exceptional in that his crime was betraying a personal confidence of the emperor. His punishment provoked heated protests. The practice was never revived in the last half of the T'ang. See Chuang Lien, Ming Ch'ing shih shih Is' ung fan (Taipei, 1972), pp. 4—5. Cambridge Histories Online © Cambridge University Press, 2008 30 INTRODUCTION according to which corporal punishments did not extend to the gentry, and the rites did not extend to the commoners. Such punishments were never imposed on officials under the Sung. The conquerors, however, disregarded this traditional Chinese privilege, and such humiliating beatings became a normal feature of government.22 The flogging of officials, particularly beatings administered to ministers at the imperial court, can be interpreted as evidence of barbarian brutality and of imperial despotic behavior. It may, however, also be viewed as the result of egalitarian tendencies under these alien regimes that negated the fundamental social and legal barriers that had strictly separated the members of the bureaucracy {kuan) and the masses of commoners (tnin) in traditional China. In general, under these dynasties a strong, personalized, arbitrary autocracy at the court and in the immediate entourage of the emperors was paralleled by an unsystematic administration characterized by the fragmentation of authority and even the laxity of routine administration. Their states were not monolithic and were weakened by diffuse lines of command. The brutality and unrelenting autocracy that the founder of the Ming dynasty so frequently displayed may have been a heritage of the sometimes brutal behavior of the Yuan rulers,23 but it is also perhaps justified to regard the first Ming emperor's despotism as a firm attempt to restore and strengthen imperial power and to get away from the unstructured, loose, and even chaotic features of Yuan government. He once himself attributed the fall of the Yuan to their negligence, laxity, and leniency, and from this perception of his predecessors he tried to preclude similar developments that might endanger the state and his power as its emperor. If one follows this interpretation, the Ming state had to be made strong because the state itself had been seriously weakened under successive alien regimes. MULTILINGUALITY The problems of governing and administering a polyethnic society are inevitably linked with the linguistic situation. It is true that alien regimes had existed previously in Chinese history, but for the foreign-dominated regimes of the Six Dynasties period we know next to nothing about the use of the conquerors' languages in the Chinese-style states that they had founded, and none of them had a written language that could be used in government and administration. From the period of conquest in the tenth to fourteenth 22 See Franke, "Jurchen customary law," pp. 2 3 1 - 2 . For a representative selection of cases under the Liao, Chin, and Yuan, see Chuang Lien, Ming Ch'ing sbih sbib ts'ung fan, pp. 1—10. 23 See F. W. Mote, "The growth of Chinese despotism: A critique of Wittfogel's theory of oriental despotism as applied to China," Orieta Extrtmul, 8 (1961), pp. 1—41. Cambridge Histories Online © Cambridge University Press, 2008 MULTILINGUALITY 31 centuries, by contrast, we have ample evidence and even surviving texts that show clearly how and to what extent the Chinese language and script were used side by side with other languages and writing systems. Never before had Chinese been challenged as the sole language of administration as during that period. The overwhelming majority of the population in China proper was, of course, Chinese, and the conquerors had to adapt their ways of government to the multilinguality of their territories. As far as the spoken language is concerned, the foreigners in the bureaucracy relied largely on interpreters if they did not themselves know enough Chinese. For this reason all dynasties of conquest had interpreters attached to their offices so that they could deal with the Chinese population, particularly in lawsuits and trials. The use of intermediaries not only complicated the procedure but frequently led to abuses. Bribing an interpreter could become a means of influencing a judge's decision without his knowing. On the other hand, linguistic competence became an important factor in rising in the bureaucracy, as can be seen in the careers of many Yuan ministers. Unlike the earlier conquerors like the Hsien-pei or T'o-pa, who never had a script of their own, each of the alien rulers between the tenth and fourteenth centuries ordered the creation of a national script. These graphic systems invented to write non-Chinese languages are of great interest to linguists, and the ongoing decipherment of some of them must be regarded as a brilliant achievement of modern scholarship. The Khitans introduced in 920 the so-called large Khitan script and in 925 the small script that seems to have been syllabic; the single elements or syllabographs are either simple Chinese characters or shaped like artificially constructed Chinese characters.24 The Jurchens, as well, developed two scripts, also called large and small, in 1119 and 1138, respectively. For some time three written languages, each with their own scripts, were used simultaneously in the Chin state: Chinese for the Chinese and Po-hai, Khitan for the Khitans, and Jurchen for the Jurchen sectors of state administration. Later in 1191-2 the Khitan script was officially abolished, and thereafter only Chinese and Jurchen were recognized as legitimate. Unfortunately, few specimens of the Khitan and Jurchen scripts have survived; we have no official documents on paper or silk, only epigraphic monuments such as inscribed steles or inscriptions on seals and metal implements, plus a few graffiti on walls or ceramics. Little progress has been made in reading Khitan. Although the Jurchen script was graphically modeled on 24 For a succinct discussion, see Daniel Kane, The Sino-Junhen vocabulary of the Bureau of Interpreter! (Bloomington, 1989), pp. n - 2 0 ; and Ch'ing-ko-erh-tai et al., Ch'i-tan hsiao nu yen chiu (Peking, 1985). Cambridge Histories Online © Cambridge University Press, 2008 32 INTRODUCTION Chinese, Jurchen characters can in only a few exceptional cases be linked with Chinese originals; rather the majority of Jurchen signs, which included both semantograms and phonograms, were inventions. They can, however, be deciphered because of the survival of a bilingual Chinese—Jurchen glossary of the early sixteenth century, produced by the Ming Bureau of Interpreters.2' The Tanguts also invented a script that at first glance looks like Chinese but is in fact entirely unrelated. It was based on exceedingly complicated principles including many ideographical compounds. The decipherment of its more than six thousand different characters was made possible by the great number of surviving specimens, including epigraphy, manuscript texts, and printed books, among which were many texts translated from Chinese. It was once assumed that the Tangut script disappeared with the extinction of the Tangut state, but in fact it continued to be among the Tanguts right through the Yuan dynasty, and the last datable specimen of Tangut writing is a Buddhist stele dated 1502.26 The Mongols, who had no national script of their own when they rose to power after 1200, had the great sense not to try to invent a complex writing system like that of their predecessors but instead used the alphabetic script of the Uighurs to write Mongolian. This is still today the official script of the Mongols in the Inner Mongolian Autonomous Region of the People's Republic of China. No decipherment was therefore needed, nor was it any problem for modern scholars to read texts in the second national script for Mongolian, which was devised by the Tibetan 'Phags-pa Lama (1235-80) and was declared to be the national script in 1269. It was designed to be a universal script for writing all languages and was. based on the Tibetan alphabet. The letters, however, were not arranged horizontally as in Tibetan but vertically so that the script could be used as an interlineary text together with Chinese. Although no official bilingual documents in Chinese and Mongolian have survived, with the exception of copies of such documents inscribed on steles, it has been ascertained that a great part of the Chinese juridical texts written in colloquial language that survive in several Yuan period collections go back to the Chinese interlineary versions that accompanied the documents' original Mongolian texts. As a result, the Chinese of these texts is ungrammatical because the words follow the very different Mongolian word order and syntax. The official use of colloquial Chinese was in itself an innovation, for until the late thirteenth century, only literary Chinese had been used in 25 See Kane, The Sino-Junben vocabulary, for the most up-to-date study. 26 See the note by Hsu P'ing-feng in Chung-kuo she-hui k'o-hsiieh yuan k'ao-ku yen-chiu so, ed., Hsin Cbung-kuo u k'ao ku fa hum ho yen chiu (Peking, 1984), p. 631; and Cheng Shao-tsung and Wang Ching-ju, "Pao-ting ch'u t'u Ming tai Hsi Hsia wen shihch'uang," K'ao kubjuehpao, 1(1977), vol. 1, PP- I 3 3 - 4 1 - Cambridge Histories Online © Cambridge University Press, 2008 MULTILINGUALITY 33 government and administration. Even under the Mongols it was, moreover, never totally displaced by colloquial Chinese, and throughout the Yuan dynasty many edicts and ordinances were still written in the literary language. We may therefore say that even within the use of Chinese language and script, some kind of bilinguality developed under the Yuan. Translation The pro-Chinese intellectual elites among the foreigners not only frequently achieved great skill in writing literary Chinese but also invariably tried to acquaint their compatriots with Chinese literature through translations. This, too, must be regarded as an innovation; earlier conquerors in the Six Dynasties period could not give their own people translations of Chinese texts, as they had no written language in which to record them. Not surprisingly, under the conquest dynasties the selection of Chinese texts for translation was largely guided by considerations of what was regarded as useful for governing the Chinese. Although information on translations into the Khitan language is scanty and fragmented, it is known that in addition to laws and medical works, some Chinese histories were translated, namely, the ninth-century general history Tung li by Ma Tsung (d. 823) and the Chiu wutai shih (the [old] History of the Five Dynasties), the latter chosen perhaps because the period of the Five Dynasties coincided with the rise of the Khitans to imperial power. Another historical work translated under the Khitan was the Chen-kuan cheng-yao (Important principles of government from the Chen-kuan period). This collection of conversations between the T'ang emperor T'ai-tsung and his ministers provides a vividly written handbook on T'ang statecraft that, because of its political style and pragmatic contents, was much appreciated by all the non-Han conquerors. It was later translated into Tangut, Jurchen, and Mongolian and, centuries later, into Manchu. Among the Khitan translations the Confucian classics are noticeably absent, which is surprising because they have always been in the eyes of the Chinese an essential guide to ruling a state and regulating social relations. Khitan emperors and ministers knew and made use of them but apparently read them in Chinese. A much fuller selection of Chinese works was translated into Jurchen. Many Confucian classics were translated, including the Analects (Lun-yii) and the Meng-tzu. A few Taoist books such as the Tao-te ching were also translated. Among the histories we find translations of the Ch'un-ch'iu (Spring and autumn annals), which was of course also one of the Confucian classics, and of such standard dynastic histories as the Shih-cbi (Records of the grand historian), the Han-shu (History of the Han dynasty), and the Hsin T'ang shu (New Cambridge Histories Online © Cambridge University Press, 2008 34 INTRODUCTION history of the T'ang). Also translated was a collection of seventy-nine model examination essays (the Ts'e-lin) by Po Chii-i (772-846), which might have served as an examination crib for Jurchen candidates. From Korean sources we know that handbooks on strategy and the primer Ch'ien-tzu wen (Onethousand-character classic) were also available in Jurchen versions. It seems therefore that the Jurchens were more eager than the Khitans were to have their nationals informed about Chinese history and civilization. Unfortunately, not even the tiniest fragment of either the Khitan or the Jurchen translations has been preserved. The case is different for the Tanguts, whose translation activities were even more comprehensive than those of the Jurchens. In contrast with Khitan and Jurchen translations, we do not have to rely on secondhand information, as the many Tangut texts that have been discovered include translations of Confucian classics like the Lun-Yii (Analects) and the Meng-tzu, encyclopedias, and works on statecraft. Military treatises such as the Sun-tzu ping-fa (Master Sun's art of war) have survived in Tangut versions, and medical handbooks were translated from Chinese together with texts on veterinary medicine, the latter an important field of knowledge for the horse-exporting state of Hsi Hsia. The selection of Chinese texts for translation into Tangut was based on considerations of practical usefulness. The same is true for translations of books into Mongolian, but with the notable absence of Chinese handbooks on military strategy, presumably because the Mongols who had conquered the whole of China did not think that much could be learned from the Chinese strategists of antiquity. Printed editions of translations from Chinese are known to have existed and included the Hsiao-ching (Book offilialpiety), the Confucian exegetical work Ta-hsiieh yen-i (Expounded meaning of the Great Learning), the Chen-kuan cheng-yao (Important principles of government from the Chen-kuan period), and several handbooks on statecraft and administration. Of these, only the Hsiao-ching in Mongolian has survived, and a small fragment of what was probably the Mongolian translation of the administrative handbook Ta-Yiian t'ung-chih (Comprehensive regulations of the great Yuan). Other translations were made under the Yuan but never printed, such as books on medicine and pharmacopoeia, works on political ethics, and the Shu-ching (Book of documents). Among the titles of unprinted translations into Mongolian are also several works of an educating and edifying nature, for example, a collection of anecdotes taken from Chinese history, and gnomic works. The Chinese originals from which some of these translations were made cannot be identified. All of this concerns what we may term secular literature translated from Chinese. In terms of quantity, however, the translated Buddhist literature Cambridge Histories Online © Cambridge University Press, 2008 MULT1LINGUALITY 35 must have far surpassed the secular literature. Although Buddhism was both popular and enormously influential under Liao and Chin, it is not known whether Buddhist texts were ever translated into Khitan or Jurchen. The Tanguts, on the other hand, produced a huge corpus of Buddhist scriptures in their own language and script, translated from the Chinese versions. The Buddhist canon in Tangut was printed in Hang-chou in 1302, long after the extinction of the Hsi Hsia state, and some volumes of that edition have survived, in addition to the great number of Buddhist Tangut texts found in Khara-khoto. Under the Yuan, many Buddhist works were translated into Mongolian, partly from Chinese and partly from Tibetan, and some fragments of printed Buddhist scriptures have been discovered in Central Asia, mainly in the Turfan region. These represent, however, only a small fraction of the corpus of Buddhist translations done under the Yuan. An impressive monument of the multilinguality among Yuan period Buddhist believers are the inscriptions at the Chii-yung Pass north of Peking. These pious texts exalting the religious activities of the emperor recorded in six languages - Sanskrit, Chinese, Mongolian (in the 'Phags-pa script), Uighur, Tibetan, and Tangut - were inscribed on the interior walls of a huge gateway that was built in 1345 under the last Mongolian emperor of China. It is perhaps a legitimate question to ask to what degree the multilinguality of the states and their administrations was perceived by the masses of the subjects. The answer can, at best, be speculative. Trials presided over by a foreign judge who knew no Chinese might well have been the most frequent occasion for a Han Chinese commoner to be confronted with a language problem. The lowest echelons of administration that had the most direct contact with the ordinary people were more largely staffed by Chinese personnel. It was only in the higher political and military offices whose activities impinged only indirectly on ordinary people that the foreigners were overrepresented in relation to their overall numbers. This was certainly the case under the Liao, Chin, and Yuan; there is practically no information available on the ethnic composition of the Hsi Hsia bureaucracy. We could advance the hypothesis that under all the dynasties of conquest, millions of rural Chinese may well have never encountered a foreigner, at least in the countryside. The cities, as a rule garrisoned by foreign soldiers, were a different case. In the same way, not many Chinese outside the cities would have ever seen a document in a foreign language and script. There was, however, at least one feature of life among the population at large that constituted a constant reminder of foreign domination. This was money. Although there were fewer coins with a Khitan inscription than coins with a Chinese inscription, such coins are known to have existed. Only one specimen of a coin with a Khitan inscription has been discovered so far, a cash Cambridge Histories Online © Cambridge University Press, 2008 36 INTRODUCTION datable to the reign period from 1095 to n o 1. As far as we know, the Jurchens never cast cash with Jurchen inscriptions but used only Chinese script on their coins. The Tanguts issued coins inscribed in both Chinese and Tangut. The paper money of Chin was exclusively printed in Chinese, with no Jurchen characters added. Under the Yiian, coins were cast that showed an inscription in Chinese but spelled with 'Phags-pa characters, and the same is true for Yiian paper money. The surviving specimens of Yiian paper money have a long inscription in Chinese, and only the official designation of the bills was repeated with a transliteration of the Chinese words in 'Phags-pa script. Everybody who handled money in the Yiian period was therefore aware that the state issuing the notes and coins was only partly Chinese. People in the market had yet another opportunity of being reminded of the multilinguality of the state: Some officially approved weights have survived bearing inscriptions in Chinese, Mongolian, and Persian (in Arabic script). THE HAN CHINESE UNDER ALIEN DOMINATION What were the effects of this long period of foreign domination on the Chinese population? Certainly the conquests themselves involved great loss of life, widespread destruction of property, and the disruption and dislocation of society at every level. This was least with the Liao, who acquired their Chinese territories by negotiation and under whom disruption and destruction affected far more widely the former population of Po-hai, and with the Hsia, who seem to have taken most of their territory in Kansu intact from the incumbent regimes. The Chin conquest of the Liao empire met with only halfhearted resistance and did not cause widespread destruction, but their conquest of Sung territories in north China involved long years of bitter warfare and much physical damage and social disruption. The Mongolian conquest of Hsia and Chin was destructive out of all proportion to these earlier campaigns. The Mongolian armies all but obliterated the central territories of Hsia in modern Ning-hsia, and in their conquest of northern China they wreaked havoc and destruction on settled populations on the same scale as they had done in Iran, Russia, and northern India, destroying cities, massacring their populations, and even contemplating turning north China into pasturelands for their herds. For a quarter of a century thereafter, north China suffered endemic warfare and administrative chaos. At the end of the Chin in 1207, its empire had had a population of about 53 million, at a time when China as a whole is usually thought to have had between 110 and 120 million people. By the end of that century in 1290, the total registered population of China had fallen to less than 60 million and remained at that level in the late fourteenth century Cambridge Histories Online © Cambridge University Press, 2008 THE HAN CHINESE UNDER ALIEN DOMINATION 37 under the early Ming. The loss was particularly severe in the northeast. The population of Hopei and Shantung fell catastrophically to little more than one-third of its former level in the eighty years between 1207 and 1290. Many factors make the interpretation of these figures difficult. But there was clearly a huge loss of population during the thirteenth century, and there is every reason to attribute much of this to the destruction wrought by the Mongolian armies in northern China, although other factors probably played their part. Southern China, by contrast, was invaded and conquered by a Mongolian Yuan regime already firmly rooted in China and inured to Chinese ways. Khubilai had every reason to attempt to take southern China intact as far as possible and with its productive base undamaged, but here too populations declined seriously over the thirteenth century, though not to the catastrophic levels in the former Chin territories. The different Mongolian policies enforced in their conquests in north China down to 1234, and those imposed in the south in the 1270s, exaggerated population trends already established under Sung. Although China's population steadily grew during the eleventh and twelfth centuries, its distribution began to change radically. In 742 under the T'ang more than 60 percent of the Chinese people lived north of the Huai River. But by the twelfth century the situation was reversed, with the majority living in the south. Even though China's population as a whole had doubled, the population of the northwest had actually declined, and that of the northeast remained static, although a sizable population came to be concentrated around modern Peking, a major metropolis under Liao and Chin. The Yuan conquest, which devastated the north but left southern China relatively untouched, accelerated this trend and left the northwest moribund and the northeast a relative backwater apart from the region around their capital Ta-tu (Peking). Parts of Hopei did not recover their mid-T'ang levels of population until the sixteenth century, and they have never recovered their relative importance. The conquests themselves were not the only destructive forces unleashed on the Chinese population by these governments. All the alien dynasties were military regimes in origin, and all engaged in frequent wars with their neighbors, both in the sedentary realms of China and Korea and in the northern steppe. Huge armies were conscripted for their campaigns, both from the tribal populations, which had always lived in a state of constant readiness for warfare and provided their cavalry, and from their Han Chinese subjects, who were used as infantry in campaigns against sedentary states where static warfare and the reduction of walled cities were necessary. Some of their campaigns were immensely costly in lives; the Liao invasions of Korea, the Hsia's constant wars with the Sung and Chin, and the abortive invasions of Japan under Khubilai are just a few random examples in which Cambridge Histories Online © Cambridge University Press, 2008 38 INTRODUCTION losses were numbered in many tens of thousands. The huge costs of constant warfare also put a great strain on the material resources of all the states of this period: Much of the benefit from the vast increase in productivity in Sung China was dissipated in supporting a huge army. The conquest dynasties subjected their settled populations to constant arbitrary levies of labor and to requisitions of goods needed by their military. Their command economies gradually developed an ever-more complex system of households designated as producers of specific products or as providers of special services, including military services, for the state. Also connected with the organization of these dynasties was a greatly increased number of slaves and semiservile dependents, both state owned and in private hands, and a broad tendency toward personal subordination in society. Slavery was especially widespread under the Jurchen Chin. Each of these dynasties seems to have permitted members of the ruling house to establish huge personal followings and to administer their own personal territories outside central control. It is realistically impossible to compare the situation of the average Han Chinese household living under the Sung with that of their contemporaries living under any of the conquest dynasties: The level of taxation differed little, and the day-to-day government was not more oppressive. Many people in the countryside probably rarely set eyes on one of their alien rulers. In the cites, however, things were different; troops and administrators were stationed there, and each of the conquering dynasties brought with them groups of non-Han merchants and clients who played some part in government, Uighurs under the Khitan and Uighurs and men from all over Western and Central Asia (se-mu) under the Yuan. For the educated Chinese elite, imbued through their education with ideas of state service as officials, accommodation with the new order was more difficult. The regions of China controlled by the Khitan and Tangut had never supported a large, well-educated elite. They had always been cultural backwaters, and even in A.D. 900 they had for centuries been under military rule, with the local literati playing a minor role. But in both states, as they grew to maturity there was great need of literate men in government service. Many of these came from the Han Chinese population. The Liao eventually had its own examination system, its own Han-lin academy (which was sufficiently important to be given a Khitan name), its own state historiographers, and its Confucian scholars who expounded the classics to its emperors and heirs apparent. The Southern Administration was almost entirely staffed by Han Chinese officials. Educated men wrote, published, and acquired surreptitiously the writings of Sung authors. Buddhism flourished mightily under imperial patronage, and many of the clergy were probably Chinese. Chinese Cambridge Histories Online © Cambridge University Press, 2008 THE HAN CHINESE UNDER ALIEN DOMINATION 39 art, too, continued. At least one early Khitan prince was a sufficiently fine painter for his works to be included in Sung Hui-tsung's imperial collections: Some survive today. The wall paintings discovered in Liao tombs are lively and impressive descendants of the vigorous Chinese tradition offigurepainting familiar from early T'ang. Liao architects laid out cities on Chinese (or Po-hai) models and built great temple complexes, some of whose stone pagodas remain standing. Clearly, Chinese cultural life - a provincial, rather old-fashioned version, perhaps, but still basically Chinese - endured and many of the Khitan nobility embraced Chinese culture side by side with their own. Many Han Chinese served in the government, a few of them reaching the highest offices. However, the question of ethnic identity was complex. Some of these successful Han families became increasingly like their rulers, intermarrying with Khitan noble clans and, if they served long at court, perforce adopting the Khitan life-style. The average Han Chinese official perhaps retained more of his cultural heritage but was excluded from real positions of power. Important decisions, especially military decisions, were a preserve for the Khitan officers at the court. The Liao government attempted to apply different rules to their Han subjects and their tribal peoples. But this did not mean that their Chinese population was misgoverned. Su Ch'e, who visited the Liao court as an ambassador in 1090, was rather surprised to find that the law was not unduly weighted against Han Chinese, though he was amazed at the degree of corruption prevalent. Nor do the Han Chinese seem to have been disaffected. They were the one major ethnic group that never once rose up against the Khitan, and at the very end, when the southern Han Chinese section of the Liao came under threat simultaneously from the Sung and the Jurchen, the population fiercely resisted the Sung and then gave up the Southern Capital to the Jurchen without even token resistance. It is more difficult to describe the situation under the Tanguts, for our sources are less full and the ethnic groups within the Hsia state were less neatly divided geographically than under the Khitan. But here again there seems to have been a strong continuation of Chinese cultural life, a great deal of publication and printing in both Tangut and in Chinese, and a lack of sharp ethnic conflict between the ruling elite and the Han Chinese population. With the Jurchen conquest the situation was altered. In both Liao and Hsia the dominant non-Han groups were not greatly outnumbered by their Han subjects. When the Jurchen conquered the Liao, they inherited the manageable northern Chinese population of the border provinces, but when they went on to conquer north China from the Sung, they found themselves Cambridge Histories Online © Cambridge University Press, 2008 40 INTRODUCTION in control of a vast and steadily increasing Chinese population of more than 40 million people, almost as many as the entire population of T'ang China in the eighth century. By 1207 their census listed 53 million people. The Jurchen were certainly outnumbered by their Chinese subjects by more than ten to one. Their response to this situation was an interesting and complex one. The Jurchen were, naturally, at some pains to preserve their racial identity. Intermarriage with Chinese was forbidden, and at first the Chinese were ordered to adopt the Jurchen costume and hairstyle. But at least after 1152 when Peking became the Central Capital and seat of government, the Jurchen elite, unlike the ruling group of the Khitan, no longer lived in their tribal homeland enjoying a peripatetic, seminomadic life-style. A Jurchen population remained in Manchuria, but the ruling house and a majority of the hundred or so great Jurchen ruling clans lived in former Liao or Sung territory surrounded by Han Chinese. The Jurchen, like the Mongols later on, differentiated between the Chinese populations conquered at different stages - between the "Northerners," formerly subjects of Liao, and the "Southerners" living in old Sung territory. Chin Shih-tsung preferred to employ former Sung officials. A new policy imposed by the Jurchen was the settlement of large detachments of Jurchen troops on extensive military colonies throughout their Chinese territories. These, together with other government lands, took up a considerable proportion of north China's farmland and were worked by Chinese dependants. But Chinese society as a whole was little disturbed: Merchant, artisan, landlord, and peasant continued their avocations. The literate elite were drawn into government service, and as the Jurchen established a Chinesestyle central government on T'ang lines, Han Chinese continued to provide the staff for most government bureaus. High culture was preserved. A huge amount of literature was written in all genres, canonical exegesis, prose and poetry, most of it reflecting conservative T'ang or eleventh-century Sung standards and unaffected by the contemporary innovations under way in Southern Sung territories, of which Chin scholars remained ignorant. While Confucianism continued along rather routine paths, independent of the Tao-hsiieh teaching that was gradually blanketing other forms of thought in the south, Buddhism and, even more, Taoismflourishedunder the Jurchen, and new literary genres aimed at a new semipopular audience - chantefables and primitive dramatic performances — also emerged. Printing continued to exhibit a high standard of excellence. Chin culture in which the highly cultivated Jurchen and Khitan participated along with Han Chinese may not have achieved the polished perfection of the Southern Sung, but it was a sturdy, independent tradition, equally firmly rooted in the Chinese past. Cambridge Histories Online © Cambridge University Press, 2008 THE HAN CHINESE UNDER ALIEN DOMINATION 41 The Mongols erupted into the Chinese scene far more violently than had the Jurchen. Their first major onslaught, on the Tangut state of Hsia, came closest to the image of the Mongolian terror that gripped the popular imagination throughout eastern Europe and Iran. The Tangut state and its sophisticated culture were almost wiped out. Next it was the turn of the Jurchen Chin, which was destroyed and its territories reduced to anarchy for almost a quarter-century. Until the accession of Khubilai, China was a sideshow for the Mongols, simply one part of their enormous empire, a rich source of booty and plunder, captives with special skills, and unrivaled artisans. It was during this period of the unbridled ravaging of China's resources that the Han Chinese population living under one of these alien regimes first suffered widespread disruption and destruction at every level of society. For the first time, too, the Chinese elite, with a few personal exceptions, was excluded from government service. The conquest of Southern China proceeded quite differently. Khubilai was determined to erect a state structure incorporating many of the features of a Chinese empire. But still the Mongols avoided the dependence on Chinese officials that had characterized the Khitan and Jurchen empires, for the Mongolian elite had members of other racial minorities, the gentry from Central and Western Asia, to act as their managers, tax farmers, and intermediaries. Some Chinese literati refused to serve their new masters and deliberately withdrew from public life to live as recluses. But nonetheless gradually some Chinese took office; many more served as clerks and minor functionaries; and the elite literati life, though no longer lived with the primary aim of service as an official, continued. Many educated men took up new professions as teachers, doctors, and business managers. Elite "Confucian" values were, as a result, spread more widely through society than had been the case before. One area into which they spread was among the ruling elite of Mongolians and their se-mu allies, many of whom themselves became accomplished writers in Chinese and full participants in Chinese culture. To a limited extent a minority of the non-Han elite was absorbed into the Chinese intellectual world. The end of the Yuan came in the mid-fourteenth century, not from invasion by yet another fresh wave of invaders, but by collapse from within. It is still far from clear what factors lay behind the many local uprisings that eventually toppled the dynasty: Natural disasters, epidemics, and a deterioration of climate all exacerbated the effects of misrule, exploitation, and administrative mismanagement. What is clear is that by the 1340s and 1350s there was massive disaffection at every level of the Han Chinese population, taking the traditional forms of banditry on a large scale, emergent sectarian move- Cambridge Histories Online © Cambridge University Press, 2008 42 INTRODUCTION ments, and mutinies among the armies. These phenomena were those traditionally associated with the fall of Chinese dynasties, but they were given a new edge by the fact that the government and many of its agents were foreigners. The twenty years of constant civil war that ended in the Ming founding must have been at least as destructive as the Jurchen conquest of north China, and more so than the Mongolian conquest of the south. Only the first phase of the Mongolian conquest of the north exceeded it in ferocity and devastation. But even before these civil conflicts broke out, Yiian China had been subjected to stresses far greater than those caused by the conquests of the earlier regimes. Did the dynasties of conquest really represent a major setback in the "natural" development of Chinese society, the Chinese economy, Chinese political institutions, and Chinese culture? Without them, would the pattern of rapid growth and rational organization that had characterized Sung China in the eleventh century have continued? Did they abort what some scholars regard as the emergence of a "modern age" under Sung? Or were these Sung developments in any case a dead end, doomed by limitations on the state's effectiveness, by China's very size and diversity, and by the failure of the Chinese elite to give due weight and attention to the practical and pragmatic? Why did the Ming, when they had finally driven the Mongols from Chinese soil, not resume the more sophisticated models of government provided by the Sung and, instead, continue so many of the institutional developments of the Chin and Yiian eras and revert to the T'ang models that all the conquerors had admired? These are complex questions, perhaps unanswerable. But they certainly suggest that the complicated and diverse periods covered in this volume deserve detailed scrutiny as a most important integral stage in China's development, a period that certainly did not close with the withdrawal of the last Mongolian armies across the frontier. Cambridge Histories Online © Cambridge University Press, 2008 CHAPTER 1 THE LIAO INTRODUCTION The founding of the Liao dynasty at the beginning of the tenth century opened a second period of extensive foreign dominance in China, a period that would last for almost half a millennium and that reached its climax with the conquest by the Mongols of all of China in 1279. Never before had China suffered such a degree of political and military domination by foreign peoples for such a protracted period. The Khitan Liao, the Tangut Hsi Hsia, the Jurchen Chin, and the Mongol Yuan in turn exercised an ever-increasing control over Chinese territory. The surviving Chinese regimes were forced to acknowledge these conquest dynasties on Chinese soil as equal if not superior powers, to establish permanent diplomatic relations with them on an equal footing, and to pay them annual subsidies or tribute. Such a state of affairs was totally at variance with the traditional Chinese worldview, which saw China as the center of the civilized world around which other peoples and nations had to orbit, and to which they were expected to display submission and deference. Nothing better illustrates the new power of the north Asian tribal peoples than the fact that throughout Eurasia the name of the Khitan, the founders of the Liao dynasty, in such forms as Kitaia, Cathaia, or Cathay, became a synonym for China. 1 In Russia and throughout the Slavonic world it still remains the standard designation for China. The Khitan in fact controlled only a small and peripheral part of China proper. But their rule lasted for more than two centuries, and because their area of domination stretched from the borders of Korea in the east to the Altai Mountains in the west, it effectively blocked China's direct communication with Central and Western Asia. As a consequence, in the west there was an In writing this chapter I have had constant recourse to the following monumental volume on the Liao: Karl August Wittfogel and Feng Chia-sheng, History of Chinese society: Liao (907-1125) (Philadelphia, •949). which is by far the most thorough and comprehensive treatment of the subject in any language. 1 See Paul Pelliot, Notes on Marco Polo (Paris, 1959), vol. i, pp. 2 1 6 - 2 9 . 43 Cambridge Histories Online © Cambridge University Press, 2008 44 THE LIAO unavoidable impression that the Khitan, now the intermediaries between East and West, were the real masters of China. The misleading and exaggerated image that the long period of Khitan rule imprinted on the minds of contemporaries elsewhere in Asia persisted long after the Khitan dynasty of Liao had disappeared. THE PREDYNASTIC KHITAN The early history of the Khitan is not well documented.2 Chinese historical sources first mention the name Khitan (Ch'i-tan) in the fourth century A.D. But there are many anomalies in these early accounts and considerable confusion about what group of people is referred to under this name. There does seem to be general agreement that the Ch'i-tan originated from the Yii-wen branch of the Hsien-pei, the group of peoples that dominated the northeastern frontier of China from the second century onward. In 345 the Yii-wen were crushed by the stronger Mu-jung group of Hsien-pei, who had founded the state of Yen, and they split into three tribes, one of which was called the K'u-mo-hsi, to which the Khitan belonged. In 388 this group again split to form the K'u-mo-hsi (later usually called simply the Hsi) and the Ch'i-tan. Wei Shou's history of the Northern Wei, the Wei shu, completed in 554, is the earliest of the standard dynastic histories to mention as independent nations the Ch'i-tan, the K'u-mo-hsi, and the Shih-wei, all of which it describes as descended from the Hsien-pei.' At that time all were nomadic pastoral peoples, the Ch'i-tan living in the grasslands of the upper valley of the Liao (Shira muren), in what is today the western part of Liaoning and Kirin provinces bordering on Inner Mongolia. The K'u-mo-hsi lived to the south and west of them in the mountainous borders adjacent to Hopei and northern Shansi, and the Shih-wei inhabited the territory north of the Ch'i-tan on the border between Inner Mongolia and western Heilungkiang. It is not entirely clear whether by Wei Shou's time Ch'i-tan was the name of a separate Khitan nation or whether these people were still a part of a larger tribal conglomeration called K'u-mo-hsi. To complicate the issue further, some of the smaller tribes that the Wei shu account of the Khitan identifies as parts of their nation also appear elsewhere in the same history as quite independent peoples. These seeming contradictions probably reflect both the imperfect sources available to Wei Shou when writing his history and also the fact that during the long Northern Wei period (386-535), the situation 2 The most detailed account of the pre-dynastic Khitan is by Otagi Matsuo, Kittan kodai shi no kenkyu, Toyoshi kenkyu sokan, no. 6 (Kyoto, 1959). 3 See Wei Shou et al., eds., Wei shu (Peking, 1974), 100, pp. 2 2 2 1 - 4 . Cambridge Histories Online © Cambridge University Press, 2008 THE PREDYNASTIC KHITAN 45 among these peoples was in constant flux: The Khitan first separated from the K'u-mo-hsi and then gradually developed into a Khitan nation incorporating other previously independent tribes. The Khitan were defeated by the T'o-pa founders of the Northern Wei and became their tributaries. In 479 a large section of the Khitan, threatened by the expansion of the Jou-jan in Mongolia, moved southeastward into the middle valley of the Liao. But they continued to grow steadily stronger into the sixth century. The precise ethnic affiliations of the Khitan and their neighbors are obscure. Not surprisingly, the Chinese sources are of little help in solving this problem. Traditional Chinese historians writing about foreign peoples were not interested in anthropology or descriptive ethnography but, rather, in the relationship of the foreign peoples with the Chinese court, even when, as in the case of the Northern Wei, this was itself the court of an alien dynasty. The foreign peoples were interesting either because their submission increased the prestige of the Chinese court or because they threatened China's national integrity. The historians' identification of the Khitan with the K'u-mo-hsi and eventually with the Yii-wen, the classification of the Khitan and their neighbors the Hsi and Shih-wei as being commonly descended from the Hsien-pei, and even the later attempts to link them with the Hsiung-nu who had ruled over northern Asia under the Han are thus little more than equations between unknowns, for we cannot define exactly the ethnic identity of any of these peoples. Their names often designated political groupings, rather than stable ethnic groups, with which the Chinese were in contact. In the nomadic society of northern Asia such political groupings were always highly fluid, with many small and intricately interrelated tribal groups that would join together in military confederations in times of crisis. These confederations were always very unstable and dependent above all on the personal charisma of their leaders and, after a short period of common purpose and unity under strong leadership, would inevitably split again. Contemporary scholars have attempted to solve the problem of Khitan origins on the basis of linguistic evidence, but studies of the Khitan language have not so far supplied any solid evidence. We know that syntactically the Khitan language resembled the Altaic languages (all the languages of the northern steppe were closely related), but this still leaves a wide range of choice among the Turkic, Mongolian, and Tungusic subfamilies of Altaic. To narrow the choice we have a meager vocabulary of some two hundred known words, half of them listed in the vocabulary appended to the Liaoshih, a work compiled as late as the fourteenth century. Unfortunately, the bulk of this vocabulary consists of names, titles, and appellations, items that tend to be readily passed from one language to another. Because the Khitan were long Cambridge Histories Online © Cambridge University Press, 2008 46 THE LIAO under the strong political influence of Turkic-speaking peoples, first as vassals of the T'o-pa, then of the T'u-chiieh, and later of the Uighurs (with whom the Liao founder's tribe had long-established marriage relations), it is quite natural that many of the tribal titles and names of Khitan offices have Turkic origins. What little we know of the basic vocabulary that tends to be least affected by word borrowings suggests links between Khitan and either Mongolian or Tungusic. Thus the Khitan may have spoken either a protoMongolian language influenced by Tungusic vocabulary or a Tungusic language influenced by Mongolian vocabulary, in either case using many Turkic loan words. Evidence from customs and material culture is equally difficult to interpret, as such things too are liable to be borrowed from other peoples. We cannot expect such evidence to offer any definite solution to the origins of the Khitan, particularly because their territories lay in the border area between the arid grasslands and deserts of the west, inhabited by Turkic or Mongolian tribes, and the forested plains and mountains of Manchuria to the east, the home of the Tungusic peoples. It is only natural that Khitan culture should have traits in common with each of its neighboring peoples. One cultural trait, however, supports the Wei shu's statement that the Khitan, Hsi, and Shih-wei shared a common origin. This is their male coiffure. As is proved by both textual evidence and Liao mural paintings, Khitan men shaved their heads in a sort of tonsure, leaving the hair on both temples hanging down to the shoulders or over the chest. This same hairstyle was used by both the Hsi and Shih-wei and also by the Hsien-pei, who were believed to be their common ancestors. The evidence from customs and material culture gives us little more than linguistic evidence concerning the Khitans' ultimate origins. It is clear enough, however, that their ancestors belonged among the heterogenous conglomeration of peoples called Hsien-pei. Beyond that we are left in the realm of conjecture. The Khitan tribes in the sixth century still formed only a weak confederation. In 553 they were disastrously defeated by the Northern Ch'i, who enslaved a great number of them and drove off much of their livestock.« At the beginning of the Sui period the Khitan were engaged in constant internecine warfare, and shortly after 586 some of their tribes became vassals of the T'u-chiieh Turks, and others submitted to the Sui. The Sui sbu (completed in 636) denounces them as the most primitive of all barbarians, a judgment that probably reflects their still unstable state of organization rather than the level of their culture. In times of peace, each tribe lived its own life, tending its 4 See Li Pai-yao et al., eds., Pei Ch'ishu (Peking, 1972), 4, p. 57. Cambridge Histories Online © Cambridge University Press, 2008 THE PREDYNASTIC KHITAN 47 flocks of sheep and horses and hunting in its own territories. Only in case of warfare did their chieftains meet and elect a temporary leader.' The Khitans' political fate was largely determined by their far more powerful neighbors and by the ever-changing balance of power between the successive regimes ruling northern China, on the one hand, and belligerent tribal neighbors to the north, northeast, and northwest, on the other. When China was strong, as under the T'o-pa Wei in the fifth century, the Khitan were drawn into its political orbit. When China was relatively weak, they became vassals of other nomadic peoples such as the T'u-chueh, the Turkic people who in the sixth century succeeded the Jou-jan as the lords of northern Asia. Some Khitan tribes in the east even paid allegiance to the state of Kogury6 centered in southeastern Manchuria and northern Korea. Nevertheless, when temporarily united, the Khitan could become a considerable military threat, as was proved in 605 when they staged a large-scale invasion of Sui territory in Hopei and northern Shansi. This provoked a major Sui punitive expedition, which inflicted grievous losses on the Khitan nation, leaving them temporarily much reduced in numbers.6 In the 620s and 630s the situation changed radically with the rise of the T'ang, not only as a powerful Chinese dynasty, but also as a power that claimed hegemony in the northern steppe following their defeat of the T'uchueh in 630. During these years the Khitan again gradually drifted into China's political orbit. One of their chieftains visited Ch'ang-an in 623, and another came to offer formal submission in 628 after calling a conference of tribal leaders. During T'ai-tsung's campaign against Koguryo in 645, some Khitan tribes fought on the side of the T'ang, and in 647 a confederation of eight tribes under the chieftain, K'u-k'o, of the Ta-ho clan submitted to T'ang suzerainty. ? It is not clear whether K'u-k'o was a temporary leader elected during a period of crisis or a new type of leader elected for life or whether he had become leader by hereditary succession. The leader who had led the tribute mission in 628 had been a member of the same Ta-ho clan. It may well be that by the 640s the Khitan confederation had become more stable and cohesive under the influence of their erstwhile overlords, the politically more sophisticated T'u-chiieh. The T'ang court decided to control them indirectly 5 See Wei Cheng et al., eds., Sui shu (Peking, 1973), 84, pp. 1881-2. 6 See Ssu-ma Kuang et al., comps., Tzu chih t'ung Men (Peking, 1956), 180, pp. 5 6 2 1 - 2 . 7 On the Khitan during the T'ang, see Liu Hsu et al., eds., Cbiu T'ang shu (Peking, 1975), 199B, pp. 5349-54 (hereafter cited at CTS); Ou-yang Hsiu and Sung Ch'i, comps., Hsin T'ang shu (Peking, 1975), 219, pp. 6167—73 (hereafter cited as HTS); Wang P'u, T'ang hut yao (Shanghai, 1935; reprinted Peking, 1955, 1957), 96, pp. 1 7 1 7 - 1 9 (hereafter cited as THY). See also Denis C. Twitchett and John K. Fairbank, eds., Sui and T'ang China, 589—906, pt. I, vol. 3 of The Cambridge history of China (Cambridge, 1979), pp. 314—16, 438—40. Cambridge Histories Online © Cambridge University Press, 2008 48 THE LIAO through a strong hereditary leadership loyal to China. To achieve this they bestowed on K'u-k'o the T'ang imperial surname Li and appointed him governor-general of Sung-mo, an office specially created for the indirect management of the Khitan tribes. By these measures the court hoped to ensure K'u-k'o's cooperation with the T'ang and the compliance of his fellow tribespeople. For a while the T'ang court's policy worked successfully, and for almost a century most of the Khitan leaders came from the Li (Ta-ho) clan. Following the pattern common to all the nomadic peoples of northern Asia, however, the succession did not always pass only to the leader's direct descendants but often went instead to mature collateral clan members such as uncles or brothers and was usually subject to "election" at periodical meetings of the tribal chieftains. In such a society, in which the ruler exercised direct personal authority, there was no way that a child, a weakling, or an incompetent could be tolerated as ruler. A similar pattern of succession continued into the Liao dynastic period. Until the 690s, successive members of the Li clan were appointed to office and noble ranks by the T'ang court, and several served with distinction as T'ang generals. Relations between the T'ang and Khitan, however, were by no means smooth. At the end of the seventh century the T'ang grip on the border regions was loosened. From the early years of Kao-tsung, when T'ang armies ranged across Asia from the Tarim basin to Korea, China had gradually retreated into a defensive posture. The Tibetans had become their main strategic concern. Meanwhile, from about 680 the T'u-chiieh again grew powerful and began to reestablish their hegemony in the steppe. The T'ang attempt to conquer Korea had ended in abject failure, and a new state of Chen (later to become Po-hai) was emerging in eastern Manchuria. Finally, the Khitan sought to rid themselves of T'ang tutelage. In 696—7 their leader, Li Chin-chung, provoked by his mistreatment at the hands of an arrogant Chinese local official, raised a rebellion and launched a massive invasion of Hopei, which drove deep into T'ang territory and captured several large cities, imposing crushing defeats on the T'ang armies. But the invasion failed, for two reasons. The T'u-chiieh khaghan had no interest in fostering a rival steppe power, and he attacked the Khitan in the rear, defeating them and driving off many of their family members and herds. Then when the T'ang court mobilized new armies against them in 697, the Hsi, who had originally joined the Khitan in their uprising, changed sides and submitted to the T'ang. The demoralized Khitan army collapsed and was routed, with massive casualties. The T'ang swiftly drove the Khitan back from Chinese territory but could not immediately regain control over them. It was not until 715, after the decline of T'u-chiieh power and the resurCambridge Histories Online © Cambridge University Press, 2008 THE PREDYNASTIC KHITAN 49 gence of T'ang expansionism under Hsiian-tsung, that the Khitan again submitted. In 716 the Khitan chieftain, still a member of the Li (Ta-ho) clan, appeared in person at the court in Ch'ang-an to present tribute. As before, the leaders were granted high offices and titular ranks from the T'ang court and, to strengthen further the bonds, a "princess" from the imperial clan was given in marriage to their leader, and members of the Khitan ruling clan were sent to Ch'ang-an as "hostage princes." An indirect administration incorporating the eight subordinate Khitan tribes was established. The next few years marked the height of T'ang influence over the Khitan. Their control, however, proved short-lived. Relations soon disintegrated, owing to internal dissension among the Khitan. Despite their investiture by the Chinese court, the leadership of the Li (Ta-ho) clan declined. A Khitan minister named K'o-t'u-yii emerged as their de facto leader, enthroning and deposing a succession of kings from the Li clan. Although he never challenged the Li clan's hereditary right to rule and never attempted to usurp their position himself, he held unchallenged authority. In the late 720s he attended the Chinese court, where he was treated rudely by the chief minister. As a result he developed a deep hatred of the T'ang and on his return in 730 killed the Khitan king, forced the Khitan and Hsi to become vassals of the T'u-chiieh, and rebelled. Although he and his puppet king were killed by the Li faction in 734, the T'ang court could not reimpose its control over the Khitan. Their attempts to impose on them a new leader chosen by themselves ended in failure. Moreover, these internal troubles led to a change of leadership. The Ta-ho clan, backed by the T'ang court, were replaced by a new ruling clan, the Yao-lien. We will return to this later. In 745 the T'ang tried unsuccessfully to restore relations by means of a new marriage alliance. But this failed; the royal bride was murdered and the Khitan again rebelled. An Lu-shan, the T'ang military governor of Fan-yang (modern Peking), then attempted to subjugate them by force. In 751 he invaded Khitan territory but was defeated with heavy losses. In 755 he again sent a larger and better-prepared force against them. This time his army was victorious and the Khitan were routed. But later in the same year An Lu-shan himself rebelled and invaded Hopei with a force including large contingents of Khitan, Hsi, and Shih-wei cavalry. A bloody and protracted civil war ensued in China, a war that threw the empire into disorder for a decade and brought the T'ang to the verge of destruction. They would never fully recover. Many Khitan, Hsi, and Shih-wei continued in T'ang service, especially in the military. Some of the Khitan and Hsi generals who had served under An Lu-shan and subsequently surrendered to the T'ang played a significant role in the semi-independent provinces of Hopei. The important province of Cambridge Histories Online © Cambridge University Press, 2008 50 THE LIAO Ch'eng-te, for example, which had a Chinese population of three and a half million before the rebellion, was ruled by a Hsi family from 762 to 781, then by three generations of Khitan governors from 782 to 820, and finally by a family of Uighur descent until the end of the dynasty. The almost equally large province of Wei-po was governed by a Hsi general from 822 to 826. We do not know what links these governors kept with their own tribal peoples beyond the T'ang borders, but they employed fellow Hsi and Khitan extensively in their armies and provincial administrations. After 755 it becomes extremely difficult to follow the history of the Khitan tribal people. They did not take advantage of China's temporary weakness to take the offensive against the T'ang and encroach on their territory, as the Tibetans were doing in the west. Perhaps internal strife and the heavy losses inflicted on them by An Lu-shan immediately before his rebellion had seriously destabilized the confederation. In any case, the border region immediately adjacent to Khitan territory was the stronghold of the powerful rebel regime. Instead of intervening, the Khitan remained peaceful and accepted the suzerainty of the Uighurs, who had succeeded the T'uchiieh as the lords of northern Asia. This did not mean breaking all links with the T'ang. Both the Khitan and the Hsi had long had a special relationship with the military provincial governors (chieh-tu shih) of Fan-yang (later renamed Lu-lung), based at modern Peking, who had traditionally held the concurrent title of commissioners for the control of the Hsi and Khitan (Ya Hsi Ch'i-tan shih). After the An Lushan rebellion, the governors of Lu-lung, now semi-independent, retained this title and its duties for themselves. Because of their considerable military power, there was little trouble on their border, and the Hsi and Khitan rarely raided the Chinese frontier. The Khitan, despite their having become vassals of the Uighurs, regularly sent tribute-bearing missions to China. When these embassies arrived at the provincial capital at Yu-chou (modern Peking), the governor would select a few dozen men to visit Ch'ang-an and keep the remainder at his headquarters. We know of at least thirty such embassies between 756 and 842. Successive emperors entertained and rewarded their visitors as custom demanded, but they no longer granted the chieftains Chinese titles or noble ranks, as the court was well aware that the Khitan were Uighur vassals. When after 840 the Uighur empire in the steppe collapsed, the situation changed again. In 842 the Khitan chieftain Ch'ii-shu renewed the allegiance to the T'ang, repudiated their former tributary relationship with the Uighurs, and requested formal investiture from Wu-tsung's court. The T'ang government renewed the Khitans' former official titles to symbolize their submission. Cambridge Histories Online © Cambridge University Press, 2008 THE PREDYNASTIC KHITAN TABLE 6 The Yao-lien khaghans I II Wa Personal name Ch'ii-lieh; puppet ruler set up by K'o-t'u-yii in 730. Killed with K'o-t'u-yii in 734. Ancestor of the tribe. Also named as Ch'u-t'e. Li Kuo-che A restored leader of the Ta-ho line and former colleague of K'o-t'u-yii. When K'o-t'u-yii was defeated by the T'ang, Li murdered him and his puppet ruler Wa and was appointed governor-general of Sung-mo by the T'ang court in 735. But in the same year he was deposed by Nieh-li (Ya-li), the li-chin (commander in chief) who abdicated power in favor of Tsu-wu of the Yaolien. Tsu-wu Originally a minister of Nieh-li, who enthroned him. Personal name Ti-lien, Chinese name Li Huai-hsiu. In 745 the T'ang court appointed him governorgeneral of Sung-mo and Prince Ch'ung-shun and married him to Princess Ching-le. Later that year he murdered the princess and rebelled. In 746 the court appointed Li Chieh-lo, who had been an ambassador to court in 722, to succeed him, but he remained at Ch'ang-an, as he could not take Tsu-wu's position. Ill- Hu-la IV Su V Su-chih VI Chao-ku VII Yeh-lan Also written Hsien-chih. Said by the Liao-shih to be identical with Ch'ii-shu, who submitted to the T'ang in 842. VIII Pa-la Also named Hsi-erh or Hsi-erh-chih. IX Hen-te-chin Also named Ch'in-te. Became khaghan in 901. -The Liao-shih notes that after the An Lu-shan rebellion the succession of khaghans is not entirely clear. Unfortunately, the T'ang historical sources break off their accounts of relations between the T'ang court and the Khitan at this crucial point and give no further information about the Khitan until the last decade of the ninth century, when the events leading up to the foundation of the Khitan state were already beginning. To fill in this crucial gap we must rely on the Liao shih, which provides a list of the Yao-lien khaghans, who ruled the Khitan throughout this period (see Table 6). The Liao shih's compilers admit that this list is not completely reliable. They also provide a more detailed genealogy of the I-la tribe, from which the Liao dynastic founder, (Yeh-lii) Apao-chi, came (see Figure i). This genealogy may have been composed as early as the 940s, and it was certainly composed with a view to establishing the legitimacy of the Yeh-lii royal house. Presumably with an eye to their many Chinese subjects, the Khitan royal house claimed descent from a Chinese culture-hero, the legendary emperor Shen-nung, protector of agriculture in the Chinese pantheon. This Chinese-style myth was, however, far less important to the Khitan Cambridge Histories Online © Cambridge University Press, 2008 52 THE LIAO themselves than was their own legendary account of the founding of their people. According to this myth, the first ancestor of the Khitan, Ch'i-shou khaghan, was once riding on a white stallion down the valley of the Lao-ho River, to its confluence with the Huang-shui River (Shira muren), where he encountered a woman riding in a cart drawn by a gray ox. The place can be identified with the ancient grazing grounds of the Khitan and the Hsi. The man and woman married and produced eight sons, who became the ancestors of the eight tribes that subsequently made up the Khitan nation. During the Liao dynastic period, statues of Ch'i-shou khaghan, his wife, and their eight sons were worshiped at the sacred Mu-yeh Mountain in this same area, the ceremonies being accompanied by the symbolic sacrifice of a white stallion and a gray bull. How ancient this tradition was is impossible to ascertain, but the number of eight tribes as the nucleus of the Khitan already appears in the earliest sources, and there seems to have been an almost-unbroken tradition of eight ancient tribes (with even some degree of continuity in their names) from the fifth century until the time of A-pao-chi. This legend of eight original tribes was closely connected with the pre-dynastic system of elected leadership. Every three years the chieftains of the eight tribes gathered together to elect (or confirm in his office) one of their number to serve as khaghan of the confederation. The khaghan's tenure of office was not for life; he might be removed from his office and replaced. In such a case he would bear no malice against his successor and would be allowed to return unmolested to head his own tribe. Such a form of political organization was not unique to the Khitan but was also found among other northern peoples. It was a device to ensure that leadership was in the hands of a capable man enjoying the confidence of his tribal peers. The "hereditary" leadership first of the Ta-ho and later of the Yao-lien clans was thus not a simple hereditary succession determined by primogeniture. Although the clan provided a pool of eligible candidates, succession did not necessarily pass to the senior clan member or even to a member of the senior generation. Rather, it was a succession subject to election and periodical confirmation by the leaders of the other tribal groups. A-pao-chi, the founder of the Liao dynasty, was not a member of the ruling Yao-lien clan, but of a tribe called the I-la (later given the clan name Yeh-lii). To give legitimacy to his claim to leadership, the official Liao genealogy traced the leadership from Ch'i-shou khaghan to a certain Ya-li who was the direct ancestor of the Yeh-lii clan. Ya-li is rather arbitrarily identified with Nieh-li or Ni-li, a partisan of K'o-t'u-yii, who is said to have murdered the Ta-ho leader, Li Kuo-che, whom the T'ang had tried to establish as leader after K'o-t'u-yii's death in 734. Nieh-li reorganized the fragmented Khitan tribes after the end of the Ta-ho confederation and led the Khitan for some Cambridge Histories Online © Cambridge University Press, 2008 THE BACKGROUND OF A-PAO-CHI S RISE TO POWER 53 time before abdicating in favor of a certain Tsu-wu from the Yao-lien clan, whose members ruled the Khitan for nine generations, adopting for the first time the title of khaghan. Meanwhile, from generation to generation, Niehli's descendants, the I-la, also seem to have held various high offices in the confederation. The point of these genealogies is that they demonstrate that the I-la tribe had an ancestral claim to the leadership that Ya-li/Nieh-li had long ago relinquished to the Yao-lien. In 906 or 907 when the last Yao-lien khaghan, Hen-te-chin or Ch'in-te, was deposed because of his ineffectual leadership, the leaders of the eight tribes elected in his place the commander in chief of the confederation's forces (the yii-yiieh), A-pao-chi, the chieftain of the I-la tribe. The old order had come to an end. THE BACKGROUND OF A-PAO-CHl'S RISE TO POWER It is impossible to understand the sudden emergence of the Khitan as a major power in northern Asia without first carefully looking at the international situation at the end of the ninth century. It is tempting to ascribe the Khitan's rise simply to the collapse of T'ang authority. But this was only one part of a complex series of changes that took place in the late ninth century. The destruction of the Uighur empire by the Kirgiz in the early 840s had freed the Khitan from their overlords and left a power vacuum in the northern steppe, which had been dominated by Turkic peoples since the sixth century. The Khitan no longer had a powerful and well-organized neighbor to the west and northwest. Although the Chinese historians on whom we depend for our information naturally concentrate on the success of the Khitan in establishing a strong dynastic regime, including the northeastern margin of traditionally Chinese territory, the real success of the early period of expansion was perhaps their wide-ranging conquests in the steppe and later in Manchuria. The Khitan's military efforts were first directed against their strongest tribal neighbors, the Hsi and Shih-wei. Not until they were subdued and the Khitan were firmly in control of the area beyond the line of the modern Great Wall did they turn to attack their more formidable adversaries, first the rich, powerful, and well-organized Chinese-style state of Po-hai (Parhae) in eastern Manchuria and then China itself. It is important to put aside the arrogant presuppositions of Chinese historians about the "normative" relationships between "barbarous" tribal pastoralists and a China that was the indisputable center, the cultural, political, and moral center alike of the human world, and to try to view the relationship through Khitan eyes. To the Khitan, China was not simply an allpowerful neighbor or a source of "superior" culture, but to all the steppe Cambridge Histories Online © Cambridge University Press, 2008 54 THE LIAO peoples it was above all immensely, unbelievably wealthy and the producer of an infinite variety of goods. Some of these were necessities (for none of the nomadic peoples were entirely self-sufficient); some were luxuries in demand among the Khitan elite; and some were articles that could be traded to the neighboring peoples or bartered into theflowof commerce that still stretched across Inner Asia to the west. Since An Lu-shan's rebellion the Khitan had seen their erstwhile overlords the Uighurs grow enormously wealthy by extracting huge subsidies from the T'ang court in return for their keeping the peace and trading their only major asset — a supply of horses for the T'ang cavalry. The great troupes of Khitan who flocked to Yu-chou under the name of "embassies" were far more anxious to have the opportunity to acquire Chinese goods for themselves, especially fine silks, than to have contact with China's superior civilization, whereas the governors of Yu-chou, who kept the majority of the members of these embassies at their own provincial capital, also undoubtedly provided them with the opportunity to trade, above all to sell the fine horses for which the Khitan were famous. The Khitan down to A-pao-chi's time had no territorial ambitions in T'ang China. They wanted Chinese wealth, Chinese manufactures, and Chinese captives, especially those with special skills. When these could not be acquired through trade, their raids across the Chinese border were swift forays aiming to carry off livestock, captives, and any portable wealth back to their own people. In the decades following the collapse of the Uighur empire, the China faced by the Khitan changed dramatically. In the 840s T'ang China remained an extremely powerful empire whose armies had repeatedly inflicted crushing defeats on the Khitan and whose central palace armies could come in overwhelming force to the aid of any border province that was critically threatened. The governors of Hopei, with whom the Khitan had the closest ties, may have enjoyed a great degree of autonomy, but they had never seriously tried to break free of the T'ang. From the mid-87os, however, the T'ang order speedily crumbled in the face of endemic disorder and massive rebellions. After Huang Ch'ao's disastrous rebellion was finally crushed in 884, the T'ang court was isolated in the area around the capital, and the empire was divided among somefiftyprovincial regimes, many of them no longer even maintaining the fiction of loyalty to the throne, and all of them highly militarized. The position was totally unstable and the emperor powerless, from the 890s a mere puppet maintained by one or other of the northern warlords. When the Khitan came into contact or conflict with the Chinese after 890, there was no longer any central power with which they could deal; instead, their dealings were with the governors of the provinces immediately across the border. Cambridge Histories Online © Cambridge University Press, 2008 THE BACKGROUND OF A-PAO-CHl'S RISE TO POWER 55 9 . MAP I . The Khitan and north China, A.D. 908 This does not imply, however, that the Chinese frontier defenses were suddenly rendered impotent. The T'ang had long depended on the virtually independent governors of Ho-pei to defend the northeastern border, and as China grew increasingly militarized, such independent provincial power centers came to support large armies. The only difference was that hardpressed frontier governors could no longer call on support from the central government in emergencies, but only on their allies, in other provinces. There was an ever-changing web of such volatile alliances among these regional regimes. At the beginning of the tenth century the border with the Khitan was controlled by two major provincial powers (see Map 1). In the west, the northern part of the modern province of Shansi formed the province of Hotung, with its capital at T'ai-yiian. This critical area had since 883 been governed by the formidable general Li K'o-yung, who had played a major part in the final suppression of Huang Ch'ao. He and many of the elite of his province - within which the T'ang government had, since the seventh century, settled many groups of tribesmen of varied origins — were Sha-t'o Turks, renowned for their valor and ferocity. In the 890s he became one of the main contenders in the bloody power struggles in northern China. Although Cambridge Histories Online © Cambridge University Press, 2008 56 THE LIAO Li K'o-yung was temporarily overshadowed by Chu Wen, in the 920s his descendants would set up a dynasty of their own, the Later T'ang. Farther to the east the Khitan faced the province of Lu-lung, the northernmost part of Hopei, which, like the other Hopei provinces, had enjoyed semi-independence for 150 years under hereditary governors who paid no revenue to the government in Ch'ang-an and refused to allow courtappointed officials in the territory they controlled. With its center at Yuchou, Lu-lung was fiercely independent, and the area had a long history of anti-T'ang feeling. From 895 until 907 it was ruled by a powerful and bellicose general named Liu Jen-kung, who owed his position to the support of Li K'o-yung. To protect its long border against both the attacks of rival warlords and tribal raids, Lu-lung, which was poor and more sparsely peopled than the other Hopei provinces and unable to support a large standing army, relied heavily on well-organized militia forces (t'uan-chieh), especially in its northern frontier districts. Most of the Lu-lung military were natives of the province, and many of their families had served in the army for several generations.8 The militiamen fought not for a distant court and a remote emperor but for their homes and families. They were noted for their bravery. They not only repulsed Khitan raids, but under Liu Jen-kung they themselves took the initiative, crossing the border to burn off the Khitans' pastures, take captives, and drive off their herds of horses, sheep, and cattle. It is thus possible that it was not so much T'ang weakness that allowed the Khitan to weld their tribes into a powerful nation at the end of the ninth century but, rather, that it was the removal of central constraints on the aggressive border governors in Ho-tung and Lu-lung, and the tough stance these provinces — Lu-lung in particular — took toward their Khitan neighbors, that provoked them to unite. The T'ang's central power may have vanished forever, but the Chinese border provinces were as powerful and as thoroughly militarized as before. It was in this situation that A-pao-chi emerged to turn the Khitan tribes into a powerful dynasty. THE RISE OF A-PAO-CHI The founder of the Khitan state was A-pao-chi (872—926), later canonized as Emperor T'ai-tsu of the Liao (reigned 907-26), although that dynastic name was not yet adopted in his own lifetime. Born in 872, he came from the I-la tribe. Later, some years after A-pao-chi's death, the tribe adopted the surname of Yeh-lii, which our sources often ascribe retrospectively but anachro8 See Matsui Shuichi, "Roryo hanchin ko," Shigaku zasshi, 68 (1939), pp. 1397—1432. Cambridge Histories Online © Cambridge University Press, 2008 THE RISE OF A-PAO-CHI 57 nistically to clan members of earlier generations. Originally they, like all other Khitan except the consort Hsiao clan in dynastic times, had no surnames. The I-la tribe had risen to prominence among the Khitan in the last years of the ninth century, and they ranked only after the Yao-lien, the clan of the khaghans. Within the I-la tribe, succession to the chieftainship (i-li-chin) did not descend directly from father to son but followed a pattern common among the nomadic peoples by which brothers rather than sons often succeeded to a title. Figure 1 shows the line of descent and the holders of the chieftainship, but their actual order of succession and dates are impossible to establish. The I-la had built up their power not simply by war and plunder but also by a system of alliances. They had established a complex pattern of intermarriage with another clan - later to adopt the surname Hsiao - who were of Uighur origin and well acquainted with the Chinese border regions. Under the chieftainship of A-pao-chi's father Sa-la-ti (posthumously canonized as Te-tsu) the Khitan had begun to practice more advanced agriculture, to foster the production of iron and salt, and to encourage weaving. Nor was the power of the I-la confined to their own tribe. A-pao-chi's paternal uncle, (Yeh-lii) Shih-lu, already held the most important office among the Khitan below that of khaghan; he was the yii-yiieh, something akin to prime minister and commander in chief of all the Khitan forces. A-pao-chi thus grew up among the commanders of a race of nomadic warriors who were already experimenting with a more varied style of life and assimilating many Chinese refugees and captives and tribal people into their society. A man of huge stature and high intelligence, he seems to have known Chinese, although he deliberately avoided using it in dealings with his fellow Khitan, fearing that exposure to Chinese ideas might cause them to lose their identity and martial spirit. A-pao-chi swiftly rose to prominence not simply because of his parentage but thanks to his military prowess, proved in constant skirmishes and campaigns against the Khitans' neighbors. By the end of the ninth century he was the ta-tna hsiieh-sha-li, the commander of the khaghan's personal guard. In 901 A-pao-chi was elected chieftain of the I-la tribe. In this role he undertook attacks on the Shih-wei to the north, on the Jurchen in the northeast, and on the powerful Hsi to the south, against whom he directed repeated forays. In 902 he led a massive raid on the Chinese border province of Ho-tung. As governor of the province, Li K'o-yung was preoccupied with preserving his province against Chu Wen, who was steadily consolidating his power in northern China and had been beaten back by Li K'o-yung from Hotung in 901 and again in 902. A-pao-chi returned from this raid with 95,000 captives and huge numbers of camels, sheep, horses, and oxen. In 903 he again raided northern Ho-tung, taking several cities and capturing Cambridge Histories Online © Cambridge University Press, 2008 58 THE LIAO much booty. Toward the end of 903, A-pao-chi turned his attention to Lulung Province, raiding the frontier districts north of present-day Peking. In the autumn of that year he was elected yii-yiieh, commander in chief. He was still only thirty-one. His campaigns continued without a break. In 904, 905, and 907 A-paochi mounted campaigns in the north against the Black Cart tribe of the Shih-wei, and in 906 there were two separate attacks on the Hsi. But his main attention was now focused on the Chinese border, where he was increasingly embroiled with the tough frontier forces of Liu Jen-kung, the governor of Lu-lung. In 9059 A-pao-chi negotiated a peace settlement with his far more formidable neighbor Li K'o-yung, the Sha-t'o ruler of Ho-tung and a major contender for power in northern China. At the head of a force of seventy thousand Khitan and tribal cavalry, A-pao-chi met with Li at Yiin-chou (present-day Ta-t'ung), where they swore brotherhood, symbolically exchanging their gowns and horses. Li K'o-yung was, of course, anxious to preserve his northern border and also to enter into a military alliance that would stand him in good stead in his ongoing struggle with Chu Wen; A-pao-chi sought Li K'o-yung's neutrality while he dealt with Lu-lung. The treaty shows that A-pao-chi was by now a major force to be reckoned with, his personal standing eclipsing that of the khaghan, his nominal ruler. In the eyes of the border Chinese at least, he was already the leader of the Khitan people. Meanwhile, hostilities continued on the border of Lu-lung, which was raided each year from 903 to 907. In one of these cross-border raids an adopted son of the governor, Liu Jen-kung, was captured. Liu, however, retaliated by crossing the border each autumn to burn off the grasslands, thereby denying the Khitan grazing for their herds. On one of these raids they even took captive a brother of A-pao-chi's wife. These counterraids caused great hardship, loss of livestock, and famine among the Khitan, so that the weak khaghan Hen-te-chin was reduced to bribing Liu Jen-kung to leave the Khitan pastures intact by giving him large numbers of horses. In 907 Hen-te-chin came up for reelection as khaghan. The tribal chieftains, shamed by his passive response to Liu Jen-kung, deposed him and elected Apao-chi as khaghan in his place.10 A-pao-chi appointed his cousin Tieh-li-t'e 9 T'o-t'o et al., eds., Liao shih, i , p . 2 (hereafter cited as LS) gives the date 905. However, Hsueh Chiicheng et al., eds., Chiu Wu-laishih (Peking, 1976), 137, p. 1828, gives the date as 907, as do Ssu-ma Kuang et al., comps., Tzu chih I'ung Men, 266, pp. 8676—9. Ssu-ma K'uang's critical note (k'ao-i) cites early sources, since lost, in support of both dates. 10 This follows the version of events in Ou-yang Hsiu, comp., Hun Wu-tai shih (Peking, 1974), 72, p. 886. Hsueh Chii-cheng, Chiu Wu-tai shih, 137, pp. 1 8 2 7 - 8 , says that A-pao-chi succeeded in the normal way by election. LS, 1, p. 2, however, gives a quite different account, according to which Hen-te-chin died ac the end of 906, whereupon the Khitan leaders who had received his deathbed injunction appointed A-pao-chi in his place. However, this must be a falsification. Hen-te-chin Cambridge Histories Online © Cambridge University Press, 2008 THE RISE OF A-PAO-CHI 59 to replace him as chieftain of the I-la," and his own younger brother La-ko became chief of the Yeh-lii clan. Whatever his military successes, A-pao-chi might never have won lasting control over the Khitan people had he not pursued a constructive strategy toward the peoples he conquered. His attacks were primarily aimed at acquiring additional manpower. Captured tribesmen were made followers and attached to his entourage. Some of them were simple herdsmen, but others, especially those from among the Hsi and Shih-wei, were skilled metal workers and artisans. Most of the Chinese captives were settled in his own territories, often in what are called "Chinese cities" (Han ch'eng). The first record of such a settlement dates from as early as 902, when A-pao-chi was still only the chieftain of the I-la tribe. I2 The city, Tung-lou (Lung-hua), was built near the Khitan founder's traditional residence and was settled with captives from northern Shansi. Later, several hundred families of Jurchen captives were also settled there. We know of some forty other cities of the same sort established later, some by A-pao-chi's younger brother An-tuan and some by other members of the nobility. Often we are told where the Chinese population came from; they seem to have been usually settled as a group taken during the same campaign. The Chinese cities were walled (to keep the populace in as much as for defense) and built on the rectangular Chinese model, with gates on four sides, lookout towers, streets, and markets with drum and bell towers. Some had Confucian, Buddhist, and Taoist temples; ancestral shrines; and postal relay stations. They were not penal colonies or places of exile; rather, they became lively centers of trade and manufacturing, and many of their Chinese populace were not prisoners but voluntary refugees from disorder and oppression in the Chinese border provinces. These settlers, voluntary and involuntary alike, were largely instrumental in laying the foundations of the Khitan state.'3 A-pao-chi's strength did not come only from captives, important as was their role. Many Khitan warriors from all the tribes joined his own personal following, and he gradually built up a personal power base that undermined the traditional tribal structure of the Khitan and the balance of power among remained a powerful chieftain for some time. Ssu-ma Kuang, Tzu chih t'ung chien, 266, p. 8678 (k'ao-i) cites a lost source recording that in 908 both he and A-pao-chi offered tribute to the Liang court. 11 LS, i, p. 3. 12 LS, 1, p. 2. 13 The most thorough study of these "Chinese cities" is by Yao Ts'ung-wu, "Shuo A-pao-chi shih tai te Han ch'eng," Kuo hsiieh chi k'an, 5 (1935), pp. 53—78; repr. in vol. 2 of his Tung-pei shih lun ts'ung (Taipei, 1959), pp. 193—216. A later study by Sechin Jagchid, "The Kitans and their cities," Central AsiaticJournal, 25 (1981), pp. 70—88, is also useful, building on the work of the Mongolian historian Kh. Perlee on the identification of the sites of many Liao cities. See also Ch'en Shu, Ch'i-tan she hui ching chi shih kao (Peking, 1963; repr. 1978), pp. 83—109. Cambridge Histories Online © Cambridge University Press, 2008 60 THE LIAO the tribes. Eventually he was able to eliminate not only the old leadership but the system of leadership itself. A-PAO-CHI BECOMES THE NEW KHAGHAN AND ASCENDS THE THRONE The basic annals of the Liao shih tell us that A-pao-chi "ascended the imperial throne" and founded a dynasty of his own in 907. Other sources date this event vaguely somewhere between 904 and 922.14 The eleventh-century Hsin Wu-tai shih gives a different account of the event, saying that A-pao-chi refused to resign from his position oiyii-yiieh or khaghan when his reelection was due and left the Khitan federation, together with his followers and Chinese subjects, to set up his own "tribe." Further confusion arises from the Liao shih itself, for elsewhere it records a second ceremony of accession in 916. The date 907 itself arouses suspicion, as it coincides with the formal end of the T'ang. In the context of their known concerns about the legitimacy of the Liao dynastic succession, it was thus a convenient and significant date for the Liao historians to choose for the beginning of their dynasty. The confusion may best be reconciled by assuming that A-pao-chi became the unchallenged leader of the Khitan in the traditional sense in 907 and that in 916, when he was due to have presented himself once again to the tribal leaders for reelection, he instead carried out a formal ceremony of accession to the throne, adopted a reign title, and took on some of the institutional trappings of a Chinese-style emperor, thus declaring his equality of status with the Chinese rulers of the Liang. A great deal happened between these two crucial dates. A-pao-chi continued his campaigns to pacify the tribal peoples. In 908 he attacked the Shihwei. In 910 and 911, uprisings among the Hsi had to be put down. In 912 he attacked the Chu-pu-ku (or Chih-pu-ku) on the distant northwestern frontier of his territories in present-day Mongolia. In 915 it was the turn of 14 The precise chronology of the Khitan before the 930s is all but impossible to establish. In the main I follow the Liao dynastic history {Liao shih), though it is often self-contradictory and at variance with other historical works dealing with the Five Dynasties and early Sung China. For this early period confusions abound: A-pao-chi's name, for example, occurs as A-pu-chi, A-pao-chin, or An-pa-chien. The date when his posthumous temple name was given appears variously as 926 or 947. The date when the Khitan state took the dynastic name Liao is given in different sources as 926, 937, 938, or 947. Later it resumed the name Ch'i-tan kuo. The date of the change is given as either 983 or 1013: The title Liao was resumed in 1066. But the Liao shih does not even record the substitution of the new dynastic title or the resumption of the dynastic name Liao. Some of these inconsistencies simply cannot be resolved. Many of course arise from the fact that until later in the tenth century, the Liao had not developed a systematic dynastic record, and when the record of the early reigns was finally compiled late in the eleventh century, its compilers confused the issue by projecting back into these early years developments that in fact date from the mid-tenth century or even later. To annotate fully every disputed name, event, or date would be impossible without overloading this account with footnotes. Cambridge Histories Online © Cambridge University Press, 2008 A-PAO-CHI BECOMES THE N E W KHAGHAN 6l the Wu-ku (tentatively identified with the Onggirad). Liao dominion was steadily spreading north and northwestward. Meanwhile, relations with the Chinese border areas remained strained. In Lu-lung, the governor, Liu Jen-kung, was deposed in 907 by his son Liu Shou-kuang, who continued his father's hostility toward the Khitan. In 909 a Khitan army led by a member of the Hsiao consort family drove deep into Hopei and defeated Liu Shou-kuang somewhere southwest of present-day Tientsin. But Liu's ambitions were set high; in 911 he proclaimed himself emperor of the independent state of Yen (once the name of An Lu-shan's rebel regime) and began to invade the neighboring provinces to enlarge his territory. But on the same day that he proclaimed himself emperor, the Khitan occupied P'ing-chou, west of Shan-hai kuan. In 912 A-pao-chi personally led an army to attack Liu Shou-kuang. Then in the next year Li Ts'un-hsii - who had been the Sha-t'o ruler of Ho-tung since the death in 908 of his father Li K'o-yung and who later became Emperor Chuang-tsung of the Later Tang (reigned 923—6) — was alarmed by Liu Shou-kuang's aggressive actions, decided to intervene, invaded Lu-lung, and took its capital, Yu-chou. Liu Shou-kuang was captured and the Yen regime was destroyed; Lu-lung was incorporated into the Sha-t'o domain, known to its contemporaries as Chin. Li Ts'un-hsii now ruled effectively over the entire border region fronting Khitan territory and was steadily consolidating a powerful regime that presented a major challenge to the Liang dynasty based in Ho-nan that had been set up by his father's old rival Chu Wen in 907. A-pao-chi had, of course, once sworn brotherhood with Li K'o-yung, but the latter had never forgiven him for subsequently trying to establish friendly relations with his hated enemy Chu Wen, emperor of the Liang. Li Ts'unhsii, in control of the powerful satrapy of Chin, which now encompassed northern Ho-pei as well as Ho-tung, was a far more powerful and threatening adversary for the Khitan than Liu Shou-kuang had been. Fortunately for Apao-chi, Li Ts'un-hsii was preoccupied with more important ambitions concerning China. For the time being, therefore, an uneasy truce prevailed on the Khitan frontier. Relations with his neighbors were also of secondary importance to A-paochi, for he faced a major problem in maintaining his supreme power among the Khitan. After his election as leader in 907, his plans to consolidate his absolute authority did not go unchallenged. The biggest threat came from his younger brothers and other members of the Yeh-lii clan, who had become the new Khitan aristocracy following the eclipse of the Yao-lien. In traditional Khitan society, succession to both the khaghanate and tribal chieftainships had commonly passed to brothers or cousins. Moreover, custom demanded that the ruler be reelected every three years, when another memCambridge Histories Online © Cambridge University Press, 2008 62 THE LIAO ber of the tribal council or another candidate from his own clan might be chosen to replace him. In 910, when his reelection was due, A-pao-chi failed to go through this procedure, and his brothers, feeling cheated of their own chances of succession, sought to prevent him from establishing a permanent dynasty based on succession from father to son, as this would have ended forever their own claims to leadership. Most resentful was the eldest of Apao-chi's younger brothers, La-ko. In 911 four of the younger brothers rebelled, and in 912 another plot to assassinate A-pao-chi engineered by the same four brothers was uncovered before it could be carried out. In 913, when A-pao-chi's second three-year term as khaghan came to an end and he again refused to put himself forward for reelection, a far more serious rebellion, led by his brothers, his uncle, and his cousin who was chieftain of the I-la, broke out and was suppressed with much bloodshed. All these rebellions failed, however, and their defeat accelerated the accumulation of power in A-pao-chi's hands. But he was not yet a complete despot. He remained sufficiently enmeshed in the Khitan tribal system that he could not simply eliminate all his rivals. His brothers' lives were spared, although his uncle and cousin and more than three hundred of their supporters were executed. To compensate the brothers and other collateral relatives and to prevent further unrest among the Yeh-lii clan, A-pao-chi combined their families into the so-called Three Patriarchal Households (san-fu fang), encompassing all the descendants of A-pao-chi's grandfather, which became one of the privileged lineage groups of the Liao empire (see Figure 1). But the dissatisfaction over a permanent hereditary leadership among the imperial family and the struggles over the succession did not cease there. In 917 La-ko again rebelled and fled to Yu-chou where he was received by Li Ts'un-hsii, ruler of the Chin, who appointed him a prefect. Later, when Li Ts'un-hsii became emperor of the Later T'ang in 923, he executed La-ko as a gesture of goodwill and friendship toward A-pao-chi. In 918 there was another short-lived uprising led by another of the brothers, Tieh-la. Disputes over the leadership and succession troubles frequently flared up among A-pao-chi's descendants. In 916, when he should yet again have presented himself to the tribal leaders for reelection, A-pao-chi took still more drastic steps to consolidate his authority on a permanent basis. First he went through a Chinese-style accession ceremony, claiming for himself the title of emperor of the Khitan and adopting a reign title,1' thus proclaiming his independence of the Liang 15 There is great confusion about T'ai-tsu's reign titles Shen-ts'e (916) and T'ien-tsan (922), which some scholars claim were invented later. The first reign title for which there is independent contemporary confirmation is that of T'ien-hsien, adopted by A-pao-chi in the last year of his life (926) and also used under his successor T'ai-tsung. See Arthur C. Moule, The rulers of China (London, 1937), pp. 91 fT. Cambridge Histories Online © Cambridge University Press, 2008 A-PAO-CHI BECOMES THE NEW KHAGHAN 63 (whose calendar the Khitan had previously used) and signifying that he was now the equal of the Chinese rulers. Perhaps most significant of all, he proclaimed his eldest son Pei (900-37; Khitan name, T'u-yii) as heir apparent. This formally ended the claims of his brothers and other clan members to the succession and also preempted the rights of the tribal elders to elect their leader in the traditional way. Pei himself was much influenced by Chinese culture and most unlikely to revert to the old Khitan custom. Yet another symbolic gesture toward the founding of a Chinese-style regime was the establishment of the first Confucian temple, which must have seemed very out of place among these bloodthirsty and violent warriors, even though a small minority of Khitan nobles were beginning to become versed in Chinese letters. In 918, in another step toward establishing a more permanent regime, Apao-chi ordered the building of a great capital city, the Imperial Capital (Huang-tu), later to be known as the Supreme Capital (Shang-ching). This was constructed at Lin-huang, north of the Shira muren (a place that later became the Mongol city of Boro Khoton), in the ancient central territory of the Khitan tribes. For its construction, mass levies of corvee labor were mobilized, during the busiest months of the agricultural year: A-pao-chi had not yet grasped the finer points of Chinese-style governance of an agrarian population. The work is said to have been finished in one hundred days, but in fact it continued for some time. Later in the same year he ordered the construction of Confucian, Buddhist, and Taoist temples at the capital. In the last year of A-pao-chi's life the capital was further extended, and a series of halls and ancestral shrines were built. Eventually the capital covered an area of 27 It: It was built on a standard Chinese plan with walls, gates, a street grid, palaces, ministry buildings, temples, courier stations, and so forth. It was in fact a dual city, for to the south was a separate Chinese city, with dense housing and markets. It also had a special quarter for the Uighur merchants, who played a major part in the trade of the north, and lodgings for envoys from foreign nations. We cannot accurately date the growth of the city, as parts of the walls were rebuilt in 931, and further construction went on well into the eleventh century. By then it was only one of five capital cities. The construction of a permanent capital symbolized A-pao-chi's rapid development of a centralized administration and institutions. Already by this time A-pao-chi seems to have begun the dualistic form of administration that continued until the end of the Liao, with a Northern Administration responsible for the tribal portion of his domain and a Southern Administration, organized more closely on the T'ang model, responsible for the sedentary portion and especially the Chinese population. As early as 910, A-pao-chi Cambridge Histories Online © Cambridge University Press, 2008 64 THE LIAO appointed his brother-in-law Hsiao Ti-lu to head his Northern Administration. This development culminated in the formal division of the empire in 947 into Northern and Southern divisions (Pei-yiian, Nan-yiian), but the process was clearly under way long before this. In the later years of A-paochi's reign, captured Chinese officials played a major role in developing the administrative system. Han Yen-hui (d. 959), a former provincial finance official from Lu-lung, devised a tax system and was largely responsible for designing the Chinese administration of the south.'6 Dating the development of changes in governmental organization in this early period is impossible. Probably much remained personal and informal. The existence of a fixed capital should not be taken to imply a permanent government structure with regular premises and a fixed court, as under a normal Chinese dynasty. Instead, government remained with the emperor's entourage, and the court was peripatetic, annually progressing on a circuit of seasonal hunting grounds (na-po) and, from time to time, accompanying the emperor on his frequent campaigns.17 The "court" was a great portable city of tents and pavilions, transported on a train of ox-drawn wagons. The entourage lived partly off the land surrounding their camp: The local inhabitants were sometimes granted tax exemptions in recompense. In the early days at least, the imperial palace at the capital was not the expected vast range of lavish buildings but, rather, the site where the emperor's tents were pitched when he was in residence. In 916 and 917 A-pao-chi once again attempted to intervene in China. At this time Li Ts'un-hsii and the Liang emperor Mo-ti (Chu Yu-chen) were locked in conflict, righting for control of central and southern Hopei. A-pao-chi seized the opportunity to invade Li Ts'un-hsii's territory in northern Ho-tung and Hopei. In 917 the Khitan besieged Yu-chou for two hundred days and were finally driven off only by the arrival of a powerful army from Ho-tung led by Li Ssu-yiian, later to become the second emperor, Ming-tsung, of the Later T'ang. In 921 and 922 the Khitan again invaded Hopei, this time at the request of a local governor nominally allied to Li Ts'un-hsii and the Sha-t'o leaders in Ho-tung. They easily overran the main frontier passes, gained control of some Chinese territory east of modern Shan-hai-kuan (then known as Yii-kuan), and penetrated as far south as Chen-chou. Li Ts'un-hsii personally mobilized an army to repel them on this occasion. 16 LS, 74, p. 1331—2. 17 On the na-po, see Yao Ts'ung-wu, "Shuo Ch'i-tan te na po wen hua," in vol. 2 of his Tung-pet sbib lun is'ung (Taipei, 1959), pp. 1—320. See also the classic study by Fu Le-huan, dating from 1942, included in revised form in his Liao shib Is'ung k'ao (Peking, 1984), pp. 36—172. Cambridge Histories Online © Cambridge University Press, 2008 A-PAO-CHI BECOMES THE NEW KHAGHAN 9 0 , , ' , , 00 '• km ' 3DOmiles SOUTHERN TUNG-TAN CAPITAL^ formerly PO-HA Khitan mdr 1111% jljll CH'IEN SHU • Ch'eng-tu J ^ isi %&$&$&&&*&• illll i|xj*Ch'u-qhou •••..••• /liiiiliit \NAIU ^—-~<\0 ^ " MAP 2 . The Khitan and north China, A.D. 924 Shortly after this the situation in China took a major turn: In 923 Li Ts'unhsii finally vanquished the Liang and set up his own Sha-t'o dynasty of Later T'ang, ostensibly as a restoration of the T'ang. He was now the undisputed master of northern China, as the Liang emperors had never been. By 925 he had established control over all northern China except for two minor border provinces south of the Ordos (Ling-chou and Hsia-chou), and in the autumn and winter of 925 he defeated the major state of Shu in Szechwan (see Map 2). He was now a formidable adversary for the Khitan. Although there were still some border scuffles, for the time being A-pao-chi remained neutral and launched no further campaigns into China on the scale of those of 917. Instead, he turned his attentions northward and eastward. In 919 the Wu-ku were finally subdued. Then in 924-5 A-pao-chi led a great expedition into the steppe, which conquered the tribes of northern Mongolia and reached as far as the old Uighur capital city Ordu Baliq, on the Orkhon River. While he himself led his army west into eastern Dzungaria, he sent another force southwest across the desert to establish his sovereignty over those of the Uighurs who had settled in the western Kansu corridor, between the Gobi and the Tibetan plateau. Meanwhile, another Khitan force, led by A-pao- Cambridge Histories Online © Cambridge University Press, 2008 66 THE LIAO chi's second son Te-kuang (Khitan name Te-chin, later to become the Liao emperor T'ai-tsung; reigned 927—47), crossed the Gobi southward and established Khitan control over the tribal peoples of the Yin-shan area and the northeastern corner of the Ordos, including remnants of the T'u-yii-hun and some minor Tangut groups. In 926, only a year after returning home from these extensive conquests, A-pao-chi set out on a still more ambitious expedition. The target this time was the powerful state of Po-hai (Parhae), which ruled over a large area of eastern Manchuria and the coastal region beyond and with which there had been border clashes in 924. Po-hai, unlike A-pao-chi's other adversaries, was not a tribal federation of nomadic pastoral peoples but a centralized state on the Chinese model that had long enjoyed stable relations not only with China but also with Korea and Japan. It was a rich country, with five capitals, fifteen superior prefectures, sixty-two prefectures, many cities, and, in the south at least, a largely sedentary agrarian population. Militarily, however, it proved no match for A-pao-chi's armies. It fell within two months, and its king and nobility were removed to the Khitan court. Instead of annexing its territory outright, A-pao-chi changed its name to the kingdom of Tung-tan and appointed as its king his eldest son, the Chinese-influenced heir apparent Pei. Tung-tan became a vassal kingdom, retaining intact for the time being its own administrative apparatus and even continuing to use its own reign titles. Why A-pao-chi acted so cautiously toward Po-hai is not entirely clear. He may well have thought that the still-immature Khitan system of government was not yet ready to cope with the very different and far more complex problems of administering a large territory mostly settled by a sedentary population and with many cities; he may simply have wished to avoid antagonizing its numerous and potentially hostile population; and he may have wished to carve out a permanent appanage for his own designated heir, who, as it turned out, was not favored by the Khitan nobles to succeed him as khaghan. Having swallowed Po-hai, A-pao-chi appears to have resumed his intention of expanding into northern China. In 926 there was a court coup in the Later T'ang capital at Lo-yang. Li Ts'un-hsii had been militarily successful, but his political organization was unstable. Early in 926 his armies in Honan and Hopei mutinied and killed him, replacing him with his adopted son Li Ssu-yiian (reigned as Ming-tsung, 926—33), a provincial commander from Hopei. The new Later T'ang emperor sent an envoy named Yao K'un to report his accession to A-Pao-chi, who was still in Po-hai. Yao later wrote a detailed account of his reception, which survives and from which we learn that A-pao-chi announced his intention of first occupying Yu-chou and HoCambridge Histories Online © Cambridge University Press, 2008 A-PAO-CHI BECOMES THE NEW KHAGHAN 67 pei and then making a settlement with the Later T'ang.18 When the envoy demurred, A-pao-chi toned down his territorial demands to simply Chenchou and Yu-chou — slightly more than the old province of Lu-lung. The envoy still refused. At this point A-pao-chi suddenly fell sick and died. In the subsequent confusion the invasion plan was forgotten, but had he lived he had clearly intended a major invasion of Hopei. At his death A-pao-chi was still only fifty-four. He had been leader of the Khitan for only two decades, but he had transformed them from a local if powerful tribal confederation into a well-organized regime controlling the nomadic peoples of Mongolia and Manchuria, as well as the former territories of Po-hai. His state had incorporated many Chinese from the border regions, established cities for their residence, encouraged a diversity of industries and settled farming, and accepted in principle the idea that the regime needed a dual form of organization, which would be able to administer the settled farming population of the south and also to govern by more traditional means the tribal peoples under their dominion. A-pao-chi had encouraged the importation of Chinese systems of belief and other aspects of culture. But at the same time he had tried to protect the Khitan culture, most importantly by providing his people with a writing system. On his accession the Khitan had been illiterate, and written Chinese was the only available medium for record keeping. In 920 the first Khitan script (the "large script," an adaptation of the Chinese script to the very different, highly inflected Khitan language) was presented, and by the end of A-pao-chi's reign this script was widely used. In 925, when Uighur envoys visited the court, the emperor's younger brother Tieh-la (whom A-pao-chi recognized as the most clever member of his family) was entrusted with their reception and, after learning their script (which was alphabetic), devised a second "small script" for Khitan. Thus by the end of A-pao-chi's reign it was possible to operate a dual system of government in which the northern tribal section conducted its business and kept documents in Khitan and the southern (Chinese) section used both Chinese and Khitan. This would help the Khitan preserve their authority and cultural identity, but it also made permanent the conflicting elements within the Khitan elite, some of whom remained intransigent in their adherence to tribal values and institutions, whereas others adopted to a greater or lesser degree the often different ideas and practices from China. The "dualistic" nature of the state created by A-pao-chi may have been a 18 For a detailed study of this fascinating document, which presents a vivid portrait of A-pao-chi, see Yao Ts'ung-wu, "A-pao-chi yii Hou T'ang shih ch'en Yao K'un hui chien t'an hua chi lu," Wen shih che bsiieh pao, 5 (1953), pp. 9 1 - 1 1 2 ; repr. with revisions in vol. 1 of his Tung-pei shlb lun u'ung (Taipei, 1959). PP- 217-47- Cambridge Histories Online © Cambridge University Press, 2008 68 THE LIAO strength, as the Khitan became more and more involved in the Chinese world, but it was also inherently divisive. THE SUCCESSION CRISIS AND THE REIGN OF T'AI-TSUNG According to the arrangement A-pao-chi established in 916, the succession should have passed automatically after his death to the designated heir apparent Pei (900—37) without discord.19 But this was not to be. The cultured and refined Pei — a skilled painter, some of whose works were later included in the Sung imperial collection; an accomplished writer in both Khitan and Chinese; a bibliophile with a large personal library and a taste for Chinese culture; and an expert also in music, medicine, and prognostication - did not appeal to the traditionally minded Khitan chieftains. A-pao-chi's personal authority had been sufficient to have him made heir apparent, in denial of all Khitan custom and precedent, but even A-pao-chi seems later to have realized that his younger son Te-kung was the better candidate, and once Apao-chi was dead it soon became clear that a simple transfer of the throne to Pei was not in the cards. The decisive factor in the succession was A-pao-chi's formidable widow, Empress Ch'un-ch'in (later entitled Empress Dowager Ying-t'ien). She had been a great power during A-pao-chi's lifetime, the first of a series of dominant empresses that gives a special character to the Khitan regime. She had played an open and active role. Early in the reign Ch'un-ch'in had devised a plan for A-pao-chi to murder some of the tribal chiefs who opposed him. Later she established her own military camp (jardo), commanded her own army of 200,000 horsemen with which she maintained order when A-pao-chi was away on campaign, and even herself organized campaigns against rival tribes. After A-pao-chi's death Ch'un-ch'in took control of all military and civil affairs. When the time for his interment came, she declined to be buried together with him according to custom, though more than three hundred persons were buried with him in his mausoleum. Instead, she cut off her right hand and had this placed in his coffin while she survived to act as regent, for, she claimed, her sons were still young and the country was without a ruler. She remained in firm control while the succession was settled and exercised great influence for many years to come. Ch'un-ch'in herself had disapproved of the choice of Pei, and she used all her influence to have him set aside in favor of his younger brother Te-kuang (902—47), whom, it seems, even A-pao-chi had eventually acknowledged 19 See Yao Ts'ung-wu, "Ch'i-tan chiin wei chi ch'eng wen t'i te fen hsi," Wen shih che hsiieh pao, 2(1931), pp. 81 —iujrepr. in his Tung-pa shih lun ti'ung (Taipei, 1959), pp. 248-82, for a general discussion of the succession problem under the Liao. Cambridge Histories Online © Cambridge University Press, 2008 THE SUCCESSION CRISIS AND THE REIGN OF T'AI-TSUNG 69 was the better candidate. Te-kuang was much more a potential leader in the traditional Khitan tribal mold, although like Pei he was educated and a capable calligrapher. He had distinguished himself on the battlefield during the invasion of Ho-pei in 921-2 and played a major role as a commander in the western campaign of 924—5. Pei seems to have sensed danger for himself and helped persuade the Khitan dignitaries at court that he should be passed over in favor of his younger brother. Months went by with the succession still in doubt. Eventually, toward the end of 927 Pei himself approached his mother and formally withdrew his claim. Te-kuang then succeeded to the throne. He is usually known by his posthumous temple name of T'ai-tsung. Pei was still, however, the ruler (Jen-huang wang) of Tung-tan, the former state of Po-hai. After the prolonged succession problem, T'ai-tsung regarded Pei as the greatest threat to his own authority and feared that he would use the wealthy and populous Tung-tan state as a power base from which to realize his frustrated claims to the Khitan throne. One at least of Pei's consorts was a member of the Po-hai royal family. In 929 T'ai-tsung ordered that the capital of Tung-tan and all its people be moved to Tungp'ing (present-day Liao-yang), which was designated the Southern Capital of the Khitan state. Tung-tan was not abolished, but it was now more closely integrated into the Khitan empire and gradually lost the special quasiautonomous status it had been granted in 926. Pei seems to have been placed under surveillance. In 930 he fled by sea to China. He was received with honor at the Later T'ang court of Ming-tsung in Lo-yang and lived there in exile until 937, when he was murdered by Shih Ching-t'ang,20 whom the Khitan emperor had supported to overthrow the Later T'ang regime and who remained a subservient puppet of the Khitan. During T'ai-tsung's reign, actions against the border regions continued. In 928 there was renewed trouble with the Wu-ku in the north. In 929 the emperor's younger brother Li-hu was sent to attack the border area north of Ta-t'ung, and in 933 he mounted an expedition against some Tangut tribes that had still not submitted. But the most important aspect of the reign was the gradual involvement of the Khitan in the political scene in north China. The regime of the Later T'ang proved to be unstable. Its emperor, Mingtsung, had undertaken important reforms of government, restored the power and influence of his bureaucrats, and built up palace armies that could outweigh the forces of his provincial generals. But his reign ended badly, with an attempted coup by one of the princes a few days before his death. His 20 Or according to another account by the last Later T'ang ruler, immediately before Shih Ching-t'ang took Loyang. See IS, 72, p. 1211. Cambridge Histories Online © Cambridge University Press, 2008 70 THE LIAO son Li Ts'ung-hou (temple name Min-ti) lasted only five months, before his adoptive brother Li Ts'ung-k'o usurped the throne and killed him. At this point the former Khitan heir apparent, Pei, who had been living under the protection of Ming-tsung, wrote to his brother T'ai-tsung suggesting that he invade the Later T'ang empire. This was in 934. In 936 Li Ts'ung-k'o ordered the powerful governor of Ho-tung, Shih Ching-t'ang, to be transferred to a post in Shantung where he would be under closer court control. Shih Ching-t'ang rebelled, and Li Ts'ung-k'o led an army to attack him in T'ai-yiian. Shih Ching-t'ang was another Sha-t'o Turk, a son-in-law of the late emperor Li Ssu-yiian, and his rebellion led to other provincial rebellions against the Later T'ang. Hard pressed by Li Ts'ung-k'o, he now appealed to the Khitan emperor for military assistance. T'ai-tsung personally led an army of fifty thousand cavalry across the border through the Yen-men pass and defeated the Later T'ang army near Shih Ching-t'ang's capital at T'ai-yiian. The T'ang regime speedily disintegrated. In the eleventh month of 936, the Khitan invested Shih Ching-t'ang as emperor of a new dynasty, the Chin. He was nothing more than a puppet of the Khitan. In 937, to curry favor with his new overlord, Shih Ching-t'ang murdered the unfortunate Pei and, later in that year, agreed with T'ai-tsung that he would treat him as his father, thus symbolically placing his dynasty on a footing of inferiority to the Khitan. The Chin monarch seems to have realized how completely he was entrapped by the Khitan and offered to pay a huge annual subsidy to the Khitan to compensate for the return of the vital prefectures of Yu-chou and Chi-chou that they had occupied. The Khitan refused, and after some difficult negotiations in the next year the Khitan received the cession of sixteen formerly Chinese prefectures, including a broad belt from Ta-t'ung to Yu-chou. This new territory gave the Khitan control of all the strategic passes that defended northern China, and a sizable foothold in Hopei (see Map 3). T'ai-tsung had achieved his father's territorial ambitions and, in the bargain, had become the nominal overlord of a Chinese emperor. For the first time a Chinese regime openly acknowledged the suzerainty of an alien dynasty. The arrangement between T'ai-tsung and his puppet lasted only a few years and collapsed after Shih Ching-t'ang died in 942. But the results were far-reaching. The Khitan would hold on to most of the Sixteen Prefectures until the end of their dynasty. Yu-chou became the new Southern Capital of the Khitan (the former Southern Capital, center of Tung-tan, now became the Eastern Capital and grew into a city even larger than the Supreme Capital). A strong Khitan administration was imposed on the former Chinese territory, and the Khitan state incorporated a very large Chinese population. Cambridge Histories Online © Cambridge University Press, 2008 THE SUCCESSION CRISIS AND THE REIGN OF T'AI-TSUNG 71 The Sixteen Prefectures ceded to Khitan, 937 *EASTERN CAPITAL SSST JK'ar-fengl (Lo-yangl- MAP 3. The Khitan and north China, A.D. 943 The northern and northeastern frontier areas of China were now part of the dominion of the border peoples and would remain beyond Chinese control until the late fourteenth century. The Khitan, for their part, were also now inextricably involved in the affairs of the Chinese world. It is doubtful whether the principal actors understood the momentous implications of these events. Shih Ching-t'ang was a Turk, not of Chinese origin, and this may explain why he was not sensitive to the issue of Chinese subservience to a non-Han regime. He was far more concerned with preserving peace with the Khitan so that he could concentrate on his own pressing internal problems than he was with such niceties of status. During his reign (936—42) he duly fulfilled his obligations as a vassal and made no objections when the Khitan entered into closer relations with the states of southern China, particularly with his neighbor and rival the Southern T'ang. He even allowed Khitan embassies to southern China to pass through his own territory by land, rather than travel by sea, as they had done before. The Khitan were well aware of events in China beyond the borders of the neighboring northern regimes. As early as 915 Ch'ien Liu, ruler of the coastal state of Wu-Yiieh in modern Chekiang, had sent an envoy by sea to the Khitan court. Wu-yiieh formally acknowledged the supremacy of the Cambridge Histories Online © Cambridge University Press, 2008 72 THE LIAO successive regimes in north China. Their motives for also opening relations with the Khitan were chiefly commercial: They wished to protect their trade interests in Po-hai and Korea. The Khitan, for their part, sought access to the seaborne trade with Southeast Asia and the Indian Ocean, to obtain exotic articles, spices, and luxury goods. The Southern T'ang also established relations with the Khitan, but in their case the motivation was political. They wished to form an alliance with the Khitan against the Later T'ang. The king of Southern T'ang and the Khitan emperor addressed one another as brothers, thus giving Southern T'ang superior status in Khitan eyes to their northern neighbors, the Later T'ang. At a single audience in 937 T'ai-tsung received envoys from the Later T'ang, from the semi-independent governor of T'ai-yiian, Liu Chih-yiian, and from the newly enthroned emperor of Southern T'ang. The Khitan were thus closely involved in the complex interstate politics of the various independent regimes in China. Relations with Southern T'ang were not purely formal. The southern court provided crucial intelligence to the Khitan about developments in the Chin in 940, 941, and 943. After the fall of Chin and the Khitan attempt to set up a regime at K'ai-feng in 947 had ended in failure and withdrawal, the Southern T'ang again suggested a military alliance against the short-lived Han regime (948-51) that succeeded the Chin. And as late as 957 they again provided the Liao with military intelligence about the Chou regime in the north, which was currently threatening the Southern T'ang. The relations between the Khitan and the southern states of Wu-Yiieh and Southern T'ang were at their height in the late 930s and 940s; for a while Wu-Yiieh even used the Khitan calendar. But T'ai-tsung's invasion showed the south what a potential threat the Khitan posed. After the accession of the Liao emperor Mu-tsung in 951, the politically inactive Khitan leader showed no interest in intervening in the everlasting power struggles among the Chinese states. Thereafter both diplomatic contacts and trade with the southern courts fell off dramatically. They deteriorated even further after 954, when Mu-tsung's uncle, who had been sent as an envoy to Southern T'ang, was assassinated. He thus refused to send any further envoys, although Southern T'ang envoys reached the Liao in 955 and again in 957, still seeking help against the Chou. After the death of Shih Ching-t'ang in 942, relations between the Khitan and the Chin rapidly deteriorated. Shih Ching-t'ang may have been a Khitan puppet, but he had quietly restored the authority of his dynasty over the fractious provinces, strengthened the structure of government, and built up a strong centralized army. His successor, Shih Ch'ung-kuei (temple name Ch'u-ti; reigned 942—6), came under the influence of a violently anti-Khitan court faction led by the commander of the imperial army Ching Yen-kuang, Cambridge Histories Online © Cambridge University Press, 2008 THE SUCCESSION CRISIS AND THE REIGN OF T'AI-TSUNG 73 and he openly repudiated the former supremacy of T'ai-tsung and his "Northern Dynasty." In 943 Shih Ch'ung-kuei abolished the privileges of Khitan merchants at the Chin capital, K'ai-feng, confiscated their property, and sent their representative, who also managed trade in Chin on behalf of the Khitan court, back home bearing an insulting letter to T'ai-tsung. T'ai-tsung decided to invade. At the end of 944, Khitan forces crossed the border of Hopei at several points, followed by T'ai-tsung's main army. The fighting dragged on for three years, and not all of it went the Khitans' way. In the late spring of 945 the invading army was seriously defeated, and T'aitsung himself had to flee the battlefield mounted ignominiously on a camel. But the Khitan persisted, wearing down the Chin armies. The province of Ho-pei, where most of the fighting took place, was devastated. The outcome was decided at the end of 946 when Tu Ch'ung-wei, the Chin commander in chief and uncle of the emperor, surrendered. T'ai-tsung was able to enter the capital, K'ai-feng, without meeting any resistance. At the beginning of 947, T'ai-tsung made a triumphal entry into K'aifeng riding in the imperial carriage, took up his residence in the palace of the Chin emperor, and held court in the ceremonial audience hall, demanding the attendance of the remaining Chin courtiers. The Chin emperor and his family were sent in exile to the Liao Supreme Capital in Manchuria. The Chin imperial armies, after the surrender of Tu Ch'ung-wei, were disarmed and disbanded, and their cavalry horses were confiscated. T'ai-tsung formally announced a general act of grace, adopted a new dynastic name for the Khitan state — now to be known as the Greater Liao — and adopted a new reign title and a new calendar (which had in fact been devised under the Chin in 939). The new reign title he chose was Ta-t'ung, "Great Unity," and this publicly proclaimed that he was determined to make himself emperor of all north China. The Liao court diarists recorded that more than a million households of the Chin population had been incorporated into their empire. But the Chinese population thought otherwise. The Khitan army had brought no adequate supplies with them and now looted the capital and plundered the countryside in search of food and forage. Oppressive levies were imposed on the citizens of K'ai-feng, and everywhere there was resentment and fear of the invaders' unbridled violence. The populace began to attack the Khitan; mutinies and uprisings broke out all over Ho-pei. The Khitan were totally unprepared to govern such a vast territory, inhabited by a hostile sedentary population that far outnumbered them. T'ai-tsung complained to his entourage: "I never knew that the Chinese could be so difficult to govern as this!" The Khitan now began to loot the capital thoroughly. It was decided to take back to Manchuria the entire body of Chin officials. This proved impossible, Cambridge Histories Online © Cambridge University Press, 2008 74 THE LIAO but in the third month of 947 they began shipping off to the Supreme Capital the personnel of the main ministries, the palace women, eunuchs, diviners, and artisans in their thousands; books, maps; astronomical charts, instruments, and astronomers; musical treatises and ceremonial musical instruments; the imperial carriages and ritual impedimenta; the weapons and armor from the arsenals; and even the copies of the Confucian classics engraved on stone slabs. While T'ai-tsung stripped bare the palace and government offices, his troops continued to pillage the city and surrounding countryside. The Khitan, already harried by popular resistance and guerrilla attacks, were now faced with a more severe threat. Liu Chih-yuan, the governor of the fiercely independent Sha-t'o stronghold of T'ai-yiian who had stood aside when the Khitan invaded Hopei, refused to acknowledge T'ai-tsung as emperor or to attend his "court" in K'ai-feng. In the second month of 947 he declared himself the emperor of a rival new dynasty, the Han. Discontented elements in the neighboring provinces rallied to his banner, posing an immediate threat to K'ai-feng and Lo-yang. T'ai-tsung was now in a precarious position, facing not only widespread guerrilla resistance, local uprisings, and mutinies throughout Ho-pei but also the threat of a full-scale military confrontation with the only major commander in the north whose forces had remained intact after T'ai-tsung had disbanded the Chin imperial army. T'ai-tsung wisely decided to withdraw to the north, to "avoid the heat of summer" as he claimed but, in reality, to avoid being trapped with his army in an indefensible position deep in hostile territory. He had enjoyed his occupation of the capital at K'ai-feng for only three months. In the fourth month the Liao armies and his vast baggage train began to withdraw, constantly harassed en route by Chinese attacks. The invasion had plainly been a major miscalculation. T'ai-tsung himself admitted that he had made grave mistakes in permitting the looting of the countryside, in imposing harsh penal levies on the cities, and in failing to deal firmly with the provincial governors, who were still a key factor in the power structure of northern China. Moreover, his campaign had never won general approval among the Khitan nobility. Never again would a Liao emperor seriously plan a campaign of conquest in China. Shortly before reaching Liao territory in northern Ho-pei, T'ai-tsung, who was still only forty-five, suddenly fell ill and died at Luan-ch'eng (south of present-day Shih-chia chuang, Hopei). The Liao, having suffered a major disaster in their invasion of China, now faced yet another succession crisis at home. Meanwhile, Liu Chih-yuan entered K'ai-feng in the sixth month and established the shortest-lived of the Five Dynasties, the Han (947-50). He left his provincial capital at T'ai-yiian in the hands of his cousin Liu Ch'ung. Cambridge Histories Online © Cambridge University Press, 2008 THE SUCCESSION OF SHIH-TSUNG 75 This provincial power base was strong enough to survive the collapse of the ephemeral Han dynasty. When that fell in 950, Liu Ch'ung declared himself emperor of the independent regional state of the Northern Han, a state that would survive until 979 and whose fortunes would remain tied to the Khitan until the end. THE SUCCESSION OF SHIH-TSUNG Immediately after T'ai-tsung's death, Yeh-lii Yuan (918—51; Khitan name Wu-yii) the eldest son of Pei, A-pao-chi's original heir apparent, declared himself emperor "before his father's coffin" at Chen-chou in Ho-pei. His uncle T'ai-tsung had loved him as if he were his own son; he had accompanied the emperor in his campaign against Chin and had taken part in the occupation of K'ai-feng, gaining a reputation among the Khitan nobles as a brave and capable commander. A dignified, generous, and magnanimous man and a skilled horseman and archer, Yeh-lii Yuan was generally respected. He reached the Liao Southern Capital (modern Peking) in the sixth month and marched north at the head of his army. Yeh-lii Yiian's assumption of the throne by hereditary right as the eldest son of A-pao-chi's eldest son roused the opposition of the still-formidable empress dowager Ying-t'ien. She supported the claim of her own favorite third son Li-hu, the younger brother of the deceased emperor. His claim was not groundless, for in 930 T'ai-tsung had appointed him heir apparent or, according to some sources, as the imperial great younger brother (huang t'ai ti). Once again there was a conflict between the Chinese style of hereditary succession and the Khitan tribal custom of succession by brothers. The empress sent Li-hu with an army to block Yiian's return to the capital. When Li-hu's army was defeated, the redoubtable old lady led her own army to confront the new emperor. On the Shira muren, just south of the Supreme Capital, the two armies faced each other for several days. The crisis was resolved by the mediation of a royal cousin named Yeh-lii Wu-chih (916-72). This time the empress dowager was unable to have her way. Wu-chih and the Khitan nobles rejected Li-hu on the grounds that he had made himself generally hated because of his cruelty. The empress dowager gave in, telling Li-hu that he had brought failure on himself.21 21 LS, 77, pp. 1255—6. An interesting source regarding these events is the account written by Hu Chiao, a Chinese official of the Chin who had been taken back on the retreat from K'ai-feng in Hsiao Han's entourage. After the latter's execution for his part in a plot in 949, Hu fled east and then eventually back to China in 953. His account is cited in Ou-yang Hsiu, Hsin Wu-tai shib, 73, pp. 904—8; Ma Tuan-lin, Wen hsien t'ung k'ao, in Sbih t'ung (Shanghai, 1936), 345, pp. 2704c—5c. For a translation, see F.douard Chavannes, "Voyageurs Chinois chez les Khitan et les Joutchen," pt. 1, Journal Aiiatiqut, 9th series, no. 9 (1897), pp. 390—411. Cambridge Histories Online © Cambridge University Press, 2008 THE LIAO -]6 The confrontation between the legitimate hereditary heir and the late emperor's brother was thus resolved in favor of the hereditary heir. But it was not his hereditary claim that had prevailed; rather, his rival had been rejected by the nobility because he was personally unacceptable as a ruler. Although the empress lost, the Khitan principle of "electing" a suitable candidate was what swayed the decision. Moreover, opposition to the new emperor, whose posthumous temple name is Shih-tsung (r. 947—51), remained powerful. Much of his short reign was spent dealing with dissidence among the royal family and nobility. Both the empress dowager and Li-hu were banished from court and sent to live in retirement at Tsu-chou, the Khitan ancestral cult center. (The empress outlived Shih-tsung and died in 953, aged seventy-four.) If the new emperor hoped that this would secure his position, he was soon disillusioned. The internal situation in the Liao remained unstable. In 948 a plot against the emperor's life was organized by T'ien-te, the second son of T'ai-tsung. The conspiracy failed, and T'ien-te was executed. Although the other conspirators were punished, their lives were spared. Among them was Hsiao Han, a nephew of the empress dowager who was married to the new emperor's sister, A-pu-li. The next year he was involved in another plot with some dissident nobles. Again, even though he was proved guilty, the emperor tried to hush up the matter and released him. Finally in 949 a letter was intercepted in which Hsiao Han was plotting another uprising, this time with An-tuan, one of A-pao-chi's surviving brothers. This time Shih-tsung had had enough; Hsiao Han was executed and the princess died in prison. THE DEVELOPMENT OF GOVERNMENT INSTITUTIONS Shih-tsung was not entirely preoccupied with this constant round of intrigue. During his brief reign there were some important institutional changes. They were not entirely innovations but, rather, the culmination of changes that had been taking place gradually for many years. It is difficult to follow the development of Liao government institutions. The Liao shib" provides a detailed, if often confusing, picture of the mature government system as it existed in the early eleventh century, but few clues to the stages by which various offices and bureaus came into being and almost nothing about how they intermeshed to form a working system of administration. Shih-tsung's reign was clearly a crucial time. Ever since the acquisition of the sixteen Chinese prefectures in 938, it had been necessary to establish more 22 LS, 45-8, pp. 685-831. Cambridge Histories Online © Cambridge University Press, 2008 THE DEVELOPMENT OF GOVERNMENT INSTITUTIONS 77 and more sophisticated government institutions to control the millions of new Chinese subjects. The temporary occupation of north China brought a huge number of former Chinese officials into the Liao system and with them came a tendency to adopt many of the techniques of Chinese administration. The most striking feature of Liao administration was the dual system of government that had been gradually emerging for many years. Since early in the tenth century, it had been customary to divide offices into "northern" and "southern." The imperial clan itself had a southern division, made up of A-paochi's close relatives of the Six Divisions, and a northern division comprising more distant relatives. A-pao-chi had appointed northern and southern prime ministers {pei-fu tsai-hsiang, nan-fu tsai-hsiang). The nature of this system was symbolized in a decree of T'ai-tsung's later years that ordered that the officials of the Northern Administration and the empress dowager — the archrepresentative of the old tribal ways - wear Khitan costume and that the officials of the Southern Administration and the emperor himself dress in Chinese style. 2 ' This northern and southern division of government was not a strictly geographical one; the "Northern Administration" was responsible for the Khitan and tribal peoples, wherever they lived, and the "Southern Administration" was responsible for the Chinese population, as had been the Chinese Office (Han-erh ssu) that A-pao-chi had set up in his early years. At the beginning of Shih-tsung's reign, immediately after his return to the Supreme Capital, he formally divided the empire into Northern and Southern Regions (Pei-mien, Nan-mien). These were true regional divisions of Liao territory. The Southern Region comprised the predominantly Chinese and Po-hai areas of the south and east. The Northern Region was the area largely settled by Khitan and dependent tribal peoples. Because the Northern Region also included stable settlements of Chinese, Po-hai, and even Uighurs, it was given a dual administrative system. It had therefore both a Khitan Northern Commission for Military Affairs (Ch'i-tan Pei shu-mi yiian) and a Khitan Southern Commission for Military Affairs (Ch'i-tan Nan shu-mi yiian). The southern commissioners were usually members of the Yeh-lii royal clan, the northern commissioners mostly members of the Hsiao consort clan. The administration of the Northern Region was mainly, though not exclusively, staffed by Khitan holding traditional Khitan titles. Its most powerful officers were the Khitan commissioners for military affairs, the prime ministers of the Northern and Southern administrations {Pei-fu tsai-hsiang, Nan-fu tsai-hsiang), the Northern and Southern Great Kings {Pei Ta-wang, Nan Tawang), both of whom were members of the royal clan, and the commander in chief {yu-yueh). These men controlled all military and tribal affairs, the 23 LS, 56, p. 908. Cambridge Histories Online © Cambridge University Press, 2008 78 THE LIAO selection of military commanders, the disposition of the tribal herds, and the allocation of pastures. Beneath them was a bewildering array of tribal officials, an office for the royal clan of the former Po-hai state, and a range of offices providing services to the imperial house: artisans, physicians, huntsmen, and commissioners responsible for the royal herds, stud farms, and stables. No one could possibly confuse the administration of the Northern Region with the orderly model of T'ang government. It was essentially a great tribal leader's personal retinue, and many of its offices were specifically reserved for members of one or another branch of the royal or consort clans and filled by hereditary selection (shih-hsiian). The government of the Southern Region was more of a deliberate creation than that of the Northern Region, which had evolved from traditional Khitan institutions. It came into being after 948, when Shih-tsung returned to the capital following the capture of K'ai-feng and the transportation to the Khitan capital of great numbers of Chinese officials. It was modeled closely on the government institutions of T'ang and the Five Dynasties. The Khitan had used many Chinese titles earlier than this, both before and after the assimilation of the sixteen border prefectures in 937. But it is unclear how far these titles had implied the existence of Chinese-style bureaus with any regular staff. In many cases they were clearly honorific titles, the Khitan emperors following the old-established practice of the T'ang court in conferring ranks and honorary offices with no real duties, as a reward for loyal services. In 947, however, the Khitan had at last created a Chinese-style dynasty, with all the outward trappings of a Chinese court. The government of the Southern Region was designed in imitation of a T'ang model. It was based, as was the government of the Northern Region, at the Supreme Capital, where it had its main offices. It had the traditional groups of elder statesmen, the Three Preceptors (san shih) and the Three Dukes (san kung) to act as imperial advisers, and a complex bureaucracy at the head of which were three ministries similar to the three central ministries (san sheng) of early T'ang. There was a Chinese Commission for Military Affairs (Han-jen Shu-mi yiian), which combined the functions of the Commission for Military Affairs (Shu-mi yiian) under the Five Dynasties with those of the T'ang Department of State Affairs (Shang-shu sheng) and controlled five rather than six executive boards; a Secretariat (at first called Cheng-shih sheng; after 1044, Chung-shu sheng) headed by a grand prime minister (ta ch'eng-hsiang) and two deputy prime ministers (ch'eng-hsiang) and including a staff of secretaries and councillors; and a Chancellery (Men-hsia sheng) responsible for drafting documents. Each of these ministries had, on paper at least, a complicated bureaucratic establishment similar to its T'ang model, but only the Secretar- Cambridge Histories Online © Cambridge University Press, 2008 THE DEVELOPMENT OF GOVERNMENT INSTITUTIONS 79 iat played any significant role in political decisions. There was also the Censorate (Yii-shih t'ai), the Han-lin Academy, the Office of Historiography, and various groups of court scholars, all organized along Tang lines. There was an imperial household department, various specialized courts (ssu) and directorates (chieri), a formal establishment for the heir apparent, and a military organization of royal guards {wet). The basic provincial organization also began to take shape along Chinese lines. Beside the Supreme Capital there was now an Eastern Capital at Liaoyang, controlling former territories of Po-hai, and a Southern Capital at modern Peking, controlling the former Chinese territories acquired in 937. A fourth Central Capital was to be added in 1007, built on the site of the old Hsi capital, at the time when the Hsi were finally assimilated into the Khitan state. Last, a Western Capital at Ta-t'ung was established in 1044. Each of these capitals was not so much an alternative seat of imperial government (as, e.g., Ch'ang-an and Lo-yang had been in early T'ang) as the regional center of a circuit, a local administrative network. Each of these circuits seems to have followed administrative procedures appropriate to its own population. The picture was further complicated during the tenth century by the fact that two of the larger groups of conquered peoples, the Hsi and Po-hai, retained a great degree of autonomy under their own leaders and paid tribute as vassals rather than taxes as subjects. Only in the early eleventh century were these populations fully incorporated into the Liao system of government. The viceroys (liu-hou) of these capitals wielded great power over their circuits, especially those of the Southern and Eastern capitals, who were among the most powerful men in the Liao system of government. They presided over a hierarchy of numerous prefectures and counties that provided the field administration for the settled regions of the empire and that in many areas coexisted with tribal organizations ruled on traditional lines. The system of government in the Southern Region was similar, at least in its outward forms, to that of the T'ang and the Five Dynasties. Many of its officials, especially at the lower and middle levels, were ethnic Chinese. Historians familiar with Chinese institutions of the ninth and tenth centuries may, however, be tempted to ascribe an unreal importance to the holders of titles which, in the Chinese system, implied great power and influence. There was one important distinction, however, between the officers of the Northern and Southern regions, apart from their different racial origins. Liao emperors were constantly on the move and resided in their Supreme Capital only for short periods each year as they traveled from one traditional seasonal hunting camp (na-po) to the next. Twice a year, in the fifth and tenth lunar months, the officers of both the Northern and Southern administrations were Cambridge Histories Online © Cambridge University Press, 2008 80 THE LIAO summoned to the emperor's camp for deliberations on state affairs. In winter the officers of the Southern Administration went south to the Central Capital to mediate the affairs of the Chinese subjects of the Southern Region. But for most of the year, as the emperor's great retinue progressed around the northern territories, making contact with the leaders of the tribal world, the emperor was still expected personally to make all the important decisions affecting the state and to mete out justice. On these peregrinations he was accompanied by most of the great officers of the Northern Administration, who lived with him on close personal terms, as much his companions (like the nokor of Mongol times) as his great officers of state. By contrast, only a handful of officials from the Southern Administration - a single prime minister and a small group of secretaries and drafting officials — formed a part of his regular entourage. Clearly the officials of the Northern Administration, by virtue of their constant access to the emperor, enjoyed far greater real power than did those of the Southern Administration. Thus the Southern Administration was essentially an executive organization for the southern areas and their settled population. The high-sounding titles of its officers should not conceal the fact that routine decision making and all military authority (southern officials were specifically excluded from decisions on military affairs at court) were concentrated in the emperor's Khitan entourage drawn from the Northern Administration. Moreover, we should not be too influenced by the official structure described in the Liao history. Many of the offices seem to have been filled only sporadically. Power in the Khitan world, in spite of the bureaucratization that began in Shih-tsung's time and continued in fits and starts into the eleventh century, had little connection with a formal and orderly government structure. It remained to the end far more dependent on an individual's personal qualities and his achievements, on his family connections, his personal relationship with the emperor and powerful ministers, his friendships, and his military following. Powerful personalities and brute force still far overshadowed institutional niceties in the Khitan world. RELATION WITH REGIMES IN CHINA Under Shih-tsung, the Liao, in spite of their withdrawal from K'ai-feng, remained embroiled in the turbulent politics of northern China. In 948 the Southern T'ang renewed their attempt to form an alliance with the Khitan against their northern neighbors, this time the new northern regime of the Han. They were rebuffed. In the winter of 949-50 Shih-tsung launched a large-scale raid into Hopei, attacking several cities well inside the Han border and taking many captives and much booty. The Southern T'ang court Cambridge Histories Online © Cambridge University Press, 2008 THE REIGN OF MU-TSUNG 8l sent envoys to congratulate the Liao on their victory, perhaps still hoping for the elusive alliance. In the winter of 950 Shih-tsung himself led another foray into Hopei. The situation in China now underwent a major change. The rickety Han regime at K'ai-feng collapsed at the beginning of 951, when its second emperor was murdered and replaced by the chief general of his imperial army, Kuo Wei (904—54), who was enthroned as emperor of the Chou. At the same time Liu Ch'ung in T'ai-yuan broke away and established himself as ruler of an independent state of Northern Han in Ho-tung. Once again the Khitan faced two separate powers on their frontier with China. The Chou got off to a bad start in their relations with the Liao. Their envoys who came to inform the Liao of the change of dynasty brought with them a letter, whose wording offended Shih-tsung, who promptly imprisoned them. Later in the year the Chou attacked Liu Ch'ung, who sent an envoy asking for Liao assistance and bearing a letter in which he humbly called himself Shih-tsung's "nephew," thus accepting Liao superiority. Shihtsung sent envoys to invest Liu Ch'ung as emperor to cement their lord and vassal relationship. The importunate Southern T'ang also renewed their request for an alliance against the Chou. In the late autumn of 951 Shih-tsung took personal command of a southern expedition against the Chou. But before the army set out, he fell victim to yet another conspiracy, this time hatched by sons of A-pao-chi's younger brothers, seeking once again to assert the claims to the succession of junior lines of the royal family. The emperor, like many of the Khitan nobles, was much addicted to drink, and when he and his entourage were helplessly drunk after having sacrificed to his deceased father in preparation for the expedition, Ch'a-ko, a son of A-pao-chi's younger brother An-tuan, murdered him. The conspirators, however, had neglected to gain the support of the courtiers and thus were summarily executed. Shih-tsung was only thirty-three, and because he had no grown son, the succession passed to the eldest son of T'ai-tsung, Ching (931-69; Khitan name Shu-lii), who is known by his posthumous temple name Mu-tsung. The southern campaign was abandoned. THE REIGN OF MU-TSUNG, 95 I - 9 6 9 The new emperor was not a distinguished monarch. Like his predecessor, Mu-tsung was a heavy drinker who would sleep off his excesses for much of the day, and his attention to public affairs was at best spasmodic. The Chinese spoke of him as the "Sleeping Prince." Problems with dissident members of the royal family continued. In 952 Cambridge Histories Online © Cambridge University Press, 2008 82 THE LIAO Lou-kuo, a younger brother of Shih-tsung, hatched a conspiracy, and one of his uncles and a prominent Chinese official plotted to defect to the Chou. The plot was crushed and Lou-kuo was executed. In 953 another conspiracy, led by a son of Li-hu named Wan, came to light. Several of the plotters were executed, though Wan himself was pardoned. In 959 Ti-lieh, one of Loukuo's co-conspirators, again plotted rebellion; and in 960 Wan's elder brother Hsi-yin, the eldest son of Li-hu, was arrested for plotting a rebellion. This time Li-hu himself was implicated and died in prison. For the rest of his reign Mu-tsung's relatives were quiet. Mu-tsung was not only inattentive to public business and self-indulgent, spending inordinate time even by Khitan standards in the hunting field. He was also violent, cruel, and capricious toward the members of his entourage, especially when drunk. Indeed, toward the end of his reign, he ordered one of his great ministers not to execute the sentences he passed when he was drunk but to let him review them when he had sobered up. The annals of his reign in the Liao shih are a sorry catalogue of casual cruelties. Events elsewhere in China made this an unfortunate time for the Liao to be under the rule of such an incompetent monarch, which virtually paralyzed the dynasty. The new Chou regime, first under Kuo Wei (r. 951—4) and then under the competent Ch'ai Jung (Shih-tsung, r. 954-9), was an altogether better-organized and more powerful state than the earlier of the Five Dynasties. They finally broke the power of the provincial governors and firmly reestablished strong central authority. At the beginning of Mu-tsung's reign in 952 Liu Ch'ung, emperor of Northern Han, asked for Liao assistance against the Chou. A force was sent under Kao Mu-han that helped repel the Chou invaders. In 954 the Chou again attacked Han, and a Khitan force was once more sent to their aid. The Liao clearly valued their alliance with the Northern Han, for in that same year they returned some Han troops who had been taken captive by mistake and also assisted the Han in putting down local uprisings against the Han in districts bordering the Liao. On more than one occasion, envoys from the Han came to Liao to discuss strategic matters. At the end of 958 the Han sent several envoys to report renewed invasions by the Chou. Then in early summer of 959 the Chou attacked the Liao in force. Their armies took the vital I-chin, Wa-ch'iao, and Yii-k'ou border barriers in the fourth month and then in the fifth month recaptured Yingchou and Mo-chou, the southernmost of the Sixteen Prefectures (see Map 4). The Liao armies retreated in the face of the onslaught. Mu-tsung roused himself to come south to the Southern Capital to take command, and the defenses were strengthened to await the Chou army. The confrontation did not occur, however. The Chou emperor fell sick and had to return to K'aiCambridge Histories Online © Cambridge University Press, 2008 THE REIGN OF 0 0 500km ' L 1 A 0 300 m i les ^ffl KHITAN TYUN / ^pHsia- F U ^ j x Yun-chou 8 • '»L * -. 83 MU-TSUNG EASTERN ^CAPITAL i jg- SOUTHERN A CAPITAL j£?: B CHOU § WESTERNS SCAPITAL % |Lo-yangl| •j i SOUTHERN T'ANG MAP 4. The Khitan and north China, A.D. 959 feng, where he died in the sixth month. The Chou invaders withdrew, and Mu-tsung returned to the Supreme Capital. Whether because of Mu-tsung's lack of spirit or whether the Khitan were anxious not to repeat the mistakes of 947, the Liao seem to have pursued a purely passive defensive strategy throughout his reign. This was not only the case on the Chinese frontier with the Chou and their successors the Sung. The same is true on the northern frontiers, where no activity is reported apart from minor troubles with the Wu-ku and Shih-wei tribes in 965. In 960 the Chou were replaced as masters of northern China by the Sung. By degrees the new dynasty restored a level of stability that China had not known since the mid ninth century. For many years after their accession to power, the Sung were concerned with consolidating their regime and restoring central authority over the various independent kingdoms that had divided China since the collapse of the T'ang. This reunification had already begun under the Chou, who had defeated the Shu kingdom in Szechwan, recovering all of its territory north of the Chin-ling Mountains, and in 957 had inflicted a crushing defeat on the Southern T'ang and recovered all the territory north of the Yangtze River. But much remained to be done. In the north the attention of the Sung court was focused not so much on the Cambridge Histories Online © Cambridge University Press, 2008 84 THE LIAO Liao as on the small but stubbornly independent Northern Han state in Shansi. The rulers of Northern Han, as we have seen, had already established good relations with Shih-tsung in the early 950s, and the Liao continued to support them against the Sung. For the Liao their state was an invaluable buffer zone and a strategic stronghold from which any attempt by the Sung to strike into the occupied prefectures of northern Ho-pei could easily be outflanked. When in 963 the Northern Han were attacked by the Sung, they immediately appealed for aid from the Khitan. In 964 a Liao army was dispatched, which helped repel the Sung invaders. The Liao also reacted against attempts by the Sung to consolidate the border gains that had been achieved by the Chou armies in 959. In 963 and again in 967 there were minor skirmishes on the border to prevent the Sung from fortifying the I-chin pass that had been overrun in 959. But there were no major hostilities. In 969 Mu-tsung was murdered. He had spent the whole of the first month of the year in a furious drinking bout during which he again abused members of his entourage. In the second month he attended to business long enough to invest as his vassal the new ruler of Northern Han, Liu Chi-yiian. But he then again began to act violently and irrationally, butchering some of his bodyguards. Finally, driven to extremes, six of his personal attendants murdered him during the night. The Liao were well rid of a bloodthirsty and totally unpredictable tyrant. The succession this time passed off without incident. All of A-pao-chi's brothers were now dead, and the energies of their descendants seem to have been exhausted in the round of conspiracies earlier in the reign. No objection was raised when the throne passed back to the senior royal line. Shih-tsung's eldest son was already dead, and the succession went to his second son Hsien (948-82; Khitan name Ming-i) who reigned from 969 to 982 and is known by his temple name Ching-tsung. THE REIGN OF CHING-TSUNG, 9 6 9 - 9 8 2 : CONFRONTATION WITH SUNG By the time the new emperor Ching-tsung came to the Liao throne, the situation in China had been completely transformed. The Chou dynasty, which had made rapid strides toward reestablishing political stability in China, had been crippled by the unexpected death of its emperor, Shih-tsung (Ch'ai Jung), in 959 and the succession of a six-year-old boy. The boy was toppled in a military coup led by a general named Chao K'uang-yin (known by his temple name T'ai-tsu; r. 960—76), who in 960 set up a new dynasty of his own, the Sung. Sung T'ai-tsu finally broke the local power of the provincial commanders, who had been the real power holders in China since the late Cambridge Histories Online © Cambridge University Press, 2008 THE REIGN OF CHING-TSUNG 85 ninth century, and gave his new dynasty a strong central government under firm civilian control. One by one T'ai-tsu eliminated and brought under Sung control the independent states that had divided up China south of the Yangtze; in 963 Ch'u in the central Yangtze basin, in 965 Later Shu in Szechwan, in 971 Southern Han in Kwangtung and Kwangsi, and in 975 Southern T'ang in Kiangsu, Anhwei, and Kiangsi. When in 976 his younger brother K'uang-i (temple name T'ai-tsung, r. 976-97) succeeded him to the throne of Sung, there remained only two independent regimes still to be incorporated into the empire: Wu-Yiieh in Chekiang and Northern Han in Shansi. Wu-Yiieh surrendered to the Sung in 978. Only the Northern Han remained. The Northern Han, the last remnant of the Sha-t'o Turkish domination of Shansi, had been closely involved with the Liao since its foundation in 951, when the first Northern Han emperor had been invested with his title by Shih-tsung of the Liao. Even the indolent Mu-tsung had understood the importance of the Han to the Liao's defensive strategy and had bestirred himself to help the Han beat off a Sung attack in the early 960s. An independent Han was greatly to the advantage of the Liao. It reduced the Liao—Sung frontier to a comparatively short stretch in the Ho-pei plain and gave Liao an ally that could threaten any attempt by the Sung to strike north at Liao across the Ho-pei plain by outflanking them from an almost impregnable base in the highlands of northern Shansi. The Northern Han was, however, a small state and quite unable, in spite of the professionalism and bravery of its troops, to withstand a full-scale war with Sung, except by relying on its alliance with the powerful Liao empire. The Han carefully cultivated this alliance. In 971, soon after Chingtsung's accession, they began to send regular monthly courtesy missions to the Liao court to ensure support. The Sung, nevertheless, were determined to conquer the Northern Han and in 974 began negotiations with the Liao to prepare a peace treaty, so as to ensure the Liao's neutrality when they attacked the Han. Beginning in 975 the Sung and Liao began to exchange regular diplomatic missions. In 977 the Sung even established five official border marshals to supervise trade with the north. Sung T'ai-tsung may have hoped to stabilize the border and also to cause a rift between the Liao and their Northern Han vassal, but if so his efforts were a failure. In the last year of T'ai-tsu's reign, 976, the Sung had invaded Northern Han. The Han asked the Liao court for assistance, and an army was sent that enabled the Han to repel the invasion. The next year saw a renewed Sung attack on the Han that brought another request for assistance. The Khitan again sent troops and cavalry to help the Han defense. Cambridge Histories Online © Cambridge University Press, 2008 86 THE LIAO In 979, after the surrender of Wu-Yiieh, Sung T'ai-tsung mounted a largescale invasion of the Han, now the last remaining independent state. The Liao sent an envoy to the Sung court demanding an explanation and were told bluntly to stay out of the conflict or they too would be attacked. In the early spring of 979 the Liao sent armies to the assistance of the Han, but the Sung armies intercepted them. The Liao suffered crushing defeats and heavy casualties. In the sixth month the Sung armies took T'ai-yiian, and the Northern Han emperor surrendered to them. The last of the independent states had been crushed and annexed. Sung T'ai-tsung, in the fullflushof victory, now, however, made a most imprudent decision. Going against the advice of all his commanders and without giving his already exhausted and overextended forces any chance to recover their strength and consolidate their position, he turned east through the passes in the T'ai-hang range and invaded the Khitan territory in northern Ho-pei, intent on recovering the Sixteen Prefectures taken by the Khitan in 937. Advancing to besiege the Liao Southern Capital (at modern Peking) Sung T'ai-tsung won some preliminary battles with the Khitan forces, but then in the seventh month the main Sung and Liao armies met in a crucial pitched battle on the Kao-liang River southwest of the capital.** It was a total disaster for the Sung, who suffered enormous casualties. The Liao took many prisoners and captured huge quantities of weapons and armor, baggage, equipment, money, and provisions. The unfortunate Sung emperor, who had been wounded, lost contact with his army, fled the battlefield alone, and made his way south to safety riding in a donkey cart. Some of his generals, thinking he was dead, wondered whether they should enthrone the son of the Sung founder in his place. What had begun as a victorious invasion of Northern Han ended in an ignominious rout. For the time being the Liao had the initiative. In 980 Ching-tsung took command in person of an offensive against the Sung in Ho-pei that recaptured the Wa-ch'iao barrier and defeated a Sung army. In 982 he launched another campaign, but this time the Liao armies were defeated, and Chingtsung was forced to withdraw. The result of these events was a complete change in the relationship between the Liao and the Sung, which no longer revolved around the buffer state of Han. The two great empires now faced each other along a continuous frontier stretching from the sea to the upper elbow of the Yellow River. And the Liao continued in possession of the Sixteen Prefectures, which continued to provoke revanchist feelings at the Sung court. It was only a matter of time before warfare broke out again. 24 On this battle, see Fu, Liao shih ts'ung k'ao, pp. 29—35. Cambridge Histories Online © Cambridge University Press, 2008 THE REGENCY OF EMPRESS DOWAGER CH'ENG-T'lEN 87 These troubles with the Sung were not the only military problems of Ching-tsung's reign. There were border problems with the Tanguts in 973, and with the Jurchen in the northeast, who invaded and looted Liao territory in 973 and again in 976. Both peoples were to cause trouble for the Liao for many years to come. In 981 there was an attempted coup aiming to enthrone a son of Hsi-yin, the son of Li-hu who had been imprisoned under Mu-tsung but later pardoned when Ching-tsung came to the throne. A group of captured Chinese soldiers managed to enthrone Hsi-yin's son, but the plot failed. Hsi-yin was forced to commit suicide, and his son was executed. In the autumn of 982 Ching-tsung, though still a young man, suddenly fell sick during a hunting trip and died in his camp. In his dying testament he left the throne to his eldest son Lung-hsii (r. 982—1031; temple name Sheng-tsung). The new emperor was only eleven years old, and so his mother, Ching-tsung's empress Jui-chih (later entitled Empress Dowager Ch'eng-t'ien), was appointed regent. THE REGENCY OF EMPRESS DOWAGER CH'ENG-T'lEN Empress Jui-chih was another of the succession of remarkable women who played a notable role in Liao public life.25 One of the reasons for this was the most unusual marriage structure of the Liao royal family, who took their wives from a single consort clan, the Hsiao, who also provided consorts for royal princesses and enjoyed the hereditary right to various influential offices.26 The royal brides, therefore, always came from families deeply involved in government and politics. Jui-chih was no exception. She was the daughter of Hsiao Ssu-wen (d. 970), appointed the northern commissioner for military affairs (pet-yuan shu-mi shih) and northern prime minister (pei-fu tsai-hsiangY1 at the beginning of Ching-tsung's reign, and she was made empress only two months after his appointment. The empress had already been influential in politics during Ching-tsung's life. Now she was left in control of the Liao empire. Although empress dowager, she was not a middleaged lady, as the title suggests. She had just turned thirty years old. The real power during the first half of Sheng-tsung's long reign, until her death in 1009, was in the hands of the empress dowager and three remark25 For his biography, see LS, 7 1 , p. 1201-2. 26 On this sytem, see Karl A. Wittfogel and Feng Chia-sheng, History of Chinese society, Liao (907— 112)), Transactions of the American Philosophical Society, n.s., vol. 36 (Philadelphia, 1949), pp. 1 9 1 - 2 , 206—12 (hereafter cited as Wittfogel and Feng); and Jennifer Holmgren, "Marriage, kinship and succession under the Ch'i-tan rulers of the Liao dynasty (907-1125)," Young Poo, 72 (1986), pp. 4 4 - 9 1 . 27 LS, 8, pp. 90. For his biography, see LS, 78, pp. 1 2 6 7 - 8 . Cambridge Histories Online © Cambridge University Press, 2008 88 THE LIAO able ministers, two of whom were Chinese. All three had been in power since the Sung invasion of 979, and the empress was accustomed to working with them. The senior figure was Shih Fang (920-94), *8 a native of Chi-chou in northern Hopei. He had been a scholarly prodigy, who was given the degree of chin-shib about 938, as the first "graduate" recorded under the Khitan. His "degree" in fact was almost certainly a personal honor, as the examination system was not permanently established for another half-century. When T'aitsung occupied K'ai-feng in 947, he was given charge of ritual and drafting edicts and subsequently held a succession of posts at the Southern Capital broken by more than a decade as a Han-lin scholar under Mu-tsung. He was highly regarded by Ching-tsung and was steadily promoted until in 979 he became northern prime minister. On Sheng-tsung's accession in 983 he attempted to retire but was refused and was given the additional post of head of the secretariat (chung-shu ling). Shih Fang became an important figure, setting the tone for a series of reforms covering the recruitment of officials and the easing of the tax burden on the people and winning wide respect. In 990 he again requested leave to retire and was permitted to reside permanently at the Southern Capital. In 993 he selected Han Te-jang as his own successor and was made honorary viceroy of the supreme capital. He died shortly afterward. Han Te-jan (941 —1011)29 was also Chinese and also from a Chi-chou family, but his background was very different from that of Shih Fang. His grandfather, Han Chih-ku,3° had been captured by the Khitan as a child and became a member of the household of A-pao-chi's empress. He soon gained A-pao-chi's confidence. The Khitan leader put him in charge of an office to control his Chinese subjects {Han-erh ssu) and gave him responsibility for court ceremonial. He and another surrendered Chinese, K'ang Mo-chi,3' who advised A-pao-chi on the establishment of Chinese cities, were each given high-sounding titles, tso p'u-yeh and tso shang-shu, respectively, and remained influential throughout A-pao-chi's reign. After K'ang's death in 926, Han Chih-ku, now president of the Secretariat (Chung-shu ling), held high office until he died early in T'ai-tsung's reign. He was the founder of the most powerful Chinese family in the Khitan state. 28 For his biography, see LS, 79, pp. 1271—2. 29 For his biography, see LS, 82, pp. 1289—91, under his later name Yeh-lu Lung-yiin. He appears in the histories under a series of names. In 1001 the emperor gave him the new personal name Te-ch'ang. In 1004 he was given the imperial surname Yeh-lii and in 1010, just before his death, the new personal name Lung-yiin. He had no son, and the descendants of his brothers, who remained influential until the fall of the Liao, continued to bear the surname Han. On his family, see Li Hsi-hou, "Shih lun Liao tai Yiit'ien Han shih chia tsu te li shih chi wei," Sung Liao Chin shih lun U'ung, 1 (1983), pp. 2 5 1 - 6 6 . 30 For his biography, see LS, 74, p. 1233. 31 LS, 74, p. 1230. Cambridge Histories Online © Cambridge University Press, 2008 THE REGENCY OF EMPRESS DOWAGER CH'ENG-T'lEN 89 His son Han K'uang-ssu (d. 98i) 3 S was a favorite of A-pao-chi's widow, the empress dowager Ying-t'ien, and became the director {hsiang-wen, a Khitan tribal title) of A-pao-chi's ancestral temple. He was closely involved with the royal family, survived participation in Hsi-yin's conspiracy under Mu-tsung in 960, and became an intimate of Ching-tsung in the 960s while he was still heir apparent. On his accession to the throne Ching-tsung appointed him viceroy, first of the Supreme Capital and then of the Southern Capital, and also commissioner for military affairs {shu-mi shih). During the Sung invasion of 979 Han K'uang-ssu was defeated and abandoned his troops. Ching-tsung wished to execute him, but the empress and her family interceded to save him. In 981 Han K'uang-ssu was appointed "commissioner for chastisement" of the southwest and died shortly afterward. He not only enjoyed strong personal influence with Ching-tsung; he also was an immensely powerful nobleman, with his own private fortress city, which later became a regular prefecture in 991. He also left five sons, who laid the foundation of a century of political power for the Han family.33 Han K'uang-ssu's two eldest sons, Han Te-yiian (d. ca. 980) and Han Tejang (941-1011), both had served in Ching-tsung's princely household before his succession. Han Te-yiian held various offices between 960 and 979 and made for himself something of a reputation for corruption before his death around 980. 34 Han Te-jang35 was chosen by Ching-tsung to succeed his father Han K'uang-ssu as viceroy of the Supreme Capital and later of the Southern Capital. He distinguished himself in the defense of the Southern Capital against the Sung invaders in 979 and was appointed commissioner for military affairs {shu-mi shih) of the Southern Administration. When Ching-tsung died, he and Yeh-lii Hsieh-chen received his will and were responsible for enthroning the young Sheng-tsung. The empress dowager favored and respected him greatly, and Han Te-jang steadily became the most powerful individual in the Liao empire. Sung sources, probably maliciously, suggest he was the empress dowager's lover. Eventually in 1004 he was given the royal Yeh-lii surname. His three younger brothers also filled high positions. The most important of them was Han Te-wei, a general who succeeded his father as punitive commissioner for the southwest and from 983 until the end of the century was mainly responsible for dealing with the Tanguts.' 6 The other most powerful persons in the first years of Sheng-tsung were 32 LS, 74, p. 1234. 33 See Lo Chi-tsu, Liao Han ch'en shih hit piao; repr. as no. 35 in vol. 4 of Liao shih hui pint, ed. Yang Chia-lo (Taipei, 1973), item 35, pp. 2—4. 34 LS, 74, p. 1235. 35 For his biography, see LS, 82, pp. 1289—91. 36 On Han Te-wei's family and their semi-Khitan status, see Wittfogel and Feng, p. 220 and n. 420. Cambridge Histories Online © Cambridge University Press, 2008 90 THE LIAO both Khitan and members of the imperial clan. Yeh-lu Hsieh-chen37 was the grandson of the commander in chief (yii-yiieh) Yeh-lu Ho-lu and had been recommended to Ching-tsung in 969 by the empress dowager's father, the commissioner for military affairs (shu-mi shih), Hsiao Ssu-wen. Ching-tsung was greatly impressed by him and married him to a niece of the empress. He distinguished himself in the war with Sung in 979 and won the respect of the empress dowager. Shortly after Sheng-tsung's accession, she organized a most unusual ceremony to ensure his loyalty. The child emperor and Yeh-lii Hsiehchen swore a pact of friendship in her presence, exchanging bows, arrows, saddles, and horses.38 The empress dowager subsequently gave Hsieh-chen many important responsibilities, making him northern commissioner for military affairs (sbu-mi shih). He remained influential until his death during the campaign against Sung in 1004. Another Khitan who helped stabilize the leadership was Yeh-lii Hsiu-ko, the commander in chief, who held this vital post from 984 until his death at the end of 998 and played a role in all the campaigns of this period.39 A measure of Han Te-jang's steadily emerging dominance can be seen in the fact that when Yeh-lu Hsiu-ko died in 998, Han succeeded to his post as yii-yiieh, and when Hsieh-chen died a year later he also took his post as northern commissioner for military affairs, holding both of these posts in addition to his original office as southern commissioner for military affairs. From 999 to i o n Han held more complete civil and military control over the Liao government, both of its Chinese and Khitan components, than any minister had before or after him.4° But while Ch'eng-t'ien was alive, there was no question who was ultimately in control; these great ministers were the empress dowager's men, and the new emperor was thoroughly dominated by his mother, who continued to browbeat and sometimes strike him in public even when he was a grown man. Immediately after the new emperor's succession she took an extraordinary step to ensure her power as regent. Before his formal installation on the throne, a Liao ruler normally went through the important Khitan religious ritual of "rebirth" (tsai-sheng), in the course of which he was symbolically reborn."" This confirmed the new emperor's right to rule in the eyes of the Khitan tribal aristoc37 38 39 40 For his biography, see LS, 8 3 , p. 1302. LS, 10, p. i n . For his biography, see LS, 83, p. 1299. See WanSsu-t'ung, Liao la ch'en menpiao; repr. as no. 33 in vol. 4 of Liaoshib hui-pien, ed. YangChia-lo (Taipei, 1973), item 33, pp. 8-9. Han held all three posts from 999 to the seventh month of 1002, when another Chinese, Hsing Pao-p'u, became the southern shu-mi ihih.XJpon Hsing's death early in 1004, however, this post reverted to Han Te-jang. 41 LS, 53, pp. 9 7 9 - 8 0 ; translated by Wittfogel and Feng, pp. 2 7 3 - 4 . According toLS, 116, p. 1537, it was supposed to be repeated every twelve years. See Shimada Masao, Ryocbo shi no kenkyu (Tokyo, 1979), pp. 339—47; Wang Min-hsin, "Ch'i-tan te 'ch'ai ts'e i' yu' ts'ai sheng i," Ku kung t'u sbu cbi k'an, 3, no. 3 (1972), pp. 31-52. Cambridge Histories Online © Cambridge University Press, 2008 C H A N G E S IN I N T E R N A L G O V E R N M E N T <pi racy. In this instance, the empress dowager herself underwent this ritual, not once but no fewer than three times, twice in 984 and again in 986. Later dowager regents also participated in the same ritual of installment in office. The empress dowager was no capricious tryant but a ruler who deeply understood the realities of power and the art of governance and who was always willing to listen to advice. She won the deep loyalty of Liao officials, Khitan and Chinese alike. She was not only an accomplished civil administrator; in the pattern of A-pao-chi's empress (the Ying-t'ien empress dowager) she was also a military commander, heading her own ordo, which could field ten thousand cavalry.42 Even when Ch'eng-t'ien was over sixty in 1005 she commanded armies in the field against the Sung. The Liao shih sums up her achievement aptly: "Sheng-tsung may be considered the most successful of the Liao emperors; most of his success must be attributed to his mother's instruction."« The reign of Sheng-tsung was the crucial period in the development of the Liao. Between 983 and 1031 a series of bloody and protracted wars stretched Khitan military capacity to its limits and ended not in total victory but in agreed settlements and an international equilibrium that ended a century of endemic warfare and instability in northern Asia. The settlement with the Sung in 1005 and the stalemate with Kory6 in 1019-20 marked the end of largescale warfare with Liao's major neighbors. Internally, too, the reign saw the final peaceful assimilation of the Hsi, the suppression of a serious rebellion by the peoples of Po-hai, and the pacification of the northwestern tribes. Only the problem of Liao's relations with the emergent state of Hsi Hsia and the continuous trouble with the Jurchen tribes remained to upset the international equilibrium during the next reigns. These events produced a balance of power in northern Asia that lasted until the second decade of the twelfth century. They took place at the same time as did the fundamental changes that were transforming the Khitan state from within. It is with these changes, which imposed a new degree of Chinese influence on the administration, that we shallfirstdeal. CHANGES IN INTERNAL GOVERNMENT The examination system There had been some haphazard examinations of candidatesforofficial service as early as T'ai-tsung's reign, but nothing regular or systematic was done 42 LS, 31, p. 367; 35, p. 404. 43 LS, 71, p. 1202. Her elder sister Hu-lien, who was married to Yen-sa-ko, Sheng-tsung's great-uncle, was also a successful military commander. After Yen-sa-ko's death, she undertook, at the head of his ordo, a pacification campaign against the Tsu-pu tribe in the far northwest, where she founded the fortress city ofK'o-tun in the Orkhon region. See LS, 13, pp. 145,149. In 1006 she was imprisoned for being part of a conspiracy, the details of which are obscure. She died in 1007 in confinement. See LS, 14, pp. 162—3; Yeh Lung-li, comp., Ch'i-lan too chih (Shanghai, 1985), 13, p. 142 (hereafter cited as CTKC). Cambridge Histories Online © Cambridge University Press, 2008 92 THE LIAO until the reign of Ching-tsung when in 977 an examination hall had been established in the Southern Capital.44 The first regular chin-shih examination was held in 988, 45 and they continued to the end of the dynasty, at first annually, as under the T'ang, and then at more irregular intervals. At first each examination produced only one or two candidates; but after about A.D. 1000 from twenty to forty candidates, sometimes more, were passed at each examination.46 Not all these chin-shih were employed: Later, under the Chin the government claimed that their Liao predecessors had given posts only to two or three out of every ten graduates.47 Perhaps the practical success of the examinations in producing a highly literate elite among the Chinese officials was at first less than their symbolic significance as the public adoption of one of the normative activities of a regular Chinese dynasty,48 and one that enshrined Chinese values. Sons and grandsons of high Chinese officials of the Liao also enjoyed the hereditary privilege of entry to office (yin) as under the T'ang, and such entrants generally enjoyed better prospects than did graduates. This practice fitted with the Khitans' widespread traditional use of hereditary succession (shih-hsiian) to office. Many offices were reserved for members of specific lineages of the Yeh-lii and Hsiao clans. Khitans were debarred from taking the Chinese examinations. Shortly after Sheng-tsung's death a prominent scholarly member of the imperial clan, Yeh-lii Shu-chen, was given two hundred lashes for allowing his precocious son to sit illegally for the chin-shih examination.49 The son was, nevertheless, employed and later gained preferment because he was also able to prove his mastery of a true Khitan skill, archery,'0 by killing three hares with three successive arrows. The subjects set for the Chinese examinations also sometimes had a distinctive Khitan flavor: In 1036 the chin-shih candidates had to compose their rhyme prose (fu) on the subject "shooting thirty-six bears in a single day"!'1 Systematic keeping of historical records Before Sheng-tsung's time there seem to have been court diarists, and some historical, or rather legendary, works had been commissioned. In 941 an 44 45 46 47 48 49 50 LS, 8, p. 64. LS, 12, p. 133. See the table in Wittfogel and Feng, pp. 491—2. T'o-t'o et al., eds., Chin shib (Peking, 1975), 51, p. 1129 (hereafter cited as CS). For a good summary of this problem, see Wittfogel and Feng, pp. 454—64. LS, 89, p. 1351. Both he and his father were skilled writers in Chinese, and the son Yeh-lii P'u-lu is said also to have been able to read the difficult large Khitan script by the age of six. His father was also one of the officials who advocated granting surnames to Khitan in addition to the Yeh-lii and Hsiao clans. 51 LS, 18, p. 217. Cambridge Histories Online © Cambridge University Press, 2008 CHANGES IN INTERNAL GOVERNMENT 93 official record was compiled of the legends surrounding the dynastic founder Ch'i-shou khaghan,'2 and under T'ai-tsung the Ch'i hsien chuan (Biographies of the seven worthies) was written." By Sheng-tsung's time there was already a Historiographical Office and a director of the national history. In 991 they presented the first Liao veritable records (shih-lu), with those for Ching-tsung's reign taking up twenty chapters. The director Shih Fang was rewarded in traditional style.'* We know also that during Sheng-tsung's reign a daily calendar (jih-li), the preliminary draft from which a later veritable record would be written, was already being compiled, as in 1003 the officials were warned not to include trivial matters in it." Rules on which matters should be reported for inclusion were made in 1011.56 By the next reign in 1044 we find as head of the Han-lin Academy and compiler of the national history one of the most remarkable scholars of the period, Hsiao Han-chia-nu,'7 who translated a variety of Chinese historical works into Khitan58 and also began the compilation of veritable records for earlier reigns together with two venerable Khitan scholars, Yeh-lii K'u-yii and Yeh-lii Shu-ch'eng.'9 It appears that in 994 the Liao produced their own calendar for the first time.60 Codification of law The gradual sinification of the Liao empire in the 980s that these developments exemplify was also marked by a series of legal measures that attempted to codify in Chinese style the confused and anomalous laws of the early period, when unwritten Khitan customary law had been applied to the Khitan and other tribal peoples and T'ang codified law was applied to the Chinese. The new emphasis on Chinese law as the standard applicable to all subjects seems to have been directly attributable to the influence of the empress dowager, who informally decided, on the basis of Chinese law alone, all disputes brought before her.6' 52 53 54 55 56 57 58 59 60 61 LS, 4 > p. 49. LS, 77, p. 1259. LS, 13, p. 141. LS, 14, p. 158. LS, 15, p. 169. For his biography, see LS, 103, pp. 1445-50. LS, 103, p. 1450. LS, 103, p. 1450; LS, 104, pp. 1456—7. LS, 42, p. 518. See Wang Ch'eng, Tung lu shih liieh; repr. in vols. 11—14 of Sang shih tzu liao ts'uipirn, 1st series, ed. Ch'ao T'ieh-han (Taipei, 1967), 123, p. 1899; and the notice of the empress dowager's death in Li Tao et al., comps., Hsu tzu chih t'ung Men ch'angpien (Peking, 1980), 72, pp. 1645—6 (hereafter cited as HCP). Cambridge Histories Online © Cambridge University Press, 2008 94 THE LIAO In 983 the T'ang legal code that had been presented by the administration of the Southern Capital was ordered to be translated into Khitan for use by the officials of the Northern Administration.62 Generally, the punishments specified in T'ang law were less harsh than those under Khitan customary usage, and T'ang law was more systematic and rational. Another indication of the supremacy of Chinese law was a ruling in 994 that any Khitan committing one of the Ten Abominable Crimes - a purely Chinese concept, grounded in Confucian ethical values — should be punished in the same way as a Chinese would.63 The compilation of a revised Chinese-style law code was ordered in 1027,64 and it was completed after Sheng-tsung's death. The employment of Chinese law for all citizens of Liao led to a hostile reaction among the Khitan. After Sheng-tsung's death in 1031, the new emperor's brother, Yeh-lii Chung-yuan, prime minister of the Northern Administration, requested that each of the five capitals set up a Khitan police commissioner to supervise the enforcement of the laws, presumably as a concession to the Khitans' feelings.65 Sheng-tsung left an excellent reputation as a just emperor, perhaps the best of the Liao rulers, but even he was liable to deal out arbitrary personal justice, particularly when drunk. In 1014, as Mu-tsung had once done before him, he instructed his ministers not to act on any decision he had taken when under the influence of liquor but to make him review his decision the next day. The state and agriculture At the end of the tenth century the Liao economy remained fragmented: The northern tribes retained their pastoral way of life, dependent on their herds and on marginal agriculture, but to the south the Hsi people had always engaged in farming, as had the population of Po-hai, and after the acquisition of the Sixteen Prefectures their long-established Chinese agrarian population became more and more the chief center of production and the most populous part of the Khitan empire. Under Sheng-tsung the government began to take some measures to improve communications, to encourage the agrarian sector, and to ensure the just collection of revenues. Throughout Sheng-tsung's reign a series of measures permitted and encouraged the exploitation and settlement of wastelands, whose farmers were to become taxpayers. In some cases, land was distributed together with plow62 63 64 65 LS, 10, p. n o . LS, 13, p. 145; LS, 6 1 , p. 939. LS, 17, p. 201. For Chung-yiian's request, see LS, 112, p. 1502. For the much-delayed implementation of the proposal in 1044, see LS, 19, p. 230. Cambridge Histories Online © Cambridge University Press, 2008 CHANGES IN INTERNAL GOVERNMENT 95 oxen. Perhaps connected with these measures was a decree of 1014 ordering that litigation not be allowed to interfere with agriculture. In 996 the military were prohibited from hunting at improper seasons and damaging crops. Inspectors were sent to examine the crops and to encourage agriculture and the planting of fruit trees. The emperor sometimes personally inspected the harvest. Similar measures continued until about 1070. One matter that caused constant contention was the attempt to irrigate land in the Chinese-settled Southern Capital circuit so as to grow rice. A proposal to do this was rejected during Ching-tsung's reign (969-79), and a renewed ban was imposed in 1064. In 1068, permission to cultivate rice was finally given, except along routes used by the army. It is clear that the government's objections were based on military considerations, as a landscape of canals and paddy fields would have been an impossible terrain for Khitan cavalry operations. Roads In Sheng-tsung's early years (984-9) serious attention was given to building roads and bridges to provide easier passage for carts and to improving the courier system, which was essential to the rapid transmission of orders and information.66 In 1027 a strip of land thirty double paces wide on either side of official highways was ordered to be kept cleared for security purposes. Taxation The history of taxation under the Khitan is all but impossible to unravel. The tribal population had traditional obligations to provide service and pay levies imposed by the emperor when these were needed. It seems certain, however, that most of the government's regular revenues were levied on the settled population of the Southern Capital circuit, where there were more tax collectors than in all the Khitan territories. Taxes there were reputed, by Sung authors at least, to have been far heavier than in the Sung empire. Labor services on the Chinese population seem to have followed no regular pattern, but manpower was mobilized ad hoc, sometimes with little regard to the annual pattern of agricultural work. Even in the last years of the dynasty the Liao's total revenues were not large by normal Chinese standards. The Khitan first imposed a tax system on its settled subjects under A-pao-chi's Chinese adviser Han Yen-hui (882—959) in 912. They continued to impose a modified version of the two-tax system 66 Wittfogel and Feng, pp. 164-5. Cambridge Histories Online © Cambridge University Press, 2008 96 THE L1AO established under the Tang, plus various surtaxes such as a levy on agricultural implements, and they frequently levied manpower for corvee work. There was a salt monopoly levied on both coastal salt and the product of inland salt lakes. There was also a tax on liquor and ferments, and various merchant taxes were levied on market transactions and goods in transit. During the reign of Sheng-tsung there were no new developments, but the expenses of the many campaigns clearly brought about somefinancialstrain. In 991 there was a land survey67 and a general census was taken for the first time, and in 997 a census was taken of the tribal population of the various ordos.68 In the 990s the government repeatedly readjusted taxes in various ways: In 991, inspectors from the north were sent to the region around Peking to look into the evasion of taxes and corvee duty.69 In 994, the government ordered that an "equitable" system of taxes be established.70 It is unclear what this meant; certainly by the end of Sheng-tsung's reign great inequities still existed, and so his successors ordered a new reassessment.71 In 995 the taxes of some areas were further increased, and in 997 the taxes of the Southern Capital circuit under the new system were reduced because they had become intolerable.72 In 998 further concessions followed,73 and in 1002 the southern areas were again given tax exemptions, and trade taxes were cut.74 The settlement with Sung in 1005 immediately reduced the strain on the Liao finances. The annual subsidy paid by the Sung government was the equivalent of a considerable proportion of the Liao's total revenues. But in the next few years new trade taxes were imposed, and the extraordinary levies continued. Coinage The Khitan had made copper cash even before the time of A-pao-chi, and sometime in T'ai-tsung's reign (927—47) an official was appointed to control the minting of cash and iron production. Shih Ching-t'ang, founder of the puppet Chin regime (936-46) and a loyal vassal of the Khitan, had supplied large amounts of copper cash to help the Liao economy. But during Shihtsung's reign, the Sung captive Hu Chiao reported that silk, rather than cash, was the main form of currency even at the capital.75 At first the Khitan 67 68 69 70 71 72 73 74 75 LS, 13, p.139. LS, 13, p. 149. LS, 82, p. 1290. LS, 13, p. 145. LS, 59, p. 925. LS, 13, p. 148. " . 13. PP- !49-5°LS, 14, p. 157. Ou-yang Hsiu, Hsin Wu-taishib, 73, p. 906; LS, 37, p. 441. Cambridge Histories Online © Cambridge University Press, 2008 CHANGES IN INTERNAL GOVERNMENT 97 depended heavily on coin imported from China, and large numbers of Sung cash have been found in what were Liao territories. The general use of metallic coin and the serious minting of cash began only at the end of Chingtsung's reign with the issue of the Ch'ien-heng coinage in 982.? 6 Another new minting, the T'ung-ho coinage, began in 983 at Sheng-tsung's accession. A large addition to the stock of coinage came in 996 when a huge cache of copper coin hidden almost a century before near Peking by Liu Jen-kung, the former governor of Lu-lung, was discovered. In 1021 a new T'ai-p'ing coinage was minted, and by the end of Sheng-tsung's reign there seems to have been an adequate supply.77 Large amounts of Sung coin circulated in Liao territory. But the quantities of coin mentioned in our sources are still very small in comparison with those of T'ang or Sung. After Sheng-tsung's time the use of money greatly increased, but the minting of coin failed to keep up with demand. By 1055 a crisis in the supply of money seems to have developed. Counterfeiting and trade in copper and iron were strictly regulated, and the export of metals to the Uighurs and Mongols was prohibited. 78 From 1056, coinage was for the first time manufactured in the Eastern Capital. 7» For the rest of the eleventh century, although new coinages were minted in 1055, 1065, 1074, 1084, 1102, and 1112 and although even the Koryo records speak of the widespread use of copper cash in Liao, there seems to have been little control over the quality of the money in circulation. The crude quality of surviving examples of Liao cash bear this out. By the 1070s the traditional official reactions to a cash shortage — the prohibition of the manufacture of copper implements (1084) and bans on exports of metals80 and of cash (1088) — begin to appear. In 1090 Su Ch'e, who was sent as an envoy to the Liao, reported that all the coin circulating there was Sung copper currency.8' By the early twelfth century, government expenditures began to far outstrip both revenue and the production of coin, and the dynasty ended with a serious money shortage. Building of the Central Capital and the final incorporation of the Hsi From the time of T'ai-tsung the Hsi peoples, after a series of rebellions and Khitan punitive expeditions during A-pao-chi's reign, had finally accepted a 76 Numismatic works claim to identify earlier coinages, but some of these are of dubious authenticity. See P'eng Hsin-wei, Chung-kuo huopi shih (Shanghai, 1958), p. 371. 77 LS, 60, p. 931; Wittfogel and Feng, pp. 181—7; Peng, Chung-kuo huo pi shih, pp. 3 7 0 - 2 . 78 LS, 60, p. 931. 79 LS, 21, p. 254. 80 LS, 60, p. 931; 22, p. 270. 81 See Su Ch'e, Luan ch'eng chi (Peking, 1984), 42, p. 938. Cambridge Histories Online © Cambridge University Press, 2008 98 THE LIAO position of semi-independence within the Khitan state not dissimilar to that of Po-hai.82 They had retained their own king, who had his own officials, and they had paid tribute as vassals, not taxes as subjects, to the Khitan government. During the frontier wars with the Sung in the last decades of the tenth century, during which their territory was invaded, giving them an opportunity to break with the Khitan had they so wished, the Hsi people remained almost entirely loyal. Between 994 and 997 a series of reforms by the Hsi king's administration were enacted, and the Hsi people's former "tribute" was discontinued. The Hsi king now became a salaried Liao official. Chinesestyle local administrative units were set up to control former Hsi territories, and Chinese settled there to farm the rich valleys. In 1006 the former residence of the Hsi king was designated the Khitan Central Capital (Chung ching).83 In 1007 the site was walled, and a Chinese population was moved from Liao-tung to settle in the new city: A Liao ancestral temple was built, together with reception hostels for envoys from Sung, Korea, and Hsia. Further preparations were made for a visit by Sheng-tsung in 1009 and more public buildings were constructed between 1018 and 1020.8« The new capital seems to have served for some time as a cult center and a site for receiving diplomatic missions: The Sung envoys Sung Pu and Lu Chen visited it in 1008 and left descriptions of it, as did Wang Tseng, who came in 1013. It had an inner and outer wall, but in Lu Chen's time it seems to have been still largely uninhabited. Unlike the four other capitals (the Supreme Capital, Eastern Capital, Southern Capital, and the Western Capital that was later established in 1044 at Ta-t'ung), all of which were considerable cities, it remained relatively small and administered only a limited area, with few subordinate local administrations and a small registered population, mainly Chinese and Hsi. But its establishment finally brought the Hsi into the Khitan state, and from the eleventh century onward they appear less and less frequently in our sources as a separate people. FOREIGN RELATIONS In spite of these wide ranging changes in the internal politics and institutions of the Khitan state under Sheng-tsung, the most important changes took place in foreign relations. At his accession, the Liao were still poised on the brink of war with the Sung, were almost isolated from the Kory6 kingdom, 82 On the position of the Hsi in the Khitan state, see Shimada, Ryochoshi no kcnkyu, pp. 8 - 1 0 ; Li Han and Shen Hsiieh-ming, "Liieh lun Hsi tsu tsai Liao tai te fa chan," Sung Liao Chin shih lun tsung, 1 (1985), pp. 2 7 7 - 9 4 . 83 On the Central Capital, see Shimada, Ryocho shi no kenkyu, pp. 443—56. 84 IS, 14, p. 163; 16, pp. 184, 18;, 188. Ancestral temples were erected for Ching-tsung in 1019 and for T'ai-tsu in 1020. Cambridge Histories Online © Cambridge University Press, 2008 FOREIGN RELATIONS 99 and still suffered constant problems with their tribal dependents, both the Jurchen in the east and northeast and the various Tangut and Mongolian tribes in the southwest and west. By the end of his long reign, stable interstate relations had been established with both the Sung and Kory6 courts, which led to almost a century of peace and stability. Renewed hostilities with the Sung, 986 The ignominious defeat of Sung T'ai-tsung's invasion of Liao in 979 left him avid for revenge and provoked violent reactions at the Sung court. In 980 Li Fang and Hu Meng advised the emperor that the Sung empire lacked the military capability to renew its assault on the Liao, but the humiliation of 979 rankled among the ministers, who constantly referred to the Khitan in colorful pejorative rhetoric as evil and uncivilized barbarians richly deserving of chastisement. By 985 the Sung were ready to launch another large-scale invasion, and this time they tried to form an alliance with the Korean king for a coordinated attack on Liao "in defense of their common civilized values."8' In 986 T'ai-tsung could contain his impatience no longer and mobilized a huge army to drive the Khitan out of the "lost prefectures." Three army groups crossed the border simultaneously, at Yen-men and Fei-hu on the Hotung border and from Hsiung-chou in western Ho-pei. The Sung armies at first overwhelmed the border defenses and made some territorial gains. Some of the Liao border commanders defected and surrendered to the Sung. But the tide soon turned, as the Liao generals lured the invaders deep into their own territories, far from their supply lines, and then surrounded and attacked them from all sides. The Liao forces won resounding victories on all three fronts; the Sung suffered heavy casualties and lost large numbers of captives and huge quantities of supplies and weapons.86 The invasion was yet another military disaster for the Sung, but it also caused serious disruption and damage in the frontier areas of Liao, where many people fled their homes; the areas south of the Southern and Western capitals, the main objectives of the invasion, were devastated and did not recover for many years. However, thousands of surrendered Sung troops were incorporated into the Liao armies, and some Sung officials and examination graduates were absorbed into the Liao civil administration. 85 See Wang Gung-wu, "The rhetoric of a lesser empire: early Sung relations with its neighbors," in China among equals: The Middle Kingdom and its neighbors, loth—iqth centuries, ed. Morris Rossabi (Berkeley and Los Angeles, 1983), p. 53. 86 For a detailed account of the fighting, see Ch'eng Kuang-yii, Sung T'ai-tsung tui Liao chan cheng k'ao (Taipei, 1972), pp. 9 5 - 1 6 1 . Cambridge Histories Online © Cambridge University Press, 2008 IOO T H E LIAO Sung T'ai-tsung and some of his court still were determined to renew hostilities. In 988 and 989 T'ai-tsung ordered his ministers to debate the options open to him in dealing with the Khitan. The ministers faced a tricky task in reconciling the practical realities with their emperor's preference for the idealistic solution of asserting Chinese superiority once and for all. A diplomatic solution to the frontier problem was suggested, though this was politically possible only as a "temporary expedient" in the face offorce majeure. But no major new campaign was planned. Liao relations with Korea During the final two decades of the tenth century, the Liao found themselves embroiled in a hostile confrontation not only with the Sung but also with the troublesome border tribes of the Jurchen, with Korea in the east, and with the rising Tangut state of Hsia in the Ordos area to the west. The Khitans' relations with Korea until the 980s had been of minor importance. The critical period of the Khitan conquest of Po-hai in 926, when conflict might have been expected, had coincided with a period of political fragmentation in Korea. The Silla state had begun to collapse in the last decade of the ninth century, and rebel leaders had set up three independent states: Later Kogury6 in the north, Later Paekche in the southwest, and a rebel regime led by Wang K6n on the west coast. In 918 Wang K6n usurped the leadership of Later Koguryd and established the Koryo dynasty in the north and northwest (he is known by his temple name T'aejo; r. 918— 43). In 926 Korea was thus split among three states and quite unable to come to the defense of Po-hai, even had they wished to do so. It was not until 935 that Silla finally surrendered to Koryo and 936 before the Kory6 king T'aejo conquered Later Paekche and reunified the peninsula (see Map 5). In the following decades, the Kory6 kingdom began a systematic expansion and strengthening of its position in the north, establishing a new "western capital" at P'y6ng-yang.8? As the self-professed heirs of ancient Koguryo, their rulers aimed to reconquer the territories north of the Taedong River, toward the Yalu Valley. There still remained no urgent cause for conflict with the Khitan, however, as this region was occupied by a variety of Jurchen and other tribal peoples and by remnants of the Po-hai population. Moreover, there was powerful opposition at the Koryd court to any territorial expansion to the north. Many of the nobles and officials strongly believed 87 This can be followed in the table showing the establishment of new local administrative centers, some with considerable garrisons, given in Chin Wei-hsien, Ch'i-tan te tungpei chmg ue (Taipei, 1981), pp. 79-81. Cambridge Histories Online © Cambridge University Press, 2008 FOREIGN RELATIONS from Tung-ching (EASTERN CAPITAL) AO * . . k'ai.rhr K'ai-chou First invasion, 993 Second invasion, 1011-12 Third invasion, 1018-19 300 km ' ioOmili MAP 5. Liao campaigns against Koryo, 993-1019 Cambridge Histories Online © Cambridge University Press, 2008 IOI IO2 THE LIAO that the Kory6 state should, as in Silla's time, restrict its aims to the control of the peninsula.88 Nonetheless, the Khitan conquest of Po-hai, followed by the resettlement of much of the Po-hai population around Liao-yang, had subtly altered the situation. Not all the former territories were incorporated into the Liao state, and not all the Po-hai people had accepted Liao rule in 926. Many members of Po-hai officialdom had taken refuge in Kory6. Po-hai itself had originally been founded by remnants of the Kogury6 ruling elite and hence looked on the northern-oriented Kory6 dynasty as distant relatives and potential allies.89 Moreover, there were three groups of Po-hai people who remained independent, beyond the Liao's eastern border: the northwestern Po-hai in the Sungari Valley in modern Heilungkiang, a group of Po-hai living west of the Yalu in Liao-ning, and, most powerful of all, a group who in 926 had established the independent state of Ting-an (Korean: Chongan) in the upper valley of the Mu-tan chiang with its capital city Wu-jo at the site of the former Po-hai Supreme Capital (modern Tung-ching, Kirin).9° In 975 Ting-an came into conflict with the Liao. A Khitan punitive expedition was sent against it but failed. In 985—6 the Liao again invaded the area, and in the 980s there was constant trouble with Ting-an and the various Jurchen tribes living in the Yalu Valley, with whom the Sung had tried to establish some sort of anti-Khitan alliance. In 991 the Liao established three forts with garrisons and military colonies in the lower Yalu Valley, to prevent communication by sea between the Jurchen and the Sung. All these activities threatened Koryo. A Liao plan to invade Kory6 in 947 had been aborted when the Kory6 king, Chongjong (r. 945—9), mobilized a large defensive army and the Liao had realized what a powerful adversary Kory6 might be. Until the 980s Khitan—Kory6 relations had been at arm's length, for the Jurchen tribes and Ting-an had provided a buffer zone between Koryo's northern frontier and the Liao border, but by 990 it was clear that Liao wanted to annex this territory. The new activities by the Liao that revived the urgency of the situation not only revealed the latent enmity between Liao and Koryo but also linked it with the broader international situation. Kory6 had maintained cultural and diplomatic relations with the Sung ever 88 For a clear analysis of the polarization of attitudes at the Koryd coutt, see Michael C. Rogers, "National consciousness in medieval Korea: The impact of Liao and Chin on Koryd," in China among equals: The Middle Kingdom and its neighbors, loth— 14th centuries, ed. Morris Rossabi (Berkeley and Los Angeles, 1983), p. 152. 89 On this relationship, see Ki-baik Lee, A new history 0/Korea, trans. Edward W. Wagner (Cambridge, Mass., 1984), p. 103. 90 On Ting-an, see Wada Sei, "Teian koku ni tsuite," Toyo gakuho, 6 (1915); repr. with revisions in his Toa shi kenkyu: Manshu hen, Toyo bunko ronso no. 37 (Tokyo, 1955), pp. 161—89; and Hino Kaisaburo, "Teian koku," in vol. 6 of Ajia rekishi jiten (Tokyo, i960), p. 388. Cambridge Histories Online © Cambridge University Press, 2008 FOREIGN RELATIONS IO3 since 962,9' as they previously had with each of the Five Dynasties in turn. In 963 King Kwangjong (r. 949—75) had been invested with his title by Sung T'ai-tsu. By the tenth century the Koreans were thoroughly imbued with Chinese cultural influence at all levels and hated, despised, and feared the Khitan. The Sung emperors were thus building on potent real sentiments when in 985 they attempted to enlist Kory6 as an ally in defending a common cultural heritage. Neither party, however, was willing to put its alliance to the test. Accordingly, during the Sung invasion of 986, Koryo had refused to attack Liao. The Khitan, for their part, believed that Kory6 posed a serious potential threat to their eastern border. Tension had been increased by Khitan attacks on the Jurchen and Ting-an. The Khitan also may well have feared that Kory6 might encourage the endemic disaffection of the large Po-hai population within Liao, a disaffection that eventually resulted in a major rebellion in 1029-30. In 9 9 2 - 3 Hsiao Heng-te, 92 viceroy of the Liao Eastern Capital, was ordered to invade Kory6. He crossed the border with a huge army, which he claimed to number 800,000, demanding the cession to Liao of the former Koguryd territories both south and north of the Yalu. The Kory6 court asked the Sung for assistance. But none was forthcoming: Once again the Sung— Kory6 "alliance" had proved to be a sham. The Koreans, however, were well able to take care of themselves and prepared strong defenses. Their king, SSngjong (r. 982-97), led a powerful army north to P'y6ng-yang. Meanwhile, after some fighting the main Liao force advanced as far as the Ch'6ngch'on River. At this point it seems that the Liao command decided that the conquest of Korea was impossible, and so negotiations began between the chief Liao general, Hsiao Heng-te, and the Kory6 commander, S6 Hui. The Liao at first offered Kory6 the choice between total surrender or annihilation, and the Kory6 was inclined to submit. But S6 Hiii's intrepid persistence enabled them to reach a negotiated settlement under which Kory6 became a nominal tributary of the Liao and broke off its long-established links with the Sung. The Korean king was invested with his title by the Liao emperor. Koryo was left free to deal with the Jurchen tribes south of the Yalu Valley, which they promptly acted upon: S6 Hui led his army into the area in 994—6 and established forts to maintain peace.93 From 994 onward there were constant exchanges of envoys. In 994 91 See Marugame Kinsoku, "Korai to So to no tsuko mondai," Chosen gakuho, 17 (i960), pp. 1—50; 18 (1961), pp. 5 8 - 8 0 . 92 For his biography, see LS, 88, pp. 1342-3. He is generally referred to in Korean histories by his style, Sun-ning. 93 Ki-baik Lee, A new history of Korea, p. 125, and map, p. 127; Han Woo-keun (Han U-gun), Thehistory of Korea, trans. Lee Kyung-shik (Honolulu, 1971), pp. 138—9. See also Rogers, "National consciousness in medieval Korea," pp. 154—6, which questions the veracity of the traditional account. Cambridge Histories Online © Cambridge University Press, 2008 IO4 THE LIAO SSngjong offered women musicians to Sheng-tsung, who, with a show of Confucian virtue, declined them. Batches of Korean students were sent to learn the Khitan language, and in 996 King S6ngjong sent an envoy requesting a marriage alliance. The Liao court permitted him to marry Hsiao Hengte's daughter (whose mother was a princess from the royal family). When S6ngjong died in 997, the Khitan court duly invested his successor Wang Chong (posthumous title Mokchong; r. 997-1009) as king in 998. During the following years the Liao were preoccupied with their invasion of Sung and the subsequent treaty negotiations. During the wars of 999— 1000 and 1004 there was no question of the Kory6—Sung alliance's being revived: In each case the Kory6 king formally congratulated the Liao court on its victories over the Sung. War with the Sung The Sung and Liao forces had confronted each other along the border since the 980s, with occasional outbreaks of localized fighting. In 994 the Sung court had made some tentative moves to establish peaceful relations, but these had been rejected. Both courts had outstanding territorial claims on the other. Sung irredentists still looked for the total recovery of the "Sixteen Prefectures" ceded to the Khitan by Chin in 938. Khitan revanchists, emboldened by their easy victories in 986-7, hankered to reconquer Mo-chou and Ying-chou, the southernmost of the prefectures that had been ceded to them by Chin in 938 and that the Chou had wrested from them in 959. This area they referred to as Kuan-nan, "South of the Barriers." The late 990s saw a change in the situation. As a result of the war with Kory6, the threat of a hostile Sung—Kory6 alliance had been removed. Meanwhile, the Tangut Hsia regime in northwest China was becoming a major irritant on the Sung border, under their warlike chieftain Li Chi-ch'ien, who had remained obstinately independent after another faction of the Tangut leadership had surrendered to the Sung in 982. In 986 Li Chi-ch'ien, having rebelled against the Sung, submitted to the Khitan and was given various offices and titles as a vassal ruler. In the spring of 989 he was given as a bride a member of the imperial clan who had been quickly granted the title of "princess," and in the next year Li Chi-ch'ien was formally installed by the Khitan court as the king of Hsia. There was thus at least a formal alliance between the Khitan and Hsia, but it was, however, rather precarious. In 992, hearing that the Hsia were negotiating in secret with the Sung, the Khitan sent a punitive expedition led by Han Te-jang's brother Han Te-wei to attack Hsia territory along the upper course of the Huang-ho River. There were further troubles with other Cambridge Histories Online © Cambridge University Press, 2008 FOREIGN RELATIONS IO5 Tangut tribes living in western territories of the Liao in 997 and in 1001. But Li Chi-ch'ien remained a Khitan vassal, albeit far from a docile one, and his forces could threaten the Sung's lengthy northwestern frontier. With the accession of Chen-tsung in 997, the Sung court was no longer led by a soldier-emperor, but by perhaps the most passive of all the Northern Sung monarchs. There also were changes in the leadership at the Liao court. In 998 Yeh-lii Hsiu-ko, the long-serving Khitan commander in chief, died. In late summer of the next year, the emperor promulgated an edict announcing mobilization for a campaign against the Sung. Just as this campaign was being prepared, the northern commissioner for military affairs, Yeh-lu Hsiehchen, also died, leaving Han Te-jang the most powerful man in the Khitan kingdom after the emperor and the empress dowager. The Liao armies attacked first toward the crucial strategic center of Ting-chou, attacking the county town of Sui-ch'eng. The first Liao assault was repulsed, but in midwinter the Liao won a second battle in the same area. Meanwhile, another Khitan force advanced farther to the east toward Ying-chou. After inflicting a heavy defeat on a Sung army near the city, this force advanced south as far as Lo-shou (modern Hsien-hsien, Hopei). The Liao forces, however, had failed to take the prefectural cities of either Ting-chou or Ying-chou, and after new year iooo, the campaign was called off and the armies disbanded. In the winter of IOOI the Liao armies, this time under the emperor's own command, again invaded Sung northeast of Ting-chou and defeated their forces at Sui-ch'eng. This time the Sung counterattacked farther east, and after this attack was beaten back, hostilities again came to an end. The summer of 1002 saw renewed border clashes, and in 1003 the Liao army under Hsiao T'a-lin renewed its attack on Ting-chou, this time penetrating as far as Wang-tu, where they captured in battle an important Sung official named Wang Chi-chung who commanded the forces in the Ting-chou region. None of these seasonal campaigns seems to have resulted in any permanent occupation of Sung territory (see Map 6). In 1004 the Liao court decided on a full-scale invasion. In the eighth month, news reached the Sung court that Khitan light cavalry had raided the area between Mo-chou and Ch'i-chou but had withdrawn, apparently unwilling to engage Sung forces in battle. 94 The prefect of Hsiung-chou, sensing serious trouble, asked that the authorities in the coastal prefectures open the floodgates along the canal and on the seacoast, flooding a wide area to prevent any Khitan advance in that area.95 A few days later the Sung court ordered the deployment of crack troops to the Ting-chou area and the general 94 HCP. 57, p. 1251. 95 HCP. 57, p. 1252. Cambridge Histories Online © Cambridge University Press, 2008 io6 THE LIAO M A P 6. The Liao invasion of Sung, 1004 Cambridge Histories Online © Cambridge University Press, 2008 FOREIGN RELATIONS IO7 conscription of young men throughout Ho-pei and Ho-tung, to be armed and trained for local defense. The invasion began in late autumn, the Liao armies being under the command of the emperor and empress dowager. A powerful army in the west first attacked along the same route as in previous years, taking Sui-ch'eng and Wang-tu.*6 But again, instead of taking Ting-chou itself they turned south to take the prefecture of Ch'i-chou, in the tenth month, and then swept rapidly southward down the western flank of Hopei. Meanwhile, another part of the army had split off and attacked Ying-chou, the crucial stronghold in the northern plain and the center of the Kuan-nan district that the Khitan hoped to reconquer. The city was besieged for more than two weeks, a siege bitterly contested by both armies, but the Khitan finally failed to take the city, and so their army withdrew and joined the advance to the south. Early in the eleventh month they routed a Sung army in Ming-chou (northeast of modern Han-tan) and advanced to Shan-yuan (sometimes read T'an-yuan), just north of the Yellow River, where the Sung had concentrated their main forces, under Chen-tsung himself, to oppose them. They were little more than a hundred kilometers from the Sung capital, K'ai-feng. The scene was set for a climactic battle. In a preliminary sortie the Khitan general Hsiao T'a-lin, 97 who had led the march through Ho-pei, was killed by a sniper using a long-range crossbow, and the Khitan attack was repulsed. Khitan forces meanwhile made a quick but destructive raid on the nearby military prefecture of T'ung-li. Negotiations to reach a peace settlement had been going on for some time. 98 The Sung were fearful of the military outcome and had been anxious for a peaceful settlement even before the 1004 invasion began. The Khitan too, even though they had resorted to armed force, were still willing to negotiate. On the Khitan side the key personality was Wang Chi-chung, the Sung official taken by the Khitan in 1 0 0 3 . " Having won the confidence of the empress dowager, he had been appointed a finance commissioner and had married a woman descended from K'ang Mo-chi, one of A-pao-chi's earliest Chinese advisers. Wang Chi-chung had not only been an important Sung commander; he had also been a personal confidant of Chen-tsung, having served in his princely household before his accession, and on his palace staff. 96 HCP, 57, pp. 1 2 6 5 - 6 . 97 Hsiao T'a-lin, whose name appears in Sung sources as Hsiao Ta-lan, was an experienced general who had fought in earlier campaigns against the Sung, in Korea, and with the empress dowager's sister Hu-lien on the Mongolian frontier. See LS, 85, pp. 1313-14. 98 There are differing versions of these events. Liao inscriptions (see Wittfogel and Feng, p. 355, n. 45, and LS, 14, p. 160) state that the Sung initiated negotiations. Sung sources state that the Liao were responsible. See T'o-t'o et al. eds., Sung shih (Peking, 1977), 7, p. 125 (hereafter cited as SS); CTKC, 7, p. 4a; HCP, 57, pp. 1 2 6 8 - 9 , gives a very circumstantial account, which I follow. 99 For their biographies see LS, 8 1 , pp. 1284—5; SS, 279, pp. 9 4 7 1 - 2 . Cambridge Histories Online © Cambridge University Press, 2008 IO8 THE LIAO His counterpart on the Sung side was Pi Shih-an (938—1005), who had served with him in the future Chen-tsung's household and was now councillor. He had encouraged Chen-tsung to go himself to the front and recommended K'ou Chun (961 — 1023) to watch for possible peace initiatives. With the Liao empress dowager's approval, Wang Chi-chung submitted a memorial to the Sung emperor through the Sung prefect of Mo-chou, stating that the Liao court wished to restore friendly relations.100 This memorial had been sent after the Liao forces had occupied the territories that they claimed in Kuan-nan. The Sung emperor was surprised and at first suspected that this was a trick, but he wasfinallypersuaded to begin negotiations. The preliminaries were, however, delayed for weeks by the obstreperous conduct of Wang Ch'in-jo (962—1025), who delayed the Sung envoy. This delay, however, worked in favor of peace: By the time the Sung envoy, Ts'ao Li-yung (d. 1029), reached the Liao headquarters, both sides were ready to negotiate. The Sung had halted the Khitan army's advance and stood opposing it in overwhelming force in a well-fortified position. The Liao had achieved their primary objective, the occupation of the Kuan-nan prefectures, and had penetrated deep into Sung territory. The Liao hope that the Sung would cede these territories in exchange for peace did not, however, materialize. Ts'ao Li-yung stubbornly rejected all Liao territorial claims and offered instead an annual payment of silk and silver, warning that the only alternative to this offer was continued warfare. That the Liao gave in and accepted the terms offered by the Sung was probably due to a realization on both sides that they had reached a military stalemate. The Khitan realized that their long-term position was far from strong, although they were within striking distance of the Sung capital. Their army was in much the same situation that T'ai-tsung had been in when he occupied K'ai-feng in 947. They were boxed into a narrow strip of territory surrounded both on the east and west by intact Sung forces, and even within their occupied area, several strategic prefectures and fortresses that had simply been bypassed by the Liao cavalry still held out. Notable among these strong points were Tingchou and Ying-chou. Indeed, the Liao army faced the real danger of being cut off from its home base and trapped in enemy territory. The treaty of Shan-yuan, 1004 After only a few days of negotiation a treaty was concluded, and the two courts exchanged oath-letters stating the terms of peace. They agreed to the following: 100 HCP, 57, p. 1268; SS, 7, p. 125. Cambridge Histories Online © Cambridge University Press, 2008 FOREIGN RELATIONS 109 1. Sung would pay the Liao annually 200,000 lengths of silk and 100,000 ounces of silver as a "contribution to military expenses." 2. The border would be carefully demarcated. 3. Both sides would take strict measures against unauthorized infringements of the border, and neither side would disturb the cultivated lands of the other. 4. Neither side would give refuge to fugitive criminals. 5. Existing border fortifications might be repaired, but no new fortifications or canals might be built along the border. 6. Both sides would observe the treaty, which was sworn with a solemn oath invoking religious sanctions in case of infringement; they would cultivate friendly relations; and they would respect each other's territorial integrity. Other agreements provided for the repatriation of prisoners and established the norms for diplomatic and commercial intercourse between the two empires. The terminology used in this treaty had deep symbolic significance. The Sung insisted that the annual payments to the Liao be called a "contribution to military expenses," to avoid their being given the humiliating title of "tribute." Likewise, the payment was to be delivered by a minor Sung official at the border prefecture of Hsiung-chou, to demonstrate that the court at K'ai-feng considered it only afinancialtransaction, not a political act implying submission. The Sung court addressed its northern neighbor as the "Great Khitan state" (Ta Ch'i-tan kuo) or the Great Liao State (Ta Liao kuo), and the Khitan referred to it as the "Southern Sung." The two dynasties spoke of each other as the "Southern Court" and the "Northern Court." Their monarchs entered into a fictitious kinship relationship as "brothers." The Sung emperor was to address the Liao empress dowager as his aunt, the Liao emperor as "imperial younger brother," and the Liao emperor referred to his Sung counterpart as "elder brother."101 This relationship involved them in a ceaseless round of ritual exchanges, in which the Khitan and Sung envoys were treated at the other's court quite differently from those of other states. Each state observed the taboos on the personal names of deceased emperors of the other state. Embassies were exchanged on such ritual occasions as the New Year celebrations, imperial birthdays, the deaths of emperors or their empresses, and the enthronement of new monarchs. 101 The kinship terms used by the two monarchs in addressing one another and the dowager empresses of the other court were determined by their actual age and generation and changed with each reign. See the table in Taojing-shen, Two sons of heaven: Studies in Sung-Liao relations (Tucson, 1988), p. 107. In practice it turned out that the Sung emperor was usually older than his Liao counterpart and thus was addressed as "elder brother," but this was the result of chance, not of design. In other reigns the relationship was that of uncle and nephew, and from 1076 until 1101 the Sung emperor Che-tsung addressed the long-lived Liao ruler, Tao-tsung, as his "junior grandfather." Cambridge Histories Online © Cambridge University Press, 2008 IIO THE LIAO The treaty of Shan-yuan was a great success of political realism over ideological pretensions. It paved the way for a century of stability and peaceful coexistence, confirmed and guaranteed by an uninterrupted flow of missions between the two courts. There were minor infringements, usually settled by the local authorities, and two more serious crises in 1042 and in 1074—6. But the settlement held. The border between the two states from the coast to the bend of the Huang-ho River was clearly demarcated and zealously guarded by both sides, constituting a genuine international frontier in the modern sense, something unprecedented in China's history. All in all, the treaty was remarkably successful and contributed greatly to the longterm stability and the economic and cultural progress enjoyed by both states during the eleventh century.Io2 The results of the treaty have commonly been misrepresented; in particular, the annual subsidies paid by Sung to the Khitan have been described as having placed a heavy burden on the Sung state. Such was certainly not the case. The annual quota of silk presented to the Khitan was equivalent to the production of only a single southern prefecture such as Yiieh-chou. And the payments must be viewed in the context of Liao—Sung trade as a whole. Trade had continued uninterrupted after the war of 979, with only brief breaks during actual hostilities, and resumed on an even broader scale after the treaty was concluded. In this trade the Sung enjoyed a massive surplus, and it has been calculated that about 60 percent of the annual silver subsidy was eventually returned to Sung as payment for various Chinese products, especially silk, for which there was an insatiable demand in the north. Whereas the subsidies were not a major outlay for the immensely wealthy Sung empire, they were far more important to the Khitan, whose revenues were comparatively meager. They used the silk for their own major internal expenses, for instance, in building the new Central Capital immediately after the treaty, and also traded much of it to their neighbors, including the Uighurs, Tanguts, Koreans, and the tribal peoples of Mongolia. The arrangement was thus a good bargain for both parties. The Sung ensured lasting peace at a modest price. The Khitan acquired a steady source of additional revenue and were able to reduce their southern border defenses to some degree and to concentrate on internal developments. 102 On the Shan-yuan treaty, see Chiang Fu-tsung, Sung shib hsin t'an (Taipei, 1966), pp. 142 ff.; Wang Min-hsin, "Shan-yuan ti meng te chien t'ao," Shih huoyiieh t'an, n.s., 5 (1975), pp. 97-107. The fullest study of the treaty in a Western language is by Christian Schwarz-Schilling, Der Friede von Shan-yuan (1005 n. Cbr.): Ein Beilrag zur Geschichte der chimsischen Diplomatic, Asiatische Forschungen no. i (Wiesbaden, 1959). See also the lengthy review of this work by A. F. P. Hulsewe in T'oung Poo, 47 0957). PP- 445—74, which provides many important corrections. Cambridge Histories Online © Cambridge University Press, 2008 FOREIGN RELATIONS III Resumed warfare with Koryd, 1011—1019 Until the death in 1009 of the empress dowager, whose funeral was duly attended by envoys from Kory6, peace continued on the eastern frontier. However, in that same year there was a violent coup at the Kory6 court. Kang Cho, governor of the Western Capital (P'ySngyang) was summoned to the main capital, Kaegy6ng, to help eliminate a clique that was plotting to depose King Mokchong. Having done as he had been bidden, Kang Cho himself murdered the king and enthroned a new ruler, Wang Sun (posthumous title Hy6njong; r. 1009—31), who was expected to rule under his tutelage. The Khitan sent an expeditionary army of 400,000 men across the Yalu River to punish the murderer of their former vassal, despite pleas from the new Korean king to halt the campaign. Its chief commanders were Hsiao P'ai-ya, elder brother of Hsiao Heng-te, and Yeh-lii P'en-nu. A first clash went Kory6's way, but in a second attack the Liao were victorious, and Kang Cho was captured and later killed. The Liao army occupied several border prefectures north of P'y5ng-yang. The Kory6 king offered to surrender, but the governor of Kaegy6ng murdered the Liao envoys and prepared to resist. The Liao army now marched south and took Kaegy6ng after winning a pitched battle outside the city. Hsiao P'ai-ya and Yeh-lii P'en-nu sacked and burned the capital, destroying the palaces, official buildings, and the Kory6 archives. The Kory6 king had taken refuge in the south, but his armies regrouped. The Liao force began to withdraw to the frontier. The surrendered districts rose against them, and in the depth of winter the army was mired down in the mountains and forced to abandon much of their armor and weapons before eventually recrossing the Yalu. The Koryfi king now sued for peace. But the Liao demanded that he come to court in person to do obeisance as a vassal and also that he cede key border areas. The Kory6 court refused, and a decade of hostility ensued. Both sides fortified the frontier regions, and Korean intransigence was strengthened by an internal coup d'etat that left the generals rather than the civil officials in control of the Kory6 court. In 1014 the Liao ordered the fortification of several border prefectures and the construction of a heavily defended permanent pontoon bridge across the Yalu. From 1015 to 1019 there was incessant warfare, with attacks on Koryd in 1015, 1016, and 1017 in which victory went sometimes to Kory6, sometimes to the Khitan, but in sum were indecisive. In 1018 a huge new expeditionary force was mobilized by the Khitan and placed under the command of Hsiao P'ai-ya. The army crossed the Yalu late in 1018 but was ambushed by a Cambridge Histories Online © Cambridge University Press, 2008 112 THE LIAO superior Kory6 force, suffering severe losses. The Kory6 army had also cut their line of retreat, and so Hsiao P'ai-ya marched south, planning to take the capital Kaegy6ng, as in i o n . But this time the Koreans had prepared defenses around the capital, and the Khitan, constantly harried by Korean attacks, were forced to retreat toward the Yalu. At Kuju, between the Ch'a and T'o rivers, they were encircled and attacked by the main Kory6 forces, which almost annihilated the Khitan army. Only a few thousand men managed to return to the Liao border. This was by far the worst defeat suffered by the Khitan during Sheng-tsung's reign, and in consequence Hsiao P'ai-ya was stripped of all his titles and offices and disgraced. In the late summer of 1019 another great army, including many tribal troops, was assembled to attack Kory6. But it was now clear to both sides that neither country could win a decisive victory. In 1020 King Hy6njong sent envoys to pay tribute to the Liao, and Shen-tsung graciously pardoned him, in 1022 sending an envoy to invest him formally as king. The tributary relationship was resumed, and envoys were regularly exchanged. When in 1031 Hy6njong died, his son and successor Wang Hum (T6kchong; r. 1031 — 4) was invested as king by the Liao court. From this date until almost the end of the Liao, Koryo remained a loyal vassal, and peace prevailed between the two states. The Kory6 court did not, however, forget the disasters they had suffered in 1010. While repairing the ravages of this decade of major warfare, they thus set their defenses on a new footing. The rebuilt capital Kaegy&ng was given a stronger outer wall in 1029, and defensive walls were constructed between 1033 and 1044 along the entire frontier from the mouth of the Yalu to Y6np'o on the Tonghae (Sea of Japan). Kory6 was taking no chances with its belligerent neighbor. The Po-hai rebellion, 1029 The end of Sheng-tsung's long reign was marked by the first really serious rebellion of a subject people against the Khitan, the great rebellion of the people of Po-hai. After the flight of its first king, Pei, in 930, the part of the old territories of Po-hai that had been incorporated into the Khitan state, the kingdom of Tung-tan, had been slowly integrated into the Liao administrative system as the Eastern Capital circuit. Much of the old Po-hai administrative structure was abolished, and some settlements of Chinese and Khitan were planted in former Po-hai territory, many of them comprising soldiers from the ordos of various members of the imperial family. But in one important respect the Pohai retained a favored position: For taxation purposes the old Po-hai territoCambridge Histories Online © Cambridge University Press, 2008 FOREIGN RELATIONS 113 ries were treated as a tribute-bearing border state, paying an annual levy of one thousand horses and 150,000 lengths of cloth. IO3 There were no monopoly taxes levied on salt or wine in the area and only minimal levies on commerce. There can also be little doubt that the campaigns against Kory6 had impoverished the Eastern Capital region by their heavy levies of grain and manpower. Then in the 1020s two successive Chinese tax commissioners at the Eastern Capital attempted to extend to this region the tax system enforced in the Southern Capital circuit and to impose further harsh levies of tax and labor service. This seems to have been provoked by conditions in the Southern Capital, which had experienced several years of famine and was suffering from a shortage of food and huge tax arrears. The Po-hai people were ordered to build boats to transport grain to the area around modern Peking. But the journey was dangerous, and many of these boats and their crews were lost. These changes caused widespread resentment. The rebellion was led by Ta Yen-lin, a distant descendant of the old Pohai royal family, who was an army commander at the Eastern Capital. In the eighth month of 1029 he imprisoned the viceroy Hsiao Hsiao-hsien and his wife, killed the offending tax commissioners and the chief military commander, and declared a new dynasty of Hsing-Liao, with its own regnal title. He reported his actions to the Kory6 court and requested their assistance. They refused to give him aid, however, and he remained isolated. Moreover, the Po-hai commander of Pao-chou, a crucial garrison on the Yalu, refused to join in the rebellion and informed the Khitan regional commandant, who executed all the potentially mutinous Po-hai troops under his command. Only a handful of the assimilated Jurchen tribes joined the rebel regime. After being defeated, Ta Yen-lin, realizing that his forces were no match for the Liao armies, retreated to defend his capital. Just a year after the rebellion began, one of his own officers betrayed him, opening the gates of the Eastern Capital to the Liao army. Ta Yen-lin was captured; the remnants of his army were quickly finished off; and his short-lived dynasty came to an end. To obviate any more trouble, the new governor of the Eastern Capital, Hsiao Hsiao-mu, quickly redressed the injustices committed by the Chinese tax commissioners and restored order with a light hand. But the old Po-hai nobility, with the exception of those who had remained loyal, were banished from the Eastern Capital and resettled along the shore of the Po-hai gulf in an area controlled from the Central Capital. Many refugees also fled across the 103 LS, 72, p. 1210. Cambridge Histories Online © Cambridge University Press, 2008 114 THE LIAO Yalu into Kory6 territory, among them not only Po-hai but many Khitan and Hsi as well, who eventually settled in Korea. IO+ THE REIGN OF HSING-TSUNG Sheng-tsung died in the sixth month 1031. He had been on the throne for almost half a century and was now sixty years old. On his deathbed he summoned his trusted ministers Hsiao Hsiao-mu and Hsiao Hsiao-hsien to supervise the enthronement of his successor. His designated heir, his oldest surviving son Tsung-chen (1016—55; Khitan name I-pu-chin, subsequently canonized as Hsing-tsung) succeeded to the throne. Hsing-tsung was still a boy of fifteen, and a regency was clearly necessary, but this raised a complicated problem. Hsing-tsung was not the emperor's son by his legitimate empress Ch'i-t'ien. Although she had borne him two sons, both had died in infancy. In 1016, however, one of his lesser consorts, Nou-chin (entitled Yuan-fei and posthumously entitled Empress Ch'in-ai), a niece of Han Te-jang and a sister of Hsiao Hsiao-mu, gave birth to a son, the future Hsing-tsung, and Empress Ch'i-t'ien adopted and raised him. 10' Although Sheng-tsung's dying testament had ordered the new emperor to preserve Empress Ch'i-t'ien's life, as soon as he succeeded, Nou-chin began plotting to remove her so that she herself could become regent. Accordingly, she had Empress Ch'i-t'ien falsely implicated in plotting rebellion with two of her most powerful supporters, the emperor's uncle Hsiao Pi-tiIo6 and her own son-in-law, the northern prime minister Hsiao Chopu.'° 7 Hsiao Pi-ti and Hsiao Cho-pu were arrested, sent to the Supreme Capital, and executed with many of their relatives and partisans. Purges of their supporters went on for months. Empress Ch'i-t'ien was banished. Shortly afterward Nou-chin sent agents to murder her, and she committed suicide.'08 Nou-chin now had herself entitled empress dowager and formally assumed the regency. Her birthday was declared a public holiday,109 and she was clearly determined to become the real ruler of Liao. On New Year's Day 1032 104 On the Po-hai rebellion, see LS, 17, pp. 203—6; Chong In-ji et al., comps., KorySsa (Tokyo, 1908— 9), 5, pp. 7 1 - 7 3 . 103 Empress Ch'in-ai also bore a second son, Chung-yuan, and two daughters. 106 Hsiao Pi-ti was an orphaned nephew of the former chancellor Hsiao P'ai-ya, who had been raised in the palace and was married to Sheng-tsung's sister. 107 Hsiao Cho-pu, who also appears as Hsiao Cho-pu and Hsiao Tsu-pu-li, was the first husband of the much-married princess Yen-mu-chin. 108 LS, 18, pp. 211 —13; 71, pp. 1202-4; 88, p. 1343. CTKC, 8, pp. 68—9, gives a somewhat different account. 109 LS, 71, p. 1203. Cambridge Histories Online © Cambridge University Press, 2008 THE REIGN OF HSING-TSUNG 115 she held court, received the homage of the emperor and the members of the court, and gave audience to the envoys from Sung. 110 She also lavishly distributed titles and offices to the members of her own family, especially to her younger brothers and their supporters. To complete her dominance, however, Nou-chin needed to replace the young emperor, who was already chafing with resentment at her actions." 1 Although the emperor was her natural son, he had been raised in Empress Ch'i-tien's family and had been closely attached to his adoptive mother. In 1034 the empress dowager, with her younger brothers, planned to dethrone him and replace him with his younger brother Chung-yuan, II2 whom she herself had brought up and who, she thought, would be more pliant to her bidding. Chungyuan, however, wanted no part of this and informed his brother what was being planned. The emperor acted at once, stripped the empress dowager of her seals of office, banished her to Sheng-tsung's mausoleum at Ch'ing-chou, and took over the reins of government himself. Hsing-tsung was not, however, able to eliminate Nou-chin's power completely. Her relatives remained entrenched in many positions of authority. In 1037 the emperor attempted a reconciliation and thereafter began to treat her with great ceremony, regularly visiting her to pay his respects. She never forgave him but, rather, resumed her part in a complex pattern of divided authority. " 3 In 1037 the emperor appointed Nou-chin's brother Hsiao Hsiao-mu as northern chancellor. Indeed, no fewer than five of her brothers subsequently held this post, and members of her family provided the majority of the northern prime ministers until the 1070s. In 1039 the empress dowager was allowed to return to the capital, where she underwent the rebirth ceremony, as Sheng-tsung's mother had done in the 980s, to reestablish her position in the eyes of the Khitan nobility." 4 The Sung court again began to send separate envoys to pay respects to her as well as those to the emperor, a practice they had discontinued when she had been banished."' In the meantime, Hsing-tsung remained on close personal terms with some members of the empress dowager's family. He also had to compensate his brother Chung-yuan for his part in defeating the coup, appointing him to the specially honored position of "Imperial Younger Brother," and thereafter Chung-yuan filled a succession of the highest-ranking posts - commander in no in H2 113 LS, 18, p. 313. CTKC, 8, p. 69. For his biography, see LS, 112, pp. 1 5 0 1 - 3 . In Sung sources his name is wrirten Tsung-yiian. As lace as 1034 she was advocating the succession of brothers, in the Khitan mode, to the Sung envoy Wang Hung-ch'en, whereas Ching-tsung was reasserting hereditary succession in the Chinese fashion. She was still, it would seem, backing Chung-yuan as candidate for the throne. See HCP, 177, pp. 4 2 8 1 - 2 . 114 LS, 18, p. 222. 115 CTKC, 8, p. 7 1 . Cambridge Histories Online © Cambridge University Press, 2008 Il6 THE LIAO chief in 1038, northern commissioner of military affairs, and viceroy of the Southern Capital from 1045 until the end of the reign. The last seems an odd appointment, as the post gave Chung-yuan authority over a largely Chinese population, even though he seems to have been much involved with the "nativist" Khitan interest at court, as we shall see later. The politics at Hsing-tsung's court was thus very complex, with both the empress dowager and, to a lesser extent, Chung-yuan supported by rival groups of kinsmen and allies. The first decade and more of his reign was passed in intricate political maneuvers to establish a balance of power between the emperor and these various groups in both the imperial clan and the consort clan of Hsiao. In general there were no striking departures from those policies laid down under Sheng-tsung that tended to favor Chinese methods of governance. The codification of law was carried an important stage further with the promulgation of the Hsin-ting t'iao-chib, the first formal codification of legislation enacted under the Liao. This was heavily influenced by Chinese models.1'6 Containing 547 articles and compiling together all the laws passed since A-pao-chi's reign, it was promulgated in 1036 and made universally applicable. Centralized legal control was further strengthened in 1046 when local administrations were ordered to report all legal cases annually to the capital.117 The codified laws were further revised in 1051.118 The new code, however, left unresolved many anomalies between Chinese written law and tribal custom. Signs of opposition to the increasing sinification of law and to the favorable treatment enjoyed by the Chinese element of the population under both Sheng-tsung and the new emperor began to appear. In 1044, at the suggestion of the pro-Khitan Chung-yuan, Khitan police inspectors were established at each of the five capitals to protect the interests of the Khitan under the new legal system. "9 In 1046 Khitan were forbidden to sell slaves to Chinese subjects,120 and in 1043 all the Chinese living in the southern parts of the empire were forbidden to own bows and arrows.121 The treatment of the recently rebellious Po-hai people was, on the contrary, relaxed. Under the influence of Hsiao Hsiao-mu, in 1041 the ban against playing polo in the Eastern Capital circuit, which was regarded as a 116 One of the persons ordered to compile it was Yeh-lii Shu-ch'eng, a skilled translator who had translated Chinese medical writings into Khitan. Another was Hsiao Te, an expert on ritual and law. See LS, 89, p. 1349; 96, p. 1400. 117 LS, 19, p. 233. 118 LS, 20, p. 243. 119 LS, 19, p. 230; 112, p. 1502. 120 LS, 19, p. 233. 121 LS, 19, p. 228. Cambridge Histories Online © Cambridge University Press, 2008 THE REIGN OF HSING-TSUNG 117 form of military training, was lifted,122 and the viceroy of the Eastern Capital circuit was specially ordered to commend to the central government able members of his local staff. "3 In 1044 the system of regional administration through the circuits attached to the five capitals was completed when Yiin-chou (modern Ta-t'ung) was elevated to the status of Western Capital, controlling the western part of the territories acquired in 938 and the region of the Yin-shan Mountains north of the great bend of the Yellow River, in what is now Inner Mongolia.124 This territory had formerly been part of the Southern Capital circuit and had a fairly large Chinese population (see Map 7). But the main problem of internal administration during Hsing-tsung's reign was the aftermath of the constant wars of Sheng-tsung's time. It is clear that these decades of warfare had caused much hardship and disruption among the population, and there is some evidence that the Khitan war machine had begun to deteriorate. The most pressing problem was the burden placed on the people, rich and poor alike, by the constant levies of manpower and military service, particularly on the distant western and northern frontiers. Sometime in the late 1030s the emperor asked his ministers for their advice on how to deal with growing distress and impoverishment, internal discontent, and banditry arising from excessive demands for corvee and military service. The court's most prominent Confucian scholar, Hsiao Han-chia-nu, presented a detailed memorial advising the withdrawal of overextended garrisons from distant frontier areas and the cessation of expansionist policies that gained useless territory and no real advantages for the Khitan. He advocated the concentration of military efforts on strengthening the really important frontiers in the south and east."5 Unfortunately, we do not know the emperor's response, but whatever the outcome at the time, similar complaints about the difficulty and expense of providing logistic support for garrisons in remote tribal areas continued until the end of the eleventh century.126 To provide a basis for levying manpower, in 1039 a general census was ordered.127 A registration of the armies and their dependents was ordered in 1046,128 and the military registers were further revised in 1051. 12 ' The army also seems to have been short of horses, once the Khitans' major resource. In 122 123 124 125 126 127 128 129 LS, 19, p. 225. LS, 19, p. 226. LS, 19, p. 231; 37, p. 438. LS, 103, pp. 1446—9; translated in Wittfogel and Feng, pp. 557—9. IS, 104, p. 1455. IS, 18, p. 221. LS, 19, p. 233. LS, 20, p. 243. Cambridge Histories Online © Cambridge University Press, 2008 Cambridge Histories Online © Cambridge University Press, 2008 n8 THE REIGN OF HSING-TSUNG THE LIAO 119 f ruan-cnou j r . » Chiang-sheng , - v . w , h • X * I?™1 chou 1-choul^ - - w e i - c h o u ^ s ^ .T'ung Shen- chou /chou ng-chou EASTERN CAPIT Yin-chou hou ; techou Cheng-i chou Huan-chou Ya-fu*P,Jurchen WESTERN CAPITAL CIRCUIT ASTERN HQoftheSW Pacification Office hou / Kuei-huachou J u . c n ( J U \ Te-chou' Feng-sheng chou . ' , Tung-shengchou Hung ' § £ t T &" ,. . WESTERN. U » ,r m S in-»u \ CAPITAL Yu-chou. r ' c h o - SOUTHERN .Ying-chou ' cnou /l-chou Shuo-chou ^ - ^ u-chou / " 2-u,'V'»v».*-' \r'"*c SOUTHERN CAPITAL CIRCUIT MAP 7. The Liao empire, ca. 1045 Cambridge Histories Online © Cambridge University Press, 2008 I2O THE LIAO 1043, therefore, the sacrifice of horses or oxen in funerary ceremonies was prohibited,130 and in 1048 commissioners were sent out to register horses.1'1 There were also misgivings about the troops' training, particularly the Chinese units. In 1035 the armies were enjoined to supervise the regular training of their catapulteers, crossbowmen, archers, and swordsmen.1'3 In 1046 the emperor watched the exercises of Chinese troops while they practiced using catapults and bows,1'' but serious concern about the inferior skills of the Chinese armies' catapulteers and crossbowmen continued through the next reign.1'4 These were skills that were irrelevant to the Khitans' traditional mobile cavalry warfare but essential to their warfare with their sedentary Chinese and Korean neighbors. Such a concern suggests that the Khitan generals had learned the lessons of the wars of Sheng-tsung's time, when the Liao armies had never succeeded in taking a fortified place of any size. Foreign relations under Hsing-tsung One of the injunctions that Sheng-tsung issued to his heir was to maintain intact the peace treaty with the Sung, 1 " which was the most important factor in dynastic stability. The regular procession of envoys between the two courts thus continued uninterrupted. The relationship with Kory6 also remained cautiously amicable. The central problem in foreign relations had shifted to the southwest, where the Hsi Hsia state had rapidly grown in power. Deeply involved with both the Khitan and the Sung, the Hsi Hsia threatened to engulf the whole northern frontier in warfare. During the early part of the eleventh century the Hsia had paid tribute to both the Khitan and the Sung, playing the great powers off against each other to secure their own independence and to gain the opportunity to enlarge their own territory and influence. Their ambitions were centered in the west, where they steadily expanded at the expense of the Uighurs (see Chapter 2). This westward expansion brought them into conflict with the Khitan, with whom they competed for control of the trade routes to the west, rather than for territory. At this time there were three independent regimes in what is now western Kansu. Liang-chou was under Tibetan warlords. In Kan-chou was a Uighur state, dominating the central part of the Kansu corridor, and farther west, 130 13: 132 133 134 135 LS, 19, p. 228. LS, 20, p. 239. LS, 18, p. 217. LS, 19, p. 232. LS, 26, p. 308. This is dated 109;. CTKC, 7, p. 66. Cambridge Histories Online © Cambridge University Press, 2008 THE REIGN OF HSING-TSUNG 121 centered at Tun-huang, was the state of Sha-chou, ruled by the Chinese warlord family of Ts'ao. This latter seems also to have passed under Uighur dominance during the early eleventh century. In 1006 the Sha-chou ruler Ts'ao Shou had sent a tribute mission to Sheng-tsung's court, which seems to have encouraged Sheng-tsung to attack the neighboring Uighur state in Kanchou. Expeditions were sent against Kan-chou in 1008, 1009, and 1010. Although the Khitan achieved some limited success, sacking Su-chou and deporting its population in 1010, these campaigns failed to conquer the area, which was at the same time under attack from the Hsia. Another expedition in 1027 laid Kan-chou under siege but failed to take the city, and it ended in disaster when the retreating Khitan army was ambushed in southwest Mongolia by the Tsu-pu tribes, who always strongly resisted Khitan attempts to expand to the west. The Hsia, meanwhile, continued with their gradual incorporation of the area known as Ho-hsi (western Kansu). In 1020 they built a new capital city at Hsing-chou (modern Yin-ch'uan, later renamed Hsing-ch'ing). By 1036 they occupied the Kan-chou area, and although Sha-chou remained at least semiautonomous until the 1050s, the Hsia claimed that even Khotan, deep in the Tarim basin, was its vassal. In 1038 the Hsia king Li Yiian-hao proclaimed himself emperor of the state of Ta Hsia and sent an embassy to K'ai-feng repudiating, in a provocative and insolent letter, his vassal relationship with the Sung. The new Hsia emperor had been married to a Khitan princess in 1031 shortly after Hsing-tsung's accession, but the royal couple did not agree, and early in 1038, when the princess died, the Khitan court sent an envoy to inquire into the circumstances of her death.1'6 For some strange reason the Liao shih makes no mention of Yuan-hao's assumption of the imperial title or the Khitan court's response. The Sung, meanwhile, took strong action. They first revoked all the titles that Yuan-hao held from the Sung court, placed a reward on his head, and suspended the border trade on which the prosperity of the Tanguts heavily depended. Gradually the relations between the Sung and the new Tangut state deteriorated, until in 1040 the border clashes escalated into full-scale war. Fighting dragged on spasmodically until 1044, broken by a series of negotiations. In spite of all their efforts the Sung armies suffered a succession of major defeats, the Tanguts proving themselves formidable adversaries. At the beginning of this confrontation the Liao court did not intervene, receiving embassies from both sides, each of whom reported their progress in the war. That the Sung were in serious trouble soon became clear, however, and after a particularly disastrous Sung defeat in 1042, the Khitan decided to 136 LS, 18, p. 220; 115, p. 1526. Cambridge Histories Online © Cambridge University Press, 2008 122 THE LIAO put pressure on them for territorial concessions in the long-disputed area of Kuan-nan. This diplomatic pressure led to a peaceful resolution. In 1042 the Khitan dropped their territorial claims in return for an increase in the annual subsidy from the Sung court to 200,000 taels of silver and 300,000 bolts of silk. The treaty confirmed the "fraternal relations" established in 1005, and apparently the Khitan promised the Sung that they would attempt to presuade Yiian-hao to conclude a peace agreement with K'ai-feng. The Sung court believed the Hsia and the Liao to be close allies. But this alliance was largely imaginary: When in 1043 Yiian-hao requested that the Khitan join him in attacking Sung, Hsing-tsung refused, and gradually a serious rift developed between the Hsia and the Khitan. For some time, Yiian-hao had shown willingness to conclude a peace agreement with the Sung, but the terms he asked were unacceptable, and his communications were considered insulting. Negotiations dragged on for two years and finally came to a conclusion after the Sung court learned that hostilities had broken out between the Hsia and the Khitan and that a first Khitan punitive expedition in 1044 had failed. The war between the erstwhile allies broke out over some Tangut tribesmen living inside the Liao border who rebelled and sought refuge in Hsia territory in early summer 1044. The Liao blamed Li Yiian-hao for having instigated the rebellion of these tribes, though they had almost certainly caused the trouble themselves by their plans to fortify their Hsia border and to ban the horse trade on which the tribesmen depended. When the rebellion erupted in 1044, the Liao immediately sent a punitive expedition into Hsia territory. The Sung court seized their opportunity to conclude their own negotiations with Hsia. In the winter of 1044 a treaty with Hsia was signed, under which the Hsia ruler accepted the status of a Sung vassal. In return the Sung agreed to pay the Hsia an annual subsidy similar to that paid to the Khitan, though on a smaller scale: 72,000 ounces of silver and 153,000 bolts of silk, plus a large quantity of tea annually. Border trade was also regularized, and markets were opened to the Tanguts. But the treaty contained one major flaw: Unlike that concluded with the Liao, it failed to define the border between the states, and consequently there were constant boundary disputes and outbreaks of warfare for the next seventy years. Although Sung had taken advantage of the outbreak of war between the Khitan and the Hsia, the Khitan invasion of 1044 had not gone well. After the initial invasion ended in heavy defeat and the loss of two of the Khitan commanders, a major force was mobilized on the frontier. In the ninth month the emperor's brother Chung-yuan and the northern commissioner for military affairs Hsiao Hui were given command of the Liao vanguard armies for a full-scale invasion. Li Yuan-hao now sent envoys and even came in Cambridge Histories Online © Cambridge University Press, 2008 THE REIGN OF TAO-TSUNG 123 person to try to make peace with the Liao emperor. But Hsing-tsung's counselors were so confident of victory that they persuaded the emperor to reject the offer of peace and to settle the matter on the battlefield. It was a disastrous decision. The Liao army was overwhelmed by the Tanguts in a pitched battle in the Ho-lan Mountains west of the Hsia capital. Many highranking Khitan, including the emperor's brother-in-law, were captured by the Tanguts. Hsing-tsung was now forced to accept Yuan-hao's earlier offer of submission, and peace was restored. The defeat upset the emperor and the court alike. In 1048, with the death of Li Yuan-hao, the Hsia throne passed to an infant prince, and the Tangut court was enfeebled by a bitter power struggle. The Khitan saw their opportunity for vengeance, and in the autumn of 1049 a large Liao army invaded Hsia and advanced on the capital in three separate columns. One group under Hsing-tsung's personal command encountered little resistance but was forced to retreat owing to a lack of water and pasture for its horses. Another column led by the emperor's brother-in-law Hsiao Hui advanced south along the Yellow River, supported by a flotilla of warships and supply vessels, but was ambushed and seriously defeated with very heavy losses, Hsiao Hui barely escaping with his life. The third column was slightly more successful. It raided a separate palace built by Yiian-hao in the Ho-lan Mountains for his favorite consort. The young widow and several high-ranking Tangut officials were captured, but this was small consolation for the heavy losses elsewhere. A second campaign in 1050 was somewhat more successful. Liao forces plundered the Hsia countryside and accepted the surrender of a Tangut general. The mother of the infant Hsia emperor petitioned the Liao court, asking to resume tributary relations, and a formal tribute embassy was sent to Liao in 1050. However, several years elapsed before the two sides finally came to terms. Liao finally dropped its demand for the repatriation of the Tangut tribesmen who had rebelled in 1044, in exchange for compensation, and so the Tanguts agreed to deliver annual tribute. The Liao kept Yuanhao's widow and retained one of the captured border fortresses. In 1053 peaceful relations were finally restored, but the relationship between the two states remained extremely cool for several decades. THE REIGN OF TAO-TSUNG In 1055 Hsing-tsung, still only thirty-nine years old, fell sick and died on one of his constant progresses. He left an empire at peace with its major neighbors. In the previous year he had granted a high-ranking office to the crown prince of Kory6, at the king's request. The king of Hsia had sent envoys requesting a dynastic marriage with a Khitan princess, and another Cambridge Histories Online © Cambridge University Press, 2008 T H E 124 LIAO bearing a solemn oath of amity. At the beginning of 1055 Hsing-tsung had received the customary embassy from Sung, which presented him with two trained elephants, and more envoys from Hsia. Internally, too, Hsing-tsung's empire was relatively peaceful and prosperous, though relations with his mother, the empress dowager, remained strained. His brother Chung-yuan, once put forward as a candidate for the throne, had been long entrenched as governor of the Southern Capital and had just had a son.'37 Hsing-tsung was succeeded by his eldest son Yeh-lii Hung-chi (10321101; Khitan name Nieh-lin or Ch'a-la, subsequently canonized as Taotsung), who for the past three years had worked together with his father in conducting the routine business of government.li% Hsing-tsung clearly intended that his son should not only succeed him but that when he did so he should also be prepared to rule. Tao-tsung ascended the throne "before his father's coffin," issued a modest and conciliatory edict to his court, and began a complex round of ritual acts and sacrifices at the various ancestral tombs and shrines. Embassies were exchanged, as custom dictated, with Kory6, Hsia, and Sung. One of the ambassadors from Sung was the famous scholar and historian Ou-yang Hsiu. The international order established under Taotsung's father thus continued undisturbed. There was no immediate dispute over the succession. The empress dowager Ch'in-ai had remained resentful about Hsing-tsung's having excluded her from a dominant role in politics in 1035 and, even after his death, could not even pretend to mourn for him, chiding his lamenting widow, "You're still a young woman — why distress yourself like this?" Only a year before she had told a Sung ambassador that she favored the true Khitan form of succession between brothers, not hereditary succession from father to son.1*' She would clearly have supported the claims of Chung-yuan, whose title, Imperial Younger Brother, carried in a Khitan context an implied claim to the throne and whom she had once plotted to enthrone in Hsing-tsung's place. Hsing-tsung himself had clearly appreciated the danger that Chung-yuan might succeed him. In 1054 he remarked to the Sung ambassador Wang Kung-chen, "I have a recalcitrant younger brother. If one day he acquires my kingdom, I fear that the Southern Court [i.e., the Sung] will not be able to sleep quietly at night!"'40 But in any case, the empress dowager took no action, even though her brothers and relatives held high positions and could certainly have influenced the succession. Chung-yuan was given a new dig137 138 139 140 LS, 20, p. LS, 2 1 , p. HCP, 177, HCP, 177, 247. 251. p. 4282. pp. 4282—3. Cambridge Histories Online © Cambridge University Press, 2008 THE REIGN OF TAO-TSUNG 125 nity, as imperial great uncle, and was granted unique personal privileges. The emperor did not address him by his given name, nor did Chung-yuan do obeisance. But having been accorded these unusual marks of respect, he was sent back to the Southern Capital to govern its people in the name of the new emperor. Empress Dowager Ch'in-ai was given the title of grand empress dowager (t'ai-huang t'ai-hou), and Tao-tsung's mother, Empress Jen-i, became empress dowager.141 Chung-yuan was appointed commander in chief (ytiyiieh) in 1056, and the grand empress dowager fell ill and died at the end of 1058. I42 In the early years of Tao-tsung's reign, two men, Hsiao Ko 14 ' and Hsiao A-la,' 44 held great influence at court. The latter was a son of Hsiao Hsiao-mu and thus was a member of the still extremely powerful clan of the empress dowager Ch'in-ai. Hsiao A-la had been brought up in the palace and had become a close friend of Hsing-tsung, whom he had served as a state counselor and as viceroy of the Eastern Capital. He also was married to a royal princess. On Tao-tsung's accession he was made northern commissioner for military affairs, and so he now shared power at court with the opportunistic Hsiao Ko. The two soon fell out. Hsiao A-la asked to retire but was sent from court to be viceroy of the Eastern Capital, probably in 1059. In 1061 he came to court at the time of a great ancestral ceremony 14 ' and offered harsh and trenchant criticism of the conduct of the government. Hsiao Ko, the target of these criticisms, denounced him to the emperor, and in spite of the empress dowager's pleas, the emperor ordered Hsiao A-la to be strangled. The Liao shih, which praises A-la highly for his unswerving loyalty and political understanding, speculates that had he not been executed, neither Chung-yiian's rebellion nor the later execution of the empress and murder of the heir apparent would have occurred. Be this as it may, the killing of A-la was a grave political error and the first serious instance of Tao-tsung's bad judgment and abiding weakness, his never-failing willingness to believe slanderous accusations even when they were leveled at those closest to him. For the time being the court was left in the hands of the self-seeking Hsiao Ko (who retired in 1062) and Yeh-lii Jen-hsien and Yeh-lii I-hsin. During these years, the court was riven by intense personal rivalries that the emperor was too weak to mediate or resolve. Moreover, basic problems persisted, chief among them the continued tension caused by the progressive sinification of the Khitan state and the imposition of central authority over 141 142 143 144 145 LS, 21, p. 252; 7 1 , p. 1204. LS, 21, p. 236. LS, 113, pp. 1 5 1 0 - 1 1 . LS, 90, p. 1355. According to LS, 90, p. 1355, the occasion was the rain ceremony (u-ie //'); according to LS, 113, p. 1511, it was the southern suburban sacrifice. The latter seems more likely. Cambridge Histories Online © Cambridge University Press, 2008 126 THE LIAO what had traditionally been tribal matters. The long-established tendency toward "rationalization," centralization, and Chinese influence continued, even though at first the new emperor was at pains to establish something of a collegial atmosphere at court, where his officials and Khitan dignitaries would feel free to offer unconstrained advice.'46 Symbolic of these tendencies, in 1055 all officials, not, as hitherto, only the emperor and members of the Chinese Southern Administration, were required to wear Chinese court dress at major ceremonies.147 Both the new emperor and his empress were highly cultivated and well educated in Chinese, and they wrote poetry. He was deeply interested both in Confucian learning and in Buddhism. Perhaps the most striking evidence of Tao-tsung's inclinations in favor of Chinese culture and institutions is the increased emphasis he placed on Chinese-style education for potential officials and on the examination system. The numbers of chin-shih passed at each examination (these were usually held at about four-year intervals) leapt from fifty or sixty under Hsing-tsung to well over a hundred.'48 In 1059 the state educational system was reformed by the establishment of prefectural and county schools,"»» as well as colleges for higher learning in the five capitals and the superior prefectures of Huang-lung and Hsing-lung.'5° In 1060 a second imperial academy {Kuo-tzu chien) was set up at the Central Capital,'5' in addition to the academy at the Supreme Capital which had been established under A-pao-chi.1'2 Finally, ritual observances were ordered in honor of various Confucian sages and teachers.'" The emperor became personally involved in the conduct of the examinations. In 1070 a new special palace examination entitled the "Examination for the worthy and good" (Hsien-liang k'o) was established, the entrants for which had to submit 100,000 words of their writings.1'4 In 1072 Tao-tsung himself wrote the questions for both the chin-shih and the palace examination for the worthy and good. 1 " Tensions over the application of Chinese-style law on the Khitan subjects of Liao surfaced again early in the reign. In 1058 tribal judges (i-li-pi) were ordered to refer their decisions in all capital cases to the local prefect or 146 For an analysis of the political polarization of Tao-tsung's reign, see Ch'en Shu, Ch'i-tan chtng chih sbih kao (Peking, 1986), pp. 137-52. 147 LS, 56, p. 908. 148 Wittfogel and Feng, p. 492 (table). 149 LS, 48, pp. 807, 8 1 1 , 817—21. 150 LS, 48, pp. 807, 811, 820. 15: LS, 48, p. 807. 152 LS, 47, p. 788. 153 LS, 2 1 , p. 258. 154 LS, 22, p. 269. 155 LS, 23, p. 275. Cambridge Histories Online © Cambridge University Press, 2008 THE REIGN OF TAO-TSUNG 127 magistrate for review. Sentence was not to be carred out unless the prefect certified that there had been no miscarriage of justice. If anyone claimed the sentence was unjust, it had to be referred to the central government for a decision. 1 ' 6 Once again tribal justice was placed under the scrutiny of a Chinese-style local administration. Chung-yuan had, as we have seen, interceded on behalf of the Khitan interest in similar circumstances in 1044. It may well have been an antiChinese nativist faction that in 1063 led him to attempt a coup d'etat. The revolt of Chung-yuan in 1063 The sources, which give confused, fragmentary and contradictory accounts of this affair,157 suggest that Chung-yiian was not the prime mover but that he was dragged into a conspiracy as a figurehead by his ambitious son Nieh-lu-ku, who had been appointed southern commissioner for military affairs in 1061, and a group of dissident noblemen. Chief among these were Yeh-lii T'ieh-pu, a grandson of Sheng-tsung, and Hsiao Hu-tu, the current husband of Hen-muchin, the daughter of Sheng-tsung and Empress Ch'in-ai. Hsiao Hu-tu was commissioner for military affairs in the Northern Administration. 1 ' 8 Nieh-lu-ku at first decided that his father should feign illness so that the emperor would come to visit him and that they should then take the opportunity to assassinate Tao-tsung. When in the early autumn of 1063 the emperor went on a hunting trip to the T'ai-tzu Mountains southwest of the Central Capital circuit (near modern Ch'eng-te), the dissidents seized their opportunity. They led a detachment of crossbowmen to ambush the emperor and attack his encampment. Tao-tsung could not at first believe that there had been a rebellion, although the empress dowager had been forewarned of the conspiracy by a loyal member of the Yeh-lii clan. 1 ' 9 But in the first encounter Tao-tsung was wounded and had his horse shot under him. Some loyal servants saved him, and his mother, the empress dowager, took command of part of the guard and helped ward off the attackers. Nieh-lu-ku, however, was killed by a stray arrow during the initial assault. One of the coconspirators, Yeh-lii Sa-la-chu, who was a member of the emperor's entourage, led a group of disaffected huntsmen to assist the rebels. But the rebels had already lost. Their real leader was dead, and they had let slip their initial 156 LS, 21, p. 256. 157 LS, 22, p. 262; 64, pp. 988—9; 112, p. 1502; 114, p. 1514 (translated in Wittfogel and Feng, p. 421); CTKC, 9, p. 88. 158 Wang Ting, Fin chiao lu (Pref. 1089), in Pao-yen t'angpi chi, ed. Ch'en Chi-ju (1606; repr. Shanghai, 1922), pp. 2b~3a, claims chat Chung-yiian was also incited by his consort, who had had a quarrel with Tao-tsung's young empress. 159 LS, 71, p. 1204. Cambridge Histories Online © Cambridge University Press, 2008 128 THE LIAO opportunity to take the royal camp by surprise. Although they proclaimed Chung-yuan emperor, when dawn came and the battle was joined in earnest, they were completely routed. Chung-yuan fled southward and committed suicide in the wilds, cursing himself for having listened to his son's plan and becoming involved in the debacle.'60 Hsiao Hu-tu, wounded in the battle, also fled and drowned himself. Yeh-lii Ming, the viceroy of the Southern Capital, had been a party to the plot and, when he received news of Chung-yiian's defeat, led a force of Hsi troops into the capital city and armed them, planning to join in the uprising. But his deputy governor mobilized the Chinese garrison to resist them, and when urgent orders arrived from the emperor, he arrested and executed Yehlii Ming. A group of envoys who were on a mission to the Sung court at the time of the uprising were also involved in the conspiracy, and as soon as they returned to Liao territory, they were arrested and sent to the capital for execution. It seems clear that this was not simply another of the recurrent opportunistic attempts to seize power by rival members of the royal family or the result of a simple family feud. Rather, it resulted from a wide-reaching conspiracy by several important and powerful people. What precisely were their motives it is impossible to say with certainty, but it seems most likely that they rebelled to prevent further encroachments on the privileges of the Khitan tribal nobility. Whatever the reasons for the uprising, Tao-tsung's reaction was swift and drastic. All the conspirators were executed together with their immediate families, among them Hsiao Ko, whose son was married to Chung-yiian's daughter. There were extensive changes among the holders of the highest offices. It was not to be the last serious incident involving the royal clan in Taotsung's long and generally peaceful reign. The dominance of Yeh-lii l-hsin and his faction Even before Chung-yiian's rebellion, power at the Liao court had been gradually shifting into the hands of a group of officials led by Yeh-lii I-hsin (d. 1083; Khitan name Hu-tu-kun). ' 6l A member of the Five Divisions (wu yuan pu) lineage of the royal clan, I-hsin had risen from an impoverished youth to become a palace attendant under Hsing-tsung and, by the end of that reign, to become a guard commander. Under Tao-tsung he was given further favors 160 CTKC, 14, p. 153, and Wang Ting, Fen chiao lu, p. 3b, claim incorrectly that he was executed. 161 For his biography, see LS, n o , pp. 1483—6. Cambridge Histories Online © Cambridge University Press, 2008 THE REIGN OF TAO-TSUNG 129 and in 1059 was appointed southern chancellor and then transferred to the northern Chancellery in the same year. I-hsin's rapid rise to power coincided with the decline of the junior patriarchal line (shao fu chang) of the Hsiao consort clan, which had held great influence ever since Hsiao P'ai-ya became northern prime minister in 1005. This office had subsequently been almost monopolized by members of the junior patriarchal line, and during Hsing-tsung's reign their power had been underpinned by the empress dowager Ch'in-ai. Her death in 1058 marked the real end of their dominance, although some individual members of the lineage continued to fill high positions. Shortly before Chung-yiian's rebellion, I-hsin had cemented a personal alliance with another powerful man, the then southern chancellor Yeh-lii Jenhsien (1013-72; Khitan name Ch'a-la),' 62 who had risen to prominence during the negotiations with the Sung in 1042. Jen-hsien had long been one of the principal opponents of Chung-yuan and his faction, who in 1060 tried to engineer his removal from the capital by having him appointed commander in chief on the troublesome northwestern frontier so as to give them a free hand at court. I-hsin managed to foil this plan by interceding with Taotsung, and later he and Jen-hsien played crucial roles in defeating the rebellion. 16 ' After the rebellion they jointly controlled the Northern Commission for Military Affairs for a while, and in 1065 Jen-hsien became commander in chief. For the next fifteen years I-hsin dominated the court and the emperor. His admittedly hostile biography suggests that although he came to exercise unrivaled influence, he was essentially a self-interested opportunist, selecting worthless and corrupt men for office, taking bribes, and allowing the military to do as they pleased. Only the empress's family refused to be dominated by him. Even Yeh-lii Jen-hsien, who at first tried to restrain some of his more outrageous presumptions, found himself endangered and willingly accepted the post of viceroy of the Southern Capital, where he proved a model governor. '6* It is difficult to form a balanced picture of I-hsin's regime. The sources for the period are virulently hostile to both him and his adherents. In the Liao shih he and his faction are given a group of unrelentingly negative biographies under the rubric "evil ministers."' 6 ' But these were based largely on the veritable record written at the command of Tao-tsung's grandson and 162 163 164 165 For his biography, see IS, 96, pp. 1395-7. IS, n o , p. 1484:96, pp. 1396—7. IS, 96, p. 1387. IS, n o , i n . See especially the preface to these chapters (p. 1483) and the historian's comment (p. •495)- Cambridge Histories Online © Cambridge University Press, 2008 130 THE LIAO successor T'ien-tso,'56 who had good reason to hate I-hsin and execrate his memory, as we shall see. Clearly, I-hsin's dominance was an interlude that aroused violent passions, the memory of which was still fresh well into the twelfth century. What does seem clear is that I-hsin's rise to power did not mean any clear change in policy or a break with the centralizing, pro-Chinese trends that had continued since Hsing-tsung's reign. Nor was there an end to the tensions between the centralist, pro-Chinese and the "nativist," pro-Khitan interests that underlay Chung-yiian's rebellion. In Tao-tsung's time, however, these conflicting interests did not result in factions of different racial composition. Some of I-hsin's supporters were ethnic Chinese, as were some of his opponents. It is difficult to perceive the real interests underlying the personal hatreds and alliances that divided the court. The emperor was somehwat aloof from these political struggles and also was weak, suspicious, and easily influenced, always ready to heed slanderous accusations of disloyalty. He provided no real political leadership but pursued his own interests. Tao-tsung was a naturally scholarly man: In 1064 he ordered a search for books lacking in the imperial collections,'67 and he continued to take a deep personal interest in the examination system, which was further developed in these years. Tao-tsung also developed an interest in history. In 1074 the government distributed copies of the Shih chi and Han shu,l68 and about this time a bureau for compiling a national history for the Liao was established, which in 1085 produced veritable records {shih lu) for the first seven reigns.'69 The emperor summoned prominent scholars to expound various canonical texts before him, and he himself wrote poetry and prose. He also continued to honor Buddhism,'70 in some cases overriding the hostility of local officials,'7' and gave lavish patronage to various monks, as did the empress. The nativist Khitan resistance to Chinese influences did not, however, wither away after Chung-yiian's abortive rebellion. Nor were the Khitan nobility weakened; in 1069, for example, the government was forced to issue orders forbidding members of the royal clan from taking advantage of their 166 This veritable record was commissioned by T'ien-tso and completed by Yeh-lii Yen in 1103. The Ch'i-tan kuo ehih, which the Liao shih compilers also used as a source, passes over the events of Taotsung's reign in almost total silence, and what it does say is full of errors. The equally hostile Fen chiao lu was written by Wang Ting (d. 1106) in 1089. Its survival is due to T'ien-tso's releasing the author from banishment to a remote frontier district, where he had been kept as a result of personal animus by Tao-tsung. See LS, 104, p. 1453. 167 LS, 22, p. 264. 168 LS, 23, p. 276. 169 LS, 24, p. 290; 104, p. 1456. 170 See the sources collected in Wittfogel and Feng, pp. 304—7. 171 LS, 24, p. 284. Cambridge Histories Online © Cambridge University Press, 2008 THE REIGN OF TAO-TSUNG 131 power to exploit the common people. I72 Tao-tsung was forced to take measures to placate the Khitan interest. For example, in 1067, although he was then engrossed in his Buddhist studies, Tao-tsung underwent the traditional "rebirth" ceremony, to reestablish his legitimacy as the leader of the Khitan nation.'" A number of significant measures began to place restrictions on the Chinese population. In 1070 they were forbidden to engage in hunting, which was regarded as a form of military training. 174 In 1064 the private publication of books was forbidden,175 a measure that can have had an impact only in those regions with an educated Chinese elite. In io63, 1 7 6 1064, 177 and 1070, 178 prohibitions were placed on the sale of metals, especially copper to the Hsia and iron to the ever-troublesome Tsu-pu tribes and the Uighurs. 179 In 1064 a ban was placed on the manufacture in the Southern Capital circuit of patterned satins, as part of sumptuary legislation aimed at its wealthy population. 180 Finally in 1070 Tao-tsung, "because he considered that the customs of the Khitan and Chinese were different and that the national laws should not be applied indiscriminately," ordered I-hsin and Yeh-lii Su to revise the laws yet again.181 This revision was a reversal of earlier attempts to produce a universally acceptable code heavily influenced by Chinese patterns. The new laws attempted to define and preserve the traditional differences between Khitan and Chinese customs. The resulting codification was almost twice the size of the Hsin ting t'iao chih of 1036, and further amendments and supplements were added between 1075 and 1085, until the code extended to well over one thousand articles. These new laws, attempting as they did to combine both Chinese and Khitan customary laws in a single codification, were so complicated and so far out of step with actual practice that they proved unenforceable. Finally in 1089 the new laws were abandoned, and the codification of 1036 was reinstated, which remained the basic body of law until the fall of the dynasty. l8s Another sign of the changing times came in 1074, when the scholarly 172 173 174 175 176 177 178 179 180 181 182 LS, 22, p. 268. LS, 22, p. 267. LS, 22, p. 270. LS, 22, p. 264. LS, 22, p. 262. LS, 22, p. 264. LS, 22, p. 270. LS, 22, p. 270. LS, 22, p. 264. LS, 62, p. 945. LS, 62, pp. 945— 6; translated in Herbert Franke, "The 'Treatise on punishments' in the Liao history," Central Asiatic Journal, 27 (1983), pp. 9 - 3 8 . Cambridge Histories Online © Cambridge University Press, 2008 132 THE LIAO official Yeh-lii shu-chen (d. 1082) suggested granting Chinese-style surnames to all the Khitan tribes, implying the adoption of the Chinese principle of surname exogamy. But the emperor rejected this proposal out of hand, declaring that "the old order should not be changed suddenly. "l83 It is clear that the differences and tensions between Khitan and Chinese, or perhaps between centralizing and pro-tribal interests, continued to influence politics and that the pro-Chinese measures of the preceding half century had reached a turning point. Natural disasters Another factor that certainly affected the government during much of Taotsung's reign was the constant threat of natural disasters and famine. From 1065 until the end of the reign, barely a year went by without some area of the Liao empire being struck by some natural catastrophe.184 At first these problems mainly affected the southern agricultural regions; later, in the 1080s and 1090s the tribal areas, too, seem to have suffered greatly. These disasters are usually recorded either because they required the payment of relief to the local population or because they forced the government to grant tax exemptions. The government must have constantly lost revenue as well as faced the great expense of providing relief. There was also widespread suffering and reports of great numbers of displaced families and vagrants. But the government could do little. Even when an opportunity arose, as with the disastrous floods in the Eastern Capital circuit in 1074, after which orders were given to build flood control works, the suggestion was opposed on the grounds that the hardship caused by the necessary levies of labor would outweigh the benefits.l8' It is impossible to give any precise estimate of the impact of these natural disasters. We have no reliable basis even for estimating the Liao population at this period. It is possible that as in Sung China during the eleventh century, there had been a large increase in population and that a crisis of subsistence was in the making. Certainly, Sung territories in northern China also suffered similar series of natural calamities, especially the devastating locust plagues that affected the whole region in the 1070s and 1080s. One disaster, however, hit the tribal population most cruelly. In the bitter winter of 1082-3 an unusually heavy snowfall killed enormous numbers (the sources perhaps exag183 LS, 89, p. 1350. 184 For a collection of some of the source material, see Witttfogel and Feng, pp. 389—95. 185 LS, 105, p. 1460. Cambridge Histories Online © Cambridge University Press, 2008 THE REIGN OF TAO-TSUNG 133 gerated by putting the number at 60 or 70 percent) of livestock and horses, the main form of wealth among the nomads.' 86 The execution of the empress and the murder of the heir apparent In 1072 Yeh-lii Jen-hsien, the only man whose influence in any way rivaled I-hsin's, died. In 1075 the heir apparent, Prince Chun (1058-77; Khitan name Yeh-lu-wo, later canonized as Shun-tsung, although he never reigned), who was both a model Khitan warrior, skilled as a horseman and archer, and a very intelligent and well-educated young man, began to participate in court affairs and was given charge of some of the business of the northern chancellery.'87 Yeh-lii I-hsin perceived in him a potential threat to his own longestablished dominance and influence over the emperor. As a first step he decided to bring about the downfall of the prince's mother, Empress Hsiian-i. l88 In 1075 t n e empress, who was a highly cultivated and well-educated lady, a poetess and a fine musician, was falsely accused by a palace slave and a minor member of the office of music of having had a sexual liaison with a musician member of her household entourage called Chao Wei-i. Yeh-lii I-hsin reported this slander to the emperor, and although Chao Wei-i strenuously denied the charge even under torture, I-hsin and his ally the learned Chinese scholar Chang Hsiao-chieh apparently fabricated evidence in the form of some erotic poems alleged to have been written by the empress to Chao Wei-i. Chao was executed together with his whole clan. The empress was ordered to commit suicide, and her corpse was returned to her family wrapped in a mat as a public mark of disgrace.' 89 Having removed the empress, I-hsin had her replaced by the sister of one of his henchmen, Hsiao Hsia-mo. This woman, usually referred to by her 186 IS, 14,, p. 288. 187 LS, 23, p. 277172, p. 1215. 188 Hsiian-i was the posthumous title given to her in iooijseeLS, 27, p. 318. Her title at the time was I-te. For her biography, see LS, 71, p. 1205. Her epitaphs are included in Chin Yii-fu, Liao lingsbih k'o chi lu (1934); repr. as vol. 1 of Liao Chin Yiian yii wen chin ts'un lit (Taipei, 1974), pp. 8 b - i o a . 189 LS, 62, p. 945; 71, p. 1205; 23, p. 277. The affair is described in dramatic detail in the only historical work surviving from the Liao, the Fen chiao lu of Wang Ting (Pref. 1089). This claims to have been based partly on eyewitness recollections. Although the Ssu-k'u editors commend its value as a supplementary source — see Chi Yiian et al., comps., Ssu-k'u ch'iian shu /sung mu t'iyao (Shanghai, 1931; repr. Shanghai, 1934), 52, pp. 1154-5 - it is obviously written from a violently partisan viewpoint. The Cb'i-tan kuo chih's biography of the empress, which is wrong in almost everything it says about her, does not mention her forced suicide. For a detailed study of this incident, see Yao Ts'ung-wu, "Liao Tao-tsung Hsiian-i huang hou shih hsiang tz'u yiian yii te wen hua te fen hsi," Wen sbib che hsiieh poo, 8 (1958), pp. 97—134. Holmgren, "Marriage, kinship and succession," suggests (p. 80) that this incident was a struggle between rival wives in Tao-tsung's harem. But this idea is not supported by the evidence cited. The only other wife mentioned in LS, 7/, pp. 1205—6, entered the harem only after the death of the Hsiian-i empress. Cambridge Histories Online © Cambridge University Press, 2008 134 THE later title Hui-fei,'9° was recommended to Tao-tsung and was introduced into the imperial harem. In 1076 another of I-hsin's enemies, the empress dowager, died. A few days after her death the new consort, Hui-fei, was formally installed as the empress.19' Hui-fei was expected to give I-hsin direct influence inside the palace and to ensure his family's future, for her younger sister was married to one of his own sons.192 With the empress Hsiian-i and the empress dowager dead, I-hsin was forced to move swiftly against the heir apparent, who had sworn revenge on him for the murder of his mother and who enjoyed the sympathy of many of the court and the people, who saw the empress's death as a flagrant injustice. Shortly after her suicide I-hsin had survived a bungled attempt at his assassination'93 and was well aware that once the heir apparent succeeded to the throne, he and his partisans would soon be brought to justice. Once again I-hsin resorted to slander and subterfuge. In the fifth month of 1077 he had a false report presented to the emperor alleging that a number of officials, all of them his own enemies, were conspiring to dethrone Tao-tsung and place the heir apparent on the throne in his place. The emperor found the accusation unsubstantiated, although for the sake of prudence he assigned the accused officials to provincial posts. I-hsin now suborned some minor palace officers to confess falsely to having been involved in the conspiracy, in order to have the case reopened. Even though the heir apparent was beaten and interrogated by I-hsin's henchmen, he denied all the charges, pointing out that because he would eventually succeed to the throne anyway, he would have nothing to gain. The interrogators thereupon falsified the evidence and claimed that he had confessed. The emperor was enraged, reduced the heir apparent to commoner status, and sent him to be imprisoned in the Supreme Capital. Shortly after, I-hsin sent emissaries to kill him and to persuade the viceroy of the Supreme Capital to report that he had died of illness. Taotsung, meanwhile, had almost immediately repented of what he had done in the heat of the moment and wished to summon the heir apparent's widow to court; I-hsin, terrified that the truth would now emerge in full, sent his men to murder her as well.'94 For the time being I-hsin was safe and secure in his position. Not only had he removed his main rival for power, but he also had managed to implicate in the alleged conspiracy a good many of his enemies, who were executed in the purge that followed. His own partisans were promoted and honored; indeed, 190 191 192 193 194 LS, 71, p. 1205. LS, 23, pp. 277-8. LS, 71, p. 1205. LS, n o , pp. 1484—5. LS, 72, pp. 1215-16; 23, pp. 279-80; n o , pp. 1485-8; 62, pp. 945-6. Cambridge Histories Online © Cambridge University Press, 2008 THE REIGN OF TAO-TSUNG 135 the court attendant whose false testimony had reopened the case was married to a royal princess. The fall of I-hsin Nonetheless, I-hsin's future remained dependent on the new empress whom he had introduced into the palace. He had naturally expected that in due course she would bear a royal heir over whom he could establish dominance. But the empress remained childless. In desperation I-hsin had her younger sister divorced from his son and had her, too, received into the royal harem. But she also failed to produce any children. To make matters worse for I-hsin, the heir apparent had left an infant son (Yen-hsi; 1072-1128; Khitan name A-kuo. Reigned as T'ien-tso huang-ti, 1101—25; n 0 temple name), born in 1072 or 1075. '95 Because the emperor had to decide on an heir, at the urging of Hsiao Wu-na he chose this grandson over the other possible candidate, his nephew Ch'un, the son of Tao-tsung's younger brother Ho-lu-wo (Ch'un, 1063—1122; Khitan name Nieh-li, reigned as emperor of Pei Liao for a few months in 1122, canonized as Hsiian-tsung). I96 This child-heir stood in the way of I-hsin's ambitions. In 1079, when the emperor was about to leave for his customary winter hunting expedition, I-hsin requested that the infant imperial grandson be left behind at the capital. Immediately various courtiers hostile to I-hsin, convinced that he planned to murder the prince, protested that the young boy would be left in danger and volunteered to remain behind to guard him. The emperor was finally persuaded to take him along in the imperial entourage.' 97 This incident seems finally to have awakened the emperor to I-hsin's devious and ambitious nature. Accordingly, in 1080 I-hsin was reduced in noble rank, stripped of the highest office, and sent away to govern Hsingchung fu. His days were numbered, however; in the winter of 1081 he was found guilty of trading prohibited goods with a foreign state and was sentenced to death. One of his allies had the death penalty reduced on account of his high status, and so he was banished to Lai-chou on the coast just north of modern Shan-hai kuan. I98 Later I-hsin was accused of storing up weapons and armor and plotting to defect to Sung, and finally, he was put to death.' 99 With the fall of I-hsin went the whole of the corrupt faction that had 195 LS, 24, p. 276, gives 1075 as his date of birth. However, it is probably wrong. LS, 30, p. 351, records his death in 102; at the age of fifty-four which would put his birth in 1072. 196 LS, 98, p. 1413. 197 LS, n o , p. 1079. 198 LS, 24, p. 286; n o , p. i486. 199 LS, 24, p. 288; n o , p. i486. Cambridge Histories Online © Cambridge University Press, 2008 136 THE LIAO supported him, the most important of them being Hsiao Yii-li-yeh and Yehlii Hsiao-chieh. Yeh-lii Hsiao-chieh was very different from the corrupt and vicious Khitan nobles who made up most of I-hsin's clique. Originally a poor scholar from a Chinese family named Chang Hsiao-chieh, Yeh-lii Hsiao-chieh had passed the chin shih examination in first place in 1055. He had enjoyed steady promotion in the bureaucracy until the early 1060s when he attracted the emperor's attention and became northern prime minister. He was granted the imperial surname after the suicide of the empress, in whose fall he had played a sinister part. After I-hsin's fall from power, Yeh-lii Hsiao-chieh, who was notorious for his insatiable greed and open corruption, was found guilty of malfeasance of official funds in 1080 and sent in disgrace to a provincial post. In 1081 he was reduced to commoner status. Later, however, he was allowed to return to the capital, and he died peacefully sometime in the late 1080s. When I-hsin was exiled, the new empress was sent away as well, reduced to the rank of Hui-fei and banished from the palace to serve at the Ch'ien-ling imperial tomb.200 Her sister was expelled from the palace and returned to her family.201 The emperor did nothing, however, to rehabilitate the unlucky victims of I-hsin's intrigues. The empress was not reburied in an imperial tomb or granted a posthumous title until 1101, after T'ien-tso's accession. The heir apparent was somewhat more fortunate. In 1083 Tao-tsung posthumously restored his rank, granted him the posthumous title Chao-huai t'aitzu, and reburied him with full imperial ritual at Yii-feng mountain.202 But the nobles and high officials who had been purged remained dishonored, and those who had been banished were left to languish. Tao-tsung was not generous to those who had once roused his suspicion and anger. The last years of Tao-tsung's reign, after the fall of Yeh-lii I-hsin, were comparatively uneventful. Tao-tsung was by now an elderly man (he turned fifty in 1082), and the active, peripatetic Khitan life-style was strenuous and demanding, even for an emperor. Only one of his predecessors had lived to reach his sixtieth birthday. The emperor continued nonetheless to pursue his 200 LS, 24, p. 287; 7 1 , p. 1205. 201 LS, 7 1 , p. 1205. Sometime afterward, in 1086, their mother Hsiao-ku, lady of Yen, was executed for having formerly seduced the prince of Liang. The prince of Liang was the title of Yen-hsi, the imperial grandson, from the third month of 1080 until the eleventh month of 1083. The accusation seems implausible if Yen-hsi is meant, because at the time of her execution he was still only eleven years old, and when he held the title prince of Liang, he was between five and eight. It is probable that the case involved not Yen-hsi but his father, the heir apparent, who had held the same title from 1063 onward. See LS, 22, p. 263. The case may therefore have resulted from the heir apparent's posthumous rehabilitation in 1083. As a result of her execution, Hsiao-ku's daughter, the former empress, was reduced to commoner status and sent to live in seclusion. See LS, 7 1 , p. 1205; 24, p. 292. 202 LS, 72, p. 1216; 24, p. 288. Cambridge Histories Online © Cambridge University Press, 2008 THE REIGN OF TAO-TSUNG 137 intellectual and religious interests; scholars were summoned to expound various of the Confucian classics, and monks to lecture on the Buddhist scriptures. In 1090 a Sung envoy commented at length on his lavish patronage of the Buddhist clergy and their all-pervasive influence in society.203 But Tao-tsung seems to have become more and more indolent and inattentive to his political duties. An anecdote tells how in the latter years of his reign he even resorted to selecting officials for high-ranking posts by having the candidates roll dice; even the historian who was later to compile the record of his reign was himself said to have been selected in this way.2"4 From the late 1080s the imperial grandson Yen-hsi, now prince of Yen, was carefully groomed to succeed to the throne. In 1086 Tao-tsung solemnly showed the young prince the armor and weapons of the first emperors T'aitsu and T'ai-tsung and described to him the hardships of the campaigns that led to the foundation of the dynasty. A few weeks later the prince underwent the "rebirth" ceremony, an important mark of his destiny as a ruler.205 In 1088 he was appointed to the first of a series of offices, to serve an apprenticeship in government. In the same year he was married,2"6 and sons were born in 1089 and 1093,2°7 a further guarantee of an orderly succession. The fall of I-hsin and his clique was not followed by any significant shift in policy. The pro-Khitan reforms of the legal code begun under I-hsin were continued, and more revisions were carried out in 1085, but these proved impossible to implement. 208 In 1090 the whole new legal system was abandoned, and the laws of 1034 were reinstated. Otherwise, the annals of these two decades record few administrative measures, except for the granting of tax exemptions and relief payments for districts stricken by natural calamities. The most noteworthy events had to do with relations with the neighboring peoples. Relations with major powers remained generally peaceful. There had been a troublesome border demarcation crisis with Sung in 1074, but after lengthy negotiations, this was settled peacefully by diplomacy in 1076. 2 0 ' The treaty itself held, and the steady exchange of ambassadors continued. The same was true of relations with Hsia and Koryo: In 1078 the king of 203 See Su Ch'e, Luan ch'eng chi, 42, p. 940. For some information on the scale of Buddhist influence during the Liao, see Wittfogel and Feng, pp. 2 9 1 - 7 . 204 LS, 98, p. 1416. 205 LS, 24, p. 292. 206 LS, 24, p. 297; 25, p. 300; 27, p. 317. 207 LS, 71, p. 1206; 25, pp. 298, 302. 208 In 1090 in a report to the Sung emperor, Su Ch'e commented that the disparity of treatment between Chinese and Khitan was largely confined to the adjudication of crimes of violence and was not so unfavorable to the Chinese population as generally believed. He also reported on the universal corruption and bribery in the Liao legal system. See Su Ch'e, Luan ch'eng chi, 42, p. 940. 209 See Klaus Tietze, "The Liao—Sung border conflict of 1074—76," in Studia Sino-Mongolica: Festschrift fiir Herbert Franke, ed. Wolfgang Bauer (Wiesbaden, 1979), pp. 127—51. Cambridge Histories Online © Cambridge University Press, 2008 138 THE LIAO Kory6 asked for the cession of territory east of the Yalu River, but this was refused without causing any break in relations.210 Tao-tsung's real problems were with the tribal peoples of the northern and northwestern frontiers. The tribal borderlands of the Liao empire were never clearly denned, constituting a zone inhabited by fragmented and unstable tribal groups who were loosely subject to the Liao court. Often some elements of the same people would be assimilated into the tribal section of the Liao empire while others of their relatives lived beyond the borders. The situation was always fluid, as groups merged and split apart and changed their allegiance. This situation was particularly complicated for two sets of peoples: the Jurchen in northern and eastern Manchuria and the Tsu-pu tribes living in the heartlands of Mongolia, the Orkhon and Selenga valleys. The Tsu-pu war, 1092—1102 The Tsu-pu were a Mongolian people either identical with or closely related to the Tatars (Ta-ta, Ta-tan).211 Like the pre-dynastic Khitan, they were a "nation" comprising various nomadic tribes who would join together when necessity demanded and could then become a powerful military force. The Liao classified them as a subordinate people, and they came to court and rendered tribute fairly regularly. Some of them lived in Liao territory, and some even served in one of the royal ordos. However, they were fiercely independent and had bitterly resisted the Khitan whenever the latter had attempted to expand to the northwest. There had been major trouble with the Tsu-pu at the turn of the eleventh century: There were outbreaks of warfare in 997-1000, 1007, 1012-23, and 1027* almost all of them coinciding with Khitan attempts to expand into Uighur-controlled territories in western Kansu. In 1069 there was a renewed Tsu-pu rebellion, which was put down by Yeh-lii Jen-hsien. After this, friendly relations had been restored, and in 1086 the Tsu-pu chieftain attended court, and Tao-tsung ordered the young prince Yen-hsi to befriend him as a potential ally. In 1089 the Tsu-pu came under the leadership of a powerful chieftain called Mo-ku-ssu. In 1092 the Khitan attacked some of their neighboring tribes on the Mongolian border, and during the conflict the Tsu-pu also became involved. In 1093, ^ ky Mo-ku-ssu, they launched a series of powerful raids across the northwestern borders of Liao, driving off many of the state herds of horses grazing there. Several other tribes joined their 210 LS, 115, p. 1522. 211 See Wang Kuo-wei, "Ta-ta k'ao," in his Kuan t'ang chi tin (repr. Peking, 1959), 14, pp. 5b—12a. Cambridge Histories Online © Cambridge University Press, 2008 THE REIGN OF T'lEN-TSO AND COLLAPSE OF LIAO 139 rebellion, including the warlike Ti-lieh (Tiriet) living around Hulun-nor (in western Heilungkiang), who had previously rebelled in 1073. The Tsu-pu invasion was a serious threat to the Khitan grazing lands, and the court entrusted its suppression to Yeh-lii Wo-t'e-la, the northern commissioner for military affairs. It took him eight years of constant heavy fighting to recover the Liao pasturelands and to force the Tsu-pu to submit. In the spring of 1100 Mo-ku-ssu was captured and sent to the capital, where he was hacked to pieces. But the war in the northwest dragged on. By the end of the year Wo-t'e-la was again at war with other northwestern tribes, and peace was not restored until 1102, when he suppressed these uprisings and beat off a new Tsu-pu invasion. This campaign was to be the last military victory of the Khitan, but Taotsung's reign ended on a successful note. When he died in I I O I , his empire still enjoyed power, stability, internal peace, and the respect of the surrounding peoples. THE REIGN OF T'lEN-TSO AND THE COLLAPSE OF THE LIAO The succession proceeded without incident: The imperial grandson Yen-hsi accepted the succession before his predecessor's coffin, announced a new regnal title, and promulgated the customary act of grace. Immediately he set about taking posthumous vengeance on I-hsin and his partisans who had brought about the deaths of his grandmother and his parents. The tombs of I-hsin and his allies were broken open; their corpses were mutilated; their posthumous offices and titles were rescinded; and all their families' possessions were confiscated and distributed to the families of those they had killed. The victims of their false accusations were restored to their ranks and noble titles; their confiscated properties were returned to their families; and those long banished were brought back to court. The deceased emperor was interred in his mausoleum together with the remains of the empress who had been falsely forced to commit suicide. The new emperor's father, the murdered heir apparent, was given a posthumous temple name as though he had actually reigned as monarch. Having vented his pent-up hatreds, the new emperor seems to have settled down into a round of routine and inactivity. This inactivity, however, may well be only a false impression created by the sources. No veritable record was ever compiled for T'ien-tso's reign, and the Liao history's basic annals for his first decade on the throne is little more than a scrappy itinerary of his hunting expeditions and travels around his empire, punctuated by the recep- Cambridge Histories Online © Cambridge University Press, 2008 140 THE LIAO tion of ambassadors from neighboring peoples.212 Reports of natural disasters continue spasmodically, and in 1105 T'ien-tso went out in disguise to see the sufferings of the people for himself. But he is harshly criticized for his excessive indulgence in hunting and, above all, for his arbitrariness, violence, and personal cruelty.2'3 There is almost no information about internal policy decisions, save that in 1105 members of merchant families were barred from taking the chin-shib examinations, a rather strange and belated adoption of an old-established Chinese social convention, quite out of place in the Khitan context, which shows the degree to which Chinese ideas had permeated government. Between 1103 and 1105, the Hsi Hsia repeatedly sent envoys asking for aid during their troubles with the Sung, and on one occasion in 1105, they asked the Khitan to attack the Sung. The Liao court wisely refused but cemented its relations with the Hsia by a marriage alliance. It also sent an envoy to the Sung court asking that they cease their attacks on Hsia and return the Hsia territories they had seized. War with thejurchen Until 1112 the Khitan state remained apparently secure. T'ien-tso had successfully resisted efforts to involve the Liao in the quarrels of the Sung and Hsia; the Tsu-pu had returned to their allegiance and sent embassies in 1006, 1110, and 1112; and relations with Kory6 remained amicable. In late winter the imperial entourage, according to custom, went on its seasonal fishing expedition to the Hun-t'ung River (the modern Sungari) somewhat west of modern Harbin. Here, again according to custom, the chieftains of the northeastern tribes, including the "wild" Jurchen from eastern Manchuria, came to pay homage. At the First Fish Feast at which they were entertained in the emperor's camp, the chieftains were expected to get up in turn and dance, as a symbolic gesture of submission. When it came to his turn, one of them, A-ku-ta, refused to do so, even after being bidden three times. T'ientso wanted to have him executed for his deliberate act of defiance, recognizing in him a potential enemy. But the influential chancellor Hsiao Fenghsien dissuaded him, belittling the harm that A-ku-ta could possibly do. It 212 The relevant section of CTKC, 10, pp. 99—100, also dismisses the first decade of his reign in a few lines, entirely devoted to embassies to the Sung court. The comparatively detailed account of the fell of the dynasty that follows mostly derived from the Chinjen wang Liao kuo, written by Shih Yuan, a man from the Liao Southern Capital who surrendered to the Sung and served as an official for many years before being sent back to the Chin. His book (long lost except for fragmentary quotations) was widely read in the Sung in the mid-twelfth century. See Fu Le-huan, Liao sbib u'ung k'ao, pp. 168-71. 213 See especially LS, 62, pp. 946—7, and the bitter denunciation made by Yeh-lii Yii-tu when he surrendered to the Chin in 1121. See also CS, 133, pp. 2847—8. Cambridge Histories Online © Cambridge University Press, 2008 THE REIGN OF T'lEN-TSO AND COLLAPSE OF LIAO 141 was to prove a fatal miscalculation and a decision that would lead to the end of the dynasty. The Jurchen were a Tungusic people whose fragmented tribes inhabited a broad belt of eastern Manchuria stretching northward from the northern border of Kory6 in the Yalu Valley through eastern Kirin and Heilungkiang and the maritime province of the Soviet Union north of Vladivostok. (Their earlier history is outlined in Chapter 3.) They had been in close contact with the Khitan ever since A-pao-chi's rise to power and particularly since the submission of Po-hai. They were powerful enough in the tenth century for the Sung to have thought of them as a potential ally against the Liao, and they had periodically caused serious trouble to both the Liao and the Kory6. By the end of the eleventh century they had long been divided by the Liao rulers into three broad groups. First were the "civilized" Jurchen (shu Nuchih), descendants of the tribes captured by the Khitan in the tenth century who had been settled in the Liao Valley and were fully assimilated. Farther north in eastern Kirin Province lived the "obedient" Jurchen {shun Nu-chih), who were considered a subordinate nation and had close and regular contact with the court. But by far the largest and most significant group were the "wild" Jurchen (sheng Nu-chih), who inhabited the middle and lower valley of the Sungari and the eastern mountains of Heilungkiang. They were nominally vassals of the Liao court, but subject to no real control. Each of these groups was divided into many tribes and clan groups living in small farming settlements but also pursuing a livelihood hunting, trapping, and herding animals. During the eleventh century one of the clans of the "wild" Jurchen, the Wan-yen, gradually established dominance over their neighbors and welded the Jurchen tribes into a powerful nation. The Liao court recognized the Wan-yen as leaders of the Jurchen and conferred on their chieftains the hereditary title of military governor of the Jurchen (Nu-chih chieh-tu shih). A-ku-ta was a prominent leader of the Wan-yen, though not yet, in 1112, their overall chieftain. In the early years of T'ien-tso's reign, relations between the Jurchen and the Liao had become increasingly strained. The Jurchen bitterly resented the way in which the Liao local officials at Ning-chiang, the main border trading post, constantly cheated them. They complained of the arrogance of the Liao envoys who abused their womenfolk and beat their village elders when traveling through Jurchen territories. And they resented their traditional obligation to provide the Liao emperors with the special gyrfalcons called hai tung ch'ing, which bred in the coastal regions, and to obtain them, the Jurchen had often to fight their way across the territories of their neighbors, the Five Nations tribes. Cambridge Histories Online © Cambridge University Press, 2008 142 THE LIAO In 1113 A-ku-ta was elected ruler of the Jurchen by the tribal leaders, succeeding his elder brother Wu-ya-shu (r. 1103-13), and was given the customary Chinese title of military governor by the Liao court. A-ku-ta immediately began to harass the Liao, raising a grievance that had rankled for years: the question of A-shu, a Jurchen chieftain who had opposed the Wanyen hegemony and taken refuge in Liao territory. A-ku-ta repeatedly and in vain demanded his return and began to build fortifications on the frontier. In the late autumn of 1114, A-ku-ta's demands having once again been refused by the Liao court, he attacked Ning-chiang Prefecture, the main frontier trading station and the place where the Liao emperors had customarily received the Jurchen leaders. At first T'ien-tso was not seriously alarmed and left the local forces to deal with the invaders, though he reinforced them with some Po-hai detachments sent from the vicinity of the Eastern Capital. This modest force was utterly defeated: The Liao had completely underestimated the strength and ferocity of the Jurchen. In the tenth month of 1114, T'ien-tso mobilized a force of select Khitan and Hsi troops under the command of Hsiao Ssu-hsien, the younger brother of his northern chancellor Hsiao Feng-hsien, but this force, too, was surprised and defeated on the Sungari River, with very heavy losses. Hsiao Ssu-hsien, in spite of his incompetence, escaped punishment, which helped demoralize the Khitan generals. By the end of the year, several border prefectures near Ning-chiang had surrendered to the Jurchen, and some of the neighboring tribes had also joined them. T'ien-tso then turned to diplomacy and sent envoys to open peace negotiations with A-ku-ta at the beginning of 1115. But late in the first month Aku-ta had declared himself emperor of a new Chin dynasty. He rejected the letters from the Liao court because they addressed him by name, not by his new title, and he continued to demand the return of A-shu and also the withdrawal of the Liao garrison from Huang-lung fu, the major administrative center in the region. Sporadic border fighting continued throughout 1115, with the Jurchen usually gaining the upper hand. Meanwhile, both sides prepared for a new round of warfare. Early in the autumn of 1115 T'ien-tso assembled a massive army under his own command, west of the Sungari. In the ninth month, before he could bring this force into action, A-ku-ta had already conquered Huang-lung fu, the easternmost major military outpost of the Liao. Then, when T'ien-tso finally crossed the Sungari River into Jurchen territory in the winter of 1115, his punitive campaign was undermined by a conspiracy to dethrone him and install his uncle, Prince Ch'un (1063-1122; Khitan name Nieh-li; posthumously canonized as Hsiian-tsung), as emperor. Cambridge Histories Online © Cambridge University Press, 2008 THE REIGN OF T'lEN-TSO AND COLLAPSE OF LIAO 143 The rebellion of Yeh-lii Chang-nu and the Po-hai uprising The conspirators were led by Yeh-lii Chang-nu, vice-commander of the vanguard, a member of the royal clan who was dissatisfied with T'ien-tso's leadership. He associated himself with Prince Ch'un's wife and son and persuaded other officers to join the conspiracy. The rebels deserted the expeditionary force and proceeded to the Supreme Capital, sending messengers to Prince Ch'un, who was then viceroy of the Southern Capital, informing him of their plan. The prince hesitated, however, knowing that many prominent and powerful members of the nobility still backed the emperor. When messengers arrived from T'ien-tso advising him to take precautions against the rebels, the prince thus decided to remain loyal. He beheaded Chang-nu's emissaries and sent their heads to the imperial headquarters. The rebellion did not end immediately, however. The rebels marched through the Khitan countryside, pillaging and rallying supporters to their ranks. But they failed to take the Supreme Capital and werefinallydestroyed by a small force of loyal Jurchen when they attacked the imperial winter camp at the Mu-yeh Mountain. More than two hundred guilty nobles were executed, and their wives and children were condemned to slavery. Yeh-lii Chang-nu was caught trying to escape to the Jurchen disguised as a messenger and was later executed by being cut in half at the waist. His quarters were sent out to be displayed at the other capitals, to discourage other potential traitors. Though Prince Ch'un had remained neutral and was not charged with disloyalty to his nephew and sovereign, the rebellion was not merely an attempt by a group of Khitan nobles to rescue the tottering empire from destruction under T'ien-tso's incompetent leadership. With the exception of Yeh-lii Chang-nu, because the leading figures in the rebellion were closely related to Prince Ch'un through birth or marriage, the uprising may be assumed to have resulted from yet another power struggle within the ruling clan. It was not the first time that T'ien-tso and Ch'un were rivals for the throne. Forty years earlier, after Yeh-lii I-hsin had had T'ien-tso's father murdered, he had unsuccessfully sponsored Prince Ch'un as the new heir apparent. When I-hsin fell from power, Prince Ch'un had also lost favor and had been banished from court. Though Chang-nu's rebellion failed, it nevertheless strengthened the prince's position. To ensure his loyalty, T'ien-tso gave him the grand title of Prince of Ch'in and Chin and appointed him commander in chief of the Liao armies, entrusting to him command of the defense operations against the Jurchen. The effects of Chang-nu's rebellion and the previous defeats in 1115 were felt immediately. Fighting spread into the adjacent Po-hai region, where Cambridge Histories Online © Cambridge University Press, 2008 144 THE disaffection had always been widespread and where a minor rebellion had occurred in the previous year. At the beginning of 1116 there was a major uprising at the Eastern Capital. The hated Khitan viceroy was murdered, and a Po-hai officer named Kao Yung-ch'ang declared himself emperor of the new state of Ta Yiian214 and requested aid from A-ku-ta against Liao punitive forces. The Jurchen relief force easily repulsed the Liao troops but then turned on the Po-hai rebels and, in the fifth month, killed Kao Yungch'ang. As a result, the whole region east of the Liao River, more than fifty prefectures, fell into Jurchen hands. This changed the entire strategic situation in the Jurchen's favor. The war was no longer confined to a distant and relatively unimportant frontier district, but now threatened the heart of the Liao empire. Great numbers of Po-hai and Khitan fled to settle in Kory6 (see Map 8). Expecting the Jurchen to advance west, T'ien-tso ordered his uncle Prince Ch'un to recruit a new army of select troops from the southern and western circuits and from the refugees from the Po-hai region. The twenty thousand men he recruited for his avenging army (yiian chtin) inflicted more damage on the civilian population than on the enemy, however, and morale collapsed everywhere. Rebellion broke out in the southern region among the Chinese population.21' When the Jurchen attacked Ch'un-chou on the Sungari River in early 1117, the northeastern army therefore melted away without offering even token resistance. At the end of the year, the Jurchen crossed the Liao River, routed Prince Ch'un's new army in battle, and conquered several prefectures west of the river. A-ku-ta adopted a new regnal title and proclaimed himself emperor of his new Chin dynasty. A bortive peace negotiations, 1118—1120 At this critical point, both sides abruptly ceased hostilities, and at the beginning of 1118 T'ien-tso initiated peace negotiations. Embassies were exchanged for two years. But the demands of the Jurchen were extremely harsh: A-ku-ta was not yet ready to destroy the Liao state, but he demanded to be invested by the Liao court as the emperor of Great Chin. Following the precedent of the Shan-yuan treaty, he further demanded to be addressed by the Liao emperor as elder brother, to have numerous Liao princes and princesses sent as hostages to the Chin court, to be paid an annual subsidy of silk and silver, and to be formally ceded the three circuits of the Supreme Capital, the Central Capital, and Hsing-chung fu. This would have left Liao in 214 According to Chong In-ji, KoryS sa, 14, p. 204. His title is given as emperor of great Po-hai in CTKC, 10, p. 108. 215 LS, 28, p. 335. Cambridge Histories Online © Cambridge University Press, 2008 ^ ™ " * - Advance of Jurchen Annies 1» Abortive Sung invasion 1122 * Flight of Ch'uns Empress Dowager 1122 •*• Withdrawal of Yeh-lii Ta-shih 1124 0 3 0 0 km | 1118 WAN-YEN Ning-chiangfi SUPREME Huang-lung fu P \ \ WbiteWE S T E R N # CAPITAL •.::•: C I R*C U I T ,* lao-chao 1118 \ Rebellion of Po-hai 1 against Khitan 1118. i Jurchen occupation 1118.V rebellion against '. >E~ASTERN Jurchen 1119, i CAPITAL MAP 8. The fall of Liao and the Chin invasions, 1017-24 Cambridge Histories Online © Cambridge University Press, 2008 146 THE LIAO control of only the Southern and Western Capital circuits and deprived of their tribal homelands. Later in 1118 A-ku-ta softened these demands somewhat, but the Liao court still balked at his terms in spite of the desperate situation in which they found themselves. Their remaining territories suffered a terrible famine in 1118, and there were more local rebellions and a constant trickle of defections to the Chin. Meanwhile, in the far west the Tsu-pu had rebelled again. A-ku-ta also faced his own internal problems. His initial victories came with unexpected ease, but his campaigns must have stretched the Jurchens' resources to the limit. Moreover, controlling the newly conquered territory proved far from easy. At the beginning of 1119 a rebellion against the Chin occupation erupted in the Eastern Capital. This had to be suppressed, and order restored. At the end of 1118 the Liao court agreed to invest A-ku-ta as king of Tung-hai, but A-ku-ta objected both to the title and to the language of the document of investiture as being humiliating, and so he angrily rejected the offer in the summer of 1119. He finally grew impatient and broke off negotiations in the third month of 1120. The hostilities resumed. Renewed fighting, 1120 A-ku-ta's first target was the Supreme Capital, which he attacked and captured in the fifth month. By doing so he showed his determination to deliver the coup de grace to the stricken dynasty. Situated in the Liao heartland, the traditional grazing grounds of the Khitan, the Supreme Capital had long since lost its political and administrative importance, but it remained the dynasty's religious and ritual center. The holiest places for the Khitan, such as the holy Mu-yeh Mountain, lay in its vicinity. The Jurchen invaders looted and destroyed the buildings at the imperial tombs and other vital religious sites. Having achieved his main objective, A-ku-ta now temporarily suspended his offensive because of the summer heat and distance from his home base, but the Liao'did not take advantage of this respite to organize its defenses. In the autumn of that year T'ien-tso went hunting as usual. The demoralized court was also once more shaken by internal dissension. Some of the most powerful men in the Liao state were tired of T'ien-tso's self-indulgence, dependence on favorites and sycophants, arbitrary justice and endless levies of manpower, and requisitions of money and goods. Above all, they resented his favorite Hsiao Feng-hsien's ability to prevent the emperor's hearing their own proposals or even from learning how truly desperate the situation was. In the spring of 1121 Lady Wen, the second wife of T'ien-tso and mother of the prince of Chin, conspired with her brother-in-law, General Yeh-lii Yii-tu, Cambridge Histories Online © Cambridge University Press, 2008 THE REIGN OF T'lEN-TSO AND COLLAPSE OF LIAO 147 to depose the emperor and install her own son in his stead. The plot was uncovered by the emperor's favorite Hsiao Feng-hsien, who was the brother of a rival consort, Lady Yuan, and hoped to ensure the succession for one of her sons. Lady Wen was forced to commit suicide, and some of the other conspirators were executed. Her son, the prince of Chin, in whom everyone had high hopes, was exonerated. Yeh-lii Yii-tu, the principal conspirator, escaped with his family and followers and in the fifth month defected to the Jurchen, who were of course delighted to receive him. He was allowed to remain in command of his troops, and in the winter of 1121—2 led a Jurchen army in an attack on the Central Capital. It and the surrounding territory fell to the Jurchen in the first month of 1122. At that time T'ien-tso was in or near the Southern Capital. Leaving Prince Ch'un in command of the Southern Capital circuit, he deserted his court and embarked on a prolonged flight from the advancing Jurchen, first heading northwest through the Chii-yung Pass into the Western Capital circuit. Hsiao Feng-hsien, still determined to ensure the succession for his nephew, now persuaded the emperor that Yeh-lii Yii-tu and his Jurchen allies were determined to destroy him and to enthrone the prince of Chin. The emperor thereupon ordered the unfortunate prince to commit suicide. Shortly afterward T'ien-tso finally tired of Hsiao Feng-hsien's manipulation and forced him to commit suicide. For the next three years T'ien-tso managed to stay always one step ahead of his pursuers. Four months after leaving the Southern Capital he had retreated to the inaccessible Yin-shan Range north of the bend in the Yellow River along the Liao—Hsia border, where he tried to recruit fresh troops from the local tribes. Hot on his heels, the Jurchen took the Western Capital in the third month of 1122 but were unable to establish firm control over all the western circuit. Furthermore, the Tanguts, fearing an invasion of their own borders, came to the support of the Liao emperor and sent troops to block the Jurchens' westward advance. A-ku-ta now arrived from Manchuria and soon defeated a combined Khitan-Tangut force near the Hsia border. To remove potential allies of T'ien-tso, to secure their own position, and to prevent further contacts with the Tanguts, the Jurchen uprooted many of the western tribes and resettled them east of the Hsing-an Range. T'ien-tso still remained in hiding in the Yin-shan, and A-ku-ta turned back east to take the Liao Southern Capital. After T'ien-tso abandoned them and fled west, severing all communication with the court, the high officials of the Southern Capital, led by the Hsi king and Yeh-lii Ta-shih, had in the third month of 1122 proclaimed Prince Ch'un their new emperor, citing as justification the precedent of T'ang Sutsung's seizure of the throne during An Lu-shan's rebellion, when Hsiian- Cambridge Histories Online © Cambridge University Press, 2008 148 THE LIAO tsung fled to Szechwan. T'ien-tso was demoted to the rank of a prince (it is for this reason that he was never granted a temple name).2'6 The Liao empire was now split: Tien-tso's authority, such as it was, was confined to the remote tribal areas of the far west. The territory controlled by Prince Ch'un was limited to the sedentary southern regions, and the Liao empire he ruled was reduced to a minor Chinese frontier state, its officials mostly Chinese, and the army, once almost entirely composed of Khitan tribesmen, now a motley mixture of Chinese troops and refugees from the east, under Khitan or Hsi commanders. To reinforce these none-too-dependable troops, Commander in Chief Yeh-lii Ta-shih attempted to recruit a new tribal army from Khitan and Hsi refugees. These recruits, however, were in such poor shape that they were nicknamed the "emaciated army" (shou chiin) and became more of a burden on the people of the southern circuit than an influx of new strength. Sung involvement Sung participation in these events had begun some years before. The Sung court was forewarned of the Jurchen threat to Liao by a defector as early as 1112. In 1117, sensing the imminent collapse of the Khitan state, it had established diplomatic contacts with the Jurchen under the pretense of trading for horses, hoping to form an alliance against Liao and divide up its territories.217 This was a shortsighted and potentially dangerous move. Not only was the court violating a solemn agreement with a powerful neighboring state for no really serious purpose, and thus forfeiting the trust of its other neighbors; it also was risking overturning the balance of power along the northern frontier that had kept the Sung secure for more than a century. The Sung court imagined that it saw the opportunity to regain the Sixteen Prefectures lost to the Khitan in 937. From the start this was an illusion: A-ku-ta proved willing to return to the Sung only Yen (the Southern Capital) and its six dependent prefectures. The Sung, moreover, were unable to intervene immediately, as they were heavily engaged elsewhere. After 1107 an uneasy peace had been restored with the Hsi Hsia. But in 1114 diplomatic relations between the two states had again collapsed: The Hsi Hsia invaded Sung territory and besieged Tingyuan. In 1115 the Sung counterattacked with a huge army and, after causing 216 LS, 29, pp. 343-4; 30, p. 352. 217 Dagmar Thiele, Der Abschluss eints Vertrages: Diplomat it zwiscbtn Sung und Chin Dynastic, 1 / 1 7 - 1 1 2 3 , Miinchener Ostasiatische Studien no. 6 (Wiesbaden, 1970), is the most detailed study in a Western language on the negotiations between the Sung and the Chin. See also the cleat account in Tao Jingshen, Two Sons of Heaven: Studies in Sung—Liao relations, pp. 87—97. Cambridge Histories Online © Cambridge University Press, 2008 FINAL DISASTER 149 widespread destruction, met with a shattering defeat. The war dragged on until 1119, and huge numbers of Sung troops were committed to the front with Hsi Hsia. In 1119 the combatants finally concluded a peace, and the Sung court could think seriously about attacking the Khitan.2'8 But in 1120—1 its attention was again somewhat distracted by the rebellion of Fang La in Chekiang.219 As long as the Jurchen attacks were directed against the tribal and Po-hai areas of the Liao empire, Sung interests had not been directly threatened. But in 1122 when A-ku-ta began to campaign against the western capital and its circuit, which included several of the Sixteen Prefectures that lay along their own borders, the Sung suddenly became alarmed. Because of their military troubles on the border of Hsi Hsia and in the south, the Sung had still not yet attacked Liao in support of the Jurchen campaigns, as they had undertaken to do. Now Liao looked like easy prey, and the Sung regime in K'aifeng hoped that the remaining officials and the population of the Southern Capital circuit would readily submit to a Sung invasion force. A large army was hastily mobilized by the Sung in the late spring of 1122. An attempt to persuade the Liao court to surrender failed, and fighting erupted in early summer. A small force of Khitan and Hsi led by Hsiao Kan and Yeh-lii Tashih, however, warded off the Sung invasion without much trouble. FINAL DISASTER Only three months after becoming emperor, Prince Ch'un died. He had no heir and bequeathed the throne to the prince of Ch'un, who was T'ien-tso's son by Lady Yuan. The prince of Ch'in, however, was in hiding with his father in the west. Prince Ch'un's wife was therefore appointed empress dowager and regent,220 but she was powerless to stem the rapid disintegration of the dynasty. In desperation, most of the Liao's Chinese subjects were now looking for any way to save their own lives, and collaboration with the Sung appeared the most attractive alternative. Li Ch'u-wen, who had helped install Prince Ch'un as emperor, secretly advocated surrender to the Sung. But the empress dowager forced him to commit suicide and executed his son for treason. Then in late autumn of 1122 Kuo Yao-shih, one of her chief generals, and Kao Feng, the commander of the vital border prefecture of I-chou, defected with their troops to Sung. Following Kuo Yao-shih's advice, Sung 218 See Evgenii I. Kychanov, "Les guerres entre les Song du Nord et le Hsi-Hsia," in fctudes Song in mmonam ttierme Balazs, ed. Franchise Aubin, ist series, no. 2 (Paris, 1971), pp. 102—18. 219 See Kao Yu-kung, "Source materials on the Fang La rebellion," Harvard Journal of Asiatic Studies, 26 (1966), pp. 211 —40. 220 LS, 29, p. 394. Cambridge Histories Online © Cambridge University Press, 2008 150 THE LIAO generals again tried to take the Southern Capital. Kuo led his troops into the city, but his Sung allies failed to halt the arrival of a loyal Khitan relief force under Hsiao Kan, which almost wiped out the invaders. The failure of the Sung offensive against the Southern Capital gave A-ku-ta the opportunity to intervene and take the region himself. The Liao empress dowager tried repeatedly but unsuccessfully to persuade him to recognize the prince of Ch'in as his vassal king. But A-ku-ta refused, secure in the knowledge that the remainder of the Liao territories was his for the taking. In midwinter of 1122 he broke through the Chii-yung pass and occupied the Southern Capital without meeting any resistance. The Liao empress dowager, high Khitan and Hsi officials, and many of the non-Chinese population had fled the capital by another route immediately before the arrival of the Chin army and were traveling north into Hsi territory.221 There they split up into two groups. The Hsi and Po-hai troops followed Hsiao Kan into the Hsi homeland, where at the beginning of 1123 he set up a short-lived Ta Hsi dynasty that lasted until his death at the hands of his own troopsfivemonths later.222 Yeh-lii Ta-shih, meanwhile, led the empress and the Khitan troops west to join T'ien-tso. When they finally met the emperor at T'ien-te (north of modern Urat South Banner, Inner Mongolia) near the Hsia border, Ch'un's unfortunate widow was executed for disloyalty, but Ta-shih, even though he upbraided T'ien-tso for having deserted his court and capital, was spared. T'ien-tso could not afford to execute so able a general, especially one with seven thousand experienced soldiers under his command. Early in the summer of 112 3 Yeh-lii Ta-shih was captured in battle by the Jurchen, who forced him to lead a raid on T'ien-tso's camp, in which they captured almost the entire imperial family and entourage.22' Only one son, the prince of Liang, managed to escape with his father into the nearby Yinshan Mountains. When T'ien-tso decided to accept a Tangut invitation to seek refuge in Hsia, the prince and his retinue objected and left the imperial camp, setting out for the Wu-ku and Ti-lieh tribes in the northern Gobi. There the prince of Liang was briefly established as the "emperor" of yet another short-lived dynasty. The Tangut ruler, Li Ch'ien-shun, had meanwhile changed his mind about giving asylum to the imperial Liao refugee. While T'ien-tso was encamped north of the Ordos waiting for a Tangut escort, the Jurchen sent a message to the Tangut court warning of the grave consequences that would ensue if Ch'ien-shun were to give him refuge. To soften the threat, the Jurchen also offered Hsia a strip of territory along the Yellow River. In a last futile attempt to win Tangut support, T'ien-tso 221 LS, 29, p. 345. 222 LS, 29, pp. 345-7. 223 LS, 29, p. 346. Cambridge Histories Online © Cambridge University Press, 2008 FINAL DISASTER 151 formally invested Ch'ien-shun as emperor of Hsia. But the Tangut ruler ignored this pathetic gesture and some months later, in the beginning of 1124, declared himself a vassal of Chin. By this time the Liao emperor had recrossed the Yellow River and sought temporary refuge with the T'u-lii-pu tribe somewhere to the north of modern Hu-ho-hao-t'e. Yeh-lii Ta-shih, who had escaped from the Jurchen late in 1123, now rejoined him. Pressed by the pursuing Jurchen, the emperor headed north into Mongolia. His fortunes had now sunk so low that his entourage had to barter their clothes for food from the Wu-ku and Ti-lieh tribes. Still, he deluded himself with plans to launch an offensive in the summer to retake the Western circuit and the Southern Capital. Ta-shih objected strongly, pointing out the absurd impossibility of such a plan. When T'ien-tso stubbornly stuck to his intention and began to attack nearby prefectures, Ta-shih tired of his delusions and decided to go his own way. In the eighth month of 1124 he declared himself king and led his followers northwest across the Gobi into northern Mongolia. "* T'ien-tso, as was inevitable, was defeated and was finally hunted down and captured near Ying-chou (modern Ying-hsien in Shansi) in the second month of 1125. He was taken to the Jurchen court in Manchuria and stripped of his imperial title, being given the title of "king of the seashore" {hai-pin wang), an ironic reference to the title "king of the Eastern sea" that he had offered A-ku-ta in n 18. Little is known about T'ien-tso's life in captivity or about the date and circumstances of his death. Both the Liao shih and Ch'i-tan kuo-shih report that he died in captivity at the age of fifty-four. Because the Liao shih gives 1075 as the year of his birth, this places his death in 1128. According to the Chin shih T'ien-tso was still alive in 1126, when one of his servants falsely accused him of planning to escape. T'ien-tso's capture by the Jurchen marked the end of the Liao, but not that of an independent Khitan nation. In northern Mongolia, Yeh-lii Ta-shih established his headquarters in K'o-tun (Chen Prefecture) on the Orkhon River, the Khitan military and administrative center of this region. He probably secured the support of the K'o-tun garrison of twenty thousand warriors and also sought the allegiance of local tribes. He set himself up as emperor and took the title of gurkhan (ko-erh-han), "universal khan." But instead of turning east to do battle with Chin, around the year 1130 Yeh-Lii Ta-shih led his horde westward in search of new territories. Within a year the Uighurs had acknowledged his suzerainty, and he had established a foothold in eastern Transoxiana, after which he gradually conquered the whole area between the Pamirs and the Aral Sea. 224 LS, 29, p. 349. Cambridge Histories Online © Cambridge University Press, 2008 Land over 20O0metres(6562ft) Land over 5 0 0 metres(164Oft) 500 1000km M A P 9. The Hsi Liao (Khara Khitan) empire, 1131—1213 Cambridge Histories Online © Cambridge University Press, 2008 FINAL DISASTER I53 Ta-shih made his capital at Balasaghun (Hu-ssu-ordo), not far from the western end of Issyk-kul. His empire, which lasted until the beginning of the thirteenth century, came to be known by the names Khara (black) Khitai or Western (Hsi) Liao (see Map 9). Evidently his authority did not extend to northern Mongolia, where he had started his westward trek. An expedition launched against Chin in 1134 ended in a complete fiasco, and this convinced Ta-shih of the futility of his attempts to restore Liao rule over its former territories. Thenceforward the history of Western Liao was associated with Central Asia and not China and is recorded almost exclusively in Arabic and Persian sources.225 225 For further details about the Khara-Khitai, consult the excellent appendix dealing with the subject in Wittfogel and Feng, pp. 6 1 9 - 7 4 . Cambridge Histories Online © Cambridge University Press, 2008 CHAPTER 2 THE HSI HSIA INTRODUCTION At its height in the middle of the twelfth century the Hsia state (ca. 9821227) embraced the Ordos and the Kansu corridor. In the northeast it abutted the Chin empire along the Yellow River; westward it extended beyond Tun-huang to Yii-men, north to Edzina (O-chi-na; Khara-Khoto) on the southern rim of the Gobi, and south to Hsi-ning, skirting Lake Kokonor and the city of Lan-chou. From his capital city of Chung-hsing on the west bank of the Yellow River in the foothills of the Alashan (Ho-lan Mountains), the Tangut emperor ruled over a multiethnic, economically diversified population numbering perhaps three million. His subjects included Chinese, Tibetans, Uighurs, Khitans, and various other Ch'iang and Turkic groups in addition to the Tangut core. It is difficult to determine how far the ethnic boundaries between these groups remained distinct or blurred. No law prohibited intermarriage. The Tangut, Chinese, and Tibetan languages enjoyed official recognition and widespread usage. The administrative apparatus bore the outer trappings of its Chinese models, though its inner workings still remain mostly unknown. Imperial revenues were derived from taxes, largely in kind, on the products of animal husbandry and irrigated agriculture, as well as on internal and foreign trade.1 As the preeminent state religion, Buddhism was lavishly patronized by the imperial family and court. The synthetic Tangut brand of northern Buddhism blended Tibetan Tantric and Chinese Mahayana scriptural traditions in a Sino-Nepalese artistic setting, often referred to as the Khara-Khoto style. This religion provided the outstanding inspiration of Tangut culture, judging by its literary and material remains. Officials and the literate classes 1 There are two comprehensive histories of the Hsi Hsia generally available: Evgenii I. Kychanov, Ocherk istorii tangutskogo gosuaarstva (Moscow, 1968), and Wu T'ien-chih, Hsi Hsia shih km (Ch'eng-tu, 1980, 2nd. rev. ed., 1983), thoroughly cover the existing scholarship. Ongoing translation projects and archaeological excavations continually increase the corpus of Tangut source materials, to which the present synthesis cannot do full justice. Some of this new material is included in a recent popular account of Hsi Hsia culture by Shih Chin-po, Hsi Hsia wen hua (Ch'ang-ch'un, 1987). 154 Cambridge Histories Online © Cambridge University Press, 2008 ETHNOGENESIS OF THE TANGUT PEOPLE 155 studied such Chinese classics as the ever-popular Hsiao ching (Book of filial piety), Lun yii (Analects of Confucius), or Meng-tzu (Mencius) through Tangut translations, but they were equally likely to read the works of Chuang-tzu or Lao tzu, military treatises, divination manuals, or Buddhist sermons on popular morality. The Tangut emperor cultivated his sacral image as a Buddhist ruler and accorded high status to the Tibetan lamas at his court.2 The Tangut empire was at peace throughout most of the twelfth century and presented a less awesome military aspect than it had one hundred years earlier, when it had fought bitterly and often with the Chinese Sung empire, and sometimes also with the Khitan state of Liao, to preserve its independence. After the fall of Liao (1125), it enjoyed reasonably cordial relations with its Jurchen neighbors to the east and presumably with its trading partners to the west and north: Kao-ch'ang, Kucha (Ch'iu-tz'u), Khotan (Yiit'ien), Khara-Khitai, and the Tatars, Naimans, and Kereyids. On the eve of the Mongolian conquests, the Tangut state occupied a position of considerable cultural and political standing in East Asia. This sophisticated state evolved gradually over many centuries. Its dynastic house traced its primacy among the Tanguts back to the Sui period ( 5 8 1 618) or earlier. An independent Tangut state came into being in 982, was formally proclaimed a "dynasty" in 1038, and was destroyed by the Mongols in 1227, 245 years after its founding. Recent studies show that remnants of the ruling clan fled from the Mongols and reestablished themselves, albeit on a much-reduced scale, in western Szechwan, where they served the Ming as t'u ssu (local tribal headmen) until they finally died out around 1700. Today, descendants of these same people are believed to inhabit certain regions of western Szechwan and to speak a dialect akin to the language of the Hsia state.' THE ETHNOGENESIS OF THE TANGUT PEOPLE The earliest Chinese notices of the Tanguts call them the Tang-hsiang Ch'iang and describe them as descendants of the western Ch'iang of Han times. Since antiquity, Ch'iang peoples had occupied the steppes around the Kokonor (Ch'ing-hai) and the mountainous regions south of the lake around the headwaters of the Yellow, Ta-t'ung, and Huang rivers. In this borderland, which traditionally formed the northeastern Tibetan province of 2 Kychanov, Ochtrk istorii, chap. 8; Wang Chung, "Lun Hsi Hsiatehsingch'i," Lishihyenchiu, 5 (1962), p. 32. 3 Wu, Hsi Hsia shih kao, pp. 127—37; Li Fan-wen, "Hsi Hsia i min tiao ch'a chi," Ning-hsia she bui k'o hsiieh, 1 (1981), pp. 38—62. Cambridge Histories Online © Cambridge University Press, 2008 156 THE HSI HSIA Amdo, commingled the ancestors of both Tanguts and Tibetans. Studies of the Tangut language have demonstrated that Tangut speakers may be classified as ancient members of the Tibeto-Burman peoples. Culturally, the Tanguts shared many traits with other Ch'iang groups. The name Tangut (Tangghud) first appears in the Orkhon Turkic runic inscriptions dated A.D. 735. Almost certainly originally an Altaic form, Tangut was probably derived in turn from a Tibetan self-appellation through Central Asian mediaries, perhaps Khotanese or Sogdian. In time, Tangut (transcribed in Chinese as T'ang-wu or T'ang-ku-t'e) became a common central and north Asian name for certain tribes inhabiting the Amdo— Kokonor region and even Kansu, and it was used until the nineteenth century. Its usage is widespread in later Chinese, Turkic, Persian, and Arabic writings and in the memoirs of nineteenth- and twentieth-century Western explorers of the Sino-Tibetan borderland.4 In their own language the Tanguts called themselves Mi or Mi-niah (Minag). Mi-nag (Chinese: Mien-yao, Mi-yao) is also their Tibetan name, first documented for the seventh century in the T'ang histories as the T'u-fan designation of Tanguts who became T'u-fan vassals. In later Tibetan literature Mi-nag denotes a kingdom in the north, the area northeast and east of Kokonor, and finally the entire Tangut realm of Hsia. During the eleventh century the Kokonor Tibetans and the Khotanese referred to the Tanguts of Hsia by this name in their correspondence with the Sung court. By the thirteenth and fourteenth centuries Mi-nag (Chinese: Mi-na-ku or Mi-nok'o) had become identified with Ho-hsi (Mongolian Khashin), both terms being used to designate the former lands and subjects of the Hsia state. Mu-ya and Mu-na are other variants of Mi-nag, found in the toponymy and lore of western Szechwan. Whatever its origin, the name Mi-nag came to be a widely applied term whose ethnic and geographical affiliations are still not entirely clear.' Throughout their history the Tanguts were influenced profoundly by their neighbors, especially by the three dominant cultures of continental East Asia: the Indie-Tibetan, Chinese, and Turko-Mongolian. From the fourth to the seventh centuries the Kokonor area came under the sway of the Hsien-pei T'u-yii-hun, whose eponymous founder of the Mu-jung clan had migrated from Manchuria early in the fourth century. The T'u-yii-hun ruled the Ch'iang hinterland, living a nomadic life in the lush pasturelands around 4 On the relationship between the Tangut and Ch'iang languages, see Wang Ching- ju, Hsi Hsia yen chiu, vol. 2 (Peking, 1933), pp. 275-88. For a more detailed discussion of the name Tangut and related issues, see Ruth Dunnell, "Who are the Tanguts? Remarks on Tangut ethnogenesis and the ethnonym Tangut," Journal ofAsian History, 18 (1984), pp. 7 8 - 8 9 . 5 The basic study is by Rolf A. Stein, "Mi-nag et Si-hia. Geographie historique et legendes ancestrales," Bulletin de 1't.cole Franfaisc a*Extreme-Orient, 44 (1947—50), pp. 223—65. Cambridge Histories Online © Cambridge University Press, 2008 ETHNOGENESIS OF THE TANGUT PEOPLE 157 Kokonor and intermarrying with the indigenous elites. By the Sui period (581—168) the Tang-hsiang had emerged as the successors of two of the T'uyii-huns' client statelets, Tang-ch'ang and Teng-chih. Their first appearance in the sources coincided with, and was linked to, the breakup of the first Turkic empire at the end of the sixth century and the rise of the Tang dynasty. By 628-30 T'ang armies had crushed the Turks and dispersed their T'u-yii-hun allies. Soon after, the armies of the new Tibetan kingdom also began to attack the T'u-yii-hun from the southwest and to pressure the Tanguts. By 680 the Tibetans had displaced the T'u-yii-hun in Kokonor, forcing many Tanguts to flee their homeland. These events engendered a series of migrations and the eastward resettlement of steppe and border peoples along or within the northwestern frontier of T'ang China. The first sizable submission of Tangut households to the Chinese border authorities, involving the Tangut chieftain T'o-"pa Ningtsung, had already taken place under the Sui in 584-5 but had required no resettlement.6 In the sixth and seventh centuries Tangut society comprised a collection of loosely allied autonomous clan or lineage groups, distinguished (or ranked) by the number of mounted warriors that each could muster. Their principal characteristic, according to the Chinese descriptions, was a stubborn independence sustained by strong mutual animosities nurtured by constant raiding and blood feuds. A military emergency might unite them; otherwise, it was said, mutual intercourse was limited to a triennial gathering for the sacrifice of animals to heaven. A semisedentary livelihood based on stockbreeding (sheep, pigs, yaks, horses, and donkeys) and raiding reinforced this preference for autonomy, which both favored and fettered the evolution of an independent state.7 That state was very much a creation of the T'o-pa Tanguts and their allies. From the start the name T'o-pa was attached to the most prominent Tangut clan. Distinct from all other early Tangut clan names found in Chinese records, T'o-pa is indisputably a princely Hsien-pei surname, also borne by the royal house of the Later Wei and by certain clans among the T'uyii-hun. That fact has led some scholars to suggest that the T'o-pa Tanguts were a Tibetanized Hsien-pei elite ruling a largely Ch'iang population, as their T'u-yii-hun predecessors had done. In the early eleventh century, the first Hsia emperor himself advanced this interpretation of their origins by claiming descent from the T'o-pa rulers of the Later Wei. Other modern 6 Wei Cheng (580-643) et al., eds., Sui ihu (Peking, 1973), 83, p. 1846. On the emergence of the Tangut during the sixth to the eighth centuries, see Paul Friedland, "A reconstruction of early Tangut history" (Ph.D. diss., University of Washington, 1969). 7 Sui shu, 83, p. 1845; Tu Yu (735—812), Tung tin, in Shih lung, comp. Wang Yiin-wu (Shanghai, 1935-7; repr. Taipei, 1965), 190, p. 10121. Cambridge Histories Online © Cambridge University Press, 2008 158 THE HSI HSIA scholars, however, regard this claim to kinship with the Later Wei imperial clan as merely a political expedient and a prop to dynastic legitimacy.8 The T'o-pa name clearly carried great prestige in both north China and Inner Asia. As the politically dominant group, the T'o-pa response to events can be traced with some certainty: first alliance with the T'u-yii-hun, then avoidance of Tibetan domination, and finally submission to the T'ang. A similar pattern is repeated in the Hsi Hsia imperial period: alliance with the Khitans, struggle against the Ch'ing-t'ang (Kokonor) Tibetans, and, despite bitter and protracted disagreements, a continued affinity for China (primarily an economic affinity). THE SURRENDER TO T'ANG AND SETTLEMENT IN THE ORDOS After Chinese armies had recovered the Ordos in 628, the Tangut chieftain Hsi-feng Pu-lai led thefirstwave of submissions to the T'ang in 628-9. T'o-pa Ch'ih-tz'u, a powerful Tangut leader allied by marriage with the T'u-yii-hun king, Mu-jung Fu-yiin, at first resisted the T'ang court's blandishments, but by 635 the T'u-yii-hun ruler had perished, and the reluctant T'o-pa leader and his followers had been placated and had submitted to the T'ang. In 631 Tangut territory was divided into thirty-two "loosely controlled" prefectures (chi-mt chou) under the overall control of a newly created Sung-chou prefectural government-general (tu-tu fu). T'o-pa Ch'ih-tz'u was appointed governorgeneral of Hsi-jung Prefecture. He was also granted the T'ang imperial surname Li, which, however, the T'o-pa did not adopt until the end of T'ang. Other Tangut chieftains likewise were appointed as prefects of their newly reorganized districts. Thus the Ho-ch'ii region around the headwaters of the Yellow River east of Kokonor, with its population of about 340,000 Tanguts, came under T'ang sovereignty. Around this time the T'ang authorities also settled 100,000 surrendered eastern Turks (T'u-chiieh) in the Ordos, also organized in loosely controlled prefectures established along tribal lines. The periodic resettlement of Turkish remnants in the Ordos and in northern Shansi continued throughout the seventh century. In time, groups of Tanguts and T'u-yii-hun joined them there. Tibetan attacks on the weakened T'u-yii-hun began in 638, and pressure on the Tanguts and other Ch'iang groups mounted as the Tibetan kingdom 8 Okazaki Seiro, Tanguto kodai shi ktnkyu (Kyoto, 1972), pp. 23—5, reviews the arguments; Wu, Hsi Hsia shih kao, pp. 2—4, 8—10, n. 3, advocates the Hsien-pei theory. Cambridge Histories Online © Cambridge University Press, 2008 SURRENDER TO T'ANG, SETTLEMENT IN THE ORDOS 159 expanded northeastward. By 680 the Tibetans had conquered all of Ch'iang territory, absorbing among other ethnic groups Tangut elements that were called the Mi-yao (Tibetan: Mi-nag; the Tibetans called their T'u-yii-hun subjects 'A-zha). Other Tangut groups, led by the T'o-pa, petitioned the T'ang authorities for permission to abandon their homeland in the grasslands of Ho-ch'ii and to move into China. The emigrants were relocated in Ch'ingchou (Ch'ing-yang, modern Kansu) and reorganized in several specially created tribal prefectures controlled by the Ching-pien government-general, a transplantation of the former Sung-chou government-general. A second migration of Tanguts into T'ang border territories followed in 692. By some accounts this group numbered as many as 200,000 persons, who were resettled in ten new tribal prefectures established in the southern Ordos, between Ling-chou and Hsia-chou.9 Toward the end of the seventh century, a powerful second Eastern Turkish (T'u-chiieh) khaghanate formed in the steppe and subsequently directed numerous raids against the Ordos and northern Shansi. Just when the T'ang had concluded a treaty with the Turks and opened a large border market in Shuo-fang in 721-22, a Sogdian-led revolt of surrendered Turks who had been settled in the Ling-chou and Hsia-chou area engulfed the Ordos region and required over a year to suppress. Although some Tanguts joined the uprising, their great chieftain T'o-pa Ssu-t'ai, governorgeneral of Ching-pien and a direct descendant of T'o-pa Ch'ih-tz'u, lent support to T'ang troops and was duly rewarded.10 The revolt may have been inspired by T'ang efforts to turn the resettled Turks into farmers. Its failure, however, weakened the formerly paramount position of the T'u-chiieh and Sogdians in the Ordos, and by the time the An Lu-shan rebellion began in 755, the Tanguts had become the dominant people on the borders of the Ordos south of the Yellow River. Moreover, by 744 the second Eastern Turkish khaghanate had collapsed, and control of the steppes had passed to one of its subordinate tribal unions, the Uighurs. In time the Uighurs became the Tanguts' chief rivals for control of the lucrative horse and livestock trade among the southern steppe, Ho-hsi, the Ordos, and China. During the An Lu-shan rebellion (755-63) and the subsequent Tibetan invasion of northwestern China, many Tanguts seized the opportunity to raid 9 For Tangut migrations and resettlements see Friedland, "A reconstruction of early Tangut history," pp. 131-36, 1 6 5 - 7 5 , 2 " . 23<>, "• 17; Wang P'u (922-82) et al., comps., T'anghuiyao(Shanghai, 1935; repr. Peking, 1955), 98, p. 1756 (hereafter cited as THY); Ou-yang Hsiu (1007-72) and Sung Ch'i (998—1061), comps., Hiin T'ang shu (Peking, 1975), 221A, pp. 6 2 1 5 - 1 6 (hereafter cited as HTS). 10 Denis C. Twitchett, "Hsiian-tsung," in Sui and T'ang China, 589—906, pt. 1, vol. 3 of The Cambridge history of China, ed. Denis C. Twitchett and John K. Fairbank (Cambridge, 1979), pp. 4 3 5 - 3 6 ; Edwin G. Pulleyblank, "A Sogdian colony in Inner Mongolia," T'tmngpao, 41 (1952), pp. 317—56; Friedland, "A reconstruction of early Tangut history," pp. 212—216. Cambridge Histories Online © Cambridge University Press, 2008 l6o THE HSI HSIA Chinese border settlements or to join local insurrections along with bands of T'u-chiieh and T'u-yii-hun. Later, members of these three groups entered the armies of both the Tibetans and the rebel Uighur commander P'u-ku Huai-en (764-5). Other Tangut groups, mainly in the Ling-chou Hsia-chou area, cooperated with the Tibetans, who overran Lung-yu (modern Kansu) and tried to conquer Ch'ang-an, the T'ang capital. Some of them subsequently returned to T'ang loyalty; others served as guides for the Tibetan invaders, accepted appointments and titles, and formed marital alliances with them. With the partial reassertion of T'ang authority in 765, the disrupted Ordos population was resettled according to a plan designed to separate the various groups from one another and from the Tibetans, who now occupied all of Lung-yu (Ho-hsi and eastern Kansu). Stringent precautions were taken to sever lines of communication between the Tibetans and the tribal inhabitants of the Ordos. In this resettlement the Tanguts of Ching-pien and the six fu of Hsia-chou were moved east of Hsia-chou and north of Yin-chou (modern Mi-chih in Shensi). T'o-pa Ch'ao-kuang, the Tangut leader at Ching-pien and the left Yii-lin grand general, was rewarded for his loyalty to the T'ang court and sent back to pacify his relocated tribespeople." He was probably a direct descendant of T'o-pa Ch'ih-tz'u, and although he himself evidently did not hold the post of Ching-pien governor-general, that office eventually was bestowed on the principal T'o-pa lineage, which he represented. Its administrative seat was moved north from Ch'ing-chou to Yin-chou, and their new official title gave the T'o-pa prestige in the building of a power base nearby at Hsia-chou. The pro-Tibetan Tanguts meantime remained at Ch'ing-chou, chief among them being the Yeh-li, Pa-li, and P'o-chou clans. One T'o-pa leader, T'o-pa Ch'i-mei, also remained at Ch'ing-chou, but neither his relationship with T'o-pa Ch'ao-kuang nor the extent of his involvement with the Tibetans is mentioned. The resulting realignment openly divided the Tanguts into two distinct and potentially hostile branches, called in contemporary accounts the P'ing-hsia group (at Hsia-chou) and the Tung-shan group (at Ch'ing-chou).12 Although some Tanguts continued to collaborate with the Tibetans, others became the targets of the Tibetans' cattle raids. Yen-chou, due west of Hsia-chou and straddling the approach to Ling-chou, seat of the T'ang Shuofang commandery, fluctuated between T'ang and Tibetan control for many years. In 786 the Tibetans invaded the southern Ordos and penetrated as far as Hsia-chou, where the Tangut prefect, T'o-pa Ch'ien-hui (the son of T'o-pa 11 HTS, 221A, p. 6217; Friedland, "A reconstruction of early Tangut history," pp. 2 1 7 - 2 6 . 12 Wu, Hsi Hsia shih kao, p. 4. Cambridge Histories Online © Cambridge University Press, 2008 THE END OF THE T'ANG l6l Ch'ao-kuang), abandoned the citadel to enemy occupation. The Tibetans retreated from Hsia-chou the following year but continued their raids on the Ordos settlements. Tangut tribes outside Hsia-chou gained a similar reputation, as a naturally predatory people whose numbers and prosperity grew apace through cattle breeding, trading, and the raiding of established frontier settlements. The T'ang government tried periodically to prohibit private commerce with these tribes, in an unsuccessful effort to stem official corruption and abuses and to halt the illicit trade of Chinese silk and weapons in exchange for Ordos livestock.'3 Cycles of oppression and retaliatory raids typified the ninth century and were caused mainly by diminishing T'ang control over its peripheral provinces. Numerous incidents reveal that the increasing tribal wealth in livestock attracted the greed of Chinese frontier officials, who exploited them by means of unfair market practices or outright seizure of their animals. In revenge, the Tanguts, often with Tibetan help, raided border prefectures in the Hsia-Yen area. The T'ang's communication lines to Ling-chou grew ever more precarious just at the same time as control over livestock production and horse supplies, so vital to the Chinese and their armies, passed into Tangut hands. ' 4 It is in this context that the name of the Nan-shan Tanguts first appeared, around 846-9, denoting the Tangut tribes dwelling in the mountains southwest of Hsia-chou and north of Yen-an, around Heng-shan. Their leaders were later identified as Yeh-li chieftains. They were described as poorer and more contentious than their P'ing-hsia cousins, with whom they constantly feuded. Perhaps exiles from Ch'ing-chou or Yen-chou, the Nan-shan Tanguts were caught between the P'ing-hsia group to their north and the Chinese to their south, each contending to assert control over the border region and its inhabitants.1' THE END OF THE T'ANG New configurations of power slowly took shape in the Inner Asian steppes and in China throughout the latter part of the ninth and most of the tenth centuries. Following the simultaneous collapse of the Tibetan and Uighur empires around 840, new bands of homeless marauders overran Ho-hsi and the Ordos. One group of Uighurs eventually settled in Kan-chou, initially under the sponsorship of local Tibetan headmen. Here a new Uighur king13 HTS, 221 A, p. 6217; THY, 98, p. 1757. 14 Friedland, "A reconstruction of Early Tangut history," p. 258. 15 Okazaki, Tanguto, pp. 71—5. Cambridge Histories Online © Cambridge University Press, 2008 162 THE HSI HSIA dom took root and gradually extended its sway in the tenth century over the neighboring oases of Su-chou and Kua-chou. Tibetan influence in Ho-hsi was reduced still further following the formation of a T'ang loyalist army at Sha-chou by a local Chinese, Chang I-ch'ao. Given the title of the Kuei-i Chun ("Return to allegiance army") by a grateful court about 851, Chang's local regime survived into the tenth century, when the Sha-chou Kuei-i Chun became an autonomous Chinese outpost in Ho-hsi under the hereditary rule of the Ts'ao family. Battles were waged between the Sha-chou regime and its Uighur rivals at Kan-chou. Early in the tenth century, however, the two groups reached a modus vivendi that served to keep open most of the time the trade routes from Central Asia to China.l6 The Uighurs also maintained friendly relations with the Tibetan-dominated regime at Liang-chou, and Sha-chou courted close ties with the Buddhist kings of Khotan (Yii-t'ien), for both regimes remained hostile to the Tibetans. For all their differences, Khotan, Sha-chou, Kan-chou, and Liang-chou had a mutual interest in the transit trade through Ho-hsi to the Tarim basin and beyond that provided them with much of their wealth. These interests were threatened by the growth of Tangut influence and ambition to control the transit trade. Although the rulers of China had lost control over Ho-hsi, their need for its horses did not diminish. On the contrary, their reliance on the import of horses from this region was pivotal in their dealings with the Tanguts. Before the fall of the Uighur khaghanate in Mongolia, Uighur merchants and envoys to China traveled a route that passed from the Yin-shan southward across the Yellow River and through the Ordos to Hsia-chou and thence to the T'ang capital. Their passage through Tangut territory threatened the Tanguts' established position in the horse trade with China. Sometime after 840 the frontier horse trade route shifted westward to pass through Lingchou, and the Tanguts and Uighurs became bitter rivals in the Ho-hsi trade. Tangut tribes dwelling along the Ling-chou route menaced imperial envoys and plundered Uighur caravans, selling the booty to other tribes. Given Ling-chou's importance to the Chinese court, notably as its principal horse procurement center, Tangut raids often prompted the dispatch of imperial troops to secure the route. The troubles besetting the T'ang dynastic house at the end of the ninth century boosted the ambitions of the P'ing-hsia Tanguts in a tangible way. 16 On the Kan-chou Uighurs and Ho-hsi in the ninth and tenth centuries, see J. R. Hamilton, Les Ouighours a I'ipoque da cinq dynasties d'apris Its documents chinois (Paris, 1955); Elizabeth Pinks, Die Uiguren von Kan-chou in der friihen Sung-Zeit (Wiesbaden, 1968); Maeda Masana, Kasei no nkishichirigakuteki kenkyu (Tokyo, 1964), pp. 333—62; and the reassessment by Moriyasu Takao, "Uiguru to Tonko," in Tonko no rekisi, vol. 2 of Koza Tonko, ed. Enoki Kazuo (Tokyo, 1980), pp. 3 0 5 - 3 0 . Cambridge Histories Online © Cambridge University Press, 2008 THE END OF THE T'ANG 163 After the Huang Ch'ao rebellion broke out in 875, north China lost contact with Ho-hsi. In 873 the senior Tangut leader at Hsia-chou, T'o-pa Ssu-kung, had already occupied Yu-chou (east of Ching-pien, Shensi) and declared himself its prefect. When Ch'ang-an fell to Huang Ch'ao at the end of 880, T'o-pa Ssu-kung led his newly raised Chinese-Tangut army to assist the loyalist forces in driving the rebels from the capital. For this service he was rewarded with the appointment in 881 as the acting military governor of the Hsia, Sui, and Yin prefectures, the former incumbent having defected to the rebels. Soon the acting title was made permanent, and in 882 his province was renamed the Ting-nan army, embracing the four prefectures of Hsia, Sui, Yin, and Yu. Later it also included Ching-pien, which became Ching-chou in the Wu-tai period.17 T'o-pa Ssu-kung continued to assist the imperial coalition, now under the command of his rival, the Sha-t'o Turkish leader Li K'o-yung. After Huang Ch'ao's defeat in 883, the court awarded the Tangut general several more high titles, conferred upon him the dynastic surname Li, and enfeoffed him as duke of the Hsia state (Hsia kuo kung). The Tanguts and Sha-t'o had first crossed paths in 847, when the Sha-t'o chieftain Chu-yeh Ch'ih-hsin (later named Li Kuo-ch'ang) supplied the vanguard for an imperial punitive campaign against a Tibeto—Tangut—Uighur raiding party at Yen-chou. Chu-yeh Ch'ih-hsin later earned the imperial surname for his role in suppressing the P'ang Hsiin rebellion in 869. In the following decade he warred with border tribes to strengthen his position at Chen-wu, northeast of Hsia-chou. At least once, in 878, he attacked the Tanguts. A stalemate resulted, broken finally by the intervention of the Tanguts' old allies, the T'u-yii-hun. Eventually, Li K'o-yung, Li Kuoch'ang's son, crushed his T'u-yii-hun opponents and extended his control over northern Shansi.18 Meanwhile, the T'o-pa clan carefully expanded its hold on the Ordos under the T'ang aegis, avoiding unnecessary confrontations or compliance with Li K'o-yung's powerful Sha-t'o organization. When T'o-pa Ssu-kung died in about 895, his brother Ssu-chien inherited the command and acquired several others as well. Two other brothers later served as military governors, but the second of these surrendered to Wang Chien, the military governor of Hsi-ch'uan (western Szechwan) who in 907 founded the state of Former Shu. By doing so he and his followers opted to leave the Ordos and return to an area bordering the Tanguts' ancient homeland.19 17 HTS, 221 A, p. 6219; Okazaki, Tanguto, pp. 79-84. 18 Sui and Tang China, pp. 700, 759, 785-6; Okazaki, Tanguto, pp. 78-9; G. M0I6, The T'u-yii-hun from the Northern Wei to the time of the Five Dynasties (Rome, 1970), pp. 195-206. 19 Okazaki, Tanguto, p. 88. Cambridge Histories Online © Cambridge University Press, 2008 164 THE HSI HSIA THE TANGUTS DURING THE WU-TAI PERIOD In 907 the T'ang dynasty formally ended with the founding of the Later Liang (907—23) by Chu Wen, the chief contender for the domination of north China. Li K'o-yung remained ruler of a powerful independent regime in Shansi and in 903 had entered into an alliance with the Khitans, thus confronting Hsia-chou with hostile neighbors to the east. Necessity drew the Tanguts into an alliance with the Later Liang, and subsequently with any faction opposed to the Sha-t'o—Khitan axis. When T'o-pa Ssu-chien died in 908, the succession passed to his adopted son Li I-ch'ang, the grandson of T'o-pa Ssu-kung. A year later Li I-ch'ang died in a mutiny in the Hsia-chou army, and the succession passed out of T'opa Ssu-kung's line into a collateral branch headed by Li Jen-fu, an uncle of Ich'ang. Li Jen-fu was a popular officer in the Tangut army at Hsia-chou and was elected by his peers to the military governorship. Shortly thereafter, Hsia-chou withstood a month-long siege by the combined forces of Li Maochen, the military governor of Feng-hsiang and Ho-hsi, and Li Ts'un-hsii, Li K'o-yung's son and later the founder of the Later T'ang (923—37), who ruled a powerful independent state in Shansi. A Liang relief force lifted the siege, and Li Jen-fu was rewarded with high titles. In 922 Li Jen-fu sent five hundred horses to Lo-yang, perhaps responding to a Liang appeal for aid against the Sha-t'o. During this period the Liao founder A-pao-chi campaigned to subjugate the tribes, including some Tanguts, inhabiting the southwestern border of Khitan territory. Some Tangut groups began to submit tribute to the Liao court, but there is no evidence of any formal contacts at this time between the governors of Hsia-chou and the Khitans. Li K'o-yung's early friendship with A-pao-chi had long lapsed by 923, when the Sha-t'o Later T'ang replaced the Liang as the dominant power in north China. Faced with a growing Khitan menace, the Tangut leader Li Jen-fu perforce acknowledged the new Later T'ang regime, receiving in return confirmation of his post, more titles, and, in 924, enfeoffment as prince of Shuo-fang {Shuo-fang wang) and other emoluments.20 Despite this surface cordiality, the Sha-t'o suspected, on the basis of apparently groundless rumors, a secret alliance between Li Jen-fu and the Khitans. Hence, when Li Jen-fu died in 933, the Later T'ang court attempted to establish direct control over the Hsia-chou area. Li Jen-fu's son and successor, Li I-ch'ao, was ordered to take up the post of military governor of Chang-wu at Yen-chou, and the Sogdian incumbent governor of Chang-wu, An Ts'ung20 Hsiieh Chii-cheng ( 9 1 2 - 8 1 ) , et al., eds., Chiu Wu-tai ibih (Peking, 1976), 132, pp. 1746-9; Ouyang Hsiu (1007—72), comp., Hsin Wu-tai ibih, (Peking, 1974), 40, pp. 4 3 6 - 7 ; Okazaki, Tangito, pp. 1 3 5 - 4 0 . Cambridge Histories Online © Cambridge University Press, 2008 THE TANGUTS DURING THE WU-TAI PERIOD 165 chin, was to replace him at Ting-nan (Hsia-chou). Naturally such a reassignment seriously threatened Tangut interests and, as the Later T'ang court had feared, aroused fierce resistance. Li I-ch'ao reported that he was "prevented" from departing Hsia-chou. An Ts'ung-chin advanced toward his new assignment with an escort of fifty thousand soldiers, who proceeded to besiege the heavily fortified citadel at Hsia-chou. The Tanguts mounted an effective defense, enlisting the aid of tribal confederates to strip the surrounding countryside bare and cut the attackers' supply routes. After more than three months, the Later T'ang troops were forced to withdraw. Li I-ch'ao then conveyed his repentance to Lo-yang and was duly confirmed as the military governor of Ting-nan. In gratitude he presented fifty horses to the Later T'ang court.21 Hsia-chou continued to flourish, politically and economically. The source of Tangut wealth was, of course, livestock, particularly the sale of horses to north China. The Wu-tai annals comment on T'ang Mingtsung's (r. 926-33) dilemma in trying to reduce the vast sums lavished on unending streams of foreign horse traders traveling to and from the capital, without jeopardizing the adequate supply of mounts for his army. Uighurs and Tanguts figured foremost among the horse dealers in Lo-yang. The Tanguts repaid Ming-tsung's ample hospitality with drunken singing bouts, regaling the Sha-t'o monarch with anecdotes from the Ordos. In 929 the T'ang court announced that in future it would conduct all horse transactions at the border markets and prohibit foreign merchants from coming to the capital. But the plan failed, and "the sheep and horses of the tribes did not cease [to fill] the roads."22 In addition, other Tangut tribes in the Ling-chou and Ch'ing-chou area continued to make a livelihood by intercepting and plundering the caravans of the Uighurs from the west. In 932 a Later T'ang army under Yao Yench'ou, military governor of Pin-chou (Pin-hsien, Shensi), punished these tribes for killing a Uighur envoy. Yao Yen-ch'ou also led the army that accompanied An Ts'ung-chin on his attempt to take up his post at Hsia-chou the following year, 933. Possibly, Later T'ang fears of collusion between the Tangut and the Khitan in the raids around Ling-chou encouraged their attempt to displace the Li (T'o-pa) clan at Hsia-chou in 933. It failed and left Hsia-chou more independent of China, yet still in a delicate position between the raiding and feuding Ordos tribes, over whom it clearly desired to increase its influence, and the greater powers beyond, that is, north China's slowly strengthening central government and the burgeoning Khitan empire in the northeast. 21 Okazaki, Tanguto, pp. 141—55. 22 Wang P'u (922—82), Wti-tai hui yao (Shanghai, 1978), 29, pp. 462—4; Chiu Wu-tai shih, 138, p. 1845; Hsin Wu-tai shih, 74, pp. 912—13. Cambridge Histories Online © Cambridge University Press, 2008 166 THE HSI HSIA Li I-ch'ao died in 935 and was succeeded by his brother (the sources disagree as to who was the elder), I-yin (his name was later changed to I-hsing, to observe the taboo on the Sung founder's personal name). Li I-yin governed Hsia-chou for thirty-two years, until his death in 967. What is known of his long tenure suggests some patterns both for the future development of the emerging Tangut state and for Sino-Tangut relations. Discord within the Li clan led to open rebellion in 943. Li I-min, the prefect of Sui-chou and younger brother of Li I-yin, plotted to overthrow his older brother. When the plot leaked, Li I-min fled with his allies (including another brother) to Yen-chou, a Chinese-controlled prefecture 340 /;' south of Sui-chou. Complying with the Hsia-chou chief's request for the rebels' extradition, the Chin court (936-46) directed Yen-chou to return Li I-min's party to Hsia-chou for execution. Over two hundred discontented clan members were implicated, resulting in a purge of the inner ranks. Li I-yin transferred the loyal official Li Jen-yu from his prefectural post at Yu-chou to replace Li I-min at Sui-chou. Not long afterward the Yeh-mu, a Ch'iang group settled in the area, killed Li Jen-yii and absconded southward into Chinese territory. In 948 Li I-yin requested permission of the Han court (947—50) to cross the frontier and punish the Yeh-mu. The court refused.23 No causes are cited for either the earlier rebellion or the Yeh-mu's grievance against Li Jen-yii. A few years later, in 952-3, the Yeh-chi (or Yeh-tse) people north of Ch'ing-chou revolted against the depredations of the Ch'ing-chou prefect. The court's attempt to pacify them enlarged the turmoil and exposed the corruption of its local officials. Here, as so often was the case, the conflict turned on the comparative prosperity of the tribal peoples, who evidently had more, particularly more to eat, than did the sedentary communities of devastated, war-torn north China. Troops sent out by the court often had to feed themselves and therefore sometimes were allowed to keep the booty of socalled pacification campaigns.24 In regard to the Khitans, the Hsia-chou Tanguts evinced a cautious attitude. When Liao armies prepared to march south in 944 against the Khitaninstalled Chin regime, Li I-yin promised the Chin court to send forty thousand troops eastward across the Yellow River into Khitan territory to create a diversion. Whether or not he actually did so, the court gave him the title of Khitan southwest pacification commissioner. It is not clear what prompted Li I-yin to send troops to the border of Yen-chou (west of Hsia-chou) in 948, ostensibly in answer to a secret summons for aid from a rebellious military governor. Perhaps one reason 23 Chiu Wu-tai shih, 132, p. 1749; Okazaki, Tanguto, pp. 157—9. 24 Him Wu-tai shih, 74, p. 913; Wu-tai huiyao, 29, pp. 354-55. Cambridge Histories Online © Cambridge University Press, 2008 THE TANGUTS DURING THE WU-TAI PERIOD 167 was rivalry with his enemy Kao Ch'ung-ch'iian, Chang-wu military governor at Yen-chou. Li's troops withdrew upon the approach of an imperial force, but this was the second time that year that Hsia-chou had flexed its muscles (compare the Yeh-mu incident). To appease the Tanguts, in 949 the Han court enlarged the Ting-nan command by the addition of Chingchou and gave Li I-yin the honorary title of chung-shu ling (director of the Secretariat): Hsia-chou responded with a gift of horses to the Han court. After the short-lived Han dynasty fell, Li I-yin negotiated with both the rival successor courts of Northern Han (951-79, at T'ai-yiian) and Later Chou (951-60) but in the end formally recognized the Chou and distanced himself from the Northern Han, who were closely linked with the Khitan. The Chou wooed Li I-yin by granting him the title of commandery prince of Lungrhsi (Lung-hsi chun wang), which had been first held by Li Jen-fu, and in 954 enfeoffed him as prince of Hsi-p'ing {Hsi-p'ing wang).1* Subsequently, relations remained cordial, between Hsia-chou and both the Chou and early Sung courts. In 962 the Tangut chieftain sent three hundred horses to K'aifeng and supported the Sung's repeated efforts to destroy Northern Han, which finally succeeded in 979. Prominent now among the local forces involved in Hsia-chou politics were the independent and generally hostile Tangut tribes living in Lin-chou and Fu-chou prefectures northeast of Hsia-chou, in territory once dominated by the Sha-t'o. Their leaders, the Che clan, were ancient rivals of the T'o-pa and later faithfully served the Sung court in thwarting the Hsia state. Blood feuds simmered between the two sides until the mid-twelfth century, when after the fall of the Northern Sung the Jurchen Chin court transported the remaining members of the clan far away to Shantung, in the interests of harmony with Hsia. 26 Upon his death in 967, Li I-yin was given the posthumous title of Hsia king {wang) by the Sung court. The Sung emperor confirmed his son Li Kuang-jui (K'o-jui) in office, despite its court's decision in 963 prohibiting non-Chinese from serving as garrison commanders {chen chiang) along the northwestern border in Shensi.27 Nevertheless, the balance of power was shifting in favor of the central government, and the loss of the buffer state between Sung and Liao along the Tanguts' eastern frontier when the Northern Han kingdom was destroyed in 979 made their position far more vulnerable. These events provoked a crisis in the Li clan that came to a head in the succession problem of 981—2. 25 Chiu Wu-tai thih, 132, pp. 1 7 4 8 - 9 ; Okazaki, Tanguto, pp. 1 6 1 - 3 . 26 T'o-t'o et al., eds., Chin shih (Peking, 1975), 128, p. 2761 (hereafter cited as CS); Han Yin-cheng, "Lin Fu Chou chien chih yii Che shih yuan liu," Ning-hsia she hut k'o hsiieb, 1 (1981), pp. 63—7. 27 To-t'o et al., eds., Sung shih (Peking, 1977), 1, p. 14 (hereafter cited as SS). Cambridge Histories Online © Cambridge University Press, 2008 l68 THE HSI HSIA THE TANGUT MOVE TOWARD INDEPENDENCE, 982-IOO2 Li K'o-jui died in 978, and his heir survived him scarcely a year. Because the latter's son was still an infant, his brother Li Chi-p'eng assumed the military governorship, apparently without the consent of other clan leaders. This caused discord among the Li elders, who were split in their attitudes toward Sung authority. Two opposing camps emerged. The Tangut prefect of Suichou forced Li Chi-p'eng into surrendering to the Sung court, by contesting the propriety of his succession and appealing for imperial intervention. His motives are not clear, but as a result Li Chi-p'eng had to accompany the Sung envoy dispatched to Hsia-chou back to K'ai-feng, and at court he formally handed over the four prefectures of Hsia-chou, Sui-chou, Yin-chou, and Yuchou (in some accounts, Ching-chou is also included) to a delighted Sung T'ai-tsung. The Sung thus achieved peacefully in 982 what the Later T'ang had failed to gain by force in 933. Li Chi-ch'ien, a cousin of Chi-p'eng renowned for his bravery and martial skills, was born at Yin-chou in 963 and grew up in the service of an elder kinsman who evidently became his benefactor after his father's death. This benefactor is said to have died in an abortive attack on Li Chi-p'eng at Hsiachou in 981. When Sung officials proceeded to round up the clan members for resettlement in the interior of China, Li Chi-ch'ien fled north with his followers into the P'ing-hsia pasturelands, where he renewed his tribal connections. Thus the ruling Tangut clan was purged of its more sinified elements, most of whom moved willingly into China, leaving the guardians of the T'o-pa tribal legacy to establish an independent Ordos state. Li Chi-ch'ien made his first base camp at Ti-chin Marsh, three hundred li northeast of Hsia-chou. Having raised a force of twenty thousand men, he began at once to harass the Yin-chou and Hsia-chou prefectures, where the imposition of Sung control had stirred up tribal unrest.28 In the northern Ordos and across the Yellow River several large Tangut tribes lived a nomadic existence. They were accustomed either to send "tribute" horses to the Sung court in exchange for gifts of tea or silk or to block and harass the Sung horse convoy route, at first in collusion with the Khitan and later as partisans of Li Chi-ch'ien. A number of their leaders, however, sought to maintain ties of fealty with Sung, primarily for economic advantage and prestige, but no doubt also to enhance their position vis-a-vis Li Chi-ch'ien, who was pressuring them to join his cause and break with Sung. To do this would have meant 28 SS, 485, pp. 13984—6; Li T'ao (1115-84) et al., comps., Hsu tzu chib t'ung Men cb'ang pirn (Chekiang, 1881; repr. Taipei, 1961), 23, pp. 7a, 14a, 16a; 24, pp. 20b—21a, 22b (hereafter cited as HCP); Okazaki, Tanguto, pp. 173—97. Li Chi-p'eng has an entry in Sung biographies, ed. Herbert Franke (Wiesbaden, 1976), vol. 2, pp. 5 2 1 - 2 . Cambridge Histories Online © Cambridge University Press, 2008 THE TANGUT MOVE TOWARD INDEPENDENCE 169 refusing to sell horses to Sung agents, which would have been directly contrary to their own interests. More troublesome still was the more numerous, heterogeneous TibetoCh'iang population scattered along his southern frontier with Sung, including the important Nan-shan Tanguts. Riven by internal divisions and conflicting loyalties, this population tended to resist all external controls. Many chieftains regularly offered armed auxiliaries to the Sung in return for gifts generously bestowed to win either their neutrality or their assistance against Li Chich'ien.2' Such tenuous allegiances, however, frequently lapsed and often had to be renewed by force. This worked to the Tanguts' advantage. Nevertheless, considering Li Chi-ch'ien's vastly inferior resources when compared with those of the wealthy and powerful Sung state, it is remarkable that he succeeded in welding such fragmented and volatile groups into a solid following. Li Chi-ch'ien's strategy to secure the allegiance of the population of P'inghsia and Nan-shan was to present his clan as their protectors against an oppressive Chinese border administration and at the same time to offer them marital alliances and the lure of power and plunder. He himself chose a wife from the powerful Nan-shan clan of Yeh-li, which thereupon became one of the leading "inner" clans of the early Hsia state. Yeh-li consorts and their clansmen acquired enormous influence and, for a while, occupied a position roughly equivalent to that of the Hsiao consort clan among the Khitan.3° Another marriage alliance was established with the Khitan dynastic house. In 986 Li Chi-ch'ien offered his submission to the Liao court and petitioned for an imperial bride. He received the princess in 989 and was granted the title of king of Hsia (Hsia kuo wang) by the Liao emperor in 990. 3 ' This new alliance was at best an uneasy one. For the Tanguts, submission to Liao never barred negotiation with Sung, and the triangular relationship that developed bristled with mutual suspicion and squabbles. Moreover, Tangut tribes residing inside the Liao border became a perennial source of friction between the two states. After concluding a satisfactory treaty with Sung in 1004-5, t n e Khitans were free to contemplate the threat posed by Tangut competition for control of the Ho-hsi trade and dominance over their Uighur clients. Still, the nominal Khitan alliance served to maintain a balance of power useful to all sides. Certainly at this time Sung was in no position to antagonize the Tanguts or to commit many troops to the Ordos frontier. Instead it relied on suasion, 29 SS, 491 and 492, the chapters on the Tang-hsiang and T'u-po, respectively; Lo Ch'iu-ch'ing, "Sung Hsia chan cheng te fan pu yii pao chai," Ch'ung-chi hsiiehpoo, 6 (1966—7), pp. 223—43. 30 Olcazaki, Tanguto, pp. 1 9 1 - 2 ; Karl A. Wittfogel and Feng Chia-sheng, History of Chinese society, Liao (907—1/25) (Philadelphia, 1949), see esp. sec. 7, "A third culture," p. 20. 31 Olcazaki, Tanguto, p. 199; T'o-t'o et al., eds., Liao shih (Peking, 1974), 115, pp. 1524-5 (hereafter cited as LS). Wu, Hsi Hsia shih kao, pp. 92—9, reviews Tangut—Khitan relations. Cambridge Histories Online © Cambridge University Press, 2008 I7O THE HSI HSIA economic pressure, and intimidation. By fortifying strategic border positions, by "pacifying" rebellious non-Chinese groups living within the border under Sung jurisdiction, and by encouraging tribute-trade contacts with the Tibetans and Uighurs, it easily exploited Tangut tribal divisions. Throughout, the two sides waged a kind of guerrilla warfare. Since 983 the Sung had restricted border trade and had substituted goods for cash in purchasing horses, so as to reduce the flow of metal to the Tanguts, who melted down coin to make weapons. In 993 the court attempted to ban the sale of superior Tangut salt along the border, hoping to impoverish Li Chi-ch'ien's partisans and induce them to surrender. But the ban provoked so much antagonism and evasion that it was promptly rescinded, although later it was nominally reimposed.'2 Although the Sung thus enjoyed an economic advantage, local punitive campaigns against uncooperative border tribes greatly reduced the attractions of serving the Sung. Li Chi-ch'ien's strategy emphasized diplomacy, military strength, and various compensations for the Sung trade blockade. As early as 984 he tested the Chinese mood with an offering of horses and camels at Lin-chou. Shortly thereafter a Sung force raided his camp at Ti-chin, and he barely eluded capture. In 985, however, he retook Yin-chou, commencing the reconquest of traditional T'o-pa territories in Ting-nan. With the aid of his Chinese adviser, Chang P'u, Li Chi-ch'ien recruited Ordos Chinese and tribal chieftains to posts in the rudimentary government that he set up in the reconquered areas. Chang P'u later spent several years as a "guest" of Sung T'ai-tsung, but that emperor failed in all his efforts to induce Li Chi-ch'ien's surrender. Unquestionably the Tangut chief's foremost concern was to establish a stable resource base. Illegal trade flourished despite Sung prohibitions, and in 1002 Li Chi-ch'ien opened his own seasonal fairs along the routes to Lingchou, attracting many customers. But the P'ing-hsia region had become a political and economic backwater after decades of warfare and unrest. Even before the Chinese razed the ancient citadel at Hsia-chou in 994, Li Chich'ien's armies were pressing westward toward Ling-chou and southwest to Lan-chou. By 1001, Tangut troops, now numbering fifty thousand, had succeeded in cutting Ling-chou's lifeline to the interior. In the third month of 1002 the city fell to the Tanguts, becoming their first capital. The surrounding area along the Yellow River provided a fertile foundation for the expanding Hsia state" (see Map 10). Ling-chou was renamed Hsi-p'ing Fu in 1003. Restoration of the canal system began at once, and Ordos residents were forcibly resettled in the new 32 HCP, 24, p. 20b. Liao Lung-sheng, "Sung Hsia kuan hsi chung te ch'ing pai yen wen c'i," Sbih huo yiieh k'an, 5, (1976), pp. 1 4 - 2 1 . 33 HCP, 5 1 , p. 5a; SS, 485, p. 13988; Okazaki, Tanguto, pp. 2 0 5 - 1 5 . Cambridge Histories Online © Cambridge University Press, 2008 Huang t'ou Uighurs Original nucleus of Hsia-chou Territory conquered 995—1005 Territory conquered 1028—50 Territory occupied 1063-81 Territory occupied 1136-7 300 km _J | 200 miles MAP I O . The growth of the Hsi Hsia state Cambridge Histories Online © Cambridge University Press, 2008 172 THE HSI HSIA capital to enlarge its farming population and tax base. That same year the Sung court recognized the fait accompli by formally returning to Li Chi-ch'ien the five Ordos prefectures surrendered by Li Chi-p'eng in 982. Meanwhile, another front had opened at Liang-chou, five hundred /*' southwest of Ling chou, the hub of tribute-trade traffic and horse commerce between the Kansu corridor and K'ai-feng. LIANG-CHOU AND TANGUT EXPANSION INTO HO-HSI For many years Liang-chou had been governed by a confederation of mixed Tibeto-Chinese tribes residing in the well-watered valleys, called the Liu-ku, of the mountains to its south. The confederation was divided into left and right wings under the nominal leadership of a great chieftain. In 1001 the fourth-generation incumbent of this position was abruptly displaced by P'anlo-chih,34 an obscure but influential figure. P'an-lo-chih's rise to power has been attributed to a supposed connection with the powerful rLangs clan of Sum-pa, which ruled P'an-chou (modern Sung-p'an, Szechwan) and had supplied high ministers to the Tibetan monarchs.35 Backed by the thirteen Che-lung tribes, another group associated with Liang-chou, P'an-lo-chih's elevation to the leadership of Liu-ku probably resulted from the support of the Sung authorities in Ch'in-chou (modern T'ien-shui, Kansu), who may well have viewed him as a stronger deterrent to Li Chi-ch'ien than his predecessor had been. Nevertheless, the previous Liu-ku chieftain remained as leader of the left wing and, characteristically, also received Sung backing as a counterweight to P'an-lo-chih. The commercial and strategic value of the triangle formed by Hsi-ning, Lan-chou, and Liang-chou figured in Tangut thinking as early as 985, when Li Chi-ch'ien sent troops to attack Hui-chou (on the east bank of the Yellow River, northeast of Lan-chou). Tangut attacks on Liang-chou began in 996 or earlier and intensified after the capture of Ling-chou. Late in 1003 Li Chich'ien occupied Liang-chou, accepting the feigned surrender of P'an-lo-chih, who had fled the city. A Tibetan ambush ensued, however, in which the Tanguts were routed and their leader was mortally wounded. He died early in 1004 near Ling-chou, at the age of forty-one.36 34 This name is given the reading Po-lo-chih by some Chinese scholars. 35 On Liang-chou, see Maeda, Kasti, pp. 383—99; Hsu Sung (1781—1848), ed., Sung bui yao chi kao (Peking, 1936; repr. Peking, 1957; Taipei, 1975), 195, fang yii 21, pp. i 4 - 2 3 b (pp. 7654-8); Iwasaki Tsutomu, "Seiryofu Hanrashi seiken shimatsu ko," Tohogaki, 47 (1974), pp. 25—41, and "Seiryofu seiken no metsubo to Soka zoku no hatcen," in Suzuki Shun sense: koki kinen Toyoshi ronso, ed. Suzuki Shun sensei koki kinen Toyoshi ronso henshu iinkai (Tokyo, 1975), pp. 7 3 - 8 8 . 36 On discrepancies in the dating of Li Chi-ch'ien's death, see Tai Hsi-chang's commentary in his Hsi Hsia chi (Peking, 1924), in vol. 4 olChung bua wen shih ts'ung sbu, ed. Wang Yiin-wu (Taipei, 1968), 3, pp. 20b—21a. Cambridge Histories Online © Cambridge University Press, 2008 THE RISE OF THE TIBETAN TSUNG-KO DYNASTY 173 Revenge soon followed. A group of Tanguts secretly joined with a Chelung splinter group and assassinated P'an-lo-chih in the middle of 1004. Thereupon the Che-lung confederation collapsed. Its loyalist faction expelled the others and elected P'an-lo-chih's youngest brother, Ssu-to-tu, to govern Liang-chou. Ssu-to-tu promptly gained Sung recognition, but his power base was considerably diminished. Plague further afflicted the area for years following 1006. Several eminent Liu-ku leaders reportedly defected to the Tanguts while other tribes gravitated toward a new locus of Tibetan authority that was forming farther south around the Tsung-ko tribes in the Huang River valley east of Kokonor at Ho-chou. THE RISE OF THE TIBETAN TSUNG-KO DYNASTY Tsung-ko designates the region of Amdo east of Kokonor (in modern eastern Tsinghai Province) (see Map 11). It was also the name of a town (Tsung-koch'eng) located midway between modern Hsi-ning (then called Ch'ing-t'ang) and Le-tu (then called Miao-ch'uan), and of a river (Tibetan: Tsong-kha). The Tsung-ko tribes first appear in Sung sources as the allies of the loyal Che-lung faction at Liang-chou." After P'an-lo-chih's death they consolidated an alliance with the Kan-chou Uighurs to protect their mutual commercial interests. The Tsung-ko escorted Uighur merchant-envoys along a route that bypassed Liang-chou far to the south and traversed Hsi-ning to reach Ch'inchou in Sung territory. Around this time, partly to counter the growing Tangut pressure on the region, the Amdo elders cast about for a new overlord of suitably venerable lineage. To the west, in a district of Kao-ch'ang, they found such a person in Ch'i-nan-ling Wen-ch'ien-pu (997-1065), whose alleged descent from the Yarlung royal house gave him a stronger claim to Tibetan loyalties than P'anlo-chih had proffered. Moreover, he was soon endowed with religious qualities appropriate to his royal station. Ch'i-nan-ling was first escorted to Hochou, where the local people named him Ku-ssu-lo,38 meaning "son of Buddha" (Tibetan: rgyal-sras) and aspired to establish him as head of a royal government. But an ambitious Tsung-ko monk, Li-li-tsun (or Li-tsun), and the Miao-ch'uan chieftain, Wen-pu-ch'i, abducted Ku-ssu-lo and took him northwest to K'uo-chou, where, around 1008, he was enthroned as king (Tibetan: btsan-po) at the age of twelve. As the new regime flourished, it was 37 Iwasaki, "Seiryofu seilcen no metsubo," pp. 73-88; Maeda, Kasei, pp. 505—9, 575-7; see Luciano Petech, "Tibetan relations with Sung China and with the Mongols," in China among equals: The Middle Kingdom and its neighbors, ioth—14/h centuries, ed. Morris Rossabi (Berkeley and Los Angeles, 1983), pp. 176—7, for relevant information from Tibetan sources. 38 This name is also read Chiao-ssu-lo by some scholars. Cambridge Histories Online © Cambridge University Press, 2008 THE HSI HSIA 174 300kr lobmilsis j. ~C|Ving:rang (Hsi-ning) MAP I I . The Tsung-ko region moved again to Li-li-tsun's seat of power at Tsung-ko-ch'eng. There Li-litsun made himself chief minister and married two (or one?) of his nieces (some accounts say daughters) to Ku-ssu-lo. In 1014 Ku-ssu-lo contacted the Sung authorities at Wei-chou (Kansu, P'ing-liang) and was granted an official appointment. Early in 1015 a joint embassy from the Tsung-ko leaders reached the Sung court. Already, reports from Ch'in-chou and Wei-chou officials had indicated that the Tsung-ko Tibetans boasted of sixty thousand to seventy thousand troops willing to oppose Tangut advances with the court's support but that on the other hand, they were also inciting border tribes to unite and resist Sung penetration of their pasturelands in the Wei River valley. Furthermore, a power struggle was coming to a head between Li-li-tsun and his protege. By 1016 Li-li-tsun had shed his clerical status and had taken wives from sixteen clans. He had repeatedly petitioned the Sung court for enfeoffment and even requested the Cambridge Histories Online © Cambridge University Press, 2008 THE RISE OF THE TIBETAN TSUNG-KO DYNASTY 175 title ofts'an-p'u (Tibetan: btsan-po) in defiance of Ku-ssu-lo's presumed authority over him. Distrustful of Li-li-tsun but believing that he might prove useful, in 1016 the court gave him, too, a nominal appointment. Between 1013 and 1016 the Tsung-ko Tibetans quarreled with their Uighur allies and blocked the route to Kan-chou. At this time Tangut troops had stepped up their offensive against the Uighurs and Liang-chou (some sources credit the Uighur soldiers under P'an-lo-chih's command with Li Chi-ch'ien's death). In 1015 a Tangut garrison again occupied that city but was ousted in 1016 by the Uighurs who killed Ssu-to-tu. Liang-chou then returned to joint Tibetan—Uighur control for another fifteen years. Tsung-ko and Kan-chou resolved their differences after a new Uighur khan was elected in 1016, but there is evidence suggesting tension and rivalry between the two peoples over the fate of Liang-chou itself. 39 In 1016—17 both Li-li-tsun and Ku-ssu-lo were implicated in a tribal uprising along the headwaters of the Wei River, provoked in part by the Sung's pushing forward its defense perimeter into this territory, building forts and wood-collecting stations. The unrest culminated in a resounding Tibetan defeat at the hands of Ts'ao Wei (973-1030), the prefect and pacification commissioner of Ch'in-chou, in the ninth month of 1016. Fighting continued sporadically into 1017. By 1018 Ts'ao Wei reported that Tibetan resistance was shattered, that Ku-ssu-lo had retreated, and that all his erstwhile subjects had submitted hostages to Ch'in-chou. According to some reports, Ku-ssu-lo held Li-li-tsun accountable for these reverses. Sometime in the early 1020s, Ku-ssu-lo finally moved to Miao-ch'uan and took its chieftain, Wei-pu-ch'i, as his new minister.40 For many years after this, Ku-ssu-lo's circumstances are obscure. Sometime later the Sung authorities in Ch'in-chou sent agents to Miao-ch'uan to regularize relations with him, in response to his occasional petitions for trade and recognition. This possibly occurred after the fall of Kan-chou to the Tanguts in 1028, although the precise dating of the Tangut conquest of Hohsi remains largely guesswork, according to Sung historian Li T'ao (1114— 83).4I Both Ku-ssu-lo and his minister Wen-pu-ch'i received Sung appointments in 1032. Just at that time, however, Wen-pu-ch'i mutinied and imprisoned his master. Ku-ssu-lo escaped, killed his captor, and reestab39 SeeSS, 41)0 passim, (esp. p. 14115), and Sung hui yao chi kao, 197, fan i 4, pp. 3 - 9 (pp. 7701—4), on the Uighur involvement with Liang-chou and Tsung-ko; see also Iwasaki Tsutomu, "Soka jo Kokushira seiken no seikaku to kito," Chuo daigaku Ajia thi kenkyu, 2 (1978), pp. 1—28. 40 Ku-ssu-lo's and Li-li-tsun's names are, however, still coupled in Sung reports as late as 1024. Li-litsun disappears from the record after 102;, when he successfully petitioned the Sung court for a monthly stipend (despite rumors that he had been killed by Ts'ao Wei's troops in 1016). 41 HCP, i n , p . 17a. Ts'ao Wei has an entry in $»»£ A/V>£ra^/«, vol. 2, pp. 1063-4; and in Ch'ang Pi-te, Wang Te-i et al., comps., Sungjen chuan chi tzu liao suo yin (Taipei, 1974-6), vol. 3, pp. 2197—8. Cambridge Histories Online © Cambridge University Press, 2008 176 THE HSI HSIA lished his headquarters further upstream at Ch'ing-t'ang. To this fortress flocked many refugees from the Tangut annexation of Ho-hsi. After Liangchou fell to the Tanguts, about 1032, the former Tibetan and Uighur allies of Ssu-to-tu also swelled Ku-ssu-lo's following.42 To the extent that is possible to date these events, 1032 stands out as a turning point in the career of Ku-ssu-lo and the rise of Ch'ing-t'ang as a vital commercial and political center between Sung, Hsia, Liao, and Inner Asia. It is also the year in which Li Yiian-hao, the most famous and formidable of the Tangut rulers, came to power, inaugurating an era of direct confrontation between the rival regimes at Ch'ing-t'ang and Hsing-chou (the new Hsia capital). LITE-MING, 1004-1032 We must now look back and retrace events among the Tanguts since the beginning of the century. Li Chi-ch'ien's eldest son, A-i (b. 989), succeeded his father in 1004 at the age of twenty-one and was known thereafter as Li Teming. Intent on pushing the Tanguts' power westward to Kan-chou and beyond, Li Te-ming first moved to secure good relations with both Liao and Sung. In 1004 the new Tangut ruler was invested by the Liao with the title of Hsip'ing wang, which was upgraded in 1010 to that of king of Hsia {Hsia kuo wang), as his father had been styled. However, apart from routine diplomatic exchanges (presumably the Tanguts sent annual tribute envoys to the Liao court), Tangut-Khitan relations began to expose their true adversarial colors. Unrest and rebellion among their northwestern Tsu-pu (Tatar) subjects prompted Liao to field armies against the Kan-chou Uighurs in 1008, 1010, and 1026, in an effort to secure its distant borders. Probably none of these campaigns achieved its intended result in spite of limited local victories. Likewise, Tangut forces also attacked the Uighurs in 1008, 1009, 1010, and thereafter. In 1015 they captured Liang-chou, only to be defeated and driven out by the Uighurs in the next years. It is not likely, however, that Liao and Hsia acted in concert against Kan-chou, in the light of their rivalry for control over Ho-hsi and its trade routes, a control that was of marginal importance to the Liao but was vital to the survival of the Tangut state.43 In 1006 the Liao court had entertained an embassy from the other major Ho-hsi regime, the state centered at Sha-chou (modern Tun-huang), whose ruler Ts'ao Tsung-shou solicited investment from both the Sung and the Liao. 42 Two recent studies of Ch'ing-t'ang and Ku-ssu-lo are Chu Ch'i-yuan, "Ku-ssu-lo cheng ch'iian hsing ch'eng ch'u t'an," Hsi-tsang yen cbiu, 2 (1982), pp. 68—77; and Sun Chii-yiian, "Ch'ing-t'ang lu chi kao," Hsi-tsang yen chiu, 2 (1982), pp. 144—55. 43 Iwasaki, "Sieryofu seiken no metsubo," pp. 79—80; Okazaki, Tanguto, pp. 239—89. Cambridge Histories Online © Cambridge University Press, 2008 LI TE-MING 177 His heir Ts'ao Hsien-shun, the last recorded Ts'ao ruler of Sha-chou, likewise contacted both courts upon his succession in 1014. The fact that around this time Sha-chou rulers began to refer to themselves as the Sha-chou Uighurs, and later in 1041 to 1042 as the Sha-chou Pei-t'ing khaghanate, has prompted speculation that Kan-chou had absorbed Sha-chou before its own conquest. Many Kan-chou Uighurs doubtless resettled in Sha-chou and Kuachou at this time, but it was perhaps more likely to have been Uighurs from the T'ien-shan area who took over Sha-chou in the early 1020s. Moreover, given the long and initmate relationship between the western Uighurs of Turfan and the Khitans, a Uighur presence in Sha-chou would block Tangut ambitions in that region, to Liao's advantage. Indeed, Sha-chou resisted absorption by the Tangut for several more decades.44 Liao did not cultivate relations with the Kokonor Tibetans until the 1050s, but evidently Li-li-tsun did solicit Khitan support, to no avail. In 1018 he asked permission to send tribute envoys through Hsia territory to the Liao court. No such embassies are recorded as having arrived, however, presumably owing to the Tanguts' opposition. Their refusal of Li-li-tsun's request, according to one account, induced the Khitan emperor to lead troops, disguised as a hunting party, on a punitive raid into Tangut territory. Li Te-ming routed the Liao host and the following year was compensated with investment as king of Great Hsia (Ta Hsia), with a gold seal and the honorary rank of director of the Department of State Affairs (Shang shu ling).4' Though its details are dubious, the episode hints at disruptive tensions between the Tanguts and the Liao court. Li Te-ming's efforts to smooth relations succeeded to the extent that he acquired a royal Khitan bride for his son and heir, Li Yiian-hao, upon the accession of the new Liao emperor, Hsing-tsung, in 1031. The conclusion of the treaty of Shan-yuan between Sung and Liao in 1004— 5 and its subsequent smooth operation had created a favorable climate for Li Te-ming's negotiations with China. He may well have delayed his own first contacts with the Sung in part to await the outcome of the Sung-Liao peace negotiations. Both sides were eager to come to terms, despite immediate and irresolvable differences. Although Li Te-ming rejected Sung Chen-tsung's initial demands, which included the return of Ling-chou and the submission of royal hostages to the Sung court (something unprecedented in Tangut history), a compromise was reached, permitting his investiture in 1006 as military governor of Ting-nan with the title Prince of Hsi-p'ing and the granting of silk, cash, and tea and annual donations of winter clothing. Thus formal tributary relations were established and Sung recognized the de facto sovereignty of the Tangut ruler. Throughout his reign Li Te-ming pressed the 44 SS, 490, pp. 14123—4; Maeda, Kasei, pp. 560—70; Moriyasu Takao, "Uiguru to Tonko," in Tonko no nkishi, vol. 2 of Kfiza Tonko, ed. Enoki Kazuo (Tokyo, 1980), pp. 3 3 1 - 5 . 45 LS, 16, p. 183; SS, 485, pp. 13991—2. Cambridge Histories Online © Cambridge University Press, 2008 178 THE HSI HSIA Chinese court for greater trade concessions and, less successfully, for the right to sell Tangut salt across the border. The first government market (ch'iieh ch'ang) was opened at Pao-an commandery (Chih-tan, Shensi) in 1007. Later, in 1026 private markets (ho shih ch'ang) were permitted in Ping-chou (modern T'ai-yiian) and Tai-chou (modern Tai-hsien) in northern Shansi. At this time because Sung emperors were committed to maintaining peace along the borders, minor frontier clashes and flourishing sales of salt and other contraband did not weaken Li Te-ming's favorable standing with the Sung throne. The Tangut royal house prospered as streams of envoys to K'aifeng took full advantage of their trading privileges in the Chinese capital to ensure a steady flow of rich gifts to their monarch. In 1020 a new Hsia capital was built northwest of Ling-chou on the opposite bank of the Yellow River, near the site of Huai-yiian-chen. Named Hsing-chou (this was changed to Hsing-ch'ing fu in 1033), the new center of government enjoyed a superior strategic position between the Alashan, guarding its rear, and the Yellow River, flanking its southern and eastern approaches. By the early twelfth century the Hsia capital was commonly called Chung-hsing, perhaps a Chinese equivalent of its Tangut name. The Mongolian name, Erighaya, was probably also derived from the Tangut name of the capital and was used to refer to the Ning-hsia region.«6 During Li Te-ming's later years his eldest son, Li Yiian-hao, commanded the Tangut armies in Ho-hsi and gained a reputation as a soldier and strategist. Many sources recount alleged conversations between father and son, in which the young prince criticized Li Te-ming's subservience to the Sung and his greed for the Chinese products obtained through border trade. He believed that his father's policies were undermining the pastoral basis of Tangut society and economy, thus threatening their cultural values and no doubt their military prowess. In particular, Li Yiian-hao denounced Li Te-ming's practice of executing trade agents who failed to obtain good bargains in the Sung markets. 47 The conquest of Ho-hsi, which was largely accomplished by Li Yiian-hao, beginning with the capture of Kan-chou around 1028, may be viewed in part as a drive to strengthen the pastoral basis of Hsia society by expanding westward away from the Chinese agricultural periphery, where further expansion was in any case impossible. 46 HCP, 96 , p. 26b; Tai, Hsi Hsia chi, 5, p. 15b: Chung-hsing occurs regularly in CS 61 and 62, and in the Tangut law code (in which the Tangut form transcribes the Chinese), thus casting doubt on Wu Kuang-ch'eng's account of the adoption of this name in 120;; see his Hsi Hsia shu shih (Pref., 1826); repr. in vol. 8 8 - 9 1 of Shih liao ts'ungpien hsii pirn, ed. Kuang-wen shu-chii pien i so (Taipei, 1968), 39, p. 11a. On the Mongolian and Tangut names, see Kychanov, Ocheri istorii, p. 56, and his "O nekotorykh naimenovaniiakh gorodov i mestnostel byvshei territorii tangutskogo gosudarstgva," in vol. 1 of Pis'mtnnyepamiatniki iproblemy istorii i kul'tury naradov vostoka: XI. godichnaia nauihnaia sissiia U IV. AN. SSSR (Ttzisy) (Moscow, 1975), pp. 4 7 - 5 1 . 47 SS, 485, p. 13993; Su Shih (1036—1101), Tung-p'o chih tin (TSCCCP ed.), 3, p. 51. Cambridge Histories Online © Cambridge University Press, 2008 HO-HSI AFTER THE TANGUT CONQUEST 179 Li Yiian-hao's later policies certainly aimed to define and preserve a distinct cultural identity for the Tangut state. Nothing indicates that Li Teming opposed his son's succession, but the latter's ambitions already provoked controversy and concern, as subsequent episodes of violent palace intrigue reveal. HO-HSI AFTER THE TANGUT CONQUEST Chinese records present a confused and incomplete account of the Tanguts' activities in Ho-hsi and tend to collapse the actual time that it took the Hsia to conquer and absorb the region. Its pacification is usually dated to the year 1036. Kan-chou was taken around 1028, Liang-chou in 1032. The westernmost of the Ho-hsi statelets, Sha-chou, however, remained autonomous at least until 1052-3, the year of its last tribute mission to Sung. A Tangut reign date of 1074 appears in a cave inscription at Tun-huang, and so by then Sha-chou was certainly under firm Tangut rule. Maeda Masana suggests that the expansion of Islamic Kharakhanid control over Kashgar and Khotan, the latter of which had traditionally been allied with Sha-chou, precipitated Shachou's submission to Tangut authority. As early as 1038, however, Li Yiianhao had claimed Khotan as his vassal, although Khotan seems not to have accepted this status. After a lapse from 1025 to 1063, Khotanese missions to Sung resumed, and the Khotanese figured prominently in political and commercial dealings at Ch'ing-t'ang.48 Any immediate commercial profits that the Tanguts expected to reap from their occupation of Ho-hsi were undercut when their rivals and prospective clients began to use trade routes passing north and south of the Kansu corridor. Merchants from the Tarim basin either traveled north, skirting the southern Gobi to reach the Khitan court, or detoured southward to Ch'ingt'ang, which developed into a thriving entrepot. The Sung court, likewise, did not ignore the importance of Ch'ing-t'ang as a source of horses, a link with the Western Regions, and also as a potential military threat as a base for joint Tibetan-Tangut operations against the Sung. After 1038 the Sung imported most of their horses from the Tibetans via border markets established in the northwest for this purpose. Court-imposed restraints on Chinese expansion in this area were completely abandoned in the 1070s, when Wang Shao was delegated to establish the Hsi-ho circuit, to "pacify" Kokonor, and to enroll more local chieftains as subsidized clients of the Sung 48 Maeda, Kaiei, pp. 565, 628—4;; Okazaki, Tanguto, pp. 2 7 0 - 9 . Tangut claims to Khotanese suzerainty are recorded 6rst in Li Yiian-hao's letter to the Sung court (SS, 485, pp. 13995-6), and later in Li Yiian's Ch'ing-t'ang lu, a Sung work that survives only in a truncated form in chap. 35 of Sbuofu, comp. T'aoTsung-i (facsim. repro., Taipei, 1972), pp. 11a—13a. Cambridge Histories Online © Cambridge University Press, 2008 l8o THE HSI HSIA court.49 Among other effects, the increasingly belligerent activities of the commanders of Ho-hsi circuit alarmed the other regional powers and brought about a gradual improvement in Tangut—Khitan relations from the late 1070s onward. Lively contacts between Hsia and Kokonor flourished at all levels. In the twelfth century the situation in Ho-hsi was altered significantly. The short-lived Chinese occupation of Kokonor (in 1099 and again from 1104 to 1119) and the Jurchen conquest of Shensi dispersed the commercial network at Ch'ing-t'ang and removed any remaining obstacle to the Tangut annexation of the area by 1136—7, when Chin formally ceded Kokonor to Hsia.'° By this time the Tanguts' trading position had improved considerably, as Hung Hao's (1088-1155) diary reveals: Uighur merchants passing through Hsia territory to Yen (modern Peking, the principal Chin capital in 1153) were obliged to render a 10 percent tax on their goods to Tangut custom officials, who took the best of their wares in payment. No doubt an end to warfare with Sung and the greater stability of Tangut central government from the end of the eleventh century onward contributed to this development. Moreover, according to Hung Hao, after the Jurchen invasion of north China, many Uighurs who had settled in the Ch'in-chou area retreated back into Ho-hsi and became Tangut tributaries.5' By the mid-twelfth century Uighurs had come to play a prominent role in the political and cultural activities of the Hsia state (from the beginning they had been employed as translators of Buddhist texts). Almost certainly, the Buddhist Khitan rulers of Khara-Khitai (western Liao) in Central Asia maintained friendly ties with their Tangut coreligionists, who had loyally supported the Liao dynastic house in its last desperate days. Finally, it is clear from the Tangut law code that in the early twelfth century, regular trade and diplomatic relations existed between Hsia and its neighbors to the north and west. LI YOAN-HAO (WEI-MING NANG-HSIAO, CHING-TSUNG), 1 0 3 2 - 1048 Li Te-ming died in the summer of 1032, shortly after his son had incorporated Liang-chou into the Tangut empire.52 Although Sung accounts of Li 49 Liao Lung-sheng, "Pei Sung tui T'u-fan te cheng ts'e," Tai-wan Sbih fan la bsiieb li shih bsiieb poo, 4 (1976), pp. 143—4. Wang Shaohas an entry in Sung biographies, vol. 2, pp. 1137—41, and in Sung jen chuanchi tzu liaosoyin, vol. 1, p. 203. For an account of his activities, see Ch'en Pang-chan (fl. 1598), comp., Sung shih chi shih pen mo (Peking: 1977), 41. 50 CS, 7 8 , p. 1772; 9 1 , p. 2017; 26, p. 653; but the evidence is obscure and contradictory. See also Wu, Hsi Hsia shu shih, 34, p. 16a—b. 51 Hung Hao, Sung-mo chi wen, repr. in vol. 3 of Liao-hai ts'ung shu, ed Chin Yii-fu (Liao-hai shu-she, 1931—4; repr. Taipei, 1971?), shang, p. 3a. On Hung Hao, see Sung biographies, vol. 2, pp. 464—5; and Sung jen chuan cbi tzu liao suoyin, vol. 2, pp. 1505—6. 52 Sung huiyao chi kao, 33, li 4 1 , p. 12b. Most accounts date his death to the winter, when the report was first received by the Chinese court. Cambridge Histories Online © Cambridge University Press, 2008 LI YOAN-HAO l8l Yiian-hao are heavily colored by rumor and legend, more is known from Chinese sources about this one Tangut chieftain than about all the other Hsia rulers combined. As a youth, Li Yiian-hao studied Chinese and Tibetan texts on Buddhism, law, divination, and military strategy. Upon coming to power he embarked on a bold program of political, social, and cultural reforms designed to strengthen the power of the Tangut ruler and to manifest the distinctive identity of his state. In addition, the proud monarch aspired to make the Sung court recognize him as an equal of the Khitan ruler. But his long, exhausting war with Sung (1039—44) ended only in a partial victory for Hsia, at the expense of the Khitan alliance. Opposition to his centralizing policies (and perhaps also to his alienation of both Sung and Liao) led finally to Li Yuan-hao's assassination and a long devolution of central authority. Yet he was unquestionably the most talented and imaginative of the Tangut rulers and left an enduring imprint on the structure of Hsia government as well as in the popular mind. A sure dating of the new ruler's activities before 1038 eludes historians, but several symbolic acts seemed to figure prominently in his scheme of things. First, he changed the royal surname from Li to Wei-ming (Nwei-mi), a native name. Thereupon all the "inner" lineages of the ruling clan adopted the new surname. The Sung authorities continued to use the Sung royal surname Chao, which their court had bestowed on Li Chi-ch'ien in 991; the Khitans had continued nonetheless to use the old surname Li. Wei-ming Yiian-hao also assumed the title of wu tsu, the Tangut equivalent of emperor or khaghan, which was glossed in Chinese as ch'ing t'ien tzu ("blue son of heaven" or "son of blue heaven").53 Next, he altered the current Sung reign title used in Hsia, in order to observe a taboo on his father's personal name. Soon an independent system of Chinese-style reign titles came into use in Hsia, along with Chinese imperial nomenclature, all having its Tangut equivalents. The most renowned of Wei-ming Yiian-hao's nativistic innovations was the head-shaving decree of around 1034: "When Yiian-hao initially enacted the head-shaving decree [t'u fa ling], he first cut his own hair and then ordered all his countrymen to cut their hair. If within three days [someone] had not followed the order, the crowd was permitted to kill him." The top part of the skull was shaved, leaving a fringe across the forehead and down the sides, framing the face. Variations of this style have been observed among other Asian peoples (Hsien-pei, Khitan, Po-hai). By contrast, the ancient Ch'iang supposedly wore their hair hanging loose over their 53 On Li (Wei-ming) Yiian-hao's name and title, see SS, 485, p. 13993; HCP, m , p. 16b; 122, p. 9b; Wu, Hii Hsia shih kao, pp. 30—3; Nikolai A. Nevskii, Tangutskaiafilologiia(Moscow, i960), vol. 1, pp. 48—9. For a discussion of Inner Asian and Chinese imperial titles, see Peter Boodberg, "Dayan, Cinggis, and Shan-yii," in the Selected works of Peter A. Boodberg, comp. Alvin P. Cohen (Berkeley and Los Angeles, 1979), pp. 85—9. Cambridge Histories Online © Cambridge University Press, 2008 182 THE HSI HSIA face. Thus it has been suggested that the head-shaving decree represented an attempt to reform a "backward" Ch'iang custom and to distinguish Hsia subjects from their Liao, Sung, and T'u-po neighbors.'4 Dress regulations were issued, prescribing costume for civil and military officials and for commoners. Tibetan and Uighur influences have been discerned in the Hsia dress code." Of far greater import was the dissemination of a Tangut script around 1036. Credit for its creation and elaboration is given to the scholar Yeh-li Jen-jung (a kinsman of two prominent brothers, the generals Yeh-li Yii-ch'i and Wang-jung), but work on the script probably had begun under Li Teming and continued over many years. Comprising over six thousand characters, the new script was adopted in government offices and schools, where translations of Chinese and Tibetan works into Tangut began at once. Over one hundred years later, in 1162, Yeh-li Jen-jung was honored with a posthumous ode addressed to "the great teacher, Iri." From the 1040s onward, Hsia communications with the Sung normally included Tangut forms of Hsia noble and official titles, which were then adopted (in Chinese transcription) in the Chinese chancellery and diplomatic documents so as to avoid any suggestion of parity between the Hsia officials and their Sung counterparts.'6 Although Tangut texts of these documents have not survived, at least many titles of offices, whose meanings in most cases remain unknown, have thus been preserved in Sung records. Wei-ming Yiian-hao's military and administrative reforms gave the Tangut state its basic political shape. It is tempting to see in the civilian and military nomenclature of Hsia government an expression of that institutional duality so well known in the Liao state. Certainly Chinese subjects readily found employment in the bureaucracy, whereas the military naturally remained the preserve of the Tangut elite's ruling power. However, the division between civilian and military is not an appropriate basis for analyzing Tangut government, which was not clearly structured along those lines. Moreover, the dualistic elaboration of the Liao government has no obvious counterpart in the Tangut state structure. This does not mean that Liao models did not inspire the architects of Tangut government, but that is a subject yet to be examined properly. 54 HCP, 115, p. 14b; Okazaki Seiro, "Scika no Ri Genko to tokuhatsu rei," Tohogaku, 19 (1939), pp. 77-86. 55 SS, 485, p. 13993; Wang Chung, "Lun Hsi Hsia te hsingch'i," p. 21; Wu, Hsi Hsiasbih kao, p. 205; A. P. Terentyev-Katansky, "The appearance, clothes and utensils of the Tanguts," in The countries and peoples of the East, ed. D. A. Olderogge (Moscow, 1974), pp. 215—24. 56 Wu, Hsi Hsia sbih kao, pp. 203, 215—17. Regarding the invention and inventor of the Tangut script, see Kychanov, Ocheri istorii, pp. 259—62; Nevskii, Tangutsiaia filohgiia, vol. 1, pp. 79—80; Nishida Tatsuo, Seikago no kenkyu: Seikago no sat kosei to Seika moji no kaidoku (Kyoto, 1966), vol. 2, pp. 5 3 9 - 4 0 of the English summary. Cambridge Histories Online © Cambridge University Press, 2008 LI YOAN-HAO 183 The early Hsia military was highly decentralized, drawing on the independent armies (//'«) of tribal chieftains for the bulk of its fighting forces. To strengthen control over his chieftains-general, the Wei-ming ruler issued a set of military regulations, presumably dealing with conscription, discipline, and service rewards. Yet he did not abandon traditional tribal customs of consultation primus inter pares. Before a military undertaking, he summoned the chieftains to participate in hunting exercises: "Whenever a catch was made, everyone dismounted and gathered around a campfire to drink and eat, cutting off fresh meat, at which time [the ruler} queried each as to his views and selected the most astute {of them]."' 7 Other measures reflect the contemporary expansion of the Tanguts' borders and armed prowess. Hsia was divided into twelve military districts called chien chiin ssu, Army Superintendancies (Kychanov translates these as MilitaryPolice Boards): six under the Left Wing (tso hsiang), headquartered east of Hsia-chou and governing the eastern half of the country, and six under the Right Wing (yu hsiang), nominally headquartered at Kan-chou and governing the western half of the country (see Map 12). This typically Inner Asian structure had also characterized the Tibetan organization at Liang-chou, which may have provided its model. Three officers, appointed from the ranks of imperial relatives and tribal elders, headed each of the twelve Army Superintendancies (for their headquarters, see Map 12). Both Chinese and Tanguts served in subordinate capacities. Presiding over the entire structure were the two chiefs of the Right and Left Wings, usually imperial kinsmen or consort relatives, whose power rivaled that of the throne. Inevitably, fatal tensions arose between these representatives of the tribal oligarchy and the Wei-ming monarch. At the height of Wei-ming Yiian-hao's reign, the Tangut army numbered from 150,000 to 300,000 troops, most of whom were assigned to garrison the border and vital interior regions, falling under the jurisdiction of one or another Army Superintendancy. When a decision was made to mobilize, central authorities dispatched messengers with silver tablets (p'ai-tzu) to the requisite general, who summoned the needed troops from lists of registered conscripts. All able-bodied males between the ages of fifteen and sixty were liable for military service.' 8 Acceptance of the silver tablet expressed compliance with the ruler's request for troops, but in addition the ruler often swore an oath with his generals on the eve of battle, as Wei-ming Yiian-hao did in 1038. Apart from the twelve Army Superintendancies, which evolved into the principal organs of regional government, the Tangut ruler restructured and 57 SS, 485, p. 13993; Wu, Hsi Hsia sbih kao, pp. 200—19. 58 On the Hsi Hsia army, see SS, 485, pp. 13994-5; 486, pp. 14028—9; Wu, Hsi Hsia sbih kao, pp. 200-19; Kychanov, Ocherk islorri, pp. 115—32. Cambridge Histories Online © Cambridge University Press, 2008 Cambridge Histories Online © Cambridge University Press, 2008 184 LI YUAN-HAO THE HSI HSIA I85 190 300 km 100 H s i 2OOmilos chou .Hamili-choul ' "%_<- U i g h ur s Y I N-^S HEI-SHU CHEN-YEN CHUN-SSU HSI-P'ING CHUN-SSU 9 A-LA- SHAN DESERT T'len-te chun ortress 4 Kua-chou Su-chou e 0 R D OS PAI-MA CHIANG-CHEN CHON-SSU H ua ng-t'ou HAN Urghurs DESERT |KAN-SU CHUN-SSU I TSO-HSIANG CHAO-SHUN [ CHUN-SSU e> 0 Kan-chou IHsuan-hiiafu) Ting-chou H S ing-ch-ingf U .^ Hua ._ choij ShunthousT* 1 "^ 011 Tibetans HSIANG-YU CHON-SSU Pai-ch Hsiachou eLing-chou pfortres (Hsi-p'ingfu) HAN-HAI Ming-shay Yen-chou S Chun,*.,. ^ ^ ^ .CH.NG-SAII ^^ung-Chou^ CHUN-SSU ®Wei-chou I CHO-LO HO-NAN CHUN-SSU The boundaries of the Hsia state (other than the border with Sung) are approximate. At some date early in the twelfth century the name of the capital was changed from Hsingch'ing fu to Chung-hsing fu. MAP 12. The Hsi Hsia state, m i Cambridge Histories Online © Cambridge University Press, 2008 ^J " ( «Shih-cho TSO-HSIANG SHENYUNG CHUN-SSU l86 THE HSI HSIA expanded the bureaucratic apparatus along Sung institutional lines. He created a central Secretariat (Chung-shu sheng), Bureau of Military Affairs (Shu-mi yuan), Finance Office (San-ssu), Censorate (Yu-shih t'ai), and sixteen subordinate bureaus (ssu) under the nominal supervision of a chancellor (shang-shu ling). Top positions in these offices could be held by Tanguts and Chinese alike. Other appointments carried Tangut titles and were clearly reserved for the ruling elite, as were the highest military appointments.59 But behind the facade of Chinese nomenclature, the actual workings of Hsia government still remain shadowy. Next to nothing is known, for example, about the government'sfiscalactivities. During the years 1035—6 Hsia armies campaigned against the Tibetans of Kokonor, the tribes around Lan-chou, and the Ho-hsi Uighurs. The longdrawn-out conquest of Ho-hsi has already been discussed. In the Huang River valley several fierce and prolonged battles took place between Tibetans and Tanguts, with heavy losses on both sides. Unable to subdue Ch'ingt'ang, Wei-ming Yiian-hao nevertheless took advantage of an estrangement between Ku-ssu-lo and his two elder sons, as well as the son of his murdered ex-minister, to isolate the Tibetan leader and force him into temporary retreat west of Ch'ing-t'ang. The Tangut army then marched to Lan-chou, to secure its rear flank against Tibetan attack and to frustrate Tibetan communications with Sung. Hsia troops penetrated to Ma-hsien-shan (north of Lint'ao hsien) and left a fortified garrison to hold the area.60 Around this time and especially after 1038, the Sung court tried to foster a united Tibetan front against Hsia, but its efforts yielded little tangible reward. Ku-ssu-lo's weakness ultimately posed a greater danger to Hsia by its failure to impede Sung annexation of the region at the end of the eleventh century. In numerous recorded instances the Wei-ming ruler dealt swiftly and summarily with opposition to his policies or his rule. As he made final preparations for his public enthronement as emperor in the autumn of 1038, his chief military leader, (Wei-ming) Shan-yii, sought asylum across the Sung border with his army but was extradited and executed at Yu-chou. The rebel and his younger brother, the monarch's paternal uncles, had been the leaders of the Right and Left Wings of the Hsia army, a clear indication of the depth of sentiment against Wei-ming Yiian-hao's imminent rejection of Sung vassal status. Whatever its misgivings about the Tangut ruler, though, the Chinese court refused to violate its treaty with Hsia by harboring the would-be defectors.6' 59 SS, 485, p. 13993; HCP, 120, p. 23b. 60 HCP, 117, pp. 17b— 18a; 119, pp. 16b— 17a; SS, 4 9 2 , pp. 14161—2. 61 On the rebellion of Shan-yu, see HCP, 122, pp. 8a~9b; Ssu-ma Kuang (1019-86), Su-sbui cbi wen (TSCCCP ed.) 12, pp. i a - 2 a . Cambridge Histories Online © Cambridge University Press, 2008 LI YOAN-HAO 187 The imperial elevation proceeded as scheduled. Already the ruler and his chieftains had sworn an oath, sealed with a libation of blood and wine mixed in a skull, ostensibly vowing first to attack the Sung's Fu-yen circuit. Two brothers of the Yeh-li clan, Wang-jung (also called Kang-lang-ling) and Yiich'i, assumed leadership of the Left and Right Wings, respectively, of the Hsia army in place of Yiian-hao's uncles who had attempted to defect. They were either uncles or brothers of the Yeh-li empress, mother of the heir apparent, and they wielded great power.62 In the tenth month of 1038, thirty-year-old Wei-ming Yiian-hao was formally enthroned as emperor of Ta Hsia. He proclaimed a new reign title and perhaps also at this time bestowed imperial titles and temple names on his father and grandfather. The new emperor then traveled to Hsi-liang Fu (the official Hsia designation of Liang-chou) to make offerings to the spirits while his generals mobilized troops. A special embassy set out for the Sung capital bearing a letter announcing the establishment of the Tangut state as an imperial power and requesting recognition as a friendly but sovereign western neighbor, something for which its author clearly expected to have to do battle. Having had some advance notice of these events, the Sung court was not entirely surprised. After declining a proposal to execute the Hsia ambassador's party, Sung Jen-tsung (r. 1022—63) proffered the usual courtesies but refused to accept the gifts of horses and camels. Likewise, the Hsia embassy rejected his edict and presents and was escorted back to the border. Jen-tsung then revoked his estwhile subject's titles and ranks and closed all the frontier markets. The Wei-ming emperor returned all his Sung paraphernalia of office with an "impudent" letter.6' The Fan [the Tanguts} and Han are each different countries, and their lands are vastly dissimilar. This is not a case of usurpation. Why is your resentment so deep? Moreover [I] was elevated by the throng and, in accordance with ancient T'o-pa tradition, aspire to be emperor. What is wrong with that? Negotiations to resolve the disagreement over the Tangut ruler's status and how to denote it thereupon commenced and concluded only in 1044. Meanwhile, the two states engaged in a war of sabotage and attrition, highlighted by three major Tangut victories. Yeh-li Wang-jung, commander of the Left Wing, acted as the Tanguts' chief negotiator with the Sung authorities at Yen-chou, first with Fan Chung-yen (989-1052) and then with P'ang Chi (988-1063). 62 The Yeh-li clan is discussed later. For Wei-ming Yiian-hao's coronation, see HCP, 111, pp. 8b, iob— ita, I4b-i5a. 63 HCP, 125, p. 1 ia-b. On the course of the ensuing war and negotiations, see Wu, Hit Hsia shih kao, pp. 59—71; Evgenii I. Kychanov, "Les geurres entre le Song du Nord et le Hsi Hsia," in £.tuda Song in memoriam ttienne Balazs, ed. Francoise Aubin, 1st series, no. 2 (Paris, 1971), pp. 106—18. Cambridge Histories Online © Cambridge University Press, 2008 l88 THE HSI HSIA Until 1042, negotiations were stalled over the Tangut ruler's refusal to call himself ch'en (subject) in letters to the Chinese court. In that year, however, Liao diplomatic intervention and a stunning Tangut rout of the Sung forces broke the stalemate.64 The Khitans pressed the Chinese for further territorial concessions in Kuan-nan, a strategic wedge of territory in Ho-pei that threatened the approaches to K'ai-feng. In the course of dealing with these demands, the Sung court urged the Khitans to mediate its conflict with Hsia by applying pressure on their vassal. Further, the Sung court secretly instructed P'ang Chi to accept the Tangut peace initiatives, agreeing that as long as the Hsia ruler agreed to accept subject {ch'en) status toward Sung, he could keep his native title of wu-tsu. Early in 1043, the Tangut ruler proposed a new formula in a communique to Jen-tsung: "The son, emperor of the Pang-ni-ting state, Nang-hsiao, submits a letter to the father, emperor of the great Sung." Pang-ni-ting appears to be a variant transcription of the Tangut name for their own country (written Pai-shang kuo in Chinese), and Nang-hsiao the new personal name of Wei-mi ng Yiian-hao. One Chinese statesman observed that this address copied the familial protocol then in use in Sung—Liao diplomatic discourse.6' The court rejected it. A few months later the Sung offered terms by which, among other things, the Tangut monarch would be invested as "ruler" (chu; greater than a king, wang, but humbler than an emperor, huang ti) and receive annual gifts. Soon the Tangut side responded with eleven demands, including an increase in the annual gifts, greater trading privileges for the Tangut envoys, and permission to sell white salt to the Chinese. It was again pointed out that here the Tanguts were imitating the Khitan example. Liao had just settled for an increase in its annual subsidy instead of the lands in Kuan-nan they had originally demanded. This apparent mimicry confirmed Chinese suspicions of collusion between their two northern neighbors Hsia and Liao, and it greatly perplexed the court when scarcely a year later the supposed collaborators went to war with each over some rebellious Tangut border tribes. After the Tangut emperor finally agreed to accept ch'en status, the Sung extended trading rights to the Tangut envoys at government hostels in K'aifeng and increased the annual presents. It would not, however, legalize sales of Tangut salt.66 Just as negotiations were completed, in the fifth month of 1044 the Khitans invaded Hsia in pursuit of some renegade border tribes. 64 Tao Jing-shen, "Yii Ching and Sung policies toward Liao and Hsia, 1042-44," Journal of Asian Studies, 6 (1972), pp. 114—22. Fan Chung-yen has an entry in Sung biographies, vol. 1, pp. 321-30, and inSungjen chuan cbi tzu liaosuoyin, vol. 2, pp. 1648—52. 65 HCP, 139, pp. 6b-ja;SS, 485, p. 13998. 66 HCP, 142, pp. 8a-i3b; Wu, HsiHsia shibkao, pp. 69-71. Cambridge Histories Online © Cambridge University Press, 2008 THE SUCCESSION TO WEI-MING YUAN-HAO 189 Perhaps the Tanguts' audacious pursuit of diplomatic parity with Liao vis-avis the Sung court also lay behind the Khitans' punitive campaign. Warned by the Liao court not to conclude peace with Hsia just yet, the Chinese warily withheld the Sung patent of investiture for Wei-ming Yiian-hao. But after learning of a Khitan defeat the Sung seized the initiative and dispatched its envoys to conclude the treaty with Hsia in winter of 1044. Besides the points already mentioned, the treaty with Sung stipulated that Hsia would receive Sung envoys at Yu-chou, not at the Tangut capital, using the same protocol (pin k'o li) as it extended to Liao envoys. Border markets were restored. The annual gifts totaled 255,000 units: 153,000 p'i of silk, 30,000 small chin of tea, and 72,000 Hang of silver.67 Clarification of the boundary between the two states, however, remained subject to discussion. This failure to clearly demarcate the border was to remain a source of bitter disputes until the two sides ceased to share a border, after Sung lost its northern territories to Chin. Meanwhile, the war with the Khitan continued. Having ably frustrated a three-pronged Liao invasion of his territory, the Tangut ruler hastened to mollify the humiliated Liao emperor.68 Relations were temporarily patched up, but the conflict broke out again several years later, and it is not even certain that the Liao court ever formally invested Wei-ming Yuan-hao's successor, Liang-tso. THE SUCCESSION TO WEI-MING YUAN-HAO The various accounts of Wei-ming Yuan-hao's death and the origins of his successor are extremely confused.69 Toward the end of the Sung war the Tangut emperor took as his own wife the intended bride of his son, a lady of the Mo-i clan. His doing so is usually castigated as incestuous, but perhaps it was an attempt to loosen the tentacles of the Yeh-li, the empress's extremely powerful clan. The Mo-i concubine bore a son. Yuan-hao's officially designated heir was Ning-ling-ko, a son of the Yeh-li empress, and the Yeh-li elders Wang-jung and Yii-ch'i, who dominated both the court and the military, doubtless viewed these developments as a serious threat to their own interests. It was thus arranged to marry Ning-ling-ko to the daughter of the 67 Huang Ch'ing-yiin, "Kuan yii Pei Sung yii Hsi Hsia ho yiieh chung yin chiian ch'a te shu Hang wen t'i," Chung hsiieh li shih chiao hsiieh, 9 (1957), pp. 19—20. 68 On the Khitan invasion of Hsia, see SS, 485, pp. 13999—14000; LS, 19, pp. 2 3 0 - 1 ; Shen Kua (1031?—95), Meng-bsi [ch'iipi t'an chiao chmg, ed. Hu Tao-ching (Peking, 1957; repr. Taipei, 1965), pp. 7 8 7 - 9 ° 69 This section draws on the following accounts: HCP, 162, pp. ia—2a; Ssu-ma Kuan, Su-shui chi wen, 9, pp. 9a-b; 10, pp. jb, 9a; 11, pp. n b - i 2 a ; Wang Ch'eng (d. ca. 1200), Tung tu shih lu'eh, repr. in vols. 11—14 of Sung shih tzu liao ts'uipirn, 1st series, ed. Chao T'ieh-han (Taipei, 1967) 127, p. 6a-b; Tai, Hsi Hsia chi, 11, pp. lib— 12a; Wu, Hsi Hsiasbusbib, 18, pp. I2b-I3a. Cambridge Histories Online © Cambridge University Press, 2008 190 THE HSI HSIA empress's uncle (or perhaps brother), Yeh-li Wang-jung. Yeh-li Wang-jung then invited the emperor to visit his tent on the eve of the wedding, intending, apparently, to assassinate him. Word of the alleged plot leaked out, and Yeh-li Wang-jung, Yeh-li Yii-ch'i, and the members of three Yeh-li lineages were executed as a result. In some sources these events are dated 1042 or 1043 and blamed on Sung subversion, but it seems certain that they took place one or two years after the conclusion of the peace treaty with Sung, in 1045 or 1046. The Yeh-li empress escaped punishment, but at some point later the emperor demoted her and installed his Mo-i concubine as empress. The former empress Yeh-li retained enough influence, however, to persuade the emperor to repent his hasty execution of her blameless kinsmen and to seek out survivors of the slaughter. As a result of this, the wife of Yeh-li Yii-ch'i, a member of the prominent Mo-ts'ang clan, was discovered and brought to the palace. When the emperor began to visit her, the former Yeh-li empress removed Mo-ts'ang to a temple in the suburbs of the capital, but the emperor continued his amours. The Sung official history asserts that in the second month of 1047 the lady Mo-ts'ang bore a male child, named Liang-tso after the river beside which he was born. Other accounts date his birth two months after the emperor's death, which occurred either at the end of 1047 or early in 1048.70 This child was brought up in the home of his mother's brother, Mo-ts'ang O-p'ang, who had served under Yeh-li Yii-ch'i's command. The Yeh-li empress's demotion (1047?) spurred the dispossessed heir apparent, with the connivance of Mo-ts'ang O-p'ang, to avenge the Yeh-li's misfortunes by attempting to assassinate the emperor later that year. After stabbing his father in the nose, the luckless young man fled to Mo-ts'ang O-p'ang's residence, where he was promptly arrested and executed, along with the remaining Yeh-li clan members. Now the Mo-ts'ang clan, whose elder O-p'ang may have fabricated the entire scenario to ensure the position of their puppet child-emperor, stepped into the power vacuum. When the emperor died the day following the assassination attempt, a council of great chieftains was convened to select a new ruler. All the tribal elders approved the heir, a cousin of Wei-ming Yiian-hao, chosen by the deceased ruler before these events and designated in his dying testament. Mo-ts'ang O-p'ang alone objected, somewhat disingenuously, on the grounds that succession by a cousin would violate the established practice of primogeniture, and he moreover claimed that the chosen candidate lacked "merit." Instead, Mo-ts'ang put forward the bastard 70 SS, 485, p. 14000; see also the sources listed in n. 69. Cambridge Histories Online © Cambridge University Press, 2008 I-TSUNG, HU1-TSUNG, AND CH'UNG-TSUNG 191 imperial infant as his own candidate, and his elevation, we are told, surprisingly aroused no dissent. Thereupon the infant became emperor (he was later given the posthumous title I-tsung and reigned from 1048 to 1068), and his mother was appointed the empress dowager. Mo-ts'ang O-p'ang became regent and palace minister (cbia hsiang), the de facto ruler of the Hsia state. Thus Wei-ming Yiian-hao failed to secure his throne against the tribal oligarchy, and there followed a half-century of rule by members of maternal clans, first the Mo-ts'ang, which lasted only fifteen years, and then the more powerful and politically adroit Liang. This pattern of maternal clan influence clearly had deep roots in Tangut (and Tibetan) sociopolitical and kinship traditions. A STATE IN PERIL: THE REIGNS OF I-TSUNG HUI-TSUNG ( I O 6 8 - I O 8 6 ) , (1048-1068), AND CH'UNG-TSUNG TO IIOO Reviled by contemporary Chinese writers as a stubborn, reckless youngster and an unworthy successor to his father, as he grew up I-tsung in fact strove against great odds to protect his own independence and to arrest the declining power and prestige of his state. The empress dowager died, the victim of an intrigue, in 1156. In 1061 I-tsung, now fourteen years old, eliminated Mo-ts'ang O-p'ang. Seeking more reliable allies than the Mo-ts'ang, he then married his accomplice, Lady Liang, formerly the unhappy wife of Mo-ts'ang O-p'ang's son, and appointed her younger brother, Liang I-mai, as his new palace minister, with hereditary tenure in office.7' For two generations Liang empresses and their brothers presided over the precarious fortunes of the Tangut state. A number of Wei-ming clan members were suppressed, exiled, or excluded from power. Others, however, were tolerated, chiefly for the prestige and loyalty that they commanded. The support of the dominant Liang faction rested in the armed strength of their various allies among tribal chieftains. They nurtured hereditary blood feuds with both Sung Chinese frontier officials and their non-Han tribal subordinates. Thus the ruling Liang faction fostered an attitude of uncompromising belligerence toward China that constantly endangered the state and became increasingly difficult to repudiate when once internal opposition to it intensified. In struggling to assert their own authority against them, the Tangut emperors, on the other hand, were forced into a posture that often appears to be pro-Chinese. I-tsung resolved a long-standing border dispute 71 HCP, 184, pp. 10b, 15b; Shen Kua, Mmgch'ipi t'an 25.3, p. 452. On the names of Liang I-mai and his son, Ch'i-pu, see Pu P'ing (pen name of Li Fan-wen), "Hsi Hsia huang ti ch'eng hao k'ao," Ninghsiashehui k'ohsiith, 1 (1981), pp. 77-8, n. 17. Cambridge Histories Online © Cambridge University Press, 2008 192 THE HSI HSIA with Sung, negotiated a new eastern boundary, and arranged for the establishment of markets previously closed by Sung because of Mo-ts'ang O-p'ang's provocative colonizing activities along the Ch'ii-yeh River. Around the time of Mo-ts'ang O-p'ang's demise, I-tsung petitioned the Sung court for permission to replace Tangut rites with Chinese court ritual and costume.72 Thus commenced a series of seesaw shifts between Han and Tangut court ceremonial that reflected fluctuations in the relative strength of the contending pro-Chinese and nativist Tangut factions. Later the empress dowager championed Tangut forms; the emperor switched to Han usage whenever the chance arose. Chinese symbols, institutions, and even individuals became weapons (and targets) in the court power struggles at Hsing-chou. I-tsung's gesture signaled more than an assertion of personal autonomy. It was also an effort to gain Sung recognition and bolster Tangut prestige in another quarter. Liao had already turned a cold shoulder on the Hsia court and had begun to woo the Tibetans, with whom the Tanguts still fought intermittently. In 1058 a Khitan princess, denied to I-tsung, was married to Ku-ssu-lo's son and eventual successor, Tung-chan.7' Sung and Hsia meanwhile waged a perpetual diplomatic tug-of-war, with the Tangut envoys repeatedly demanding the same dignities accorded their Khitan counterparts. The Sung officials in their turn complained of the inferior quality of the Tangut emissaries and the defiant tone of their communications; both sides regularly exchanged insults. In 1067, by means of a ruse, a Chinese border official seized the Hsia town of Sui-chou.™ In retaliation the Tanguts killed a Sung prefect, who was exposed as a double-dealing traitor. After lengthy debate, the Sung court decided to keep Sui-chou, greatly embittering the Tanguts who thereafter continuously harassed the area. At the end of the year I-tsung died at the age of twenty, presumably from wounds received in battle. His eldest son, sevenyear-old Wei-ming Ping-ch'ang (Hui-tsung) was enthroned that winter. I-tsung had initiated a number of reforms, in the process abandoning some of the precedents governing Chinese-Tangut relations that had been established by his father.7' This, more than his alleged character defects, may explain Chinese rancor toward him and his policies. 72 SS, 4 8 5 , p. 14001. Possibly he did so ac the instigation of the Liang clan, which Shen Kua claims to have been of Chinese origin (Mtng ch'i pi t'an, 2 3 . 3 , p. 452), even though Liang is attested as a venerable Tang-ch'ang clan name. 73 HCP, 188, pp. 2b-3a. 74 P'eng Pai-ch'uan (fl. 1113), T'ai-p'ing chih chi t'ung lei, repr. in vol. 10 of Shib yuan ts'ung shu, ed. Chang Chiin-heng (1917; repr. Taipei, 1966), 15, pp. ib-2a. 75 HCP, 196, p. 23b. Cambridge Histories Online © Cambridge University Press, 2008 THE REIGN OF HUI-TSUNG 193 THE REIGN OF HUI-TSUNG ( 1 0 6 8 - 1 0 8 6 ) Hui-tsung's troubled reign, or more accurately that of his mother, the empress dowager Liang, opened with an inconclusive war in 1070—i, sparked by the Sung refusal to negotiate a settlement of the Sui-chou affair. Thereafter the Chinese commenced a massive program of border fortification in the west, aimed to displace the Tibetans in Ch'ing-t'ang. Treaty relations with Hsia were repeatedly broken; competition for military merit and its generous rewards inflamed the ranks of Sung officers with war fever. In an expansionist mood, the Sung mounted offensives against Hsia in 1081—3, 1091—3, and 1096—9. The Chinese recaptured Lan-chou in 1081 and advanced into Ch'ing-t'ang. As the war continued and the theater of battle expanded, the Liang clique became more and more dependent on increasingly independent generals. Chief among these were members of the royal Wei-ming clan, who dominated the center and southeast of the state, and their allies the Jen-to clan, who controlled the southwest area bordering western Shensi and Kokonor. Meanwhile, Tangut-Khitan relations began to improve, and the hardpressed Tibetans also began to cultivate closer ties with Hsia. In 1072 a sister of the Tangut emperor was married to Tung-chan's son, Lin-pu-ch'ih (Rinpoche)> The young Tangut emperor had grown up under the complete dominance of his mother the empress dowager Liang and of her younger brother Liang I-mai. He was married to one of Liang I-mai's daughters to ensure the continuation of the Liang family's hold on imperial power. In 1080 Huitsung, chafing under his mother's dominance, discarded Tangut court ritual, which had been restored in 1070 by the empress dowager, in favor of Chinese ceremonies. A year later the Hsia court uncovered a plot alleged to have been hatched between the young emperor and his Chinese favorite, Li Ch'ing, to turn over the southern Ordos to Sung. Li Ch'ing was executed, and Huitsung was temporarily imprisoned in a fortified dungeon five //' from the palace. Instantly the emperor's adherents rallied their forces to oppose Liang rule. Chief Minister Liang I-mai in vain dispatched emissaries to their camps with the silver tablets used to summon them to arms, attempting to make them affirm their loyalty. For several months Sung officials breathlessly awaited a resolution of the internal crisis in the Hsia state. Finally the Sung emperor ordered the mobilization of a great army under the command of the eunuch-general Li Hsien, to launch afive-pronged"punitive" attack on the Tangut capital. Thefivear76 HCP, 233, pp. 6b-7a. Cambridge Histories Online © Cambridge University Press, 2008 194 THE HSI HSIA mies, however, failed to converge at the appointed time because their leaders soon fell to squabbling among themselves. But in 1081 Li Hsien managed to capture Lan-chou, which had been under Tibetan domination since the eighth century and had fallen into the Tangut orbit about 1063.77 Moreover, Li and the other generals inflicted considerable damage and devastation throughout the southern provinces of Hsia, though not without heavy cost to themselves. Empress Dowager Liang and her generals at first withdrew to defend the capital and Ling-chou, while in the southwest Jen-to Ling-ting led the proWei-ming forces in defending, with heavy casualties, the palaces and government buildings of the T'ien-tu mountain region against Li Hsien's onslaught. The Jen-to-Wei-ming alliance had its headquarters at the Cho-lo Army Superintendancy north of Lan-chou.78 Later in 1082, the two rival parties cooperated to defeat the Sung at Heng-shan in northern Shensi. In 1083, mounting losses finally induced the Sung to accept the Tanguts' peace overtures. But the Chinese court's refusal to discuss a return of captured territories ensured the continuation of hostilities, especially around Lan-chou, where the Sung soon began to consolidate and extend its gains. A sobered Hui-tsung had evidently resumed his throne in 1083. In 1085 Liang I-mai died, and his son, Liang Ch'i-pu, succeeded him as the new state minister. The warrior empress dowager died later that year. In 1086 Huitsung too died, bequeathing the throne of Hsia to his three-year-old son, Wei-ming Ch'ien-shun (Ch'ung-tsung, 1086-1139), under the regency of his widow, the new empress dowager Liang, who was Liang Ch'i-pu's younger sister. Thus power at the capital Chung-hsing, although now openly contested, passed from one generation of Liang siblings to the next. A simultaneous change of leadership in the Sung capital brought a brief restoration of treaty relations and the return of a few forts. But the border was rife with rumors of a coup in Hsing-chou, as conflict soon broke out between the empress dowager and her brother the state minister Liang Ch'i-pu. Bad feeling also developed in the Liao court, which disliked the new regents. Liang Ch'i-pu found an ally in the Tibetan leader A-li-ku, the adopted son and unpopular successor of Tung-chan (d. 1083). A-li-ku's regime at Ch'ingt'ang was challenged by that of his rival, Wen-ch'i-hsin, at Miao-ch'uan, adjacent to the Tangut southwestern province governed by the Jen-to faction at Cho-lo. Long-standing differences between the two Tibetan centers had resulted in friendly relations developing between Miao-ch'uan and the Jen-to 77 P'eng P a i - c h ' u a n , T'ai-p'ing chih chi t'ung lei, 1 5 , p p . 12a—26a. O n Lan-chou under Tangut control, seeTai, Hsi Hsiachi, 13, p. 9b; and HCP, 226, p. 3a. On Li Hsien, see Sung jtn chuan chi tzu liao 10yin, vol. 2, p. 912. 78 Maeda, Kasei, pp. 593—613. Cambridge Histories Online © Cambridge University Press, 2008 THE REIGN OF HUI-TSUNG 195 royalist leaders at Cho-lo while the rival Tibetan regime in Ch'ing-t'ang enjoyed an alliance with Liang Ch'i-pu." Extreme instability prevailed in the Kokonor region throughout the 1090s. A-li-ku's death in 1096 set in motion a confused succession contest that the Sung exploited to extend their control over Kokonor in 1099. At every step they met Tangut resistance. The long-awaited coup at the Tangut court apparently took place late in 1094. Jen-to Pao-chung, Wei-ming A-wu, and their followers slew Liang Ch'i-pu and exterminated his clan. The empress dowager, protected by her own still-considerable military power, evidently sided with the assassins because she had suspected her brother of plotting against her son, the twelveyear-old emperor, and herself.80 But the full restoration of Wei-ming royal authority was not possible in the face of a grave military emergency. The Sung had mounted offensives against Hsia in 1091 and 1093. Then in 1096 they launched an all-out campaign to destroy Hsia and occupy Kokonor. The war dragged on until 1099. Alarmed at the ferocity of the Sung depredations, the Liao court, which was itself embroiled in a longdrawn-out war with the Tsu-pu in Mongolia, to the north of Hsia, three times issued stern notes urging the Sung to desist, but to no avail.8' These four years of violent warfare, along with the long closure of border markets, brought further deprivation and misery to the Hsia people, ruining their livelihoods and their lands. The empress dowager died in 1099, poisoned, rumor had it, by a Liao envoy for her failure to assist the Khitans to suppress a rebellion of vassal tribes.82 For months the Sung court spurned Tangut envoys sent to announce her death and to sue for peace. Determined to end the war and at last free of Liang dominance, the Wei-ming elders went to extreme lengths to appease the Sung court. While the Tanguts were thus negotiating peace in the Chinese capital, they continued to resist the Sung advance into Kokonor. Nevertheless, the Sung seized and fortified T'ien-tu and took Hui-chou. While Chinese armies marched up the Huang River valley in the fall of 1099, the various Tibetan factions at Miao-ch'uan, Tsung-ko, and Ch'ing-t'ang were in constant agitation, alternately surrendering and resisting. The Sung command staff was in almost equal disorder, as generals and their subordinates impeached one another, were demoted, cashiered, and reinstated in dizzying succession. Tangut armies also entered the fracas and the Tibetan forces, numbering sixty 79 Maeda, Kasci, pp. 6 0 6 - 9 ; HCP, 4 0 2 - 4 , 444 passim; 467, pp. 8a-b. 80 The data are elusive. See for example, Sung huiyao chi kao, 175, ping 8, pp. 31b—32a. Wu, Hsi Hsia shu shih, 29, pp. 15a-16a, provides the only account of the coup that assigns it a fixed date; contemporary corroboration remains to be found. 81 IS, 115, p. 1528; HCP, 492, pp. 8b-9a; 507, pp. 3b-4a. 82 Wu, Hsi Hsia shu shih, 31, p. ib. Cambridge Histories Online © Cambridge University Press, 2008 196 THE HSI HSIA thousand or seventy thousand, were still large enough to foil the Sung's initial attempts at occupation.83 After the first surrender of Ch'ing-t'ang in the ninth month of 1099, the town was retaken by the Miao-ch'uan faction, which installed its own candidate as A-li-ku's successor. The Sung court confirmed this person in office, and in 1102 he contracted a marriage alliance with the Tangut royal house. Between 1102 and 1104 Sung armies reoccupied Ch'ing-t'ang, which had been renamed Shan-chou in 1099 and was again renamed Hsi-ning in 1104.84 By 1109 the Sung government had registered all the Tibetan towns of Kokonor under Chinese names, but it never effectively controlled this area. Until the end of Northern Sung (1128), it remained a zone of constant conflict between the Chinese, on the one hand, and the Tibetans and Tanguts, on the other. This account may give the impression that in the late eleventh century Hsia was an unstable society, embroiled only in factional squabbles and sporadic warfare with its neighbors. However, there were two major developments that also deserve attention. The first was the rapid growth of Buddhism as the state faith under imperial auspices. This began in Li Te-ming's time, if not before. The Liang empresses devoted considerable attention to the translation of sutras, and by the end of the eleventh century the entire Tripitaka had been rendered into Tangut. This in itself was a monumental achievement. Numerous temples, either newly built or restored, were dedicated. There was a regular exchange of religious adepts and materials with foreign Buddhist centers such as Liao, Turfan, and Tibet.8' Second, the Hsia economy and food supply, vulnerable to prolonged disruption of its vital trading links with the Chinese, had evolved into a rather sophisticated mechanism. Once freed from the extraordinary demands placed on it by constant warfare, it could more than adequately sustain the growing imperial establishment and permit the remarkable cultural flourishing of the twelfth century. The Tanguts became famous not only for their livestock, hunting falcons, and other animals but also for the products of their local industries. These included their prized camel-hair carpets; printed and illustrated books comparable in quality with the best Sung publications; rhubarb and other herbal plants; and high-quality salt, which served as a barter currency in Hsia, much as textiles did in Liao. The Tanguts' own economic philosophy is perhaps summed up in a Tangut maxim: "Possessing sheep, 83 HCP, 514, pp. 7a-2oa; 515, pp. 7 a - i 3 a ; 516, pp. 3b-22b. 84 SS, 4 9 2 , p. 14167; 87, pp. 2 1 5 4 - 7 0 . 85 Shih Chin-po, " 'Hsi Hsia i ching t'u chieh," Wen hsien, 1 (1979), pp. 215—29; Shih Chin-po, Hsi Hsia wen hua, pp. 7 5 - 6 . Evgenii I. Kychanov, "From the history of the Tangut translations of the Buddhist canon," in Tibetan and Buddhist studies commemorating the 200th anniversary of the birth of Alexander Csoma de Kijros, ed. Louis Ligeti (Budapest, 1984), pp. 377—87. Cambridge Histories Online © Cambridge University Press, 2008 HSIA COMES OF AGE: CH'UNG-TSUNG AND JEN-TSUNG 197 one slaughters cattle in the mountains of the Tangut land; possessing money; one seeks [profits in} trade with Chinese merchants."86 HSIA COMES OF A G E : C H ' U N G - T S U N G ( 1 0 8 6 - 1 1 3 9 ) A N D JEN-TSUNG (1140-II93) Ch'ung-tsung and his mentors guided the state through its last period of strife with the Sung and supported the Liao court against its Jurchen conquerors, the Chin, until its fall in 1124, when necessity forced them to come to terms with the new masters of north China. Even more important, they initiated an overhaul of the Tangut government. Although by the end of Ch'ung-tsung's reign in 1139 the Shensi border with Chin had still not yet been stabilized, and Jurchen failures to honor certain territorial promises continued to irritate the Tanguts, who from time to time asserted their claims to disputed districts along the Hsia—Chin border, the Hsia state, which was now more or less completely cut off from Sung, enjoyed a long period of peace. The domestic policies of Ch'ung-tsung and his successor Jen-tsung aimed primarily to strengthen central authority over the army and its leaders, the still-powerful tribal aristocracy, who naturally resisted any reduction of their traditional privileges. Notwithstanding imperial efforts at Confucian indoctrination, the loyalties of those leaders to the dynasty were secured by a contractual compromise between the throne and the great clans, by which each obtained a guarantee of its position. This compromise was formally expressed in legal documents and institutional arrangements worked out in the succeeding decades of the twelfth century.87 In the early years of his real reign (i.e., after his mother's death in 1099) Ch'ung-tsung ruled in collegial partnership with the Wei-ming elders who had returned the young monarch to his throne and were determined to tighten their clan's hold over it. First, they had to tame the military bureaucracy, beginning with their own erstwhile allies. In 1103-4 t n e emperor 86 Wu, Hsi Hsia shih kao, pp. 170—88; Kychanov, Ocherk istorii, pp. 79—99; Evgenii L. Kychanov, Vnov' Sobrannye dragotsennye pamye izrecheniia (Moscow, 1974); the Tangut text appears on p. 155 and the Russian translation on p. 90. 87 See Ruth Dunnell, "Tanguts and the Tangut state of Ta Hsia" (Ph.D. diss., Princeton University, 1983), chap. 5, "The structure of Tangut government," esp. pp. 202—52, much of which is based on Kychanov's translation and study of the Tangut law code, all four volumes of which have now been published. See Evganii I. Kychanov, Izmennyi i zanovo utverzbdennyi kodeks deviza tsarstvovaniia nebesnot protsvetanie (1149-1169), vol. 1, Commentary (Moscow, 1988); vol. 2, Facsimile, translation and annotation, chaps. 1—7 (Moscow, 1987); vol. 3 , chaps. 8—12 (Moscow, 1989); vol. 4, chaps. 13—20 (Moscow, 1989). This work includes the original Tangut texts, a Russian translation, annotations, and a commentary on the annotations. A brief description of the code's contents may be found in Evgenii I. Kychanov, "Monuments of Tangut Legislation ( i 2 t h - i 3 t h centuries)", in ttudes TiMtaines, Actes du XXIXe Congres international des Orientalistes, July 1973 (Paris, 1976), pp. 29—42. Cambridge Histories Online © Cambridge University Press, 2008 198 THE HSI HSIA stripped the Jen-to clan of its military power. Jen-to Pao-chung was demoted and summoned to the capital on suspicion of treason. Ch'ung-tsung then granted his own younger brother, Wei-ming Ch'a-ke, a noble title and command of the Tangut army. An extremely competent general, Wei-ming Ch'a-ke won several stunning victories over Sung troops during the next decade and remained one of the most powerful figures in Tangut government until his death in about 1156.88 During the Chen-kuan reign period (1101 —13) — its auspicious reign title deliberately chosen to invoke the much-admired model of T'ang T'aitsung — Ch'ung-tsung issued the military code entitled "The Jade Mirror Code of the Chen-kuan Reign" (Chinese: Chen-kuan yii ching t'ung; the original is in Tangut), of which fragments survive.89 The emperor also established a state school with three hundred pupils supported on government stipends and endeavored to promote qualified persons to office, especially those with literary skills. According to a late source, imperial Prince Wei-ming Jenchung was literate in Chinese and Tangut, held an inner-court appointment, and was ennobled in 1120. He became the spokesman for a "civilian" faction in the government and often denounced the commander in chief, Wei-ming Ch'a-ke, for corruption and abuse of power.9° Ch'ung-tsung, and his successor after him, shuffled appointments to play the two groups off against each other. Consort relatives were evidently kept under strict control. In 1105 the emperor was married to a Liao princess, but there is no record that either she or Jen-tsung's mother, the Chinese concubine Lady Ts'ao, were ever elevated to the rank of empress. Grateful to Liao for firm support against its Chinese adversaries, the Tangut royal house cultivated close relations with the Khitans, even when it must have been clear that their cause was lost, and aided the last Liao emperor in his flight from the Jurchens. In 1125 Ch'ungtsung's Khitan wife and her son reportedly died of broken hearts after the last Khitan ruler's capture by his Jurchen pursuers.9' In the penultimate year of his reign (1138) Ch'ung-tsung is reported in a very late source, the nineteenth-century Hsi Hsia shu-shih by Wu Kuangch'eng, to have at last taken an empress, in the person of the daughter of Jen Te-ching. Jen Te-ching was a surrendered Chinese holding office in Hsia who later became chief minister, dominated the court for twenty years, and eventually tried to carve out his own state in the eastern half of the country. Although 88 According to Wu, Hsi Hsia shu shih, 36, pp. 10b—1 i b ; SS, 486, pp. 14019-21. 89 Evgenii I. Kychanov, "Svod voennykh zakonov Tangutskogo gosudarstva 'Iashmovoe zertsala upravleniia let tsarstvovannia Chzhen'-kuan' (1101 —1113)'," Pis'mmnyepamiatniki vostoka, 1969 (Moscow, 1972), pp. 2 2 9 - 4 3 . 90 SS, 486, p. 14109; Wu, Hsi Hsiasbusbih, 32, p. 12b; 33, p. 8a; 34, pp. I 5 b - i 6 a . 91 Wu, Hsi Hsia shu sbih, 33, p. 17b. Cambridge Histories Online © Cambridge University Press, 2008 HSIA COMES OF AGE: CH'UNG-TSUNG AND JEN-TSUNG 199 the career of Jen Te-ching forms one of the outstanding episodes of twelfthcentury Hsia history, many of its details as related by Wu Kuang-ch'eng lack corroboration in earlier sources. The daughter who gave him his entree to court power by becoming first a royal concubine, then empress, and soon after empress dowager, is nowhere else mentioned. But the story certainly suggests a comparison between Jen Te-ching's rise to power and the careers of his eleventh-century Liang predecessors, despite very altered circumstances. Ch'ung-tsung died after fifty-three years on the throne, in the sixth month of 1139, at the age of fifty-six, and he was succeeded by his sixteen-year-old son Wei-ming Jen-hsiao (temple name Jen-tsung) who would also reign for more than half a century. These two long reigns at last gave some stability to the Hsia royal house. After his enthronement, Jen-tsung honored his mother, Lady Ts'ao, with the title of "state mother" (kuo mu) and appointed as his own empress a certain Lady Wang, who came from a venerable Tangut clan.92 The first years of Jen-tsung's reign were troubled with uprisings and widespread banditry, the result of the serious ravages of war and natural disaster throughout the north and northwest of China in the previous three or four decades. According to Wu Kuang-ch'eng (again his account is not corroborated elsewhere), in 1140 a group of discontented Khitan exiles rebelled under the leadership of a certain Li (or Hsiao) Ho-ta. The rebels surrounded Ling-chou before being crushed that winter. Famine and earthquake sparked serious unrest in the environs of Hsia-chou and Hsing-chou in 1142—3. Jen-tsung enacted tax remissions and relief measures and was greatly obliged to his Chinese commander, Jen Te-ching, for suppressing Li Ho-ta and pacifying the tribal uprisings. Jen-tsung reportedly wanted to reward Jen Te-ching with a palace appointment but was persuaded by his skeptical counselor, Wei-ming Jen-chung, to keep Jen Te-ching away from court in a field command post at Ling-chou. 9 ' In the following year, 1144, Jen-tsung began to introduce certain Confucian institutions into the government. First he decreed the establishment of schools throughout the country, and a secondary school (hsiao hsiieh) was opened up inside the palace for imperial scions aged seven to fifteen. Another school, the Ta Han t'ai-hsueh (Superior school of Chinese learning), was created in 1145. A Confucian cult was formally inaugurated, and orders were given that temples be built and sacrifices maintained therein throughout the land. In 1147 examinations for the selection of candidates to office were instituted.*» Although it is known from other sources that the Hsia state granted examination degrees, nowhere does the Tangut law code discuss 92 55, 486, p. 14024. 93 Wu, Hsi Hsia shu shih, 35, pp. 9a— n b , 16a; CS, 134, p. 2869. 94 55, 486, pp. 14024—5. Cambridge Histories Online © Cambridge University Press, 2008 200 THE HSI HSIA examinations of this sort as a method of official recruitment, whereas it does devote many articles to the regulations governing the inheritance of office and rank, which probably remained the more common path to office. Finally in 1148 Jen-tsung established an Inner Academy {Nei-bsiieh), of which no description survives in the Sung history's terse account, and staffed it with renowned scholars. Speculation on the possible motives for these measures must take into account the contemporaneous rise of the Chin state as the Tanguts' new eastern neighbor. In the 1140s and 1150s the Jurchen rulers undertook similar measures, but on an incomparably vaster scale, as it was their ambition to create a Chinese-style court and bureaucracy to match the supreme position they had achieved in East Asia by means of war and diplomacy. Tangut envoys traveled yearly to the Chin court and so were reasonably well informed about Jurchen activities, and vice versa. If the often-faulty Sungshih chronology is correct, however, then the Tangut reforms preceded those of the Jurchen by five to ten years; for example, the Chin ruler Wan-yen Liang (r. 1149—61) decreed the construction of Confucian temples only in 1156, a decade later than those established by the Tanguts.95 It may be supposed that the Tangut ruler was motivated more by domestic considerations, namely, the development of an influential "Confucian" following at Hsing-chou and a commitment to a more rationalized and ritualized exercise of imperial prerogative and to the bureaucratic process as a means of combating the perceived weakness and corruption in the Tangut government and also the deeprooted influence of the military elite. Representative of this military elite, Jen Te-ching evidently had a patron in the army chief, Wei-ming Ch'a-ke, through whom he sought a position at the capital. At first, as we have seen, his efforts were blocked by Wei-ming Jenchung, but after the latter's death Jen Te-ching reportedly bribed his way to court and was made the director of the Department of State Affairs (Shang-shu ling). Soon he became director of the Central Secretariat (Chung-shu ling), though it is not certain how long he actually held this post or indeed what duties it entailed in the Hsia state. After the death of Wei-ming Ch'a-ke in 1156, Jen Te-ching began to concentrate power in his own hands, to appoint his own relatives to office, and to suppress detractors.96 The Sung shih confirms that in 1160 Jen Te-ching obtained from the Tangut monarch the noble title of prince of Ch'u, the first and only Chinese so far known to have been enfeoffed by a Hsia ruler. Many courtiers must have viewed this as a grave breach of dynastic precedent and a dangerous 95 Tao Jing-shen, Tbejurcben in twtlftb-century China: A study of linicization (Seattle, 1976), pp. 41—4. 96 Wu, Hsi Hsia shu shih, 36, pp. 3b—7a, 12a. Cambridge Histories Online © Cambridge University Press, 2008 HSIA COMES OF AGE: CH'UNG-TSUNG AND JEN-TSUNG 2OI arrogation of power, as such honors were normally reserved for meritorious royal scions. Later that year the chief minister reportedly denounced the newly established schools as useless Chinese institutions inappropriate to Hsia society, wasting meager resources in the support of parasitic scholars. Jen-tsung's response to this attack on his closest associates, the scholars and monks, is unknown, but Jen Te-ching was clearly defeated, for the schools remained intact. In 1161 the emperor went further and also established the Han-lin Academy to compile the dynasty's historical records. This innerpalace agency, along with the Censorate and the schools, became centers of opposition to the chief minister, whose domains in the Secretariat and Military Bureau were moved outside the inner court in 1162.97 In 1161-2 Hsia became peripherally involved in the Chin war against the Sung. The Sung authorities in Szechwan tried unsuccessfully to solicit Tangut support against the Jurchens, while Hsia troops briefly occupied both Sung and Chin territories in Shensi, which Hsia claimed as its own. Jen Teching probably had a hand in these activities, considering his control over the army and his subsequent efforts to enlist the support of Szechwan officials for his private schemes. From 1165 to 1170 the chief minister labored to create for himself an independent satrapy in northern Shensi and the Ordos, with his headquarters at Ling-chou and the nearby Hsiang-ch'ing Army Superintendancy. He further meddled in the troubled affairs of some Chuang-lang (Hsi-fan) tribes whose homeland now unhappily straddled the ill-defined border region between Sung, Chin, and Hsia in the T'ao River valley. In the event, a jurisdictional dispute arose between Chin and Hsia, presaging the unrest that would engulf this area in the early thirteenth century. Jen Te-ching then tried without success to curry favor with the Chin emperor, Shih-tsung (r. 1161 — 89), who shrewdly parried the minister's veiled overtures. Finding no encouragement in that quarter, Jen Te-ching began to exchange secret messages with the Sung pacification commissioner in Szechwan. A Hsia patrol captured one of the latter's agents carrying a letter to the chief minister and forwarded the incriminating document to his emperor, who passed it on to the Chin court.98 Before obtaining this firm evidence of the Hsia minister's treachery, the Chin ruler had received reports from captured Sung spies, among others, of suspicious activity along his southwestern border. The Chin court also learned that Jen Te-ching had sent a large number of troops and laborers to 97 SS, 486, p. 14025; Wu, Hsi Hsia shu shih, 36, pp. I 3 b - i 4 b . 98 On the Chuang-lang, see CS, 9 1 , pp. 2016—18. On Hsia contact with Szechwan, see SS, 34, pp. 6 4 3 - 4 ; 486, p. 14026; Chou Pi-ta (1126—1204), Wen-cbang chi (Chou l-kuo Wen-chung kung wen chi) (SKCP ed.) 61, pp. I7b—i8a; 149, pp. i6a-i7a; and CS, 61, p. 1427. Cambridge Histories Online © Cambridge University Press, 2008 2O2 THE HSI HSIA repair and fortify Ch'i-an-ch'eng (the former Chi-shih), the Tangut outpost in Chuang-lang territory. Shih-tsung dispatched officials to investigate, but they were too late to halt the construction and unable to confirm rumors of a Sung-Hsia conspiracy. In reply to their inquiry, the Tanguts (i.e., Jen Teching himself) asserted the purely defensive nature of the fortifications." The death of Jen Te-ching's daughter, the empress dowager Jen, in 1169 or 1170, supposedly spurred the chief minister to pressure Jen-tsung to grant him the eastern half of the country, a statelet that he named Ch'u. To ratify the deed, Jen Te-ching further persuaded his sovereign to request a patent of investiture for him from the Chin court. The Chin emperor Shih-tsung expressed his profound disapproval of the whole affair and privately wondered why the Hsia ruler could not discipline his unruly subject. He denied the patent, returned the embarrassed embassy's gifts, and promised to send an official investigation. This proved unnecessary. In the eight month of 1170 Jen-tsung's own men secretly rounded up and executed the chief minister, his clan, and his adherents. A Hsia delegation delivered a letter of gratitude from Jen-tsung to the Chin emperor, politely averring that no further assistance would be required, except to maintain the peace and integrity of their common border in the area where the former minister had caused clashes with the Tibetans.100 Even without any reliable sources regarding Jen Te-ching, it is possible to venture some explanations of the episode, and especially of Jen-tsung's conduct. First, the Tangut emperor was not an absolute monarch; he was subject to the restraining influence of traditional tribal practices. One important institutional restraint was the special position of the chief counselor, especially when he was a member of the consort clan (which may or may not have been the case here). Both in Tibet and among the Uighurs, chief ministers wielded significant powers, and the influence of their models on Tangut government is beyond doubt.101 Another important point is that Jen-tsung was the first Tangut emperor who did not grow up on the battlefield and who did not cultivate close personal ties with the army. Instead he delegated control of military affairs first to his uncle Wei-ming Ch'a-ke and then to Jen Te-ching. For a long time this was a convenient, and from a military point of view, a successful arrangement. But when need arose, the emperor had to look elsewhere for support; he could not challenge the military head-on. 99 CS, 91, p. 2017-18. 100 CS, 134, p. 2869—70; Wu, Hsi Hsia shu shih, 37, p. 13a. 101 See Sato Hisashi, Kodai Cbibetto shi kenkyu (Kyoto, 1958—9), vol. 2, pp. 11—14, 28-9, 711-38; Elizabeth Pinks, Die Uiguren von Kan-chou in dtr frubcn Sung-Zeit, pp. 106—7, ' ' 4 ~ ' 5 J Abe Takeo, "Where was the capital of the West Uighurs?" in the Silver jubilee volume of the Zimbuti kagaku kenkyusho (Kyoto, 1954), p p . 439—41. Cambridge Histories Online © Cambridge University Press, 2008 HSIA COMES OF AGE: CH'UNG-TSUNG AND JEN-TSUNG 203 Jen-tsung apparently found that support in the educational and cultural institutions in which he had been raised and that he fostered all of his life. He cultivated the civilian norms of Confucianism and propagated the imperial persona of an aspiring Bodhisattva. Jen-tsung's patronage of Buddhism and cultivation of bodhichitta (enlightened mind) are nowhere alluded to in Chinese chronicles, but extensive Tangut materials reveal the magnitude and significance of his Buddhist activities, activities that, after all, were the traditional occupation of every Tangut ruler. That Jen-tsung busily engaged in accumulating good merit to attract support, to heighten and display his prestige and moral authority, and thus to undermine his rival in no way • diminishes his religious sincerity. He waged an ideological battle with Jen Te-ching in which he held all the weapons and ultimately forced his minister into treachery, repudiating the very values that he needed to legitimize himself as an independent ruler. When Jen-tsung's minister was finally exposed as a traitor, his elimination was a routine matter; whatever justifiably frustrated elements had banded to his cause now deserted him. There certainly existed an impulse to split up the Tangut empire into a Chinese-oriented Ordos state and a steppe-oriented Ho-hsi state. This impulse reflected deep-rooted geopolitical and cultural realities, not simply a tribal tendency toward decentralization. But ultimately much stronger was the contrary impulse to maintain the integrity of the realm, which sprang from another overarching geopolitical reality: the existence of a system of interstate relations in which Hsia, Sung, and Liao (later replaced by Chin) acted as a stable tripod of power in continental East Asia. Neither Sung or Chin would tolerate the creation of another independent kingdom in north China; witness the earlier failure of Chin to rule north China through its puppet regimes of Ch'i or Ch'u. If Jen Te-ching represented a conservative element in Hsia society discontented with the direction of change in government policies, Jen-tsung, on the other hand, embodied thefirmlyentrenched legitimacy of Wei-ming dynastic rule and the established territorial integrity of the state. His rule represented a civilian government headed by a semidivine Buddhist ruler, and his power was based on a compromise with the military establishment (i.e., with the tribal aristocracy) by which hereditary privileges were confirmed by the state, in exchange for loyalty to the throne."" Many of these points are elaborated in the Tangut law code: "The revised and newly endorsed code of law of the age of 102 See, for example, the argument in Shimada Masao, Ryocho kansii no kinkyQ, Toyo hoshi ronshu no. 1 (Tokyo: Sobunsha, 1978), English summary. One of the first scholars to comment on the Buddhist aspect of Tangut rulers was Paul Serruys, "Notes marginales sur le folklore des Mongols Ordos," Han Him: Bulletin du centre d'ttuda Sinologiques de Pikin, 3, nos. 1—2 (1948), p. 172. On the Buddhistinspired native honorifics for Hsia emperors, see Pu P'ing, "Hsi Hsia huang ti ch'eng hao k'ao," pp. 70-82. Cambridge Histories Online © Cambridge University Press, 2008 204 THE HSI HSIA celestial prosperity" (T'ien-sbeng chiu kai hsin ting chin ling), which, perhaps not coincidentally, was issued at the end of the T'ien-sheng reign period (1149— 70), about the time of Jen Te-ching's execution.IO3 The man who succeeded Jen Te-ching as the dominant minister, Wo Taoch'ung, came from a Tangut family that for generations had supplied historiographers to the Tangut court. A Confucian scholar and teacher of Tangut and Chinese, Wo Tao-ch'ung translated the Lunyii (Analects) and provided it with a commentary, also in Tangut. He wrote as well, in Tangut, a treatise on divination, a topic of enduring interest to the Tanguts. Both works were published during his lifetime and were still extant during the Yuan period. Upon Wo Tao-ch'ung's death Jen-tsung honored him by having his portrait painted and displayed in all the Confucian temples and state schools.IO4 The Tangut emperor Jen-tsung was a superb propagandist who, like his Jurchen counterpart Shih-tsung, mastered the public role of a virtuous ruler. But whereas Chin Shih-tsung gained the Confucian reputation of a "Little Yao or Shun," Jen-tsung's path to perfection is strewn with allusions to Buddhist sainthood.IO5 Jen-tsung oversaw and participated in the editing and revision of all the Buddhist translations undertaken at the courts of his predecessors. Thus, although it was further refined in the Yuan period, by the end of his reign the Tangut Tripitaka was virtually completed and was printed in its entirety early in the fourteenth century.106 Religious zeal prompted the emperor's most eloquent and extensive propaganda acts. Throughout his reign the emperor and members of his family, notably his second consort empress Lo (who was of Chinese descent), sponsored massive printings and distributions of favorite Buddhist texts on various commemorative occasions. The most spectacular of these publications occurred in 1189. Celebrations in that year honoring the fiftieth anniversary ofJen-tsung's 103 See n. 87. 104 Yii Chi (1272 —1348), Tao-yiian bsiieb ku lu (KHCPTS ed.) 4, pp. 83—4; Ch'en Yuan, Western and Central Asians in China under the Mongols: Their transformation into Chinese, trans. Ch'ien Hsing-hsi and L. Carrington Goodrich (Berkeley and Los Angeles, 1966), p. 128, in which the surname Wo is misread as Kuan; Wu Chi-yu, "Sur la version tangoute d'un commentaire du Louen-yu conserved a Leningrad," T'oungpao, 55 (1969), pp. 298—315. 10; See, for example, Nevskii Tangutsiaia filologiia, vol. 1, p. 82; and the text of Jen-tsung's dedication to a newly restored bridge over the Hei-shui in Kansu, in Wang Yao, "Hsi Hsia Hei-shui ch'iao pei k'ao pu," Chung yang min tsu htiiih yuan hsu'eh pat, 1 (1978), pp. 51—63. This stele inscription is registered in Chung Keng-ch'i, comp., Kan-choufuchih(1779; repr. in Chung-kuofangchihts'ungsbu: Hua-pei ti fang no. 56i;Taipei, 1976), 13,pp. 1 ib—12a, but did not come to the attention of either Wu Kuang-ch'eng or Tai Hsi-chang. Jen-tsung's dedication is translated from the Chinese by E. Chavannes, "Review of A. I. Ivanov: Stranitsa iz istorii Si-sia (Une page de l'histoire du Si-hia; Bulletin de I'Academie imperiale des sciences de Saint-Peiersbourg, 1911, pp. 831-836)," T'oung pao, 12 (1911), pp. 441—6. 106 Wang Ching-ju, Hsi Hsia yen chiu, vol. 1 (Peking, 1932), pp. 1—10; Heather Karmay, Early SinoTibetan art (Warminster, 1975), pp. 35—45. On Tangut Buddhist activities and the Tangut Tripitaka before and after 1227, see Shih Chin-po, Hsi Hsia wen hua, pp. 64—105. Cambridge Histories Online © Cambridge University Press, 2008 THE HSIA STATE AND THE MONGOLIAN CONQUEST 205 accession were highlighted by the printing and distribution of 100,000 copies of the Kuan Mi-le p'u-sa shang sheng Tou-shuai t'ien ching (Sutra on the visualization of the Maitreya Bodhisattva's ascent and rebirth in Tushita Heaven) and 50,000 copies of several other sutras, printed in both Tangut and Chinese. The year 1189 was a year of changes in East Asia; it also saw the death of Chin Shih-tsung and the abdication of Sung Hsiao-tsung. Hence the Tangut ruler had cause for a liberal demonstration of his gratitude to the Buddha. Despite occasional incidents with the Jurchens, peace had prevailed for most of his long reign. In general the two courts maintained fairly cordial relations, though some disagreements did arise, increasingly so toward the end of the twelfth century, because of economic and minor territorial disputes. The Jurchens charged that at border markets the Tanguts traded worthless gems and jades for their good silk textiles, recalling Northern Sung complaints about Khotanese envoys flooding Chinese markets with poor jade. Consequently, the Chin closed the border markets at Lan-chou and Pao-an in 1172 and did not reopen them until 1197. Next complaints arose of illegal trafficking across the Shensi border, so that the Sui-te market was also shut, leaving only a market at Tung-sheng and one at Huan-chou. Drought and famine had swept north China in the 1170s; Tangut border raids increased during the same period, culminating in 1178 with a Tangut sack of Lin-chou (which was now in Chin hands). In 1181 the Chin emperor finally reopened the Sui-te market and granted three-day trading privileges to Tangut envoys visiting the Chin capital.107 In 1191 some Tangut herdsmen strayed into Chen-jung prefecture and were chased by a Chin patrol, which they took captive. They then ambushed and killed a pursuing Jurchen officer. Jen-tsung refused to extradite the guilty parties, assuring the Chin that he had already punished them. These, however, were comparatively minor disturbances of generally peaceful relations. After the deaths of Shih-tsung of Chin in 1189 and Jen-tsung in Hsia in 1193, however, the short reigns of their respective successors proved to be the prelude to an era of internal disorder and conflict between the states, the principal cause of which was the growing power and unification of the Mongols under Temujin (the future Chinggis khan). THE LAST YEARS OF THE HSIA STATE AND THE MONGOLIAN CONQUEST When Jen-tsung died in 1193 at the age of seventy, his eldest son by Empress Lo, Ch'un-yu (Huan-tsung; r. 1193—1206), ascended the throne at the age of 107 CS, 134, pp. 2870-1; Yii-wen Mou-chao (13th c ) , Ta Chin kuo chih (KHCPTS ed.), 17 passim. Cambridge Histories Online © Cambridge University Press, 2008 2o6 THE HSI HSIA seventeen. Very little is known about Huan-tsung's reign, but certainly the most crucial event came with the first Mongolian raid into Hsia territory in 1205. From 1206 until its destruction in 1227 by Chinggis khan, the Tangut royal house endured a period of unprecedented instability. But neither internal decay nor inherent weakness finally undid the dynasty. Rather, the Hsia state, like its more powerful neighbors, was destroyed by the Mongols, the new steppe power whose emergence fatally destabilized the Sung—Chin—Hsia balance of power in East Asia. Anti-Chin and anti-Mongolian factions formed at the Hsia court, with usurpation and abdication occurring for the first time in the state's history. Since the 1170s, stirrings from the steppe had occasionally affected Hsia and Chin and appear in the official records. One reason that the Jurchens closed three of their western border markets with Hsia was their suspicion of the Tanguts' espionage activities and their possible dealings with the KharaKhitai far to the west, which they considered hostile to Chin interests.108 It is also known that a Kereyid prince who was overthrown by Temiijin's father, probably in the 1170s, found refuge in Hsia, never to be heard of again. Another Kereyid prince reportedly spent some time in exile among the Tanguts, who favored him with the honorary epithet of Jakha Gambu (roughly, "elder state counselor"), by which name he is known to posterity. Jakha Gambu's fickle loyalties were tolerated by Temiijin because his brother To'oril (Ong khan) was Temvijin's own sworn father and because Temiijin's family took several of Jakha Gambu's daughters in marriage. One of them, the famous Sorghaghtani Beki, mothered Mongke, Khubilai, and Hiilegii. Another of Jakha Gambu's daughters evidently married the Tangut king, and her beauty is said to have inflamed Chinggis khan in his final onslaught against the Tangut country.109 It seems possible, then, that the Tangut ruling house had already entered into the steppe network of marriage alliances, which might help explain their special status later in the Mongolian empire. The Kereyid connection with Hsia did not end there. After To'oril's eventual defeat by Temiijin in 1203, the Kereyid leader's son, Ilkha Senggiim, fled through Edzina to northeast Tibet, whence he was driven to the Tarim basin and finally killed by a local chief.'IO Although the Tangut 108 CS, 50, p. 1114; SS, 486, p. 14026. 109 Rashid al-DIn, Sbornik letopiiei, vol. i, pt. 2, trans. O. I. Smirnova(Leningrad, 1952), pp. 109-10, 127; Paul Pelliot and Louis Hambis, trans., Hiiloire da campagna de Gengis Khan, Cbthg-wou Ti'inTcheng Lou (Leiden, 1951), pp. 230, 261; Iurii N. Rerikh (George N. Roerich), "Tangutskii titul dzha-gambu Kereitskogo," Kratkiesoobsbcbeniia iiutituta narodov Azii, 44 (1961), pp. 4 1 - 4 . n o Rashid al-DIn, Sbornik letopisti, vol. i , p t . 2 , p . 134; Sbengwuch'in chengluchiaochu, in Meng-kusbib liao ssu cbung, ed. Wang kuo-wei (Peking, 1926; repr. Taipei, 1962, 1975), p. 107. The notice in Sung Lien et al., eds., Yiian shih (Peking, 1976), i , p . 23 (hereafter cited as YS) is misplaced under the year 1226 as an explanation for Chinggis's invasion in that year. Cambridge Histories Online © Cambridge University Press, 2008 THE HSIA STATE AND THE MONGOLIAN CONQUEST 207 authorities had apparently refused to grant refuge to the Kereyid fugitive, his flight across Tangut territory became the pretext for a Mongolian attack on Ho-hsi in 1205. Several fortified settlements were plundered and many cattle taken away.'' * In 1206 Temiijin was proclaimed Chinggis khan, and a coup in Chunghsing brought a new ruler to the Tangut throne. Huan-tsung was deposed by a cousin, Wei-ming An-ch'iian (Hsiang-tsung; r. 1206-11), and died in captivity a month later. Grand Empress Dowager Lo was coerced into persuading the Chin court to grant the usurper a patent of investiture.112 Thereafter she disappears from the record, presumably dispatched to a monastic retirement. The following year Chin lost the allegiance of the Onggiid and the tribally mixed border guards {jiiyin; Chinese: chiu) on their northwest frontier, and the nearby Tangut garrison of Wu-la-hai was sacked by the Mongols. "3 The Mongols could now invade with impunity both Shansi and the Ordos. The Mongolian raiding party did not withdraw from Wu-la-hai until spring of 1208. The Hsia sent a succession of embassies to the Chin capital, presumably seeking a united front against the Mongolian advance, but unfortunately for both states, the Chin monarch died that winter without an heir and was succeeded by a far less competent relative known to history as Wei-shao wang (deposed in 1213). This prince refused to cooperate with the Tanguts and is said to have commented: "It is to our advantage when our enemies attack one another. Wherein lies the danger to us?"1'4 Whatever actually happened, Tangut—Jurchen relations deteriorated rapidly from this point onward. In the autumn of 1209, after receiving the voluntary submission of the Uighurs of Turfan, Chinggis khan launched a major invasion of Hsia. Entering Ho-hsi through a pass "north of Hei-shui-ch'eng and west of Wu-la-haich'eng," Mongolian troops overcame a Tangut army led by an imperial prince and captured its deputy commander. Advancing to Wu-la-hai, the Mongols again overwhelmed that garrison, whose commander surrendered, and took prisoner a high official called Hsi-pi Ou-ta. From Wu-la-hai the Mongolian army turned south and marched to K'o-i-men, the garrison situated west of in Wu Kuang-ch'eng's tale that becuase of their deliverance from the Mongolian menace on this occasion the Tanguts changed the name of their capital from Hsing-ch'ing to Chung-hsing, must, however, be rejected as a fiction. See n. 46; YS, i , p . 13; Rashld al-Dln, Sbornik Utopiiti, vol. i , p t . 2, p. 150; Sbeng wu ch'in cbeng lit chiao chu, p. 118. 112 CS, 134, p. 2871. 113 Paul Buell, "The role of the Sino-Mongolian frontier zone in the rise of Cinggis Qan," in Studies on Mongolia: Proceedings of the first North American Conference on Mongolian studies, ed. Henry G. Schwarz (Bellingham, 1979), pp. 66—8. See also Igor de Rachewiltz's comments on the Jiiyin (chiu) troops in "The secret history of the Mongols: Chapter eleven," Papers on Far Eastern History, 30 (1984), pp. 105-7. 114 CS, 62, p. 1480; 12, p. 285; see Ta Chin kuo chih, 21, pp. 2 3 - 4 , for the Jurchen ruler's remark. Cambridge Histories Online © Cambridge University Press, 2008 2O8 THE HSI HSIA Chung-hsing to guard the capital. Here, in the rugged Alashan fastness of K'o-i-men, Mongolian troops engaged a formidable opposing army led by another imperial prince. Eventually defeating this force and imprisoning its commander, they proceeded to invest the capital city (see Map 13). Employing an ancient stratagem, the Mongols diverted waters from the Yellow River canals to flood the city. But the dikes broke, inundating their own camp as well and forcing them to lift the siege and withdraw. First, however, they sent the captive Hsi-pi Ou-ta into Chung-hsing to negotiate on their behalf. In 1210 the Tangut ruler nominally submitted to Chinggis khan and handed over a daughter in marriage to the Mongolian chief, as well as delivering to him a large supply of camels, falcons, and woven textiles."3 Thereby Chinggis secured his western flank before attacking the Chin state. Not long after, the Tanguts started to harry Chin border prefectures, whereupon diplomatic exchanges between the two neighbors abruptly declined and ceased altogether after 1212. In 1211 the Tangut emperor died under obscure circumstances and was replaced on the throne by a cousin, Wei-ming Tsun-hsu (Shen-tsung; r. 1211-23; d. 1226). This prince held the distinction of earning the first native chin-shih degree awarded to an imperial scion, naturally standing first in his class.1'6 The dilemma of how to deal with the Mongols and whether or not to abandon the long-standing alliance with Chin had certainly thrown the Tangut court into turmoil. Shen-tsung appears to have played all sides at once. To appease the Mongols, he launched a war against Chin's southwest provinces (Lin-t'ao circuit), providing the Tanguts with an opportunity to avenge old territorial grievances. Control over Lin-t'ao became critical to the Jurchens after 1214, when in the face of overwhelming Mongolian pressure they moved their principal capital from Yen-ching (Peking) south to Pien (K'ai-feng). Because the Sung court had halted delivery of their tribute payments, Chin was in a desperate financial situation, and their territories in Shensi became a vital source of food, manpower, and horses for the beleaguered Jurchen hemmed in in Honan. At the end of 1214 a rebellion supported by Hsia engulfed Lan-chou, seriously undermining the Chin hold on the area for years. Moreover, in 1214 the Tangut court initiated negotiations with the Sung authorities in Szechwan, proposing joint military action against Chin's western territories. These negotiations continued until 1221 but led to only one abortive joint action in 1220. "7 The Jurchens were only once able to field an army against 115 YS, 60, p. 1452; i , p. 14; 169, p. 3977; Rachewiltz, "The secret history," pp. 84—5. 116 CS, 134, p. 2871. 117 SS, 486, p. 14027; 40; pp. 774-5; Li Hsin-ch'uan, Cbien-yen i lai cb'aoycb isa chi, repr. in vols. 21— 2 of Sung shih tzu liao ts'ui pien, 1st series, ed. Chao T'ieh-han (Taipei: 1967), 19, /' cbi, p. 8b. Cambridge Histories Online © Cambridge University Press, 2008 THE HSIA STATE AND THE MONGOLIAN CONQUEST 209 H HEI-SHAN WEI-FU CHUN-SSU A-LA-SHAN DESERT I PAI-MA CHIANG-CHEN CHUN-SSU O R DOS DESERT Sheng-kuei 1 fortress K'o-i mens/ Chung-hsingfu, , Ting-chou Pai-ch'ih fortress Ching-chou Ling-chou Yen-chou HSI-SHOU PAO-CH'IN CHON-SSU *\ f U " N Route of Mongolian forces SC) 100 150km 50^ MAP 13. The first Mongolian invasion of Hsi Hsia, 1209 Cambridge Histories Online © Cambridge University Press, 2008 lOOmiles 2IO THE HSI HSIA Hsia, in the latter half of 1216, when Hsia gave passage to a Mongolian army that crossed the Ordos to attack Chin territory in Shensi, and provided auxiliary troops to assist in the Mongolian operations."8 The war with Chin was politically unpopular and economically ruinous for Hsia. Opposition to it mounted at the Tangut court, and peace feelers were issued from time to time, but with no result. It also is clear that powerful voices at the Tangut court disdained the policy of cooperating with the Mongols, despite a steady stream of Tangut defections to the Mongolian camp. In the winter of 1217 or early in 1218 a Mongolian force approached the Hsia capital, either to enforce Tangut compliance with an earlier promise to support Chinggis's Central Asian campaign against Khwarazm or to punish a Tangut refusal to render such aid. Although it is not clear whether a battle actually took place, there is no doubt that the Tanguts had defied the Mongolian order and that a certain Asha Gambu gained notoriety for his outspoken contempt for the great khan's pretensions. "9 Chinggis khan deferred reckoning with the Tanguts until after his Central Asian expedition, on which he embarked in 1219, leaving Mukhali in charge of operations in north China. The Hsia war against Chin, meanwhile, showed no sign of slackening. Then, in 1223, everything changed: Mongolian pressure on both Chin and Hsia temporarily eased after Mukhali's death in the third month. At the end of the year Shen-tsung abdicated to his second son, Wei-ming Te-wang (Hsien-tsung; r. 1223—6), and with a new Chin emperor also on the throne, negotiations to end the war with Chin began late in 1224. The peace agreement, formally proclaimed in the ninth month of 1225, established a fraternal relationship between the two states, with Chin as the elder and Hsia the younger brother. Each court retained the use of its own reign titles in correspondence. Subsequent talks cleared up disputes regarding points of ritual, border markets, and so forth.120 This unorthodox arrangement represented a diplomatic triumph for Hsia, albeit a short-lived one. The accounts of Chinggis's last campaign against Hsia and of his death are extremely confused and full of conflicting details. Most likely Chinggis set out in the winter of 1225 and invaded Ho-hsi in spring of 1226. Since withholding auxiliaries for the Mongolian Central Asian campaign in 1219, the Tanguts had committed further offenses, refusing to send a hostage 118 CS, n o , p. 2421; 14, pp. 3 1 8 - 2 2 . 119 YS, 1, p. 20, following CS, 15, p. 334; Rachewiltz, "The secret history of the Mongols: Chapter eleven," pp. 95—6. In this passage of the "Secret history," Chinggis khan's envoy reminds the Hsia king that in 1209 he (actually his predecessor) had promised to act as the khan's right hand (i.e., his western Bank). 120 CS, 17, pp. 375—6; 38, p. 869; n o , pp. 2424, 2433—4; 62, pp. 1487—8. Cambridge Histories Online © Cambridge University Press, 2008 THE HSIA STATE AND THE MONGOLIAN CONQUEST 211 prince to the Mongolian khan's court and conspiring with various unnamed powers to resist the Mongols. 121 The Secret history of the Mongols relates an incident, nowhere else recorded: Progressing toward Tangut territory in the winter of 1225, Chinggis stopped to hunt wild horses and suffered serious injuries when his horse bolted and threw him. The march was halted and camp pitched to await the khan's recovery. As a face-saving measure, Chinggis sent envoys to inform the Tangut ruler he was being called to account for his reckless impudence. From the Tanguts' response, Chinggis would judge whether to withdraw or proceed with his invasion. When the Mongolian envoys were received at the Tangut court, Asha Gambu was again present, for he claimed responsibility for the reckless words he had spoken earlier and further mocked the Mongols, challenging them to battle. Upon hearing this reply, Chinggis pledged to avenge the insult, "Even though we may die, let us challenge their boasts!"'22 What lies behind the episode we can only speculate, but it appears that the Tangut court was again dominated by a powerful minister rather than by its emperor. Certainly Asha Gambu's defiant refusal to negotiate or compromise provoked the Mongols' devastatingly thorough obliteration of the Tangut state. The Mongolian armies first reduced the Tanguts' Ho-hsi provinces, isolating the capital. Hei-shui (Edzina) and neighboring locales fell in the second month of 1226. From his summer camp at Hun-ch'ui-shan (perhaps the Ch'ilien Mountains), the Mongolian khan directed operations against Su-chou and Kan-chou, and the Mongolian general Siibetei led a separate force to subdue the Sa-li Uighur prince and other tribes inhabiting the mountains south of Kan-chou and Sha-chou (see Map 14). Two Tangut officers in Mongolian service, Ch'a-han and Hsi-li Ch'ien-pu, participated in the campaigns. At Su-chou, Hsi-li Ch'ien-pu's elder brother was commander of the garrison and put up a spirited defense despite his sibling's pleas to surrender. When the city fell, the Mongols spared Hsi-li's relatives but massacred the rest of the population. I23 At Kan-chou, Ch'a-han likewise tried to persuade his own younger brother and father, who commanded the city's defenses, to avoid the fate of Su-chou. The loyalist deputy commander killed them both and refused to surrender. 121 YS, 1, pp. 23—4; Paul Pelliot, Notes on Marco Polo (Paris, 1959), vol. 1, pp. 3 0 9 - 1 1 ; Igor de Rachewiltz, "The secret history of the Mongols: Chapter twelve," Papers on Far Eastern History, 31 (1985), pp. 21—5, 4 4 - 5 5 , and "The Hsi-yu lu by Yeh-lii Ch'u-ts'ai," Monumtnta Serica, 21 (1962), pp. 6 3 - 4 , n. : 3 8 . 122 Rachewiltz, "The secret history of the Mongols: Chapter twelve," p. 23. 123 YS, 121, p. 2977; 122, pp. 3008—9, 3011; 120, p. 2955. On Hsi-li Ch'ien-pu, see Pai Pin and Shih Chin-po, "Ta Yuan Su-chou lu ye k'o ta lu hua ch'ih shih hsi chih pei," Min tsu yen chiu, 1 (•979). PP- 6 8 - 8 0 . Cambridge Histories Online © Cambridge University Press, 2008 ,_-'—\ 100 200 100 300 km 200 miles MAP 14. Chinggis's invasion of Hsi Hsia, 1226-7 Cambridge Histories Online © Cambridge University Press, 2008 T H E HSIA S T A T E A N D T H E M O N G O L I A N C O N Q U E S T 213 Nevertheless, after Kan-chou's subjugation Ch'a-han managed to save its populace, owing to his great prestige and length of service in the Mongols' cause. Only thirty-six men who had resisted surrender were executed. I24 In autumn the main Mongolian army passed southeast to Hsi-ling (Liangchou), an important urban center. Its defender, Wo Cha-tse (a descendant of Wo Tao-ch'ung), surrendered after a brief resistance, and the districts of Ch'o-lo and Ho-lo quickly followed suit. I25 From there Chinggis khan proceeded to cross the Yellow River and march downstream northeastward toward Ling-chou. By this time it seems that the Tangut emperor had died of fright, at the age of forty-five. A luckless kinsman, Wei-ming Hsien, inherited the throne. When the Mongols invested Ling-chou in the eleventh month, the Tanguts dispatched one last ambassador to the Chin court, requesting the Jurchens to withhold further embassies. A large Tangut army under Wei-ming's command marched southwest to relieve the siege of Ling-chou. Chinggis hastened across the frozen river to meet the advancing relief force and routed it in an icy battle. Yeh-lii Ch'u-ts'ai witnessed the collapse and sack of Ling-chou in December: "All the officers contended with each other to seize children, women and valuables. His Excellency {i.e., Yeh-lii] took only a few books and two camel-loads of rhubarb." 126 After reducing Ling-chou, Chinggis khan set up his camp nearby at Yenchou-ch'uan. Leaving behind an army to close in on the Hsia capital, he himself then moved southward, again crossing the Yellow River, and advanced west through Lin-t'ao to Chi-shih-chou. In spring of 1227 the Mongol khan swept through Lin-t'ao, conquering its major districts, and halted to "escape the summer heat" in the Liu-p'an Mountains. I27 The khan was an ailing man, and this is probably where he died. The siege of Chung-hsing lasted for six months. Finally, the hopelessly surrounded Tangut monarch capitulated in the sixth month of 1227. News of Chinggis khan's death was concealed until Wei-ming Hsien emerged from the capital walls according to the agreed terms of his surrender. The Mongols slew him forthwith and sacked the city. Ch'a-han managed to avert a total massacre and rescued scattered survivors of the debacle.128 It has been suggested that the Mongols' treatment of the Tangut royal 124 YS, 120, pp. 2 9 5 5 - 6 . 125 YS, 134, p. 3254; 146, pp. 3 4 6 5 - 6 ; 1, p. 24. On the reading ch'o, see Paul Pelliot, Review of E. Haenisch, "Die letzten Feldziige Cinggis Han's und sein Tod. Nach der ostasiatischen Ueberlieferung," Alia Major, 9 (1933), pp. 503—51, and Toungpao, 31 (1934), p. 161. 126 YS, 146, p. 4355; Rachewiltz, "The His-yu lu," p. 65, n. 142. Tangut rhubarb was famed for its medicinal properties. 127 Pelliot, Notes on Marco Polo, vol. 1, pp. 310— 11, failed to identify Chi-shih chou as the Chin prefecture in Lin-t'ao circuit south, not north, of the Yellow River. 128 YS, 120, p. 2956. Cambridge Histories Online © Cambridge University Press, 2008 214 T H E H S I HSIA family and the unusually sanguinary massacre of the Chung-hsing population "were meant to provide Chinggis with a suitable escort in the afterlife," in which the "virtue" of the enemy king was transformed into a beneficial protective spirit serving the deceased khan." 9 Despite the picture of total devastation painted here, many Tangut survived the fall of Hsia and entered Yuan service. As se-mu (Western and Central Asians) they enjoyed a status and privileges superior to those accorded to either the Jurchens or the Khitans, let alone the Chinese. I3° Central China (Hopei and Anhwei) harbored small communities of Tanguts who evidently continued to use their script until the end of the Ming period. Branches of the royal clan broke away and emigrated to western Szechwan, northern Tibet, and perhaps even northeast India, becoming local rulers of secular or Buddhist establishments.'J1 From their T'o-pa roots to their post-Mongolian manifestations, in legend and in life, the creators of the Hsia state left behind a complex historical legacy that is still far from being fully understood or appreciated. Fortunately, archaeology has supplied an abundance of Hsia texts and other artifacts, far more than now exist from Khitan or Jurchen society. As this material becomes available in translations and facsimiles published by scholars in the Soviet Union, China, and Japan, it will one day be possible to write a cultural and social history of Hsia that does not rely almost exclusively on Chinese chronicles. 129 Rachweiltz, "The secret history of the Mongols: Chapter twelve," pp. 49-50. 130 Elizabeth Endicott-West, Mongolian rule in China (Cambridge, Mass., 1989), pp. 66-7: "The sons and brothers of Mongolian, Muslim, Uighur, Naiman, and Tangut ta-lu-hua-ch'ib [darughachi] were governed by separate and more favorable yin regulations than the relatives of Khitan, Jurchen, and northern Chinese ta-lu-hua-ch'ih." 131 Shih Chin-po and Pai Pin, "Ming tai Hsi Hsia wen ching chuan ho shih chuang ch'u t'an," K'ao ku hsiieh pao, 1 (1977), pp. 143-64; Guiseppe Tucci, Tibetan painted scrolls (Rome, 1949), vol. 1, p. 164; Stein, "Mi-flag et Si-hia," p. 237, n. 2; Wu, Hsi Hsia shih kao, pp. 127-37. Cambridge Histories Online © Cambridge University Press, 2008 CHAPTER 3 THE CHIN DYNASTY GENERAL REMARKS It has long been recognized that the dynastic pattern is not a very satisfactory criterion for periodizing Chinese history. A certain exception must, however, be made for dynasties of conquest in which the rule of an imperial house is at the same time a period of foreign domination. One of these is the Chin dynasty ( n 15—1234), which first overthrew the Khitan state of Liao and continued alien rule over large parts of northern China, although under a different ruling group, the Jurchen people. It is too easily forgotten that the region of modern Peking was continuously under non-Chinese rule - with all the resulting social and anthropological influences - for more than four hundred years. First the Khitans invaded the north of China, then the Jurchens came, and finally, from the early thirteenth century onward the Mongols, who, in contrast with Khitans and Jurchens, were successful in subjugating the whole of China. The rule of the Chin dynasty and the Jurchens is therefore a link in the chain of non-Chinese invasions from the northern border regions into China proper; at the same time, however, the Chin state had many distinctive features. Not the least important of these were the methods by which the Jurchens dominated the Chinese. The Khitans, the Jurchens, and the Mongols in turn each practiced a different way of government and accepted Chinese civilization to varying degrees. A comparison of these features may therefore lead to a comparative typology of alien rule over China. There are reasons for regarding the Chin dynasty as little more than just a traditional stage in the foreign domination of northern China. With the Jurchens, a Tungusic people appears for the first time in world history as an identifiable entity and as a great political power. After the downfall of the Jurchen state of Chin, Tungusic tribes did not disappear altogether from the scene. Centuries later the Jurchens' direct descendants, the Manchus, repeated the achievements of their forefathers and rose to the status of military dominance. This time, in the seventeenth century, they succeeded where the Jurchens of the twelfth century had failed: They subjugated the 215 Cambridge Histories Online © Cambridge University Press, 2008 210 THE CHIN DYNASTY whole of China. We should add that the dynastic name of Chin (Chinese for "gold") was the first name of a Chinese dynasty to become known in medieval Europe. Marco Polo mentions a "gold king" (rot d'or), which is a faithful rendering of the name that the Mongols had given to the ruler of the Chin state; the Mongolian words Altan khan mean "Golden khan." The account by Marco Polo is, however, fantastic, as he also tells of an encounter between the "gold king" and the legendary Prester John. It is therefore one of the many ironies of historical transmission that the first Chinese dynastic name to be encountered in a European medieval source was that of a non-Chinese dynasty. THE JURCHEN PEOPLE AND THEIR PREDYNASTIC HISTORY The origin of the Jurchen people is singularly complex. The difficulties begin with their ethnic name, which in various sources is rendered in. several different Chinese orthographies. This chapter uses the form Jurchen, which is warranted by the Chinese transcription Chu-li-chen and seems to be the original form. Today, however, one frequently encounters the form Jiirched or Jiirchid in Western scholarly literature. This is in reality the Mongolian form of the Jurchen name (-d is a Mongolian plural suffix). The form Niichen appears in the early tenth century and is obviously related to the form Lii-chen, said to be the Khitan pronunciation of the ethnic name. Under the Liao the syllabic element chen was prohibited because it occurred in the personal name of the Liao emperor Hsing-tsung, and the Jurchen name was therefore officially changed to Nii-chih. These inconsistencies have led to much confusion, particularly in older Western literature. In the sixteenth century the Jurchen called themselves Jusen, a form clearly derived from the older form, Jurchen.1 The linguistic affinity of the Jurchen language is, however, clear. The Jurchens spoke a Tungusic language and therefore belonged to a linguistic group that, some scholars believe, forms together with the Turkic and Mongolian languages, a genetically related family of languages. Today, Tungusic languages are still spoken in parts of northeastern Siberia and eastern Manchuria. There also is an isolated pocket of Manchu speakers in the Autonomous Region of the Sibe, west of Kuldja in Sinjiang. That the Jurchen language, which had already become extinct by the end of the sixteenth century, was a close relative of the Manchu language was early discovered in i The various forms of the name and their phonetic problems have been exhaustively studied by Paul Pelliot, Notes on Marco Polo (Paris, 1959), vol. 1, pp. 3 7 6 - 9 0 . Cambridge Histories Online © Cambridge University Press, 2008 THE JURCHEN AND THEIR PRE-DYNASTIC HISTORY 217 the West,2 and the Manchus themselves were certainly conscious of their Jurchen affinities. The Jurchen idiom is also the first Tungusic language ever recorded in writing. Our chief source for linguistic studies of the Jurchen language of the twelfth and thirteenth centuries are the words preserved in Chinese orthography in the Chin shih. Chinese—Jurchen glossaries compiled during the sixteenth century reflect a later stage of phonetic development, although they are useful also for a study of the "Old Jurchen" idiom spoken under the Chin dynasty. The original homeland of the Jurchens was the densely forested mountain areas of eastern Manchuria and what is now the Soviet far eastern province south of the Amur River. Relatively early - that is, in the tenth century the Jurchens seem to have spread into the Manchurian plains where the Sungari River region was a center of their settlements. We should note here that the Jurchen way of life, as opposed to Chinese sedentary and agricultural civilization, cannot be described by the simple formula of pastoral nomadism. In this respect the Jurchens certainly differed from the Khitans and even more from the Mongols, both of whom corresponded much more closely to the classical description of pastoral steppe nomadism (no permanent settlements, living in tents, horses and camels as the chief domestic animals, little or no agriculture). It seems that the Jurchen peoples had adapted their way of life and economy to their respective habitat: Hunting and fishing were predominant among forest people, whereas among those living in the plains, cattle raising and agriculture prevailed. The Jurchen raised and even exported horses, but their chief domestic animals were oxen. Important products of their country were hunting birds such as falcons and hawks, eagerly sought by the Khitans and even the Chinese. A study of the tribute and gift lists in Liao and Chinese sources shows which of the Jurchens' products were exported: horses (for which China proper was an inexhaustible market), falcons, gold and pearls, and forestry products (beeswax, pine seeds, ginseng, the last being an important item in the Chinese pharmacopaeia). Obviously the Jurchens' contributions to China's foreign trade were small, and in the system of tributary relations they were for a long time a minor element. But there can be no doubt that it was just these occasional contacts with more developed states that generated in the Jurchens a desire to emulate the power and splendor of such courts as were visited by their envoys from the backwoods of Manchuria. All Chinese sources agree that the Jurchens were a part of the Mo-ho (Korean Malgal) tribes who used to live along what is now the border between Korea and Manchuria. The Mo-ho themselves were not a uniform 2 This discovery was made by Claude de Visdelou (1651 —1737). See his Hiitoin abrigk de la Tartarie, in vol. 4 of Bibliolblque oriaitale (.Maest'icht, 1780), p. 288. Cambridge Histories Online © Cambridge University Press, 2008 2l8 THE CHIN DYNASTY ethnic entity but, rather, a conglomeration of various elements. In T'ang times they had become subjects of the Po-hai kingdom in southern Manchuria. Some of the Mo-ho tribes formed a group that later, in the tenth century, was known as the "Five Nations" (wu-kuo) in the region northeast of modern Chi-lin and that became an element of the Jurchen people, or, rather, federation. Another part of the Mo-ho were the seven tribes of the "Black Water Mo-ho" who settled on the middle and lower Amur River ("Black Water" (Hei-shui) was one of the names of the Amur). It was from these Black Water Mo-ho that the first ancestors of the Jurchens originated, if we are to believe the ancestral traditions current among the Jurchens in the early twelfth century. Before describing in more detail the gradual growth of the Jurchen tribes into a federated nation, a word must be said about their earlier predynastic history as reflected in Chinese sources. It is not clear at what date the Jurchen name first appears in Chinese records. The name of the Mo-ho occurs already during the late fifth century, but the earliest record of a name that could perhaps be related to the Jurchens is A.D. 748, when an embassy of the great and the Little Ju-che people presented gold and silver to the T'ang court. These Ju-che were said to be one of the nine tribes of the Shih-wei federation, which in T'ang times had extended their domination over parts of northern and western Manchuria and therefore might well have included some Tungusic-speaking elements.' Otherwise, the Jurchen name disappears until the early tenth century. It seems that the disintegration of the T'ang empire in the late ninth century facilitated the formation of new political units in more than one border region of China proper, and certainly the Jurchens were no exception. They came as tribute bearers to the Liao court but also sent their embassies first to the court of Later T'ang (an embassy is recorded in A.D. 925) and then to the Sung court from 961 onward, mostly by sea via the Liao-tung peninsula. The general political situation in the borderlands of China can explain why the Jurchens appeared just at the time they did. During the ninth century the Po-hai kingdom, which stretched from the Gulf of Liao-tung to the northeastern parts of Manchuria, had monopolized the lucrative trade and tribute relations with China. In 926 the Po-hai state was absorbed into the Khitan state of Liao, and because of the loose structure of the Liao empire, the more distant peoples and tribes had a chance, with the disappearance of Po-hai authority, to establish their own relations with other countries. After 926 those Jurchens who lived in the Manchurian plains, in particular the Liao 3 Paul Racchnevsky, "Les Che-wei etaicnt-ils des Mongols?" in vol. i of Milanga de linologie offeru a Monsieur Paul Demitville, Bibliotheque de l'Institut des Hautes Etudes Chinoises, vol. 20 (Paris, 1966), pp. 235, 246-51. Cambridge Histories Online © Cambridge University Press, 2008 THE JURCHEN AND THEIR PRE-DYNASTIC HISTORY 219 River region, became the direct subjects of the Liao. They are known as the "Civilized Jurchen," as opposed to their relatives outside Liao domination, the "Raw" or "Uncivilized Jurchen," who still followed their old way of life as forest settlers, hunters, and fishermen. It is not at all clear from which of these two different Jurchen groups came the succession of envoys and traders whose arrival in K'ai-feng is recorded during the period from 961 to 1019. The Jurchen envoys who brought "fine steeds" by sea via Teng-chou in Shantung may have come from the Civilized Jurchen who raised horses in the flat plains of Manchuria and who had succeeded in evading the control of their Khitan overlords. Others may have been emissaries from the Uncivilized part of the Jurchen nation. We know for certain that the Khitans tried to prevent the trade and tributary relations of their Jurchen vassals with the Sung. In 991 they cut off the land route by building palisades near a place through which travelers from Manchuria had to pass. But Sung—Jurchen relations continued by the sea route until the beginning of the eleventh century. In IOIO the Khitans waged a campaign against Koryo in which the Jurchens took the Korean side. The Khitans suffered a heavy defeat and retreated. But for the Jurchens, this victory of their Korean ally had the result that their emissaries for quite some time thereafter reached Sung China only as members of Korean embassies. The data of the Chin sources, although they give a highly stylized picture of the early history of the Jurchens, can be reconciled to some extent with this outside information just summarized. The tribal chiefs who had been ancestors of the ruling clan of the Chin, the Wan-yen clan, were accorded imperial designations posthumously in 1136-7, and their advisers are represented as "chancellors." This is a feature recurrent in Chinese history, in which there had always been a tendency to make imperial ancestors, however humble or barbarian their origins might have been, appear as noble and sinified dignitaries. As the Jurchens had no written records, their earlier history had to be transmitted orally. We have no dated evidence for before the time of the chieftain Wu-ku-nai (1021—74), under whom the consolidation of the various Jurchen clans and their transformation into a nationlike federation began. Tradition says that Wu-ku-nai was a sixth-generation descendant of the founder of the Wan-yen clan. This clan ancestor, by the name of Han-p'u, must therefore have lived sometime around the year 900, which is about when the Jurchens appear on the diplomatic scene. The social life of the Jurchens during the tenth and eleventh centuries was still largely determined by tribal independence and individualism. Although the father of Wu-ku-nai had already received a military title from the Liao court, such appointments had little if any meaning and conferred no power or actual functions. They signified little more than the recognition of a loose sovereignty of the Liao Cambridge Histories Online © Cambridge University Press, 2008 22O THE CHIN DYNASTY and the acceptance of a formal vassal status. Even the Chin sources tell us that in the middle of the eleventh century the Jurchens still had no script, no calendar, and no offices. Another feature that points to a great measure of clan independence is the fact that not a few place names in eastern Manchuria in those times had names like "the town of such and such a person." There was much resistance among the tribal chiefs against the rise to dominance of the Wan-yen clan, and it needed a person of the stature of Wuku-nai, the grandfather of the founder of the Chin dynasty, to overcome the opposition of other clans. Wu-ku-nai is described as a brave warrior, a great eater and drinker, and a lover of women. He finally succeeded in extending his domination over the whole territory of eastern Manchuria from the Ch'ang-pai shan, the "Ever-White Mountain," a center of religious worship on the Korean-Manchurian border, to the "Five Nations" in the north. The Liao appointed Wu-ku-nai as military governor (chieh-tu-shih) of the Uncivilized Jurchen, and the Liao emperor even once received him in audience. One of the actions that contributed to his growing power was the planned increase in armaments through the purchase of iron and iron weapons from among the other clans (the Jurchens had a reputation for being skillful blacksmiths). But the chief reason that Wu-ku-nai was able to forge the recalcitrant tribes and clans into a more organized power was the desire common to all Jurchens to remain as independent as possible of their Khitan overlords. One reason for the Jurchens' discontent was the bad treatment they had suffered from the Khitans. The Jurchens were supposed to deliver annually to the Liao court products of their country, such as pearls, hunting falcons, and sable furs. It seems that much extortion and violence took place during these annual meetings of Liao envoys and Jurchen tribute bearers at the trading posts of the frontier regions. The fact that Wu-ku-nai himself and all his successors held titles of military governors conferred by the Liao court did little to make the Jurchens faithful vassals. On the other hand, the contacts with the Liao state and the resulting acquaintance with the ways of a more organized and structured type of government led to a growing awareness that the traditional tribal organization would be insufficient if the Jurchens wanted to match the Khitans. It was a grandson of Wu-ku-nai who achieved both formal and factual independence for the Jurchens: A-ku-ta, later canonized as T'ai-tsu. THE REIGN OF A-KU-TA AND THE FOUNDING OF THE CHIN DYNASTY Under A-ku-ta's predecessor, his elder brother Wu-ya-shu (r. 1103-13), the Jurchens had already achieved sufficient power to be able to stabilize their Cambridge Histories Online © Cambridge University Press, 2008 A-KU-TA AND THE FOUNDING OF THE CHIN DYNASTY 221 border with Korea and at the same time to rally more and more tribal groups and clans to the support of the house of Wan-yen. The Wan-yen clan had their original home in the valley of the An-ch'u-hu River (modern A-la-ch'u-k'o River, east of Harbin, a southern tributary of the Sungari). This region was for many years the center of the Jurchen state and became the site of the Chin Supreme Capital (Shang-ching, situated near the modern town of A-ch'eng southeast of Harbin). After the death of Wu-ya-shu, A-ku-ta was proclaimed ruler of the Jurchens by the tribal elders in 1113 and was also appointed as military governor by the Liao, following the usage of the past. The tribal election of rulers had been practiced for a long time among the Jurchens, and there existed no strict rules for inheritance of the chieftainship, but the choice was limited to the members of the Wan-yen clan. Soon afterward, a full-scale war broke out between A-ku-ta's forces and the Liao. At first the trained soldiers under A-ku-ta's command did not exceed a few thousand, but with each of his victories more leaders joined their forces with the Wan-yen armies. The war was occasioned — but this was not much more than a pretext on the Jurchen side — by A-ku-ta's demands for the extradition of A-shu, a Jurchen leader who had deserted to the Liao many years before. The Liao refused to surrender A-shu and ignored all subsequent demands from the Jurchen side. Within a surprisingly short time A-ku-ta succeeded in eliminating the Liao armies sent against him and in establishing himself as the undisputed master of most of Manchuria. The title of emperor (huang-ti) and the dynastic name Chin, "Gold," were assumed in the spring of 1115. The name Chin was taken from the name of the An-chu'u-hu River; an-ch'u-hu meant "golden" in the Jurchen language.4 In adopting a river name as their dynastic designation, the Jurchens were following the example of the Liao, who had taken their name from the Liao River in southern Manchuria. At the same time a Chinese-style reign title (Shou-kuo, "receiving statehood") was introduced, and A-ku-ta himself took a Chinese personal name in addition to his native name. Ironically, his Chinese name was Min, "compassionate." Thus A-ku-ta had assumed all the regalia inherent in the position of a Chinese emperor. But we are informed in the existing sources that the mastermind behind this decision was a man of Po-hai origin, Yang P'o, who as a young man had obtained the chin-shih degree under the Liao. In a memorandum he pointed out that A-ku-ta should not be content with only the de facto independence of the Jurchen tribes under his leadership but should aspire to 4 The name of the river can perhaps be reconstructed as "ancuqu, a word that has survived in the Manchu language as ancun, meaning "earring" and, in some combinations, "gold." This word is not related to the Manchu word aisin (gold), which formed part of Aisin Gioro, the family name of the Manchu emperors. Cambridge Histories Online © Cambridge University Press, 2008 222 THE CHIN DYNASTY the higher dignity of an emperor. He also wished to secureformalrecognition of A-ku-ta's imperial position and outlined the necessary steps in another memorandum, which is of considerable interest because it enumerates the qualities thought to be indispensable for legitimate emperorship. Yang P'o invokes Chinese traditions of establishing a new dynasty by pointing out that the founders of a state obtained their rank either by a voluntary cession of power from a preceding dynasty or by the receipt of a document of investiture. Therefore he drafted a list of ten demands to be submitted to the Liao emperor. These were, first, recognition of A-ku-ta's title Great Holy and Great Enlightened and of the dynastic name of Great Chin. The use of a jade-ornamented state carriage, of imperial robes and caps, and of a jade seal should be conceded. Both states, Liao and Chin, should communicate like elder and younger brothers, that is, adopt the model of pseudofamilial relationships that had gradually evolved in the multistate system since the eleventh century in continental East Asia. Regular embassies should be dispatched for the New Year festival and the rulers' birthdays. This, too, had become diplomatic routine since the treaty concluded between Sung and Liao in 1004. If granted by the Liao, all these demands would have resulted in an uneasy coexistence between Liao and Chin and would have permitted the continued existence of the Khitan state. But other demands put forward by A-ku-ta on the advice of Yang P'o could hardly have been granted by Liao because they would have affected the material basis of their state: These were the unconditional cession of the two circuits of Liao-tung and Ch'un-chou, and the promise to pay annually to Chin the amount of 250,000 ounces of silver and 250,000 bolts of textiles. These sums amounted to practically the whole tribute paid annually to Liao by the Sung. It seems therefore that A-ku-ta and his advisers, from a very early time on, intended to eliminate the Liao state completely.' At the Liao court the danger posed by the Chin was soon recognized, and the years following the rise of Chin to imperial status were full of halfhearted attempts to placate the Chin and of abortive military actions against the Jurchens. In spite of the war between the two states, diplomatic communication was not completely severed. However, the more military successes the Jurchens achieved, the more their demands were increased. After a defeat of the Liao armies, or rather their wholesale desertion in the beginning of 1117, A-ku-ta again tried to impose new demands on the tottering Liao. In 1118 he outlined his conditions for a peace. This time A-ku-ta demanded to be recognized by the Liao emperor as elder brother, which meant 5 For a detailed account of early Chin—Liao relations, see Herbert Franke, "Chinese texts on the Jurchen: A translation of the Jurchen monograph in the San ch'ao pei metig hui pirn," Zentralasialiscbe Sluditn, 9 (i975). PP- 15'-66- Cambridge Histories Online © Cambridge University Press, 2008 A-KU-TA AND THE FOUNDING OF THE CHIN DYNASTY 223 a reversal of the status proposed in 1115. Liao should cede three more circuits in Manchuria and send a prince, a princess, and an imperial son-in-law as hostages to the Chin court. Above all, the diplomatic documents referring to the relations between Liao and Sung, Hsi-hsia, and Koryo should be submitted to the Chin court. This would have allowed the Chin to become the legitimate receivers of the tribute paid hitherto by Sung to Liao and to assume the status of a major power vis-a-vis the flanking states in the east and west, Koryo and Hsi-hsia. 6 This dilatory attitude on the part of T'ien-tso of Liao was not well received by A-ku-ta and his court. Repeatedly A-ku-ta refused to accept diplomatic letters from Liao because they did not sufficiently take into account his newly acquired imperial status. He could afford to do this because he had in the meantime conquered the rich and fertile lowlands of southern Manchuria, including Liao-yang, the "Eastern Capital" of Liao, which was a center of Pohai civilization. Many of the Liao military leaders, not only Khitans, but also members of other tribes such as the Hsi (originally a Turkic-speaking people living in the southwestern highlands of the Liao empire), defected to the Chin and were frequently incorporated into the Chin armies together with their subordinate troops. Sometimes these defectors proved unreliable and rebelled against their new overlords, but A-ku-ta could crush these shortlived revolts very quickly. At this point it should be remembered that from the beginning, the Jurchen armies had included many non-Jurchen elements, chiefly drawn from tribes and ethnic groups who had joined them earlier. A-ku-ta was able to resolve the problem of military leadership over a multinational army, whereas the last ruler of Liao proved unable to handle his own ethnically mixed forces. All the attempts at compromise on the part of Liao failed in the end. When the Liao court promised to invest A-ku-ta as "King of the State of the Eastern Sea," he resisted fiercely, because the title of king (wang) was plainly inferior to that of emperor of the great Chin, which he had already assumed years before. The longer the abortive negotiations were protracted, the stronger A-ku-ta's position had become in the meantime, and compromise no longer seemed necessary. The total overthrow of the Liao empire, although it might perhaps not have been A-ku-ta's major political objective in the early stages of his rise to power, was now definitely in sight, certainly after 1119. All these successes achieved by A-ku-ta, however impressive they were, must be regarded as still within the reach of any determined and able military leader and did not require superior diplomatic skill, except perhaps 6 T'o-t'o et al., eds., Liao shih (Peking, 1974), 28, pp. 336—7 (hereafter cited as LS). Cambridge Histories Online © Cambridge University Press, 2008 224 THE CHIN DYNASTY in the handling of his followers. But A-ku-ta was more than just a ruthless and able general who could score victories against an adversary weakened by inept leadership, severe internal dissent, and disorganization. He emerged as a brilliant diplomat and strategist when after 1117 the bilateral Chin-Liao relations were replaced by a triangular power system that included the Sung. The fourth powerful state with which the Chin had to reckon, the Hsia state of the Tanguts, was at that time not yet directly involved. The Tanguts supported the Liao to some extent before n 2 4 and it was only afterward, when Chin and Hsia become contiguous, that the Chin claimed suzerainty over Hsia, having supplanted the Sung as their direct neighbors. The formalization of Hsia—Chin relations was achieved under A-ku-ta's successor Wuch'i-mai and secured for Chin an additional rise in status.7 The Jurchens had, as we have seen, already established peaceful relations with the Sung court during the early part of the Northern Sung dynasty, and their rise to hegemony in the north had been followed closely by observers in K'ai-feng. Recall that after 1005, Liao—Sung relations had been based on a treaty arrangement whereby the Sung had bought peace on their northern border through annual payments to the Liao and by formally recognizing the Liao conquest of the Sixteen Prefectures in north China, including the region of Yen (Peking). The growing weakness of the Liao having become apparent, the Sung looked for potential allies to help them recover the never-forgotten Chinese territories that had been lost to the Khitans during the period of political disunity in the first half of the tenth century. Now the Chin state, with its deep-seated grudge against the Liao, became in the eyes of Sung politicians a natural ally against the Liao, and hopes of recovering the lost territories with the assistance of Chin began to be entertained in the Sung capital. In n 17 a Sung envoy was sent to the Chin under the pretext of buying horses, but in reality to negotiate a Sung—Chin alliance against Liao. Between 1117 and 1123 Sung envoys traveled altogether seven times to the Chin court, and six Chin embassies went to K'ai-feng, not counting the steady flow of diplomatic correspondence. Three major points of negotiation emerged during these diplomatic contacts: a joint Sung-Chin attack against the tottering Liao empire, the territorial problems (return of the Sixteen Prefectures to Sung), and the problem of the annual payments formerly paid to Liao. The bargaining position of the Sung during those years deteriorated rapidly because A-ku-ta had soon discovered that in spite of some preparations in the Sung empire, the Chin armies would not need to wait for the assistance of Sung troops to conquer for themselves the southern part of the 7 For a chronological account of Chin—Hsia relations, see chaps. 61 and 62 of T'o-t'o et al., eds., Chin shih (Peking, 1975) (hereafter cited as CS), and for additional information, the monograph on Hsi Hsia in chap. 134. Cambridge Histories Online © Cambridge University Press, 2008 A-KU-TA AND THE FOUNDING OF THE CHIN DYNASTY 225 Liao state, with Yen (Peking) as its center. The Sung, on the other hand, thought that they could make their own territorial claims acceptable to the Chin only by offering a military alliance. Soon the Chin also put forward their claim that they should be regarded as the legitimate successors of Liao with regard to the annual Sung "tribute" payments. By the time a formal treaty was finally concluded in 1123 - the first one between the two states — the military situation had radically changed. Sung attempts to capture Yen had failed, whereas the Chin armies had not only forced the Western and the Central capitals of Liao to surrender but, by the end of 1122, had also been able to conquer Yen, its Southern Capital. The Liao emperor had already become a fugitive in the far west; the end of his empire was in sight. With Yen in the possession of the Chin, A-ku-ta's position had become unassailable, and the Sung were forced on the defensive. They had to accept A-ku-ta's conditions concerning the return of the Yen region to the Sung. Not all of the six prefectures of the Yen region were to be handed back to Sung, and in addition to the annual payments formerly granted to the Liao, the Sung had to pay the Chin a huge compensation for the loss of tax income from the Yen prefectures. One might ask why A-ku-ta consented at all to conclude a formal treaty with Sung, as his military position was so strong. A possible answer is that he wanted to formalize the question of annual payments. These amounted to 200,000 taels of silver.and 300,000 bolts of silk, a sum perhaps not ruinous to Sung China's treasuries but important as a source of income for A-ku-ta's own young state of Chin. The treaty was, in addition, a way to have his own imperial status solemnly recognized. The treaty text, as was usual in Chinese diplomacy since early times, was couched in the form of two parallel and identical oath-letters, and no differentiation in status between the parties is implied in the texts of the respective documents. A-ku-ta is addressed as "August Emperor of the great Chin," and the Sung ruler accordingly. This meant that Chin had now achieved a state of equality with Sung, only ten years after the formal foundation of the Chin as a dynasty. These oath-letters were exchanged during the third and the fourth month of 1123. Their texts do not, however, contain any details about how the cession of the six prefectures to Sung should be carried out. Other details, such as delimiting the borders, were also left to be worked out in corollary agreements. The general instability of the situation, together with the points left undecided, became the reason that instead of leading to a stabilization of China's northern border, the treaty of 1123 marked the beginning of a long period — lasting for almost twenty years — of violent and devastating warfare that more than once threatened the very existence of Sung China. By the time A-ku-ta concluded his alliance with Sung, the Liao state had Cambridge Histories Online © Cambridge University Press, 2008 226 THE CHIN DYNASTY suffered fatal defeats. In 1122 the Liao Central Capital (modern Tsaghan Suburghan in Jehol Province, Manchuria) fell to Chin armies. The Liao emperor T'ien-tso fled west. In the Southern Capital (modern Peking) another member of the Khitan imperial clan was proclaimed emperor but died soon afterward. Chin armies reached Peking. The Khitans were unable to put up organized resistance despite local attempts to withstand the onslaught of the Chin forces. In 1125 the last Liao emperor T'ien-tso was captured and degraded to the status of a prince. This marked the formal end of the Liao state that had been such a formidable adversary of Sung China. But in its place the Sung now had as their northern neighbor a state that was at least as dangerous an opponent as Liao had ever been. A-ku-ta did not live to see the final extinction of Liao and the humiliation of Sung; he died a few months after the treaty of 1123 had been concluded. The foundations he laid remained, however, the basis for the future exploits of Chin. FROM WAR TO COEXISTENCE: C H I N - S U N G RELATIONS BEFORE THE TREATY OF I I 4 2 A-ku-ta was succeeded by his younger brother Wu-ch'i-mai (1075—1135), whose posthumous temple name was T'ai-tsung. The new ruler, like his deceased brother, displayed great military and diplomatic astuteness in the rather confused situation that confronted him in 1123. Although an agreement had been reached with the Sung, the task of implementing the terms of the treaty was still unresolved. The Sung state was evidently the Chin's most formidable potential adversary. So Wu-ch'i-mai tried to strengthen the Chin's position on its other borders. The final annihilation of the remnants of Liao had been relatively easy, but there was still the state of Hsi Hsia on the Chin's western border, a state that was certainly not strong enough to be a serious menace but that could always stir up trouble in the western parts of the Chin state. Wu-ch'i-mai adopted a conciliatory policy toward Hsi Hsia and concluded a peace treaty in 1124 by which Hsi Hsia acknowledged the Chin's suzerainty by declaring itself an "outer vassal." At the same time the Chin ordered the repair of its frontier defenses against Kory6, its eastern neighbor. Some local unrest in the Po-hai districts was also quickly brought under control, so that the Chin was in a position to concentrate on its relations with the Sung. These were, at first, outwardly normal. Despite the outbreak of fighting in 1124, the Sung, even as late as 1125, sent an embassy to the Chin court in order to offer formal congratulations on Wu-ch'i-mai's succession to the throne. The situation was full of tension but had not yet degenerated into a full-scale war. Perhaps the chief reason that this occurred was the affair of Cambridge Histories Online © Cambridge University Press, 2008 CHIN-SUNG RELATIONS BEFORE 1 1 4 2 227 Chang Chiieh, a case that illustrates the fluid and chaotic conditions resulting from the collapse of the Liao state and the advance of Chin. Chang Chueh had been a Liao military governor residing in P'ing Prefecture (modern Yung-p'ing east of Peking). He assassinated the former Liao commander of Yen (the Eastern Capital) who had joined the Chin side, under the pretext that this official had caused unnecessary hardships to the population of Yen, a great portion of which had been evacuated to P'ing. At the same time Chang Chiieh declared his allegiance to the Sung and was reappointed by them to his former office. All this had happened before the death of A-ku-ta, but his successor was not slow to react. A Jurchen army attacked P'ing Prefecture. Chang Chiieh fled to Yen where Kuo Yao-shih, a former Liao commander, had likewise joined the Sung side. But the Chin regarded Chang Chiieh as a traitor and rebel and demanded his extradition. The Sung court agreed and gave orders for his execution; his head was sent to the Jurchen. Kuo Yao-shih thereby realized that the Sung was not in a position to protect those who had prematurely declared their allegiance to this state and so submitted to the Chin, who reinstated him as governor of Peking. The decision by the Chin court to wage a full-scale war against the Sung was taken in autumn of the year 1125, and although the Chang Chueh incident was not its only motivating force, it certainly contributed to this decision. Within a few months the Chin armies occupied the greater part of Shansi and Hopei, and early in 1126 they crossed the Yellow River and laid siege to the Sung capital of K'ai-feng (see Map 15). The Sung court found itself in a desperate situation. No auxiliary armies that could have forced the Chin to lift the siege were in sight. On the other hand an attack against the walled capital of Sung was a military venture that could have cost the Chin armies heavy losses in men and materiel. The Chin commander, Wo-li-pu, the second son of A-ku-ta, therefore accepted the Sung offer to negotiate a withdrawal of the Jurchen forces. Wo-li-pu had to conduct his negotiations with the Sung without the possibility of contacting his ruler Wu-ch'i-mai in faraway Manchuria, but we may assume that he had full plenipotentiary powers in both military and diplomatic affairs. The terms demanded by the Chin side for their withdrawal were exorbitant. The three prefectures of T'ai-yiian, Chung-shan, and Ho-chien were to be ceded to the Chin (this implied the loss of the greater part of the modern provinces of Shansi and Hopei). Moreover, an enormous war indemnity was to be paid in addition to the normal annual payments, and a Sung imperial prince was to be sent as a hostage to the Jurchen camp. Two versions of the oath-letter in which the Sung acceded to the demands of Chin have been preserved, and both reflect the hopelessly inferior position of the Sung. The Cambridge Histories Online © Cambridge University Press, 2008 lung-sl 112£xii. Chin army withdraws 1127with two cap tiveSungEmperore > Campaigns 1125,x.—1126.V. -=<•<> Return route • Campaigns 1126.x.—1127, iv. - " " • Return route Western campaigns 1127—29 Ju-chou 1127.XM. 100 MAP 15. The Chin invasions of Northern Sung Cambridge Histories Online © Cambridge University Press, 2008 CHIN-SUNG R E L A T I O N S BEFORE I I 4 2 229 annual payments were increased to 300,000 taels of silver, 300,000 bolts of silk, and 1 million strings of coins. The war indemnity imposed on the Sung was so high that it amounted to not less than 180 years of annual payments. As hostage the prince of K'ang, a younger brother of the Sung emperor Ch'in-tsung, went to Wo-li-pu's camp. He was accompanied by Chang Pang-ch'ang, a high-ranking official who had been in favor of appeasing the enemy. The negotiations had lasted for about a week, and after Wo-li-pu had consented to the revised version of the Sung oath-letter, he gave orders on 10 February 1126 to lift the siege of the Sung capital. The continued existence of the Sung state as a power on equal footing with Chin had been bought at great expense, but, as it turned out, only for a brief period. One of the immediate results of the military and political disaster had been the elimination of the chancellor Ts'ai Ching and his followers, who had dominated Sung politics for many years. Another had been the abdication of the Sung emperor Hui-tsung (r. 1100—26) in favor of Ch'in-tsung (r. 1126—7). A side effect of the new situation was that the king of Kory6 declared himself a vassal of Chin in the summer of 1126. The two major states bordering on Chin — Koryd and Hsi Hsia — had thus now recognized the overlordship of the Chin emperor. The war between Sung and Chin soon flared up again. It must remain a matter for speculation if the reason was that Chin had realized how weak the Sung was militarily arid therefore resolved to annihilate the Sung empire once and for all, or whether the reasons that Chin sources give us for the resumption of fighting were real reasons and not just facile pretexts. As it was, the Chin soon complained that Sung was breaking the armistice and was even trying to induce some former Liao generals to attack the Chin. Once again the Chin armies marched south and crossed the Yellow River in the eleventh month of 1126 to lay siege to K'ai-feng. This time the resistance of the city had to be overcome by force. On 9 January 1127 the victorious Jurchen army entered the city after heavy fighting and pillaged relentlessly. Emperor Ch'in-tsung and ex-Emperor Hui-tsung, together with a host of courtiers and family members, were made prisoners; the prince of K'ang, enthroned as the new emperor (Kao-tsung) tried to organize resistance in regions not yet threatened by the enemy. The de facto surrender of Sung was total and became formalized through the degradation of Ch'in-tsung and Hui-tsung to the status of commoners. In May 1127 the two former emperors were marched north with their entourage, thus leaving a sort of interregnum in the central parts of the Sung empire. The Chin at first did not try to incorporate the territories under their control into their state but instead set up a new Chinese dynasty under Chang Pang-ch'ang. The name given to this state was Ta-Ch'u, "Great Cambridge Histories Online © Cambridge University Press, 2008 230 THE CHIN DYNASTY Ch'u," reminiscent of the ancient state of Ch'u in the Yangtze valley. The capital was Chien-k'ang (Nanking). Chang Pang-ch'ang's role of a quisling appointed by the Chin did not, however, last long. He was killed, or rather forced to commit suicide, on the order of the Sung court. This showed that Chin domination was still by no means absolute. Also, in the northern parts of China, numerous towns under loyal commanders still held out, and in many places the resistance against the Chin was organized by volunteers. This taxed the military resources of the Chin so much that for the moment a continued attack into the regions south of the Yangtze River was impossible. Another factor that slowed the Chin aggression was the death of Wo-li-pu and of Wo-lu, another military leader, in 1127 and the death of Che-mu in 1129. They all were close relatives of the emperor. Wo-li-pu was a son, Wo-lu a cousin, and Che-mu a brother of A-ku-ta. Contrary to Chinese and, in particular, Sung custom, members of the Chin imperial clan were prominent in both military and political affairs. Clan affiliation was thus an extremely important factor in the Chin empire, as contrasted with the more nonfamilial and, at times, partisan composition of the ruling circles in the Sung state. In the next few years following the Sung defeat in 1127, Chin troops tried repeatedly to gain strategic footholds south of the Yangtze River (see Map 16). This forced the Sung court to abandon Chien-k'ang (Nanking), which had become the temporary capital in 1129, and to withdraw to Shao-hsing in Chekiang Province. But even Chekiang was raided by advance forces of the Chin. The chaotic conditions prevailing in China north of the Yangtze, however, prevented the Chin from permanently incorporating these areas into their empire. The whole of the Central Plains was a battleground where Chin troops, Sung generals, and independent militia leaders fought with no conclusive results. The Chin tried again to resolve their problems by creating a buffer state, not only for protection against the Sung, but also in the hope of attracting more military and civilian defectors from Sung to a state under nominal Chinese rule. Another reason for this attempt to create a new state in north China was that the Jurchens lacked the trained personnel with which to exploit these regions for their own benefit. Their choice fell on Liu Yii (1079-1143?), a native of Hopei who had passed the chin-shih examination in about 1100 and had served the Sung in various provincial and metropolitan offices. When he was prefect of Chi-nan (Shantung), he changed sides in 1128 and joined the Jurchens. He was favored by T'ai-tsung and some leading Jurchen generals and finally, toward the end of 1129, was appointed as emperor of great Ch'i. The capital of this state was first at Ta-ming in Hopei, but Liu took up residence in Tung-p'ing (Shantung), his Eastern Capital, and later moved into the former Sung Cambridge Histories Online © Cambridge University Press, 2008 100 2O0 100 30Okm 200miles Chinarmydefeatedby HanShih-chung H30,iv. iTing-hai 'ang-kuo MAP I 6. Chin raids south of the Yangtze, 1129—30 Cambridge Histories Online © Cambridge University Press, 2008 232 THE CHIN DYNASTY capital at K'ai-feng in 1132. He tried hard to organize north China by establishing a workable administration and reviving the stagnant economic life of his territory. Compulsory military service and heavy taxation were imposed on the population. His troops fought alongside the Jurchens against the Sung and even scored a few victories, such as the capture of the strategic town of Hsiang-yang in 1135. But the Sung counteroffensive under Yiieh Fei in 1134—5 ended in the reconquest of much of the lost territory. This turn of events did much to discredit Liu Yii's military value to the Jurchens. In 1135 his protector, the Chin emperor T'ai-tsung, died and his successor, a grandson of A-ku-ta, later canonized as His-tsung (1119—49), proved much less favorably inclined toward Liu Yii. Late in 1137 the state of Ch'i was abolished and Liu Yii was demoted from the rank of emperor (huang-ti) to that of prince (wang). It seems that he was even suspected of having opened secret negotiations with Yiieh Fei. Liu Yii was transferred first to Hopei and then to the town of Linhuang in northwestern Manchuria where he was allowed to end his life in supervised retirement. The experiment of organizing the Jurchen conquests as a Chinese puppet state with the help of Chinese defectors had thus failed, and the Chin were faced with the alternative of either trying to achieve coexistence with the Sung or continuing their policy of aggression and annihilation of the Sung. It is hard to say when the Chin finally realized that the Sung empire could not be conquered. There had been abortive peace talks as early as 1132. One factor that perhaps influenced the Jurchen decision to come to terms with the Sung was the death of the former Sung emperor Hui-tsung in n 35, in Wukuo-ch'eng, a town in the Sungari region of Manchuria where he had been held as a captive together with the other members of his former court. The Chin government had appreciated that the imperial prisoners in their custody were a diplomatic asset of the first order, and therefore their treatment had been relatively lenient. The gradual improvement of their fate can be shown from Chin sources (Sung sources remain silent on this point). In the beginning, in 1127, Hui-tsung and Ch'in-tsung had been degraded to the status of commoners and in 1128 were forced to pay obeisance to the ancestral spirit of A-ku-ta in the latter's mausoleum, wearing mourning dress — a ritual of expiation imposed on supposed war criminals. Then the former emperors were formally enfeoffed as marquises (bou) with the insulting titles of Hun-te (Muddled Virtue) and Ch'ung-hun (Doubly Muddled). Six Sung princesses were given as wives to members of the imperial Wan-yen clan of Chin. In 1137, the Sung court was formally notified of Hui-tsung's death, and when a peace treaty was in sight (1141), Hui-tsung received the posthumous rank of prince of T'ien-shui chiin; his surviving son Ch'in-tsung, that Cambridge Histories Online © Cambridge University Press, 2008 C H I N - S U N G R E L A T I O N S BEFORE I I 4 2 233 of duke (kung) of T'ien-shui chiin — even the Chin respected the status difference of generations. Note also that the name of their new fief was a perfectly neutral one and not derogatory, as their former titles had been. T'ien-shui was a commandery at the upper course of the Wei River in the eastern part of what is now Kansu Province. A few months later Ch'in-tsung was given the emoluments due his ducal rank. After the treaty had been concluded, the male relatives of the two emperors who were still in Chin captivity were accorded emoluments, a privilege that was extended in 1150 to the female descendants of the former Sung emperors. In other words, the Chin treated their captives as hostages who could always be used to bring pressure on the Sung. The death of Ch'intsung in 1156, however, deprived the Chin of their principal hostage on whom they could rely to prevent the Sung from breaking the peace treaty. This treaty of 1142, which was to regulate Sung-Chin relations for almost twenty years, was the result of protracted negotiations. The Chin had the advantage of being able to use the return to the Sung of the coffins of Huitsung, his empress, and the emperor's mother as a bargaining chip. In addition they kept up the military pressure by repeatedly sending troops into the territory south of the Yellow River and in 1140 again conquered the whole of Honan and Shensi, which had been returned to Sung control early in 1139 after the conclusion of a preliminary peace. But the conclusion of a peace treaty would not have been possible if the revanchists had remained in power at Hangchou, where the Sung had finally established their capital in 1138. The elimination of Yiieh Fei, the most successful and popular of the Sung generals, by his adversary Ch'in Kuei opened the way toward a final agreement. In 1141 Yiieh Fei was ignominiously put to death in his prison, a foul deed that made the advocate of coexistence, Ch'in Kuei, forever a bete noire in Chinese history. Almost at the same time, negotiations between Sung and Chin were initiated. These were extremely involved and protracted. It seems that the Chin side, through the commander Wan-yen Tsung-pi, signaled to the Sung that a peace could be obtained by agreeing to make the Huai River the border between the two states. This was in October 1141. Wan-yen Tsung-pi was the fourth son of A-ku-ta and had been entrusted with operations in central China. Two months later the Sung agreed in principle. Extracts from state letters of both sides have survived in Sung sources; their dates range from October 1141 to October 1142. But the texts of the treaty itself, or, more correctly, the oath-letters of Chin and Sung, have not survived. What we have is an abbreviated version of the Sung oath submitted already by the end of 1141. The terms of peace were harsh. The Sung agreed to have the middle course of the Huai River as a border, which meant that the whole of the Cambridge Histories Online © Cambridge University Press, 2008 234 THE CHIN DYNASTY Central Plains was lost to the invaders. Also, the two strategically located prefectures of Tang and Teng (in modern Hupei), which were to play a significant role in the war of 1206, were ceded to Chin. From 1142 onward, annual tributes of silver and silk amounting to 250,000 ounces and bolts, respectively, were to be paid each year in the last month of spring; the payment was to be delivered to the Chin border town of Ssu-chou on the northern bank of the Huai River. Additional terms concerned security along the border. Fugitives from Sung to the north could not be pursued, and no large garrisons could be stationed in the Sung border prefectures. At the same time Sung promised not to hide fugitives from the north but to surrender them. The wording of the Sung declaration is characterized by extreme humility, which acknowledged the Sung's new vassal status. Chin is addressed as "Your superior state," whereas Sung termed itself "our insignificant state." Equally humiliating was the use of the word "tribute" (kung) for the annual payments. But the worst loss of face was that Sung was no longer regarded by Chin as a state in its own right, but rather, as a vassal, and it is understandable that no Sung source has preserved the text of the Chin decree that invested Kao-tsung, whose personal name was Chao Kou, as ruler of the Sung state. It has been preserved, however, in the biography of Wan-yen Tsung-pi, along with the covering letter accompanying Kao-tsung's oathletter.8 The Sung emperor must have regarded the extorted documents as the nadir of his career. To call himself "Your servant Kou" must have taxed to the extreme his gift for self-denial. The Chin diplomat who presented the investiture document to Kao-tsung was a Chinese who had formerly been in the service of the Liao and found employment at the Chin court. He was received by Kao-tsung in a formal audience on 11 October 1142, a date that must be regarded as the actual end of hostilities and the beginning of a new period of coexistence. The Chin withdrew their armies and promised to return to the Sung the coffins of Huitsung and the empresses. It is surprising that the surviving documents of the negotiations and the short version of Kao-tsung's oath-letter do not mention the resumption of trade between the two countries, but this is certainly due to the deficiency of the sources. In fact, licensed border markets were established, the most important one being Ssu-chou. Trade soon began to flourish again. For the Chin the stabilization of their southern border and the final 8 CS, 77, pp. 1755—6. For a general study of treaties between Sung and Chin, see also Herbert Franke, "Treaties between Sung and Chin," in Etudes Song in memoriam Elienne Balazs, ed. Franjoise Aubin, 1st series, no. i (Paris, 1970), pp. 5 5 - 8 4 . Cambridge Histories Online © Cambridge University Press, 2008 T H E P O L I T I C A L H I S T O R Y OF C H I N AFTER II42 235 conquest of the Central Plains brought about a gradual shift of their political and economic centers from north to south (see Map 17). More and more Jurchens settled in north China; the Chin had at last become a state that was, ethically and economically, to a great extent Chinese. For the Sung, too, the consolidation brought about by the treaty proved a considerable asset. Despite the formal vassal status that he had had to accept, Kao-tsung had stabilized the situation and had, moreover, fulfilled the moral duties of filial piety by recovering the body of Hui-tsung and securing the release of his mother. The Chin refused, however, to free Ch'in-tsung, a refusal that was probably not entirely unwelcome to Kao-tsung, whose position as emperor would have become precarious had his older brother returned. T H E POLITICAL HISTORY OF CHIN AFTER II42 A period of peaceful coexistence thus seemed to lie ahead after 1142. It was interrupted twice during the following seventy years, once by the Chin and once by the Sung, thus demonstrating that revanchism had not died out with the 1142 treaty and remained a constant matter of controversy in court circles. The years immediately following the treaty were, however, peaceful for both states. The Chin state had asserted itself as a power in China proper and continued to transform itself into a Chinese-style political entity. This transformation from a more tribal, feudal society into a bureaucratic organization did not take place without some resistance from the more conservative faction among the Jurchen grandees. The ruler Hsi-tsung (r. 1135-50) himself had been enthroned when he was a boy and had never taken a prominent part in the diplomatic and military activities that occurred after his accession. He left all this to the imperial clan members who held the highest military and civilian offices. The strong personal leadership shown by T'ai-tsu and T'ai-tsung was absent in their successor, who was, moreover, not a very capable person and even more addicted to drinking than was usual among the hard-drinking Jurchens. As long as the Chin state was not confronted with any critical situation, a ruler like Hsi-tsung might be tolerated by the more farsighted members of the imperial clan, and indeed there was not much to interfere with his pursuit of personal pleasure. It is true that some border warfare broke out with the unruly inhabitants of the northwestern steppe regions, but following the Sung example the Chin now adopted a policy of appeasement. In this context the Mongols played a prominent role. It seems that already in the mid-twelfth century their tribe had been consolidated enough to be regarded by the Chin government as the potential partner of an agreement. Cambridge Histories Online © Cambridge University Press, 2008 Cambridge Histories Online © Cambridge University Press, 2008 236 T H E POLITICAL HISTORY O F C H I N AFTER THE CHIN DYNASTY 300 km 260mil ... ': Heng-chou /Hsi-pei-lu Chao-t'ao-ss CAPITAL ."; "11 f K * : I UNli -r UKIMU LU / TAingfti.•:.?•...?ningfi / •Liao-yahgfu J -.. QPien'il./ . \ M >.,... EASTERN PO/-SU- Hsim-chou^-..jfhou = . : W . /Ting-hai^' .: chou / xv*>: liarigL i N - T A o a u * L.•••;"••&*•' . . ••—^JCh'in-chou, T, ;. ^ ^ T ^ / -Lung Tao-chou Lchou FENG-HSIANG LU :; T'ang-. . .;' . # '' •: Ts'ai-chou 1 Cheng-chou 2 Chiin. chou MAP 17. The Chin empire Cambridge Histories Online © Cambridge University Press, 2008 II42 237 238 THE CHIN DYNASTY Chinese sources of Sung provenance record that in 1146-7 the state of the Mongols was "pacified" and that its chieftain Ao-lo Po-chi-lieh was invested as assisting state ruler of the Meng. Then peaceful relations were established, and Chin annually gave them very generous presents. Thereupon Ao-lo Po-chi-lieh called himself Ancestral Originating Emperor [Tsu-yiian huang-ti] and proclaimed the era t'ien-hsing [Heavenly Rise]. The great Chin had used military force but eventually could not subdue them [the Mongols] and only sent elite troops that occupied some strategic points and then returned.9 It is not clear to which Mongolian chieftain the term Ao-lo Po-chi-lieh refers. It is a composite title; the second half is the Jurchen word bogile, "leader, chief." The first half of the title could be a transcription of the Mongolian word a'uruigh), "base, camp," so that the whole title would mean something like "chief of the camp." A modern Japanese scholar has suggested that the title Ao-lo Po-chi-lieh refers to Khabul khan, the grandfather of Chinggis khan, who indeed, as the Secret history of the Mongols tells us, "ruled over all the Mongols."10 In other words, it seems that around 1146—7 the chieftain of the Mongols had become an "outer vassal" of the Chin state and had been accorded a title appropriate to his rank. It should not surprise us that both the Secret history and the Yuan-shih (Yuan history) remain silent on this episode. The Chin shih (Chin history) also omits to mention it, perhaps because it was compiled under the Mongolian Yuan dynasty and therefore tended to pass in silence over anything that could point to a vassal status of the Mongols under the predecessors of Chinggis khan. It is also significant that all our information on these early Mongolian-Chin relations comes from Sung sources, which did not have to consider political prohibitions imposed by Mongolian rule. In any case it remains a fact that after 1146 the Mongols had already become a major power in the steppe regions that the Liao government had formerly found difficult to control. This was to some extent a political configuration similar to that a generation earlier when the Jurchens themselves had been vassals of the Liao on their eastern frontier and tried to win formal and factual independence from their overlords. At the same time, in 1146, the Chin tried to win the allegiance of the Western Liao or Khara Khitai, who under Yeh-lii Ta-shih had founded an empire in Central Asia. But this diplomatic initiative failed completely, and the chief envoy was killed on his way to the far west. This same envoy had successfully estab9 This agreement with the Mongols is not mentioned in the Chin sbib, but it appears in Yii-wen Mouchao, Ta Chin kuo chih (KHCPTS, ed.), 12, pp. 99—100; and in Li Hsin-ch'uan, Chien-ytn i lai ch'ao yeh tsa chi (KHCPTS ed.), 19, p. 591. 10 Jicsuzo Tamura, "The legend of the origin of the Mongols and problems concerning their migration," Ada Asiatica, 24 (1973), p. 12. Cambridge Histories Online © Cambridge University Press, 2008 THE POLITICAL HISTORY OF C H I N AFTER I I 4 2 239 lished relations in 1144 with the Uighurs to the west of the Hsi Hsia state. Unlike Sung, Kory6, and Hsi Hsia, however, the Uighurs did not send regular courtesy embassies for the New Year festival and the emperor's birthday every year but appeared only irregularly to pay homage to the Chin with an offer of local products. The situation of Chin in the intricate multistate system of East Asia was thus firmly established. What created an element of instability was the Chin emperor's own personality. Apart from his conduct he seems to have suffered from persecution mania and repeatedly had high officials and even members of his own clan killed on flimsy pretexts. Inevitably a faction developed against him, and finally the conspirators murdered him on 9 January 1150. The conspiracy was headed by the emperor's cousin Ti-ku-nai, whose Chinese name was Wan-yen Liang (1122-61). He was duly enthroned as emperor, but the Chin shih does not recognize him as such; the sources always refer to him as prince of Hai-ling, Hai-ling wang, and in 1180, many years after his death, he was even posthumously demoted to the rank of a commoner. The Hai-ling wang episode In the rogues' gallery of Chinese rulers, Hai-ling wang occupies a place of honor. Sung and Chin sources alike describe him as a bloody monster. Indeed, in this respect he proved far worse than Hsi-tsung had ever been. For him it became standard procedure to murder his opponents, including those from the imperial clan itself. The fact that he transferred the wives and concubines of the murdered princes into his own harem added, in the eyes of Chinese historians, lechery to bloodthirstiness. In later centuries he even became an antihero in popular pornography, where his exploits are embellished with gusto. But it would be wrong to judge his personality from the point of view of the moralizing sources alone. Hai-ling wang was in fact a far more complicated person than the cruel and ruthless usurper he appears to have been at first sight. There was method and purpose behind his seemingly senseless acts of violence. He marks the last phase of transition from a more collective and clan-dominated leadership to monarchic autocracy. At the same time he was, strange as it may sound, a great admirer of Chinese civilization, and some part of his ruthless extermination of Jurchen grandees can be interpreted as a fight against the advocates of the old tribal and feudal ways of life. Another motivation was the elimination of Wu-ch'i-mai's descendants because he wanted to keep the succession to the throne in A-ku-ta's lineage. Hai-ling wang was an avid reader and had studied the Chinese classics and histories. Greatly impressed by the many Sung Chinese whom he had met after the resumption of diplomatic relations and by their ways, he Cambridge Histories Online © Cambridge University Press, 2008 240 THE CHIN DYNASTY had adopted such typically Chinese customs as playing chess and drinking tea, so that in his youth he had been given the nickname of Po-lieh-han, a Jurchen phrase meaning something like "aping the Chinese."11 Under Hai-ling wang's rule, many reforms that tended to sinify the Jurchen state and society were introduced, in ritual and ceremony as well as in fiscal policy and administration. No longer content with the fact that the political center of the Jurchen state was still to a large extent in the underdeveloped region of Manchuria, he began to shift the existing central agencies to the south. Yen-ching (Peking), which had hitherto been the Southern Capital, was reconstructed, and new palaces were built there. In 1152 Hailing wang took up residence in Yen-ching and had it renamed the Central Capital. Some years later, in 1157, he even ordered the destruction of the palaces and mansions of the Jurchen chiefs in the Supreme Capital in northern Manchuria and had the status of the town reduced to that of a simple prefecture. He also gave orders to build an imperial residence in K'ai-feng, the former Sung capital, and made it his Southern Capital. All this shows how much Hai-ling wang wanted to become a Chinese ruler instead of a Jurchen leader. But his aspirations went beyond the sinification of his realm. He regarded himself as the potential overlord of all of China and considered his legitimacy as the ruler of China to be as good as that of the Sung. After eliminating, chiefly through murder, those of his opponents who were in favor of continuing the coexistence policy with Sung, Hai-ling wang began preparations for a new war of conquest. His pretexts were not subtle: In 1158 he accused the Sung of violating the 1142 treaty by illicitly purchasing horses at the border markets. From 1159 on, Hai-ling wang thoroughly organized preparations for an all-out attack against Sung. In order to eliminate a possible diversion by unrest on the border with Hsia, the minister of war was dispatched to inspect the border and its delimitation. Horses were requisitioned in great numbers; the total is reported to have been 560,000 animals. A huge store of military weapons was brought together and provisionally stored in the Central Capital (Peking). The emperor, having realized that a huge campaign could not take place with the Jurchen soldiers alone, ordered the registration of Chinese soldiers throughout the whole country. Such measures met with local resistance by the Chinese population and the Chin shih records several minor revolts led by Chinese, particularly in the southeastern region, which bordered on Sung. The recruitment among the population continued until the summer of 1161. Hai-ling wang had also foreseen that an advance into the Sung territory 11 Ta Chin kuo chih, 13, p. 103. Cambridge Histories Online © Cambridge University Press, 2008 T H E POLITICAL HISTORY OF C H I N AFTER I I 4 2 241 would have to rely largely on river transport and that naval battles with the Sung navy were inevitable. For this reason, thirty thousand sailors were registered, and a great number of barges were confiscated to serve for transport and as warships. Following the tradition of his ancestors, Hai-ling wang assumed personal command over his troops and, in the seventh month of 1161, proceeded from the Central Capital to the Southern Capital (K'aifeng). Shortly afterward he tried to assert his legitimacy as ruler over the whole of China by an action that was as desperate as it was cruel. He ordered the liquidation of all surviving male descendants of the Yeh-lii clan and the Chao family, thus exterminating any potential pretenders from the former ruling families of the Liao and Sung empires. It is reported that over 130 persons were killed during the summer of 1161. Hai-ling wang's ruthless action aggravated the unrest among the Khitans, who consequently resorted to open rebellion. A force of ten thousand soldiers had to be diverted to Manchuria in order to quell the revolts. Another atrocity ordered by Hai-ling wang was the murder of his own stepmother, the empress dowager, nee T'utan, together with ten of her court ladies. His reason was that she had dared to criticize the campaign against the Sung. Apparently there was never a formal declaration of war. Normal diplomatic intercourse in the form of the usual courtesy embassies for congratulating the other state for the New Year and the emperor's birthday continued uninterrupted throughout the period when Hai-ling wang was already preparing his campaign, although the mobilization certainly did not go unnoticed by the Sung court. In the summer of 1161 the Chin dispatched the usual embassy to congratulate Sung Kao-tsung on his birthday, and the envoys were received in audience as late as 14 June 1161. According to the Sung shih (Sung history), one of the envoys "behaved insolently" and gave the Sung court to understand that Hai-ling wang would attack soon. This act of treason later cost him his life. In any case the Sung were sufficiently warned and found time to strengthen their border defenses. Hai-ling wang had concentrated his troops on the Huai River front. He left K'ai-feng on 15 October; shortly afterward the Chin armies crossed the Huai (on 28 October) and advanced toward the Yangtze River. The Sung, who regarded the Yangtze as their most important defense line, were unable to stop the invaders and to prevent the Chin columns from reaching the river. They concentrated huge forces on the southern banks of the Yangtze in order to ward off any attempt to cross the river. On other sectors of the front, however, the Sung were able to score some local successes by conquering some Chin prefectures along the western part of the border. This destroyed any hopes for the "Blitzkrieg" that Hai-ling wang might originally have cherished. Cambridge Histories Online © Cambridge University Press, 2008 242 THE CHIN DYNASTY While Hai-ling wang set up his camp near Yang-chou, some of his subordinate generals tried to cross the Yangtze River some sixty miles upstream at Ts'ai-shih (south of modern Ma-an shan in Anhui Province). This attempted crossing failed on 26 and 27 November 1161. Sung historiographers later viewed the Chin army's failure to advance south of the river as a victory of gigantic dimensions; it was compared with the battle on the Fei River in A.D. 383 when the southern state of Chin had allegedly defeated the northern invaders led by Fu Chien, the ruler of the Ch'in state. The Fei River battle was given in Chinese historiography the dimension of a paradigm, namely, that the defenders of Chinese civilization were able to crush an invasion by barbarian northerners. Modern scholarship has shown that the Fei River battle was, if not a downright myth, at least greatly exaggerated in importance.I2 It is a question whether the Ts'ai-shih victory by the Sung might be a similar case of idealized exaggeration. Sung sources report that the Sung had only 18,000 soldiers at Ts'ai-shih, whereas the Chin are said to have had 400,000 men. This is an obvious misrepresentation. Although 18,000 might be a relatively correct figure for the units concentrated around Ts'ai-shih, the total strength of the Sung armies on the middle Yangtze front is elsewhere given as 120,000, and so the 400,000 men of the Chin can refer only to the total strength of the Chin armies under Hai-ling wang, a figure that also included the many noncombatants attached to the fighting units. Even the Sung sources disagree as to the attackers' losses, either by drowning when crossing the river in their boats or in combat after having reached the southern bank. It seems safe to assume that the Chin losses did not exceed more than about four thousand men. In other words, the defeat at Ts'ai-shih did not by any means fatally affect Chin power. It must also be remembered that the defenders enjoyed considerable advantages. Not only were their warships larger than those of the Chin, a fact admitted on the eve of the battle by a Jurchen commander, but also the time needed by the Chin leaders to bring together a sufficient number of boats and soldiers was so great that a surprise attack was out of the question. Moreover, the Chin were not in a position to use their most dreaded weapon, their cavalry. A modern Chinese scholar carefully analyzed the background and the battle itself and concluded that Ts'ai-shih was a relatively small scale military engagement. '* But the psychological effect of the Ts'ai-shih battle was by no means negligible. The Sung regained confidence and felt strong again. The eventual outcome of Hai-ling wang's reckless campaign was, however, 12 Michael C. Rogers, "The myth of the battle of the Fei River (A.D. 383)," T'oung Poo, 54 (1968), pp. 50-72. 13 T'ao Chin-sheng, Chin Hai-ling ti u fa Sungyii Ts'ai-shih chan i te k'ao shih (Taipei, 1963). Cambridge Histories Online © Cambridge University Press, 2008 T H E POLITICAL HISTORY OF C H I N AFTER II42 243 not decided on the battlefield but by other factors. Over the years, Hai-ling wang had made himself hated even among his own entourage. Not surprisingly, therefore, he was murdered, together with five of his concubines in his camp tent near Yang-chou on 15 December 1161 by a group of officers. The widespread dissatisfaction with Hai-ling wang's autocratic rule, among aristocratic Jurchens as well as Khitans, Po-hai, and Chinese, had resulted in a conspiracy by a comparatively moderate faction of the imperial clan members to dethrone him and to replace him with his cousin Wu-lu. The latter had already been proclaimed emperor on 27 October 1161 in Liao-yang, some weeks before Hai-ling wang was killed. The news of the coup in Liao-yang had certainly reached Yang-chou by mid-December and thereby prompted the assassins' action. The new ruler, later canonized as Shih-tsung (1123— 89), was faced with an extremely difficult situation: Khitan rebellion, unrest in the Chinese provinces, and the uncertain issue of the campaign against Sung. It is no small credit to Shih-tsung's abilities that he proved himself able to overcome all these problems and to achieve a stature that not only places him among the great Jurchen leaders but also secures him respect as one of the most remarkable persons who ever occupied the Chinese throne. The golden age of Shih-tsung (1161—1189) The new emperor Yung (originally Wan-yen Wu-lu, b. 1123, r. 1161-89; posthumous temple name Shih-tsung) was a grandson of the founder of the dynasty, and his father Wan-yen Tsung-fu had distinguished himself during the early years of the Chin state. His mother did not come from one of the Jurchen clans from which the Wan-yen customarily chose their consorts but was of Po-hai descent. The Po-hai element in the state of Chin formed one of the culturally advanced segments of the population. Already during Hai-ling wang's lifetime the future Shih-tsung had been one of his open opponents and was therefore not given an executive metropolitan command post but was appointed as vice-regent of the Eastern Capital, Liao-yang, a central Pohai settlement. When the Khitan rebellion erupted in the summer of 1161, Shih-tsung was able to subdue unruly Khitan elements in the area under his control. Because Shih-tsung was supported chiefly by the Po-hai people, and also to some extent by Jurchen grandees dissatisfied with Hai-ling wang's reckless war against the Sung, after he was proclaimed emperor he hastened to the Central Capital (Peking). The first order he gave early in n 6 2 was to withdraw the Chin armies from the Yangtze front, and in spring of that year he sent envoys to the Sung asking for a resumption of normal relations. But Shih-tsung was cautious enough not to demobilize his forces, and there was Cambridge Histories Online © Cambridge University Press, 2008 244 THE CHIN DYNASTY repeated border warfare both in the Huai region and on the southwestern front where the Chin state faced the Sung province of Shu (Szechuan). Finally in 1165 a peace treaty with Sung was concluded that improved the terms for the Sung. They no longer had to regard themselves as vassals of Chin; instead, the relationship was to be that of uncle and nephew. This still gave the Chin a slight ritual and ceremonial advantage over Sung, but at least the term "vassal" had disappeared. Also, the word "tribute" (kung) was replaced by the more neutral word "payment" (pi). The amounts to be paid remained the same and the Huai border was not changed. As a precaution Shih-tsung ordered sixty thousand soldiers to be garrisoned on the border with Sung when the border markets were reopened. For forty years after 1165, ChinSung relations remained peaceful. That the conclusion of a peace with the Sung was delayed for some years certainly resulted from Sung hopes to take advantage of Chin's internal difficulties, namely, the Khitan rebellion. The Khitans, together with the Hsi, a Turkic tribe, had refused to mobilize for war against Sung and had taken up arms against their Jurchen overlords. The center of the rebellion was the northwestern province in the region near the modern Great Wall. Some Khitan leaders had tried to establish contacts with the Sung in the hope of obtaining support from them. Shih-tsung, however, was able to defeat the rebellious forces in the autumn of 1162; some scattered Khitan warriors fled to Sung. The existing Khitan military units were disbanded, and their individual members were distributed among the Jurchen units. Only those Khitan leaders who had remained loyal were allowed to keep their former ranks. For protection against local elements of doubtful loyalty and also for defense against the ever-unruly Mongolian neighbors in the northwest, dozens of country towns in that area were fortified during the following years. Within a few years Shih-tsung had thus succeeded in stabilizing the internal and external situation of his state. The remainder of his long reign saw many reforms in the field of administration, economy, and education. Apart from some border fighting against the Mongols, the Chin state enjoyed uninterrupted peace for more than twenty-five years. Shih-tsung tried to be a just and economical ruler, and he was given the nickname of "Little Yao and Shun."14 Unlike his predecessor, however, he was not an unqualified admirer of Chinese civilization but had great nostalgia for the frugal Jurchen ways of life as he saw them embodied among those Jurchens who still lived in their ancestral homes in Manchuria. He viewed with concern the growing sinicization of Jurchen aristocracy and commoners and attempted to preserve their national identity through edicts prohibiting them from adopting Chi14 CS, 8, p. 204. Cambridge Histories Online © Cambridge University Press, 2008 THE P O L I T I C A L H I S T O R Y OF C H I N AFTER II42 245 nese names or dress. The old residence in Hui-ning was again rebuilt and raised to the status of Supreme Capital, and in 1184-5 Shih-tsung spent almost a full year there, reveling in the memory of the glory of his forefathers and listening to the recitals of bards praising the exploits of A-ku-ta. It is certainly due to the long period of internal stability and relative prosperity under Shih-tsung that the Chin state was able to maintain itself against overwhelming odds in later years. Decline set in under his successor, not so much through neglect by the Jurchen ruling minority or incompetence at the highest levels as through the attacks of an enemy who proved even more dangerous to Chin than the Jurchen had been to the Chinese. The last years of the twelfth century saw the rise of the Mongolian federation under Chinggis khan, the first stage in their rise to supremacy and dominion over all of Eurasia. Sung revanchism: the war 0/1206—1208 When Shih-tsung died in the beginning of 1189, he was succeeded by his grandson, as the prince originally appointed as heir apparent had died in 1185. The new ruler, later canonized as Chang-tsung (1168—1208), was not as strong a personality as his grandfather had been but was full of good intentions and was also able to maintain some sense of unity among the members of the imperial clan and the Jurchen grandees who were prominent in high military and civilian posts. Despite all the prohibitions against the sinicization of the Jurchens, Chinese ways were adopted by more and more Jurchens and also Khitans. But there were indications of imminent decline. The Mongolian menace that had been more a nuisance than a real threat in previous years now grew to considerable proportions. Another element that tended to undermine the internal stability of Chin was a factor more or less beyond the control of even the best of governments. The Yellow River, which had repeatedly inundated the plains in Hopei and Shantung in previous years,floodedvast regions in 1194 and formed two new courses (see Map 18). These catastrophes affected the most fertile and economically important regions of the state, with the usual consequences: displaced farmers, disrupted communications, and occasional unrest among the population. Recent research has attributed the decline of Chin power largely to these natural causes, which debilitated the economic foundations of the Chin state.1' The internal conditions of the Chin state were not unknown to the Sung, although it remains an open question as to how much the Sung knew about 15 Toyama Gunji, Kincho shi kenkyii, Toyoshi kenkyu sokan no. 13 (Kyoto, 1964), pp. 5 6 5 - 9 2 , gives a detailed account of the successive inundations of the Yellow Rivet duting the twelfth and early thirteenth centuries. Cambridge Histories Online © Cambridge University Press, 2008 246 THE CHIN DYNASTY Northern course of Huang-ho until 1194 ^ H B Main course after 1194 Branch channels after 1194 — — — Eastern course in late 11th century S^ Areas suffering from serious flooding 1166-94 FLOODS 1166-68 Tung-p'ing f u Liangshan po 'i-chou MAP 18. The change of the Huang-ho's course, 1194 the rising power of the Mongols. The Chin government under Chang-tsung at last realized the extent of the Mongolian danger, and after 1192, largescale fortifications were constructed on the northwestern border against the invasions of the Tatar, Onggirad, Salji'ud, and other Mongolian tribes. These defense measures, together with the repeated "punitive" campaigns into Mongolian territory, severely taxed the resources of the Chin just at a time when the Yellow Riverfloodshad hit their agricultural surplus areas in north China. The government resorted to confiscating Chinese-owned lands, chiefly those of tax evaders, and gave them to reliable Jurchens, with the inevitable result that relations between the Chinese and Jurchens became strained. In these circumstances the Chin had no intention whatever to abandon their coexistence policy toward the Sung. Their agricultural resources had Cambridge Histories Online © Cambridge University Press, 2008 T H E POLITICAL HISTORY OF C H I N AFTER 1 1 4 2 247 been affected not only by the Yellow River floods but also by a series of droughts and devastations by locusts in the provinces bordering on the Sung. The key area of Shantung was affected even more badly than other regions were. The Sung court was well aware of the precarious conditions in the Chin state: A regular source of information was the embassies that twice a year traveled to the Central Capital (Peking) and crossed just those parts of Chin territory that had suffered most from the natural calamities. It is perhaps not sheer coincidence that a relatively great number of the Sung commanders taking part in the war had been at one time or other envoys to Chin. From 1204 on, local violations of the border in the Huai region were committed by Sung military leaders. The Sung minister responsible for reviving this revanchism was Han T'o-chou, who had been instrumental in enthroning Emperor Ning-tsung in 1194 and had twice (in 1189 and 1195) served as an envoy to the Chin. He either covertly encouraged or connived at raids by the Sung army into Chin territory, so that the years from 1204 onward are characterized by a gradual escalation of hostilities. These did not at first interrupt either normal diplomatic intercourse or the annual payments by the Sung, but the psychological climate at the ceremonies accompanying the audiences for foreign envoys deteriorated considerably. At the New Year's audience at the Sung court in 1206 (5 February), the Chin envoy mistakenly believed that the Sung courtiers had violated the taboo on the personal name of the Chin emperor's father and protested strongly. This was, of course, interpreted by the Sung as arrogant insolence. All this points to a growing irritation and instability of the formal relations between the two powers. By the summer of 1206 the Sung minister Han T'o-chou thought that at last the time was ripe to resort to war, and the deputy minister of war, Yeh Shih (1150-1223), a prominent literatus, was ordered to draft the declaration of war. He refused, however, because he opposed the war and as a result was demoted to a provincial post. In his place Li Pi, a follower of Han T'ochou and member of the revanchist party, composed the formal decree declaring war on the Chin, and this was promulgated throughout the Sung empire on 14 June 1206. Six days later the beginning of the northern campaign was solemnly proclaimed to Heaven and Earth, to the imperial ancestors, and to the gods of the land, a ritual that made war irrevocable. In the meantime the Chin, too, had realized that war was inevitable and resorted to the usual formalities such as mobilizing their whole armies and solemnizing their campaign by declaring to their ancestors and gods the breach of the existing treaty of 1165. The text of the Sung decree announcing the war has been preserved in its entirety. It is a document combining moral indignation and accusations Cambridge Histories Online © Cambridge University Press, 2008 248 THE CHIN DYNASTY against the Chin, with the assertion that the Chin, through their evil actions and incompetence, had lost the Mandate of Heaven and thus the legitimate claim to rule their country. At the same time the document expresses the firm hope that the Chinese population of the Chin state would rise against the Jurchens and ally themselves with the Sung. The Sung forces deployed on the most important front, that along the Huai River, totaled about 160,000 men. This figure seems to be realistic. The Chin armies mobilized against the Sung in December 1206 consisted of 135,000 men distributed from east to west, including their forces in Shensi Province facing the rich Sung province of Szechwan. It seems, therefore, that the Sung had in the beginning a slight numerical advantage over their adversaries. But the Sung campaign was soon doomed to fail. It is true that the Sung were able to take the border town of Ssu-chou on the northern bank of the Huai River, but this cannot be regarded as a great victory. The town was, as we know from a Sung observer who visited the place, hardly defended at all, and its walls consisted of only a low mud rampart. The attempts by Sung commanders to attack the strategic Chin prefectures of T'ang and Teng in northern Hupei failed miserably. Tens of thousands of Sung soldiers deserted and dispersed, due partly to the weather, however. It had rained heavily for days; the soldiers had to camp in tents that were soon flooded by water; and provisions did not arrive in time so that the soldiers suffered from hunger. The hay for the Sung horses was wet and rotted. Contemporary Sung sources agree that the Sung campaign was badly organized and incompetently led. On the other hand, in the autumn of 1206, the Chin armies penetrated deeply into Sung territory and laid siege to a great number of prefectural towns. They also advanced in the west and occupied several Sung strongholds in southern Shensi. The Sung hopes that the Chinese population in the Chin state would gladly rise up and join the Sung proved to be an illusion. A large-scale defection of Chinese did not occur. On the contrary, the Sung governorgeneral of Szechuan, Wu Hsi, whose family had held high offices in Szechuan for several generations, declared his allegiance to the Chin, who invested him as prince of Shu. This was a serious blow to the Sung defenses in Szechuan, because Wu Hsi had seventy thousand soldiers under his command. Wu Hsi's defection took place in December 1206 and threatened the total collapse of the western front. The Sung were spared the worst, however, when on 29 March 1207 a group of loyal officers murdered Wu Hsi. The war continued, nonetheless, even though no major and decisive engagements took place after April 1207. The first feelers for a resumption of peaceful coexistence were initiated by the Sung. As in the 1140s, when revanchists and appeasers had struggled for Cambridge Histories Online © Cambridge University Press, 2008 THE POLITICAL H I S T O R Y OF C H I N AFTER II42 249 supremacy, there now emerged forces at the Sung court in Hang-chou that advocated a policy of coexistence with Chin. Han T'o-chou was dismissed from his office and soon afterward was murdered (27 November 1207). He and his clique were held responsible for a war that had been begun irresponsibly. The Chin were equally disinclined to engage in a prolonged war and favored a return to normal relations and, above all, the resumption of the Sung annual payments. They were additionally motivated toward making peace by the Mongolian menace on their northern frontier. During the negotiations in the second half of 1207 and in 1208, the Chin had consistently regarded Han T'o-chou as the chief "war criminal" and had repeatedly demanded his extradition. When they were informed of his death, they asked for the delivery of his head as a token of expiation. The Sungfinallyaccepted the Chin's demands. In July 1208 the Chin proclaimed the end of the war, and the Sung formalized the new peace treaty on 2 November 1208. The annual payments from Sung were raised by fifty thousand ounces of silver and fifty thousand bolts of textiles, and Han T'o-chou's head was duly dispatched to Chin, preserved with lacquer and packed into a box. This gruesome souvenir of the war was displayed in the ancestral temple of the Chin imperial family.l6 Emperor Chang-tsung lived to see peaceful relations with Sung restored. He died on 29 December 1208. Throughout his reign, which had lasted for almost twenty years, he had tried hard to strengthen Chin's defenses and at the same time to promote the introduction of reforms that were planned to convert the Chin state into a body politic modeled on T'ang and Sung precedents. He was instrumental in introducing a law code in order to unify the existing disparate legislations. This code, the T'ai-ho lii, was proclaimed in 1202. It is significant that at about the same time the emperor adopted the "power" or "element" earth from among the five agents {wu hsing) as the emblem of his dynasty. This was the outcome of long and protracted discussions among his political advisers. In traditional political thought, each legitimate dynasty was assigned one of the five powers. The Sung had adopted the element fire, and the emblematic color of their dynasty was red. The formal adoption of a cosmic ruling power was therefore an act of major political significance. It meant that from then on the Chin dynasty had claimed a place in the legitimate succession of Chinese imperial states. On the political level the adoption of the earth power signaled to the world at large, and to the Sung in particular, that Chin considered itself legitimized to rule All under Heaven, that is, the whole 16 For a detailed study of the 1206 war, see Corinna Hana, Bericht iibtr die Verteidigung der Stadt Te-an wdhrend der Periode K'ai-hsi, 1205 bis 1208 (Wiesbaden, 1970). See also Herbert Franke, Studien und Texte zur Kriegsgachicbte der siidlichen Sungzcit, Asiatische Forschungen no. 102 (Wiesbaden, 1987). Cambridge Histories Online © Cambridge University Press, 2008 250 THE CHIN DYNASTY civilized world. The Sung could therefore no longer claim to be the only state on Chinese soil to exercise legitimate rulership, and it even seems to be a reasonable assumption that the deterioration of Sung-Chin relations was partly due to this new self-image of the Chin as a rival legitimate state in the Chinese sense.17 Altogether, Chang-tsung's reign marks an apex of sinicization and a reversal of Shih-tsung's well-meaning but sometimes impractical nativism. The war of 1206—8 had shown that the military strength of the Chin state was largely unimpaired and that it continued to be a formidable adversary of the Sung. But this, together with internal reforms aimed at reconciling the Chinese population with Jurchen rule, was not enough to guarantee the survival of the state under the onslaught of the Mongols. The distintegration of the Chin state The external threat to the Chin state after 1208 coincided with a period during which the throne was occupied by a ruler whose right of succession was disputed. It had originally been the custom to have brothers succeed each other before passing on to the next generation. When Chang-tsung died, even though he was survived by an elder brother, a court clique enthroned another prince, Wan-yen Yun-chi, the seventh son of Shih-tsung, born of a Po-hai concubine. The histories do not accord him the title of emperor, and hence he is known merely as Wei-shao wang (r. 1208—13). Mongolian conquests and internal dissension (1208—1215) Already under Chang-tsung the Chin had made their preparations against Mongolian invasions by strengthening their northwestern border. A line of fortifications connected by walls and ditches was constructed, partly using defense works that had been built during the early reigns of the Chin. It stretched from the region of modern Tsitsihar in a southwestern direction parallel to the great Hsingan Mountains as far as Lake Dalai-nor in modern Inner Mongolia and finally joined the original Chinese Great Wall. The military value of these fortifications was, however, impaired by the fact that many of the garrison units stationed along them were of Khitan and Hsi descent rather than reliable Jurchen troops. The reasons that Chinggis khan chose the Chin state as a target for his attacks seem to go back to events in the late twelfth century. The tribe of the Tatars (the Tsu-p'u of the Liao and Chin sources) had been an enemy of the 17 These problems are exhaustively studied in Hok-lam Chan, Legitimation in imperial China: Discussions under the Jurchen—Chin dynasty (1115—1234) (Seattle, 1984). Cambridge Histories Online © Cambridge University Press, 2008 T H E POLITICAL HISTORY OF C H I N AFTER II42 251 Mongols for a long time, and bloody feuds had flared up between them time and again. The Tatars were loosely allied with the Chin, but around 1190 they renounced their allegiance. A punitive expedition against them was organized by the Chin under the imperial clansman Wan-yen Hsiang in 1196, and the Mongols joined this campaign to take their revenge against their old enemies. Jurchen and Mongolian forces together penetrated deeply into Mongolia and at last, in the eighth month of 1196, succeeded in inflicting a crushing defeat on the Tatars, whose tribal chief perished on this occasion. The Mongolian contingents had fought together with their allies from the Kereyid tribe under To'oril. Their assistance in quelling Tatar power and aspirations was rewarded by the Chin emperor; To'oril was given the rank of prince (wang) and was henceforth known as Ong khan, whereas Temiijin (the name by which Chinggis khan was called before his enthronement as khan in 1206) was invested with a lesser title, probably of Khitan origin. In any case he had to regard himself from then on as an outer vassal of the Chin, although the title he had received from the Chin court must have enhanced his prestige among the other steppe tribes. It is self-evident that after his rise in 1206 Chinggis khan was no longer content to be treated as a Chin vassal and that he aimed at formal independence from the Chin, however loose his vassal status might have been. An additional motive might have been the conquest of Chin territory, a country that must have seemed full of incredible riches to the steppe nomads. A third motive was perhaps revenge for the death of Ambaghai khan. Ambaghai had been proclaimed as the successor of Khabul khan and the leader of the Mongolian federation. He was a cousin of Khabul and the founder of the Tayichi'ud lineage of the Mongols. He, too, was on bad terms with the Tatars, who in one of the constant raids on each other took him prisoner and extradited him to their Chin overlords. The Chin, it seems, then had him killed cruelly. Chinggis khan, who considered himself a legitimate successor of Ambaghai khan as leader of the Mongols, perhaps resented the ignominious death of Ambaghai at the hands of the Chin, but this must remain speculative in view of the deficiency of our sources. Finally, yet another reason for Chinggis's hatred of Chin may well have been personal biases of the Chin ruler himself. When he was still a minor Chin prince, Wei-shao wang had accepted the customary tribute presents from Chinggis khan and, in the eyes of the Mongolian ruler, had behaved insolently toward him. When they received an order from the ruler of Chin, the Mongolian tribute bearers should have kowtowed before the representatives of the Chin state. But when Chinggis khan learned that the new Chin emperor was Wei-shao wang, who had earlier insulted him, he flew into a rage and in 1210 broke off tributary relations with Chin, deciding on an allCambridge Histories Online © Cambridge University Press, 2008 252 THE CHIN DYNASTY out attack against his Jurchen overlords.'8 This decision was certainly prompted by information that the Chin state at that time was suffering from a severe famine. In the spring of 1211 the Mongols advanced on the Chin borders with two armies, the eastern one led by Chinggis khan himself (see Map 19). The border fortifications constructed some years earlier under Chang-tsung were easily overrun, and soon the Mongolian forces were able to take the strategic pass of Chii-yung kuan that protected the Chin Central Capital (modern Peking) in the north. The reinforcements sent by the Chin high command were easily defeated by the Mongols, who then plundered and ravaged the countryside around the capital. But there was no attempt to invest and take the city at that time. On the western front the Mongols advanced into Shansi and thereby prevented Chin auxiliary troops in Shensi, where strong garrisons protected the border with Hsi Hsia, from coming to the rescue of the Chin armies on the eastern front. All of this shows the strategic ability of Chinggis khan, but apparently the campaign of 1211 did not yet have conquest as its objective; it was, rather, a series of reconnaissance raids and expeditions to pillage the countryside. During the winter the Mongols withdrew their armies, which gave the Chin a chance to reorganize their northern and northwestern defenses. The Mongols renewed their attacks in the autumn of 1212. In the next spring, they again took Chii-yung kuan and this time penetrated even more deeply southward into northern China, devastating parts of Hopei, Shantung, and Shansi and taking several towns. During the winter of 1213—14 the Mongols effectively blockaded the Central Capital. All this happened at a time when the Chin court itself was in a state of chaos. In the eighth month of 1213 Wei-shao wang, a ruler who despite his irregular succession to the throne had shown a keen concern for the defense of his state, was killed in a conspiracy. The leader of this plot was Hu-sha-hu from the Hoshih-lieh clan, a Jurchen grandee who had been the commander of the Western Capital (Ta-t'ung in Shansi Province), which had fallen to the Mongols. He enthroned the elder brother of Chang-tsung, Wu-tu-pu (1163-1223), later canonized as Hsiian-tsung (r. 1214-23). One of the reasons that led Husha-hu to do away with Wei-shao wang and to install an emperor whom he believed he himself could control was apparently that he feared disgrace and punishment for having lost the strategic areas under his command. These events happened when Mongolian troops were advancing toward the Central Capital and their cavalry detachments were already raiding the northern parts of Hopei, Shantung, and Shansi provinces. There was, however, 18 This is the version given in Sung Lien et al., eds., Yiian sbib (Peking, 1976), 1, p. 15 (hereafter cited as YS). See also Chapter 4 in this volume. Cambridge Histories Online © Cambridge University Press, 2008 :::;*;*;* Border of Chin and Hsi-Hsia V ™ Mongolia » Campaigns 1211—12 ^ ' - £ Jebes campaign into Manchuria - - - * • Campaigns 1213-14 1211-12 * Samukha's campaigns, 1216—17 Chinggis returned to the North after the fall of Chung-tu, 1216-17 from Mongolia^ 1911 0211,v. Hua-chou* 12tg,ix.-^—^. Ching-chao MAP 19. Chinggis's campaigns against Chin Cambridge Histories Online © Cambridge University Press, 2008 n _300km 200 miles 254 THE CHIN DYNASTY another factor besides the dissension at the Chin court that had contributed decisively to the defeats suffered by their armies. Repeated droughts in north China had led to widespread famine and thereby undermined the logistics of the Jurchen war machine. In a last-minute effort the government tried to rally all population groups against the Mongols by abandoning the existing differentiations among the races. Military and civilian posts were opened to Khitans and Chinese without the former numerical limitations. In spring of 1214 the Chin sent envoys to the Mongols to ask for peace and also offered in marriage a daughter of Wei-shao wang to Chinggis khan. The Mongols withdrew from the Central Capital. The situation in the north remained precarious, however, and so Hsiian-tsung decided to transfer his court to the Southern Capital (K'ai-feng), which was not only in the center of the agriculturally developed Chinese plains but was also protected from the north by the Yellow River. This transfer of the capital was interpreted by Chinggis khan as a preparation for the resumption of war, and so he decided to march again against the Central Capital. On 31 May 1215 the city surrendered to the Mongols and former subjects of the Chin such as Khitans and Chinese who had defected to them. The capital was by far the most populous and important city conquered so far by the Mongols in East Asia. At about the same time, diplomatic relations between Chin and Hsia collapsed, after many years of increasing strain, and a period of intermittent warfare and intense hostility characterized the decade from 1214 onward. This unfortunate development replaced the formerly good relations between Chin and Hsia and was largely the result of factionalism and power struggles at the courts of both states, which undermined their ability to turn back the Mongols. Rebellions in Shantung The disastrous loss of the Central Capital, a major Chin administrative center and military stronghold, was paralleled by serious setbacks in other parts of the state. In 1214 the Chin had asked the Sung to deliver the payments stipulated by the treaty of 1208 one year in advance in order to make up for losses suffered in the past. Instead, the Sung refused to pay at all and thereby aggravated the fiscal problems in the tottering Chin state. This coincided with the outbreak of local rebellions in Shantung, a part of China that had throughout history been a hotbed of social and religious unrest (see Map 20). The first leading figure of the rebellion was Yang An-kuo. He came from a fortified village in eastern Shantung inhabited by members of the Yang clan, who specialized in manufacturing boots and other leather goods, an occupaCambridge Histories Online © Cambridge University Press, 2008 I L I A O | Rebel regimes / ITA-CHEN WAN-NU PlbjHSIEN 1215-33 H Chin territory 1223 \ Vassal of Mongols HS I - HSIA OCCUPIED BY conquered by Mongols MONGOLS BY 1223 AIN I NG TERRITORY LI-CHUAN attacks Sung 1230 .killed 1231 500 km 300miles MAP 20. The rebellions in Manchuria and Shantung Cambridge Histories Online © Cambridge University Press, 2008 Oppressed 1233 by Mongols preparatory to invasion of KORYO 256 THE CHIN DYNASTY tion that earned Yang An-kuo the nickname of Yang An-erh, "Yang the saddler." Yang, a strong and ruthless man, was the village headman and already during the war of 1206—8 had emerged as a leader of a group of local rebels against the Chin authorities. After the war had ended, Yang An-kuo returned to his allegiance and was rewarded by the Chin government with a minor prefectural post and a military command. When the Mongolian attacks in 1213 and 1214 disrupted the administrative machinery of the state, Yang An-kuo resumed his rebellious activities and began to plunder country towns in the eastern part of the Shantung peninsula. In the summer of 1214 he even proclaimed himself emperor in a small coastal town (modern P'englai) and adopted the reign title of Tien-shun, "Heavenly Obedience." This was too much for the Chin authorities who, despite being pressed hard by the Mongols in the north, mobilized an army against him under P'u-san Anchen. This expedition was successful, and by late autumn 1214 the strongholds of Yang An-kuo had been reduced. Yang tried to flee by boat but was captured and killed at the beginning of 1215. His "empire" had lasted hardly more than three months. Yang was not the only rebel in Shantung; two more had risen independently in central Shantung. They, too, were eliminated during P'u-san Anchen's punitive expedition in 1215. But the general unrest in Shantung persisted, and the former lieutenants of Yang An-kuo and the other rebels continued guerilla warfare in the inaccessible mountainous parts of the peninsula. Around 1215 the rebel leaders adopted the color red for their uniforms, and they were known from then on as the Red Coats (Hung ao). Red has always been an auspicious color in China, signaling good luck and hope. The choice of the color red had, in all probability, nothing to do with the insurgents' pro-Sung feelings (red was the emblematic color of the Sung, correlated with the element of fire). Nor can we assign any religious or sectarian motivations to the various rebel leaders; the aims and objectives of all these condottieri who disturbed Chin rule in Shantung seem to have been purely selfish and secular. Each of them tried to carve out for himself a region in which to build up a center of local power. When in 1217 yet another devastating flood of the Yellow River prevented Chin military actions in Shantung, the Sung government tried to use the Red Coats movement to their own advantage. They offered official ranks to the rebel leaders and promised them material assistance. One of them was Li Ch'iian, a colorful person who had become famous for his physical strength and ruthlessness and was a relative by marriage of Yang An-kuo, either his brother-in-law or his son-in-law. In 1218 Li Ch'iian was nominated as a commanding prefect in Shantung by the Sung court, but in fact he was more or less independent and administered his territory as he pleased. By 1219 the Cambridge Histories Online © Cambridge University Press, 2008 THE POLITICAL HISTORY OF C H I N AFTER II42 257 Chin government had lost its control over eastern Shantung where Li Ch'iian reigned on his own. His submission to the Sung, which in any case had been a formality, did not last long. From 1225, as the Mongolian armies advanced into Shantung, Li Ch'iian realized that he would do well to come to terms with the invaders. In 1227 he declared his allegiance to the Mongols and later turned against his former protectors, the Sung. In 1230 he even advanced with his troops deep into Sung territory and attacked Yang-chou on the Yangtze River. But this expedition failed, and Li Ch'iian was killed on 18 February 1231. His death marked the end of the Red Coats. In 1231 his adopted son Li T'an (Marco Polo's "Liitan sangon") inherited his office and continued the warlord career begun by his father. His loyalty turned out to be as fickle as that of Li Ch'iian: When in 1262 he tried to surrender Shantung to the Sung, Khubilai khan had him executed.19 In later traditional Chinese historiography and in modern times the Red Coats "movement" has frequently been labeled as nationalistic and patriotic and as indicative of antiforeign feelings among the lower classes. The Shantung insurgents were not, however, motivated by such modern concepts as nationalism but, rather, were simple adventurers who tried to ally themselves with whichever major power could enhance their own prestige and emoluments. In normal times none of them would have been able to resist the Chin state for long, but in the turmoil following the Mongolian invasions, their rebellions could succeed to a limited extent and thus eliminate Jurchen control over the eastern part of what remained of their state. The loss of Manchuria: Yeh-lii Liu-ko and P'u-hsien Wan-nu The Manchurian homelands of the Jurchens, where many of them still lived, and, in particular, the comparatively prosperous region of Liao-tung, could have been an area for retreat for the Chin government. Indeed, a Jurchen minister had advised Hsiian-tsung to withdraw from the Central Capital (Peking) to the Eastern Capital (Liao-yang) instead of K'ai-feng. However, whereas the Liao-tung region was still under the firm control of Chin when the Mongols attacked in 1211, northern and central Manchuria had already been lost because of the insurrection of Yeh-lii Liu-ko. Liu-ko was a scion of the Liao imperial family and, like so many other Khitan insurgents, had cherished hopes of gaining independence from their Jurchen overlords. With 19 On the Red Coats movement under Yang An-kuo, see CS, 102, pp. 2243—5; a n ^ Franchise Aubin, "The rebirth of Chinese rule in times of trouble: North China in the early thirteenth century," in Foundations and limits of state power in China, ed. Stuart R. Schram (London and Hong Kong, 1987), pp. 113-46. On Li Ch'iian, see T'o-T'o et al., eds., Sung shih (Peking, 1977), chaps. 476 and 477 (hereafter cited as SS); and his biography by Franchise Aubin in Sung biographies, ed. Herbert Franke (Wiesbaden, 1976), vol. 2, pp. 542—6. Cambridge Histories Online © Cambridge University Press, 2008 258 THE CHIN DYNASTY his followers, mostly Khitan cavalry and soldiers, he declared his allegiance to Chinggis khan in 1212, quickly gained control of central and northern Manchuria, and was even allowed to adopt the title emperor of Liao in 1213. A Chin punitive expedition against him in 1214 failed. Liu-ko's puppet state survived until 1233 when the Mongols destroyed it. The general responsible for the abortive Chin campaign against Yeh-lii Liu-ko was Wan-nu, a member of the Jurchen P'u-hsien clan. After his defeat by the Khitan rebels, Wan-nu retired with his troops to the region of the Eastern Capital in southeastern Manchuria. Like so many others he realized that the end of Chin was near and therefore tried to carve out a portion of territory for himself from the remnants of the once-great empire. In the spring of 1215 Wan-nu, too, declared himself independent, adopting the title of king and naming his state Ta-chen. This name was not a geographical name, as practically all Chinese state names had been before this (including that of Chin itself, though in this case there were overtones of Chinese cosmological symbolism). Ta-chen is a highly literary expression standing for "gold" in Taoist texts. This name, therefore, was meant to proclaim that Wan-nu regarded himself as the true successor of the Chin, and to underline this point he also adopted the clan name of Wan-yen, the ruling house of Chin. The Taoist connotations of the state name and other features of Wan-nu's regime were the result of the influence of a very curious person, the Chinese Wang Kuei. He came from the region of modern Shen-yang and had been a specialist in fortune-telling and the exegesis of the l-ching (Book of changes), at the same time being an adherent of the Taoist religion. Although he lived as a hermit, Wang Kuei's reputation as a sage must have been great because he had been summoned to the Chin court as long before as 1190. He had refused and had again refused in 1215 when Hsiian-tsung had invited him to court and offered him high office. Instead he became the chief adviser of Wan-nu, whom he continued to serve until he was well over ninety years old. Wan-nu saw no chance of regaining the plains of central Manchuria, which were at that time firmly held by Yeh-lii Liu-ko in alliance with the Mongols, and so he turned eastward and to the north. His state covered the eastern, forested, and mountainous part of Manchuria and also included the region of the former Supreme Capital on the Sungari. Wan-nu's territories thus bordered on Koryo, and he would certainly have liked to extend his domination in that direction, but his invasions into Koryo led to no lasting results. The state of Ta-chen existed for over eighteen years until the Mongols, during their campaign against Koryo, advanced against Wan-nu's strongholds and took him prisoner in 1233. Wan-nu's political role can be compared with Cambridge Histories Online © Cambridge University Press, 2008 THE ANNIHILATION OF CHIN 259 that of the insurgent Li Ch'iian in Shantung: Both established themselves in border regions far away from the center of the Chin state, and both tried to remain independent of the advancing Mongols with whom, however, they sometimes nominally allied themselves. For the Chin state the loss of Manchuria, first to Liu-ko and Wan-nu and subsequently to the Mongols, was a severe blow because it cut off the remains of their state in China from their main horse- and cattle-breeding areas and from those regions with a substantial Jurchen population, on whose loyalty they could have relied. As the situation was in 1215, Chin had lost not only the grain surplus—producing areas of northern Hopei but also the regions from which they had obtained a great number of their cavalry horses. It is surprising that despite these formidable, and indeed fatal, losses, Chin was still able to survive as a state for some years. One reason was certainly that from 1219 onward Chinggis khan directed the greater part of his forces westward in order to attack western Asia; another factor may well have been the fear of the Mongols that united loyal Jurchens and Chinese against a common foe. T H E A N N I H I L A T I O N OF C H I N , 1215-1234 The events of 1215 had reduced the Chin territories to the region around the Yellow River and transformed it into a buffer state hemmed in by the Mongols, Hsi Hsia, Li Ch'iian and his Red Coats in Shantung, and, of course, the Sung in the south. Although the strategic situation seemed hopeless, the Chin court in K'ai-feng decided to attempt to compensate for its losses in the north by means of a southern campaign against the Sung. In 1217 it decided to attack the Sung on the Huai River front, but the Chin troops could not advance as deeply into Sung territory as they had been able to do in 1206-7. At the same time Hsi Hsia attacked the western borders of Chin; here, however, the Chin were able to push back the invaders. There followed a confused series of battles for border towns in the Huai region, with no decisive results. Repeatedly the Chin sued the Sung for peace (which always implied the demand for continued payments), but in 1218 the Sung did not even allow the Chin envoys to enter their territory. Another attempt to invade Sung followed, this time with some tactical gains but no strategic success. In the meantime the Mongols, this time under Mukhali (d. 1223), one of Chinggis khan's most able and trusted generals, maintained their relentless pressure, chiefly in Shansi, where the strong strategic fortress-city of T'aiyiian was taken (see Map 21). Hsiian-tsung seems to have planned further Cambridge Histories Online © Cambridge University Press, 2008 ' Campaign in Shansi, 1218 1 Campaign in Shansi, 1219 Campaigns in Hopei and Shantung 1220 Campaigns in Shansi andShensi 1221-23 .•••" 1221x ;" Feng-chou Mongol forces fail to take cities 300 km ~~2<Somiles MAP 2 i . Mukhali's campaigns against Chin Cambridge Histories Online © Cambridge University Press, 2008 THE ANNIHILATION OF CHIN 261 actions against Sung and therefore tried to negotiate for peace with the Mongols. In 1220 the minister Wu-ku-sun Chung-tuan was sent on an embassy to Chinggis khan, who was encamped at that time in Transoxania, and offered to recognize the Mongolian khan as his elder brother in return for a cessation of hostilities. This attempt to include the Mongolian ruler into the network of pseudofamilial relations that had existed among the states of continental East Asia since the tenth and eleventh centuries failed. A second embassy of the Jurchen grandee was equally unsuccessful. This time, Chinggis khan recommended to the Chin representative that Hsiian-tsung renounce his imperial rank and instead become king of Honan (Ho-nan wang) under Mongolian suzerainty. Chin, however, rejected the offer to be invested with an inferior rank by the Mongols, and the peace talks therefore came to an end in 1222. In 1223 Hsiian-tsung died and his third son, Ning-chia-su (b. 1198; Chinese names Shou-li and Shou-hsii; r. 1223-34) succeeded him. He was the last ruler of Chin and was later canonized as Ai-tsung, "Pitiable Ancestor. " The ten years of his rule saw the final collapse of the Chin state and Jurchen rule. When Ai-tsung assumed the throne, his government had lost control over practically all the territories north of the Yellow River. Apart from Honan, the former empire of Chin consisted only of parts of Shantung and Shansi and the province of Shensi. After Mukhali's death, the Mongolian attacks and raids lost some of their previous vigor while Chinggis khan himself was engaged in the west. One of Ai-tsung's first actions was to make peace with Sung (1224). Chin formally gave up the claim to the annual payments, and Sung agreed to a cession of hostilities. The ceremonial embassies for the New Year and the rulers' birthdays were discontinued. This meant the end of the normal diplomatic intercourse that had governed Sung-Chin relations for almost a century with occasional interruptions (1160-5 a n d 1206-8). With regard to Hsi-hsia, Aitsung favored reconciliation after a period of constant border warfare, military actions that had sometimes been carried out with the Mongols' assistance. In 1224, negotiations with Hsi Hsia were initiated, and in the ninth month of 1225 a peace treaty was concluded. The Hsia ruler was acknowledged as the younger brother of the Chin emperor; both states also agreed to use their own reign titles in diplomatic correspondence, which resulted in a rise in status for Hsi Hsia because they were no longer considered to be vassals of Chin. The border trade was also resumed, a vital matter for the Chin because their cavalry had to rely largely on the import of Tangut horses now that the grazing grounds of Manchuria had been lost to them. The willingness of the Tanguts to cease their incursions into Chin territory on the Shensi border was Cambridge Histories Online © Cambridge University Press, 2008 262 THE CHIN DYNASTY TOLUI'S ARMY 1230 OGODEIS AIN ARMY 1230-31 Sung Emperor leaves K'ai-feng for Kuei-te 1223,ii. Emperor retreats to Ts'ai-chou 1233,viii, Siege 1233,xii. City falls 1234, ii. M A P 2 2 . The destruction of Chin, 1234 certainly motivated by renewed Mongolian attacks against their own kingdom. The Chin, on their part, had given up all hopes for an expansionist policy and were content to stabilize their state within its existing borders. They even achieved some local victories over the Red Coats in Shantung. In 1227 Chinggis khan died while his campaign against Hsi Hsia was still in progress. Ai-tsung tried to appease the Mongols by sending an embassy offering formal condolence, but the Mongols refused to receive the envoys in their camp. Already in 1226, diplomatic relations between Hsi Hsia and Chin had ceased; the last embassy from the Tangut court arrived in the Chin capital on 6 November 1226 and announced the death of the Tangut ruler. Chin dutifully dispatched a mission of condolence four weeks later, but the Mongolian attack against Hsi Hsia prevented its entry into Tangut territory. After the annihilation of Hsi Hsia in 1227 and the death of Chinggis khan on 25 August 1227, the Chin enjoyed a brief period of respite from the Mongols. The new great khan Ogodei resumed his operations to eliminate the Chin who, despite their tremendous setbacks, were still able to resistfiercely(see Map 22). In 1230 and 1231 Ogodei organized a large-scale campaign against Cambridge Histories Online © Cambridge University Press, 2008 THE ANNIHILATION OF CHIN 263 the Chin capital of Pien (K'ai-feng). Two columns were set in march, one under Ogodei's command in Shansi and the other under Chinggis khan's youngest son Tolui in Shensi. The strategic aim was a pincer attack on K'aifeng from the north and south. When both armies met in the winter of 1231-2, they were put under the command of Siibetei, a distinguished warrior whose forces ten years later would spread terror in Galicia and Hungary. Although the Chin command had sent thirty thousand soldiers to protect the northern flank of the capital at the Yellow River fords, on 28 January the Mongols were able to cross the river, and on 6 February the first Mongolian cavalrymen appeared under the walls of the capital. The Chin court tried feverishly to mobilize all able-bodied males in the capital and to organize resistance against the attackers, who began their siege operations on 8 April 1232, about two weeks after they had asked for formal surrender and hostages. Throughout these weeks the Chin government had tried desperately to come to terms with the Mongols, and several further peace talks took place during the summer of 1232. These came to a definitive end when on 24 July two Chin officers murdered the Mongolian envoy T'ang Ch'ing in his hostel, together with some thirty other people. After this act of treachery the Mongolian attacks were renewed with increased energy. The situation in the besieged capital became chaotic and hopeless, particularly after the outbreak of an epidemic in the summer of 1232. The provisions stored for emergencies soon proved inadequate, and despite ruthless requisitions of food among the population, the capital suffered from severe famine. A graphic description of life in the capital during the siege has survived; it is an eyewitness account by a Chinese intellectual who had held offices under the Chin. 20 His moving account gives evidence of the total disorganization within the government. Nominations, promotions, demotions, and executions of suspected traitors followed one another ceaselessly. On the other hand, it is surprising that the city could be defended at all, for it seems that the Jurchen and Chinese soldiers were able to put up an effective defense against the Mongolian forces and their Chinese allies. The siege of K'ai-feng is also of some interest for the history of military technology, because gunpowder was used by both parties, if not as a propellant for projectiles, then certainly for grenades hurled by catapults. These bombs were used by the defenders of K'ai-feng against men and horses, with deadly results. Another weapon credited to the inventiveness of Chinese artisans was a flamethrower (or rocket?), called a "fire lance." Sixteen layers of strong yellow paper were pasted together and formed into a pipe over sixty centime20 Liu Ch'i, comp., Kuei ch'ien chih, 11. This has been translated into German by Erich Haenisch in Zum Untergattg zweier Rticht: Berichit von Augenztugtn am dm Jahrtn 1232—33 und 1368-70 (Wiesbaden, •969). PP- 7-26. Cambridge Histories Online © Cambridge University Press, 2008 264 THE CHIN DYNASTY ters long. This pipe was filled with a mixture of charcoal made from willow wood, iron filings, powdered porcelain, sulphur, and niter and was fastened to a lance. The soldiers handling these weapons carried a small iron box with glowing embers and, in battle, ignited the fire lances, which ejected a flame over three meters long. When the explosives were spent, the pipes could be reloaded." In the winter the emperor decided to leave the city while it was still possible. Followed by a host of loyal Jurchen and Chinese officials, he left for Kuei-te in Honan where he arrived on 26 February 1233, and later in that summer, on 3 August 1233, he found refuge in Ts'ai-chou. The capital was thus left in the hands of the commanding generals. One of these was Ts'ui Li. He planned to avert the worst for the capital and for himself by preparing to surrender, because if K'ai-feng had been taken by storm, indiscriminate slaughter and pillage would have resulted. The officials and generals who were still loyal to the absent emperor were eliminated, and on 29 May the city gates were opened to the soldiers of Subetei. The capital was plundered in a "normal" way, but it seems that soon barter trade between the inhabitants and the northerners was permitted; the townspeople gave their last possessions, valuables, and silk in exchange for rice and grain transported from the north. Some slaughter occurred nevertheless. Over five hundred male members of the Wan-yen clan were marched out of the city and massacred. Ts'ui Li, who might have cherished hopes for a high position in the Sino-Mongolian hierarchy, did not enjoy the fruits of his coup, as he was assassinated by a Chin officer whose wife he had allegedly insulted. The fall of K'ai-feng still left the Mongols to administer the final coup de grace to the remnants of the Chin imperial court. Ai-tsung's situation was so desperate that envoys were sent to the Sung to ask them for grain and to point out that the Mongols were a great danger and that they would destroy the Sung in their turn. The Sung commanders of course refused any assistance and continued to prepare a joint attack with the Mongols against the last Chin strongholds. But even so, the prefectural town of Ts'ai-chou held out for some time after the attacks began in December 1233. After an unsuccessful attempt to flee from the town, Ai-tsung ceded his "throne" to a distant relative and committed suicide. This man, too, fell in the street fighting when Mongolian soldiersfinallyentered the town on 9 February 1234." The Chin state and the 21 CS, 116, p. 2548. For the bombs or grenades, see CS, 113, p. 2 4 9 5 - 6 . For a more recent study, see Jixing Pan, "On the origin of rockets," Toungpao, 73 (1987), pp. 2 - 1 5 . 22 The account of the events in Ts'ai-chou in the Chin sbib is largely based on the Ju-nan i sbib, a text written by an eyewitness, Wang E. On the author, who lived from 1190 to 1273, see Hok-lam Chan, "Prolegomena to the Ju-nan i sbib: A memoir of the last Chin court under the Mongol siege of 1234," Sung Studies Newsletter, 10, suppl. 1 (1974), pp. 2 - 1 9 ; and Hok-lam Chan, "Wang 0 ( 1 1 9 0 - 1 2 7 3 ) , " Papers an Far Eastern History, 12 (1975), pp. 4 3 - 7 0 . Cambridge Histories Online © Cambridge University Press, 2008 GOVERNMENT STRUCTURE 265 Wan-yen dynasty had come to an end. The Sung, at last, had their revenge, but as the envoy from Chin had warned them, they had won as a neighbor an even more formidable adversary than the Jurchens. The end of a dynasty has always been a favorite topic for Chinese historians and historical philosophers. They usually try to explain the fall of a state by deviations from the moral principles embodied in the ethical code of Confucianism and so do not have many good things to say about the Chin state. But even orthodox historians would have to admit that the cardinal virtue of loyalty was alive during the last stages of Chin. There were defectors and opportunists, but a surprisingly great number of leaders and soldiers, Jurchen and Chinese alike, remained faithful to the bitter end. GOVERNMENT STRUCTURE The rise, decay, and fall of the Chin dynasty are to a great extent linked with the history of their institutions. The basic feature of their government and administrative system was the complex interplay between native Jurchen traditions, features inherited from the Liao state and Chinese (Sung) influence. Further institutional changes took place during the dynasty as a result of the exigencies of the political situation. The interplay of these factors make the institutional history of Chin a complicated field of research, so that only the broad outlines of development can be given here. From tribal council to a Chinese-style government In their predynastic and early dynastic periods, the early ways of government among the Jurchen people were very different from the highly hierarchical and departmentalized Chinese bureaucracy. The following is a description of Jurchen political structure in the period before the founding of the Chin state: "They have no overlord and no name for their state. They live scattered in the mountains and valleys and elect for themselves a brave and valiant person as chieftain. A smaller chieftain has one thousand households, a great one several thousand households."23 Even long after A-ku-ta and his predecessors had won supremacy and allied the sometimes-recalcitrant Jurchen tribes under unified leadership, many of these features dating back to the tribal period still could be found, particularly in military matters: From the commanding general down to the soldiers, everybody managed cases of disobedience on his own. They all had millet gruel and roast meat for food, and there was no difference in 23 Peifengyangshalu, in Shuofu, ed. T'aoTsung-i (Taipei, 1963), 25, p. 24b. Cambridge Histories Online © Cambridge University Press, 2008 266 THE C H I N DYNASTY quality between high and low. When their country is involved in great affairs [war], they all go out into the wilderness and sit down in a circle, drawing in the ashes. Then they deliberate, starting from the lowest one present. When the council has come to an end, they wash away [the charcoal], and not a human voice is heard - such is their secrecy. When the army is about to march, a great reunion with a banquet is held, at which strategic proposals are offered. The generalissimo listens and then selects among these what is appropriate; then immediately a special leader is appointed for its execution. When the army returns after a victory, another great reunion takes place, and it is asked who has won merits. According to the degree of merit, gold is handed out; it is raised and shown to the multitude. If they think the reward too small, it will be increased.24 It was a long time before the vestiges of these semiegalitarian customs disappeared. A-ku-ta, for example, did not expect his officials to kowtow before him. The gulf that existed in Chinese hierarchical thinking between an emperor and his subjects was unknown under the early Chin rulers, and the growing autocracy under Hsi-tsung and Hai-ling wang was, in a certain respect, nothing but an adoption of Chinese ways. Even as late as 1197, when the state structure has been patterned after the Chinese model, we find a curious case of imitation of the old tribal council customs. The question was whether or not the Mongols should be attacked. A vote was taken among the highest officials, and the court historians have duly recorded the outcome of this poll: Out of eighty-four, only five favored an attack; forty-six were for a defensive strategy, and the rest preferred alternating between attack and defense.2' On the other hand, some sort of central control became imperative as soon as the range of government actions was enlarged, through diplomatic contacts but chiefly through the acquisition of new territories. A-ku-ta therefore created a system of what may be called prime ministers. These were called pochi-lieh in Chinese; the Jurchen word was something like bogile. It survived in the Manchurian language as beile and was used by the Manchus until the beginning of the twentieth century as the designation for a high dignitary. The original meaning of bogile seems to have been "leader, chief," and it was already in use in the predynastic period, because A-ku-ta was elected supreme bogile in 1113 when he succeeded his brother. The assumption of this title supplanted the honorary designation as military governor, which had customarily been conferred on him by the Liao rulers, and illustrates the high prestige attached to the rank of bogile. Another proof of this prestige is the fact that only close relatives of the emperor from the Wan-yen clan bore this title. Usually the various bogile titles established in 1115 were prefixed by the word gurun (Chinese: kuo-lun), "state, nation." 24 Pa fmg yang sha lu, 25, p. 25b. Sec also Ta Chin kuo cbib, 36, pp. 2 7 8 - 9 , for a brief summary of the military practices of the early Jurchens. 25 CS, 10, p. 242. Cambridge Histories Online © Cambridge University Press, 2008 GOVERNMENT STRUCTURE 267 The highest office was that of the "great leader," held by the presumptive heir to the throne. Other bogile were called "commanding leader," "first leader," "second leader," "third leader," and "assistant leader"; these meanings have been transmitted through both the Jurchen words (transcribed with Chinese characters) and Chinese glosses. The rank of assistant leader was not as high as the others and was mostly conferred on generals during campaigns. The description of the various bogile offices in the contemporary sources indicates that there was already some degree of functional differentiation. The commanding leader was the head of political affairs in general, with the second and third leaders as his deputies. There was also a bogile whose chief duties were in the diplomatic field, namely, the i-shih bogile (the first part of this word is still unexplained). Although this differentiation may be regarded as the beginning of a specialized bureaucracy (all the bogile had their own subordinate staff), it would be erroneous to regard these titles as offices in the strict sense. They were far more a distinction conferred ad personam, because some bogile offices were abolished when their incumbents died. Many changes were made in the bogile system, which in its later stages showed Chinese influence even in its nomenclature, until it was abolished altogether shortly before T'ai-tsung died ("34-5)By that time, the Jurchens had expanded their domination not only over the former territory of the Liao state but also over great parts of northern China, chiefly Hopei and Honan. They were thus faced with the problem of how to rule a state comprising many different ethnic groups with different economic and social backgrounds. In terms of numbers the Chinese certainly were a majority, including both the Chinese inhabitants of the former Liao state and those in the newly conquered regions. At first the Jurchens followed the example set by the Khitans under the Liao dynasty, in which a marked dualism had existed: The Khitans and their related tribes continued to be administered by their own tribal organization, whereas the Chinese were subjected to an administration modeled on Chinese patterns, largely inherited from the T'ang dynasty. A similar dualism was practiced after the conquest of the Central Plains. The Jurchen people were organized into units of their own (the meng-an mou-k'o; see the next section). For the administration of the newly conquered regions in China proper, a new authority was created in 1137, named the Mobile Presidential Council (Hsing-t'ai shang-shu sheng). This organization existed from n 37 to 1150 and was revived as a measure of military expediency after 1200. The Mongolian Yuan dynasty later took it over from Chin and transformed it into a full system of provincial administration. The term sheng for province, which now forms part of the administrative system of the Chinese Cambridge Histories Online © Cambridge University Press, 2008 268 THE CHIN DYNASTY People's Republic, thus goes back to a Chin institution that had lasted through the Yuan, Ming, and Ch'ing dynasties into the Republican period after 1911. The Chin name Mobile Presidential Council indicated that it was originally mobile, that is, not located a priori in a definite town, as all Chinese administrative units usually were, but assigned to whatever place seemed politically preferable. At the same time, the organization was not conceived as an independent unit. It was a branch office of the Presidential Council (Shang-shu sheng) and thus subordinated to a metropolitan agency. The transition from the rule of Jurchen generals over the newly conquered regions and their population to a more centralized type of administration was carried a great step forward with the creation of this instrument of centralized control. One of the many duties of the organization was the recruitment of bureaucratic personnel through civil service examinations. The dissolution of the state of Great Ch'i in 1136—7 opened the way for former civil servants of Ch'i to enter the new bureaucracy of Chin. The leading positions remained, however, a prerogative of Jurchen dignitaries. The same is true for the Presidential Council itself. It had been established as early as 1126 in the Supreme Capital in Manchuria at a time when the campaign against Sung was still in full swing. It soon grew into a fullfledged prime minister's office and remained the chief policymaking agency under the Chin. The name of the council itself as well as those of its various subordinate offices were Chinese. The leading personnel were mostly members of the imperial clan and other Jurchen dignitaries. In later years some Khitans, Hsi, and a few Chinese and Po-hai also rose to these most powerful ranks in the bureaucracy. The highest office in the Presidential Council was that of left prime minister (tso ch'eng-hsiang). Out of the sixteen persons who held that office over the years, no fewer than eleven came from the imperial Wan-yen clan; four were from other Jurchen clans; and one was a Po-hai. The office of right prime minister was held consecutively by five imperial clansmen, two other Jurchens, two Po-hai, three Khitans, and two Chinese. The lower echelons of the Presidential Council showed a larger proportion of Khitans and Chinese.26 The preponderance of imperial clansmen is interesting because it contrasts with the usage under Chinese dynasties, in which under both T'ang and Sung, members of the imperial clan were rarely if ever promoted to senior ministerial posts. Clan affiliation was, therefore, for the Jurchen state of Chin, a far more powerful guarantee of loyalty than was the abstract behavioral code of Confu26 Mikami Tsugio, Kindai uiji seido no kenkyu, Kinshi kenkyu no. 2 (Tokyo, 1970), p. 217, provides a table of the highest offices broken down on the basis of the nationality of the incumbents. Cambridge Histories Online © Cambridge University Press, 2008 GOVERNMENT STRUCTURE 269 cian state ethics. In regard to the political machinery of the Chin state, the age-old antagonism between aristocracy and bureaucracy in China was decidedly resolved in favor of the aristocracy and tribal connections, at least among the highest echelons of executive power. There were two more councils at the highest level, the Secretarial Council (Chung-shu sheng) and the Court Council (Men-hsia sheng), but these two always remained overshadowed by the Presidential Council. For some time there was also a superchancellor's office, the director of the three councils {ling san-shengshih), as part of the hierarchy, but this office was abolished under Hailing wang in 1156 together with the Secretarial Council and the Court Council. After that date, only the Presidential Council remained. All these offices had a deep influence on the institutions of the Mongolian Yuan dynasty, in which the Secretarial Council became the highest political decision-making agency, whereas the Presidential Council existed only temporarily. Another central bureau that had a long tradition under Chinese dynasties was the Censorate (Yii-shih t'ai). It was established in 1138 and continued to exist until the end of the dynasty. The Censorate increased in importance under Hai-ling wang and Shih-tsung and was enlarged and raised in status in 1172 and again in 1181. There were also changes introduced in the reign of Chang-tsung, all tending to enhance the position of the Censorate as a body for policy criticism and supervision of all branches of the bureaucracy. It is notable that Shih-tsung in particular preferred holders of the chin-shih degree as members of the Censorate because he considered them as better suited to the delicate tasks of their office. Under the Presidential Council there existed from the 1140s onward the classical Six Boards, which had been an integral part of the Chinese central bureaucracy for centuries: the ministries of Revenue, Civil Appointments, Rites, Works, Punishments, and War. Their organizational structure and functional differentiation followed the Chinese model prevailing under the T'ang and Sung (and also the Liao) and therefore need not be described in more detail. Another central agency to be mentioned here was the Chancellery (Shu-mi yuan). It had been created in 1123 and, during the campaigns against the Sung, was transferred to the south and for some time was responsible for the affairs of the Chinese population such as taxation, the recruitment of personnel, and labor service. The Chancellery later developed into a kind of imperial general staff and constituted the highest central military authority. In this respect it resembles its Sung counterpart, but in contrast with the Sung government structure, the Chin Chancellery always remained subordinate to the Presidential Council. The proliferation of offices was greatest under Hai-ling wang, who did so much to transform the Chin state from a tribal and aristocratic body politic Cambridge Histories Online © Cambridge University Press, 2008 270 THE CHIN DYNASTY into a Chinese bureaucracy. Toward the end of the twelfth century there was hardly a central government office that did not have its counterpart in the Sung state. The names were perhaps different, but the functions were the same. This is also true for such eminently traditional Chinese offices as those for astronomy and astrology, national historiography, and the many bureaus and departments dealing with the administration of the imperial household and ceremonial affairs. In one respect, however, the Chin dynasty was a faithful follower of LiaoKhitan (and Po-hai) precedents. Unlike a proper Chinese dynasty, which normally had one capital, the Liao had five capitals, as did the Chin. In both cases this can be interpreted as a remnant of the times when even the rulers had no fixed abode, but it was also a remnant of a ritualized system of seasonal sojourns. On a more practical level, the system of multiple capitals also provided the means to establish centralized agencies in more than one locality. The five-capital system of the Chin is particularly complicated because names like Southern or Central capital were given to different towns in different periods (see Table 4). The shift of the main centers of power is clearly reflected in the transfer of names. Yen-ching (Peking) was the Southern Capital until Hai-ling wang made it the center of the state. Thereafter it became the Central Capital. After Peking had been abandoned to the Mongols, Lo-yang became the Central Capital. Local government in the predominantly Chinese parts of the state closely followed the Chinese models as developed under the Sung and T'ang dynasties, and it is therefore without characteristics peculiar to the Chin government system. Counties (hsien) and prefectures (fu or chou) were the basis of local administration, and they were administered more or less as they were in contemporary Sung China. The next-higher administrations, corresponding to provinces, were the routes (/«), of which there were nineteen. The only difference between Sung and Chin administration on the local and provincial level was in the partly military, partly tribal organization of the border areas. These will be described briefly in the section on the military organization of the Chin. Selection ofpersonnel Even the short outline of Chin government structure in the preceding chapter showed that the bureaucratic framework of the state called for a great number of officials. For the later periods of Chin we havefiguresthat give some idea of the bureaucracy's numerical strength. In 1193 there were 11,499 officials, of whom 4,705 were Jurchens and 6,794 w^re Chinese. Thisfigureis said to have Cambridge Histories Online © Cambridge University Press, 2008 GOVERNMENT STRUCTURE 271 risen in 1207 to a total of 47,000. It seems, therefore, that the Chin had at least as many officials as the Sung had in their early northern period (12,700 in 1046).27 How, then, were these great numbers of officials recruited? As the Liao had done before them, the Chin practiced a dualistic recruitment policy. They established an examination system on Chinese lines, selecting candidates on the basis of merit. At the same time the selection and promotion of personnel was handled according to segregationist principles based on the individual's group affiliation or personal position. Such institutions as the operation of the principle of protection (yin), hereditary offices, and the transfer of military personnel to the civilian bureaucracy played a major role in recruitment. In both examinations and the principle of favored social groups the Chin state tried to find means to secure the preponderance of Jurchen personnel. At the beginning of the dynasty, former Liao officials were simply incorporated into the Chin administration when the Jurchens invaded their territory, and the regularization of recruitment policy evolved slowly. The Chin examination system began in 1123 when the first examinations were held. From 1129 onward, triennial examinations for the chin-shih degree were held; later they were held annually. Originally the examinations given in the north differed from those given in the south (the newly incorporated Sung and Ch'i territories). The northern examination subjects concentrated on poetry and prose literature (which was regarded as easier); those in the south, on the Chinese classics. One reason for this regional differentiation may have been the desire to make the examinations easier for the northern Chinese, who, as former subjects of Liao, were perhaps regarded as more reliable than the southerners. The examinations on the classics were abolished for some time but then were revived and reorganized in 1188—90. In addition to the Five Classics (the books of Changes, Rites, Songs, and Documents and the Spring and Autumn ~nnals), the texts used were the Analects of Confucius, the Meng-tzu, the Hsiao ching, the Yang-tzu (Yang Hsiung's Fa-yen), and the Taoist classic Tao-te ching. Although the positions with executive power, particularly at the highest levels, were mostly held by Jurchens, Chinese received an important means of access to the bureaucracy through the chin-shih examinations. Over time, more and more of the high-ranking Chinese were men who had attained their ranks through their status as chin-shih graduates, rather than through other means such as conferment of office adpersonam or military achievements. The non-Chinese and non-Jurchen elements (Khitans, Hsi, and Po-hai) did not 27 For the Sung figure, see Edward A. Kracke, Jr., Civil service in early Sung China, 960-1067, Harvard— Yenching Institute Monograph Series no. 13 (Cambridge, Mass., 1953), p. 55. The Chin figures are recorded in CS, 55, p. 1216. Cambridge Histories Online © Cambridge University Press, 2008 272 THE CHIN DYNASTY play a significant role in the examinations. Only five Po-hai and one Khitan seem to have obtained the chin-shib degree during the whole history of Chin. Emperor Shih-tsung must have realized that whatever its shortcomings, the examination system provided the government with dependable servants. He therefore opened a new career for Jurchens by establishing in 1173 a Jurchen chin-shih degree (until this date there had been no degree requirement for Jurchen officials) and encouraged his countrymen to take the degree. The object of introducing a separate Jurchen degree may have been twofold: It was in harmony with the general attitude of Shih-tsung, who wished to preserve the Jurchen language and customs, and it was perhaps a calculated measure to bring more Jurchen commoners into the bureaucracy in place of the sometimes-overbearing and recalcitrant Jurchen aristocrats. But unlike the Chinese, who eagerly seized the opportunities offered them through the examinations, the Jurchens as a whole continued to advance their careers without degree qualifications. Only 26 out of 208 high-ranking Jurchen officials were chin-sbih degree holders. For them, their group privileges and hereditary privileges remained the chief means of entering and rising in the hierarchy. The protection privilege (yin) was an important prerogative of those who already held office and contributed to the self-perpetuation of the officials as a class. In the beginning and until the reign of Shih-tsung there was no limitation on the number of family members whom the higher officials from the seventh rank upward could nominate and "protect." Then a gradation was introduced that fixed a maximum of six proteges for officials with the highest (first) rank, proportionately fewer for those holding lower ranks, and that denied the privilege altogether for those of the eighth rank and below. This rule favored, of course, the high officials, the majority of whom were Jurchens. Hereditary selection as practiced under the Chin had similar effects; for example, Jurchens from the imperial Wan-yen clan had the privilege of entering the palace service without formal protection. Jurchen commoners were also recruited for palace and palace guard service and from such beginnings could make an official career. This is a clear parallel to the Mongolian system of imperial guards (kesig). Also, the Jurchen meng-an mou-k'o system (see Chapter 4), with its hereditary offices, was a form of hereditary selection based on group status. Last, the practice of transferring meritorious military leaders to the civil bureaucracy also favored the Jurchen elements of the population, as the military organization remained very much a Jurchen preserve during most of the dynasty. Discrimination was not confined to recruitment. After they had joined the ranks of officialdom, the Chinese normally rose to higher ranks more slowly than did their Jurchen counterparts. Promotion was formalized Cambridge Histories Online © Cambridge University Press, 2008 GOVERNMENT STRUCTURE 273 and based on both individual merit and seniority. Merits were calculated according to a complex rating system, in an attempt to achieve objectivity. The selection and promotion of personnel under the Chin dynasty therefore show many dualistic features. But we should emphasize that there was no monopoly of office for the Jurchens, nor was there any general war on the recruitment of Chinese. Rather, the Chin state aimed at a compromise and tried to shape its institutions for recruitment in a way that could be used to balance the influence of the different groups that made up its population. The adoption of the civil service examinations for the Chinese, while at the same time adding some checks and maintaining preferential methods of promotion for the Jurchen, certainly helped bring about social stability. It also is certain that examinations played a greater role in personnel selection under the Chin than under the two other dynasties of conquest, Liao and Yuan. 28 Military organization: meng-an mou-k'o and border administration The meng-an mou-k'o system was a sociomilitary organization typical of the Jurchens. It has been much studied, not only because of its inherent interest, but also because it was in many respects a precursor of the Manchu banner (niru) system that the Manchus used for establishing military control after their conquest of China in the seventeenth century.29 The Chinese syllables meng-an mou-k'o are a transcription of two Jurchen words: Meng-an means "thousand" and is a loan word from Mongolian (mingghan; Manchu: minggan). Originally the leader of one thousand men in war was called a meng-an; later the word also came to designate the unit under his command. Mou-k'o is explained by the Chin shih as the leader of one hundred men. The word is not, however, a numeral but is related to the Manchu word mukun, for which the dictionaries give the meanings "clan, family, village, herd, tribe," and so on. The meng-an mou-k'o system was based on the tribal divisions of the Jurchens and was not a purely military organization but a comprehensive social system into which in principle the entire Jurchen population was organized under A-ku-ta. It soon became the most important military and political means of control over the subjugated population. The basic unit was the mou-k'o. The number of households attached to one unit varied. In theory it should have been three hundred households but in reality was frequently smaller. Nor did the meng-an have one thousand households, as the name implies. Normally, a meng-an was composed of seven to ten mou-k'o. 28 For a detailed analysis of the Chin recruitment system, see Tao Jing-shen, "The influence of Jurchen rule on Chinese political institutions," Journal of Asian Studies, 30(1970), pp. 121—30. 29 On the meng-an mou-k'o system, see Mikami Tsugio, Kinshi kinkyu, vol. 1, pp. 109—417. Cambridge Histories Online © Cambridge University Press, 2008 274 THE CHIN DYNASTY The mou-k'o itself was subdivided into p'u-li-yen (there also are several other transcriptions of this word), a word probably related to Manchu feniyen, "flock, herd." P'u-li-yen, like the other words, designated both the unit and the title of its leader. The p'u-li-yen had fifty households under his command. All able-bodied males in a household had to serve as soldiers. Male household servants were also conscripted and served as auxiliary soldiers (a-li-bsi; cf. Manchu ilhi, "subordinate, assistant"). Each fully equipped soldier was entitled to be accompanied by an auxiliary soldier during a campaign. The mou-k'o in the Manchurian homelands of the Jurchens settled in and around stockaded villages, and most bore the name of their original geographic location, which they usually retained even after they had been transferred to other parts of the state. The formal introduction of this system is said to have taken place under A-ku-ta in 1114, but it certainly goes back to much earlier times. It underwent many changes in the following years. When the Jurchens conquered the Liao state, they used the meng-an mou-k'o system to organize those Khitan, Hsi, Chinese, and Po-hai who had surrendered. Hereditary office in the system was a considerable inducement for Khitan leaders to join the Jurchens with their subordinates. The number of households in a Khitan mou-k'o unit was, however, smaller than among the Jurchens and amounted only to 130. We do not know how many households were normally attached to a Po-hai or a Chinese mou-k'o. In one case at least, a Chinese mou-k'o consisted of only 65 households.30 The formation of new Chinese units came to a stop in 1124, but the number of Chinese serving in the Chin army must already have been quite considerable, for during the campaign against Sung in 1126—7, several Chinese divisions each of ten thousand men fought against their countrymen under Jurchen command. It is not quite clear how many of these were simply conscripts drafted for the campaign and how many were part of regular Chinese meng-an mou-k'o units. The number of soldiers outside the meng-an mou-k'o system usually varied according to the needs of the military situation. They were conscripted from the civilian population when an emergency arose and were disbanded when the campaign had come to an end. During the last years of the dynasty, when the meng-an mou-k'o system had seriously declined, the Chinese population, including even high officials and dignitaries, were subjected to ruthless conscription for military service. Hereditary office for Chinese and Po-hai unit commanders was abolished in 1145 but was retained for the Khitan and Hsi units. At the same time, the existing meng-an mou-k'o were graded into three classes. The first-class units 30 CS, 44, p. 993. Cambridge Histories Online © Cambridge University Press, 2008 GOVERNMENT STRUCTURE 275 were those commanded by imperial clan members; second-class units were commanded by other Jurchen; and third-class units were those composed of Khitans, Hsi, Chinese, and Po-hai. This attempt to give different status to units of different origin was, however, abolished under Hai-ling wang in 1150. This ruler, who, as we have seen, tried to curb the power of the Jurchen aristocracy, also ordered the transfer of those units still commanded by imperial clan members from the Supreme Capital to other towns in the southern parts of the state. The whole system suffered a severe blow when the mobilization against Sung was resisted by some Khitan and Hsi units. These were mostly stationed on the northwestern borders and had good reason to fear for the safety of their dependents if all their warriors were mobilized for the campaign, because Mongolian raids were a constant menace in the area. They rebelled in 1161. After the rebellion was crushed, many of their units were disbanded, and the households were distributed among the Jurchen units. Only those who had remained loyal were retained and accorded the privilege of hereditary office as before. Another factor impairing the efficiency of the entire system was economic. As the meng-an mou-k'o units were also administrative and economic units — similar perhaps to the military colonies under Chinese dynasties — they had been assigned land for agriculture and were supposed to be economically selfsufficient. Many Jurchens had only limited experience of farming and were unaccustomed to farming in a Chinese environment. Some of them hired Chinese to work on their land, and they themselves led an idle life, drinking excessively and neglecting their military skills. Some units were also given government land of poor quality. Unable to compete with the more skillful Chinese and exploited by usurious moneylenders, a great number of the Jurchen commoners in the units were reduced to poverty. They were exploited not only by Chinese but also by their own richer and more powerful countrymen, particularly by members of the imperial clan who had managed to acquire huge land hold ings at the expense of the less fortunate Jurchens and, of course, the Chinese. In sharp contrast with earlier times, when the meng-an mou-k'o warriors, chieftain and commoner alike, lived together like "fathers, sons and brothers"*1 and when a frugal way of life was normal, a deep gulf now arose between the rich and poor Jurchens. Emperor Shih-tsung showed great concern for the deteriorating conditions of his impoverished countrymen. Relief measures were undertaken; government grain was distributed to impoverished units; and agricultural techniques were promoted. Sumptuary regulations and laws against luxury were enacted to inhibit drinking and extrava31 Ta Chin kuo chih, 36, p. 278. Cambridge Histories Online © Cambridge University Press, 2008 276 THE CHIN DYNASTY gance, and orders for regular military exercises were given. Military colonists were moved from poor lands to better regions and there was an attempt to concentrate those Jurchens living scattered among the Chinese into more closed and compact groups. A general census of the whole men-an mou-k'o population was taken in 1183, in which not only the people but also their lands, cattle, and slaves were registered. The results showed such a gross disparity between rich and poor that Shih-tsung's government resorted to a redistribution of lands and confiscation of excessive landholdings. These measures temporarily improved the situation. To social historians, the census figures are of interest. Apart from the imperial clan, whose holdings were registered separately, the whole meng-an mou-k'o population was 6,158,636 persons living in 615,624 households. Of these persons, 4,812,669 were commoners (the majority of them Jurchen), and the rest were slaves attached to individual households. The number of meng-an was 202, that of mou-k'o 1,878.3* Under Shih-tsung's successors, the system apparently lost its effectiveness, and when the Mongols invaded, the Chin government had to rely more and more on conscripted troops. But up to the end the meng-an mou-k'o remained the basic organization of the Jurchen military machine. The emperor and the crown prince had their own mou-k'o. This imperial guard was called ho-cha mou-k'o {ho-cha is the transcription of a Jurchen word perhaps related to Manchu hasban, "protection, screen"). The members of this regiment, numbering several thousand men, were recruited from among the normal units. Candidates had to befivefeet, five inches tall and to pass a military test. Within the regiment there was a small elite unit, that of the "close attendants," numbering two hundred warriors. They alone had the privilege of bearing arms in the presence of the emperor. The members of this personal bodyguard had to be at least five feet, six inches tall. The higher command structure of the Chin armies was relatively simple. Several meng-an mou-k'o formed a wan-hu, literally "ten thousand households." The next-highest office was that of the chief commander (tu-t'ung). The commander in chief(/» yiian-shuai) acted as generalissimo, but this office was activated only in times of war. In many respects the higher military hierarchy of the Chin was modeled on that of their Liao predecessors. This is also true for the tribal units that had existed under the Liao and that had been taken over by the Chin, sometimes without even changing their names. These units were mostly stationed on the northwestern borders and consisted of Khitans, Hsi, or members of other tribes. Unlike the agricultural meng-an 32 For an analysis of the demographic aspects of therafng-rf/Jmoa-i'osystem, see Ping-ti Ho, "An estimate of the total population of Sung—Chin China," in ttutits Song in mmoriam ttitnne Baldzs, ed. Francoise Aubin, series i, no. I (Paris, 1970), pp. 33—45. Cambridge Histories Online © Cambridge University Press, 2008 SOCIAL STRUCTURE 277 mou-k'o of the Jurchens proper, these tribal groups were nomad cattle breeders, a fact that is, for example, reflected in the title commissioner of herds {ch'tin-mu shih) for some of these units. But like the Jurchen units they were both military units and self-sufficient socioeconomic communities. Under the Chin, there were altogether twelve of these commissioners of herds. Some of them had under their command former members of the imperial camp guards (ordo) of the Liao and their descendants, but there was also one commissioner commanding Jurchen tribesmen. It seems that these commissioners were organized formally at a rather late date, under Emperors Shihtsung and Chang-tsung, in connection with the preparations for defense against the Mongols. Another feature inherited from the Liao were the chiu units. These were originally detachments of frontier soldiers. Under the Chin there were nine chiu units, mostly stationed in northeastern Manchuria. Finally, there were eight special offices named tribal commanding prefects {tsu-pu chieh-tu shih), whose names suggest the partly Tangut, Mongolian, Khitan, and Hsi affiliations of their subordinate populations. They were stationed along the western and northwestern borders of the state and, like the other organizations, were military organizations for border defense. SOCIAL STRUCTURE It is strangely ironical that the Chin shih, although the official history of a "semibarbarian" state, has preserved much clearer information about the system of population control and census taking than most of the other Chinese dynastic histories have. 33 Even for the Sung we have, despite a wealth of specific data, no very clear picture of the definition of age groups and similar registration policies. From the relevant chapter in the Chin shih we have, however, unambiguous information about not only the age groups but also the methods by which the population was enumerated every three years. The population registers were based on enumerations made at the lowest level, that is, by village headmen and, for the meng-an mou-k'o population, by the stockade overseers. The number of headmen per village varied with the number of households; those with fewer than fifty households had only one headman, whereas larger villages with three hundred or more households had four. In the towns and cities the corresponding overseers were in charge of urban wards or quarters. At the beginning of a census year, these local overseers had to visit all households and list the names, ages, and sex of the family members. The figures obtained were added together and then sent 33 This point is stressed by Ho Ping-ti in ibid. Cambridge Histories Online © Cambridge University Press, 2008 278 THE CHIN DYNASTY TABLE 7 Total population of Chin Year Households Individuals Persons per household 1187 1190 1195 6,789,499 6,939,ooo 7,223,400 8,413,164 44,705,086 45.447.900 48,490,400 53.532,15' 6.59 6.55 6.71 6.33 1207 on to the higher authorities. They had to reach the Ministry of Revenue within three months after the census had begun. Contrary to the obscurity of age-group definitions under some other dynasties, the Chin had clearly defined age groups. Everybody between the ages of seventeen and sixty was considered an adult (ting); the physically handicapped or disabled were, however, not considered adults. The figures for three national censuses have been preserved in the Chin shih (1187, 1195, and 1207). They are of interest not only for the history of Chin but also because they must be used along with the corresponding Sung figures if we wish to estimate the total number of Chinese in the twelfth and thirteenth centuries (see Table 7). We do not havefiguresforthe earlier years of Chin, and so the population increase can only be computed over a period of twenty years. The average annual increment (rate of growth) was 0.9 percent, which must be considered normal if viewed against other periods of Chinese history (e.g., the average annual rate of increase was 0.87 percent between 1779 and 1794). The average number of individuals per household was somewhat higher under the Chin than under other dynasties, when it was normally betweenfiveand six. The larger households under the Chin dynasty certainly resulted from widespread slavery. The figures for the meng-an mou-k'o population in 1183, for example, show that their households averaged 7.8 persons and that the average number of slaves per household was not fewer than 2.18. The households of the imperial clan had a considerable slave population, over 163 per household. If all these factors are taken into account, it suggests that the size of the nuclear family as a unit of communal living and consumption was approximately the same as in other periods of Chinese history. The population of the Chin state was quite large. In the twelfth and thirteenth centuries, Chin alone, not to mention the Sung, was far more populous than any contemporary European country, with well over 53 million people. In 1207 the Chin territories supported almost as many people as did the whole of Tang China in 742. We do not know exactly how these Cambridge Histories Online © Cambridge University Press, 2008 ETHNIC GROUPS 279 many millions were distributed geographically, but the geographical monograph of the Chin shih gives the number of households for each prefecture. Unfortunately, this source does not say to which year these figures refer, but it may be assumed that it was shortly before the Mongolian conquests and the loss of Manchuria to P'u-hsien Wan-nu in 1215, because the total number of households is even larger than that of the 1207 census. The geographical distribution of the population over the area of the Chin empire is shown on Map 23. This distribution shows that almost a quarter of the entire Chin population lived in the Yellow River plains around K'ai-feng (modern Honan). Another region of great population density was eastern Shantung. The third most heavily peopled area was Peking and its surroundings. It also is clear that the Manchurian homelands of the Jurchens were only sparsely populated, although these low figures might be partly due to difficult communications and corresponding deficiencies in census taking in these remote districts. It is also remarkable how few people lived in the strategic area bordering on the Hsi Hsia state in present-day Kansu. By far the biggest city in the whole Chin empire was the Southern Capital, K'ai-feng, with 1,746,210 households in the metropolitan district. The second largest city was the Central Capital (Peking), with 225,592 households, whereas the old Supreme Capital (Huining) in Manchuria contained only 31,270 households. The Eastern Capital (Liao-yang) was only slightly larger, with 40,604 households. ETHNIC GROUPS Although we can get a rather clear picture of the distribution of the Chin population, at least for one single year, we know far less about the relative strength of the ethnic groups within the Chin state. No statistics show us the precise percentage of these groups even locally. The figures given for the meng-an mou-k'o population cannot be used for this purpose because these military units included not only the Jurchens but also other ethnic groups. Only a very rough estimate is possible. If in 1183 there were over 4.8 million free military colonists, it might perhaps be assumed that the greater majority, say 80 percent of them, were indeed Jurchens and the rest Khitans, Pohai, or Chinese, so that the Jurchen population must have been something like 4 million, considerably less than 10 percent of the total. Not all Jurchens could consider themselves superior to other ethnic groups. Those living in the military colonies were segregated from the surrounding Chinese population, and the privileges for Jurchens were most strongly felt in the bureaucracy, in which Jurchens not only occupied most of the important positions but also enjoyed speedier promotion than others did. Cambridge Histories Online © Cambridge University Press, 2008 1 dot represents 5000 households M A P 2 3 . Population distribution in Chin, ca. 1211 Cambridge Histories Online © Cambridge University Press, 2008 ETHNIC GROUPS 281 Intermarriage with Chinese and others was made legal as late as 1191, although it must have occurred quite frequently earlier. The official nationality policy of the Chin state, however, changed considerably over the years. In the early period of conquest, the Jurchens tried to impose their clothing and hairstyle on the Chinese. Like the Manchus in the seventeenth century who ordered the Chinese to adopt the queue as their hairstyle, their forefathers in 1126 also ordered the Chinese to change their clothing and hairstyle, an ordinance that was reinforced in 1129. But it does not seem that this rule was strictly observed, and under the pro-Chinese emperor Hai-ling wang the Chinese in Honan were allowed to wear their own costume. Emperor Shih-tsung reversed this policy, not in an attempt to convert the Chinese into Jurchens, but in order to preserve the Jurchens' national identity. Contrary to earlier state policy, many Jurchens in his time (1161—89) seem to have adopted Chinese ways of behavior and to have forgotten their national traditions, including their own language. Therefore the emperor prohibited the Jurchens from dressing in the Chinese way or from adopting Chinese family and personal names. The imperial princes who had been given Chinese names had to resume their Jurchen childhood names. Only the Jurchen language was to be spoken in the palace, and members of the imperial bodyguard who had forgotten Jurchen had to relearn it. Jurchen bards were ordered to perform before the emperor to keep alive the old traditions. National pride was the reason for another ordinance under Changtsung: In 1191 the Chinese were forbidden to refer to Jurchens as "barbarians" (fan). But in spite of all these well-meaning attempts to segregate the Jurchens and preserve their identity, more and more of them merged with the Chinese majority of their state. Only in the backwoods of Manchuria where a compact Jurchen population lived, did their language and customs remain alive. The national crisis and the repeated disasters after 1200 led to their further assimilation. Equal treatment of meritorious Khitan soldiers was prescribed in 1201, and differential treatment of non-Jurchen military colonists was abolished in 1215. One reason for this gradual disappearance of the Jurchen people's national characteristics was that the garrisons in which they lived were dispersed over the whole state. The fact that even under the Chin the Khitan people remained much more of a close ethnic entity can be attributed to the remote and marginal areas in which they lived, where they could maintain their traditional tribal ways of life. To the Jurchen minority not all Chinese seemed alike. They made a clear distinction between the "northerners," that is, the Chinese who had been former subjects of Liao, and the "southerners," the former Sung subjects in Honan and Shantung. This is evident from some of the addresses given by Shih-tsung, who was a shrewd observer. For him, the northerners were Cambridge Histories Online © Cambridge University Press, 2008 282 THE CHIN DYNASTY unreliable and cunning and always adapting themselves to circumstances, whereas he regarded the southerners as straightforward and honest. "Among the people in the south there are many who are unyielding, dare to speak up and offer honest criticism."34 This is a striking difference from the usual cliche regarding regional characteristics. In our century the northern Chinese were commonly credited with the characteristics attributed in the twelfth century to their Honan ancestors, and the "real" southerners, particularly in Canton, with those of the twelfth-century Pekinese. Despite the rapid sinicization of the Jurchens, there must have been a major language problem in the administration of the empire. Chinese petitions had to be translated into Jurchen during the court proceedings, which caused delay and not-infrequent misunderstandings, particularly during the earlier reigns when the knowledge of Chinese was still rare even among the educated Jurchens. A further complication was the script. A special script for the Jurchens was created in 1119, the so-called large script, which was apparently based on the Khitan large script. Another script was invented in 1138, the "small script," which has survived on a few stone inscriptions and continued to be used in Manchuria into the sixteenth century. Together with the Khitan script and Chinese there thus existed three completely different writing systems, so that there was a script barrier as well as a simple language barrier between the ethnic groups and within the bureaucracy. It is interesting that the creation of a national Jurchen script system did not mean the abolition of the other two scripts. This is easy to understand in the case of Chinese, which was, after all, the language of the large majority of the population and of their cultural elite. The diplomatic correspondence of the Chin state with Kory6, Hsi Hsia, and Sung seems to have been conducted entirely in Chinese. But within the Chin bureaucracy, the Khitan script also continued to be widely used, if not the large Khitan script, which is unusually complicated, then at least the smaller Khitan script, which was semialphabetical. It was ruled in 1138 that diplomas of appointment for Chinese and Po-hai were to be written in Chinese, for Jurchens in the smaller Jurchen script, and for Khitans in Khitan. Throughout the following decades the Khitan script continued to be used, even in such sensitive offices as the Bureau of National Historiography. And the greatest nationalist of all Chin emperors, Shih-tsung himself, praised the Khitan script for being able to express profound and subtle poetry much better than the Jurchen script could. This reflects the alphabetic and phonetic character of the smaller Khitan characters, which could easily render the sounds of Jurchen. Even more important is the fact that the Khitan script remained for a long time 34 CS, 8, p. 184. Cambridge Histories Online © Cambridge University Press, 2008 ETHNIC GROUPS 283 the medium through which Chinese literature was propagated to the Jurchen. Chinese works were translated (or transcribed?) into Khitan and then from Khitan into Jurchen. It was only under Shih-tsung's successor that the use of Khitan was discontinued. In 1191-2 the use of Khitan script was proscribed, and those officials in the Bureau of National Historiography who knew only Khitan were dismissed. It seems strange that in Sung China no efforts were made to become acquainted with the script of their Jurchen adversaries. When documents or other materials in Jurchen script were captured, no-one could read them. Later Chinese dynasties, especially the Ming and Ch'ing, developed an intricate system of translators' offices and interpreters' schools. But the Sung, perhaps out of national pride, never attempted to do this. In Korea, on the other hand, things were very different, and the Jurchen language continued to be taught and learned for centuries, long after the Chin state had disappeared.35 Social classes Turning from the ethnic problems of the multinational Chin state to its class structure, we are faced with a certain difficulty. Like all other dynastic histories, our main source, the Chin shih, records persons, events, and structures from the metropolitan and bureaucratic point of view. We do not know very much about the social conditions and changes among the Chinese population under the Chin. But perhaps we can assume that the reason for the relative silence of our sources is that no noticeable changes occurred in comparison with conditions under the Northern Sung. Life in the Chinese countryside must have been much the same as it was under the Sung, and it is easy to imagine that city life, too, in a place like K'ai-feng after the violent events of the actual conquest, was very much what it had been before, the only difference being the absence of the Sung court and its dignitaries. It cannot be said that the Jurchen conquest of northern China led to a total social upheaval. Many Chinese, particularly among the upper class, suffered personally from the invaders, but the class structure of the Chinese population was not radically altered. Rich, educated, and influential Chinese continued to live alongside the exploited masses of laborers, small tenant-farmers, and poor landowners. Nor can it be proved that exploitation of the peasantry was any greater under the Chin than under the Liao or Sung. Only in one 35 The complicated situation with regard to the three different scripts used for the Chin written language is summarized by Karl A. Wittfogel and Feng Chia-sheng, History of Chinese society, Liao ( 9 0 7 - j 125), Transactions of the American Philosophical Society, n.s., vol. 36 (1949), p. 253. For studies of the Jurchen language in Korea, see Hiu Lie, Die Mandscha-Spracbkunde in Korea, Indiana University Publications, Uralic and Altaic series, vol. 114 (Bloomington, 1972), pp. 7 - 1 0 , 1 5 - 1 9 . Cambridge Histories Online © Cambridge University Press, 2008 284 THE CHIN DYNASTY respect was a new factor introduced by the Jurchen conquest, namely, the appearance of widespread slavery. As to the upper end of the social ladder we have much information about the changes within the Jurchen population and its stratification. The highest social class in the Chin state was, without doubt, the imperial Wan-yen clan. There were also other Wan-yens, the descendants of nonnoble families of that tribe, but the family of the chieftains that later became the ruling house of Chin enjoyed immensely greater prestige and power than did the other Wanyens. They were a small elite minority within the national minority of the Jurchens, but as shown in the preceding chapter, they held many of the highest offices and, during the early years of the dynasty, practically conducted both military and political affairs. Thefiguresof the 1183 census show that there were 170 households containing 982 clan members. To these must be added 27,808 slaves, or over 163 slaves per imperial clan household. As the possession of slaves was an important social and economic factor that also influenced the property tax assessment, the wealth of these households can be gauged when compared with the average of two slaves per normal meng-an mou-k'o household. The same disparity is revealed when comparing the landholding of the rank-and-file military colonists. The imperial clan household had, on average, 2,166 mou of land; the rank-and-file household had only 274 mou. The wealth and arrogance of imperial clan members, who were mostly absentee landowners and lived in the capital, and their idleness and extravagance caused resentment, not only among the population at large, but also to a stern ruler like Shih-tsung. One of the recurring reproaches was that these imperial relatives were town dwellers, imitated the Chinese way of life, and had lost their ethnic identity and forgotten their former military values. The influence of Chinese civilization on the emperors and the imperial clan must have begun very early. Nothing would be more wrong than to assume that A-ku-ta and his entourage were just savages. One important indicator of the degree of sinicization under the Chin and other dynasties of conquest is the system of personal names. In all Jurchen families a child was given a Jurchen personal name at birth, at least as long as the language and the cultural elements of their past were still alive. Chinese names, however, appear quite early. Chinese names were also given in addition to Jurchen names, even in A-ku-ta's own generation. And what is more, these names conform to the system called p'ai-hang in Chinese, which assigns to all male members of the same generation certain Chinese characters or character combinations according to a predetermined sequence (sometimes taken from a specially composed poem). For example, the generation following A-ku-ta usually had Chinese names whose first character is tsung (ancestor). This was certainly a conscious imitation of Sung usage, which had, too, a strict Cambridge Histories Online © Cambridge University Press, 2008 ETHNIC GROUPS 285 onomastic system based on the p'ai-hang principle in use among the imperial Chao clan.' 6 The age-old Chinese custom of conferring the imperial clan name on meritorious outsiders, particularly non-Chinese allied chieftains, also occurs under the Chin, who granted this inexpensive distinction in thirty cases. But other Jurchen clan names were also sometimes given with imperial approval to non-Jurchen officials. Chinese personal names are, in any case, evidence of the degree of Chinese influence within a Jurchen clan. The number of Jurchen clans {hsing) is given differently in the sources. For the predynastic period, Chinese sources speak of the "thirty clans"; another figure given is seventy-two (this is certainly a round number, seventy-two being a cosmologically relevant number and therefore sometimes meaning only several dozen). A long list of the Jurchen clans, ninety-nine in all, appears in the Chin shib. Counting the imperial Wan-yen family, who are listed separately, the total would be one hundred. This, too, looks like systematizing numerology, and even more clan names appear in the body of the history. A curious distinction is made in the clan list. Eighty-three clans are given white titles (pai-hao) and sixteen, black titles {hei-hao).*1 It is not clear to what this distinction between black and white refers. It is possible that the white clans were regarded as more ancient or superior, because among the Jurchens, as with the Mongols, white was an auspicious color.38 Of the eighty-three white clans, twenty-seven, including the Wan-yen clan, had received fiefs in the Jurchen homelands near the Gold River in Manchuria; thirty in Hopei; and twenty-six in Kansu. The sixteen black clans were enfoeffed in Honan and northern Kiangsu, that is, in the southernmost part of the state. Although no explanation is given in the Chin shih passage, this differentiation must in some way be connected with the transfer of the original meng-an mou-k'o organization to the conquered territories. It is not clear whether these fiefs were only titular ranks or whether they implied territorial jurisdiction or actual landholding. 36 For genealogies of the Jurchen clans and a list of the various Chinese orthographies of personal and clan names, see Ch'en Shu, Chin shih shih pu i wu chung (Peking, i960). On the Chinese p'ai-hang system of personal names, see Wolfgang Bauer, Der chinesische Personenname: die Bildungsgestze un hauplsachlichsten Btdcutungsinhalte port Ming, Tzu und Hsiao-ming, Asiatische Forschungen, Monographienreihe zur Geschichte, Kultur und Sprache der Volker Ost-und Zentralasiens, vol. 4 (Wiesbaden, 1939), pp. 200—10. No systematic study of Jurchen personal names has yet been made, or, for that matter, of Manchu names that could help explain Jurchen names. 37 The list of ninety-nine clan names appears in CS, 55, pp. 1229—30. A Yuan author, Yao Sui ( 1 2 3 9 1314), gives a different figure; see his Mu an chi (SPTK ed.) 17, p. 21b. He says that there were 68 "white" and 44 "black" clans, a total of 112. Instead of the term "title" (bao), he uses the term white or black "writing" (shu). The exact meaning of writing in this context is as enigmatic as is the term "title" (or "number") used in the Chin shih. 38 On the significance of black and white, see Ch'en Shu, "Ho-la Ch'i-tan shuo chien lun T'o-pa kai hsing ho Yuan tai Ch'tng tai te kuohao," Li shih yen chiu, 2 (1956), pp. 67—77. On p. 71 he says that black sometimes signifies inner, paternal clans and white outer clans linked by marriage. It appears to be the reverse here. Cambridge Histories Online © Cambridge University Press, 2008 286 THE CHIN DYNASTY One more observation can be made. Not all of the one hundred clans were in fact Jurchens. Among the white clans there are listed some non-Jurchen clans or tribes, for example, the Khitan Yeh-lii, the Turkic Onggiid, and the Mongolian Onggirad. In addition to the social stratification within the Jurchen ruling minority, we therefore have also ethnic differences, although the inclusion of non-Jurchen clans in the list certainly means that they had been to some extent politically assimilated and integrated in spite of their previously different ethnic background. All this shows the rather fluid character of the aggregation of tribes that was to become the Jurchen nation. Another social differentiation within the Jurchen nation was the fact that as a rule, the imperial Wan-yen clan intermarried with only eight clans, all of indubitable Jurchen origin.39 These eight clans enjoyed considerable social prestige, and we find many of their members in the highest offices. The marriage customs of the Chin ruling house therefore fall in between the Chinese usage, in which in principle there was no limitation on the families from whom consorts might be taken, and that of the Liao rulers, who chose their empresses from only one family, as did the Mongolian Yuan emperors. We earlier mentioned slavery as a characteristic of Chin social structure. They were the lowest stratum in Chin society. But within the misera contribuens plebs there again existed a distinct stratification from afiscalpoint of view. The families in the state, apart from the privileged Jurchen nobility and the tax-exempt Chinese bureaucrats, were classified as follows: taxpaying families, non-taxpaying families, basic families and miscellaneous families, regular families and controlled families, government families, slave families, and families paying double land tax.40 This enumeration is rather unsystematic because it mixes fiscal, ethnic, and socioeconomic categories, but it gives a good overall view of the gradations within the masses of the population. Taxpaying and corvee-liable families were those who owned land. Nontaxpaying families were those who were too old, too young, or invalid. Basic families (pen-hu) refer to the Jurchens, and miscellaneous (tsa-bu) to Khitan, Chinese, Po-hai, or other ethnic groups, a distinction introduced in 1195, perhaps in order to avoid the having to pin down the nationality of individuals. Regular families (cheng-hu) is a term designating families who had been slaves ofameng-an mou-k'o and had subsequently been freed but still remained under the administration of their respective unit commanders. The meng-an mou-k'o population as registered in the census of 1183 there39 These eight clans, according to CS, 64, p. 1528, were the T'u-tan, T'ang-kuo, P'u-ch'a, Na-Ian, P'usan, Ho-shih-lieh, Wu-lin-ta, and Wu-ku-lun. Of these, the T'ang-kuo and P'u-ch'a were "black," and the rest were "white." This list is, however, incomplete because in addition to these eight clans, the P'ei-man clan also had supplied empresses for T'ai-tsu and Hsi-tsung (CS, 63, pp. 1502—3). 40 This enumeration is given in CS, 46, p. 1028. The distinction between the Jurchen and other ethnic groups (basic and miscellaneous families) is mentioned in CS, 46, p. 1036. Cambridge Histories Online © Cambridge University Press, 2008 ETHNIC GROUPS 287 fore included all those formerly enslaved persons who were presumably Chinese. The term cheng-hu could therefore also be rendered as regular families. Controlled families (chien-hu) were those who were controlled by the Board of Palace Registers; they were government bondsmen who had been conscripted from the free population to serve in government agencies, mainly the palace administration. Government families (kuan-hu) were persons who had already been slaves and had then been forced to work in imperial workshops. They were distinguished from "normal" slave families owned by private individuals. The double taxpayers (erh-shui-hu) finally were perhaps more a kind of slaves than simply a group of persons who were taxed doubly. This group consisted of families who had been given to Buddhist monasteries by the emperors of the Liao dynasty. They had to pay both rent to the monastery and land tax to the government. For all practical purposes they were slaves of the monasteries. Their numbers must have been considerable, because even at the end of the twelfth century, temple slavery was denounced, and a decree was issued that permitted them to be set free. The pious donation of people to monasteries was a relatively humane procedure if compared with the wholesale enslavement that occurred during the military campaigns of the Chin state. It can be assumed that most of these slaves were people who had been taken captive. Another reason for slavery common in all periods of Chinese history was self-sale in periods of famine and distress, or the sale of children at such times. All these reasons for enslavement as a private slave (as contrasted with a conscripted government slave) are well attested during the Chin dynasty. The biggest slave owners were, of course, the members of the imperial clan. When still a prince, Shihtsung owned over ten thousand slaves. Slave status was hereditary, so that the unhappy victims of military enslavement not only had to suffer personally but their descendants also remained slaves. Marriage between a free woman and a slave resulted in slave status for the woman; she could, however, sue for divorce if she had been unaware of her husband's slave status. If sons or daughters of freed slaves who had been born while the parents still were slaves married a free person, they were regarded as free and could even take the examinations.•»' Slavery did not mean in every case absolute poverty at the barest subsistence level. Sometimes a slave could, perhaps as a majordomo, achieve some influence and status. For example, a decree issued in 1190 forbade slaves of the imperial clan to encroach on merchants or to extort debts under false pretenses. The redemption of slaves, which was in theory always possible, was handled differently under different reigns. In the early period, return to a free 41 A detailed description of marriage rules for slaves is given in CS, 45, p. 1021. Cambridge Histories Online © Cambridge University Press, 2008 288 THE CHIN DYNASTY status was more or less left to the individual largesse of the owner. In 1116 it was ruled that a slave could be freed by compensation in kind: One person could be freed if redeemed with two persons. Later, in 1141, a decree ordered that the government could free slaves if three bolts of textiles were paid for one adult male, and two for a woman or child. At a still later date, perhaps around 1200, payment for redemption could be made in money: The ransom rate, if the slave had been paid for in money, was fifteen kuan (strings of cash) for an adult male, and half this sum for a woman or adolescent.43 It seems, however, that the redemption of slaves by payment was limited to cases in which slavery had been due to self-sale and similar reasons, not to military enslavement. In any case, the widespread practice of slavery under the Chin continued under the Mongols and remained a characteristic of thirteenthand fourteenth-century social structure. It goes without saying that the overwhelming majority of the slave population was of Chinese origin, although there must have also been Jurchens and other tribal slaves as well. It should now be clear that the basic unit of Chinese society under the Chin, as in other periods, was the family. The family system of Chin, at least among the Chinese population, was certainly not different from that of contemporary Sung China. Chin legislation concerning marriages and family status is frequently referred to in our sources. It is not, however, always clear whether the decrees and ordinances concern the Jurchen population or Chin subjects in general. Most of the rather detailed rulings seem to point to a conflict between original Jurchen or other non-Chinese custom, and Chinese tradition. Among the Jurchens as well as the Po-hai people, marriage by elopement was quite common, a custom that was forbidden in Shih-tsung's reign. Another custom conflicting with Chinese usage was levirate and sororate. It had been customary among the Jurchens for sons to marry the deceased father's concubines; nephews, those of uncles; and brothers, the other brothers' widows. Both these ancient traditions were abolished or modified under Shih-tsung: Elopement was forbidden, whereas sororate and levirate were allowed only for Jurchens but prohibited for Po-hai and Chinese.43 Another concession to Chinese mores was the introduction of exogamy. Originally every Jurchen could marry within his own clan, but already under A-ku-ta, nonexogamous marriages were frowned on and could be ended by divorce. Under his successor even the sons or daughters of stepfathers or stepmothers, who were not at all consanguineous, were forbidden to marry. Concubinage was legal, but in 1151 it was decreed that officials could have only two concubines. It must remain an open question whether this limita42 On the practice of redemption by supplying substitutes, see CS, 2, p. 29; by paying money, 58, p. '35343 CS, 6, p. 144. Cambridge Histories Online © Cambridge University Press, 2008 ETHNIC GROUPS 289 tion ever became effective. An indication that adultery - in other words, sexual freedom for women — was tolerated among the Chin elite is a decree of 1170, which ruled that officials' wives who had committed adultery should not enjoy their husbands' rank. Those women, however, who had obtained their rank because of their sons' positions were not affected by this rule. It is easy to imagine that staunch Confucianists would have denounced such behavior. A similar conflict between tribal customs and Chinese traditions occurred in law. The old law of the Jurchens was, to a great part, based on the ius talionis (eye for an eye) and on the principle of indemnity (Wergeld in Germanic law). Lighter crimes were punished by flogging. Murderers were executed, and 40 percent of their personal property went to the state (ruler or chieftain), 60 percent to the victim's family, and the murderer's family members were enslaved. But redemption was possible by giving horses, oxen, or other property to the victim's family. In this case, the only punishment inflicted on the culprit was to cut off his ears or nose in order to identify him as a criminal. In the transition from tribal to Chinese codified law, several phases can be distinguished. Under T'ai-tsu the old customs remained largely unchanged, and also under his successor T'ai-tsung, Jurchen legal customs were applied along with some laws of the Liao and Sung. The laws were extremely harsh, and capital punishment for theft of goods exceeding a certain value was common. The second phase is characterized by attempts to codify the existing statutes and the compilation of an eclectic mixture of Sui, T'ang, Sung, and Liao legal precedents (1145). This compilation was, however, not a comprehensive code like the T'ang lit shu i (T'ang code with commentary) or the Sung hsing t'ung (Encyclopedia of Sung penal law). These laws were regarded as excessively harsh and cruel. The third phase in the transitional period was the reign of Emperor Shihtsung. He took a great interest in legal matters and ordered the compilation of a comprehensive collection of laws and statutes. The result was a handbook of 1,190 articles in twelve chiian; the emperor, however, was dissatisfied because the text was not always clear and some legal terms had become obsolete. A thorough revision was therefore ordered. The final stage in the sinicization of Chin law took place under Emperor Ch'ang-tsung. After some preliminary codifications the T'ai ho code (T'ai ho lit) was proclaimed and became effective in the fifth month of 1202. The T'ai ho code as such has not survived, but the Chin shih gives a detailed description of its contents.*« The code had 563 articles (the T'ang 44 SeeCS, 45. Cambridge Histories Online © Cambridge University Press, 2008 290 THE CHIN DYNASTY code had 502) and was accompanied by a collection of 713 ordinances and a handbook containing imperial edicts and regulations for the six ministries. This massive codification of Chinese law under Ch'ang-tsung shows the energy with which the legal scholars (all of them Chinese) had been at work. It must be regretted that the T'ai ho code is lost as a whole. But, out of the 563 articles of the code, the legal content of 130 is known through citations in later juridical works, above all, in the Yuan juridical handbook Yuan tien chang, so that for roughly one-quarter of the T'ai ho code a comparison with T'ang and Sung laws is possible. Some of the differences between codified T'ang law and Chin law can be explained by economic developments. In the T'ang code the value of illicit goods or gains was assessed in silk, whereas under the Chin the value was measured in cash, which shows that monetary economy had become universal. From other differences it can be deduced that the Chin code tried to strengthen the authority of both the state and the family elders. For example, the punishment for an official who did not carry out his duties within the prescribed time was more severe under the Chin code. We also find rules in Chin law that are more severe with regard to offenses threatening the authority of the head of the family and the husband. Such an offense was not punishable if a husband "according to reason" had beaten his wife if she had committed a crime and the result had been fatal. The power of a slave owner over his slaves was enhanced by the Chin code. If a slave cursed his master, the T'ang code prescribed exile, but in Chin law it was a capital offense. Also, some types of sexual crimes were more severely punished than under T'ang and Sung law. Of particular interest are those provisions of Chin law that reflect the multinational character of the state. The ethnicity principle was expressly given priority. Offenses committed against each other by persons of the same nationality (t'ung-hi) were to be tried according to the respective customs of that nationality. Some of the peculiar Jurchen marriage customs also were permitted by Chin law. Inheritance law differed among the nationalities. The separation of households during the lifetime of parents or paternal grandparents was punishable under T'ang law. But among the Jurchens, sons could set up their own households as soon as they were able to support themselves, a custom also to be found among the Mongols. The Chin code expressly permitted Jurchen sons or grandsons to live separately in their own households while their fathers or grandfathers were still alive. The premature division of family property inherent in this custom may have contributed to the impoverishment of the Jurchen military colonists, a fact already noted by a Jurchen minister in the Ta-ting period (1161-89). When the Chin state was absorbed into the Mongolian empire, the T'ai ho Cambridge Histories Online © Cambridge University Press, 2008 ECONOMIC CONDITIONS 291 code remained in force for the Chinese population in the newly conquered territories. It was formally abrogated only in 1271, the year when the Mongolian ruler Khubilai adopted the dynastic name of Yuan. In sum, the legal development under the Chin from unlimited vendetta to a preponderantly Chinese system after 1202 must be seen as a parallel to Jurchen societal evolution from an unstructured clan society to a multinational state modeled on Chinese precedents. It could perhaps be said that the controlled harshness of Jurchen justice was, over the years, replaced by the uncontrolled harshness inherent in the traditional Chinese legal systems. The formal abrogation of the T'ai ho code therefore marks the end of an important transitional period in the legal history of northern China.« ECONOMIC CONDITIONS Agriculture and pastoralism Land, under the Chin, was in principle a commodity that could be inherited, sold, or mortgaged, and there were no general prescriptions regarding what the individual farmer or tenant had to grow, except for mulberry trees. The military colonies had a special status, and sometimes it is not quite clear whether our sources are speaking of land tenure in general or are referring to the meng-an mou-k'o land. In addition to the lands in private hands, the government owned a considerable proportion of the arable land. Either these fields were regarded as public fields, or they were attached to offices to provide the officeholders with income in kind. We have no exact figures for the relative proportions of privately held land, meng-an mou-k'o land, and government lands, only some stray indications. For example, in 1221 roughly one-quarter of the arable land in Honan belonged to the government in one or the other form. Also, all lands near the Great Wall, near other fortifications, and alluvial lands on the banks of the Yellow River were considered to be government property. This large reserve of land in the hands of the government was mostly used for distribution to military colonists, but ordinary farmers were also encouraged to apply for a grant of government land if it were uncultivated or untenanted. After the disasters of 1214—16, over half a million military colonists sought refuge in Honan and Shantung and applied for government land there. It seems that land was frequently 45 As yet there exists no comprehensive study of the Chin legal system in a Western language. A standard account is given by Niida Noburu in Chugoku hosei shi ktnkyu: Keiho (Tokyo, 1959), pp. 4 5 3 - 5 2 4 . Equally useful is the study by Yeh Ch'ien-chao, Chin lii chih yen chiu (Taipei, 1972). For early Jurchen legal customs, see also Herbert Franke, "Jurchen customary law and the Chinese law of the Chin dynasty," in State andlawin East Alia: FestschriftKarlBiinger, ed. Dieter Eikemeier and Herbert Franke (Wiesbaden, 1981), pp. 215—33. Cambridge Histories Online © Cambridge University Press, 2008 292 THE CHIN DYNASTY extorted from its legal owners by government agencies or by Jurchen grandees, because laws were repeatedly passed against such abuses. The standardized land allocation policy of earlier centuries (which had lasted until the early and middle T'ang period) no longer existed under the Chin, except in the case of the military colonies. For these, afixedquota was set for the allotment of land, which varied according to the times and the locality. The normal allotment for an adult individual was, under Shihtsung, 4 ch'ing, 4 mou (1 ch'ing is roughly equivalent to 14.34 acres, 1 mou to 0.14 acres) plus three plow-oxen. A limit for the number of oxen (and, consequently, the amount of land) was introduced as well, but it seems that this was valid only in cases of new allotments or the redistribution of lands and cattle, because a huge disparity had developed over the years between poor and rich military colonists, just as among the nonmilitary farming population. We know that in 1183 the military colonists owned around 1,690,3 80 ch 'ing of land. This was a considerable percentage of the total land under cultivation in the Chin state. The total of cultivated lands can be figured only indirectly through the tax figures. Land tax was 10 percent of the yield. A high-class field yielded 1.2 shih (1 shih equals approximately 599.4 liters) per mou, a middle-class field 1 shih, and a low-class field 0.8 shih. In 1171 the annual income from land tax was about 9 million shih of grain. If we take the average yield of 1 shih per mou as our basis, the total taxed lands must have amounted to something like 900,000 ch'ing, or well over 13 million acres. Although our figures for 1171 and 1183 are separated by twelve years, we can therefore conclude that in the golden age of Shihtsung the greater part of all cultivated lands was in the hands of the military colonists. The intensity of agriculture obviously varied from region to region. Honan, in particular the area around K'ai-feng, was the definite center of agricultural production. In 1219, when the territory of Chin had been drastically reduced, Honan had 1.97 million ch'ing of arable land, of which less than half, only 960,000 ch'ing, was cultivated, certainly because of the widespread flight of the farming population and the military insecurity in the border areas. The total grain yield (millet and rice) in the state can be estimated at about 90 million shih per annum, one-tenth of which went to the government as land tax. The annual government expenditure in grain was over 9 million shih in 1192 (7 million shih of millet and 2 million shih of rice). This was used for officials' stipends and army personnel. We are told that the average grain consumption of an individual was 5 tou per month, or 6 shih per year. In other words, the average yield from the lands in the whole state could just feed the population adequately, but good harvests were Cambridge Histories Online © Cambridge University Press, 2008 ECONOMIC CONDITIONS 293 needed if a food reserve was to be built up. For food production, the Chin state certainly could not compete with the Southern Sung, where in most areas more than one annual harvest of rice was possible. This precarious food situation was early recognized by the Chin government, and great attention was paid to increasing the cultivated area by means of irrigation, particularly during Chang-tsung's reign. Magistrates who increased the irrigated area in their districts were rewarded with promotion. Another measure for increasing the yield was terraced cultivation (small-plot farming), meaning that mountain slopes could be farmed. It seems that all these measures were taken relatively late and applied only locally, so that the Chin's overall food situation did not improve radically. It explains also why rice was one of the commodities imported from Sung into Chin. Sericulture also played a certain role. Mulberry trees were ordered to be planted in relation to the landed property of the farming household. We know the figure for military colonists: One mou out of every forty mou was to be planted with mulberry trees; another passage even mentions 10 percent of the area as obligatory mulberry plantations. Although the great centers of silk production were in the Sung state and silk was one of the export goods from Sung, Chin apparently had a textile production of its own, sufficient for most basic needs. Cattle breeding on a large scale was concentrated in central and western Manchuria, northern Shansi, and Kansu, including areas that now belong to Inner Mongolia. These pasture grounds had been taken over from the Liao, and the Liao herds had become property of the new overlords. Jurchen officials were assigned as commissioners of herds, whereas all the minor herdsmen were Khitan or other tribesmen. The commissioner of herds and their staffs were recruited from the meng-an mou-k'o population, including slaves. These officials were held responsible for the number of cattle under their supervision (horses, camels, oxen, and sheep). If the number decreased and more than a fixed percentage of the animals died, they would be punished and degraded; a higher-than-average increase would be rewarded. It was regarded as a first-class result if every year the increase was 2 horses, camels, or oxen, or 4 sheep per every 10 animals and if the death rate among horses was fewer than 15 out of 100. The Khitan rebellion of 1160—2 reduced the herds of Chin to near extinction; of the nine herding grounds the animal stock of five was completely broken up and appropriated by the rebels, who had more livestock than did their Jurchen adversaries. In the remaining four, only a small part of the herds remained. A long time was needed to restore the former number of cattle in these regions. Finally, after some plentiful years, in 1188 the herds had again reached a considerable size. Cambridge Histories Online © Cambridge University Press, 2008 294 THE CHIN DYNASTY The government controllers then counted 470,000 horses, 130,000 oxen, 4,000 camels, and 870,000 sheep.46 Pasturelands existed not only in the northern parts of the state but also in the former Chinese southern provinces, although there they were much smaller and limited by settled agriculture. In Honan, around K'ai-feng, 63,000 ch'ing of land (only a small percentage of the cultivated land) were used as pasture, and 35,000 ch'ing in Shansi Province. In view of the great importance of horses for warfare, all the horses in the state were liable to confiscation in times of emergency. The herds of the military colonists were normally replenished from the north so that the loss of the Manchurian plains in 1215 resulted in a marked decrease of war potential. Hunting, which had been one of the main occupations of the primitive Jurchens, became more and more the sport of the ruling minority when the court as the center of the state was moved south. Originally the Chin emperors until Hai-ling wang's time imitated Khitan customs by holding hunts in all the four seasons: fishing and hunting wild geese in spring, deer in autumn, and tigers in winter. These seasonal expeditions, however, were no longer possible after the capital had been moved to Peking because it was felt that these large-scale battue hunts would interfere with agriculture. So hunting expeditions were limited to one month in winter. Also, for the military colonists, hunting was limited to two periods of ten days each winter. Manufacture and crafts Although our usual sources do not inform us specifically, we can safely assume that the arts and crafts of the civilian population in the former Sung territories of Chin continued to be the occupation of many inhabitants of the towns. The social structure of Chinese society under the Chin certainly changed more at the top than at the bottom and intermediate levels, and the same must have been true for the economic activities of the Chinese. We are much better informed about government manufacture and the production of commodities for which a state monopoly existed. These monopoly goods were salt, wine, yeast, vinegar, tea, incense, alum, cinnabar, tin, and iron. These were partly, as in the cases of salt and wine, produced under government supervision and traded through government agencies, or as with tea and cinnabar, they could be imported and sold only with a government license. 46 The number of horses in the Chin state was less than half the number the Liao had had one century earlier in 1086, when one million horses were counted in a cattle census (LS, 24, p. 291). Cambridge Histories Online © Cambridge University Press, 2008 ECONOMIC CONDITIONS 295 The most important commodity from the fiscal point of view was salt. The center of salt production was in Shantung, mostly the production and marketing centers that had already flourished under the T'ang. There were also some salt pools and lakes in Manchuria and Mongolia that produced salt for local consumption, but even their small yields had already been taxed under the Liao. Salt production was extended when the Jurchen forces conquered the Chinese Central Plains, and new monopoly offices had to be established. Altogether seven administrations controlled the production and distribution of salt; those in Shantung brought in the greatest profit. Salt was sold only against vouchers (for large quantities) and tickets (for smaller quantities), which had to be officially endorsed. The weight measures (cases and bags) differed by region. We have some figures on the retail price of salt: It varied between thirty and forty-three copper cash per Chinese pound {chin). With this we may compare the price for one tou (approximately six liters) of rice in approximately the same period (around 1180): three hundred copper cash. In other words, if measured by weight, salt was almost as expensive as rice.47 Retail trade was sometimes in the hands of local tycoons who made a business out of monopolizing the local market, to the detriment of small shopkeepers and peddlars. This was made possible because the salt vouchers and tickets were transferable instruments like bank checks or paper money and were not personalized or rationed. But despite these encroachments by private money-makers, the government made an enormous profit from the salt monopoly. The annual proceeds from salt were fixed by a quota system that regulated production and sales figures and furnished the biggest item of money income in the national budget. The annual quota before 1198 had been not less than 6,226,636 kuan, or strings of a thousand cash. This was later raised to 10,774,512 kuan, a sum that accounted for almost half of the money revenue. Another commodity produced in government shops was wine. A wine monopoly was introduced by former Sung officials along with other fiscal institutions in 1125. Private wine production was prohibited, just as private salt production and trading was. The wine monopoly also followed the quota or, rather, target system, but the few figures in our sources show that the proceeds from wine were far below those from salt. The chief monopoly office was in the Central Capital, Peking; it yielded only a few hundred thousand 47 There is little information on retail prices under the Chin dynasty. Only some scattered data on prices can be found; for example, some prices are given in Lou Yao's (1137—1213) description of a Sung embassy to the Chin in 1169—70, the Pei hsingjih lu (TSCC ed.). In a Chin prefecture! town near the border, Lou Yao had to pay 210 cash for one pound of wheat flour, 120 cash for one peck (tou) of millet or some other grain, and 240 cash for one peck of rice (Pei hsingjih lu, shang, p. 12b). For one bolt of good silk, the price was 2,500 cash in Hopei; for one ounce of floss silk, 150 cash. Horses were even more expensive than a first-class donkey, which cost 40,000 cash (Pei hsingjih lu, hsia, p. 8b). Cambridge Histories Online © Cambridge University Press, 2008 296 THE CHIN DYNASTY kuan annually. It is clear that the law prohibiting private wine making was frequently disobeyed, particularly in the Jurchen grandees' households. On the other hand, some exemptions from the law were legal, such as the production of wine for special occasions like marriages and funerals. Yeast was treated like wine from the fiscal point of view; it was subjected to the monopoly laws mainly because it was an indispensable ingredient for making wine. A curiosity is that yeast formed a part of an official's salary in kind, which also points to private wine making on a large scale. Like other Chinese dynasties the Chin had many government workshops. These produced military weapons and such consumer goods as textiles and embroidery. Printing offices were also run by the government. The government workshops could recruit skilled labor from the population, because in principle every artisan was registered and liable to labor conscription. Whereas we are not well informed about labor conditions in nongovernment industries, we do have detailed figures for industrial wages in the state-run workshops. For example, conscripted workers in the arsenals received daily 100 copper cash and about one liter of rice; printers were better paid, with a daily allowance of 180 cash and gifts of textiles. Foremen and supervisors received correspondingly more, according to afixedwage-schedule. Surprisingly, mining was largely left to private enterprise. There were foundries for gold, silver, copper, and iron. In earlier reigns, gold and silver production was taxed but became tax exempt under Shih-tsung. In 1192 the silver smelteries were again taken under government management. An iron monopoly was introduced rather late, after the loss of Manchuria and the Peking region to the Mongols in 1219. Metal and coal mining seems to have been quite highly developed under the Chin.48 To the silver mined in northern China (chiefly in Hopei) must be added the considerable import of silver into the state through the annual payments from Sung, so that the government treasuries were able to hoard considerable amounts of silver. Gold and silver were, however, like all other commodities, subject to a commercial tax introduced in 1180. We do not have figures that would allow us to compute the commercial turnover for the whole state in a given year, but we do have such figures for the region of Peking. Gold and silver sales were taxed ad valorem at 1 percent, other commodities at 3 percent. For gold, this rate was later raised to 3 percent, and for other commodities, to 4 percent. The figures available allow us to calculate the total industrial and commercial turnover for Peking in 1196 at over seven million kuan, a rise of one-third if 48 A detailed description of the coal and mining industries is given in Robert Hartwell, "A cycle of economic change in imperial China: Coal and iron in northeast China, 750—1350," Journal of the Economic and Social History of the Orient, 10(1959), p p . 102—59. Cambridge Histories Online © Cambridge University Press, 2008 ECONOMIC CONDITIONS 297 compared with that during the reign of Shih-tsung in the 1180s. 49 But such figures are meaningful only if we can compare them with those of other parts of China and their commercial turnover. Communications and foreign trade After their conquest of north China, the Chin could take over the existing transport system of waterways and road networks. Communications were vital chiefly because large urban centers like Peking depended on grain transport for food. The situation was better in K'ai-feng because this city was situated in the midst of a densely cultivated surplus area. The waterways were much more important than road transport, and there was a large network of canals and riverways in Honan, Shantung, and Hopei. The northwestern provinces had to rely more on road transport. Although roads linked the prefectural and county towns in all parts of the state, road transport was more expensive than water transport. We know the exact figures for the transportation costs of grain, rice, salt, copper coins, and other commodities. They show that road transport for some of these goods was up to two or three times more expensive than water transport. Road transport costs also differed according to whether the road went through flat or mountainous country and also according to the season; the wages for porters were higher in summer and autumn than in winter or spring. They were still low enough, being 90 and 114 cash per day, respectively. All these figures are for government transportation of taxes in kind and local tributes or proceeds from government trade, but they give an indication of the relative costs of water and road transport. Road transport for private traders and their caravans was limited to coolies and ox carts because they were forbidden to use horses for transportation. Government-owned boats and whole flotillas could be hired on a long-term lease over several years. Although the hiring costs were equal to the value of goods transported, payment could be stretched over five years or more. The hire was highest during the first year (20 percent), so that the goods became more expensive by one-fifth through transport costs alone. If the water and road transport system was thus partly under government management and only partly private, the courier system was for government use only. A courier network was established very early (in 1124) in connection with the campaigns against Liao and Sung. Every fifty // a postal station was established where fresh horses awaited the dispatch riders. During the campaign of 1206 an express courier system was introduced; it was consid49 CS, 49, p. 1106. Cambridge Histories Online © Cambridge University Press, 2008 298 THE CHIN DYNASTY ered very fast because the riders could cover three hundred // per day. The horses were commandeered from the civilian population - much like the well-known postal system under the Yiian dynasty. The same abuses were also common under the Chin, namely, the illicit use of the state posts for private purposes. An important aspect of the communication system was the transportation of goods to and from the established border markets with Sung, Kory6, and Hsi Hsia. Foreign trade was an important factor in the Chin economy, which extended beyond the annual payments from Sung. The border between Sung and Chin did not constitute an "iron curtain," and no Great Wall separated the two parts of China. Foreign trade was, therefore, now what had formerly been domestic trade among different provinces of the same country. Although trade between Sung and Chin had begun already on a smaller scale after the establishment of the buffer state ofCh'i (which profited accordingly), commercial relations between the two states were regularized after the treaty of 1142, and its corollary agreements had provided licensed border markets on both sides. For the Sung the main center was the county town of Hsii-i in northeastern Anhui, where the Huai River met the Pien River on which K'ai-feng was situated. On the Chin side the center was the prefectural town of Ssu, but there were other licensed centers, nine more on the Sung side and eleven on the Chin side. One of these was in Shantung, perhaps for trade by oceangoing ships. Sung—Chin trade was interrupted only during the war of Hai-ling in 1161—5 and the revanchist war of Sung in 1206—8 and continued more or less until the last campaign of 1217-18 and the downfall of Chin. For both states, foreign trade was a sort of state monopoly. Unauthorized transactions were forbidden, and Sung goods were sold to Chin merchants at fixed prices through the government. Wholesale merchants from Sung could not enter Chin territory, only small traders with capital or goods valued at one hundred strings of cash or less. They needed a permit to enter Chin territory and, after their return from Chin, a sale certificate for which the government charged them a tax. The tax was 20 percent of the value; 2 percent more went to the government broker, and 0.4 percent as transportation cost to the coolies. The charge on the Chin side was 30 percent. Moreover, each Sung merchant had to spend three strings of cash, for which he was entitled to food and lodging. These arrangements were a source of considerable income to both governments, but they also led to continual attempts to evade the cumbersome legitimate procedures by smuggling. Another reason for dissatisfaction was the bribery practiced by the market overseers. The annual proceeds from Ssu prefecture during the Ta-ting period (1161— 89) was 53,467 kuan; the target was raised in 1196 to 107,393 kuan and thereby doubled. The quota or target system implied that a certain Cambridge Histories Online © Cambridge University Press, 2008 ECONOMIC CONDITIONS 299 amount of goods had to be traded in order to meet the quota. The largest import item was tea. It seems that everybody in the Chin state, including the farmers, drank tea, which of course became scarce if trade was interrupted for some reason. Attempts to grow tea in Honan had failed, so that Chin was dependent on imports from the Sung. 50 The annual average of imported goods in Ssu Prefecture during the Ta-ting reign reads like a grocery store's stock: 1,000 crates of green tea, 500 pounds each of lichees and longans, 6,000 pounds of kumquats, 500 pounds of olives, 300 cases of dried bananas, 1,000 pounds of sappan wood (used for dyeing), 7,000 cases of tangerines from Chekiang, 8,000 cases of oranges, 300 pounds of granulated sugar, 600 pounds of ginger, and 90 bags of gardenia seeds (also a dye). Other goods, for which the quantities are not given, were rhinoceros horn, ivory, and cinnabar.51 Although the Sung had prohibited the export of rice and copper coins to Chin, it seems that rice, as well as cattle, was sold over the border to Chin. Chin also banned exports of coins, grain, iron weapons, and armor. The goods exported from Chin were northern pearls from Manchuria, drugs like ginseng, textiles, and - if available for export - horses, although in theory the export of horses was forbidden. Another export commodity from Chin must have been antiquities, because in 1157 their export was banned. The sale of antiquities to Sung can be seen as a result of the widespread growth of art collecting among the Sung intelligentsia. It is difficult to say whether Chin had an adverse or favorable trade balance with Sung if all items, including the illegal border trade, are taken together. Similar border markets existed between the Chin and Hsi Hsia. Hsi Hsia purchased chiefly textiles and silk and exported horses and jade from Inner Asia. Licensed markets had been established in 1141 as part of a mutual agreement between the two states. There was also some trade with Koryo and the Mongols, but not much is known about commercial relations with these two nations. Currency The money of Chin is a good illustration of Gresham's law. Chin money plays a great role in the history of paper money because there were several attempts to establish a paper currency, which resulted in accelerated inflation. The basic problem with Chin money was the shortage of copper. Copper was a monopoly item; the manufacture and sale of copper utensils were in the 50 The economic role of trade between the Sung and the Chin was studied exhaustively by Kato Shigeshi in his Shina ieizai shi kosbo, Toyo bunko ronso no. 34 (Tokyo, 1953), vol. 2, pp. 2 4 7 - 3 0 4 . 51 CS, 50, pp. 1 1 1 4 - 1 5 . Cambridge Histories Online © Cambridge University Press, 2008 3OO THE CHIN DYNASTY hands of the state. When copper began to become scarce, private smelting and manufacture were allowed, but the government fixed the sale prices. At the beginning of the dynasty, the coins of Liao and Sung, later also those of Ch'i, were used, and it was only in 1157 under Hai-ling that thefirstcopper coins of Chin were cast. The more the economy expanded, chiefly during the peaceful reign of Shih-tsung, the more acutely the shortage of copper coins was felt. In spite of heavy punishments, the population began to counterfeit coins, but these were of worse quality than the government coins. To meet the shortage, iron coins were issued, but these were discontinued in 1193 because they were impractical. One objective of casting iron coins, which circulated chiefly in the southern provinces, was to stop the flow of copper currency into Sung territory. The problem was to have enough copper cash to serve as legal tender throughout the state for both taxes and private commerce. There were frequent complaints that not enough coins were circulating, mainly because they were being hoarded by private individuals. We know that the total amount of coins in circulation in 1178 was over sixty million strings of cash (kuan). This is not much considering that the state had then a population of over forty million people and that the economy was nourishing. But copper coins were not the only metal in circulation, because silver ingots were the most common means of payment, at least for larger transactions. Paper money was first printed in 1157, after the capital had been moved to Peking from Hui-ning in Manchuria. It was modeled on Sung paper money. The period of circulation was limited to seven years, after which the money was either withdrawn from circulation or exchanged for new notes. In 1189 this limitation was abolished, and only regional restrictions were made. The denominations followed the coinage system; the bills were issued with a face value of strings (kuan) and cash (wen) and could be converted into actual cash. The government tried to keep the amount of paper money issued within reasonable limits. The total face value of notes was not to be more than the amount of cash in circulation. At the same time legislation was introduced in order to limit the amount of coin in private possession. A new kind of paper money was introduced in 1197, which could be exchanged for silver. The government treasuries had hoarded a large stock of silver, a part of which was cast into ingots of fifty ounces each. Money taxes could be paid either in silver or with the new notes; in some cases half had to be paid in silver and half in notes. This complicated the currency situation because the old copper coins continued to circulate, and the exchange rate for these forms of legal tender was subject to changes. A further complication was that certain kinds of paper money were restricted to the Central Capital, Southern Capital, and other cities. The silver-based notes were a relatively Cambridge Histories Online © Cambridge University Press, 2008 ECONOMIC CONDITIONS 301 sound currency. As long as the government accepted tax payments in paper, its value could be considered stable. This situation changed, however, when war broke out in 1206. The enormous costs of the campaign severely taxed the financial resources of the state and apparently led to the excessive printing of paper money, especially when the Mongols attacked. Denominations as high as one thousand kuan were printed and put into circulation. From then on, the monetary history of Chin is a series of desperate attempts to stabilize the currency by repeated issues of paper money. Every few years a new type of money was issued with a new high-sounding name, but the value in each case sank so low that, for example, in 1221, eight hundred kuan in paper currency was equivalent to one ounce of silver. One reason for the rapid depreciation was that with every new issue the old notes continued to circulate, so that the national economy was flooded with paper money. Silver, of course, kept its value but was obviously hoarded by everybody who could find it. In the four years between 1217 and 1221, paper money had been devalued at the rate of 40,000 to 1. The effect on the private economic sector can easily be imagined, and we hear repeatedly that the merchants and tradespeople were forced to shut their shops because they were unwilling to sell in exchange for worthless paper.' 2 Even in the very last moments of the dynasty, when the court of the emperor had found temporary refuge in Ts'ai Prefecture, a new paper currency, in theory convertible to silver, was issued (1233), but a few months later the Chin state disappeared. The monetary chaos during the last two decades of Chin was not due to unsound fiscal policies but, rather, was the consequence of military defeats and the resulting losses in revenue and in general economic productivity. For many years, chiefly under Shih-tsung and Chang-tsung, Chin money was certainly as sound as Sung money. The experience of the Chin did not, in any case, deter the fiscal policymakers of the Yuan dynasty, who for many years based the Yuan monetary system on a paper currency, much to the surprise of travelers like Marco Polo, who could not believe their eyes when they saw that a piece of printed paper could serve as money. Taxation and the national budget No state can survive without taxation of some kind. The problems, in Chinese history as elsewhere, were how much to tax and how to match 52 It should be mentioned as a curiosity that the money of 1223 was printed not on paper but on silk, a feeble attempt to give it some intrinsic value, however small. A few notes and a printing block for paper money from this period have survived to this day; see Joseph Mullie, "Une planche a assignats de 1214," Toungpao, 33 (1937), PP- I 5 ° - 7 - Cambridge Histories Online © Cambridge University Press, 2008 3O2 THE CHIN DYNASTY income to expenditure. The scholars who compiled the Chin shih in the 1340s under the Yuan did not have a high opinion of Chinfiscalpolicy. They wrote a clear and succinct summary of the economic development of Chin as they saw it. 5 ' They pointed out that the desire for quick profits caused longterm damage to the people and that the Chin combined in their economic policy the laxity of Sung with the harshness of Liao and therefore imitated those features that had brought about the ruin of these two states. The Yuan authors, of course, had to justify somehow the collapse of Chin, and their judgment appears overly severe to an unbiased modern observer. The real decline of Chin set in rather late, after 1200, and it was due much less to bad legislation and poor civil policies than to foreign politics in which Chin had fended off ambitious or revengeful neighbors on all sides. Indeed, a great, and perhaps the greatest, part of the state revenue went to maintain the military machine. During the first reigns, the state's economic situation must be regarded as excellent. The Chin had inherited the accumulated wealth and reserves of both the Liao capital and the Sung capital. After the conquest of K'ai-feng the booty was enormous: 54 million bolts of silk, 15 million bolts of brocade, 3 million ingots of gold, and 8 million ingots of silver.'•» When the artificial buffer state of Ch'i was dissolved in 1137, its treasuries contained 98.7 million strings of cash, 2.7 million bolts of silk, 1.2 million ounces of gold, 10.6 million ounces of silver, and 900,000 bushels (shih) of grain." A part of these huge reserves, particularly the textiles and precious metals, may have already been in store under the Sung and then transferred to Ch'i, but in any case it constituted a formidable accumulation of capital. This was steadily increased over many years through the annual payments from the Sung and through the taxes collected from the Chin population. We may well ask what became of these vast sums and how they were expended. In 1191 the Chin treasury had only 60,000 ounces of gold (1,200 ingots) left and 552,000 ingots of silver. It seems that an inordinate amount of government expenditure was taken up by gifts. On every possible occasion, gifts were handed out lavishly according to status. Funerary gifts, rewards for meritorious soldiers from general down to the last assistant mou-k'o, and marriage gifts to members of the imperial clan and the court are mentioned over and over again in the Chin history. In 1142 an imperial relative who had been a victorious general received 1,000 slaves, 1,000 horses, 1 million sheep, 2,000 ounces of silver, and 2,000 bolts of textiles. At the opposite end of the ladder we read about 53 CS, 46, pp. 1027-31. 34 Ta Chin kuo cbih, 32, p. 236. 55 Liu Yu shih chi (Hsiieh hai lei pin ed.), p. 36b. Cambridge Histories Online © Cambridge University Press, 2008 ECONOMIC CONDITIONS 303 gifts of a few strings of cash. When it was reported to the emperor in 1167 that the city jails in Peking were empty, he ordered that three hundred strings of cash be spent on entertainments for the officials of the capital. These generous gifts from the imperial treasuries meant that a great part of the nonconsumable goods (money and precious metals) was in circulation instead of in reserve, and thus eventually the government spending affected even the small shopkeepers or the musicians who played at banquets. The same is true for the money that went into salaries. A part of this money, whether gifts or salaries, did find its way back to the public treasuries by way of taxes. The real problem was that the state economy was based not only on money but also on taxation and expenditure in kind, particularly in grain and rice. The production of these basic commodities could not easily be increased and was subject to the whims of nature (drought and flood). Large grain stocks were, however, vital, to expenditure in kind in normal times, and more especially in times of crisis when a huge army had to be fed. There are a few figures available for normal years that show the margins within which public expenditure operated. In 1171 the total stocks of grain amounted to 20.7 million shih. The annual revenue in grain was 9 million shih, of which 7 million went toward normal expenditure, mainly official salaries. Of the surplus, 1 million was spent on relief to the population in areas hit by natural disasters. This means that the government granaries held a reserve sufficient for two years. In 1180 the money income from taxation was 20 million strings, of which 10 million were spent. This left a considerable surplus on the money side. A few years later, in 1192, the grain and rice stores were 37.863 million and 8.1 million shih, respectively, amounting to a reserve adequate to supply the officials and the army for five years. The money in hand had reached by this time 30.343 million strings, a sum that could cover all normal expenditures for slightly more than two years. But a series of bad harvests or a war, or a combination of both, could quickly draw down these reserves, and this is what happened a few years later. We have already discussed the importance of the monopoly revenue for public finance. We now shall describe briefly the major taxes. The land tax was a tax in kind that was collected twice a year, in summer and in autumn. The tax rate was determined according to the land held by the taxpayer. For every tnou of fields of the best category, 5.3 sheng (about 3 liters) of grain had to be paid in one year, 0.3 sheng in summer and 5 sheng in autumn, plus a bundle of straw weighing 15 pounds. It is not clear whether the straw (or hay) was used as fodder for animals or for building and repair purposes. Tenants of government land had to pay a rent instead of tax, but this was a difference only in name. Also, tenants of government-owned buildings in towns had to pay rent. Cambridge Histories Online © Cambridge University Press, 2008 304 THE CHIN DYNASTY In addition to the land tax, a property tax in money (wu-li ch'ieri) was levied. It was based on an assessment of the total property, not only fields, but also gardens, orchards, and tree plantations; and number of houses, horses, oxen, and other cattle. The property tax for military colonists was based on the number of oxen (niu-t'ou shut). Contrary to the usual privileges of tax exemption for officials, this property tax had to be paid by everybody from the highest officials down, and there was no privilege for the Jurchen population, either. The property tax caused much resentment because everything depended on the official evaluation of property. The property census was taken mostly by government officials, but once by local worthies because of the complaints about the officials' ruthless assessments. The property tax system also suffered from the fact that the condition of the property could change in the period between two evaluations. Frequently we read that households had been impoverished but were still assessed on their former property, whereas the newly rich had to pay tax on only the much smaller property with which they began. Changes in economic status were therefore common and indicate the existence of some degree of social mobility. We do not have figures on how much had to be paid in respect to what kinds of property, but it is known that in one year (1198) the proceeds from this tax totaled 2.5 million strings. This was considerably less than the expected target sum of over 3 million strings. Almost one-fifth of the expected income had to be written off for reasons of poverty and inability to pay. A commercial tax was introduced in 1180. The tax rate was 1 percent on gold and silver and 3 percent on all other commodities. This was later raised to 3 percent for gold and 4 percent for all other commodities. An additional monetary tax was levied in times of military emergencies, for the first time in 1163. Nofiguresare known, but our sources indicate that extortion must have been rampant. Finally, the exemption payments in lieu of labor service or postal horses were a source of money income for the government, but here too no figures are available. There can be no doubt that within the monetary sector, by far the biggest item of revenue was the proceeds from the salt monopoly, but the survival of the state depended largely on grain production, which was the most variable and precarious factor in the public economy. SCHOLARSHIP, LITERATURE, AND THE ARTS The dazzling brilliance of Southern Sung civilization that even after the Mongolian conquest of China impressed foreigners like Marco Polo certainly overshadows the achievements of Chin China. We may, however, ask to what extent this might reflect a value judgment of Chinese intellectuals in later Cambridge Histories Online © Cambridge University Press, 2008 SCHOLARSHIP, LITERATURE, AND THE ARTS 305 centuries when, particularly under the Ming, the period of foreign domination in north China was regarded as a barbarian interruption between Sung and their own time. It was only under another foreign dynasty, that of the Manchus, that Chin authors received more attention and their works were reprinted or collected from various sources. One indicator of the position of the Chin period in Chinese intellectual and literary history is the inclusion of Chin works in later anthologies and florilegia. Here we find a notable absence; it does not seem that any of the Chin Confucians was accorded a place in the Great Tradition of Confucian scholarship. It is hard to decide whether this is due to the bias of later generations or to a real difference in quality. In sheer bulk the output of Chin intellectuals was considerable, exegetical scholarship as well as poetry and prose.' 6 Unfortunately, the great majority of literary works produced under the Chin are known to us only by their titles; the works themselves have been lost. This again points to a conscious neglect in regard to posterity, and it remains a matter for speculation whether or not this was justified. Tradition, after all, also implies selection, and the selecting process in Chinese tradition bypassed the Chin. The towering figure of Chu Hsi (1130—1200) eclipsed whatever the more conventional exegetes in north China had to contribute. Intellectual conservatism, content with rehashing T'ang and Northern Sung thought, seems to have been a characteristic of Chinese philosophy under the Chin. Although the states of Sung and Chin were not hermetically sealed off from each other, the free flow of communication and scholarly contacts was drastically reduced, and certainly not many Southern Sung books became known among Chin intellectuals. The basic works by Chu Hsi were, in fact, introduced to the north only after the fall of Chin in 1235, when a Sung scholar, Chao Fu, had been taken prisoner by the Mongols.57 The provincialization of the northern scholars was, therefore, partly the result of this lack of communication. But this was not the only factor that might explain the relative unproductivity of the north. Apart from the destructive effects of prolonged warfare during the first decades of the Chin state, the brain drain that followed the transfer of the Sung capital from K'ai-feng to Hang-chou must be taken into account. K'aifeng, for two centuries an imperial capital, was reduced to the status of a provincial town, and for many years Sung visitors who passed through K'ai56 The new two-volume edition of the Chih shih published by the Kuo-feng yen-chiu yiian (College of National Defense) in Taiwan in 1970 contains in vol. 2 a bibliography of works written under the Chin. This bibliography, compiled by Yang Chia-lo, lists no fewer than 1,351 titles (including inscriptions on stone). 57 On the slight acquaintance of Chin scholars with Southern Sung Neo-Confucianism, see Wing-tsit Chan, "Chu Hsi and Yiian Neo-Confucianism," in Yiian thought: Chinese thought and religion under the Mongols, ed. Hok-lam Ch'an and Wm. Theodore de Bary (New York, 1982), pp. 199—200. Cambridge Histories Online © Cambridge University Press, 2008 306 THE CHIN DYNASTY feng deplored the decay of ancient splendor. The depressed intellectual climate that prevailed under these conditions changed only gradually. Under Emperor Hsi-tsung, Confucian state rituals were introduced in which the emperor himself took part, and in 1140 a forty-ninth-generation descendant of Confucius was enfeoffed as a duke. The adoption of a Chinesestyle bureaucracy considerably increased the employment opportunities for Chinese intellectuals during the following years, and toward the end of the twelfth century, intellectual and artistic life had recovered to a great extent. By the end of the twelfth century, Chinese, sinicized Jurchen, and Khitan scholars emerged who were the product of the Chin examination system and who had profited from the long period of peace under Shih-tsung. These men represented Chinese culture in its broadest sense when the Mongols invaded the country. The gradual debarbarization of Mongolian rule in northern China is the great achievement of these men who had been formed under the Chin and kept the Chinese tradition alive. Even though none of them may have reached the intellectual stature of their Southern Sung contemporaries, the survival of Chinese values in a period of unprecedented violence and turmoil must be credited to the Chin intellectuals, regardless of their ethnic origin. In the natural sciences as in philosophy, the Chin contribution was more in the line of conservative traditionalism than of innovation. No new theoretical approaches changed the scientific heritage taken over from the Northern Sung. Astronomy, which because of its inherent cosmological elements had always been a practical science in China, found its place in the Astronomical Bureau. Several calendars were established under the Chin, and the last one, adopted in 1180, continued to be used long into the Mongolian period before it was replaced by a new system in 1281. Several works on geography were compiled, and a few travelogues on parts of the Chin empire written, but here, as in the field of astronomy, the emphasis was on description rather than on new theoretical approaches. In contrast, Chinese medicine flourished under the Chin (and early Yuan), a development that may be linked with the general interest in Taoist speculation in the north (see the next section).'8 Chinese literature in the Chin state has suffered from the same neglect as has Chin scholarship. Much was lost in later centuries, and only a handful of collected works by individual authors have been preserved. To these must, however, be added the many individual poems included in compilations of 58 The Chin contribution to the natural sciences was studied by Mikhail V. Vorob'ev, "O estestvennykh naukakh v chzhurchzhen' skom gosudarstve Tszin'," in Iitoriia, kul'tura, yazyki narodov Vostoka, ed. Iu. A. Petrosian (Moscow, 1970), pp. 145-9. Chinese medicine under the Chin is described in Jutta Rail, Die pier grown Medizinischulen der Mongolenzeit: Stand und Entwicklung der chinesischen lAedizin in der Chin- und Yiian-Zeit (Weisbaden, 1970). Cambridge Histories Online © Cambridge University Press, 2008 SCHOLARSHIP, LITERATURE, AND THE ARTS 307 the Ch'ing period. Altogether well over 5,500 poems by almost four hundred authors have been transmitted in these collections, a remarkable output if one considers the relatively short duration of the Chin state and its smaller territory and population in comparison with the Sung. Poetry and prose in the classical language continued to follow the examples set by Northern Sung literary figures, above all by Su Tung-p'o, who was highly regarded by Chin writers throughout the twelfth century. Literary fashions developed in Southern Sung China apparently did not spread to the northern state of Chin. In style and form, Chin poetry continued T'ang and Northern Sung patterns. Chinese literary historians have pointed out that the Chin poetic achievement reached its peak at a time when the Chin state was already on the decline. Yuan Hao-wen (1190-1257), a distinguished Chin writer who lived into the Mongolian period after the fall of Chin, compiled the anthology Chung-chou chi (Collections of the central districts), in which he assembled over two thousand poems by Chin authors. It includes not only works by authors born under Chin rule but also those by Sung authors who had taken the Chin side by accepting official appointments from the Jurchen, a fact that later exposed Yuan Hao-wen to blame by overzealous literary critics. The great esteem in which the famous eleventh-century literati of Northern Sung were held perhaps cannot be explained by aesthetic considerations alone; there might well be underlying political reasons. It was not fortuitous that writers like Su Tung-p'o, Ssu-ma Kuang, Ou-yang Hsiu, and Huang T'ing-chien, to name only a few, belonged to the so-called conservative party that opposed the new policies advocated by Wang An-shih and his followers. During Hui-tsung's reign the conservatives were not only out of power, but even their works were, for some time, proscribed. When the Chin took the Sung capital in 1127, they attributed the defeat of their adversaries to the disastrous and, in their eyes, insincere politics of Ts'ai Ching and his clique, who were partisans of the reform policy inaugurated by Wang An-shih. After the fall of the capital, Chin emissaries looked for books and manuscripts of the conservatives, whereas they threw away the works by Wang An-shih that they found in the metropolitan libraries." Yuan Hao-wen himself refused to serve the Mongols and thus continued to consider himself a loyal subject of the now extinct Chin. His refusal may even have enhanced his moral stature among the Chinese intellectuals, so that he became the leading figure in north China during the early years of Mongolian rule. He began to compile the Chung-chou chi shortly before the fall of the 59 Toyama, Kinchoshi ientyu, pp. 5 9 4 - 6 1 8 . The Chin had wished to set up Ssu-ma P'u, a grandnephew of Ssu-ma Kuang, as puppet emperor. Only when he refused did they appoint Chang Pang-ch'ang instead. Cambridge Histories Online © Cambridge University Press, 2008 308 THE CHIN DYNASTY Chin state, taking as his basis two earlier collections of Chin poetry. His method, collecting the works of the writers of a single dynasty, was an innovation, as was the inclusion of short biographies of the authors. Many later anthologies followed his model. The chief aim of Yuan Hao-wen's anthology was to perpetuate Chinese literary values through a time of troubles and to supply biographical information on persons whose works had, in his opinion, enough merit to be transmitted to posterity. In other words, not only literary values but also moral and political criteria determined for Yuan whom he should include in his anthology. In this latter respect he acted like the historian that he was, and indeed some of the biographies supplied by him were later incorporated into the Chin sbib.60 Apart from its value as a historical source, the Chung-chou chi allows an appraisal of Chin poetry, both regular poems and songs {tz'u). Chinese critics in later centuries have had an indifferent opinion of the literary qualities of Chin poetry. It is hard to decide where the bias ends and the objective criticism begins. There seems, however, to be general agreement that Chin poets were, as a rule, straightforward and simple, rejecting overartistic formal experiments. A new note appeared after the Mongolian conquests of 1215, when the Chin state was reduced to a fraction of its former territory and the fall of the dynasty loomed darkly on the horizon. To sensitive minds the end of the world seemed imminent, a total absorption of civilized humankind into unspeakable barbarism. Later Chinese literary critics have labeled the spirit of the poetry that was occasioned by the apocalyptic events between 1215 and 1234 as death and chaos (sang-luan). There are few poems in Chinese literature that emanate such hopelessness and helplessness as those written by Yuan Hao-wen and some of his contemporaries in 1233 when the Chin state finally collapsed. The collection and anthologies mentioned earlier do not, however, include one remarkable corpus of poetry that cannot be left out of consideration if we wish to get a clearer view of the poetic achievements under the Chin. This is the poetry of the Taoist writers, which is preserved not in the usual literary collections but in the Taoist canon. Written sometimes in a curious mixture of classical and vernacular Chinese, these poems are mystical in content and quite different in subject matter from the poetry of non-Taoist writers. This kind of religious poetry has not yet been studied and is never even mentioned in the histories of Chinese literature. But it deserves our interest as both poetry and an expression of religious feelings among the followers of a 60 On Yuan Hao-wen and his Chung ebou chi, see Hok-lam Chan, The historiography of the Chin dynasty: Three studio (Wiesbaden, 1970), pp. 67-119. Cambridge Histories Online © Cambridge University Press, 2008 SCHOLARSHIP, LITERATURE, AND THE ARTS 309 religion that had been widely adopted by the masses and even by nonofficial intellectual circles. Another unorthodox genre of literature under the Chin was the chantefables or medleys. The Chinese term is chu-kung tiao, "various modes," a name that points to the strong musical element in the genre. They consisted of shorter or longer suites of songs, each sandwiched between an introduction and a coda. The songs were grouped by mode and did not just repeat a single tune. Moreover, spoken and sung parts were interspersed so that in some ways the chantefables may be regarded as precursors of Yuan opera. The origin of the chantefable in China is not known, but certainly they go back to the eleventh century. In contrast with the arias of Yuan operas, the songs of chantefables are not so much lyrical first-person expressions as dramatized descriptions. They were usually performed by female professional entertainers, and indeed they must have been an urban art form, a sort of music-hall entertainment with a strong mimetic element in addition to the song and percussion accompaniment. The surviving fragments of the chantefables are also characterized by their satirical and comical elements. We know the titles of a great number of chu-kung tiao, and it seems clear that historical romance and erotic themes constituted their subject matter, but unfortunately only one text has been preserved in total, a chantefable version of the famous love story of the "western pavilion," Hsi-hsiang chi, attributed to a certain Tung Chieh-yiian. 6 ' All other Chin chantefables are known only through excerpts and fragments, but one at least, the story of Liu Chih-yuan (895-948), the founder of the Later Han dynasty in the tenth century, has been sufficiently well preserved to justify a translation.62 Many studies have been devoted in recent years to the chantefables of the Chin period, not only because they are a "missing link" leading to Yuan operas, but also for their intrinsic literary value. They represent a genre of literature that combined colloquial and literary language and that must have been extremely popular among the urban audiences in Chin China. Another literary form that can be traced back to the Chin period are the playlets known under the Chinese name of yuan-pen. They were not exclusively found in northern China, but it seems that they, like the chantefables, flourished under the Chin. The name yuan-pen means "scripts from the playhouses" and is frequently applied to Yuan operas as well. They were originally a variety entertainment combining dance, songs, and burlesque, 61 See Tung Chieh-yiian, Master Tung's western chamber romance (Tung hsi bsiang chu kung tiao): A Chinese chantefable, trans. Li-li Ch'en (Cambridge, 1976). 62 M. Dolezhelova-Velingerova and James I. Crump, rrans., Liu Chih-yuan chu-kung-tiao: Ballad of the hidden dragon, Oxford Library of East Asian Literatures (Oxford, 1971). The manuscript of the Chinese text was discovered in the ruins of Khara Khoto in Ning-hsia Province in northwestern China. Cambridge Histories Online © Cambridge University Press, 2008 3IO THE CHIN DYNASTY sometimes with a strong admixture of parody. The little we know about the Chin yuan-pen shows that they were mostly farcical to the point of bawdiness, and it must be regretted that of some seven hundred titles known today only a tiny fraction has survived, and these incompletely.63 Many of the comical sections of Yuan opera may be regarded as an elaboration of earlier Chin yuanpen, again a sharp contrast with the more serious and detached mood of Chin literature in classical language, in which humor is notably absent. Whereas the classical literature of the Chin did not greatly influence later ages, its colloquial literary forms and performing arts did become an integral part of China's literary heritage. The Jurchen minority did not contribute to these developments. There were a few educated Jurchen who had fully absorbed Chinese culture and who distinguished themselves as poets in the classical language, but their number was insignificant, and none of the great literary figures of the Chin was of Jurchen origin. It seems that the Jurchens' acceptance of Chinese culture was eager but more passive than active. There was also no attempt to preserve the Jurchens' oral poetry by translating it into Chinese, and so the songs of the Jurchen bards have been lost forever. Only here and there can we get a glimpse of Jurchen folk poetry, for example, from a shaman's enigmatic curse on a murderer.64 The song that Emperor Shih-tsung improvised in 1185 in his native language was a poem of praise for his ancestors and their hardships in founding the empire, but its Chinese version recorded in the Chin shih reads like a stiff and formal composition made up by a Chinese translator, and it must be presumed that the original Jurchen text was much more colorful and epic. 6 ' On the other hand, several Chinese works were translated into Jurchen. None of the original Jurchen books has survived, and we know only the titles of the translated Chinese books. The titles show that in addition to the Confucian classics, the translated works deal for the most part with Chinese statecraft and strategy. In other words, the principle of selection was to provide the educated Jurchens with those books from which they could learn the essentials of Chinese scholarship and ethicopolitical rules of conduct. This excluded belles lettres and fiction. Nevertheless, translation activities under the Chin were certainly more extensive than they had been under the Liao and were to be later under the Mongols, a fact that shows the eagerness of the Chin court, particularly that of Shih-tsung, to acquaint their own people with the essence of Chinese culture. A word should be added here on book printing under the Chin. No 63 J. I. Crump, "Yuan-pen, Yuan drama's rowdy ancestor," East and West, 14 (1970), pp. 473—91. 64 CS, 65, p. 1540. 65 CS, 39, pp. 8 9 1 - 2 . Cambridge Histories Online © Cambridge University Press, 2008 SCHOLARSHIP, LITERATURE, AND THE ARTS 311 Jurchen print has been preserved, but there are quite a few extant specimens of Chinese books printed under the Chin. These show that the craftsmanship of carvers and printers retained the high standard it had reached under the Northern Sung. Indeed, some Chin prints can stand comparison with the finest books printed in Southern Sung times.66 The same cannot be said for painting and calligraphy,67 although a definitive appraisal of the pictorial arts under the Chin is still lacking. Art historians in China, Japan, and the West have always been attracted by the indubitable excellence and variety of styles that mark the Southern Sung. But this attraction is in itself evidence that northern China in the twelfth and early thirteenth centuries did not have much to offer that could be compared with Southern Sung art. A fourteenth-century work on painting lists forty-seven painters of the Chin period, but none of them figures prominently in the current histories of Chinese painting. Moreover, this work seems to have been based on written records rather than on actual inspection of the paintings.68 The father of Emperor Chang-tsung and the Hai-ling emperor are said to have painted. Another imperial clan member among the painters was Wan-yen T'ao69 (1172-1232), who was also a distinguished poet and a friend of literati like Yuan Hao-wen. A few more Jurchen painters are listed and also two Khitans, one of them Yeh-lii Lii (n31-91), the father of Yeh-lii Ch'u-ts'ai. But the great majority of painters were Chinese, as could be expected. Emperor Chang-tsung was a great patron of the arts. He took a lively interest in the former Sung imperial collections that had been captured in 1127. Many surviving T'ang and Northern Sung paintings still today bear the seals of Chang-tsung, for example, the well-known scroll "Admonitions to the imperial ladies" preserved in the British Museum. Chang-tsung also appointed as director of his collections a well-known artistic and literary figure of his time, the painter-poet Wang T'ing-yiin (1151-1202).?° The emperor himself was an active calligrapher who tried to emulate the Sung emperor Hui-tsung as both an artist and a patron of the arts, and he even imitated Hui-tsung's style of handwriting, as is shown by the extant colophons written by him. While waiting for a definitive appraisal of Chin painting, we suggest that the styles imitated under the Chin were those of 66 K. T. Wu (Wu Kuang-ch'ing), "Chinese printing under four alien dynasties," Harvard Journal of Asiatic Studies, 13 (1950), pp. 4 5 3 - 9 . 67 For a basic survey of Chin painting and calligraphy, see Susan H. Bush, "Literati culture under the Chin (1122-1234)," Oriental Art, n.s., 15 (1969), pp. 103—12. 68 Hsia Wen-yen, T'u huipao Men (KHCPTS ed.), 4, pp. 9 3 - 9 6 , 129. 69 Susan Bush reads the personal name of Wan-yen T'ao as Shou; see her "Literati culture," p. 112, n. 5. 70 On Chang-tsung as a collector and calligrapher, see Bush, "Literati culture," pp. 103—4; and Toyama, Kincho shi kenkyu, pp. 660—75. Cambridge Histories Online © Cambridge University Press, 2008 312 THE CHIN DYNASTY the great masters of Northern Sung, above all the academy painters, in other words, the painters whose works had been transported to the north and incorporated into the Chin imperial collections. One thing is certain, however: The ideal of the gentleman painter, of an educated nonprofessional artist who combined literary training with the art of the brush, seems to have become just as fashionable under the Chin as in southern China. This kind of painting - scroll painting on paper or silk - was in every respect an art of the elite. No self-respecting gentleman painter would ever have condescended to adorn the walls of a temple with a fresco. This remained the reserve of professional painters, but just because of their professionalism, most of the men who created such works have remained anonymous. It is sometimes even impossible to date accurately wall paintings and other decorative pictorial works such as woodcuts. The same is true for sculpture. Unlike the situation in Japan, the sculptor has mostly remained an anonymous artist in China. And yet the production of sculpture in northern China must have been considerable, particularly in view of the many Buddhist and Taoist temples that were built and that had to be decorated and provided with statues. The Liao court had previously patronized Buddhism, and Buddhist art consequently flourished. This development continued after the founding of the Chin dynasty. An interesting feature of Chin sculpture is the frequent use of stone and marble, which finds no counterpart in Southern Sung China. Stylistically, Sung traditions were maintained in north China. Two major tendencies can be distinguished. One was a simple and archaistic style that tried to imitate T'ang art, sometimes so successfully that Chinese sculptures manufactured under the Chin or Yuan and even Ming are passed off by art dealers in our own century as T'ang works. The other was more pictorial, broad, and fluent; one art historian has termed it "a kind of plastic baroque."7' Since 1949 some ancient monuments from the Chin dynasty, both sculptural and architectural, have been excavated or recovered and described in Chinese archaeological journals, so that a more representative corpus of dated or datable works of art is now waiting for a scholarly treatment by art historians. It could also be pointed out that some of the more spectacular parts of modern Peking were constructed in the Chin period. What is known now as the former imperial city with its beautiful landscaped gardens and lakes was a summer residence built in 1179 where both Shih-tsung and Chang-tsung frequently spent the hot months of the year. Under Khubilai khan this former Chin summer resort was made the winter palace of the Yuan emperors 71 Osvald Sirln, "Chinese sculpture of the Sung, Liao and Chin dynasties," Bulletin of the Museum of Par Eastern Antiquities (Ostasiatiska Samlingarna) 14 (1942), pp. 45-64. This study is based chiefly on publications by Japanese archaeologists and art historians. Cambridge Histories Online © Cambridge University Press, 2008 RELIGIOUS LIFE 313 and has remained a part of the Forbidden City ever since.72 As yet there exists no history of the arts under the Chin. It seems, however, not premature to characterize the visual arts in Chin China as conservative and traditional, perpetuating the stylistic heritage of the Tang and early Sung. RELIGIOUS LIFE Already before the establishment of their own state the Jurchen had come into contact with Buddhism, in the Po-hai region. As early as the tenth century a leading Jurchen, A-ku-nai, the elder brother of Han-p'u, the "first ancestor" Shih-tsu, was a Buddhist. When the Jurchens invaded the Liao state, they encountered a flourishing Buddhism receiving considerable patronage from the Liao court. This greatly influenced the attitude of the Jurchen imperial clan toward the Buddhist religion and also the politics of the Chin government. In the imperial family, not a few of the empresses and consorts were pious Buddhists, and the mother of Shih-tsung even became a nun in her later years. In his younger years Shih-tsung himself was attracted by Buddhism but later became somewhat detached, although he continued to favor monasteries and monks. The same is true for Changtsung. The esteem in which institutionalized Buddhism was held by the Chin emperors can be measured by their donations. Both secular and Buddhist sources frequently mention pious donations to the clergy and monasteries. The amount of these was sometimes considerable. When the Hai-ling emperor granted on one occasion 500 bolts of silk, 50 bolts of other textiles, and 500 ounces of silver, this was still a relatively minor gift compared with what, for example, Shih-tsung donated in 1186: 2,000 mou of fields, 7,000 chestnut trees, and 20,000 strings of cash. Agricultural land was frequently given to monasteries in addition to cash, so that some religious communities became huge landowners. Also, the number of temple bondsmen (the double taxpayers mentioned earlier) was large. Feeding monks was another way for the lay population to acquire religious merit, and this too was practiced by the Chin court. The granting of ordination diplomas can also be regarded as a kind of donation because normally the number of candidates for ordination as a monk was restricted. Sometimes thousands of monks were ordained by imperial privilege in one ceremony. All this set an example of patronage for other members of the Jurchen aristocracy and certainly also for well-to-do Chinese. 72 For an historical account of the Chin palaces in Peking, see George N. Kates, "A new date for the origins of the Forbidden City," Harvard Journal of Asiatic Studies, 7 (1943), pp. 180-202. Cambridge Histories Online © Cambridge University Press, 2008 314 THE CHIN DYNASTY On the other hand, this imperial patronage was linked with a strict state control of Buddhism (and of Taoism). In this the Chin followed not only the example of Liao but also that of most Chinese dynasties of the past. The Liao had forbidden the unauthorized ordination of monks as early as 991, and the Chin repeated this prohibition in 1130. Chin legislation, like that of the Sung, included detailed regulations concerning the ordination of monks and nuns.73 Full ordination was possible only after the novices had passed an examination. They had to show that they were able to read one hundred words, selected out offiveset texts from among the Buddhist scriptures. All these sutras were Mahayana texts, including the ever-popular Lotus sutra. The number of candidates was restricted to eighty, and after 1190 the examination was to be held only every three years. A state-controlled hierarchy of Buddhist monks existed, analogous to the civil bureaucracy. For each administrative unit, a monk of exemplary behavior was selected by the government authorities as a monk-official, whose term of office was limited to three years. In cases of minor offenses, these monk-officials had complete jurisdiction over the monks and nuns in their districts, although crimes calling for a punishment heavier than flogging had to be referred to the higher echelons of monastic jurisdiction. Another aspect of state control was the prohibition of the private foundation of temples or monasteries without government authorization. All these limitations and controls were conditioned by the fact that monks were exempted from taxation and corvee labor. On the other hand, the state held a widespread sale of ordination diplomas in times of crisis when the public treasuries needed money. The first cases of that kind are reported from the year 1160 when the Hai-ling emperor was preparing for his invasion of Sung. The price for a diploma varied between one hundred and three hundred strings of cash, a considerable sum of money. Buddhism under the Chin was very much Chinese Buddhism and in this respect followed the heritage of Sung. Hardly any attempt was made to make contact with the Buddhist centers in Tibet, Central Asia, or India, and no pious monks set out from Chin to collect holy scriptures or to visit the places where the Buddha had lived and taught. Also, very few monks from outside China seem to have visited northern China under the Chin. The case of an Indian monk who preached and performed miracles on Wut'ai shan in the 1130s is quite isolated.74 He seems to have been a follower of Tantric Buddhism. The schools that flourished most under the Chin were Ch'an (Zen) and the Pure Land school (Ch'ing-t'u), both long recognized in 73 CS, 55, p. 1234. For monastic jurisdiction, see also Ta Chin kuo chih, 36, p. 275. 74 Nien-ch'ang, Fo ttu li tai t'ung tsai, in vol. 49 of Taisho shinshu daizo kyo, ed. Takakusu Junjiro and Watanabe Kaikyoku (Tokyo, 1929—34), 20, pp. 685b-c. Cambridge Histories Online © Cambridge University Press, 2008 RELIGIOUS LIFE 315 China as orthodox and acceptable to the secular authorities (unlike some other sects). The contribution of Chin to Buddhist scholasticism was negligible. No new sutras were translated from the Sanskrit, and none of the authors of Buddhist works received into the Buddhist canon of the Ming (the one we have today) lived under the Chin. But this does not mean that Chin Buddhism was intellectually static. There was a marked tendency toward syncretism, even among the Buddhist clergy. The leading figure was Hsing-hsiu, also known under his fancy name Wan-sung (1166-1246). He was equally well versed in the Buddhist scriptures, the Confucian classics, and Taoist thought and enjoyed favor from the Chin court. Some of his works have survived but were not included in the Ming canon. He was the spiritual teacher of Yeh-lii Ch'u-ts'ai and also of another remarkable lay Buddhist, Li Chih-ch'un (1185-1231). Li was not only a versatile writer and poet but also a thinker of some depth. He wrote the Ming-tao chi-shuo (Collected opinions for propagating the Truth), a book in which he assembled passages from Confucian and Taoist writers in order to show that their teachings were compatible with the basic Buddhist tenets. This conspectus is perhaps the most distinguished specimen of religious thinking among the intellectual elite of Chin. 75 Lay Buddhism and not imperial patronage was also responsible for printing the Buddhist canon under the Chin. Between 1148 and 1173 printing blocks were cut in Chieh-chou (Shansi) and financed by subscriptions raised through a pious association active in Shansi and Shensi provinces. The Chin canon comprised over seven thousand chtian, of which about five thousand were discovered in 1933 in a monastery in Chieh-chou. 76 Buddhism as practiced among the scholar-officials and at court must be contrasted with the lively sectarian movements. Some of these sects, such as the White Cloud sect and the White Lotus sect, belong to the ever-present undercurrent of messianic movements in China, which lasted right into the nineteenth century. It. has also been said that they were to some extent influenced by Manichaeism (the color white is frequently associated with the Manichaeans). These sects were regarded as heterodox by the established clergy and government authorities alike. Another sect, forbidden in 1190, was perhaps Tantric because its believers worshipped the Sun Buddha Vairocana, the highest of the five Dhyani Buddhas in esoteric Buddhism. 77 75 The text of Ming too chi shuo is preserved inNien-ch'ang, comp., Fotsulitai t'unguai, vol. 49, 20, pp. 695C-99C. 76 Wu, "Chinese printing under four alien dynasties," pp. 4 3 6 - 7 and plate 4. After 1949 the set was transferred to the National Library in Peking. Some texts from the Chin canon, which represent a high level of the printer's art, have been reproduced in facsimile. 77 CS, 9, p. 216. See also Igor de Rachewiltz, "The Hsi-yulu by Yeh-lu Ch'u-ts'ai," Monummta Serica, 21 (1962), p. 40, n. 13, for a bibliography. Cambridge Histories Online © Cambridge University Press, 2008 316 THE CHIN DYNASTY But the most active and, it seems, the most actively proscribed Buddhist sect was the Dhuta sect. Sanskrit dhuta (Ch: t'ou-t'o) means "religious observance," and it seems that the adherents of this movement observed special rituals and taboos of their own. The sect was considered heterodox, and under the Mongols it was even regarded as a separate religion. Unfortunately we know nothing about its intellectual or scriptural background because all the information we have on the Dhutaists comes from hostile sources.?8 Their followers were accused of destroying Buddhist images and oppressing monks, corrupting morals, and, above all, destroying filial piety. But these accusations do not mean much, except that they point to a dissatisfaction with established ways of worship and piety. Many of their followers came from the artisan and merchant classes, and the persecution they suffered (they were proscribed in 1188) was probably caused by the movement's egalitarian tendencies. The contempt shown by their enemies among the clergy and in the government is also expressed by the term that was used to designate the Dhutaists, k'ang-nieh, which might be translated as "chaffy pests." The virulent sectarianism in north China under the Chin is, in any case, evidence of religious vitality as well as socioeconomic disequilibrium among the masses of the population. There were also monk-led rebellions under the Chin. Some of these may have been caused by excessive government control that resulted in resentment among the monks; others were perhaps nationalistic in origin and directed against Jurchen rule, but this remains highly speculative. Messianic rebellions, chiefly of worshipers of Maitreya, the Buddha of the coming millennium, had flared up intermittently in China since the fifth century, and some occurred also under the Chin. A rebellion in 1161 in Hopei may have been a reaction against oppressive government measures imposed in connection with the war against Sung. The revolt in Hopei and western Shantung that broke out in 1171 was certainly messianic in origin and based on a curious interpretation of a passage in the Lotus sutra.^ With regard to Taoism, we can similarly distinguish between the established religion tolerated and controlled by the state, on the one hand, and popular movements that were regarded as heretical, on the other. The same restrictions applied to the Buddhist clergy were also applied to Taoist monks and nuns: state-supervised ordinations, examinations, and a monastic hierarchy with defined jurisdiction. All this did not differ from the corresponding regulations under the Sung. Even the five set texts for the examination of novices were the same as under the Sung: the Tao-te ching and four other holy 78 For a current bibliography on the Dhuta sect, see Rachewiltz, "The Hsi-yu tu," pp. 38—40. 79 CS, 88, p. 1961. Cambridge Histories Online © Cambridge University Press, 2008 RELIGIOUS LIFE 317 scriptures. But popular religious fervor was perhaps even stronger in Taoism than in Buddhism, and several new sects or religions appeared in north China during the twelfth century. They had no connection and little in common with Taoism in the Sung state, in which the magicoreligious school flourished under the "Celestial Masters" (the Taoist popes). The northern schools of Taoism were a quite independent development. The most important and best-known sect is the Ch'iian-chen sect (Integral Realization).80 It left such strong impressions on its contemporaries that already under the Yiian some of the Ch'iian-chen patriarchs appear in dramas as miracle-working saints. This is not surprising because the history of the movement teems with bizarre personalities. The founder was Wang Che (1112—70), a native of Shansi who had failed the chin-shih examination and spent many years in solitary meditation. Because of his eccentricities he was sometimes called "Crazy Wang." In 1167 he took up residence in the mountains in the northeastern part of the Shantung peninsula and attracted a large following. It is perhaps significant that he as well as his major pupils had received a literary education and came from the middle classes. There was nothing proletarian about the Ch'iian-chen patriarchs. Wang's successor was Ch'iu Ch'u-chi (1148-1227), who became the undisputed leader of this school and indeed the most prominent Taoist in Chin China. He is famous for his interview with Chinggis khan in 1219, which resulted in a privileged position of Taoism under Mongolian rule. But Ch'iu, who is better known by his assumed name of Ch'ang-ch'un, "Eternal Spring," had long before been an important figure. He had even once been received in audience by the Chin emperor Shih-tsung, in 1188. Although the Ch'iian-chen sect was basically Taoist, we find a strong element of syncretism in its teachings. From each of the three religions one text was selected as embodying the basic truths of enlightenment: from Confucianism the Hsiao ching (Book of filial piety), from Taoism the Tao te ching, and from Buddhism the Heart sutra, a concise version of the huge sutra of Transcendental Wisdom (Prajnaparamita). Ascetic practices, such as communal fasting, the control of physical desires, and abstention from wine, meat, and women played a great role. All this was to enable a return to original celestial integrity. Purification of the soul and meditation were 80 "Integral realization" is the translation of Ch'iian-chen adopted by Igor de Rachewiltz. Other authors have proposed different renderings of the Chinese term; for example, "perfect realization" (Holmes Welch) and "completely sublimated" (Arthur Waley). On the Ch'iian-chen sect, see the article by Igor de Rachewiltz cited in n. 77. See also Paul Demi£ville, "La situation religieuse en Chine au temps de Marco Polo," in Oriente Poliano: Studi e conferenzt ttnutt all' ls.M.E.0. in occasion! del VII centenario delta nascitadi Marco Polo (1254-1954) (Rome, 1957), pp. 193-236; esp. pp. 196-201. The first Western scholar to study the Ch'iian-chen sect in some detail was Arthur Waley, in his The travels of an alchemist (London, 1931), pp. 13—33. Cambridge Histories Online © Cambridge University Press, 2008 318 THE CHIN DYNASTY considered necessary for eventual salvation and transformation into a "fully realized man," chen-jen, who had left the fetters of the world behind and joined the ranks of the genii. It is important to note that self-discipline and self-realization were fundamental elements of the way to salvation, rather than chemical concoctions or magic spells and charms, as in other schools of Taoism. One would expect that these severe requirements for becoming an adept would have turned people away rather than attracted them, but in fact the sect soon gathered a huge following. The Ch'iian-chen movement found followers even among the social elite, and it is reported that a "mad" adherent of the sect even became the personal tutor to a high-ranking Jurchen.8' The fanatical elements of this new religion made it suspect, particularly in Buddhist circles, and in 1190 the Ch'iian-chen sect was formally proscribed. But this ban did not have the slightest effect, and when the Chin government began to founder under the Mongolian invasions, the movement even gained strength, perhaps because in a period of general instability and chaos, the proposed road to salvation seemed to offer an escape from the somber realities of life. It has been suggested that the Ch'iian-chen sect was a nationalistic movement, directed against the Jurchens and their rule, and therefore a kind of spiritual resistance.82 But if this were true, it was a passive resistance based on noncooperation and voluntary withdrawal and not an activist movement that could easily have led to armed rebellion. The strong intellectual element in the movement can also be inferred from the many writings that the Ch'iian-chen patriarchs have left, including poetry. The Ch'iian-chen writers contributed to the perpetuation of Chinese literary values no less than did their counterparts among the scholar-officials. The corpus of Ch'iian-chen writings was incorporated into the Taoist canon of the fifteenth century.8* Thanks to the literary activities of the Ch'iian-chen patriarchs, we are well informed about the history of the whole movement. This is not so for the other heterodox sects that were founded in north China in the twelfth century, the Grand Unity (T'ai-i), the Primordial Origin (Hun-yuan), and the True Great Way (Chen-ta-tao) sects. The first two were the subject of an ordinance restricting their activities in 1191; the last continued to exist underground and reappeared under the Ming when it was again subjected to government persecution.84 81 CS, 119, pp. 2 6 0 2 - 3 . 82 Ch'en Yuan, Nan Sung ch'u Ho-pti hsin too ehiao k'ao (Peking, 1941; Peking, 1958; repr. Peking, 1962). 83 The monastery where Ch'ang-ch'un took up residence in 1224, the White Cloud Monastery (Pai yiin kuan) in Peking, is indeed the place where one of the two surviving copies of the Taoist canon were preserved and rediscovered in 1933. 84 Rachewiltz, "The Hsi-yu lu," p. 42, n. 19. Cambridge Histories Online © Cambridge University Press, 2008 CONCLUSION 319 Finally, a word should be said about foreign religions under the Chin. We know nothing about Islam or Nestorian Christianity, although these religions may have been practiced by foreigners who had come to north China. But there is evidence of the existence of a Jewish community under the Chin. In 1163 a synagogue was built in K'ai-feng, a fact recorded in a fifteenthcentury inscription. We may perhaps assume that these Jews had come from the Middle East (Persia) via the caravan routes in Central Asia, rather than by sea.8* CONCLUSION The survey of Chin history as given in this chapter may, to some readers, appear too long in view of the relatively short duration of the dynasty. But what is true for the history of any given period or region in China is also true for the Chin: A closer study reveals a surprising wealth of significant and sometimes conflicting details and eventually results in an aversion to generalizations about "China." There existed no "China" as a whole in the twelfth and thirteenth centuries; rather, there was Chinese civilization that took on very different shapes in the north and in the south. Each study of Chin history thus must be a study of regionalism. Seen against the broad horizon of history, the Chin is a period of transition in some respects, of novel experiences in other fields like religion and vernacular literature, and of conservatism in still others. It is hard to decide which characteristic was stronger, the transitional or the conservative. Traditionalism certainly contributed much to the emergence of a feeling of a separate northern identity. Once the Jurchen had given up trying to conquer the south, a sense of growing stability must have pervaded the intellectual elite, and it is strange that there were no widespread defections to the south, to the national Chinese state of Sung. It seems that the Chin state and its ruling elite developed a strong sense of their own legitimacy. They considered themselves to be the guardians of the "real" Chinese traditions of the T'ang and Northern Sung. The surprising endurance of the Chin against overwhelming odds after 1206, the survival of a state sandwiched between the revanchist Sung and the invincible Mongols, can perhaps be partly explained by the increased feeling of legitimacy that must have underlain the loyalty of officials and soldiers, many of whom preferred death to surrender. The Chin confirmed their own inclusion in the legitimate succession of Chinese dynasties in 1203 when the government proclaimed that henceforth 85 For details on K'ai-feng's Jewish community, see Donald D. Leslie, The survival of the Chinese Jews: The Jewish community of K'ai-feng (Leiden, 1972). However, the written sources of the Chin dynasty remain silent, and the building of the synagogue is not mentioned in contemporary sources. Cambridge Histories Online © Cambridge University Press, 2008 32O THE CHIN DYNASTY the element earth would be assigned to the Chin dynasty, succeeding the Sung whose element had been fire.86 This might appear to the modern mind as a senseless speculation, but to every Chinese in the Middle Ages it meant much more: At the latest in 1203 the Jurchen state of Chin had, in its own eyes, become fully Chinese and a legitimate link in the chain of successive dynasties on the highest, if rarefied, level of cosmological speculation. This had taken less than a century to accomplish. But in that century the Chin had traveled the whole way from a rustic tribal society to a state that in many respects could be considered a fully legitimate element in the Chinese world order. Modern historians, too, might well consider the Chin as more than just a barbarian interlude in Chinese history. There can be little doubt that the achievements of Chin, and the conviction of Chin intellectuals that they represented true Chinese values, contributed much to the cultural vitality that enabled them to perpetuate Chinese ways of life under the crushing onslaught of the Mongols. 86 CS, i i , p. 260. For a general study of this problem, see Chan, Legitimation in imperial China. Cambridge Histories Online © Cambridge University Press, 2008 CHAPTER 4 THE RISE OF THE MONGOLIAN EMPIRE AND MONGOLIAN RULE IN NORTH CHINA MONGOLIA A N D T E M O j I N , CA. II50-1206 Tribal distribution Toward the end of 1236 Mongolian armies under the direction of the great general Siibetei crossed the Volga in force, the right wing moving north into the Bulghar lands and the Russian principalities, and the left wing into the north Caucasus and the western Qipchaq steppe. By the time the campaign was called off in 1241, the princes of Russia had been subdued, and perhaps even more important from the Mongolian point of view, the numerous Qipchaq tribes, the last of the nomads of Eurasia to resist them, had been brought under their control. All of the "peoples of the felt tent" from Manchuria to Hungary were now members, through choice or compulsion, of a vast nomadic imperium. The unparalleled unification of the steppe tribes under the aegis of the Mongols in the thirteenth century stands in sharp contrast with the division and discord of the twelfth century (see Map 24).'The level of political and social integration in this period was most often the individual tribe or, at best small, unstable confederations of tribes. The strongest of these confederations, the Qipchaq in the west and the Khara Khitan in Jungaria, were able, it is true, to dominate sections of the steppe and its immediate hinterland, but they were nonetheless pale and imperfect imitations of the great nomadic empires of the past, such as those created by the Hsiung-nu, Turks, or Khazars. This lack of political unity was equally characteristic of the eastern end of the steppe. Some tribes (irgen) of the Mongolian plateau did maintain their internal cohesion, but others disintegrated into their constituent elements - clans (obogb), which then became independent entities competing with one another for pastures, political leadership, and the favors of their sedentary neighbors. Although historical data on the principal tribes of Mongolia, which served as the initial building blocks of the Chinggisid empire, are limited, their geo321 Cambridge Histories Online © Cambridge University Press, 2008 Land over 2000metres (6562 ft) Wffk Land over 500metres (1640ft) 0 500 1000 km MAP 24. The steppe world, 1190 Cambridge Histories Online © Cambridge University Press, 2008 MONGOLIA AND TEMtJJIN 323 graphical distribution and the level of their internal integration are known in broad outline.1 The westernmost of the tribes, the Naiman, were probably of Turkic origin. They inhabited the southern slopes of the Altai Range and the upper course of the Irtysh River. The Naiman were a fairly cohesive and stable group with permanent leaders {khans) until the end of the twelfth century when a rivalry within the ruling family effectively destroyed their unity. Culturally, the Naiman were generally more advanced than were the tribes of central and northern Mongolia, owing to their close proximity to the centers of Uighur civilization located in the Turfan depression and on the northern slopes of the T'ien-shan Range. The Naiman learned various administrative techniques from their sophisticated, sedentary neighbors to the south, and they shared as well a common religious heritage, a form of Nestorian Christianity much influenced by indigenous shamanistic practices. The Kereyid, to the east of the Naiman, also professed Nestorian Christianity under the influence of their neighbors. Throughout the twelfth century they enjoyed stable leadership and some degree of political unity. The core of their territories was in the upper reaches of the Selenga and Orkhon River valleys, a region that for both strategic and ideological reasons had long played a pivotal role in the formation of all successful nomadic confederations in the eastern steppe. The southeastern zone of the Mongolian plateau, the heart of the Gobi region, was inhabited by the Turkic-speaking Onggiid. Their principal settlement, T'ien-te — Marco Polo's Tenduc — was located just north of the loop of the Yellow River near the strategic Ordos Desert, which formed the frontier between the Chin dynasty and the Tangut, or Hsi-Hsia, kingdom. The wellestablished Onggiid princely house, who were firm adherents of Nestorianism, considered themselves, at least nominally, vassals of the Jurchens. The Onggirad, or Khonggirad, to the north of the Onggiid, occupied the western slope of the Great Khingan Range. They were in contact with the Chin dynasty by the late twelfth century and appear at that time to have been rather loosely organized under several different chiefs. The Onggirad regularly exchanged brides with the Mongols, their immediate neighbors to the west, a practice that was continued after the founding of the empire. The steppe region to the south of the Keriilen River was the domain of one of the more powerful and aggressive tribes of the Mongolian plateau, the Tatars, who, at the instigation of the Chin, played an active role in the politics of the steppe. In their efforts to keep the nomads divided and their 1 Louis Hambis, Gengii khan (Paris, 1973), pp. 7—22, provides a succinct discussion of the history and distribution of the peoples of Mongolia in the twelfth century, on which I have drawn freely. Though this work is a popular survey, it rests on extensive research. Cambridge Histories Online © Cambridge University Press, 2008 324 THE RISE OF THE MONGOLIAN EMPIRE own frontiers secure, the Jurchens assiduously promoted conflict between the Tatars and neighboring tribes, particularly the Kereyid and Mongols. Owing to the great success of this policy, feuding among these tribes, carried on with a murderous intensity, became endemic by the mid-twelfth century. Chinggis khan's own tribe, the Mongols, lived between the Keriilen and Onan rivers, that is, just to the north of the Tatars. Throughout the third quarter of the twelfth century they were sharply divided among themselves and thus frequently fell victim to the depredations of their neighbors (see section "Early history of the Mongols"). Of all of the nomads of the eastern steppe, the Mongols were perhaps the most divided and the least likely, it would seem, to provide the leadership that would unify the "peoples of the felt tent." To the northwest of the Mongols were the lands of the Three Merkid. Divided, as their name suggests, into three branches, each with its own leader, the Merkid ranged along the lower course of the Selenga, south of Lake Baikal. Though they occasionally combined forces to undertake raids on their neighbors, the Three Merkid, like other tribes in or near the forest zone - such as the Kirgiz of the upper Yenesei and the Oyirad living immediately west of Lake Baikal - did not possess a high degree of internal cohesion. The social order These tribes of Mongolia, as was the case among the steppe nomads in general, were constructed from a varying number of hypothetically related lineages, obogh, that traced their ancestry back through the paternal line to a putative founder.2 Because its membership was deemed to be of one bone (yasun), that is, descended from a common progenitor, the lineage was an exogamous unit that regulated marriage. Its leadership determined migration routes, distributed pasturelands, organized hunts and raids, and made political decisions concerning entrance into or withdrawal from tribal confederations. A distinctive feature of these lineages is the frequency and ease with which they bifurcated: When lineages increased in number or experienced internal discord, they segmented into sublineages that in turn could multiply and develop into new lineages. Because sublineages were frequently in the process of splitting off the original stem and becoming lineages in their own right and, further, because large, militarily successful lineages acquired many of the characteristics of a tribe, there is considerable vagueness and 2 On Mongolian society and economy, see Sechin Jagchid and Paul Hyer, Mongolia's society and culture (Boulder, 1979), pp. 19—72, 245-96; B. Vladimirtsov, Le rigimesocialdes Mongols: Le Feodalisme nomade (Paris, 1948), pp. 3 9 - 1 5 8 ; and Elizabeth E. Bacon, Obok, a study of social structure in Eurasia (New York, 1958), pp. 4 7 - 6 5 . Cambridge Histories Online © Cambridge University Press, 2008 MONGOLIA AND TEMOjIN 325 confusion in the sources regarding social nomenclature, and this makes it difficult to establish the exact status of an individual segment or its relationship to other segments at any given time. Though defined in genealogical terms, the lineage and the tribe were essentially political entities composed of individuals whose ties of blood were more often fictive than real. In the steppe, common political interest was typically translated into the idiom of kinship. Thus, the genealogies of the medieval Mongols (and other tribal peoples) were ideological statements designed to enhance political unity, not authentic descriptions of biological relationships. This explains why political formations based on such lineages and tribes, themselves arbitrary and temporary constructions, were by nature dynamic, flexible, and unstable. It also explains why nomadic confederations and empires coalesced with such lightning speed and then just as rapidly disintegrated as a consequence of internal tension or external pressure.3 Below the level of the lineage and sublineage was the nomadic camp, the ayil. This was the basic production unit in the Mongols' pastoral economy, normally consisting of a single extended family with its own tents (ger) and herds. For purposes of cooperative labor or local defense, several ayil might temporarily come together to form a giire'en, literally a "circle," that is, a lager or encampment encircled by tents and wagons. Besides the division into descent groups, Mongolian society was separated into several loosely constructed estates - nobles, commoners, and dependents. The nobles advanced claims to such status as the direct descendants of a lineage's eponymous progenitor. This estate provided political leadership for lineages and tribes. There were, however, no strict rules of succession or appointment to positions of authority, and there was considerable latitude in selecting leaders. In the main they were chosen on the basis of their personal attributes and experience, through an informal consensus of prominent members of the lineage. Proper genealogical credentials were, of course, an asset but not a necessity; noble antecedents could always be fashioned to accommodate an able and successful leader. For elevation to the rulership of a tribe or confederation, a more formal procedure was adopted — the convocation of a diet, or khuriltai, composed of nobles and worthies. The junior and collateral lines of the descent group formed the commonality, called the "black hairs" or "black heads," that made up the bulk of the population. The nobles normally possessed larger herds and enjoyed access to the best pasturelands, but no sharp social distinction was drawn between the two estates, nor was there any dramatic difference in life-style. At the bottom 3 See the discussion in Rudi Paul Lindner, "What was a nomadic tribe?" Comparative Studies in Society and History, 24 (1982), pp. 689-711. Cambridge Histories Online © Cambridge University Press, 2008 326 THE RISE OF THE MONGOLIAN EMPIRE of the social scale were found the bo'ol, slaves or bond servants, usually acquired in raids on nearby tribes or sedentary peoples. Both single persons and entire descent groups could be made the dependents of others; that is, an individual could become the personal bond servant of his captor, and a lineage, or part thereof, defeated in battle might collectively become the dependents, or clients, of the victorious obogh. Bo'ol, whether individuals or parts of lineages, were obliged to work for their masters as domestics, herders, or agricultural laborers and to take up arms on their behalf in time of war. Though clearly in a subordinate position, bo'ol were often treated as part of the family and achieved de facto freedom even without formal manumission. The nb'kb'd (singular, nb'kor), or "companions," of major lineage chiefs or tribal khans were another important estate in medieval Mongolian society. They formed the retinue of an aspiring chief or khan, providing him with military and political advice and undertaking in general any commission desired by the lord, from tracking down stray animals to acting as his personal emissaries in diplomatic negotiations. In return for their service, the nb'kb'd received protection, provisions, and food. True boon companions, they fought, lived, ate, and drank with their master. The nb'kb'd were recruited from all social strata. Some were members of the nobility who by free association attached themselves to a ruler of a tribe or lineage not his own, and some were bo'ol who had demonstrated their ability and loyalty on the battlefield, for example, the famous commander Mukhali, whom Chinggis khan elevated from dependent to companion status. Though socially diverse, the nb'kb'd did share one common characteristic: None, so far as we know, were blood kin of their masters. Structurally, then, the tribes of twelfth-century Mongolia were fairly complex entities. Typically, the core of such a tribe was composed of lineages and sublineages that for political purposes claimed a common ancestry based on a communally recognized but artificially contrived genealogy. Attached to the core were various nonkinsmen: lineages associated through marriage, dependent individuals and client lineages made subordinate through military defeat and capture, and nb'kb'd recruited from various external sources. Economic conditions The primary occupation of the inhabitants of the Mongolian plateau was herding domesticated animals. Each of the five principal types of animals kept by the Mongols — horses, sheep, camels, cattle, and goats - had its specific uses, and was valued according to a well-established order of precedence. Horses, the prized possession of the pastoral nomads, were used as military mounts, for transportation, and for herd control. Without them, the Cambridge Histories Online © Cambridge University Press, 2008 MONGOLIA AND TEMOjIN 327 extensive, mobile economy of the steppe nomad would have been impossible. In second place and by far the most numerous of the herd animals were sheep, which together with goats, the lowest category, supplied meat and wool. Third in precedence were camels, employed as beasts of burden mainly in the arid Gobi region to the south. The fourth-ranked long-horned cattle, also found in substantial numbers, contributed meat, hides, and transportation. The famous large-wheeled carts (ger tergen) that carried the tents of tribal leaders were pulled by teams of oxen. All of the animals provided milk, the by-products of which, such as ayiragh (fermented mare's milk, the Turkic kumis), yoghurt, and various kinds of cheese, were staples in the Mongols' diet. Even the droppings of the animals were used, serving when dried as the major source of fuel in the barren steppe. The frequent controlled movement of the herds in search of water and fodder was neither aimless nor unbounded. There was a well-established annual cycle from spring to summer to winter camp; the latter, often shared by several related ayils, was usually a more permanent facility situated in a protected river valley. Because their herds were complex, composed of animals with different rates of movement and divergent food and water requirements, the herder at migration time had to make very fine calculations concerning daily distances traveled, routes taken, anticipated weather conditions, and the like to accommodate the disparate needs of his beasts. Any major migration of their complex herds (together with people and possessions) was thus a complicated problem in logistics requiring careful planning and execution - training that the Mongols were later to put to good use in their far-ranging military campaigns. Given the harsh environmental conditions and the consequent limited capacity of the Mongolian plateau to sustain herds of beasts, it was essential that the nomads distributed themselves evenly over all the available pasturage. One of the crucial functions of the lineage was to facilitate a peaceful distribution, to adjudicate internal disputes over grazing land, and to protect its members from outside competitors. Individual herdsmen therefore thought in terms of guaranteed seasonal access to portions of the lineage's territory rather than of personal, permanent ownership of land - in other words, usufruct rather than proprietary rights. Although the Mongols had a strong commitment to pastoral nomadism, hunting also played a role in their economy. It augmented their food supply, provided furs and hides for clothing or trade, and helped control populations of predators, especially wolves, that were a constant threat to their herds. Large-scale cooperative hunts on the lineage or sublineage level functioned as a form of military training, sharpening individual skills and promoting coordination among formations drawn from various kinship groupings. Cambridge Histories Online © Cambridge University Press, 2008 328 THE RISE OF THE MONGOLIAN EMPIRE In the forested regions of southern Siberia, the relative importance of hunting in the tribal economy increased substantially, so much so that the medieval Mongols always distinguished the "peoples of the forest," hoi-yin irgen, from those who lived in the steppe. Though hunting was their mainstay, the tribes of Siberia - Oyirad, Buriats, Khori Tumad, and so forth had horses, pursued a nomadic (albeit nonpastoral) life-style, and were always considered part of the basic manpower pool on which expanding, steppebased tribal confederations customarily drew. Agriculture was not an independent branch of the nomad's domestic economy, but it was not unknown among the peoples of Mongolia: The Siberian tribes, at least those in the Yenesei region, cultivated fields, as did the Onggiid along the Great Wall. In fact, none of the pastoralists of the Eurasian steppe could claim a purely nomadic economy, unconnected with and untouched by the sedentary world. Indeed, pure nomadism is a hypothetical construct, not a social reality. Pastoral nomadism can most usefully be viewed as a continuum that ranges from near sedentary transhumant communities to a theoretically possible, but never realized, "pure" form of nomadism, that is, a society deriving everything it uses or consumes from its own herds.4 The need for supplemental winter food and forage for the herds and the desire for luxury goods such as tea and textiles was ever present among the nomads. And because their own economy could never fully meet the demand for these goods, nomads were necessarily compelled to turn to their sedentary neighbors for agricultural products. In the case of the tribes of Mongolia, this meant continuous economic interaction with China. The preferred means of acquiring the desired products was the payment of "tribute" in the form of furs, hides, horses, or whatever to the Chinese in return for "bestowals," such as grain, metal implements, and luxury items. If the Chinese, who were largely self-sufficient, refused the proffered exchange, the nomads would threaten force. In short, the steppe people used war and the threat of war to extort the right to offer tribute to the Middle Kingdom. This economic exchange always involved the nomads in an intricate web of political relationships with the Chinese, who used the tributary system as a means of controlling or manipulating the barbarians for their own ends. Thus, from the Chinese standpoint, the purpose of the bestowals — goods, noble titles, or brides - was, on the whole, political rather than economic' Interaction of this nature provided an important impetus, though unintended on the part of the Chinese, for state formation among the steppe 4 Douglas L. Johnson, The nature of nomadism: A comparative study ofpastoral migrations in southwestern Asia and northern Africa (Chicago, 1969), pp. 1-19, discusses the concept of a nomadic continuum. 5 These points are brought out with great clarity by Sechin Jagchid in "Patterns of trade and conflict between China and the nomads of Mongolia," Zentralasiatische Studien, 11 (1977), pp. 177-204. Cambridge Histories Online © Cambridge University Press, 2008 MONGOLIA AND TEMOjIN 329 peoples: Nomads formed larger political unions — confederations — and created formal state structures primarily for conducting relations with and fending off threats from settled states. This is the reason that major nomadic empires normally evolved near the frontiers of sedentary polities and not in the innermost recesses of the Eurasian steppe.6 Early history of the Mongols The broad contours of the ethnogenesis and early history of Chinggis khan's own tribe, the Mongols, are fairly well established.7 The first mention of the ethnonym "Mongol" (more properly, Mongghol) is found in the T'ang dynastic histories in the form Meng-wu. In these texts the Meng-wu are always described as a branch of a larger ethnic grouping known to the Chinese as the Shih-wei. In their earliest identifiable homeland, the area south of the Amur River in the general vicinity of the Lesser Khingan Range, the Shih-wei eked out a meager existence through a combination of primitive agriculture, pig raising, hunting, fishing, and pastoralism. Though they had a reputation as fearsome warriors, the Shih-wei seldom posed a threat to their neighbors, owing to the absence of any form of centralized, tribal authority: Lacking a supreme ruler, political leadership was widely diffused among innumerable heredity chieftains whose power, such as it was, did not extend beyond the local community. According to the T'ang sources, the Shih-wei were counted as dependents of the Turk khaghanate, which held sway over the Mongolian plateau between 553 and 745. Sometime during the tenth century the Meng-wu component of the Shihwei began a movement westward toward the Argun River, where they became subjects of the Khitan (Ch'i-tan), to whom they were linguistically related.8 They continued their migrations westward and southward and fi6 In "The Hsiung-nu imperial confederation: Organization and foreign policy," Journal of Asian Studies, 41 (1981), p. 45, Thomas J. Barfield argues persuasively that nomadic confederations form in response to external stimuli provided by neighboring sedentary states with which the nomads continuously interact and not as a consequence of internal evolution (i.e., as the result of changing class structure). For the contrary view, see S. Nacagdorz, "L'organisation sociale et son deVeloppement chez les peuples nomades d'Asie centrale," Etudes Mongo/es, 5 (1974), pp. 135—44. 7 The most extensive discussion of the ethnogenesis of the Mongols and related groups is L. L. Viktorova's Mongoly: Proiskhozhdenie naroda i istoki kul'tury (Moscow, 1980). See also Evgenii I. Kychanov, "Mongoly v VI—pervoi polovine XII v.," in Dal'nii Vostok i sosednie territorii v srednie veka, ed. V. E. Larichev (Novosibirsk, 1980), pp. 136—48; Paul Ratchnevsky, "LesChe-wei, £taient-ils des Mongols?" in vol. 1 of Milanges de sinologie offerts a Monsieur Paul Demihiille (Paris, 1966), pp. 2 2 5 - 5 1 ; and Louis Hambis, "L'histoire des Mongols avant Gengis-khan d'aprts les sources chinoises et mongoles, et la documentation conservee par RasTd-al-DIn," Central Asiatic Journal, 14 (1970), pp. 125—33. 8 On the early history of the Mongolian language and its relationship to Khitan and Hsien-pi, see Louis Ligeti, "Lc Tabghatch, un dialecte de la langue Sien-pi," in Mongolian Studies, ed. Louis Ligeti (Amsterdam, 1970), pp. 265—308; and D. Kara [Gyorgy Kara], Knigi mongol'skikh kocbevnikov (Moscow, 1972), pp. 8—13. Cambridge Histories Online © Cambridge University Press, 2008 33O THE RISE OF THE MONGOLIAN EMPIRE nally established themselves in the Onan—Keriilen area during the eleventh century. The movement of the Meng-wu from northern Manchuria to eastern Mongolia required an increased commitment to the pastoral elements of their mixed economy. New animals, camels and sheep, were added to their herds of cattle and horses, and the part-time, limited pastoralism of the forest zone gave way to the full-time, extensive pastoralism typical of the steppe. The Mongols' own legendary account of their origin does not indicate their original home and only hints at the migration that brought them to the Onan—Keriilen region. According to this creation myth, which is contained in the Secret history,» the progenitors of the Mongolian people were a bluegray wolf, whose birth was ordained by Heaven, and a fallow doe, whose origin is left unexplained. Departing from an undisclosed clime, the couple crossed a sea or lake, also unnamed, and then occupied the region around Burkhan Khaldun, a mountain now identified with Khentei Khan in the Khentei Range near the headwaters of the Onan and Keriilen. Here was born the only offspring of this union, Batachikhan, a human male, from whom all the numerous Mongolian lineages originated. In the eleventh generation, we are told, a descendant of Batachikhan named Dobun Mergen married a young woman, Alan Gho'a, of the Khorilar lineage. She bore her husband two sons during his lifetime, and after Dobun Mergen's demise she gave birth to three additional sons fathered by a supernatural being riding on a moonbeam. The youngest of the three, Bodenchar, was the founder of the Borjigin obogh, the most ancient of the Mongolian lineages, into which Temiijin, the future Chinggis khan, would later be born. This genealogy of Chinggis khan's early ancestors, although full of fanciful and mythical elements, reveals several interesting features of Mongolian social structure that have important historical implications. First, the line between Batachikhan and Chinggis khan is not reckoned solely on the basis of paternal descent, as one would expect. A woman, Alan Gho'a, by the Mongols' own "official" accounting, is a vital link in the genealogical chain leading from the mythical past to the historical present. Her prominent and honored position in an otherwise exclusively male line of descent points up the high status of women in Mongolian society and anticipates the crucial role they later were to play in the emergence and consolidation of the empire. Second, tribes as well as lineages had mythical ancestors. Although in theory 9 See Francis Cleaves, trans., The secret history of the Mongols: For the first time done into English out of the original tongue, andprovidedwithexegeticalcommentary (Cambridge, Mass., 1982), sec. 1-42 (pp. 1-10) (hereafter cited as Secret history). For a comparison of the Mongols' creation myth with those of the Turks and other Inner Asian peoples, see Denis Sinor, "The legendary origin of the Turks," in Folklorica: Festschrift for Felix J. Oinas, ed. Egle Victoria Zygas and Peter Voorheis (Bloomington, 1982), pp. 223—57. Cambridge Histories Online © Cambridge University Press, 2008 MONGOLIA AND TEMOjIN 331 all members of the Mongolian tribe and its senior lineage, the Borjigin, were consanguines, the membership of each included groups of nonkinsmen. The integration of alien elements - client lineages or groups to whom they were politically allied - was accomplished by the simple expedient of "discovering" common ancestry in the dim and legendary past. Thus as we argued previously, tribes and lineages were to a large degree artificial creations, manufactured by means of genealogical manipulation, the mythical ancestors providing fictive ties of kinship to biologically unrelated segments. Such manipulation, allowing as it did the ex post facto creation of shared lines of descent (and therefore mutual obligations) between disparate kin and ethnic groupings, was a common and essential ingredient in the process of state formation among the steppe nomads. The first of the ancestors of Chinggis khan about whom there seems to be authoritative information is Khaidu, a great-great-grandson of Bodenchar in the senior line. According to data preserved by Rashid al-Din,IO Khaidu exercised a loose suzerainty over several Mongolian lineages and successfully brought the Jalayirs, a tribe living a nomadic life along the Keriilen, under his control. No dates are given, but Khaidu appears to have been active during the last decades of the Liao dynasty (907-1125). It is, however, Khaidu's grandson Khabul who is usually given credit for founding the first Mongolian "state." He took the title khaghan and "ruled all the Mongols."11 During his reign, Rashid al-Din informs us, tensions developed between the Mongols and the Jurchens, the latter apparently viewing the growth of Mongolian solidarity as a threat to their security. To neutralize the danger from this quarter, the Jurchens decided on negotiations to bring the Mongols into their tributary network. They invited Khabul to court, where a diplomatic fiasco ensued. The Mongolian leader, after first gorging himself at a banquet, began drinking heavily and, in an outburst, laid hands on the imperial person, tweaking the emperor's beard, and had to be subdued. Initially, the Chin emperor (the Altan or "Golden" khan of the Mongolian and Persian sources) decided to let the outrage go unpunished and allowed Khabul to depart for home. Having second thoughts, however, he sent officials in pursuit to bring the upstart chieftain back to court for chastisement. The emperor's agents caught up with Khabul only to be enticed into a trap and killed. The Jurchens were deprived of revenge when Khabul died shortly after these events. There is no direct confirmation in the Chinese sources concerning this episode, but the Chin annals note that between 1135 and 1147 the Mongols 10 Rashid al-Din, Jam? al-Tavarikh, ed. B. Karlml (Tehran, 1959), vol. 1, pp. 177-97, supplies the most detailed account of Chinggis khan's immediate ancestors (hereafter cited as Rashld/Katimi). 11 Secret history, sec. 52 (p. 11). Cambridge Histories Online © Cambridge University Press, 2008 332 THE RISE OF THE MONGOLIAN EMPIRE were continuously harassing the northern frontiers, necessitating extensive military countermeasures.12 Although a connection between the Jurchens' abortive diplomatic negotiations with Khabul and the Mongolian raids seems likely, this cannot be established with certainty: The Chin records never associate Khabul by name with the frontier disturbances and one can only infer, on the basis of general chronological considerations, that he was the Mongolian ruler during the decades in question and hence the likely instigator of the raids. In any event, animosity between the Jurchens and Mongols continued unabated following Khabul's death. Shortly afterward, the Mongols, through their matrimonial ties to the Onggirad, became embroiled in a bitter quarrel with the Tatars. The Jurchens lent their support to the latter in an effort to weaken and divide their unruly neighbors. Confronted with these formidable opponents, the Mongols hurriedly convened a khuriltai and selected Ambaghai as the new khan. The founder of the Tayichi'ud lineage, Ambaghai was a descendant of Charakhai, the second son of Khaidu, and thus was a cousin of the deceased Khabul. His reign, apparently a short one, ended when he was captured by the Tatars and dispatched to the Chin court, where he was nailed to a wooden donkey and left to die a slow and painful death. The Mongols then settled on Khutula, the third son of Khabul, as Ambaghai's successor. The return of the office of khan to the Borjigins left the descendants of Ambaghai, the Tayichi'ud, embittered; this was the source of the deep enmity that developed between the two lineages in later years. Of Khutula's reign we know only that he launched a series of unsuccessful attacks against the Tatars in revenge for their part in Ambaghai's execution. Nothing is said of the circumstances surrounding his demise, nor is there any mention of a successor. The decline and disintegration of Mongolian power and solidarity, whatever its precise causes, can be dated with fair assurance to the 1160s, the decade of Temiijin's birth. Although the unity temporarily achieved under Khabul, Ambaghai, and Khutula constituted a step in the direction of state formation, it was nonetheless a hesitant and tentative one, the significance of which is easily overstated. The three were essentially tribal war leaders, or khans, to whom authority was granted for the duration of hostilities; they were not permanent emperors, or khaghans, as anachronistically reported in the Secret history. And certainly none of the available evidence even hints at the emergence at this time of any kind of administrative machinery or lines of authority indepen12 Tamura Jitsuzo, "The legend of the origin of the Mongols and problems concerning their migration," Acta Asiatica, 24 (1973), pp. 9 - 1 3 , collects, translates, and analyzes most of the relevant Chinese material on these conflicts. For additional notices, see Yii-wen Mou-chao, Ta Chin kuo chih (KHCPTS ed.) 10, p. la; 12, pp. 2a, 3a. Cambridge Histories Online © Cambridge University Press, 2008 MONGOLIA AND TEMOjIN 333 dent of and in competition with the traditional kinship structure. 13 The experience and memory of this brief unity may have contributed to the consolidation of the Mongolian nation, but it bequeathed nothing in the way of institutional foundations on which the later empire of the Great Mongols could build. The preliminary work would have to be done anew. Temiijin's family and youth Yesiigei, Chinggis khan's father, was the third son of Bartan Ba'atur, the second son of Khabul khan. A member of the Kiyad yasun, a sublineage of the Borjigin founded apparently by his immediate ancestors, Yesiigei is first mentioned in connection with the Mongols' struggle against the Tatars. At the time of Ambaghai's capture and execution, which must have occurred in the 1150s or 1160s, we are informed that Yesiigei abducted his brideto-be, Ho'eliin, of the Onggirad lineage, from a Merkid tribesman to whom she was betrothed. After marrying the willing Ho'eliin, he participated in the raids against the Tatars, raids presumably initiated by Khutula, his uncle. Upon his return from one of these attacks, Ho'eliin gave birth to Temiijin, the first of her four sons, somewhere in the Onan River valley. The sources provide scant additional data regarding Temiijin's early childhood, not even a precise birth date, although it is virtually certain that he came into the world in the year 1167. 14 At the age of eight Temujin was taken to the Onggirad, his mother's lineage, in search of a suitable wife. For his son, Yesiigei selected Borte, the nine-year-old daughter of Dei Sechen, one of the Onggirad chieftains. Leaving Temujin behind with his future father-in-law as a gesture of good faith, Yesiigei departed for his own camp. On his way he encountered a group of Tatars and joined in their feast. His hosts recognized him as the leader of a costly raid against them and secretly decided on revenge. His food was poisoned, and when Yesiigei reached home three days later, he fell ill and died before his eldest son could be brought to him. Yesiigei's demise soon proved to be more than just a personal loss for his family. During his lifetime Yesiigei, though not a wealthy or influential figure, had acquired a small following of retainers and had, apparently, advanced some claim to leadership among the Borjigins. Following his death, which presumably took place in 1175 or 1176, his relatives and 13 The case for a Mongolian "state" in the twelfth century is argued by N. Iszamc, "L'Etat ftodal mongol et les conditions de sa formation," fctuda Mongoles, 5 (1974), pp. 127—30; the contrary case is made by Jean-Philippe Geley, "L'Ethnonyme mongol a l'epoque pre-cinggisqanide (XII siecle)," ttudts Mongolts, 10 (1979), pp. 5 9 - 8 9 ; and Nikolai Munkuev, "Zamerki o drevnikh mongolakh," in Tataro— Mmgoly v Azii i Evrope, ed. S. L. Tikhvinskii, 2nd ed. (Moscow, 1977), pp. 377—85. 14 On Temiijin's birth date, see Paul Pelliot, Notes on Marco Polo (Paris, 1959), vol. 1, pp. 281—8. Cambridge Histories Online © Cambridge University Press, 2008 334 THE RISE OF T H E MONGOLIAN EMPIRE supporters began to melt away. The Tayichi'ud, who had harbored a grudge against the Borjigins since Khutula's election, were the first to decamp. Yesiigei's near relatives and dependents shortly joined the migration, urged on by Ambaghai's aggrieved widows. This attempt to isolate Temiijin's family, seemingly a deliberate one, was in the end successful despite Ho'eliin's determined efforts to stop the defections. Ho'eliin, her four sons, and a few loyal retainers were left behind in the region of the Upper Onan to fend for themselves. Deprived of their herds and the economic support of their kinsmen, they were reduced to fishing and grubbing roots for sustenance. Temiijin and his brothers thus grew to early adulthood in conditions or poverty and extreme want. It was during this period of trial that the young Temiijin first demonstrated his capacity for cold-blooded calculation and ruthless, decisive action in a violent confrontation with his two half-brothers, Begter and Belgiitei (both of whom were offspring of a concubine of Yestigei). Their difficulties arose initially over a disputed catch of fowl. After a second such incident in which Begter and Belgiitei deprived Temiijin and his full brother Jochi Khasar of a fish that they had hooked, the deeply aggrieved Temiijin determined to have his revenge. With the aid of Jochi Khasar, he trapped Begter on a hill overlooking their camp. Approaching from opposite directions Temiijin and Jochi Khasar coolly shot with their arrows the offending halfbrother. Ho'eliin roundly denounced the two for murdering her stepson, mainly, it seems, on the grounds that this act weakened the family at a time when they were vulnerable to outside attack. Temiijin quietly took his mother's abuse but never expressed any sorrow or feelings of guilt over this grisly affair. As he was to show frequently in later life, Temiijin was not a man to suffer lightly any affronts to his pride; those who challenged his dignity or impeded his path to power were always made to pay a heavy penalty for their imprudent behavior. The period of forced isolation, five or six years in duration, came to a temporary and unwanted end shortly after Begter's death. As Ho'eliin had forseen, the Tayichi'ud, fearing retribution, returned to check on Yesiigei's progeny. Finding that five of the six sons had survived and grown into young adulthood, the Tayichi'ud surrounded the family and demanded that they give up the eldest, Temiijin, presumably as a hostage to guarantee the behavior of the others. Temiijin, who was fourteen orfifteenat the time, fled his mother's camp, concealing himself in a dense wood for nine days. Hunger finally drove him from the forest, and the waiting Tayichi'ud immediately descended on him. Placed in a cangue, he lived as a prisoner among the Tayichi'ud for some time, being rotated to a different ayil every fortnight. This humiliating and extremely uncomfortable existence came to an end Cambridge Histories Online © Cambridge University Press, 2008 MONGOLIA AND TEMOjIN 335 when Temiijin managed to escape with the aid of a friendly Suldus tribesman, Sorkhan Shira, who was a client of the Tayichi'ud. Temujin made his way back to his family, and reunited they sought sanctuary in the mountains south of Burkhan Khaldun. Here life remained difficult - their diet consisted largely of marmots and field mice - and none too secure. Their possessions, meager as they were, soon attracted the attention of some passing bandits, who made off with several geldings. Temujin set out in pursuit of the miscreants and, after various tribulations, returned in triumph with the stolen horses to his appreciative family. This venture is particularly noteworthy in that Temujin began the process of building a personal retinue by acquiring in the course of this quest his first follower and companion, Bo'orchu. The episode, moreover, seems to have considerably increased Temiijin's self-confidence. In any event, shortly afterward he traveled to the Onggirad territory to claim his wife-to-be, Borte, who had been betrothed to him years before. Dei Sechen unhesitatingly recognized his pledge to the long-deceased Yesiigei and gave his daughter in marriage to the determined suitor. By this union Temujin not only gained a wife whose intelligence and strength of character was a major asset in his rise to power, but he also had reestablished ties with an old ally, the Onggirad. His family's period of isolation and rejection was over at last. Ong kban as ally Immediately following his wedding, Temujin turned his attention to the establishment of ties of friendship and alliance with another nearby tribe, the Kereyid. He presented himself to their leader, To'oril, more commonly known by his title Ong khan, using Borte's dowry, a black sable coat, as his offering. Temujin reminded the chieftain that as a former anda, or "sworn friend," of his deceased father, Yesiigei, Ong khan was now "like a father" to him. 1 ' Ong khan acknowledged the legitimacy of this claim and offered to help gather up the scattered people of his new protege, Temujin. The very fact that a person of Ong khan's stature made such a pledge had immediate and beneficial effects: Old servitors who had strayed at Yesiigei's death began to return to the fold, swelling the numbers of Temiijin's followers. Shortly after his return from the meeting with Ong khan, Merkid tribesmen, led by Toghto'a, raided Temiijin's camp and carried off Borte in rather belated retaliation for Yesiigei's abduction of Ho'eliin. Temujin, who had fled 15 On this institution, see Fujiko Isono, "A few reflections on the anda relationship," in vol. 2 of Aspects of Altaic civilization, ed. Larry V. Clark and Paul H. Draghi (Bloomington, 1978), pp. 81—7. Cambridge Histories Online © Cambridge University Press, 2008 336 THE RISE OF THE MONGOLIAN EMPIRE ignominiously to the safety of Burkhan Khaldun when the raid commenced, naturally turned to his patron for assistance in securing Bone's safe release. Ong khan agreed and suggested that they also seek the support of Temiijin's boyhood friend and anda, Jamugha, the leader of the Jadaran, one of the Mongolian lineages. Detailed plans were made, and in 1184, approximately nine months after Borte's capture, the united forces of Temiijin, Ong khan, and Jamugha moved against the Merkid, who were camped on the Kilgho River, an affluent of the Selenga. With little advance warning of the enemy's approach, the Merkid were unable to mount effective resistance and were routed. Borte was recovered unharmed, and much booty was taken. Many Merkid were captured or killed, but the main force under Toghto'a managed to evade the pursuit and found refuge in the forests of southern Siberia. In the aftermath of victory, people in substantial numbers began to flock to Temiijin's standard. Some were relatives, members of the same sublineage, the Kiyad yasun, and others nonkinsmen seeking a position in his fastgrowing retinue. Temiijin and his anda Jamugha at first planned to continue their military partnership and to travel together. However, competition for retainers and latent rivalry over the right to reunify the scattered Mongolian lineages caused tension between the two, and at Borte's insistence Temiijin finally broke with his old comrade and took an independent course. The decision of the Tayichi'ud, Temiijin's former tormentors, to align themselves with Jamugha further served to widen the breach. Following the rift, Temiijin moved to the Upper Keriilen, where his adherents, in the main his own kindred, convened a khuriltai on the shores of the Koko Naur, the "Blue Lake." At this convocation, held sometime between 1187 and 1189, Temiijin assumed the title of khan16 and made known his pretension to the leadership of all the Mongolian peoples. Although in strict genealogical terms, Temiijin, whose line traced its descent through Bartan Ba'atur, the second son of Khabul khan, was nominally junior to the descendants of Okin Barkhagh, the eldest son of Khabul khan, this did not constitute a serious obstacle on the new khan's path to power. His own family's credentials were sufficiently prestigious to establish a basis for his claims of leadership, claims that could be tested and validated in political and military competition with rival lines. To indicate his seriousness of purpose to friend and foe alike, the new khan immediately set about organizing a household establishment appropriate to his new status. From among his companions (nokod) he appointed cooks, 16 The Secret history, sec. 123 (p. 55), states that Temiijin was given the more grandiose title of Chinggis khan (Oceanic Ruler) at this time, but it seems more likely that the latter title was conferred only in 1206, after he had unified all the tribes of Mongolia. Cambridge Histories Online © Cambridge University Press, 2008 MONGOLIA AND TEMOjIN 337 herders, quiver bearers, wagoneers, and chamberlains. Two of his earliest companions, Bo'orchu and Jelme, were given supervisory posts. The household and its staff tended the new khan's personal needs and economic interests and served as the kernel around which the imperial guard and imperial administration later formed. News of Temiijin's elevation was sent to Ong khan, who received the tidings with pleasure, and to Jamugha, who did not. A clash between the two rivals was not long in coming. Jamugha, outraged at the slaying of a younger relative by one of Temiijin's followers, marched his forces on the Mongolian leader's ordo (camp). The outcome of the battle, fought at Dalan Baljud, a locale in the Keriilen basin near the Senggiir River, is variously reported in the sources, but in all probability the day belonged to Jamugha. His victory, however, was by no means decisive; it was only the first round of what developed into a lengthy and bitter struggle. Moreover, immediately after the battle Jamugha's behavior alienated many of his own supporters, in consequence of which, the leaders of the Mangghud and Uru'ud at the head of their respective lineages came over to Temiijin. Their defection must be reckoned as a political gain for the Mongolian chieftain that gave him and his supporters a measure of consolation in the aftermath of defeat. In the years following the battle of Dalan Baljud, which took place in the late 1180s, little is known of Temiijin's activities or his whereabouts until he reappears on the scene in 1196 as an ally of the Chin. What transpired in the intervening years is difficult to establish owing to the vagueness and confusion of the sources. There is, however, one episode of importance that is generally dated to these years — Ong khan's temporary fall from power. According to this reconstruction of events, Ong khan, who had once before experienced such a decline of fortune in Yesiigei's lifetime (i.e., before 1176) was again forced to flee his domain in the face of a rebellion in his own family. On this occasion he left Mongolia and found sanctuary with the Khara Khitan. After a lengthy but unspecified period of exile, Ong khan returned home friendless and in great need. Because of his past services Temujin received the hapless refugee with kindness and helped him restore his authority over the Kereyid sometime in 1195 or 1196. 17 Whatever the precise dates of his exile, Ong khan was certainly back in Mongolia by 1196, for in that year he and Temujin launched a joint attack on the Tatars. 18 The Chin dynasty had 17 See Hambis, Cengis khan, pp. 47, 57. 18 Although the date of the attack on the Tatars is confirmed by Chin sources - see Paul Pelliot and Louis Hambis, trans., Histoin da campagna de Gengis khan (Leiden, 1951), vol. 1, pp. 1 9 5 - 9 — t n e timing of Ong khan's restoration is still in dispute. Paul Ratchnevsky, dnggis-khan: Sein Leben und Wirken (Wiesbaden, 1983), pp. 4 8 - 9 , argues that Ong khan's restoration did not occur until 1197 and that therefore he did not participate in the 1196 campaign against the Tatars. Li Tse-jen, Ch'eng- Cambridge Histories Online © Cambridge University Press, 2008 338 THE RISE OF THE MONGOLIAN EMPIRE turned on their old allies in 1195, and here was a glorious opportunity to wreak vengeance on a common enemy, the Tatars, who had, as Temiijin put it to Ong khan, "made an end of [our] grandfathers and fathers."'9 The proposal was accepted, and in the spring of 1196 the combined forces of the Kereyid, Mongols and Jurchens administered a major defeat to the Tatar tribes in the Onan—Keriilen area. Victory brought in its wake the opportunity to plunder the defeated and to recruit additional followers. It also brought rewards of a less tangible character: In return for their service to the Chin, the Jurchens granted to Ong khan the title "prince" (Chinese: wang; Mongolian: ong) and bestowed on Temiijin a lesser title, ja'ud khuri, the meaning of which is somewhat obscure.20 Although the titles were clearly honorary in nature, they conferred on their holders a measure of prestige, and in the political culture of the steppes, anything that enhanced personal stature was an asset. Once the campaign was concluded, Temiijin for the first time undertook an offensive (ca. 1197) without Kereyid support, a punitive expedition against the Jiirkin, an erstwhile ally who had reneged on their pledge to join the attack on the Tatars. He soundly defeated them, executed their leaders, and made dependents of the survivors. Not only had he taught a memorable lesson in loyalty and military discipline, but in exterminating the Jiirkin ruling strata, who were descendants of Khabul khan's eldest son, Okin Barkhagh, Temiijin had rid himself of the only Mongolian noble line senior to his own (see Table 6). Unification of the steppe peoples Although Temiijin had successfully dealt with the Jiirkin unaided and by now enjoyed a certain stature independent of his relationship with Ong khan, he still required his mentor's support in tackling bigger game. Around 1199 the two launched a joint campaign against the numerous, but divided, Naimans in the west.21 Jamugha, ostensibly reconciled with h'isanda, accomchi-ssu ban hsin chuan (Taipei, 1970), pp. 104—; 107—8, contends that Ong khan participated in the attack on the Tatars, was then exiled, and was subsequently restored to power by Temiijin in 1198. In my view, the exile and restoration of Ong khan took place before 1196, not after. In any event, he assuredly took part in the 1196 assaulc on the Tatars. 19 Secret history, sec. 133 (p. 62). On the importance of vengeance in Mongolian culture, see Larry V. Clark, "The theme of revenge in the Secret history ofthe Mongols," in vol. 2 of Aspects of Altaic civilization, ed. Larry V. Clark and Paul A. Draghi (Bloomington, 1978), pp. 37—57. 20 This somewhat puzzling title, perhaps of Khitan origin, is discussed at length by Paul Pelliot in Notes on Marco Polo, vol. 1, pp. 291—5. 21 There is considerable confusion in the sources concerning the chronology of Chinggis khan's campaigns between 1199 and 1202. I have followed the reconstruction in Hambis, Gengis khan, pp. 6 1 76, throughout this section. Cambridge Histories Online © Cambridge University Press, 2008 MONGOLIA AND TEMOjIN 339 panied them. They confronted one of the two Naiman khans, Buyirugh, who ruled the mountain lineages, at Lake Kishil Bashi on the northern slopes of the Altai. Naiman resistance was soon overcome, and Buyirugh fled northward to the Upper Yenesei. Tayang khan, the ruler of the Naiman of the steppe, who had earlier refused to assist his brother and rival, Buyirugh, now felt himself menaced and belatedly dispatched reinforcements to fend off further Mongolian— Kereyid encroachments. The victors, now on their way home, encountered the fresh Naiman contingents along the Bayidaragh, a river in the southern foothills of the Khangai Range. Because it was dusk when the opposing forces met, the battle was postponed until morning. If the account in the Secret history can be believed, Jamugha, who still secretly nursed a grudge against his anda, cunningly persuaded the weak-willed Ong khan to abandon Temiijin, leaving the latter to face the Naiman troops alone. The Kereyid stealthily departed under cover of darkness, and only at daybreak was their treachery discovered. The conspirators' plans, however, badly miscarried. The Naiman reinforcements did not attack Temiijin as anticipated but moved against the retiring Kereyid chieftain, catching him unawares. Forsaken by Jamugha, the hard-pressed Ong khan had to request help from his recently betrayed comrade. Temiijin mounted a rescue operation that succeeded in extracting Ong khan from his difficult predicament. The latter was naturally grateful to his savior and begged forgiveness. Inexplicably, Temiijin accepted his apologies, restored the Kereyid's properties and peoples, and the two settled their differences for the time being. Following the Naiman campaign, Temiijin fought a brief, indecisive battle with the Merkids, led by his old nemesis, Toghto'a, and then turned his attention to the Tayichi'ud. In 1200 the Mongolian leader, with the assistance of Ong khan, attacked his near kinsmen along the banks of the Onan, inflicting heavy losses on them. Those of their leaders who fell into Mongolian hands were summarily executed. Those spared, mainly women and children, Temiijin distributed among his followers. In the face of Temiijin's continuing successes, Jamugha, again an open enemy, now forged an opposing coalition with himself at its head. Granted the title oigiir khan (universal khan) along the Arghun River in 1201 by an alliance of twelve tribes, which included, among others, the Jadaran, Merkid, Naiman, Oyirad, and the remnants of the Tayichi'ud, Jamugha planned an immediate offensive. Temiijin, forewarned of his intention, went out to meet his enemy in the company of his old but none-too-reliable ally, Ong khan. In the ensuing confrontation, fought along the Keriilen in a blinding storm, the Mongolian chieftain launched a determined and timely spoiling attack that thoroughly disorganized the opposition's advance guard commanded by the Cambridge Histories Online © Cambridge University Press, 2008 34O THE RISE OF THE MONGOLIAN EMPIRE Naiman Buyirugh khan. By the time Jamugha arrived on the field with his troops, the issue had been decided in Temiijin's favor. Seeing the hopelessness of the situation, Jamugha turned on his defeated and demoralized allies, despoiled them, and made off toward the Arghun River. Ong khan pursued the retreating Jadaran, while Temiijin devoted his energies to tracking down and settling with the remaining Tayichi'ud. Battle was finally joined along the Tula River. Though severely wounded in the engagement, Temiijin directed his troops to a clear-cut victory. Vanquished again, their leadership systematically exterminated, the Tayichi'ud were broken once and for all. In the winter of 1201-2 Temiijin contented himself with resting his troops in anticipation of afinalshowdown with the hated Tatars. On the eve of the campaign he issued explicit orders to his army, sternly reminding them that their first duty was to kill Tatars, not to acquire booty. Anyone caught looting before the fighting ended would be subject to harsh punishment. The attack came in the autumn of 1202 and resulted in a crushing defeat for the Tatars along the Khalkha River. The survivors were rounded up, penned, and then mercilessly put to the sword. Only women and children were spared. Another old score was satisfactorily settled; Yesiigei had been avenged; and his son was now master of all of eastern Mongolia. Temiijin's spectacular rise to prominence and power called for a further readjustment in his relationship with Ong khan. The Kereyid, for once reading the situation correctly, at last realized that the Mongolian chieftain was no longer a subordinate, but an equal whose long-term aspirations could be achieved only at the expense of his own interests. Rather than fight it out with his former protege, Ong khan sought a compromise. At Ong khan's behest he and Temiijin renewed and reaffirmed their ties as father and son in a solemn ceremony held on the banks of the Tula in early 1203. In formalizing Temiijin's status as an adoptive son, Ong khan was offering to make the Mongolian chieftain his legitimate heir. Thus the aging Kereyid could live out his remaining days in tranquillity and honor, and his ambitious "son" would in time inherit overlordship of the peoples of central Mongolia. Temiijin promptly accepted, but not surprisingly the offer excited the jealousy of Ong khan's natural son and heir presumptive, Senggiim. The latter tried to dissuade his father from finalizing the accord and vociferously opposed a corollary agreement that would have joined in marriage his younger sister, Cha'ur Beki, with Temiijin's eldest son, Jochi. When his initial remonstrations were rebuffed by his father, Senggiim fell under the influence of the ubiquitous Jamugha, who urged him to act on his own behalf to secure his rightful inheritance. Senggiim concluded that Temiijin would have to be eliminated but hesitated to act without his father's consent. He therefore pressed his rather spineless father on the issue until the weary Ong khan gave Cambridge Histories Online © Cambridge University Press, 2008 MONGOLIA AND TEM0JIN 341 way, agreeing, tentatively at least, to his own son's proposal to kill the ambitious Mongolian chieftain. Senggiim thereupon set his plans in motion. He announced in the spring of 1203 that he was no longer opposed to his sister's wedding to Jochi, in order to lure Temiijin to a celebratory feast at which he intended to assassinate his unwanted new "brother." Temiijin, however, learned of the plot on his way to the celebration and escaped before Senggiim's agents could strike. The sequence of events at this point is somewhat jumbled in the sources but can plausibly be reconstructed in the following fashion.22 Upon receiving information of the plot against his life, Temiijin and a few followers took refuge on the Baljuna (variously reported as a river or a lake) located south of the Khalkha River. Though greatly outnumbered by the pursuing Kereyid forces, his retainers never wavered. Moved by their loyalty, Temiijin drank from the bitter waters of the Baljuna and swore an oath to remember always their steadfastness and fidelity. From there they moved to a nearby locality, the Khalakhaljid sands, where, reinforced by Mongolian contingents, Temiijin engaged the Kereyid troops. The Mongols emerged victorious but suffered heavy casualties. Perhaps still outnumbered, they withdrew downstream along the Khalkha River.23 Temiijin passed the summer in northeast Mongolia searching out new recruits, rallying his forces and parleying with his opponents, whom he reproached for their double dealing. In the autumn he moved back to the west, established his camp between the Onan and Keriilen rivers, and prepared for a decisive encounter with the Kereyid. He surprised his foes, who were camping, it appears, somewhere in the southern Bayan Ula Range and, after a three-day battle, overwhelmed them. Ong khan escaped, only to perish later at the hands of the Naiman. The whole of his people were made dependents and distributed among the victorious army. With central Mongolia, that is, the former Kereyid territories, in his grasp, Temujin now controlled the strategic Orkhon River valley, which gave his nomadic armies easy access to the Ordos Desert and China beyond, as well as to the migration and invasion routes leading through the Altai to Jungaria and thence to the western Eurasian steppes.24 Mongolian dominion over Ong khan's former lands also meant that the Naiman had a new and menacing 22 My description of the events from the time of Chinggis khan's flight from Senggiim to the Mongolian leader's final defeat of Ong khan some months later is based on Francis W. Cleaves's exhaustive study, "The historicity of the Baljuna covenant," Harvard Journal of Asiatic Studies, 18 (1955), pp. 3 5 7 - 4 2 1 , particularly pp. 378—81, 387—92. 23 See also the reconstruction by Ratchnevsky in dnggis-khan, pp. 64—8, who argues that Temujin, on discovering Senggiim's plot, first fled to the Khalakhaljit sands, where he was defeated by the Kereyid; then, with his few remaining supporters, he took refuge at Baljuna, where the oath was sworn. 24 On the strategic importance of this region in steppe history, see Larry Moses, "A theoretical approach to the process of Inner Asian confederation," ttudes Mongols, 5 (1974), pp. 115—17. Cambridge Histories Online © Cambridge University Press, 2008 342 THE RISE OF THE MONGOLIAN EMPIRE neighbor on their eastern frontier. The Naiman khan, Tayang, feeling threatened, proposed an alliance with the Onggiids, hoping to interest the latter in a two-pronged attack against the Mongols. However, the Onggiid ruler, Ala Khush Digid Khuri, wisely declined and soon informed Temiijin of the Naiman's intentions. Despite hesitation on the part of some of his advisers, Temiijin insisted on war with the Naiman, and set about reorganizing his forces on a decimal basis, creating units often, a hundred, and a thousand, for the forthcoming campaign. In May 1204, after appropriate shamanist ceremonies in honor of their battle standard (tugb), the Mongolian contingents marched west to do battle with the Naiman. The armies met on the southern slopes of the Altai Range. The Naiman and their allies — the Merkid and Oyirad, as well as Jamugha and his Jadarans - suffered a major defeat, and the Mongols added many new recruits to their ranks. Tayang fell in the fighting, but Jamugha, who seems always to have abandoned his allies on the eve of a crucial battle, did so again and temporarily avoided capture. He remained at large, wandering in the steppe for several months until he was betrayed by his few remaining followers and brought to Temiijin. After personally inquiring into the prisoner's actions and attitudes, the Mongolian chieftain had Jamugha, his anda and rival, put to death. CHINGGIS KHAN AND THE EARLY MONGOLIAN STATE, 1 2 0 6 - 1227 The khuriltai of 1206 Although dissident elements among the Merkid and Naiman continued to resist25 and the forest tribes had yet to be subjugated, Temujin was effectively the master of the Mongolian plateau by 1205. To formalize his status and to proclaim the formation of this new and powerful confederation, a great khuriltai was convoked in the following year, the Year of the Tiger. The site selected for this solemn occasion was the source of the Onan. No further geographical details are provided, but it seems certain that the diet was held in the vicinity of Burkhan Khaldun, the holy mountain of the Mongols, where their mythical progenitors, the blue-gray wolf and the fallow doe, begat Batachikhan, the founder of all the Mongolian lineages. Unfortunately, there is no authoritative register of the participants. However, from a lengthy roster of individuals appointed to high military office at 25 Merkid and Naiman resistance continued in various forms until 1219, when it wasfinallysuppressed. Cambridge Histories Online © Cambridge University Press, 2008 CHINGGIS KHAN AND THE EARLY MONGOLIAN STATE 343 the conclusion of the khuriltai, which is perserved in the Secret history, it is apparent that representatives of all the recently submitted tribes and lineages were in attendance, as well as Temiijin's n'o'kod and immediate kinsmen. Once assembled, the first order of business was raising Temiijin's white battle standard flying nine horses or yak tails, an act full of symbolic meaning to the steppe peoples. White, perhaps under Manichaean influence, was considered the most auspicious color among the nomads, and the number nine, following indigenous tradition, had long been associated with good luck and other magical properties. Thus the unfurling of the tugh signified and advertised Temiijin's personal good fortune, or charisma, the possession of which, according to the political notions of the steppe, gave him the right to exercise sovereignty over the "peoples of the felt tent." If Rashid al-Din's account of the proceedings is accepted,26 the khuriltai then granted Temiijin the title of Chinggis khan, usually interpreted to mean "Oceanic Ruler," that is, universal sovereign. The title, according to the Persian historian, was publicly conferred on Temiijin by Teb Tenggeri, the chief shaman. 27 Now formally enthroned, Chinggis khan addressed his ethnically diverse adherents - henceforth to be known collectively as Mongols28 — and expressed his gratitude for their aid and loyalty. The ceremonies at an end, Chinggis khan turned to the crucial business of organizing his domain. Administration When on the eve of the Naiman campaign of 1204, Chinggis khan organized his armies on the decimal system, he also created a personal bodyguard (kesig). As initially constituted, the guard consisted of a day watch (turgha'ud) of seventy men, a night watch (kebte'ud) of eighty, and a detachment of braves (ba'atud) numbering one thousand. The kesig as an institution emerged directly out of the household establishment that Chinggis khan had first formed in the late 1180s. Its personnel, like that of the household, was recruited from 26 The Secret history, sec. 123 (p. 55), notes that Temiijin received this title long before the khuriltai of 1206. If this is correct, the latter diet simply reconfirmed a long-held title. However, I prefer Rashid al-Din's version of events, for two reasons: First, it seems more logical that Temiijin would have received such a grandiose title after, rather than before, his unification of Mongolia, and second, the Secret history often employs titles anachronistically. 27 Rashld/Kariml, vol. 1, p. 308. A Mongol of the Khongkhotan lineage, Teb Tenggeri, also known as Kb'kdchii, was an early supporter and influential adviser of Temiijin. He fell from grace shortly after the khuriltai of 1206, when his efforts to provoke discord between Chinggis khan and his brother Jochi Khasar were exposed. By Chinggis khan's explicit order, the once-powerful shaman was killed in a wrestling match with another of the Mongolian ruler's brothers, Temiige Odchigin. On this episode, see Jean-Paul Roux, "Le chaman gengiskhanide," Anthropos, 54 (1959), pp. 424-7. 28 It is typical of nomadic polities that the leading tribe of the confederation gives its ethnonym (which has become politicized) to the whole, even though its membership includes many heterogeneous, unrelated ethnic groupings. Cambridge Histories Online © Cambridge University Press, 2008 344 THE RI SE OF THE MONGOLIAN EMPIRE his nokod. And organizationally, it is difficult if not impossible to distinguish between the two bodies, as guardsmen {kisigden) served simultaneously as protectors of the khan's person and as domestics who tended his private needs and looked after his possessions. In this latter capacity, kisigden held appointments as chamberlains (cherbi), stewards (ba'urchi), quiver bearers (khorchi), doorkeepers (e'utenchi), and grooms (aghtachi). The guards, moreover, supervised the activities of the female attendants and minor functionaries such as camel herders and cowherds; took care of the khan's tents, carts, weapons, musical instruments, and regalia; and prepared his food and drink.29 As Chinggis khan's power and possessions continued to grow, so naturally did the kesig's administrative and economic responsibilities. His dramatic expansion of the kesig from 1,150 to ten thousand men in 1206 was not therefore motivated exclusively by considerations of security or prestige; it was primarily designed to meet the growing administrative needs of the nascent Mongolian empire. And because the guard/household establishment provided both personal service and the machinery through which Chinggis khan administered his rapidly multiplying subjects, territories, and economic interests, it accompanied him wherever he went — on a campaign or on a hunting trip. Thus the "central government" of the early Mongolian state, in essence the imperial guard, was located wherever its sovereign chose to alight. Chinggis khan also created at this time a new office, that of chief judge iyeke jarghuchi), to supervise and coordinate the activities of the recently expanded administrative system. He selected for this post the Tatar foundling Shigi Khutukhu, a member by adoption of his immediate household. The duties of the chief judge were manifold. By Chinggis khan's express order, he was to divide and apportion subject people, that is, decide how they were to be distributed among military units and the khan's family. As his title suggests, Shigi Khutukhu was the highest legal authority in the empire; in cooperation with select members of the kesig he was to try all evildoers and was granted the power of life and death over the guilty. Of equal moment, he was instructed to institute and maintain a "Blue Register" (Koko Debter) in which all judicial decisions, including Chinggis khan's own legal pronouncements (jasagh), were to be preserved in order to establish precedents for future legal decisions. All matters relating to the apportionment of peoples were recorded in it as well. The Koko Debter was therefore a combination of law code and population register. 3° 29 Hsiao Ch'i-ch'ing, The military establishment of the Yuan dynasty (Cambridge, Mass., 1978), pp. 3 4 - 8 . 30 Pavel Poucha, "Uber den Inhale und die Rekonstruktion des ersten mongolischen Gesetzbuches," in Mongolian studies, ed. Louis Ligeti (Amsterdam, 1970), pp. 377—415; and Liu Ming-shu, "Yuan tai chih hu k'ou ch'ing ts'e," Chung-kuo wen hua yen chiu hui k'an, 7 (1947), pp. 101—6. Cambridge Histories Online © Cambridge University Press, 2008 CHINGGIS KHAN AND THE EARLY MONGOLIAN STATE 345 The introduction of regular record-keeping procedures in 1206 was rendered possible by Chinggis khan's farsighted decision, made several years earlier, to transcribe his native tongue into a written form. In 1204, when the Naiman were defeated, an Uighur official, T'a-t'a T'ung-a (Tatar Tonga?), who served at their court, fell into Mongolian hands. After an extended interview with the educated prisoner, Chinggis khan ordered him to adapt the Uighur script to Mongolian and then teach the new alphabet to his sons. Shigi Khutukhu, Chinggis khan's "fifth son" by adoption, must have been among the first to master the alphabet and to write in his native language. T'a-t'a T'ung-a also gave instructions concerning the use of seals in the conduct of official business, which the Mongols quickly adopted.' 1 The military establishment The decimal system that Chinggis khan introduced into the Mongolian armies in 1204 followed well-established steppe traditions. The scale on which it came to be applied was, however, unprecedented. No figures are available on the number of units formed in 1204, but there is a complete order of battle of the Mongolian army as of 1206 preserved in the Secret history and a similar list for 1227 contained in Rashid al-Din's "Collected chronicles."' 2 According to the former source Chinggis khan created ninety-five units of a thousand (.mingghan) from the greatly enlarged manpower pool available to him following the defeat of the Naiman and the consequent surrender of sundry other tribes in 1205—6. To this figure must also be added the 10 mingghan that formed his personal bodyguard. The figures for 1227, the time of Chinggis khan's death, indicate that the number of mingghan mobilized from the tribes of Mongolia had grown to a total of 129. It is highly unlikely that these units maintained for long their paper strength, but in theory at least, the strength of the purely Mongolian component of Chinggis khan's army ranged between 105,000 and 129,000. As the Mongols expanded territorially, the same system was imposed on subject peoples — both nomadic and sedentary — so that by the mid-thirteenth century the size of the Mongolian army, though unrecorded anywhere, was certainly many times the 1206 or the 1227 figure. The mingghan, not the more famous tiimen (unit of ten thousand), was the basic structural element of the army in Chinggis khan's day. When the need arose, ten mingghan were grouped together to form a temporary tiimen. A commander of one of the subordinated mingghan was entrusted with the 31 Sung Lien et al., eds., Yiiansbih (Peking, 1976), 124, p. 3048 (hereafter cited as YS). 32 Sant history, sec. 202 (pp. 141-2); and Rashld/Karimi, vol. 1, pp. 3 9 9 - 4 1 3 . Cambridge Histories Online © Cambridge University Press, 2008 346 THE RISE OF THE MONGOLIAN EMPIRE command of the larger formation, while, however, continuing to lead his own unit of one thousand. All of Chinggis khan's generals, it appears, held permanent commissions as commanders {noyan; plural, noyad) of one thousand and continued to function as such, even when placed in charge of large field armies. Most of the commanders were companions of the Mongolian chieftain and members of his household. For this reason many of the highestranking officers in the army bore such seemingly insignificant titles as shepherd (khonichi), steward, and quiver bearer. As a fighting force, the success of the Mongolian army rested on its mobility, discipline, and maneuverability. They enjoyed no technological superiority over their enemies; they possessed no secret weapon. By nature, all nomadic armies are mobile, but none was as disciplined as the forces of Chinggis khan. The legal pronouncements (jasagh) of Chinggis khan, as far as they have come down to us, seem mainly concerned with military discipline. The extant fragments indicate that the punishments for disobedience were severe and that military units were held collectively responsible for the behavior of individual members. The ability of commanders to coordinate effectively the movement of major formations on the battlefield was another hallmark of the Mongolian military machine. This maneuverability was achieved by frequent peacetime exercises, usually in the form of vast hunting expeditions in which the army participated as units. It was also a matter of discipline. Mongolian field commanders were expected to adhere strictly to prearranged operational plans. If a unit failed to appear at the proper place at the specified time, its commanding officer was promptly punished, no matter what the excuse offered. 33 The Mongolian army, in addition to its primary task, had important administrative functions. The membership of the mingghan created in 1204 and 1206 included soldiers on active service together with their families and dependents. Each mingghan was then a military formation and an organ of local government under the control of the same official, the noyan. The imposition of the decimal system afforded Chinggis khan an opportunity to undermine tribal authority and loyalty and to replace it, to some extent, with military discipline and solidarity. Old enemies such as the Tatars, Kereyid, or Naiman were systematically broken up as tribal groups and assigned piecemeal to composite units of one thousand or distributed as dependents (bo'of) of mingghan made up of other peoples. Only old tribal 33 On the weaponry, training, and discipline of the Mongolian army, see Denis Sinor, "The Inner Asian warrior," Journal of the American Oriental Society, 101 (1981), pp. 133—44; a n ^ Denis Sinor, "On Mongol strategy," in Proceedings of the fourth East Asian Atlantic conference, ed. Ch'en Chieh-hsien (Tainan, 1975), pp. 238-49. Cambridge Histories Online © Cambridge University Press, 2008 CHINGGIS KHAN AND THE EARLY MONGOLIAN STATE 347 allies of demonstrated loyalty were allowed to form their own ethnically homogenous mingghan. The Onggirad, for instance, were allowed to form into mingghan as a tribe with their own chief as commander. But even in such cases, the tribe, whatever its record of fidelity, was nonetheless encased within a new institutional framework and subject to rigid military discipline. The mingghan was as much an instrument of social control as it was of military mobilization. The Mongols' ideology During the early thirteenth century the Mongols elaborated an ideological system that, to their own satisfaction in any event, legitimized the sovereignty of the Chinggisid line and justified their expansionist policies. The system was certainly in place by the 1240s, but the precise time at which these doctrines were first articulated and propagated remains unclear. It seems reasonable to assume, however, that Chinggis khan and his advisers devoted some attention to the matter of legitimation at his enthronement in 1206. Although other components of Mongolian ideology may have been added later, it is convenient to summarize the entire system at this juncture. The Mongolian conception of sovereignty, like that of many other Eurasian peoples, was rooted in the notion of sacral kingship. In the Mongols' own formulation, sovereign power was conferred on an earthly ruler by Eternal Heaven (Mongke Tenggeri), the sky god and chief deity of the steppe nomads. As Heaven's chosen instrument, Chinggis khan enjoyed the protection and assistance of Mongke Tenggeri, which guaranteed the longterm success of his military and political ventures. The special good fortune that accompanied his rise to power was thus a manifestation of heaven's favor. The narrow escapes, the timely warnings of danger, and the unexpected victories on the field of battle all served to identify Chinggis khan as the earth's sole legitimate sovereign. His white battle standard with nine tails, as noted earlier, symbolized and affirmed his good fortune and hence his sovereign rights. Chinggis khan's pretensions were further buttressed by his control of the Orkhon River valley and its surrounding mountains, the Otiiken-yish of the Turkic inscriptions, the core territory of all previous nomadic polities of the eastern steppe. According to pre-Mongolian Turkic traditions, good fortune (qut) and imperial power were strongly associated with possession of these holy mountains. The Mongols, it is true, had their own holy mountain, Burkhan Khaldun, where their khans had to be invested, but it is significant that when Chinggis khan chose a site for the imperial capital he selected Khara Khorum, located in the heart of the Otiiken-yish, a decision designed Cambridge Histories Online © Cambridge University Press, 2008 348 THE RISE OF THE MONGOLIAN EMPIRE presumably to harness the good fortune that inhered in the region and to sway opinion in the nomadic world at large.'4 The sovereignty bestowed on Chinggis khan and subsequently bequeathed to his successors was universal in character. In the orders of submission customarily sent to neighboring states before initiating hostilities, the Mongols claimed the right, if not the duty, to bring all the world under their dominion. All nations outside their frontiers were considered members of the Mongolian empire-in-the-making, and all were required to accept Mongolian suzerainty without question or hesitation. Because in Mongolian eyes their expansion was divinely sanctioned, anyone refusing to submit was thereby thwarting the will of Heaven and deserving of the harshest punishment.35 The claim of a Mandate of Heaven and universal sovereignty echo wellknown Chinese political doctrines, but exact Turkic parallels can also be found for all elements of Mongolian ideology. Although direct Chinese influence cannot be ruled out, it seems more probable that the Mongols were introduced to these ideas, whatever their original provenance, through the mediation of the Turks, especially the Uighurs, whose influence on the Mongolian state in its formative years was extensive.36 Early conquests Besides serving to confirm and consolidate Chinggis khan's hold on the eastern steppe, the khuriltai of 1206 laid plans for various new military and diplomatic initiatives. Soon after its deliberations concluded, a mop-up operation was mounted against Buyirugh khan and his followers, which resulted in the Naiman's death. In the following year, 1207, Chinggis khan dispatched emissaries to the forest tribes of southern Siberia. His orders of submission had the desired effect: The Kirgiz of Upper Yenesei, the Oyirad, and other forest peoples of the Baikal region peaceably capitulated, presenting tribute of furs, falcons, and geldings to their new sovereign. With their northern border secure and fresh conscripts in the offing, the Mongols could now direct their attention toward their neighbors in the south. By 1207 there were unmistakable signs that the Chin frontier defense 34 Chinggis khan designated Khara Khorum as his capital in 1220, but construction did not begin until 1235 during the reign of Ogodei. See YS, 58, p. 1328. 3 ; The pioneering work on Mongolian ideology is Eric Voegelin's "The Mongol orders of submission to the European powers, 1245—1255," Byzantion, 15 (1940—1), pp. 378—413. See also the important article by Igor de Rachewiltz, "Some remarks on the ideological foundations of Chinggis khan's empire," Papers on Far Eastern History, 7 (1973), pp. 21-36. 36 For the Turkic parallels, see Peter B. Golden, "Imperial ideology and the sources of political unity amongst the pre-cinggisid nomads of western Eurasia," Anhivum Eurasiae Medii Aevi, 2 (1981), pp. 37-76. Cambridge Histories Online © Cambridge University Press, 2008 CHINGGIS KHAN AND THE EARLY MONGOLIAN STATE 349 system was beginning to erode. The Jurchens had been unable to prevent the unification of the steppe peoples under Mongolian auspices, and their chief march warden in the Gobi, the Onggiid Ala Khush Digid Khuri, was openly treating with Chinggis khan. Their problems were further exacerbated when the Jiiyins (Ch: Chiu), an ethnically mixed people inhabiting the sensitive Chin—Tangut—Onggiid border regions, who frequently served the Chin as military auxiliaries, revolted against their overlords, complaining of inequitable treatment. Believing that the Onggiid prince was at the center of the turmoil, the Jurchens had Ala Khush Digid Khuri assassinated, in the hope that he would be replaced by a member of the Onggiid princely house loyal to the Chin court. Their plans, however, misfired. The successor, a nephew of the murdered prince, immediately broke with the Jurchens and formally acknowledged Mongolian suzerainty. ' 7 With the strategic Onggiid territories under his control, Chinggis khan was now in a position to strike a major blow at either the Chin dynasty or the Tangut kingdom of Hsi Hsia. He decided to settle first with the Hsi-Hsia, whose territories he had previously reconnoitered in 1205 and 1207. A large army was assembled and marched south, entering Tangut territory in the late spring of 1209. The Mongolian forces, under Chinggis khan's personal command, defeated the Tanguts' frontier garrisons and pushed on to the Hsi Hsia capital, Chung-hsing (present-day Yin-ch'uan, Ning-hsia) on the Yellow River, which they placed under siege in October. When direct assaults proved unavailing, the Mongols endeavored to divert the waters of the Yellow River in order to inundate the besieged city. Their plan, however, miscarried. The irrigation canals of the Yellow River plain burst their dikes, flooding the Mongolian positions as well as the Tangut capital. In the face of this unexpected turn of events, both sides decided to seek an end to hostilities. The negotiations, which commenced in January 1210, produced an acceptable compromise: The Tangut ruler acknowledged his subordination to the Mongols and pledged to support their future military operations with troops. In return Chinggis khan lifted the siege and withdrew his armies from Hsi Hsia territory. The Mongolian khan then returned to the steppe with a new wife, a Tangut princess, and enhanced prestige; for the first time, his armies had bested the forces of a major sedentary power. Shortly after arriving home Chinggis khan accepted the submission of two other sedentary peoples, the Uighurs and Qarluqs. In this case, however, their submission was voluntary, not compelled. The Uighurs, long dependents of the Khara-Khitai kingdom (centered in Jungaria and Semirechie) 37 Paul D. Buell, "The role of the Sino-Mongolian frontier zone in the rise of Chinggis qan," in Studies on Mongolia: Proceedings of the first North American conference on Mongolian studies, ed. Henry G. Schwartz (Bellingham, 1978), pp. 63—8. Cambridge Histories Online © Cambridge University Press, 2008 35O THE RISE OF THE MONGOLIAN EMPIRE had for some years grown increasingly restive under the oppressive rule of their overlords. In 1209 the Uighurs, provoked beyond endurance, rose in rebellion and killed the much-hated and tyrannical Khara-Khitan political resident in their capital. Seeking Mongolian protection, Barchuq Art Tegin, the Uighur ruler, hastened to communicate a pledge of loyalty and service to Chinggis khan. The latter was delighted and commanded Barchuq to present himself to the Mongolian court with appropriate tribute. After some delay, occasioned by Chinggis khan's operations against the Tanguts, the Uighur ruler finally had his audience in the spring of 1211 beside the Keriilen River. As the ruler of the first sedentary state to join the empire voluntarily, Barchuq was made Chinggis khan's honorary "fifth son" and was ranked first among the princes (wang) of dependent states.*8 Second in rank was the Qarluq Turk Arslan khan, the ruler of Qayaliq, a city in the Hi valley, whose audience followed Barchuq's. Like his Uighur counterpart, Arslan khan had become disenchanted with Khara-Khitan overlordship and willingly changed masters when the opportunity presented itself. Both contributed auxiliaries to the Mongolian army, but most importantly, the Uighurs, a numerous and culturally sophisticated people, provided their new masters with a reservoir of experienced administrative and clerical personnel on which Chinggis khan and his successors drew extensively. Their heavy dependence on the Uighurs for services of this nature accounts for the substantial Turkic influence on Mongolian chancellery practices, financial institutions, and political doctrine. Initial attacks on the Chin The Chin dynasty, the Mongols' next opponent, possessed a large and effective army, but the demands placed on it were extensive. In the west they were engaged in a border war with the Tanguts, and in the south they confronted the Sung dynasty, which had never relinquished its hopes of recapturing the north. As recently as 1206-8 the two powers had clashed, and though the Chin had emerged victorious, the security of its southern frontier was a continuing source of concern. Chinggis khan, fully informed of Chin troop dispositions, took the first step toward hostilities in 1210 when he broke off tributary relations (begun around 1195) with the Chin court. He delayed the actual test of arms, however, until the following year, by which time he had subdued the Tanguts. Having no other immediate enemies, Chinggis khan could now 38 On this episode, see Thomas T. Allsen, "The Yuan dynasty and the Uighurs of Turfan in the 13th century," in Chine among equals: The Middle Kingdom and its neighbors, 10th—14th centuries, ed. Morris Rossabi (Berkeley and Los Angeles, 1983), pp. 246—8. Cambridge Histories Online © Cambridge University Press, 2008 CHINGGIS KHAN AND THE EARLY MONGOLIAN STATE 351 afford to concentrate his forces against the Chin, his most powerful neighbor. 39 The Mongolian formations departed from the Keriilen early in the year and by spring reached the Onggiid territory, which they used as a staging area for their forthcoming invasion. The center and left, that is, eastern, wing of their army, led by Chinggis khan, assaulted and captured many fortifications along the Chin's northern frontier, including Chii-yung kuan, a pivotal garrison guarding the approaches to its capital, Chung-tu (modern Peking). The Jurchen court dispatched sizable reinforcements to their endangered borders, but these contingents were defeated piecemeal as they moved to the north. The Chin defenses were so disorganized by these setbacks that elements of the Mongolian army were able to reach and pillage the environs of Chung-tu. In the meantime, the right, that is, western, wing of the Mongolian army under Chinggis khan's sons advanced into Shansi, taking a few cities, ravaging the countryside, and, most important, tying down enemy troops. When the order to withdraw came in the beginning of 1212, both wings of the Mongolian army returned to the north, abandoning most, if not all, of the Chin territories that they had occupied. By all available indications, the campaign of 1211 had as its immediate goal booty and information, not the acquisition of territory.40 The Chin forces quickly reoccupied their northern frontier regions and prepared for the next onslaught. In the fall of 1212 the Mongols returned and began pressuring the.Jurchens' outer defenses. Key garrisons such as Chiiyung kuan had to be reduced a second time, and this was achieved in 1213 only after Chinggis khan committed additional forces to the task. Once the frontier defenses were pierced, the Mongols struck rapidly south, penetrating much deeper into Chin territory than they had done previously. When they reached the agricultural areas north of the Yellow River, the army divided into three groups that spread devastation throughout Shantung, Hopei, and shansi. Some cities were taken and looted, but in general the Mongols concentrated their attention on the open countryside, bypassing strong points whenever possible. In late 1213 the Mongolian armies, having wreaked great destruction in the Chin heartland, began moving back to the north. This time, however, they retained control of all major frontier passes and left a force around Chung-tu to enforce a close blockade. Efforts to invest the city proved unsuccessful, but the alarmed Chin emperor was induced to negotiate. He offered the Mongols much tribute — gold, silk, and horses — in return for an 39 On the campaigns against the Chin, see Henry D. Martin, The rise o/Chingis khan and hii conquest of North China (Baltimore, 1950; repr. New York, I 9 7 i ) , p p . 113-219. 40 Secret history, sec. 248 (pp. 184—5); and Sechenjagchid, "Patterns of trade and conflict between China and the nomads of Mongolia," Zentralasiatische Studien, 11 (1977), p. 198. Cambridge Histories Online © Cambridge University Press, 2008 352 THE RISE OF THE MONGOLIAN EMPIRE end of hostilities. They accepted these terms and, as agreed, discontinued their blockade in the spring of 1214. The Jurchen court, unnerved by the experience, used this respite to depart from Chung-tu for K'ai-feng, which they established as their new capital in the summer of 1214. When the departure of the ruling house became known to Chinggis khan later in the fall, he immediately ordered his forces back to the recently besieged city. All attempts to take Chung-tu by storm failed, owing to the dogged resistance of its garrison. Finally Chinggis khan arrived on the scene in January 1215 and took direct control of the operations. When it became apparent that the Mongols had turned aside Chin relief armies, the garrison's morale gave way, and the city surrendered to the attacking forces at the end of May. In the weeks following its capitulation, the capital was systematically sacked and partially destroyed by fire. His immediate military goal accomplished and the vast booty properly inventoried, Chinggis khan left Chung-tu for Mongolia, leaving behind garrisons in the captured Chin territories. The loss of its capital was not, however, the only serious setback suffered by the Chin at this time. In 1212 the Mongolian general Jebe penetrated the Liao River valley and toward the end of the year temporarily seized Tungching (modern Liao-yang), the eastern capital of the Jurchen. The occupation of this city, a major defeat in itself, helped, in turn, ignite a widespread rebellion among the Khitan (Ch'i-tan), another Manchurian-based people, who had been unwilling subjects of the Chin since the fall of their own dynasty, the Liao, in 1115. Taking full advantage of the growing discomfiture of their opponents, Mongolian armies in 1214 successfully attacked Chin positions east and west of the Liao River. Tung-ching was again occupied in 1215 and subsequently became the main base for the Khitan rebel leader, Yeh-lii Liu-ko, who now formally acknowledged Mongolian suzerainty.41 By the following year a large part of Manchuria, the Jurchen homeLand, was in enemy hands (see Map 25). A concentrated Mongolian assault at this juncture might well have brought the Chin dynasty to the point of collapse, but events unfolding in Turkestan would soon cause Chinggis khan to direct the bulk of the Mongols' military efforts westward for nearly a decade. Campaign in the west Mongolian involvement in the Western Regions, the Hsi-yii of the Chinese, began in 1208 when a punitive expedition was mounted against a coalition of 41 On Khitan uprisings against the Chin, see Sechen Jagchid, "Kitan struggles against Jurchen oppression: Nomadism versus sinicization," Zentralasiatische Studim, 16 (1982), pp. 163-83. Cambridge Histories Online © Cambridge University Press, 2008 CHINGGIS KHAN AND THE EARLY MONGOLIAN STATE 353 Return route to Chinggis's camp ontheKerulenR." Chinggis's encampment 1214-15 Chinggis makes treaty with Khitan rebel Yeh-lii Liu-ko 1212,iii. Rebellion of Chang-chiu n V chung. £ TungChing 1215 , •v X.*V— A« 4 (Liao-yang) eked by Jebe 1212,ii. Taken byYen-lu # ^ / Peninsula Jebe's campaign 1211,ii—1212.U. Jochi Khasar's Northern campaign 1214,x—1215, iii. Mukhalis campaigns 1214—16 300 km M A P 2 5 . The campaigns in Manchuria, 1211—16 Cambridge Histories Online © Cambridge University Press, 2008 354 T H E RISE OF THE MONGOLIAN EMPIRE renegade Merkid and Naiman tribesmen who had established a base of operations along the upper Irtysh in western Mongolia^2 (see Map 26). The Mongolian forces overwhelmed the rebels and killed their leader, Toghto'a, Chinggis khan's long-time enemy. The surviving Merkid first fled to the Uighur realm and subsequently to the Qipchaq steppe, while the remaining Naiman escaped into the territory of the Khara-Khitan, or Western Liao. The leader of the latter group was Giichuliig, whose father, Tayang khan, had died fighting the Mongols in 1204. At the time of Giichiiliig's flight, the Khara-Khitan state was engaged in a struggle with Khwarazm, a Muslim state embracing most of western Turkestan and Khurasan (Afghanistan and northern Iran). The Naiman chieftain fished in these troubled waters for a while, finally allying himself with the Khara-Khitan sovereign. He soon emerged as the ruler's chief adviser and used this position of trust to seize power within the Khara-Khitan realm in 1211. The waning fortunes of the Western Liao revived rapidly under the energetic leadership of Giichuliig. He forced the Khwarazmshah, Muhammad, to pull back in the region of the upper Syr Darya and in 1213—14 extended his sway over the predominately Muslim population of the Tarim basin. The success of the Naiman usurper began to draw Chinggis khan's attention westward just as word of the Mongolian ruler's triumphs in China reached Sultan Muhammad. In 1215 the Khwarazmshah sent a diplomatic mission to Chinggis khan in north China in order to gather intelligence about the new power in the east. Chinggis khan cordially welcomed the ambassadors, indicating his desire for peaceable relations and commercial intercourse with his neighbor to the west. To this end, the Mongolian ruler dispatched a return embassy to Muhammad that arrived in Khwarazm in the spring of 1218. Negotiations commenced, and several days later the sultan agreed to a treaty of peace and friendship with Chinggis khan. The amicable phase of their relationship was, however, soon brought to an abrupt end by a dramatic incident at Utrar, a Khwarazmian city on the upper Syr Darya. Shortly after the treaty was concluded, the city's governor, apparently with the foreknowledge of the sultan, massacred the members of a large Mongoliansponsored trade caravan and seized their merchandise. In response, Chinggis khan hastened an embassy to the Khwarazmshah, demanding the punishment of the offending official and the return of the confiscated goods. For reasons that are hard to fathom, given his knowledge of the Mongols, Muhammad 42 The western campaign is discussed in detail by Wilhelm Bart hold, Turkestan down to the Mongol invasion, trans. T. Minorsky, 3rd ed. (London, 1968), pp. 355—457; and Ilia P. Petrushevskii, "Pokhod mongol'skikh voisk v sredniuiu Aziiu v 1219—1224 gg. i ego posledstviia," in TarlaroMoagoly v Azii i Etmpe, ed. S. L. Tikhvinskii (Moscow, 1977), pp. 107-39. Cambridge Histories Online © Cambridge University Press, 2008 Assembly area for Chinggis's armies *?; x '•'•ft'*^Bemian»J\EBrwan Region controlled by Chinggis Jebe's 1218 expedition against former territories of Khara Khitai Chinggis's main army Subsidiary Mongolian armies ..- .. \, / , Jebe andSiibetei pursue Mohammed of Khwarazm. After his death (1220,xi) they continue into the Caucasus and Southern Russia, returning to join Chinggis in 1223. Jebe dies during return through Qipchaq in 1223. MAP 26. Chinggis's western campaigns Cambridge Histories Online © Cambridge University Press, 2008 356 THE RISE OF THE MONGOLIAN EMPIRE rebuffed these entreaties and contemptuously put Chinggis khan's emissary to death. The Mongolian ruler, infuriated by these outrages, prepared for war. Before confronting his new enemy, however, Chinggis khan had to complete operations on two other fronts. First, in the north, a rebellion of the forest tribes - Oyirad, Tumad, and Kirgiz - required attention. The uprising began in 1217 and continued to spread until the winter of 1218—19 when Jochi, in command of the right wing of the Mongolian army, reached southern Siberia and quickly pacified the rebellious tribesmen.« Second, east Turkestan had to be cleared of Giichiilug and the Khara Khitan. Operations in this quarter had begun in 1216 under Jebe, who was now ordered to dispose of the Naiman without further delay, a commission he carried out successfully. By the end of 1218 Giichiilug was dead, and the Khara-Khitan territories, which fronted on those of the Khwarazmshah, were occupied by Mongolian forces. His lines of communication to the west now secure, Chinggis khan assembled a large force along the Irtysh. Composed of Mongolian regulars and numerous auxiliaries mobilized from vassal states, the army moved against Khwarazm in the summer of 1219. The main force under Chinggis khan advanced on the enemy's population centers south of the Syr Darya, while a covering force led by Jochi entered the steppe regions north of the river, engaging the nomadic Qipchaq and Qangli, who were closely allied to the Khwarazmian ruling house. Rather than meet the invader in the open field, Muhammad, contrary to the advice of his generals, used his army, which was numerically superior to that of the Mongols, to garrison the major cities of the realm. Although the decision probably cost him any chance of victory, it did force the Mongols into a series of time-consuming and destructive siege operations: The reductions of Utrar, Khwarazm, Harat, and Samarkand were particularly bloody affairs, in which the Mongols drove unarmed prisoners of war against heavily defended walls to provide "human camouflage" for their assaulting forces. Moreover, to add to the carnage, once taken, several cities rose up against their new masters and had to be subdued a second time. In such cases, savage reprisals, normally in the form of mass executions, were visited upon the population. When he became aware that his strategy had failed, Muhammad panicked and fled to an island in the Caspian Sea, where he died around 1221. His son and successor, Jalal al-Din, continued to offer a spirited resistance with the meager forces at his disposal. In pursuit of the energetic and dashing Jalal alDin, Mongolian contingents tracked him from northern Iran through Af43 The Secret history, sec. 239 (pp. 173-4), mistakenly places Jochi's campaign against the forest peoples in 1207. For a discussion of the confusion in the Secret history's chronology, see Paul Pelliot, Notes critique d'histoire Kalmouke (Paris, i960), vol. 1, pp. 5, 57 (n. 39), 60 (n. 58). Cambridge Histories Online © Cambridge University Press, 2008 CHINGGIS KHAN AND THE EARLY MONGOLIAN STATE 357 ghanistan into India and then back to Iran and Azerbaijan. To the Mongols' frustration he always managed to elude capture, but heroics of this nature could not long forestall the collapse of the Khwarazmian state. By 1223 Turkistan and Khurasan had been subjugated and Mongolian garrisons and governors (darughachi) installed in all urban centers. Despite the hopelessness of his cause, Jalal al-Din refused to capitulate and continued his futile defiance until his death at the hands of Kurdish bandits in 1231. With organized resistance coming to an end in the Khwarazmian realm, the Mongols began preparations for their next series of conquests. Siibetei and Jebe, who were campaigning in Georgia and Azerbaijan at this time, asked permission to cross the Caucasus and attack the Qipchaqs. Chinggis khan readily consented, and in 1221 Siibetei commenced his famous raid, or, more accurately, reconnaissance in force, into western Eurasia. Accompanied by an army of three ttimen, he moved into the south Russian steppe and in the late spring of 1223 defeated the combined forces of the Russian princes and western Qipchaqs at the battle of the Kalka River (a small stream flowing into the Sea of Azov). Siibetei next scouted the Russian principalities as far west as the Dnieper and then turned back east, fighting a brief engagement with the Volga Bulghars before returning to western Mongolia in 1224. The necessary intelligence having been gathered, Jochi was ordered to launch a follow-up campaign to bring the western steppe under Mongolian dominion. Chinggis khan in the meantime had withdrawn the bulk of his armies from Turkestan, reaching the Irtysh in the summer of 1224 and central Mongolia in the spring of 1225. Back at home, he planned yet another campaign: In 1223 the Tangut ruler had without warning withdrawn his forces, which had been supporting Mongolian operations against the Chin, and the Mongolian leader was determined to exact a heavy penalty for this faithlessness. Mukhali's campaigns against the Chin When Chinggis khan reached the Keriilen in late 1215 or early 1216, Mongolian operations against the Chin were scaled down temporarily but not halted. Mukhali, one of Chinggis khan's most able and trusted generals, continued his efforts to clear the Liao River valley of Jurchen forces, a task that he completed in 1216. After occupying the major cities of the region, Mukhali went to Mongolia in the fall of 1217 to report to his sovereign. Pleased with his work, Chinggis khan gave him the title of kuo-wang, or "prince of the realm," and placed him in overall command of a new campaign to seize those parts of north China still in Jurchen hands, that is, the land south of the T'ai-ho Range. Cambridge Histories Online © Cambridge University Press, 2008 358 THE RISE OF THE MONGOLIAN EMPIRE Mukhali returned to the south in the same year and set up militaryadministrative headquarters at Chung-tu (now renamed Yen) and Hsi-ching (modern Ta-t'ung). The forces at his disposal included 23,000 troops of the left wing of the Mongolian army, augmented by 77,000 Chinese, Jurchen, and Khitan auxiliaries that had either surrendered or defected to the Mongols during the earlier fighting with the Chin. As a matter of policy the Mongols encouraged and rewarded such defections, and the results had been gratifying: Many Chin commanders, especially those of non-Jurchen origin, came over with their units intact. It was the addition of these crucial auxiliaries, which comprised three quarters of the troops available to Mukhali, that enabled the Mongols to maintain unremitting pressure on the Chin even after the greater part of their army, the center and right wing, had been withdrawn from north China and committed in the west.1''* In the initial phase of the new campaign, the Mongols launched a threepronged attack from Chung-tu and Hsi-ching designed to wrest Shansi, Hopei, and Shantung from Chin control. Pushing into Hopei with the center and main column, Mukhali soon encountered stiff resistance. Cities had to be taken by direct assault, at high cost to both sides. On several occasions cities won at such a high price were lost and had to be retaken. Though the going was difficult, Mukhali was nonetheless making slow progress. In 1218, leaving the Chin defector Chang Jou behind to consolidate Mongolian gains in Hopei, Mukhali shifted his attention to Shansi. T'ai-yiian, the main Chin bastion in the northwest of the province, was taken in October, and the Mongols were then able to drive steadily to the south. By the end of 1219 only the southernmost strip of Shansi remained outside Mongolian control. Mukhali now returned to central Hopei and received the surrender of the remaining Chin-controlled cities, including the key garrison at Ta-ming, during the summer and fall of 1220. Thereafter he pressed on into western Shantung, taking its chief city, Chi-nan, in October, without a fight. The relatively easy campaigning of 1220 was made possible by the Chin dynasty's ill-advised military involvement in the south. In 1217, during the lull in the fighting with the Mongols, the Chin emperor had foolishly consented to open a campaign against the Sung, which had suspended its tribute payments to the Jurchen court three years earlier. The series of annual 44 The number of troops available to Mukhali was carefully calculated by Huang Shih-chien in "Mu-hua-li kuo wang hui hsia chu chiin k'ao," Yuan shih lun ts'ung, i (1982), pp. 57—71. For accounts of the campaign, see Igor de Rachewiltz, "Muqali, Bol, Tas and An-t'ung," Papers on Far Eastern History, 15 (1977), pp. 45—55; and Martin, The rise ofChingis khan, pp. 239—82. On the role of the Sung in the Mongolian—Jurchen conflict of 1217—25, see Charles A. Peterson, "Old illusions and new realities: Sung foreign policy, 1217—1234," in China among equals: The Middle Kingdom and its neighbors, 10th14th centuries, ed. Morris Rossabi (Berkeley and Los Angeles, 1983), pp. 2 0 4 - 2 0 . Cambridge Histories Online © Cambridge University Press, 2008 CHINGGIS KHAN AND THE EARLY MONGOLIAN STATE 359 offensives launched by the Chin between 1217 and 1224 were often successful on the local level, but they never brought decisive victory. The Sung, refusing to negotiate despite initial setbacks, continued to resist and on one occasion in the summer of 1219 even managed to destroy a major Jurchen army in the Han valley. The Chin decision to divide their military effort proved costly. The marginal gains at the expense of the Sung in no way compensated them for their losses to the Mongols in the north and in the long run clearly undermined their ability to cope with Mukhali's forces. Undaunted, however, in 1220 the Chin mobilized a new army and prepared a counterthrust in the hope of recouping some of their losses. Though initially formed, it appears, to strike at eastern Shantung, where an army of anti-Jurchen Chinese rebels (the Hung-ao or Red Coats) had been organized, the new force soon attracted Mongolian attention. Once he became aware of its existence, Mukhali moved south from Chi-nan in late 1220 and attacked the new Chin army at Huangling kang, a ford on the south bank of the Yellow River not far from K'aifeng. He decisively defeated his foes, and with this victory the Mongols extended their control over most of the Jurchen territory north of the Yellow River, except for eastern Shantung, which remained in the hands of the Red Coats, and Shensi, which remained under Chin authority. Placing Chinese defectors in charge of the surrendered areas, Mukhali returned to the north, conducting mop-up operations along the way. In the meantime, the Chin court, its counteroffensive having failed, sent an embassy headed by Wu-ku-sun Chung-tuan to Chinggis khan in the west to discuss possible terms. The Mongolian demands that the Chin emperor accept the title of prince (wang), and thus recognize Chinggis khan as his sovereign, and that Shensi be evacuated were, however, considered excessive, and so the hostilities continued. Mukhali renewed pressure on the Chin by initiating a major campaign in Shensi and eastern Kansu (Kuan-chung) in mid-1221. After first crossing the Ordos (with the acquiescence of the Tanguts, who also contributed auxiliary contingents numbering fifty thousand), Mukhali spent the remaining part of the year and the beginning of the next reducing the major cities of northern and central Shensi. In the spring of 1222 he left one of his lieutenants, Monggii Bukha, in charge of operations in Shensi and crossed the Yellow River into Shansi to forestall a new Chin offensive in this quarter. In the fighting that followed, the Mongols took Ho-chung and other fortified cities along the river. In Shensi, meanwhile, Monggii Bukha had become bogged down in extensive blockade operations. Even after the return of Mukhali and his troops to Shensi in the fall of 1222, the Mongols were still unable to force the capitulation of many key cities, including Ch'ang-an and Feng-hsiang. Cambridge Histories Online © Cambridge University Press, 2008 360 THE RISE OF THE MONGOLIAN EMPIRE The sudden withdrawal of the Tangut auxiliaries at this critical juncture further undermined the Mongols' military position. His striking power now much reduced, Mukhali lifted the siege of Feng-hsiang in early 1223, and following a brief retaliatory attack on the Hsi Hsia frontier, he moved back to Shansi, where he shortly after fell ill and died (in March or April). The deceased commander was immediately replaced by his brother, Dayisun, but the Mongolian offensive had lost its momentum. Making the most of this opportunity, the Chin hastily ended hostilities with the Sung, moved its troops back into southern Shansi, and recovered some of the territory previously lost to the Mongols. Supported by the Sung, with whom they were loosely allied, the Red Coats also took advantage of the situation to extend their hold in Shantung and, briefly, to seize territories in Hopei. This latter move precipitated a rebellion by Wu Hsien, a former Chin commander who had recently defected to the Mongols. In 1225 he changed sides once again, this time throwing his lot in with the Sung. In the face of these setbacks and Chinggis khan's determination to deal next with the perfidious Tanguts, the Mongols had to content themselves with a holding operation in north China for the following few years. Administration of north China The Mongols, as Chinggis khan himself recognized, knew little of the "laws and customs of cities" and were ill equipped to undertake the administration of complex sedentary societies on their own. It was therefore necessary to recruit numerous technical specialists, particularly people with experience in government or commerce who were willing to help the Mongols administer and exploit the agricultural and urban populations under their control. Even before his invasion of the Chin, Chinggis khan began building up a cadre of such specialists from among the Khitan and Chinese officials who, for a variety of reasons, left Chin service and submitted to the Mongols.45 By the time operations commenced against the Jurchens in 1211, Chinggis khan had in his entourage a body of advisers who were intimately familiar with both the Chin administrative system and conditions in north China. As the Mongolian campaign gathered momentum, the number of defectors increased markedly. Officials of Chinese origin were most numerous in this second wave, but for the first time a few Jurchens also came over to the Mongolian camp and offered their services. Civil officials who defected or surrendered without resistance were routinely left in their old posts adminis45 In preparing this section I relied heavily on Igor de Rachewiltz's excellent study, "Personnel and personalities in North China in the early Mongol period," Journal ofthe Economic and Social History ofthe Orient, 9 (1966), pp. 8 8 - 1 4 4 . Cambridge Histories Online © Cambridge University Press, 2008 CHINGGIS KHAN AND THE EARLY MONGOLIAN STATE 361 tering hsien (counties) and chou (prefectures). Their principal responsibilities were to maintain order and mobilize the resources of their areas for the benefit of the Mongolian military machine. Chinese and Khitan military commanders who switched sides participated in operations against the Chin either independently or in association with Mongolian armies. Such commanders were confirmed in office by Chinggis khan or later by Mukhali. They received official titles in Chinese or Mongolian and were given emblems of authority (Mongolian: gerege or baisa; Chinese: p'ai-tzu) as an indication of their new status. To coordinate their military and administrative efforts, the Mongols, drawing on Chin precedents, created a series of Hsing-t'ai-chung-shu-sheng, Branch Secretariats. Formed initially in the early twelfth century as branches of the Shang-shu sheng, the Secretariat of State Affairs, the supreme administrative organ of the Chin government, the Hsing-sheng, as they were called for short, were established mainly in newly conquered territories and later on in frontier areas threatened with attack. The officer in charge, also termed a hsing-sheng, was given plenipotentiary powers within his jurisdiction, which in Chin times normally corresponded to a lu, or "circuit" (which was divided into several prefectures, or chou). The Mongols were quick to adapt this institution to their needs. In 1214 the first Branch Secretariat was established, whose head was the Mongolian general Samukha Baghatur. The Khitan Shih-mo Ming-an was appointed hsingsheng ofYen-ching (Chung-tu) in 1215 following its capture. After 1217 when major campaigning was resumed, there was another wave of Chinese defections, and for the first time some Han Chinese were made hsing-sheng. Like their Chin counterparts, the Mongolian-appointed hsing-sheng received full discretionary powers upon assuming office. Most were military men, and given the importance of the post, they were carefully screened before being appointed. Although the office of hsing-sheng was, on the surface at least, alien to the Mongolian tradition, it was effectively integrated into the Mongolian sociopolitical system.46 Chinese or others appointed to this office were given an appropriate military rank and, in some cases, granted a commission in the guard, becoming thereby a companion (nb'kb'r) to Chinggis khan or his prince of the realm, Mukhali. To guarantee their fidelity, the sons of such officials were held as hostages in units of the kesig. Thus a Chinese hsing-sheng, in terms of the extent of his combined military—administrative authority, formal rank, and relationship to the court, was a rough equivalent to a senior Mongolian commander (noyan) of a mingghan or ttirnen. Because ties of loyalty in the Mongolian system of governance were highly 46 Rachewiltz, "Personnel and personalities," pp. 128—32. Cambridge Histories Online © Cambridge University Press, 2008 362 THE RISE OF THE MONGOLIAN EMPIRE personal, an office of any type or importance was normally hereditary. This was true of the hsing-sheng: Sons succeeded fathers, and over time these commands were transformed into private domains. In the long run this process of "feudalization" was to have undesirable consequences from the standpoint of the Mongolian court, but in the short term it was an efficient way of consolidating its hold on newly won territories in north China. At the apex of the Mongolian administrative system in occupied China stood the viceroy, Mukhali. He undoubtedly communicated with Chinggis khan, but on the whole he enjoyed extensive autonomy. Chao Hung, a Sung emissary who visited Mukhali's camp in 1221, compared him with a Chinese emperor, though he knew that in truth he was not a sovereign ruler. *~* Certainly no other Mongolian commander was ever entrusted with as much authority and freedom of action as that granted to the kuo-wang. In the course of his lengthy campaign against the Chin, Mukhali naturally became familiar with some aspects of Chinese culture. Chao Hung noted that the kuo-wang's clothes and regalia were Chinese, as were the court procedures followed at his camp. On the other hand, the same observer reported that women were conspicuous at Mukhali's court and that they drank and conversed freely with men, all of which testifies to the strength and persistence of Mongolian social customs.48 Not surprisingly, the Mongolian regime in north China as it developed in the first half of the thirteenth century was composed of a complex blend of Chinese, Jurchen, Khitan, Uighur, and Mongolian administrative techniques and social usages, which was typical of the hybrid polities that had formed along China's steppe frontier since the collapse of the Han dynasty. Mongolian policy in north China The Mongolian campaigns against Chin caused widespread destruction, loss of life, and social dislocation. Their tactic of isolating major cities by devastating their hinterlands meant that both urban and rural populations suffered grievous casualties and privation. An envoy of the Khwarazmshah Muhammad who reached Chung-tu shortly after its surrender in 1215 encountered there a most grisly scene: The area around the former Chin capital for the distance of several days' march was strewn, he reports, with the bones of uncounted dead, and disease, spread owing to the great number of unburied corpses, continued to claim new victims, including members of his own 47 Chao Hung, Meng-ta pei In, in Mtng-ku shih liao ssu dung, ed. Wang Kuo-wei (1926; repr. Taipei, 1962, 197;), 5b; and Erich Haenisch and Yao Ts'ung-wu, ed., Meng-tapci-lu undHti-ta shib-liieb: Cbinesische Gesandtenberichte iiber ditfriihen Mongoltn, 1221 und 1237 (Wiesbaden, 1980), p. 35 (hereafter ciced as Cbinaische Gesandtenbericbte). 48 Chao H u n g , Meng-ta pei tu, 13a; Cbinaische Gesandtenbericbte, p . 79. Cambridge Histories Online © Cambridge University Press, 2008 CHINGGIS KHAN AND THE EARLY MONGOLIAN STATE 363 party.49 As Chinese officials entered Mongolian service in greater numbers after 1217, conditions improved somewhat, especially after Mukhali was persuaded to order his armies to cease the wanton destruction of life and property. Nevertheless, throughout the 1220s, north China was still an active theater of operations, and the death toll among the civilian population remained extremely high. Those who survived the military onslaughts, pestilence, and food shortages and who came under Mongolian administrative control faced many new trials. The Mongols, from the inception of their state, always made heavy demands on their subjects. One of the chief obligations of surrendered peoples was the provision of auxiliary troops to support further Mongolian expansion. Because siege and blockading operations were new to the Mongols and, moreover, needed a great deal of manpower, Chinese formations were quickly mobilized and utilized for this task. Some of these Chinese units came over to the Mongols intact with their officers, whereas others were newly recruited from the civilian population by the Chin officials who had retained their posts under the new regime. By 1213 Chinese auxiliaries, variously called Han chiin (Han armies) or hei chun (black armies), were in action against the Jurchens. These forces were steadily built up during the campaign and at the time of Mukhali's death substantially outnumbered the Mongolian contingents. In addition to military conscription, the Chinese were forced to supply their overlords with a lengthy list of goods and services. Before Ogodei's reign, there is no evidence of the existence of a unified system of revenue collection in any of the sedentary regions of the Mongolian state, including north China. Although data on the situation in former Chin territories between 1211 —1227 are scarce, it appears that, as was their practice elsewhere in active combat zones, the Mongols simply demanded what they required from the surrendered population as the need arose. Thus, tax collection was an ad hoc, irregular arrangement, essentially an endless series of extraordinary requisitions and exactions to meet the contingencies of war.' 0 For the most part, the Mongols collected taxes in kind, for example, in grain, clothing, mounts, and weapons (or metal implements that could be converted into weapons). In the terminology of this era, all obligations to the state - whether military service, corvee, or taxes in kind or coin — were subsumed under the name of requisition, alba 49 Minhaj al-DTn Juzjani, Tabaqat-I Ndiiri, ed. W. Nassau Lees (Calcutta, 1964), pp. 335—6; and Minhaj al-Dln Juzjani, Tabaqal-i niiiri, trans. H. G. Raverty (New Delhi, 1970), vol. 2, p. 965. 50 YS, 153, p. 3609, records that Liu Min was placed in charge of tax collection in the Yen-ching (Chung-tu) area in 1223. This may indicate that a more orderly and familiar system of exactions was introduced into certain localities at this time, but it certainly provides no evidence for the existence of a unified, centrally controlled program of regularized revenue extraction in north China as a whole before the fiscal reforms of the 1230s. Cambridge Histories Online © Cambridge University Press, 2008 364 THE RISE OF THE MONGOLIAN EMPIRE khubchiri (Chinese: ch'ai-fa). The only category of the population of north China exempt from these various imposts was religious dignitaries. In 1219 the Ch'an monk Hai-yiin obtained a grant of immunity for his Buddhist followers, and in 1223 the Mongolian court extended this privilege to the Taoist Ch'iian-chen sect and subsequently to other major religious groups Muslims, Christians, and so on in their domain.'1 As we noted, the harsh conditions in north China were eased slightly when Chinese influence in the administration increased after 1217. Attempts were made to rebuild destroyed facilities, restore agriculture, and revive social and educational institutions. But such efforts were local in nature and never received the active backing of the Mongolian regime. Significant improvements would have to await the final conquest of the Chin dynasty and the reforms of Yeh-lii Ch'u-ts'ai in the early and mid-12 30s. The Tangut campaign and the death of Chinggis khan When Chinggis khan departed for Khwarazm, he asked the Tanguts for auxiliaries, a request that, contrary to their earlier pledge to the Mongols, was not fulfilled. Several years later the Tanguts, having second thoughts, sent troops to help Mukhali wrest Shensi from the Chin in order to restore their standing with the Mongols. But in early 1223, in yet another reversal of policy, which reflected deep divisions within their court, the Hsi Hsia unexpectedly withdrew these formations. Such fickle behavior on the part of a dependent state was both a military menace and a challenge to Mongolian prestige and could not be tolerated; the Tanguts would have to be called to account and brought permanently into line. To test the mood of the Hsi Hsia court, or perhaps to divide it further, Chinggis khan in the spring of 1225 made a proposal that afforded the Tanguts a final opportunity to subordinate themselves peaceably to the Mongolian state: Their sovereign, Wei-ming Te-wang, was forthwith to send a son to Chinggis khan's camp as a hostage to guarantee his future fidelity. The Hsi Hsia, however, failed to respond to the offer and in the fall of 1225 further provoked the Mongols by signing a peace treaty with the Chin. ' 2 War was now inevitable. Unlike the invasion of 1209, which saw a quick thrust toward Chunghsing, the campaign of 1226 had as its initial objective the conquest or neutralization of the western portion of the Hsi Hsia kingdom in order to 51 Tao-chung Yao, "Ch'iu Ch'u-chi and Chinggis khan," Harvard Journal of Asiatic Studies, 46 (1986), pp. 201-19. 52 Martin, The rise of Cbingis khan, pp. 283-308; Pelliot, Notes on Marco Polo, vol. 1, pp. 304-30; and Evgenii I. Kychanov, "Mongolo-Tangutskie voiny i gibel' gosudarstva Si Sia," in Tartaro-Mongoly v Azii i Europe, pp. 46—61. Cambridge Histories Online © Cambridge University Press, 2008 THE REIGNS OF OGODEI AND G O Y O G 365 isolate the capital and court from the rest of the country. Hostilities commenced in the spring of 1226 with an attack on Khara Khoto - the Etzina of Marco Polo and the Hei-shui of the Chinese — an important Tangut outpost in the western Gobi. The Hsi Hsia garrison there was soon overrun, and the Mongolian forces moved south into the Kansu corridor to reduce the cities of Su-chou and Kan-chou. By the end of summer both had fallen, the former with great loss of life. Chinggis khan, who had been directing both operations from the coolness of his headquarters in the foothills of some nearby mountains, now reorganized his forces, sending one column westward to Kua-chou and the other eastward to Hsi-liang. The latter, one of the main urban centers of the Hsi Hsia kingdom, capitulated without a fight in August. The victorious and still-fresh Mongol force was then ordered to cross the Yellow River and press toward the Hsi Hsia capital, Chung-hsing. In late 1226 they reached and besieged Ling-chou, an important fortified garrison south of the capital. When the Tangut ruler, reacting to this threat, dispatched a large army to relieve the beleaguered city, Chinggis khan hurriedly crossed the Yellow River with reinforcements and routed the advancing relief force. By early 1227 Chung-hsing itself was under siege, and by the end of the summer it was on the verge of collapse. Once his troops had surrounded Chung-hsing, Chinggis khan struck south in the direction of the Wei River valley, attacking positions along the western frontier of the Chin throughout the spring and summer of 1227. In August, however, the Mongolian chieftain became ill and soon died, apparently from complications arising from injuries that he had suffered in a riding accident in the fall of 1225. His death, which occurred somewhere south of the Liu-p'an Mountains, was temporarily kept secret, and in accordance with his deathbed wishes, operations against the Hsi Hsia capital continued until September, when the city was taken and sacked. Once the Tangut kingdom was destroyed, the body of Chinggis khan was taken back to Mongolia and buried on Burkhan Khaldun. Forces were left behind to consolidate recent gains, but further offensive operations were curtailed while the royal clan, their chief advisers, and commanders repaired to the Mongolian homeland to mourn the loss of their leader and to put in order the affairs of the realm. THE ORGANIZATION OF THE EMPIRE: THE REIGNS OF OGODEIAND G O Y O G The Chinggisid dispensation and the khuriltai of 1229 Chinggis khan had first faced the issue of succession in 1219 on the eve of his invasion of Khwarazm. On the advice of a junior wife, Yesiii, who pointed Cambridge Histories Online © Cambridge University Press, 2008 366 THE RISE OF THE MONGOLIAN EMPIRE out the many personal hazards he would face in the forthcoming campaign, the Mongolian leader decided to resolve this crucial matter without further delay. In the ensuing court discussion, a violent quarrel soon broke out between the two leading candidates, his eldest sons, Jochi and Chaghadai. The latter, seeking to advance his claim on the throne, openly questioned Jochi's paternity, calling attention to the fact that his elder brother had been born immediately after Bbrte's release from a period of captivity among the Merkid. An altercation followed, and it was apparent to their father that their personal differences were irreconcilable and that neither would ever willingly accept the elevation of the other. To avoid a disputed succession, Chinggis khan consequently turned to his third son, Ogodei, as a compromise candidate. This solution proved acceptable to his other three sons, all of whom publicly and solemnly swore to their father that they would respect this decision and that at the appointed hour they would unhesitatingly give Ogodei their loyalty and support. To eliminate any possibility of doubt on this score, on his deathbed some eight years later Chinggis khan reaffirmed Ogodei's right to the throne.53 As an additional measure to minimize tension and conflict among his descendants, in the years before his death Chinggis khan assigned each of his sons territories and peoples of their own. In theory each would maintain himself on his respective lands but would recognize the sovereignty of Chinggis khan's chosen successor and actively collaborate with the latter in further expanding the empire's frontiers. Jochi, the eldest, as Mongolian custom dictated, was the first to receive his territory, the Irtysh region, sometime in 1207 or 1208. This, by order of his father, was later extended to include the steppe regions of western Eurasia and the Russian principalities. When the other sons were assigned their lands is not known, but it was probably in the early 1220s. Under this dispensation Chaghadai received west Turkestan, the Tarim basin, and the western Tienshan region. Ogodei obtained Jungaria and the western slopes of the Altai, and Tolui, the youngest son and therefore the guardian (pdchigiri) of his father's hearth, received the Mongolian homeland.'* North China, as far as can be determined, was not included in any of the assignments at this time; presumably it was among the territories that Chinggis khan retained under his own authority and subsequently passed on to his successor. Of equal importance, Chinggis khan also made provisions to divide the 53 Secret history, sec. 254—5 (pp. 189-97); Rashid/Karimi, vol. 1, p. 443; and Rashld al-Din, The successors of Genghis khan, trans. John A. Boyle (New York, 1971), p. 18 (hereafter cited as Rashld/Boyle). 34 'Ata Malik Juvaynl, Ta'rlkh-i jahan gusha, ed. Mirza Muhammad Qazvtni (London, 1912), vol. 1, p. 31 (hereafter cited as Juvayni/Qazvinl); 'Ata Malik JuvaynT, The history of the world conqueror, trans. John A. Boyle (Cambridge, Mass., 1958), vol. 1, pp. 42—3 (hereafter cited as Juvayni/Boyle); and Bartold, Turkestan down to the Mongol invasion, pp. 3 9 2 - 3 . Cambridge Histories Online © Cambridge University Press, 2008 THE REIGNS OF OGODEI AND GOYOG 367 native Mongolian component of his armies among his sons and other relatives. He bequeathed 4,000 troops to each of his three eldest sons and lesser numbers to various other kinsmen. The remainder of the army, 101,000 men, not previously apportioned, came into the possession of Tolui, who, as the odchigin, by nomadic convention received his father's residual estate. 55 Tolui, of course, was expected to place these armies, the heart of the Mongolian military machine, at the disposaal of the empire and, initially at least, did so. In later decades, however, Toluid control of this force would greatly affect the power relations among the increasingly competitive Chinggisid lines. The most vexing issues of succession and inheritance having been anticipated and disposed of during Chinggis khan's lifetime, the first transfer of power within the Mongol empire was accomplished with minimal discord. Tolui was put forward as an alternative candidate, but his cause was not pressed vigorously. It may well be that his candidacy was never intended as a serious challenge to Ogodei but was advanced, rather, as a means of laying the foundation for future Toluid claims on the Khaghanate. In any event, no open breach occurred, and preparations for Ogodei's enthronement slowly moved forward. First Chinggis khan had to be properly buried and imperial kinsmen and military commanders called in from the far reaches of the empire. For the duration of the interregnum, Tolui, as guardian of the Mongolian homeland, was placed in charge of the affairs of state, that is, became regent. 56 When the accord was achieved and the preliminaries were completed, the khuriltai was at last convened at Kode'e Aral, an area near the Keriilen River, sometime in the fall of 1229. Following Chinggis khan's directive, those assembled formally offered the throne to Ogodei, who, after several polite and ritualistic refusals, was finally "persuaded" to accept, assuming the old Turkic title khaghan, or emperor, as a means of distinguishing himself from his brothers, who now bore the title of khan. To demonstrate their acceptance of his elevation, Ogodei was lifted onto the throne by his chief potential rivals for the crown, his brothers Tolui and Chaghadai and his uncle Temiige Odchigin. Then, in the words of the Secret history, the guards and quiver bearers were "delivered unto Ogodei khaghan"; that is, the reins of govern55 Rashid/Karlmi, vol. i,pp. 399—417, provides a complete breakdown of the division, unit by unit. It needs to be emphasized, however, that this listing includes only those formations raised in Mongolia proper. Although specifics are lacking, it is clear that in addition to the four thousand "Mongol regulars," who served as a kind of cadre, each of the eldest three sons also had at his disposal locally recruited auxiliary contingents. Thus, the strength of the armies under the ocntrol of Jochi, Ogodei, and Chaghadai in 1227 was actually much greater than the figures reported by Rashid al-Din. For further discussion, see John Masson Smith, "Mongol manpower and Persian population, "Journal of the Economic and Social History of the Orient, 18 (1975), pp. 273-5. 56 YS, 115, p. 2885. Cambridge Histories Online © Cambridge University Press, 2008 368 THE RISE OF THE MONGOLIAN EMPIRE ment were placed in his hands. The enthronement ceremonies at an end, a large celebratory feast was held at which the new khaghan dispensed gifts and thanks to the assembled notables.57 Renewed expansion and the fall of the Chin Ogodei, who had spent most of his life on campaign, began his reign with a burst of martial energy. The frontiers of the empire, in accordance with the decisions reached at the recently concluded khuriltai, were to be pushed vigorously outward in a number of directions. One of the most pressing tasks left over from his father's reign was the conquest of the Qipchaq steppes and the Russian principalities. Chinggis khan had given this important assignment to Jochi as early as 1221 or 1222, but the latter, despite his father's mounting anger, never devoted his full attention to the task. After Jochi's death in 1227, a ^ ew months before that of his father, operations in this quarter, halfhearted as they had been, came to a complete halt. Upon becoming khaghan, Ogodei immediately restarted this campaign. In 1229 he dispatched three fresh ttimen to clear out the lower Volga region in preparation for a major assault on the western end of Eurasia. The eastern Qipchaq tribes who occupied the area between the Volga and the Ural rivers put up unexpectedly fierce resistance to the invading armies, thus disrupting and delaying for several years the Mongols' plans for follow-up operations west of the Ural Mountains. In 1235, after discussion at the highest levels, the old warrior Siibetei was sent to the Volga with reinforcements. Arriving on the scene in 1236, he soon overwhelmed the opposition and then struck westward into Russia and the Qipchaq steppes, which were brought under Mongolian sovereignty by 124158 (see Map 27). In conformity with Chinggis khan's earlier directives, Ogodei dutifully assigned this vast tract of territory to Jochi's sons. Orda, the eldest, received the lands between the Irtysh and the Urals, and Batu, the second son, the Russian principalities and the western Qipchaq steppe. There were unfinished tasks, too, in the Middle East. In 1230 Ogodei placed one of his guardsmen, Chormakhan, in charge of Mongolian forces in the area and directed him to track down the elusive Jalal al-Din, who was still trying to organize an anti-Mongolian coalition in western Iran. After first consolidating his hold on Khurasan, Chormakhan moved into the 57 Secret history, sec. 269 (p. 209); Juvaynl/Qajvinl, vol. 1, pp. 144—9; Juvaynl/Boyle, vol. 1, pp. 183— 9; and YS, 2, p. 29. On Ogodei's new tide, see Igor de Rachewilez, "Qan, qa'an and the seal of Giiyiig," in Documents barbarorum: Festschrift fiir Walter Heissig zum 70. Geburtstag, ed. K. Sagaster and M. Weiers (Wiesbaden, 1983), pp. 2 7 2 - 8 1 . 58 Thomas T. Allsen, "Prelude to the western campaigns: Mongol military operations in the Volga-Ural region, 1217—1237," Archivum Eurasiae Medii Aevi, 3 (1983), pp. 5—24. Cambridge Histories Online © Cambridge University Press, 2008 500 km !S*V BALTIC Baidar annihilates Polish and German army under Duke Henry of Silesia Vladimir •Hii.1238 SEA Moscow* Kolomna BATU assembles his armies in the \ Bulghar Steppe 1238 Chmielnik Defeat of Polish Army 12.iii.1241 Pacification of Bulghars 1236-7 BAIDAR & KHAIDU SUBETEI crushes Hungarian army Southern army KHADA'AN Return route 1242 M A P 2 7 . Batu's invasion of Russia and Europe Cambridge Histories Online © Cambridge University Press, 2008 37° THE RISE OF THE MONGOLIAN EMPIRE Transcaucasus in pursuit of the last Khwarazmshah. In 1231 when news reached the Mongolian commander that this troublesome rebel had been killed by bandits, he redirected his energies toward Asia Minor, where the Seljuqs of Rum held sway. They, like the eastern Qipchaqs, gave a good account of themselves, and it was only in 1243, following long and arduous campaigning, that the Mongolian army in west Asia, now under the command of Baiju, Chormakhan's successor, could claim effective control of the Seljuq domains. In northeast Asia Ogodei mounted a sizable campaign against Korea (see Map 28). The Mongols had first entered the peninsula in 1218 during their conquest of Manchuria. Unable to withstand the invading force, the Koryo court agreed to pay an annual tribute in return for the withdrawal of the Mongolian army. Upon receipt of the first installment of tribute, the Mongols, as agreed, departed. In 1225, however, the Mongols' chief tax collector died under mysterious circumstances, and so Ogodei used this incident to justify launching a new attack on Korea in the fall of 1231. By December Mongolian troops had surrounded Kaegyong, the capital, forcing the Korean king to offer his submission. The Mongolian demands on their new subjects were onerous and unending, and in the summer of 1232 the Koreans revolted, killing the Mongolian agents (darughachi) stationed in the north of the country. The Korean ruler, realizing that Mongolian retaliation would not be long delayed, abandoned Kaegyong for the security of Kanghwa, an island just off the coast in the Yellow Sea. In response to these challenges, the Mongols launched another series of campaigns to compel Korean acceptance of their dominion. With time out for a truce between 1241 and 1247, the struggle went on until 1259, when the Koreans finally acquiesced to the foreign occupation." Despite these heavy commitments elsewhere in Eurasia, Ogodei was also determined to deal with yet another piece of unfinished business - the destruction of Chin. When Mukhali died in 1223, the Chin had rallied and retrieved some territories occupied by the invaders. Intent at the time on punishing the Tanguts, the Mongols were forced to restrict their operations in north China. Although sporadic fighting continued through the later 1220s over various frontier regions, the Mongols made no attempts to deal a death blow to the Chin dynasty in their last bastion, Honan. Ogodei began preliminary operations against the Chin in 1230. The Mongols were soon made painfully aware that the Jurchen state, though much battered, was still capable of mounting a vigorous defense and that new plans 59 Gari Ledyard, "The Mongol campaigns in Korea and the dating of the Stcrtt history of the Mongols," Central Asiatic Journal, 9 (1964), pp. 1 —16. Cambridge Histories Online © Cambridge University Press, 2008 THE REIGNS OF OGODEI AND Gt)Y(JG Liao-yang (Liao-tung Hsing-sheng) KANGHWA Seat of royalcourt (1232—70) Invasion of 1231 Invasion of 1236-39 "* Invasion of 1253-54 MAP 28. The Mongolian invasions of Koryo Cambridge Histories Online © Cambridge University Press, 2008 371 372 THE RISE OF THE MONGOLIAN EMPIRE would have to be devised and additional forces gathered to deliver the coup de grace. The Mongols therefore reorganized their armies and in 1231 put into action a new strategic plan. The left wing of the army, led by Siibetei, took up position in Shantung, the center, under Ogddei, moved into Shansi, and the right wing, commanded by Tolui, advanced into Shensi. The latter force subsequently entered northern Szechuan and asked the Sung government's permission to cross its territory preparatory to a surprise attack on K'ai-feng from the southeast. This huge enveloping movement was designed primarily to avoid the necessity of reducing the formidable T'ung-kuan forts that commanded the western approaches to the Chin capital. Their request for rights of transit denied, the Mongols simply forced a passage along the Sung border. In early 1232 Tolui's troops routed a major Chin force at San-feng and then resumed their march on K'ai-feng. By summer all three columns were concentrated around the capital, and Siibetei assumed control of the combined Mongolian armies. Following the transfer of command, both Ogodei and Tolui became seriously ill. The khaghan, who was on his way back to Mongolia when the sickness struck, recovered, but his brother, who had accompanied him north, weakened and died later that year. Siibetei, in the meanwhile, increased pressure on the beleaguered city throughout the fall, and in February 1233, the Chin emperor, Ai-tsung, fled the capital. Several weeks later, the abandoned and demoralized garrison capitulated, and in May the gates of K'ai-feng were opened to the Mongols. The Chin emperor, now in desperate straits, took refuge in Ts'ai-chou in southwest Honan and appealed to the Sung court for assistance. The latter, hoping to regain territories long lost to the hated Jurchens, rejected the Chin overtures and instead negotiated an alliance with the Mongols. Despite increased pressure, the city continued to hold out. In the fall of 1233 the combined Mongolian—Sung armies converged on Ts'ai-chou; after several months of resistance Ai-tsung, realizing that there was no escape and no hope of victory, killed himself. Shortly thereafter, on 9 February 1234, Ts'ai-chou fell and, with it, the Chin dynasty. In the aftermath of the fighting, the Sung, eager to benefit from the Jurchen collapse, made an ill-advised attempt to occupy the whole of Honan. The Sung forces, woefully inadequate to the task, were soon put to flight by the Mongols, who had no intention of sharing the spoils of victory with their recent ally. Administrative restructuring To govern this huge and continually expanding domain, the new emperor, following precedent, fashioned his central administrative apparatus from the Cambridge Histories Online © Cambridge University Press, 2008 THE REIGNS OF OGODEI AND G O Y O G 373 personnel of the guard/household establishment, the majority of whom came to him as part of his father's legacy.60 At the head of his central chancellery, Ogodei selected the Nestorian Christian Chinkhai (Chen-hai), originally a chamberlain in his father's household establishment. A Kereyid (though in some sources he is said to be a Uighur), he had been in Mongolian service at least since 1203. In his early career he was given a series of military and administrative commissions, which he successfully discharged. It was not, however, until the beginning of Ogodei's reign that Chinkhai suddenly rose to prominence as chief minister of the empire. Many other key officials in his administration, most notably Yeh-lii Ch'u-ts'ai, were recruited in a similar manner. Not surprisingly, then, the style of government at the center nonetheless remained much as it had been under Chinggis khan. On the regional level of government, however, the acquisition of numerous sedentary subjects, both urban and rural, necessitated some significant adjustment of the machinery through which the empire was administered. In Chinggis khan's day the Mongols had been content to place newly conquered sedentary populations under the control of the responsible theater commander, who functioned as an all-powerful governor-general, as, for example, Mukhali had done in north China. Nonnative specialists, for example, Uighurs in China and Chinese in Turkestan, were brought in to assist the Mongolian administration, but this did not eliminate the conquerors' heavy dependence on native administrative institutions and personnel, which they systematically co-opted for their own purposes. To ensure that these local elites loyally obeyed orders and advanced imperial interests, the Mongols placed special officers, darughachi, in key population centers, in auxiliary military commands, and at the courts of dependent states. In the early years of the empire these officers, who oversaw census taking, tax collection, military recruitment, and the like, were selected from among the grand khan's nokod. The first darughachi mentioned in the sources is Jabar Khoja (Cha-pa-erh Huo-che), who was posted to Peking sometime between June 1214 and May 1215. 61 The institutional roots of this office are not known with certainty, but it has been connected with the Chin office of hsing-sheng and with the basqaqs of the Khara Khitai kingdom, whose duties paralleled those of the later darughachi. A word of Turkic origin, basqaq is an 60 Secret history, sec. 269 (p. 204), says that those who guarded Chinggis khan were transferred to Ogodei upon his elevation. Although it is true that the bulk of the kesig went to Ogodei, the Shu 'ab-ipanjganah (an unpublished genealogical supplement to Rashld al-DIn's_/am/ 'al-Tavarikh), folio 105I— io6r and I27r-1, which provides a lengthy list of the individuals inherited by all four sons, makes it clear that the "personal thousand" of Chinggis khan, the most elite contingent in the imperial guard, was transferred to Tolui. Compare also Rashid/Karimi, vol. 1, p. 555; and Rashld/ Boyle, p. 163. 61 YS, 120, p. 2961. Cambridge Histories Online © Cambridge University Press, 2008 374 T H E RISE OF THE MONGOLIAN EMPIRE exact semantic equivalent of the Mongol darugbachi, both of which words mean "one who presses."62 Ogodei's major innovation in the system that he inherited was to reduce the administrative responsibilities of the theater commanders, whose primary duty after all was military conquest, and to turn over these tasks to full-time "civilian officials," who could devote their energies exclusively to fiscal and administrative matters. In 1229 he created the first two such civilian-led Branch Secretariats, one for eastern and western Turkestan and the other for occupied north China. Later on, once a firm Mongolian bridgehead had been established in the Middle East in the 1240s, a third secretariat was established in northern Iran. The purpose behind this restructuring was to assert the grand khaghan's primacy over the riches of the empire's sedentary sector, particularly over tribute and revenues that were being siphoned off by members of the imperial family at the local level. The interests of the latter, though certainly subordinated, were by no means ignored under the new system, and by the end of Ogodei's reign it had become established practice to allow interested regional khans and other imperial princes to place their personal representatives on the staffs of the Branch Secretariats and to have a voice in the selection of their chief executive officers. The work of these administrations was thus to some extent a joint undertaking of the entire Chinggisid line, with the grand khan enjoying the status of primus inter pares. This joint system of governance, it appears, first evolved in the Turkestan administration, in consequence of disputes that arose between Ogodei and the regional khan, Chaghadai, and was later applied in China and Iran.6* The first head of the Turkestan chancellery was Mahmud Yalavach, a Turkish-speaking Khwarazmian merchant who entered Mongolian service in 1218 as a diplomatic envoy (hence his name, Yalavach, Turkic for messenger). All of the sedentary regions between the Aral Sea and the Tangut lands fell within his jurisdiction. Mahmud Yalavach's appointment established an enduring family tradition of administrative service to the Chinggisid dynasty; for at least four generations his numerous descendants were continuously in the employ of various Mongolian princely lines in Turkestan and China.6" 62 On the office of darugbachi, see Francis W. Cleaves, "Darugba and gerege," Harvard Journal of Asiatic Studies, 14 (1951), pp. 4 9 3 - 5 2 6 . On the possible relationship between darughachi and basqaq, see Istvan Vasary, "The origins of the institution of basqaqs," Ada Orientalia Acadtmiae Scitntiarum Hungaricae, 32 (1978), pp. 2 0 1 - 6 . 63 Paul Buell, "Sino-Khitan administration in Mongol Bukhara," Journal ofAsian History, 13 (1979), pp. 121— 51, was the first to point out the collegial nature of these Branch Secretariats, which he calls "joint satellite administrations," and to trace their evolution. 64 For a brief history of this family, see Thomas T. Allsen, "Mahmud Yalavach," in vol. 1 of Yuan personalities, ed. Igor de Rachewiltz and Hok-Iam Chan (forthcoming). Cambridge Histories Online © Cambridge University Press, 2008 THE REIGNS OF OGODEI AND GOYOG 375 Mahmud Yalavach's counterpart in north China was the renowned Yeh-lii Ch'u-ts'ai, a sinicized Khitan descended from the Liao ruling house. A man of broad intellectual and spiritual attainments, he was an adherent of both Confucianism and Ch'an Buddhism. Like his father before him, he led an active public life, serving the Chin dynasty in a variety of administrative capacities. Yeh-lii Ch'u-ts'ai was in Chung-tu when it fell in 1215, and three years later, by imperial command, he traveled to Mongolia for an audience with Chinggis khan. Greatly impressed with the Khitan, the Mongolian ruler placed him in his retinue as a scribe (.bichechi) and court astrologer. In 1219 Yeh-lu Ch'u-ts'ai accompanied his master to Central Asia and did not return to China until 1226. During the interregnum the regent, Tolui, sent him to the former Chin capital to quell a disturbance that had broken out in the vicinity. He soon accomplished his mission and returned to Mongolia in time for Ogodei's enthronement. Here, in 1229, north China was given into his charge.6' Yeh-lii Ch'u-ts'ai and his reforms As a loyal servitor of the Mongols for over a decade and as a man intimately familiar with the Chinese scene, Yeh-lii Ch'u-ts'ai possessed all of the requisite qualifications for the new post of chief tax collector for north China created by Ogodei. His appointment, however, did not sit well with some elements of the Mongolian hierarchy, who rightly feared that the khaghan's main purpose in placing a man with Yeh-lii Ch'u-ts'ai's background and views into such an office was to assert direct imperial control over revenue collection at the expense of their own interests. This concern was shared as well by the Chinese military commanders in Mongolian service. Even though the latter at times actively supported Yeh-lii Ch'u-ts'ai's efforts at reviving Chinese cultural life, they, too, had become accustomed to governing their domains with minimal external interference in the chaotic conditions of the 1220s. Like their Mongolian counterparts, they therefore tended to view with deep suspicion any attempt at administrative or fiscal centralization.66 Yeh-lii Ch'u-ts'ai's disturbing proposals for fiscal reform first emerged during debate concerning the general direction of imperial policy that developed at the beginning of Ogodei's reign.6? At these discussions Begder (Piehtieh), a court official and spokesman of an extremist faction, seriously put 65 Igor de Rachcwiltz, "Yeh-lii Ch'u-ts'ai (1189-1243): Buddhist idealist and Confucian statesman," in Confucian personalities, ed. Arthur F. Wright and Denis C. Twitchett (Stanford, 1962), pp. 1 8 9 - 2 0 1 . 66 Hsiao Ch'i-ch'ing, "Yen Shih, 1182—1240," Papers on Far Eastern History, 33 (1986), pp. 119—22. 67 On Yeh-lii Ch'u-ts'ai's fiscal and administrative reforms, see Rachewiltz, "Yeh-lu Ch'u-ts'ai," pp. 2 0 1 - 7 ; and Nikolai TS. Munkuev, Kitaiskii istochnik 0pervykh mongol'skikh kbanakh (Moscow, 1965), pp. 34-6. Cambridge Histories Online © Cambridge University Press, 2008 376 THE RISE OF THE MONGOLIAN EMPIRE forward a plan to depopulate north China and convert its agricultural lands into pasture for Mongolian herds! Yeh-lu Ch'u-ts'ai was able to turn aside this appalling proposal by arguing forcibly that if a regular scheme of taxation were introduced into the area, revenues would increase dramatically, much to the long-term benefit of the imperial coffers. Persuaded that such a program, if successful, would enhance his power as khaghan, Ogbdei offered the Khitan an opportunity to implement his policy recommendations on a trial basis. Yeh-lii Ch'u-ts'ai had won this round in what proved to be an ongoing debate at court and now had to demonstrate that his measures would generate revenue at the promised levels. The first formal step in the implementation of Yeh-Lii Ch'u-ts'ai's plan was taken toward the end of 1230, when Ogbdei appointed revenue commissioners (cheng-shou k'o-shui shih) for each of the ten circuits (lu) in occupied China.68 All were of Chinese origin, for the most part former officials of the Chin dynasty. They were to oversee the collection of taxes on the basis of an entirely new system devised by Yeh-lii Ch'u-ts'ai. Under the new regime every adult would pay a fixed tribute (ch'ai-fa) assessed in silk yarn, the rate for villagers being considerably higher than that levied on urban dwellers. Cultivators also had to pay a set amount of grain per household, regardless of the amount and quality of their land, whereas those in the cities paid a supplemental levy in silk yarn to provide food for official couriers in transit. For the most part, taxes assessed in silk yarn were commuted into silver for payment to the government. Though hardly equitable, this system did establish firm categories of taxation and did specify clearly the basis of assessment.69 As intended by their author, taxes - at least in theory - were now regular and predictable. More important from the Mongolian point of view, they did indeed produce more revenues. Ogodei was so delighted with the results that in 1231 he named Yeh-lii Ch'u-ts'ai his chief of the Secretariat {chung-shu ling); that is, he was given responsibility for the overall administration of north China. In his new capacity, however, he was still subordinate to the Central Secretariat headed by Chinkhai, who had to countersign all official documents issued by the chung-shu ling before they could be considered valid.i° Nevertheless, Yeh-lii Ch'u-ts'ai's fiscal success had strengthened his hand, and he used his increased influence to push for yet more reforms. 68 YS, 2, 36. 69 The fullest account of early Mongolian fiscal practices is contained in the Hei la shih liieh, in Meng-ku shih liao ssu chung, ed. Wang Kuo-wei (1926; repr., Taipei, 1962, 1975), I3a-b, a work prepared by P'eng Ta-ya and Hsu T'ing, two Sung envoys who traveled in north China between 1234 and 1236. For a translation and analysis of the relevant passages, see Herbert F. Schurmann, "Mongolian tributary practices in the 13th century," Harvard Journal of Asiatic Studies, 19 (1956), pp. 312—18. 70 P'eng, Hei ta shih liieh, 10a; and Cbinesische Gesandtenbericbte, p. 133. Cambridge Histories Online © Cambridge University Press, 2008 THE REIGNS OF OGODEI AND GOYOG 377 Although conditions in north China had been improved somewhat by these early measures, many problems remained unresolved, and new ones soon surfaced. For one thing, the Mongols were not long satisfied with the tax rates as originally set; between 1231 and 1234, for example, the grain assessment jumped from two to four bushels per household. Further, the Mongols were unable to refrain from their old habit of exacting extraordinary levies above and beyond the regularly scheduled taxes. These difficulties arose in part because it was the practice of the Mongolian authorities first to calculate state expenditures and then to establish tax quotas to meet their budgetary requirements. As expenses increased, new tax rates were set, or special levies were introduced at short notice to make up the expected shortfall. Such deficits could be generated by changes in imperial policy, by new military campaigns, for example, but many were simply the result of the personal greed of the Mongolian leadership. Tribute in the form of silver that came into their possession was regularly invested with Hui-hui (Inner and Central Asian Muslim) merchants, who then either used this capital to buy goods for trade or lent it to the populace at extremely high interest rates. These partnerships (called ortaq in Turkic, wo-t'o in Chinese, and ortogh in Mongolian) between princes and merchants were frequently very profitable, and so members of the Mongolian hierarchy were always anxious to acquire additional capital for new commercial ventures. 71 Among the Mongolian leadership there was no worse offender in this regard than the khaghan himself. Ogodei freely gave huge sums of cash to ortogh merchants for investment. On several occasions he reportedly gave individual merchants five hundred ingots of gold or silver (balhh in Persian) for this purpose. To indicate the magnitude of this sum, it must be remembered that in 1230 Yeh-lii Ch'u-ts'ai collected only ten thousand silver ingots (Chinese: ting) from his entire jurisdiction. Despite the protests of his ministers, particularly Mahmud Yalavach, the khaghan continued this practice to the end of his reign. 72 With monies leaving the treasury at this rate, it is not surprising that there was always pressure to readjust the tax quotas upward. The misery caused by higher taxes was not, however, the only mischief caused by the ortogh merchants. As agents of the khaghan or of a powerful prince, they used their connections to extort goods and monies from the population. A frequent trick employed by these merchants in north China was to claim falsely but vociferously that goods they had purchased with a prince's money had been stolen and thus to force the local population, who feared official retaliation, to make good their "losses." 71 P'eng, Her la shih liieh, I5a-b; and Chinaische Gaandtenberichtt, p. 152. 72 Juvaynl/Qazvlni, vol. 1, pp. 166—6, 170—7; and Juvayni/Boyle, vol. 1, pp. 209-10, 213-15. Cambridge Histories Online © Cambridge University Press, 2008 378 THE RISE OF THE MONGOLIAN EMPIRE The fact that there continued to be a large floating population in north China after 1231 was a further and sure sign that conditions had not yet returned to normal. Despite the reforms, many people (one contemporary estimate says 50 percent of the total population) still had a great incentive to abandon their homes simply to avoid the seemingly endless demands of the state and the depredations of its agents. Others had disappeared from the tax rolls because they were made slaves and retainers of high Mongolian officials. Yeh-lii Ch'u-ts'ai, sensitive to this problem, proposed to Ogodei in 1234 that a census be taken in north China to discover the hidden and floating population in order to return them to their homes and the tax rolls. The khaghan agreed and deputed Shigi Khutukhu, the original keeper of the Blue Register, to conduct an enumeration. The census, which got under way in 1234, occasioned new debate at court over future revenue policies. Even though he was satisfied with the established categories of taxes, Yeh-lii Ch'uts'ai argued for fundamental changes in the methods of assessment. In his opinion, tribute should henceforth be collected on the basis of the household, not individual adults, as had been the practice in China since 1231. The Mongols, however, preferred the system of individual assessment, a method that was still in effect in Turkestan, where Mahmud Yalavach had introduced his own fiscal reforms in the early 1230s. In the end, a compromise was reached whereby the older poll tax assessed on individual adults was greatly diminished and a new household tax, also to be paid in silk, was introduced. On the whole, Yeh-lii Ch'u-ts'ai had won his point. In regard to the agricultural tax, his suggested revision was accepted as it stood. The levy on grain, initially assessed at a flat rate on each household, would now be based on the quantity and quality of land owned. When the census was completed in 1236, the new system was introduced. The results were gratifying: With more equitable methods of assessment in place, the burden on individual households was substantially reduced, in some cases by as much as 90 percent. But thanks to the expansion of the tax base following the conquest of Honan in 1234 and the registration of the hitherto hidden or floating population, total government revenues were maintained. Yeh-lii Ch'u-ts'ai was now at the height of his power and influence, but trouble was brewing on the horizon. The eclipse of Yeh-lti Ch'u-ts'ai Following an initial period of imperial activism, from the mid-i23os onward, Ogodei became less and less interested in the administration of his realm. As he began to devote himself to drinking, dalliance, and extravagance, local and regional interests energetically reasserted their influence. Cambridge Histories Online © Cambridge University Press, 2008 THE REIGNS OF OGODEI AND G O Y O G 379 The turning point in this struggle between the proponents of princely autonomy and the partisans of centralized imperial control were the years 1236 and 1237. The first unmistakable indication that Yeh-lii Ch'u-ts'ai was beginning to lose the khaghan's ear came in 1236, when Ogodei decided to enlarge dramatically the system of princely appanages (Mongolian: khubi; Chinese: fen-ti) in north China. By imperial decree all senior princes and princesses were to receive large tracts of agricultural land as a supplemental sources of personal income. The sons of Jochi, for example, were granted 41,302 households in P'ing-yang, and Chaghadai was given 47,330 households in T'ai-yiian. 7 ' Although Yeh-lii Ch'u-ts'ai did manage to extract another decree from Ogodei that stipulated that tax collection and military recruitment within individual appanages were to remain under the jurisdiction of the imperial court, the granting of khubi on this vast scale constituted a major setback for the advocates of centralization. As Yeh-lii Ch'u-ts'ai had no doubt foreseen, the central government, without an active and energetic khaghan intent on enforcing his will, did not in practice exercise much authority within the appanages, and the members of the attached households, lacking any form of protection, were subject to endless abuse and exploitation. Yeh-lii Ch'u-ts'ai had undeniably lost a vital battle but persisted nonetheless in his efforts at reform. This time he turned his attention to administrative reorganization. The Mongols themselves had already taken a step in this direction. With the collapse of the Chin dynasty in 1234, the Mongols soon became aware that their own law, the jasagh of Chinggis khan, had limited utility in regulating a sedentary society, and they therefore decided to apply throughout their Chinese territories the Chin statutory code, the T'ai-ho lii, first promulgated in 1201 and designed on the basis of T'ang models. Although this was a welcome development, Yeh-lii Ch'u-ts'ai now had in mind plans of much greater scope that would lead eventually, he hoped, to the full restoration of Confucian-style government in north China. To achieve this objective, Yeh-lii Ch'u-ts'ai first sought Ogodei's permission in 1237 to hold civil service examinations throughout the north as a means of restoring the Chinese literati, who had suffered great privation and loss of status in the preceding decades, to their accustomed role in government. The khaghan gave his blessing for this phase of the program, and Yeh-lii Ch'u-ts'ai in the following year organized examinations in all circuits. Over four thousand applicants passed (one-fourth of whom were slaves or prisoners of war at the time of the examination), but few of the successful candidates, much to the 73 YS, 2, p. 35195. P- 24I4- Cambridge Histories Online © Cambridge University Press, 2008 380 THE RISE OF THE MONGOLIAN EMPIRE Khitan minister's bitter disappointment, were ever called for actual service.74 Instead, for the most part they were used as administrative advisers in their home localities. The Mongols had no intention of turning over the governance of north China (or any other conquered territory) entirely to indigenous officials. In fact, the role of foreign administrative specialists, principally Uighur and Turkestanis, in the government of north China continued and even grew in the coming years. Yeh-lii Ch'u-ts'ai's inability to gain acceptance of his administrative program was a further demonstration of the limits of his influence. Future events would show that not only was he unable to extend his reforms but that the measures already in place also were by no means immune from attack, particularly his fiscal policies so painfully implemented over the preceding decade. At the center of this assault stood the Inner and Central Asian merchants whose exploitative practices Yeh-lii Ch'u-ts'ai had never been able to curb effectively. Their influence among the Mongolian hierarchy, always strong, steadily increased at Yeh-lii Ch'u-ts'ai's expense, and in 1239 Ogodei was persuaded to farm out the taxes of north China to a Muslim merchant, Abd al-Rahman (Ao-tu-la-ho-man, An-tu-la-ho-man), thus bypassing the regular system of revenue collection. The tax quota established for that year, not unexpectedly, was raised substantially to 44,000 ingots of silver. Early in the following year, the khaghan, pleased with the merchant's results, placed the tax farmer in charge of the tax bureaus of north China.75 The victory of the merchant faction was now temporarily complete. Some idea of the baleful impact of the new policy on China's population can be gleaned from an edict that Ogodei felt constrained to issue toward the end of 1240. According to this document, the populace and officials had been compelled to borrow large amounts of money from the ubiquitous Inner and Central Asian (Hui-ho) merchants in order to meet their increased tax obligations. The rates charged were so high that the interest often equaled the principal within a year's time. In light of this, the khaghan "magnanimously" decreed that henceforth the interest could not legally exceed the original amount borrowed.76 Under Abd al-Rahman's regime, the ortogh merchants - operating simultaneously as tax farmers and moneylenders - were clearly reaping huge dividends at the expense of the Mongols' Chinese subjects. It is difficult, in fact, to imagine a more ruinous and exploitative economic system than the one 74 On the plight of the Confucians in this period, see Makino Shuji, "Transformation of the shih-jen in the late Chin and early Yuan," Ada Asiatica, 45 (1983), pp. 1—26. 75 YS, 2, p. 36; and Shmg wu ch'in cbeng lu cbiao cbu, in Meng-ku shih liao ssu chung, ed. Wang Kuo-wei (1926; repr. Taipei, 1962, 1975), p. 106b. 76 YS, 2, p. 37. Cambridge Histories Online © Cambridge University Press, 2008 THE REIGNS OF OGODEI AND GOYOG 381 prevailing in north China at this time: Tax farmers placed high bids with the court for the right to collect taxes, thus forcing the tax quotas ever upward. The merchants, of course, collected as much as possible over the set quotas to maximize their profits. Because few people could pay such elevated rates, they were forced to turn to ortogh merchants (themselves tax farmers) for loans on usurious terms. The capital of the latter, it should be remembered, was provided by the Mongolian court or by imperial princes who had obtained it in the first instance in the form of taxes from their hapless subjects! Thus, by the end of Ogodei's reign, Yeh-lii Ch'u-ts'ai's influence at court had evaporated, and his program of reform was in shambles. He retained his title oi chung-shu ling and continued on as royal astrologer, but he was no longer consulted on affairs of state. The most revealing indication of his eclipse came in 1241, when Ogbdei was finally moved to reverse his destructive financial policies in north China. Conditions had so deteriorated that the khaghan determined to remove Abd al-Rahman from office and reinstitute a more regular system of tax collection. Having come to this conclusion, however, it is significant that Ogodei turned not to Yeh-lii Ch'u-ts'ai but to another Central Asian Muslim, Mahmiid Yalavach (Ya-lao-wa-ch'ih), a Turkic-speaking Khwarazmian, who until 1239 had been head of the Branch Secretariat of Turkestan. Evidently, Ogodei felt that he could not, in the prevailing political climate at court, return the administration of north China to Yeh-lii Ch'u-ts'ai or any other champion of Chinese interests. It should not be assumed, however, that in bypassing the Khitan for Yalavach the court had opted for a continuation of Abd al-Rahman's policies, even in modified form. Mahmiid Yalavach, though he never stood high in the valuation of his Chinese associates, was a reformer in his own right. He had regularized the collection of taxes in Turkestan, opposed the extravagance of Ogodei's court, and fought against the extension of the appanage system within his former jurisdiction. Although it is true that he was unable to curtail the rampant official corruption and ceaseless quarreling among the appanage holders during his short tenure in office — late winter of 1241 to spring of 1242 - his appointment nonetheless represented a change of imperial policy in north China. 77 In sum, then, even though Yalavach's policies paralleled those of Yeh-lii Ch'u-ts'ai in many respects, his assignment in place of Abd al-Rahman clearly signaled the end of the Khitan's personal influence. Yeh-lii Ch'u-ts'ai died in 1243 in Khara Khorum during the regency of Toregene (1241-6), honored but unheeded. 77 For a brief and generalized description of conditions in north China in this period, see Yao Sui, Mu an chi (SPTK ed.), 15, p. 4a. Cambridge Histories Online © Cambridge University Press, 2008 382 THE RISE OF THE MONGOLIAN EMPIRE The death of Ogodei and the regency of Toregene Following the completion of the Chin campaign, in which the khaghan had actively participated, the imperial princes urged Ogodei to remain in Mongolia, to live the good life, and to enjoy the vast bounty generated by the Chinggisids' successful imperial venture. The khaghan, desiring to lead the forthcoming campaign against the Russian principalities, at first rejected their entreaties but was "persuaded" in the end to heed their advice.78 From the mid-i23os onward, Ogodei, bowing to the princely pressure as well as to his personal inclinations, led a life of luxury and leisure, spending most of his time hunting and drinking. The resulting political vacuum was eagerly filled by the khaghan's second wife, Toregene, who soon consolidated her authority at court and began issuing imperial edicts in the name of her rapidly failing husband. By the end of his reign Ogodei was drinking so heavily that a special court functionary was named to count and thereby control the number of wine goblets that he emptied daily. Such measures, however, were to no avail, and on 11 December 1241, while on a hunting trip, Ogodei succumbed at age fifty-six, following a lengthy drinking bout. The second khaghan of the Mongolian empire was buried apparently in or near his personal appanage in Jungaria, and not with his father at Burkhan Khaldun.79 According to Mongolian social custom, when a male head of household died, his widow was to administer his estate in trust and to assume his authority until the eldest son came of age. The Mongolian ruling house continued to respect this principle, long sanctioned by social convention, utilizing it as a mechanism for the transference of political authority within the early Mongolian empire. The empire itself, in other words, was looked upon as the patrimony of the imperial family and treated accordingly. Thus, when Ogodei died his widow, Toregene, already in firm control of the court, also took up the administration of his estate, that is, acted as regent of the empire until such time as a new male head of the imperial family, a khaghan, could be confirmed in office by a khuriltai.8o Once Toregene, known in the Chinese sources as Liu Huang-hou (Sixth Empress Dowager),81 assumed office, she endeavored to engineer the en78 Juvaynl/Qazvinl, vol. i , p p . 156-7; and Juvaynl/Boyle, vol. i,pp- 198—9. 79 See John A. Boyle, "The burial place of the great khan Ogedei," Ada Orimtalia, 32 (1970), pp. 45—50. 80 This principle was not applied following the death of Chinggis khan because his principal wife, Borte, the mother of his potential successors, had predeceased him. Furthermore, Chinggis khan's final illness was of brief duration, and none of his secondary wives had the opportunity to establish her ascendancy at court. 81 In all likelihood, her actual title was not liu buang hou (sixth empress dowager) - she was in fact his second empress — but ta huang hou (great empress dowager), which is a direct translation of her Mongolian title, yeke khatun. Apparently this confusion arose among Chinese writers of the thirteenth Cambridge Histories Online © Cambridge University Press, 2008 THE REIGNS OF OGODEI AND G O Y O G 383 thronement of her son Giiyiig (r. 1246—8), in the face of considerable opposition. Those resisting his candidacy, led by Batu, Jochi's second son and khan of the Golden Horde, were in the end unable to prevent Giiyiig's elevation, but they did manage to delay a final decision on the matter for about four and one-half years. And throughout this artificially prolonged interregnum, the able and determined Toregene ruled over the affairs of the empire in her deceased husband's name. In the military sphere, however, Toregene's authority was somewhat circumscribed. Major military undertakings in progress at the time of Ogodei's death, such as the invasion of central Europe, were automatically halted, as the Mongolian princes and most high-ranking commanders had to return home to confirm a successor. Some campaigning was resumed under Toregene, but in all cases the objectives as well as the scale of the operations were clearly limited. Baiju, for instance, was allowed to complete his destruction of the Seljuqs of Rum, and an operation of equally narrow scope was mounted against Sung-held territories south of the Huai River, which culminated in the Mongol occupation of Shou-chou (modern Shou-ch'un in Anhui) in 1245. 82 In administrative and fiscal matters, however, there are many indications that the regent enjoyed wide powers and that she chose to exercise this power to its fullest extent. In 1244, for example, Toregene authorized the issuance of new coinage in the Transcausasian provinces in order to facilitate the collection of the poll tax, an initiative that anticipated Mongke khaghan's (r. 1251—9) more systematic and extensive efforts to monetize imperial revenues in the next decade. 8 ' That the regent was no mere caretaker is evidenced further by her personnel policies. She dismissed Chinkhai (Chen-hai) and many other central government officials from office and made sweeping changes, too, in the leadership of the Branch Secretariats: Korgiiz in Iran, Masud Beg in Turkestan, and Mahmud Yalavach in China, all appointees of Ogodei, were soon removed. Yalavach, much hated by Toregene, was forced to flee for his life from her agents, finally finding sanctuary with Koten, Ogodei's second son. Toregene replaced the departed Khwarazmian with the previously dismissed Abd al-Rahman. China was once again delivered into the hands of the rapacious tax farmer. Although little is known of the economic and social policies pursued in north China during Toregene's regency, it is a reasonable century, because the characters for six (//'») and great (to) are graphically similar. See Igor de Rachewiltz, "Some remarks on Toregene's edict of 1240," Papm on Far Eastern History, 23 (1981), pp. 42-3. 82 YS, 2, p. 38. 83 M. A. Seifeddini, "Monety s nadpis'iu 'uluq mangul ulus bek'," Numizmatika i Epigrafika, 9 (1971), pp. 115-21. Cambridge Histories Online © Cambridge University Press, 2008 384 THE RISE OF THE MONGOLIAN EMPIRE conjecture that they were similar in character to those of Abd al-Rahman's first administration. The elevation ofGilyiig and dhsidence among the imperial princes Although her personnel changes, undercutting as they did the collegial character of the Branch Secretariat, displeased some of the Mongolian princes, none openly challenged Toregene on this issue. Her power, as her opponents recognized, was transitory in nature, and her more unpalatable policies could always be reversed at some future date. On the question of succession, however, Toregene encountered strong resistance. Because the Mongols' weakly developed system of determining succession, nomination by the ruling khaghan, was easily circumvented in practice, any transference of imperial power inevitably became the focal point of intense political struggle and in the long run led to armed conflict among the princes, all of whom, in principle at least, were eligible for the throne.84 Therefore, because the long-term interests of the entire imperial family, as well as the personal ambitions of its individual members, were at stake, in order to obtain her ends Toregene was forced to negotiate this critical issue with all the interested parties. The major sources of opposition to Toregene's plan to place her eldest son, Giiyiig, on the throne were two. First, because Ogodei had nominated his grandson Shiremiin (who was the eldest son of Kochii, the deceased khaghan's third son), there was discontent and resistance within the regent's own family.8' This explains why Toregene'e enemies, such as Mahmud Yalavach, were able to find safety with other Ogodeid princes and why she felt compelled to remove her husband's ministers. Although Shiremiin's candidacy was successfully derailed by Toregene, a second and more formidable line of resistance immediately surfaced among collateral branches of the royal house. Giiyiig had made powerful enemies among his fellow princes, the most important of whom was Batu, the khan of the Golden Horde and the acknowledged leader of the Jochids. The two openly despised each other; the cause of their deep-seated enmity went back to personal disagreements that broke out during the campaign of 1236—41 in western Eurasia. At issue was a question of seniority and precedence, which became so heated that Ogodei was forced to intercede to settle the matter in Batu's favor.86 This naturally 84 On the nature of succession struggles among steppe peoples, see Joseph Fletcher, "Turko-Mongolian monarchic tradition in the Ottoman empire," Harvard Ukrainian Studies, 3 - 4 (1979—80), pp. 2 3 6 - 5 1 . 8; Rashld/Karimi, vol. 1, p. 443; and Rashid/Boyle, p. 19. 86 Secret history, sees. 275—6 (pp. 215-17). Cambridge Histories Online © Cambridge University Press, 2008 THE REIGNS OF OGODEI AND GtJYtJG 385 further embittered the relationship between the two princes and between Giiyiig and his father and may have been at the root of Ogodei's decision to bypass his son in favor of his grandson Shiremiin. When, in 1241, word reached the still-feuding Mongolian princes in the west that Ogodei was dangerously ill, Giiyiig immediately left the campaign and hastened to his father's side, presumably to patch up their differences and thereby advance his own claims to the khaghanate. The emperor died before his son could reach him, but Toregene had already taken positive action on her son's behalf. With perhaps unseemly haste, given her husband's recent demise, she hurriedly convoked a khuriltai in the fall of 1241 in the hope of securing quuick approval of Giiyiig's succession. Although he harbored no imperial ambitions of his own, Batu was determined to deny the throne to his hated adversary, Giiyg. He therefore declined the invitation to join the khuriltai, claiming a severe attack of gout as an excuse. By such tactics the Jochids managed to delay Giiyiig's enthronement for several years.8? In the meantime, Toregene continued to press her case, using her position and influence to acquire support for her son. After much intrigue and political infighting, the regent secured the necessary consensus within the imperial house, and in the suummer of 1246, a new khuriltai held on the Keriilen duly elevated Giiyiig to the emperorship. Batu, bitterly resentful, resisted to the end, refusing to appear at Giiyiig's enthronement, again claiming illness, and sending in his stead his older brother, Orda, to represent his family at the official ceremonies. Open conflict between the antagonists had been avoided, but the Jochids' acceptance of the new khaghan had been tendered grudgingly and reluctantly. The extent of the opposition to Giiyiig's elevation is revealed in the statement by the author of the Tartar relation, a contemporary of these events, that the new emperor had been elected "by a majority of one." 88 This, of course, was not literally true, but it did accurately mirror the growing political tension among the Mongolian princes. Further evidence of the division within the ruling house is the fact that another claimant for the throne, Temiige Odchigin, the youngest brother of Chinggis khan, hoping to profit from the widespread dissension, tried to seize the khaghanate for himself without benefit of a properly constituted khuriltai. As the Franciscan friar Carpini phrased it, because "he wanted to rule without an election," he was put to death with the concurrence of the assembled princes.89 For the 87 Rashld/Kariml, vol. i, pp. 523—4; and Rashid/Boyle, p. 120. 88 R. A. Skelton et al., The Vinland map andthe Tartar relation (New Haven, 1965), p. 84. 89 Christopher Dawson, ed., The Mongol mission: Narratives and letters of the Franciscan missionaries in Mongolia and China in the thirteenth and fourteenth centuries (New York, 1955), p. 25 (hereafter cited as Mongol mission). Cambridge Histories Online © Cambridge University Press, 2008 386 THE RISE OF THE MONGOLIAN EMPIRE first time a member of the royal clan had been killed as a consequence of internal power struggles. Giiyiig thus assumed office in a climate of hostility and suspicion. Subsequent succession crises would prove even more costly in human life and in the end would undermine the unity and power of the empire of the Great Mongols. Giiyiig's administration of the realm At the time of his enthronement, Giiyiig was forty years of age. Like most Mongolian princes, his youth had been spent on military campaign; he had fought against the Chin and participated in the conquest of the northern Caucasus in 1239-40. As far as one can tell from the meager evidence, he had had little or no civil administrative experience before his elevation. Giiyiig commenced his reign by showering his supporters, from princes of the blood to lowly scribes, with an array of costly gifts — jewels andfinery,as well as grants of money. In emphasizing his extraordinary generosity on the occasion of his enthronement, the Persian sources leave the distinct impression that this was no mere show of imperial munificence but a political payoff on a gigantic scale for debts incurred in acquiring the khaghanate.'0 Indeed, throughout his brief and generally ineffectual reign, Giiyiig continued the practice of lavishing presents on members of the royal clan and the army, much to the detriment of the imperial coffers. As an additional gesture of goodwill, the new emperor restored to office most of the administrative personnel dismissed during Toregene's regency. Chinkhai was returned to his post as head of the Central Secretariat, together with his former associates. The administration of Turkestan was once again entrusted to Masud Beg, and his father, Mahmud Yalavach, was placed in charge of the Branch secretariat for north China, with the title of great judge (yeke jarghuchi).'*1 Abd al-Rahman, Toregene's agent in China, was stripped of his responsibilities and subsequently put to death. Although the previous regime had been reconstituted for the most part, Giiyiig was able to place a few of his own men in positions of power, principally his atabeg, or tutor, Khadakh, a Nestorian Christian of Naiman origin who, according to Carpini, held the office of "procurator of the whole empire."92 This restoration of the collegial principle in the governance of the empire, most certainly a concession wrung from the Ogodeids by other princely lines 90 Juvaynl/Qazvinl, vol. I, p. 209; and Juvaynl/Boyle, vol. 1, pp. 254—5. 91 Ch'eng Chii-fu, Hsiieb-lou chi (Hu-pei hsien cheng i shu ed.), 2 5 , p. 17b. 92 Mongol mission, pp. 66—7. For his title, atabeg, see Juvaynl/Qazvinl, vol. i , p . 2i3;andJuvaynI/Boyle, vol. 1, p . 2 5 9 . Cambridge Histories Online © Cambridge University Press, 2008 THE REIGNS OF OGODEI AND GOYtJG 387 as the price for their support of Giiyiig's candidacy, appears to have been honored in both spirit and practice by the new khaghan. When, for instance, irregularities were reported in the management of imperial land grants in Taming circuit (contemporary Hopei) during the first year of Giiyiig's reign, the incident, a matter of some significance, was investigated jointly by the Tangut Hsi-li Ch'ien-pu, a representative of the imperial court, and by Bujir (Pu-chih-erh), a long-time servitor in the Toluid household. It is interesting to note that the crucial documents in the case, the papers of the Ta-ming censor, were placed in the custody of the Toluids' agent, not the khaghan's.93 Giiyiig's princely partners took an active interest in the business of the empire. Unfortunately, further data on the administrative arrangements and measures in north China are extremely scarce for this period. Even the activities of its chief judge, Mahmud Yalavach, are unknown. From the available records, it appears that Giiyiig's major administrative undertaking was an effort to compile fresh population figures for the empire. An order for a census of the Chinese territories was issued late in 1247, and similar endeavors, though clearly limited in scope, are reported for Iran and the Russian principalities at about the same time. The results and actual compass of the enumeration in China are nowhere recorded, perhaps because they were never completed, if indeed they were even begun.94 In general, under Giiyiig's stewardship the imperial government lacked vigor and exhibited a pronounced tendency toward decentralization, if not fragmentation. Local officials, in consequence, enjoyed ample opportunity to exploit the subject population, and this in turn led to outbreaks of "banditry" and rebellion in the countryside and cities of north China.9' The erosion of central authority was due in part to the Jochids' intransigence, but the situation was aggravated by Giiyiig's own failings as a leader. His father's son, he was a dedicated drinker from an early age; chronic illness also took its toll, and by the time he ascended the throne he was in such a weakened condition that he had neither the vitality nor the inclination to take an active part in managing the empire's affairs. He developed no general plan of action around which to mobilize the energies of his people, and his policy initiatives, few in number, were never forcibly implemented. Initially, for instance, he made a great show of calling in all unauthorized imperial decrees (jarligh) and badges of authority {gerege) issued under Ogodei and Toregene, 93 YS, 122, p. 3012; and Yao Sui, Muancbi, 19, pp. l o b - n a . 94 For details and documentation on this registration, see Thomas T. Allscn, "Mongol census taking in Rus', 1245—1275," Harvard Ukrainian Studies, 5 (1981), pp. 36—38. 95 Hu Chih-yu, Tzu-shan ta ch'iian chi (SKCS ed.) 151, pp. 20b— 2 ia, describes several such outbreaks in the period betwen 1247 and 1249. Cambridge Histories Online © Cambridge University Press, 2008 388 THE RISE OF THE MONGOLIAN EMPIRE but he himself soon became equally lax in this regard, and his successor, Mongke, had to bring such abuses under control once again. What energies Giiyiig did possess he seems to have expended in partying, hunting, and keeping a wary eye on Batu. To add to the woes of the empire, Giiyiig was profligate in the extreme. His ceaseless bestowals of money and rarities on his backers assumed enormous proportions and soon alarmed his more responsible ministers. Court criticism on this score, however, carried little apparent weight with the khaghan, for by his own declaration he desired above all to surpass the record of munificence and openhandedness of his father.'6 In this he certainly made a most promising start: The Persian chronicles record that by the time of his death Giiyiig had written drafts against the imperial treasury totaling 500,000 silver ingots (Jbdlish) to pay for the luxury items that he distributed to his supporters.9' These numbers may be exaggerated, but it is clear that Giiyiig was heavily involved with and financially indebted to the ortogh merchants. The extensiveness of his dealing with the ortoghs and their consequent influence at court suggest that the abusive practices of these merchants so common during Ogodei's reign - tax farming, usury, peculation, and extortion — were again the order of the day. The death of Giiyiig and the regency of Oghul Khaimish During Giiyiig's short reign, military operations were few and restricted in scope. In 1246—7 there were limited attacks on Sung positions in Hupei and Anhwei, and about the same time Eljigidei, the new commander in west Asia, undertook some minor campaigning in Iran. The Mongols' uncharacteristic lack of aggressiveness in this period is obviously connected with the continued tensions among the princes of the blood. Batu's grudging acceptance of Giiyiig's sovereignty was, of course, at the center of their divisions, and because an open conflict between the two seemed imminent, the princes were unwilling to commit large numbers of their troops to a new round of external expansion. The empire, in brief, was edging toward civil war, and the potential antagonists were husbanding their resources for the inevitable clash of arms. Though the sources are somewhat vague, the available evidence seems to indicate that Giiyiig, with uncustomary decisiveness and resolve, made the first move to force a showdown with Batu. In the fall of 1247 the khaghan 96 Rashld/Karlml, vol. i, p. 574; and RashTd/Boyle, p. 188. 97 Juvayni/Qazvinl, vol. 3, pp. 8 3 - 5 ; and Juvaynl/Boyle, vol. 2, pp. 603—4. Cambridge Histories Online © Cambridge University Press, 2008 THE REIGNS OF OGODEI AND G O Y O G 389 left Mongolia for his appanage in the Hi River region, ostensibly on a tour of inspection. In fact, as his subsequent actions demonstrate, the true purpose of the "tour" was to position himself unobtrusively in Jungaria for a surprise strike at his rival's territories, which lay farther to the west. What event, if any, triggered his decision to move against the Jochids at this particular time is unknown. In any case, once he reached Jungaria, Giiyiig immediately set about reorganizing and expanding his force preparatory to the attack. By imperial decree the khaghan ordered that "one man from every one hundred Mongol households be enlisted as 'braves' " (pa-tu-lu; Mongolian: badur).** Because the latter were an important element in the imperial guard, functioning as the khaghan's advance guard and as elite shock troops," it is clear that Giiyiig was contemplating offensive action in the near future. His preparations completed, in the spring of 1248 the khaghan set out from his base in Jungaria with a large army, heading in a westerly direction. Batu was at this time camped at Ala Qamaq, a locale south of Lake Balkash, and was therefore in the general line of march of Giiyiig's massed troops. At this juncture, Tolui's widow, Sorghaghtani Beki, who was outwardly on good terms with the Ogodeids, secretly dispatched a messenger to alert Batu to the khaghan's movements and hostile intentions, an action obviously calculated to win Jochid support for her carefully orchestrated campaign to place her son Mongke on the imperial throne. Batu received the timely warning and readied his forces for the expected onslaught. The long-awaited confrontation was, however, averted when Giiyiig's frail constitution finally gave way and he died at Khum Sengir (Heng-hsiang-i-erh), a place one week's march from Besh Balikh, the Uighur summer capital on the northern slopes of the T'ien Shan Mountains. With his demise the campaign was called off. Giiyiig's body, on the orders of his widow, Oghul Khaimish, was taken back to his appanage in Jungaria for burial. Despite his open hostility to the Ogodeids, Batu readily conceded Oghul Khaimish's right to act as regent of the empire until a khuriltai confirmed a successor. Her powers, however, were circumscribed, for Batu stipulated that she had to retain the services of the deceased khaghan's ministers and officials and that massive purges of administrative personnel, as had occurred under Toregene, would not again be tolerated.100 Unlike the opposition, which quickly rallied around Mongke, Tolui's el98 YS, 2, p. 39. Giiyiik's plans for an attack on Batu are also noted in Yuan Chiieh, Ctiingjungchiishih chi (SPTK ed.), 34, pp. 240-253. 99 See Hsiao, The military establishment of the Yiian dynasty, p. 36, for details on the "braves." 100 Juvaynl/Qazvlnl, vol. 1, pp. 217—18; and Juvaynl/Boyle, vol. 1, p. 263. Cambridge Histories Online © Cambridge University Press, 2008 39° THE RISE OF THE MONGOLIAN EMPIRE dest son, the Ogodeids had difficulty settling on their own candidate for the vacant throne. Oghul Khaimish was forced to spend most of her time contending with the discord inside her own family and preventing the elevation of the Toluid standard bearer. Under these circumstances, Juvaynl states, very little governmental business was transacted during the interregnum except "negotiation with merchants, the provisional allocation of sums of money to every land and country and the dispatch of relays of churlish messengers and tax gatherers."101 Specifics are lacking on the categories and rates of taxation in force in the sedentary sector of the empire during the regency of Oghul Khaimish, but information on the tax (khubchir) levied on the nomads' herds is available in the "Records of the horse administration of the great Yiian."102 According to edicts preserved in the collection, the schedule for this tax, which constituted the main fiscal obligation of the nomadic population to the central government, was fixed initially at one animal per hundred in 1234. Under this system, herders with fewer than one hundred animals paid no tax at all. Unaccountably, by an edict issued in July 1250, Oghul Khaimish and her advisers raised the rate to one animal in ten, an increase of dramatic proportions. The rationale behind this measure is difficult to fathom, as it presumably had the effect of weakening rather than strengthening the support of the nomadic population — the vital military core of the empire - for the Ogodeid cause at the very height of the struggle over the throne. Their shortsightedness in this regard is perhaps symptomatic of the overall ineptitude displayed by the Ogodeids in their fruitless efforts to keep the khaghanate within their own family. THE EMPIRE AT ITS APOGEE: THE REIGN OF MONGKE, 1251-1259 Mb'ngke and his rivals The Ogodeids, divided by internal rivalry and unable to close ranks behind a single candidate of their own, shortly found their claims on the throne vigorously contested by another Chinggisid line, the Toluids. The Toluids had long been preparing for this day under the leadership of Tolui's widow, Sorghaghtani Beki (d. 1252), a woman of great force and political acumen 101 Juvaynl/Qazvini, vol. i, pp. 219; and Juvaynl/Boyle, vol. 1, pp. 264—65. The translation is taken from Boyle. 102 Ta Yuan ma chtng chi (K'uang-t'ang hsiuh chu'n ts'ung shu ed.), 1, pp. 29b—30a. For a translation of the relevant passages, see C. R. Bawden and S. Jagchid, "Some notes on the horse policy of the Yuan dynasty," Central Asiatic Journal, 10 (1965), pp. 2 5 4 - 5 . Cambridge Histories Online © Cambridge University Press, 2008 THE REIGN OF M5NGKE 391 who had presided over her family's fortunes since the death of her husband in 1232. 10 3 Sorghaghtani Beki's campaign to place her son Mongke on the throne was meticulously planned. As a first step she patiently sought to create for the Toluids an image of unstinting and disinterested service to the empire. To give moral substance to this claim, she insisted that her family remain officially neutral in the princely disputes that broke out in the 1230s and 1240s and that they accept the decisions of the khuriltais with good grace. On her advice, too, the Toluids always cooperated with the reigning khaghan and actively supported the empire's campaigns of conquest with their large armies. Though somewhat contrived, this carefully fashioned reputation for unswerving loyalty to the Chinggisid legacy would later be invoked by the Toluids as clear evidence of their qualifications for and moral right to the empire's highest office. Sorghaghtani Beki also took the lead in seeking from the collateral lines support for Mongke's candidacy. While maintaining friendly and correct relations with the Ogodeids, she and her family covertly cultivated the friendship of the Jochids, showing great concern for Batu's failing health and paying him the deference that was his due as the head of the senior Chinggisid line. Given his hostility toward the Ogodeids and his own lack of interest in the khaghanate, Batu was a natural ally of the Toluids in their bid for the throne. Sorghaghtani Beki's timely warning to Batu on the eve of Giiyiig's planned strike to the west was thus a natural by-product of the secret understanding that bound together the two families. The Jochids, for their part, openly expressed their solidarity with the Toluid cause, immediately after Giiyiig's death. Determined that no more Ogodeids would sit on the throne, Batu hastily called for a khuriltai at Ala Qamaq (where he had been awaiting Giiyiig's armies), making it clear to all that Mongke was his choice for emperor. The Ogodeids refused to attend this khuriltai, claiming that the selection of a new khaghan had to take place in the Onan-Keriilen area. In this they received the support of the sons of Chaghadai and so were able to delay the diet's opening. But beyond preventing the immediate confirmation of the Toluid candidate, there was little else on which the Ogodeids did agree. Two of Giiyiig's sons, Nakhu and Khoja, made public their claims on the throne, as did their cousin Shiremiin. All three established separate courts, each with its own group of supporters, and Oghul Khaimish was, in consequence, unable to forge a common front against the opposition. 103 For a sketch of her remarkable career, see Morris Rossabi, "Khubilai khan and the women in his family," in Studia Sino-Mongolica: Festschrift fur Herbert Franke, ed. Wolfgang Bauer (Wiesbaden, '979). PP- 158-66. Cambridge Histories Online © Cambridge University Press, 2008 392 THE RISE OF THE MONGOLIAN EMPIRE In the meantime, the supporters of Mongke, undeterrred by earlier rejection, persisted in their efforts to convince the Ogodeids to participate in the diet. Finally, after much cajoling, the agents of the Toluids managed to extract an agreement from Khoja and Nakhu at least to send representatives to the khuriltai even if they would not come in person. On this basis the Ala Qamaq diet, dominated by the lines of Tolui and Jochi, got under way in 1250. 1 O 4 Coming straight to the point, Batu opened the proceedings by proposing Mongke for the throne and urging his immediate confirmation. The khuriltai was about to do so when a messenger unexpectedly arrived from Oghul Khaimish to argue the cause of Shiremiin, who was now put forth, rather belatedly, as the Ogodeid candidate. While pleading his case, when the emissary noted that Shiremiin had been duly nominated by Ogodei as his successor, the Toluids responded by pointing out that the deceased khaghan's instructions on this score had been blatantly violated years ago by his own family and were thus no longer in force.105 Unable to counter this argument, Shiremiin's spokesman sat down. Batu and Uriyangkhadai, the son of Siibetei, next addressed the throng, ordering those assembled to acclaim Mongke as khaghan, which they did without further debate. After a proper show of reticence, Mongke assumed the throne. Sensitive, however, to Mongolian tradition and the criticisms of their opponents, the Toluids decided to convene a second khuriltai in the Onan-Keriilen area to confirm officially and enthrone formally the new emperor. During the interim, Sorghaghtani Beki and Batu again tried to persuade members of the opposing lines to relent and accept Mongke's elevation. After a year of effort a few Ogodeids and Chaghadaids came over to the Toluids; with these defectors in hand, the second, formal khuriltai was convened in the summer of 1251 at Kode'e Aral, where Chinggis khan had been enthroned in 1206. Batu himself did not attend but sent his brothers and sons with a large security force. Also present were descendants of Chinggis khan's brothers and the few Ogodeid and Chaghadaid princes. Technically, then, all branches of the imperial family were represented. To the surprise of no one, the gathering confirmed Mongke's first election without a word of dissent. A celebration followed, in which the new khaghan was especially solicitous to those who had broken with their families to attend; all were handsomely rewarded for the sacrifices that they had made on his behalf. Those who had boycotted the two khuriltais were now faced with a difficult 104 On the date of this khuriltai, variously reported as 1249 or 1230, see Paul Pelliot, Les Mongols et la papauti(Paris, 1923; repr. Peking, 1939; repr. Peking, 1959), vol. 3, pp. 199—201, n. 3. 10; YS, 3, p. 44; and Rashid al-Din,Jami'al-Tapariib, ed. A. A. Alizade (Moscow, 1968), vol. i , p t . 1, pp. 140—1 (hereafter cited as Rashld/Alizade). Cambridge Histories Online © Cambridge University Press, 2008 THE REIGN OF MONGKE 393 choice: either to acquiesce in the fait accompli or to undertake to depose the khaghan. After consultation with their families, Nakhu and Shiremiin, the belated Ogodeid candidate for the throne, opted for the latter alternative and decided on a desperate and dangerous course of action: Publicly announcing that they were coming to congratulate Mongke on his coronation and thereby to acknowledge his right to the throne, they secretly resolved to assassinate the unsuspecting khaghan and his immediate supporters. Audacious and well prepared, their plot was close to success when it was suddenly exposed by mere chance. While searching for a lost animal, a falconer in Mongke's household accidentally came across a wagon from Shiremiin's baggage train that had broken down and been left behind. Discovering that weapons had been carefully concealed in the cart, he hurried back to the imperial camp to warn his master of a probable attack. Once the disbelieving khaghan was convinced that this intelligence was true, he sent Menggeser, the head of his personal guard, to deal with the threat. The conspirators, unaware that their plans had been revealed, were easily disarmed and taken into custody.106 Disunity and ineptitude had cost the Ogodeids the throne, and their ill-starred attempt to recoup their losses would soon cost many of them their lives. Purge, consolidation, and legitimation Fearful of further plots against his life and thirsting for revenge, Mongke initiated a ruthless and wide-ranging purge in the wake of the princes' plot. Troops of the Jochids and Toluids were formed into a huge jerge (Mongolian for a line of soldiers drawn up in a semicircle) and ordered to sweep Mongolia, Jungaria, and Turkestan for accomplices of the accused princes. The latter, already in custody, were tried by the khaghan himself. Khoja, Nakhu, Shiremiin, and others later implicated in the conspiracy were found guilty, and all, after an initial period of exile or confinement to a military camp, were put to death. Oghul Khaimish and Shiremiin's mother, Khadakhach, were also tried. Taken to Sorghaghtani Beki's camp, they were accused of treason and black magic. After enduring much humiliation and physical abuse, they were either executed or, as some sources have it, allowed to commit suicide. The ministers of Ogodei and Giiyiig, Chinkhai, Khadakh, and their associates, were brought before Mongke's chief judge, Menggeser. Convicted of inciting Nakhu and Shiremiin to rebellion, most paid for their crimes with 106 This episode is recounted in a variety of independent sources, for example, YS, p. 1351; Mongol mission, pp. 147—8; Juvayni/Qarvini, vol. 3, pp. 39—47; and Juvaynl/Boyle, vol. 2, pp. 574—9. Cambridge Histories Online © Cambridge University Press, 2008 394 T H E R'SE OF THE MONGOLIAN EMPIRE their lives. As long-time servitors of the rival Ogodeid line they could not simply switch allegiance and join the new regime: A Mongolian prince and his retainers rose together and fell together. Smaller fry were tried by special traveling tribunals (jargbus) that Mongke dispatched throughout the empire. In Harat, Afghanistan, for instance, a pro-Chaghadaid scribe was identified by Mongke's agents and, following extensive interrogation to determine the extent of anti-Toluid feeling among the Mongolian-appointed officials in the area, was put to death. '°7 Even in far-off Iraq the tribunals managed to ferret out and punish dissidents and rebels. The blood purge was by no means limited to the imperial family and government officials: It touched as well the heads of all dependent states, each of whom was required to travel to Khara Khorum for a personal interview with the new khaghan. Those who had remained neutral or had sided with the Toluids retained their positions, whereas those who had exhibited Ogodeid sympathies were quickly eliminated. Salindi, the Uighur iduq-qut, was discovered to have been in league with Oghul Khamish and, on Mongke's command, was beheaded. The iduq-qut's brother and successor, Ogriinch, who had demonstrated his fidelity to the Toluid cause, served as Salindi's executioner. The same fate befell the ruler of Klrman, a satellite state in Iran. The total number of victims is not known, but it was unquestionably large: Menggeser claimed that he had himself tried and executed seventyseven ministers and officials, and from the testimony of contemporary witnesses, including Mongke himself, it is certain that the ranks of the Ogodeid and Chaghadaid princes were thinned considerably.Io8 Apparently, few opponents slipped through Mongke's nets, for there were only isolated instances of princely insubordination throughout the remainder of his reign. The victory of the Toluids led, naturally, to major adjustments in the relations among the imperial princes. First, Mongke in essence created two new regional khanates in 1251 when he placed his younger brothers Khubilai and Hiilegii in charge, respectively, of China and west Asia. By this move he solidified Toluid control over the most economically productive regions of the empire and thereby strengthened his hand in dealing with the other princely lines. Those among the Ogodeids who had come to terms with the khaghan were given widely scattered territories in Inner Asia. Khaidu (Hai-tu), a grandson 107 Sayf ibn Muhammad, Ta'rikh-i nimah-i Harat, ed. Muhammad Zubayr al-Siddiql (Calcutta, 1944), pp. 172-81. 108 Mongol mission, p. 203; and Gandzaketsi Kirakos, Istoriia Armenii, trans. L. A. Khanlarian (Moscow, 1976), p. 236. Cambridge Histories Online © Cambridge University Press, 2008 THE REIGN OF MijNGKE 395 of Ogodei, was granted the city of Qayaliq south of Lake Balkash, and his brother Melik (Mieh-li) was given land along the Irtysh. The "tame" Ogodeids, in consequence, had minimal resources at their command and were carefully separated from one another. It would be nearly two decades before the fortunes of the family were revived by Khaidu. In regard to the Chaghadaids, the reigning, pro-Ogodeid khan, Yesii Mongke, was deposed and executed and replaced by his nephew Khara Hiilegii, who had held the office previously from 1242 to 1246. Dismissed by Giiyvig in favor of Yesii Mongke, the embittered Khara Hiilegii had become an early adherent of the Toluid cause. A sickly man, he died before reaching his ordo (camp) in Turkestan and, by imperial decree, was succeeded by his young son Mubarak Shah, with his widow, Orghina, acting as regent. Thus the two rival lines were temporarily reduced to impotence and subordinated to the will of the khaghan. Of necessity, Mongke's relations with Batu were conducted on a much different basis. The Jochid was, genealogically speaking, senior to Mongke and had been his chief supporter in the struggle for the throne. The khaghan therefore accorded Batu special esteem and was properly thankful for his aid. But this does not mean, as is sometimes assumed, that Batu was an all-powerful kingmaker who obtained, in return for securing the throne for the Toluids, the western half of the empire for himself.109 The Toluids, in truth, came to power largely through their own efforts. Sorghaghtani Beki was the real kingmaker, and Toluid control of the bulk of Chinggis khan's army tipped the balance of power in their favor. In fact, Mongke never conceded any of his imperial prerogatives to Batu in diplomatic or military matters. Foreign emissaries entering Jochid territory were always sent on to Khara Khorum to negotiate directly with the khaghan. And the emperor could and did commit sizable contingents of Jochid troops to campaigns that benefited other princely lines, as, for example, in the case of Hiilegii's attack on Baghdad in 1257—8. Internally, the affairs of the Golden Horde were governed by the two in partnership: All major administrative operations - census taking, revenue collection, and so forth — were jointly undertaken by agents of the khaghan and the Jochids. Theoretically, the principle of joint administration was operative in the other regional khanates as well, but in practice the partnership was equal only within the confines of the Golden Horde. Elsewhere in the empire — Central Asia, China, and Iran - the emperor clearly held a dominant position vis-avis the regional khan. At the same time that they were busily consolidating their power, the 109 This is the argument of Wilhelm Barthold in litoriia Turktstana, repr. in his Sochinmiia (Moscow, 1963), vol. 2, pt. 1, p. 148. Cambridge Histories Online © Cambridge University Press, 2008 396 THE RISE OF THE MONGOLIAN EMPIRE Toluids devoted considerable attention to the problem of legitimation. The succession had obviously been disputed, and Mongke felt compelled to justify his elevation to the Mongolian people at large. As the centerpiece of the argument the Toluids stressed the "constitutionality" of the two khuriltais that had confirmed Mongke in office. They pointed out that all branches of the royal house had been duly represented and that thus the convocation represented a true consensus. Conversely, they emphasized the Ogodeids' treacherous betrayal of the decisions of these properly constituted khuriltais. To get this message across, the purge trials were used as forums; properly encouraged, the Ogodeid princes fell all over themselves at court confessing their sins and acknowledging the legality of the two diets and therefore Mongke's right to the crown. To buttress their claim on the throne, the Toluids portrayed Mongke as the living embodiment of the Chinggisid tradition, a man who was uniquely qualified for the office by virtue of his rigid adherence to the letter and spirit of the jasagh. Mongke's own proud assertion that he "followed the laws of [his] ancestors {and} did not imitate the ways of other countries"110 neatly summarizes this side of the Toluid case. In a further effort to wrap themselves in the mantle of the founding father, in 1252 the Toluids began official worship of Chinggis khan. At the same time Tolui, who was buried next to his father, was retroactively made a khaghan and also became the object of officially sponsored veneration. To counter the Toluid claims, the Ogodeids simply argued that Chinggis khan had intended, by nominating Ogb'dei as his successor, to reserve the khaghante for his third son's family, and consequently, Mongke, whatever his qualifications or endorsements, could not be considered a legitimate emperor. Their premise, of course, was disputed by the Toluids, who attributed statements to Chinggis khan and Ogodei that contradicted this claim. Whether or not this or any other Toluid assertions were true is beside the point; what is important is that they established a credible case in favor of their legitimate succession, which they broadcast widely and effectively. Administrative apparatus Long before his accession Mongke had formed a shadow government in anticipation of the day when imperial power would pass into Toluid hands." 1 The new Central Secretariat that replaced the regime of Chinkhai and Khadakh was headed by Menggeser, the commander of the imperial guard. A Mongol of the Jalayir tribe, Menggeser, who held the title oiyeke jarghuchi, n o YS, 3, p. 54. i n This section is based on my article "Guard and government in the reign of the grand qan Mongke," Harvard Journal of Asiatic Studio, 46 (1986), pp. 500-21. Cambridge Histories Online © Cambridge University Press, 2008 THE REIGN OF MONGKE 397 was responsible for the day-to-day operations of the central government and also looked after the management of princely appanages. Next in line was the Kereyid Bulghai, a Nestorian Christian. He was in charge of scribes and chamberlains, scheduled audiences, and recorded memorials. When Menggeser died in 1253, Bulghai assumed his deceased colleague's duties, without, however, relinquishing any of his original functions. Under both administrations, the Central Secretariat, called Chung-shu sheng in Chinese, was subdivided into various departments, each with its own chief. There were departments for sacrifices and shamans, for control of the ortogh merchants, for the postal relay stations (jam), and for the imperial treasuries and arsenals. In these upper echelons of the administration the responsible officials were almost entirely Mongols. On the clerical level, however, the majority were non-Mongolian, as each senior minister needed a host of scribes fluent in the principal languages of the empire, that is, Chinese, Tibetan, Uighur, Tangut, Persian, and Mongolian, to assist him in his work. This clerical staff was so large and the record keeping so extensive that a third of Khara Khorum was set aside for their exclusive use. They lived and worked in a number of large tents and always accompanied the khaghan on his travels. Following a well-established practice, Mongke recruited his central government personnel from his own guard/household establishment, the core of which had come to him as an inheritance from his father. The careers of Menggeser and Bulghai illustrate this pattern of recruitment. The forebears of both men had first entered service as low-ranking members of Chinggis khan's guard. When Chinggis khan died, their families were inherited by Tolui, under whom Menggeser and Bulghai commenced active careers as officers in the guard. Both then served Sorghaghtani Beki after her husband's demise and finally rose to prominence in Mongke's household establishment in the years before his accession. This pattern of recruitment and service holds true as well for low-level officials. All of the Chinese scribes in the Central Secretariat whose careers can be traced also started out as bichechi in either Tolui's or Mongke's kesig. In short, the competence and loyalty of all of Mongke's officials had been tested and confirmed by years of service in the Toluid household before the reins of government were placed in their hands. Because of this recruitment practice, it is difficult or perhaps meaningless to try to make any clear-cut institutional distinction among guard, household, and central government in the early Mongolian empire. For all intents and purposes, they were a single body, one that functioned simultaneously as a security force, elite military unit, court, and imperial government. Below the Central Secretariat Mongke retained the system of regional administrations first instituted by Ogodei. Called appropriately "Moving" Cambridge Histories Online © Cambridge University Press, 2008 398 THE RISE OF THE MONGOLIAN EMPIRE Secretariats (Ch: Hsing shang-shu sheng), they were in effect mobile branches of the Central Secretariat. The men that Mongke selected to head these secretariats — Mahmud Yalavach in China, Mas'iid Beg in Turkestan, and the Oyirad Mongol Arghun Akha in Iran - all were holdovers from Giiyiig's reign. All three survived the change in regime because they were experienced administrators and, in theory at least, were the representatives of the entire Chinggisid line in the sedentary sector of the empire. Mongke respected the principle of collegiality at the regional level, allowing interested Mongolian princes to name representatives (called nok'dr in the Persian sources) to serve on the staffs of the Branch Secretariats. In China, for instance, one of Mahmud Yalavach's principal assistants was Chao Pi, a Chinese associate of Khubilai. The Khwarazmian's other assistant, we should note, was Bujir, a long-standing member of the khaghan's own household. The same situation obtained in Iran, where Batu, Hiilegii, Sorghaghtani Beki, and Mongke all had their designated n'okb'd on Arghun Akha's staff. On the local level, native elites, for example, the nobility in the Russian principalities or the monastic organizations in Tibet, were normally left in place to carry out governmental functions under the watchful eye of darughachi (usually termed basqaqs in the western territories), who were appointed by the khaghan from among his retainers. The system of governance was thus essentially the same as had evolved under Mongke's predecessors. Mongke in fact was no innovator: He simply used the institutions bequeathed to him. He was, however, a forceful and tireless administrator who successfully dominated the governmental apparatus. He was able, of course, to exercise immediate and firm control over his Central Secretariat, as he and his staff lived, worked, ate, and drank together. And under his direct supervision the Central Secretariat rigorously monitored the work of its regional branches from above, continuously auditing their account books and interviewing their officials. In addition, Mongke's riokod attached directly to the Branch Secretariats monitored these bodies from within, and his darughachi carefully scrutinized their activities from below, reporting back any irregularities to the throne. Extremely well informed about, and deeply involved in, all phases of the administration of his realm, from planning through implementation, Mongke was in most instances able to push his policies through against the opposition of princely or other parochial interests. Imperial design Unlike his predecessor Giiyiig, Mongke came to the throne with a wellconceived program of action in mind. His grand design for the empire, Cambridge Histories Online © Cambridge University Press, 2008 THE REIGN OF MONGKE 399 which envisioned both internal reform and external expansion, fully engaged, as intended, the attention and energy of the Mongolian people, nobility and commonality alike, and contributed significantly to the success of his reign. 112 The khaghan first announced his reform measures shortly after the khuriltai of 1251. Their purpose was to ensure the central government free access to the resources of the empire and to curtail the irregular exactions of princes and officials.113 As a first step, Mbngke rescinded all the tablets of authority, seals, and jarlighs issued since Chinggis khan's day. This was designed to deprive his competitors for the resources of the empire of the means of continuing their unwarranted appropriation of goods and services. He also severely restricted their use of the jam, the postal relay system. Princes, officials, religious dignitaries, and ortogh merchants had become accustomed to using the jam for their personal needs, thus clogging the empire's lines of communication and adding to the already heavy burden on those households of the populace assigned to maintain the relay stations. Henceforth only officially authorized personnel would be allowed access to this system. The powers of the appanage princes also came under close scrutiny; they would no longer be permitted to summon or tax the people in their personal domains without first consulting the imperial court. Last and most important, Mongke instituted a unified and regular system of assessing and collecting taxes throughout the entire realm. And as a corollary measure to ensure maximum tax yields, the khaghan decreed that territories already under Mongolian dominion be economically rehabilitated and that in active theaters of operations, destruction and killing be kept at a minimum; abandoned agricultural lands and devastated cities contributed nothing of value to the imperial treasuries. The effectiveness of this series of reforms is difficult to measure. Certainly, destructive practices continued, as witnessed by the sack of Baghdad in 1258. But even though old habits died hard, it it equally true that Mongke was seriously endeavoring to stem unnecessary damage to the sedentary sector. Mongolian officers who grossly violated his instructions were, in fact, severely punished. On one occasion in 1258 the emperor had the attendants of his son Asudai (A-su-tai) flogged for interfering with agricultural work in China. At the same time he meted out the death penalty to other officers who had seized vegetables from the gardens of Chinese peasants. In another case, a 112 Rubruck, who visited Mongke in 1254, notes the Mongols' pride in their accomplishments and their expectation that soon everyone in the world would surrender to them. See Mongol mission, pp. 149—30. 113 Consistent accounts of these measures are available in YS, p. 45; Juvaynl/Qazvlnl, vol. 3, pp. 75—8; and Juvaynl/Boyle, vol. 2, pp. 598—9. Cambridge Histories Online © Cambridge University Press, 2008 4OO THE RISE OF THE MONGOLIAN EMPIRE Mongolian tiimen commander accused of murdering a Persian civilian was put to death before the gates of Tus, a city in Khurasan, where the crime had been committed. "•» The program of revitalization also produced some practical results. The economic life of Turkestan's urban centers, severely damaged in the period 1219—23, was restored to preconquest levels, and certain areas in China, such as Hsing-chou in Hopei, were also rebuilt. In regard to the rest of his reforms, Mongke brought the ortogh merchants under control, drastically reduced court expenditures, and successfully implemented and enforced his new tax measures. Data on conditions within the appanages are limited, but the khaghan intervened in these domains whenever he felt that imperial interests were at stake (see the section "Khubilai and north China" for details). In alleviating and equalizing the burdens of his subjects, Mongke was not motivated by ethical considerations, but by the desire to gain unfettered access to and control of the resources necessary to fuel his plans for expansion. He intended to launch simultaneously major invasions of Sung China, Korea, and west Asia, and so anyone, from predatory prince to peasant tax evader, who prevented or impeded the flow of resources earned Mongke's wrath and enmity. The subject populace was still to be squeezed, but on a regular and systematic basis and only under the supervision of imperial agents. By combining internal reform with a new round of expansion, Mongke accomplished two major objectives: First, he kept the divided Mongolian ruling elite busy at all times preparing for and executing military operations, and second, he was able to concentrate authority at the center and thereby consolidate his hold on the throne, in the name of procuring resources for military campaigns. Under these circumstances, it was difficult for disgruntled princes to challenge Mongke's authority at a time when the empire was engaged in carrying out Chinggis khan's mandate to his people to conquer the world. Most Mongols, whatever their other differences, generally agreed that the principal business of the Mongolian empire was to conquer and subdue. However self-serving the policies may have been, it should not be assumed that Mongke was merely manipulating the words and memory of Chinggis khan for his personal benefit. He seems to have sincerely believed that the Mongols had a great mission — universal dominion — and that it was incumbent on him to make his grandfather's mandate a reality. 114 YS, p. 51; and RashidMlizade, p. 154. Cambridge Histories Online © Cambridge University Press, 2008 THE REIGN OF MONGKE 4OI Mobilization In order to secure the resources needed for further campaigns of conquest, Mongke called for a new and definitive census of the entire realm. Designed to identify and mobilize the empire's wealth and talent for war, this census tabulated animals, orchards, and sources of raw material (e.g., iron and salt deposits), as well as people. In both extent and thoroughness, Mongke's inventory far surpassed the earlier efforts of Ogodei and Giiyiig. 11 ' The enumeration commenced in China in 1252 under the general direction of Bujir, Mongke's agent on Mahmud Yalavach's staff. Supplemental censuses were taken in 1255, 1257, and 1258 to uncover the displaced population and to register the inhabitants of newly surrendered areas. In west Asia the census was conducted by Arghun Akha. In close association with the khaghan's personal representatives, he surveyed Afghanistan, Iran, Iraq, and the Transcaucasus between 1253 and 1258. The registration in the Golden Horde was first ordered in 1254. Actual counting began in the north Caucasus in 1254 and culminated in Novgorod, the northernmost of the Russian principalities, in 1259. In the latter case, as in all others, the census teams were composed of representatives of the khaghan, the regional khan, and other interested members of the imperial family. The data accumulated were entered into registers that, when completed, were sent directly to Mongke. With up-to-date census figures at their disposal, the Central Secretariat could make a reasonably accurate estimate of the quantities of taxes and numbers of military recruits that they could expect from a given province or dependent state, thus making it difficult for local authorities to withhold resources from the center. The tax obligations of the populace were individually fixed at the time of the registration according to a system pioneered by Mahmud Yalavach in Turkestan during the 1230s. Under his scheme there were three basic categories of taxes: a head tax (khubchir) in coin imposed on all adult males, an agricultural tax (khalan) paid in kind by the rural population, and a commercial tax {tamgha) on all business transactions collected in cash in urban markets and at numerous customs stations." 6 The khubchir, as formulated by Muslim chroniclers of the period, was assessed everywhere in the empire in accordance with an individual's ability to pay, the original rate per annum varying between one dinar for the poor to eleven dinars for the wealthy. In practice, this formula was applied only in the 115 See Allsen, "Mongol census taking in Rus'," pp. 3 8 - 5 2 . 116 On the evolution of the categories of taxation, see John M. Smith, "Mongol and nomadic taxation," Harvard Journal of Asiatic Studio, 30(1970), pp. 4 6 - 8 6 . Cambridge Histories Online © Cambridge University Press, 2008 402 THE RISE OF THE MONGOLIAN EMPIRE Islamic portions of the empire, where the system had originally been devised and tested. In China the principle of graduated taxes was retained, but the khubchir (Chinese: k'o-ch'ai or ch'ai-fa) was modified in several important respects to conform to Chinese customs. It was collected there on the basis of households, not individuals, and payment was accepted in precious goods, especially silk floss, as well as in silver. Further, the original maximum rate of six Hang of silver and one-half chin of silkflossper year per household was soon reduced. Thus, with the exception of the increased importance of the silver component, the k'o-ch'ai under Mongke was much the same as it had been following Yeh-lii Ch'u-ts'ai's revenue reforms of 1236. To facilitate the collection of this tax, Mongke made a concerted effort to place more currency into circulation throughout the empire, at least in those areas that had prior experience with a monetary economy."7 In China a Superintendency of Paper Money (Chiao-ch'ao t'i chii-ssu) was established by imperial order in 1253. There is no doubt that the superintendency actually issued bills, for Rubruck examined specimens in Khara Khorum in 1254, but the extent to which this currency circulated and its role in the payment of k'o-ch'ai are not indicated in the Chinese sources. However, as Rubruck expressly states elsewhere in his work that the Chinese daily paid their Mongolian overlords fifteen hundred iascot (Turkic, yastuq, "pillow" or ingot) of silver in addition to an unspecified amount of silk," 8 it seems permissible to infer that paper money was not widely used for tax purposes, at least in the early years of Mongke's reign. In Turkestan, Iran, and the Transcaucasus there is sure evidence that by the mid-12 50s local mints were producing large quantities of dinars and that the khubchir was in fact collected in this coin. Within the confines of the Golden Horde, only the Volga Bulgars paid tribute in coin; in the Russian principalities, which had no tradition of minting coin, the Mongols were content to collect the khubchir (dan' in Russian) in furs or other high-value commodities. Among the nomadic population, the khubchir was still calculated according to the size of herds. Under Mongke, the excessive one-in-ten rate inaugurated by Oghul Khaimish was quickly abrogated, and the former rate of one per hundred head was restored. As for agricultural taxes, the Mongols for the most part followed local usage. In China they retained the system elaborated by Yeh-lii Ch'u-ts'ai, in which the tax was assessed on the basis of the quantity and quality of land. And in west Asia the customary Islamic tithe prevailed. In summary, then, Mongke established three categories of taxes to be 117 Bart hold, Istoriia Turkatana, p. 149, was the first to point out Mongke's efforts to monetize the taxes. 118 Mongol mission, pp. 144, 169—70. Cambridge Histories Online © Cambridge University Press, 2008 THE REIGN OF MONGKE 403 applied in all the sedentary regions of the empire. The basis of assessment and rates were clearly stated for each category, and in the case of the tribute and agricultural levies it was stipulated that they were to be collected only once a year. Although inequities continued and extraordinary levies were never completely eliminated, a sustained effort was made to adjust the tax schedules to local traditions and economic realities. But even allowing for such accommodations, which were normally introduced at the regional level, a fairly uniform and regular system was imposed on the empire's extremely heterogeneous population. The recruitment of manpower for the Mongols' war effort also was intimately tied to the census. Following the registration, the population of north China was divided into three basic categories: military households, civilian households, and appanage households. Those in the military category were organized into tiimen (Chinese: wan-hu, literally "ten thousand families"), groups that in theory could field ten thousand infantry to fight as auxiliaries alongside Mongolian regulars. Although the total number of new troops mobilized during Mongke's reign is unknown, it must have been considerable, judging from the figures reported for various localities: Twenty thousand troops were raised in Shantung in 1255, and Georgia yielded ten tiimens by 1257. The Mongols' field armies at this time were thus both large and international in character. The armies that invaded the Sung between 1253 and 1259 contained, besides Mongolian and Turkic contingents, Chinese, Tanguts, Koreans, Uighurs, and Alans, the last a north Caucasian people impressed into service following the census of 1254. In addition to regular military recruitment, the census was used to identify skilled craftsmen, who were then put to work in imperial arsenals, mines, and the like or employed as specialist troops in the army. Some idea of the effectiveness and flexibility of the Mongols' manpower mobilization system can be gleaned from the fact that a unit consisting of one thousand "crews" of Chinese artillerymen accompanied Hiilegii to the west and helped breach the walls of Baghdad in 12 5 8. Mongke was able to supply his brother with these troops because all the metal workers, carpenters, and gunpowder makers in north China had been registered as catapult operators (p'ao-shou) during the census of 1252 and placed on separate rolls. When it was subsequently determined that such specialists would be needed in west Asia, Mongke and his advisers had only to consult the proper rolls to find the desired number of qualified personnel. New conquests At the khuriltai of 12 51, agreement had been reached and planning begun for a series of campaigns in west Asia, Korea, and southwest China. Determined Cambridge Histories Online © Cambridge University Press, 2008 404 THE RISE OF THE MONGOLIAN EMPIRE to fulfill Chinggis khan's mandate and confident that his mobilization measures would produce the necessary troops and materiel, Mongke moved swiftly to implement the new program of conquest. By the following year, preparations were well enough advanced to start all three operations at once. Contingents assigned to the west Asian campaign were assembled in western Mongolia. In the summer of 1252 the advance guard departed, and in the following year, the main force. Though nominally under the control of the emperor's brother Hulegii, the conduct of field operations once the armies reached their destination had been entrusted to Ked Bukha, the commander of the vanguard and a steward in Mongke's household.11' Because of the great distances involved and the large quantities of men, animals, and siege equipment in his train, the logistical problems were immense, and Hiilegii's progress to the west was necessarily measured. The bulk of his forces reached Khurasan in 1256, where they were reinforced by contingents from the Golden Horde and the Chaghadai khanate and by numerous local auxiliaries - Persians, Armenians, and so on — mobilized on the basis of the recent census. The formidable weight of these combined forces fell first on the Ismailis (known to the Europeans as the Assassins), an Islamic religious order famed and feared for their use of assassination as a political weapon. Ensconced in a network of castles in the mountains of northwestern Iran, the Assassins initially offered fierce resistance, but after months of continuous battering they began to give way; by early 1257 most of their strongholds had capitulated and their leader was in Mongolian hands. Mongke, cognizant of their special skills and wary of attempts on his life, ordered all surviving Ismailis, starting with the grand master of the order, to be put to the sword. Hiilegii next moved his massed armies toward Baghdad, the seat of the 'Abbasid Caliphate. Mustasim, the caliph, refused Hiilegu's order of submission and thereby condemned to death himself and most of the citizens of Baghdad. The Mongolian columns converged on the city in late 1257 and established a close siege. Baghdad fell to Ked Bukha's assault forces in February 1258 and was then given over to pillage and massacre for seven blood-filled days. In the meantime, several subsidiary campaigns were under way: The secondary cities of Iraq and western Iran were under attack, and sizable expeditions had been sent into Kashmir and Slstan. With the completion of these operations in 1259, Hiilegii had fulfilled Mongke's commission to secure a firm Mongolian (and Toluid) hold on west Asia. Hiilegii's efforts to extend his 119 This campaign is treated in detail by John A. Boyle in "Dynastic and political history of the IIkhans," in The Saljuq and Mongol periods, vol. 5 o{ Tht Cambridge history of Iran, ed. John A. Boyle (Cambridge, 1968), pp. 340-52. Cambridge Histories Online © Cambridge University Press, 2008 THE REIGN OF MONGKE 405 newly created realm still farther in the direction of Syria resulted, however, in a major disaster when in September of the following year, the Egyptian Mamluks crushed the invading armies at 'Ain Jalut and made the Mongolian commander, Ked Bukha, their prisoner. I2° At the other end of Asia, the campaign against the Koreans commenced in late 1252 under the direction of Prince Yekii (Yeh-ku), the eldest son of Jochi Khasar, Chinggis khan's brother. Mongke's choice of commanders in this instance was a poor one: Unreliable and insubordinate, the prince badly mismanaged matters and so in 1253, on the khaghan's orders, was replaced by Jalayir (Cha-la-i-erh), an officer in the imperial guard. The latter made some progress in the next five years but was unable to overcome the Koreans' spirited resistance, forcing Mongke to send large reinforcements to the peninsula in 1258. Even with fresh units in the field, it took another year of strenuous campaigning before the Koreans reluctantly acknowledged Mongolian overlordship. On the southern Chinese front, large Mongolian forces were also in motion. In order to avoid a costly frontal assault on the Sung, which would have required a risky forced crossing of the lower Yangtze, Mongke decided to establish a base of operations in southwestern China, from which a flank attack could be staged. To this end, in the late summer of 1252 he ordered his brother Khubilai to invade and occupy the Nan-chao, or more properly the Ta-li, kingdom ruled by the Tuan family in Yunnan, which adjoined the Sung's more weakly defended western and southwestern frontiers.121 Starting from Shensi, the Mongolian armies reached the T'ao River, an affluent of the upper Yellow River, in the fall. The advance force under the Chinese general Wang Te-ch'en then penetrated the Szechwan basin, defeated the local Sung garrisons, and established a major Mongolian base in the city of Li-chou (later called Pao-ning). The path to the south now cleared and his lines of communication secure, Khubilai, with the main force, advanced on the Ta-li kingdom. In the fall of 1253, after traversing wild and mountainous terrain, Khubilai set up headquarters on the Chin-sha River in western Yunnan. Here he divided his troops into three columns and marched on Ta-li, the capital of the kingdom (see Map 29). Between December 1253 and January 1254 the kingdom was subdued, and even though its ruler had rejected Khubilai's submission order, the capital and its inhabitants were spared. As they had done on many other 120 John M. Smith, " 'Ain Jalut: MamlGk success or Mongol failure?" Harvard Journal of Asiatic Studies, 44 (1984), pp. 3 ° 7 - 4 5 121 The fullest account of this campaign in the Chinese sources is found in the biography of Uriyangkhadai, the field commander in charge of the opeation. See YS, 121, pp. 2 9 7 9 - 8 1 . See also Otto Franke, Gtschichtt dcs chinaischtn Riichts (Berlin, 1948), vol. 4, pp. 316—19. Cambridge Histories Online © Cambridge University Press, 2008 406 THE RISE OF THE MONGOLIAN EMPIRE Eastern Army Wang te-ch'en Central Army Khubilai Western Army Uriyangkhadai Uriyangkhadais subjugation of Ta-li Boundaries of Ta-li kingdom follow Tan Ch'i-hsiang Chung-kuq li-shih li-fuchi Vol6.p,75-6, Both the western and southern boundaries are probably far beyond the area actually controlled by Ta-li. Boundaries of Fu and Chun in Ta-li MAP 29. The conquest of Ta-li Cambridge Histories Online © Cambridge University Press, 2008 THE REIGN OF MONGKE 407 occasions, the Mongols left the native dynasty in place under the supervision of Mongolian officials. The king, Tuan Hsing-chih, was later sent to Mongke and invested with the title of Mo-ho-lo-ts'o, or Maharajah (literally, "great king"). At the end of 1254 Khubilai returned to Mongolia to consult with his brother the khaghan. Siibetei's son Uriyangkhadai, whom Mongke had assigned to Khubilai as expedition commander, remained in the southwest conducting campaigns against the local Man and Lolo tribes. His efforts at pacification proved successful, and following a clash with Sung troops he returned north to Kansu in early 1257. From here he sent emissaries to Mongke's court informing his sovereign that Yunnan was now firmly under Mongolian control. Pleased, the emperor honored and generously rewarded Uriyangkhadai for his fine achievement. Uriyangkhadai subsequently returned to Yunnan and began preparing for the first Mongolian incursions into Southeast Asia. In late 1257 he entered Annam, forcing the ruling Tran dynasty to seek safety on an offshore island. In the spring of the next year the Annamese king, recognizing the futility of further resistance, recognized Mongolian suzerainty by sending his son as a hostage to the imperial court. The stage was now set for Mongke's own appearance on the scene and the commencement of the long-planned conquest of the Southern Sung. Khubilai and north China The administration of north China during Mongke's reign was conditioned to a considerable extent by his relationship with his younger brother Khubilai, the basic characteristics of which are best revealed by comparing the latter's position in China with Hiilegii's somewhat analogous status in west Asia. The similarity of their situations came about through the conscious design of the emperor: Khubilai and Hiilegii were granted their princely jurisdictions at the same time, in the fall of 1251, and as Rashid al-DIn notes, the khaghan specifically and pointedly paired together the two princes on this occasion as his right and left wings. 122 Their commissions, too, were identical. Hiilegii was to extend and consolidate Mongolian-Toluid power in west Asia, and Khubilai was to do the same in east Asia. To carry out their tasks, both were given Mongolian field armies commanded by officers, Ked Bukha and Uriyangkhadai, whom Mongke had selected from his personal guard. For administrative purposes, each prince was allowed to build up a cadre of local 122 Rashld/Kariml, vol. 2, p. 685. Cambridge Histories Online © Cambridge University Press, 2008 408 THE RISE OF THE MONGOLIAN EMPIRE specialists, mainly Central Asian Muslims in Hiilegii's case and Chinese in Khubilai's, to help them rule their domains. In both instances, however, Mongke reserved the right to appoint and discharge key officials within his brothers' jurisdictions. Both, in other words, were il-khans, subordinate khans, whose rights to their territories were entirely contingent on the emperor's will. And neither, because their lands had been granted to them by their elder brother in a secondary dispensation, had the stature of the other regional khans, such as the Jochids and Chaghadaids, whose lands and titles had been allotted to them in Chinggis khan's original dispensation. Although their status was essentially the same and their functions likewise identical, the two brothers' personal relationships with Mongke nonetheless differed in several important respects. Hiilegii, for his part, was content with his station. He advertised his subordination to the emperor by formally adopting the title tl-khan and, as far as the records show, had no serious disagreements with Mongke. Khubilai, on the other hand, was much more restive and willing to test the khaghan's authority. Harboring ambitions of his own and closer to the seat of imperial power, and therefore to the scrutiny of the central government, Khubilai frequently clashed with the khaghan and his agents. In 1252, for example, he had a dispute with Bujir, whom he found too willing to impose the death penalty on Chinese subjects for minor breaches of the law. And prompted, no doubt, by his Confucian advisers, who were trying to prevent the introduction of Muslim financial practices into their homeland, Khubilai in the same year expressed to the throne his displeasure over the reappointment of the Khwarazmian Mahmud Yalavach as head of the Branch Secretariat of north China. Subsequent conflict over administrative procedures between Mahmud and Chao Pi, Khubilai's representative on his staff, merely reflected the ongoing tension between the khaghan and his younger brother. Their most serious confrontation, however, erupted in 1257 over Khubilai's administration of north China. Beginning in 1251, the year he obtained princely jurisdiction over the region, Khubilai launched a series of reforms in Honan, Ching-chao (in Shensi), and Hsing-chou (in modern Hopei) designed to reinstitute Chinese-style government in these areas and to restore their economies. In 1252, when Mongke announced his decision to distribute new appanages to the imperial family, Khubilai, at the urging of one of his most trusted Confucian counselors, Yao Shu, requested and received the strategic and fertile Wei River valley as his personal share (kbubi). Two years later, following the Yunnan campaign and his meeting with the khaghan, Khubilai went to his appanage and, with the aid of his advisers, continued his efforts at administrative reform and economic revitalization. His work in Cambridge Histories Online © Cambridge University Press, 2008 THE REIGN OF MONGKE 409 Ching-chao, as well as in Honan and Hsing-chou, met with considerable success and progressed undisturbed until 1257. In this year a major investigation of Khubilai's activities was suddenly and unexpectedly launched by Mongke, who correctly suspected that his brother was endeavoring to build up in Ching-chao and Honan an independent power base from which he might eventually challenge the throne. The inquiry, initiated ostensibly by complaints of corruption in Shensi, was headed by 'Alamdar (A-lan-ta-erh), the assistant garrison commander of Khara Khorum. His investigating team, moving swiftly through the provinces of Honan and Shensi, seized a multitude of administrative records and subjected local officials to harsh interrogation. On the basis of their findings and with, of course, the knowledge and approval of the emperor, the board of inquiry dismissed Khubilai's chief pacification officer (hsuan-fu shih) in Shensi and summarily executed some of his underlings. As a further display of his imperial power, Mongke ordered the imposition of extraordinary levies on the province, which, after some minor adjustments in the methods of payment and delivery, were exacted in full. Although outraged by this turn of events, Khubilai was in no position to defy the emperor at this time. On the advice of Yao Shu he exercised his only viable option and traveled to Mongolia in early 1258 to reaffirm his loyalty and beg his older brother's indulgence. The latter, placated by this show of submissiveness, drastically curtailed Khubilai's administrative authority in north China but took no further punitive action against him. His preemptive purge a complete success, Mongke now considered the matter closed, and as an indication that his trust in Khubilai had been restored, the emperor invited his contrite brother to help plan and execute the second phase of his assault on the Sung.12* The assault on the Sung and the death of Mongke Once the success of Uriyangkhadai's campaign in Yunnan seemed assured, Mongke had begun to give serious consideration to a large-scale and direct attack on the Southern Sung. Starting in 1254, Mongolian raids and reconnaissances along the Sung's northern borders were stepped up, and a program of political warfare was initiated, designed to induce Sung frontier officials to defect. In the summer of 1256, citing as his casus belli the imprisonment of Mongolian envoys by the Sung, the khaghan formally announced his intention of attacking the Sung and held preliminary discussions on a plan of 123 On Khubilai's reforms and his subsequent confrontation with the emperor, see Hsiao Ch'i-ch'ing, Yiian tai shih hsin fan (Taipei, 1983), pp. 2 8 5 - 9 4 ; and Hok-lam Chan, "Yao Shu (1201-1278)," Papers on Far Eastern History, 22 (1980), pp. 25—8. Cambridge Histories Online © Cambridge University Press, 2008 4IO THE RISE OF THE MONGOLIAN EMPIRE military action. In early 1257 assault troops were deployed to the south, and Mongke made known his resolve to lead in person one of the three invasion armies committed to the campaign. A year later the emperor moved south through the Gobi, where he met with Khubilai and other princes to go over the details of the forthcoming operation. By the spring of 1258 Mongke and his force reached the Liu-p'an Mountains of Kansu and established a temporary camp. Here he stayed throughout the month of May, making final preparations for his drive into Szechwan. Although it had been partially occupied by the Mongols at the time of Khubilai's attack on Ta-li, Szechwan still contained a formidable Sung presence. The Sung commanders committed their available forces to defend a number of fortified cities scattered throughout the province. Mongke therefore divided his troops into three separate columns and began the task of reducing these garrisons one by one. At the head of one column, the emperor penetrated northern Szechwan in the fall of 1258, taking several cities by storm, and then proceeded to Li-chou, which had been under Mongolian control since 1253. After inspecting its defenses he moved his army into the heart of the Szechwan basin, where he spent the rest of 1258 besieging various cities and accepting the surrender of numerous Sung officials. In early 1259 he moved to western Szechwan to attack Ya-chou (now Ya-an), which he took in January. Mongke then turned back east and advanced on Ho-chou, a major garrison city on the Chia-ling River, around which he concentrated a considerable part of his forces. Assaults on the city began in February and were intensified throughout the spring and summer. Ho-chou, however, put up a stout defense, successfully repelling all the Mongolian assaults, and in July, unable to take the city, Mongke ordered the siege lifted. Leaving behind three thousand men to blockade Ho-chou, the emperor sent the bulk of his troops against Ch'ung-ch'ing in the south of Szechwan. In the meantime two additional Mongolian armies had entered the conflict, invading Sung territory from different directions. Khubilai, commanding the central army, opened up a new front in Hupeh in August 1259, moving against the Sung city of O-chou on the middle Yangtze, while the eastern wing, commanded by the Mongolian prince Taghachar, made for northern Huai-nan (modern Anhwei Province). In conjunction with these operations, Uriyangkhadai struck northeast into Sung territory from southern Yunnan, with the objective of joining Khubilai's forces on the Yangtze. Now simultaneously assailed from the north, west, and south by Mongolian field armies, the Sung defenses were gravely threatened, and the very survival of their dynasty was in serious jeopardy. The dire situation in which they found themselves was, however, shortly alleviated. On 12 August 1259, Cambridge Histories Online © Cambridge University Press, 2008 THE EMPIRE ON THE EVE OF CIVIL WAR 411 Mongke died, of either dysentery or wounds inflicted by a Sung catapult, in the mountains outside Ho-chou.124 His commanders halted operations in Szechwan, returned his body to the north, and informed the imperial family of his death. When Khubilai received word of his brother's demise in the following month, he terminated his siege of O-chou and immediately called off all further campaigning in China. For the Sung dynasty, Mongke's death meant a twenty-year reprieve, and for the Mongolian empire, a new and divisive succession crisis, from which it would never fully recover. EPILOGUE: T H E EMPIRE ON THE EVE OF CIVIL WAR The Mongolian empire reached the high point of its power under Mongke. He effectively centralized imperial authority and, in so doing, procured the resources to complete the conquests of Chinggis khan and Ogodei in west Asia and Korea, to extend Mongolian rule into southwest China, and to begin the subjugation of the Sung. Yet for all his concrete achievements, Mongke also bears heavy responsibility for the discord and strife that immediately followed his death. Through errors of commission and omission Mongke, the advocate of centralization, helped sow the seeds of a civil war (1260—4) that forever shattered the unity of the Mongolian empire. By far the most damaging of these errors was Mongke's failure to provide for an uncontested line of succession within his own family. So far as is known, Mongke never explicitly stated his choice of successor. He may well have meant to indicate his preference for his younger brother, Arigh Boke, by leaving him in command at Khara Khorum when departing for the south in 1258, but if this indeed was his intention, then he unwisely and unaccountably took no further steps to advertise and solemnize his choice or to strengthen Arigh Boke's position. The resulting ambiguity led, of course, to the contending claims of Khubilai and Arigh Boke, and the split among the Toluids in turn gave disgruntled members of collateral lines the opportunity to reassert and advance their own disparate and often-conflicting claims and interests. The surviving Chaghadaid princes, still smarting from Mongke's purge of their family, moved swiftly to establish their independence in Turkestan. In 1260 Alghu, a grandson of Chaghadai, easily deposed his cousin Mubarak Shah, the Toluid appointee. Henceforth, the Chaghadai khanate, which had designs on Afghanistan and east Turkestan, was almost constantly at war with the Il-khans of Iran and the grand khans of China. 124 On the cause of his death, see Sergei A. Shkoliar, Kilaiskaia doognatnl'naia artilleria (Moscow, 1980), PP- 3 3 6 - 7 . n. 18. Cambridge Histories Online © Cambridge University Press, 2008 412 THE RISE OF THE MONGOLIAN EMPIRE The Ogodeids, even more embittered by the events of 1251, displayed the same independent attitude. Khaidu, a grandson of Ogodei, was the first of his family openly to defy imperial authority. In 1256 he had arrested and then refused to return to court an envoy sent to him by the khaghan. Mongke, presumably preoccupied with the approaching campaign against the Sung, took no action against the rebellious prince. Then, when the struggle for the succession between Khubilai and Arigh Boke broke out, Khaidu did what he could to encourage the conflict, hoping, of course, to bring about the ruin of the house of the Toluids.12' As an Ogodeid, he naturally felt that the office of khaghan belonged exclusively to his family and, at the first opportunity, organized (1269) a coalition of Mongolian princes in Central Asia that contested Khubilai's and his successors' right to the throne into the early years of the fourteenth century. From the perspective of Khubilai, who was forced to wage a long and costly war against Khaidu, Mongke's failure to crush the Ogodeid champion at the very first sign of rebelliousness must have seemed an inexplicable and unpardonable blunder. Trouble was brewing, too, in the Jochid realm. Batu died around 1255 and, with Mongke's approval, was succeeded by two short-lived descendants, first by a son, Sartakh (r. ca. 1255—7), and then by a very young grandson, Ulaghchi (r. 1257). Upon the death of the latter, Berke, a brother of Batu, came to the throne, again with Mongke's consent. A recent convert to Islam, Berke also sided with Arigh Boke in 1260 and made war on Khubilai's ally, the Il-khan Hiilegii. Some Muslim chroniclers of the era126 attribute his enmity toward the Il-khan to pious outrage over the destruction of Baghdad, long the spiritual center of the Islamic world, but the real source of Berke's dissatisfaction with Hiilegu was the disposition of territories in the Transcaucasus. In 1252 Mongke had granted Georgia (Ch'ii-erh-chih)127 to Berke but Hiilegii, covetous of this populous country, ignored the Jochid prince's rights in the region, an affront that prompted Berke's invasion of Iran in 1262 and his alliance with the Il-khans' principal adversaries, the Mamluks of Egypt. For the first time a Mongolian prince of the blood had allied himself with a foreign power in a dispute with another Mongolian khan. The succession crisis of 1260 thus swiftly exposed the latent personal animosities and territorial rivalries among the Mongolian princes that had for the most part lain hidden beneath the surface during Mongke's reign. Khubilai's decisive victory over Arigh Boke in 1264 did nothing to eradicate 125 YS, 153, p. 3619; and Pelliot, Notes on Marco Polo, vol. i, pp. 126-7. 126 Minhaj al-Din JuzjanI, Tabaqat-i nasiri, ed. W. Nassau Lees (Calcutta, 1964), pp. 430-1; and Tabaqat-iNasirT, trans. H. G. Rafferty(New Delhi, 1970), vol. 2, pp. 1255-7. 127 YS, 3, p. 45. This Chinese form, Ch'ii-erh-chih, goes back to Gurj, the standard Persian and Arabic name for Georgia; see Pelliot, Notes on Marco Polo, vol. 2, pp. 738-9. Cambridge Histories Online © Cambridge University Press, 2008 THE EMPIRE ON THE EVE OF CIVIL WAR 413 these deep divisions and produced at best a temporary and exceedingly tenuous restoration of unity. For all intents and purposes, there were now four independent regional khanates in existence, each pursuing its own separate interests and objectives. Only the Il-khans remained loyal to Khubilai, but even they, while still recognizing the supremacy of the khaghan in China, exercised extensive autonomy in the conduct of their internal affairs. The four khanates continued to interact with one another well into the fourteenth century, but they did so as sovereign states; they formed alliances, fought one another, exchanged envoys, and traded commercial products. And in the case of Yuan China and Il-khan Iran, there was an extensive program of cultural and scientific interaction. But the four khanates never again pooled their resources in a cooperative military endeavor. The campaigns against the Southern Sung and the 'Abbasid Caliphate were the last unified military ventures of the Mongolian empire. The Mongols' mandate to bring the known world under their dominion was never to be realized. Cambridge Histories Online © Cambridge University Press, 2008 CHAPTER 5 THE REIGN OF KHUBILAI KHAN THE EARLY YEARS In 1229, when Khubilai's uncle Ogodei, rather than his father Tolui, had been selected as the successor to Chinggis khan, it had seemed unlikely that Khubilai would play more than a subsidiary role in Mongolian history. Few could have foreseen that he would eventually become the most powerful figure in the Mongolian domains. One of those who did was his remarkably competent and intelligent mother Sorghaghtani Beki. Like her near contemporary Eleanor of Aquitaine, Sorghaghtani raised four sons — Mongke, Khubilai, Hiilegii, and Arigh Boke — who became monarchs. Both women dedicated themselves to their sons' careers and did not halt their efforts until their sons were enthroned. Sorghaghtani's contemporaries regarded her as one of the greatest women of her age. The Hebrew physician Bar Hebraeus, quoting a poet of his era, wrote about her that "if I were to see among the race of women another woman like this, I should say that the race of women was far superior to men."1 Without her political savoir-faire and the training she provided them, her sons would not have succeeded in replacing the house of Ogodei as the main Mongolian royal line. Sorghaghtani profoundly influenced her sons. First of all, she ensured that they were literate, an indispensable skill for those who aspired to rule a great empire. In addition, by her own example, she taught them some basic political principles. In her appanage in north China, for example, she did not exploit her Chinese subjects, nor did she plunder the region. She recognized that tax revenues would increase if she promoted, instead of interfered with, the native agrarian economy. Her policy of religious toleration also made an impression on her sons. Though she herself was a Nestorian Christian, she contributed funds and other tangible support to Buddhist monasteries, Taoist temples, and Islamic religious academies (madrasa).2 She assumed that 1 E. A. Wallis Budge, trans., The cbronography of Gregory Abu'l Faraj the son of Aaron, the Hebrew physician commonly known as Bar Hebraeus (London, 1932), vol. 1, p . 398. 2 John Andrew Boyle, trans., The successors of Genghis khan: Translated from the Persian of Rashid al-Din (New York, 1971), pp. 199—200. 414 Cambridge Histories Online © Cambridge University Press, 2008 THE EARLY YEARS 415 patronage of the religions in her realm would make it easier for her to govern. In administering her lands, she recruited several Chinese advisers to devise suitable institutions to regulate her Chinese subjects. The political lessons were not lost on her subjects. Sorghaghtani gave birth to Khubilai on 23 September 1215, the year that Chinggis khan had seized Peking. The sources regarding Khubilai's childhood, education, and travels are limited. It seems clear, however, that his upbringing was left to his mother, for his father Tolui was away on military campaigns in. Central Asia or China during Khubilai's childhood and adolescence. Sorghaghtani recruited a Uighur named Tolochu to teach Khubilai to read and write Mongolian.3 She made sure that he was exposed to Chinese ways through her Chinese advisers, although strangely he was never taught to read Chinese. She was also responsible for Khubilai's first official position. She cajoled her brother-in-law Ogodei into granting Khubilai as his appanage jurisdiction over Hsing-chou, a region in Hopei, which in 1236 had a population of ten thousand households. Following his mother's example, Khubilai sought good relations with his Chinese subjects by fostering agriculture and pursuing a policy of religious toleration. Also like his mother, he gathered around him a coterie of advisers, most of whom were Chinese.4 Later in his career, he would consult Nestorian Christians, Tibetan Buddhists, and Central Asian Muslims and would not limit himself to Chinese counselors. Khubilai's earliest advisers were an eclectic lot. The monk Hai-yiin (1205-57) introduced him to the precepts and practices of Buddhism and arranged a meeting between him and Liu Ping-chung (1216-74), who proved to be one of Khubilai's most invaluable counselors. Chao Pi (1220—76) lectured Khubilai on Confucianism. But Yao Shu (1201—78) was unquestionably the most influential of his Confucian advisers, for he was a "practical and versatile scholar who skillfully adapted his classical learning to changing circumstances."' His pragmatism and realism appealed to Khubilai and the Mongol nobility. Yao Shu's advice was always couched in logic that the Mongols could understand. A few Confucian advisers served Khubilai only reluctantly. Chao Fu, one of these less-thanenchanted counselors, was once summoned by Khubilai. During their con3 On Tolochu and the influence of other Uighurs on the Mongols, see Igor de Rachewiltz, "Turks in China under the Mongols: A preliminary investigation of Turco-Mongol relations in the 13th and 14th centuries," in China among equals: The Middle Kingdom and ill neighbors, lolh-14th centuries, ed. Morris Rossabi (Berkeley and Los Angeles, 1983), pp. 281-310. 4 On these advisers, see Hsiao Ch'i-ch'ing, "Hu-pi-lieh shih tai ch'ien ti chiu lii k'ao," Ta-lu tsa chih, 25, no. 1 (July 1962), pp. 18—20; 25, no. 3 (August 1962), pp. 2 4 - 6 ; and Yao Ts'ung-wu, "Hu-pi-lieh han tui yii Han hua t'ai tu te fen hsi," Ta-lu tsa chih, 11, 1 (July 1955), pp- 22—32. 5 Chan Hok-Iam, "Yao Shu (1201-1278)," Papers on Far Eastern History, 22 (1980), p. 45. Cambridge Histories Online © Cambridge University Press, 2008 416 THE REIGN OF KHUBILAI KHAN versation, Khubilai asked him how to subdue the Southern Sung. Chao replied, "The Sung is my mother and father land. There is no way that I will lead others to attack my mother and father. "6 Despite such occasional resistance, Khubilai developed a kind of "kitchen cabinet" of advisers. He could seek their counsel on practical problems or could discuss Confucian morality with them. No other Mongolian noble had recruited so many Chinese Confucians. Such a large retinue was not needed to help him rule his own appanage in Hsing-chou. Clearly, Khubilai foresaw a more important role for himself in the Mongolian domains, particularly in China. His advisers, for their part, served him for a variety of reasons. Some were attracted by monetary rewards or exemption from labor service or other duties. Others hoped to restore unity and order to China and thought that the Mongols had the best chance of unifying north and south China. Still others worked for Khubilai in order to improve the lot of their people and to seek to sinicize the Mongols. Khubilai did not, however, limit himself to Confucian scholars as advisers and administrators. His Chinese subjects might not be entirely trustworthy, and they were not competent to help Khubilai fulfill all his responsibilities. For military advice and assistance, for example, he relied on Mongolian commanders. He entrusted his current and his later military campaigns to the Mongols. He turned to Uighurs and other Turks as translators and interpreters, local governors, and secretaries. By the time of Mongke's death in 1259, therefore, Khubilai had recruited advisers and administrators representing a variety of different religions, ethnic groups, and occupations. Although he was not the first Mongol to seek advice and assistance from the peoples they had subjugated - Ogodei and Mongke had done so before him — he was unique in having such a large coterie of advisers. One of Khubilai's influential advisers was his wife Chabi. The sources yield few details about her life and career, but they do reveal that she was a suitable mate for a man who aspired to be a ruler of a great empire. She advised Khubilai to prevent his Mongolian retainers from converting the fertile farmland in his appanage to grazing areas for their flocks. Her reason was that if he encouraged such conversions, he would not only disrupt the native agrarian economy but would also alienate his Chinese subjects. Chabi was also a fervent Buddhist and was, in particular, attracted by Tibetan Buddhism. Her firstborn son was given the name Dorji (from the Tibetan rDorje). She no doubt prompted Khubilai to invite such Buddhist monks as Hai-yiin to his territory and to discuss the complexities of Buddhist dogma 6 Sung Lien et al., eds., Yuan shih (Peking, 1976 ed.), 189, p. 4314 (hereafter cited as YS). Cambridge Histories Online © Cambridge University Press, 2008 THE EARLY YEARS 417 and doctrine with them. Her enthusiasm for Buddhism surely motivated his patronage of that religion. In sum, Khubilai considered her views seriously. Yet this distinguished group of advisers, including his wife Chabi, did not bring Khubilai into prominence until his brother Mongke's accession to the khaghanate in 1251. Khubilai remained a shadowy figure and is scarcely mentioned in the Mongolian, Chinese, or Persian sources of that era. With his older brother's usurpation of power, however, Khubilai began to assume greater responsibilities and is accorded more attention in the histories of that period. He appears to have helped persuade Mongke to challenge the house of Ogodei for control of the Mongolian domains. With his brother's success in 1251, Khubilai joined the inner circle of decision makers at court and was, in part, responsible for the relatively lenient treatment accorded to Mongke's opponents in the struggle for the khaghanate. Khubilai's first significant assignment during Mongke's reign entailed a military expedition. Mongke wanted to continue the expansionist policies of his predecessors, and he ordered his younger brother Hiilegii to extend Mongolian control into the Middle East. Khubilai, the third brother, was to command an expedition against the kingdom of Ta-li in the modern province of Yunnan. Control over this region in southwest China would offer the Mongols still another base from which to attack the Southern Sung dynasty. Khubilai received Mongke's order to launch this expedition in July 1252, but he did not move toward Ta-li until September 1253.7 The preparations for his campaign against Ta-li were particularly important, for it was his first major assignment. Finally, at the age of thirty-six, he was granted responsibility for a vital military objective. He did not wish to spoil this opportunity to prove himself as a military commander. In the late summer of 1253 Khubilai was prepared to undertake the mission that Mongke had assigned him. Starting from Shensi, he marched his troops toward Ta-li. Before launching an attack, Khubilai dispatched three emissaries to demand that Ta-li submit. Tuan Hsing-chih, the king, and Kao T'ai-hsiang, the leading minister, who was the real power behind the throne, responded by executing the three envoys. Khubilai thereupon unleashed a three-pronged assault on the kingdom, and his troops defeated the enemy, compelling a retreat to the capital. The Chinese sources credit Yao Shu with preventing unnecessary bloodshed. He reputedly persuaded Khubilai to order his men to fashion a silk banner with a message assuring 7 On this campaign and its significance, see YS, 4, pp. 59—60; Hsia Kuang-nan, Yuan tai Yun-nan sbih ti ts'ung k'ao mu lu (Shanghai, 1935), p. 107; Yang Shen, Nan-chao yeh sbih (repr., Taipei, 1969), pp. 184-6; and Constantin M. d'Ohsson, Histoin da mmgols depuit Tchinguiz-Khan jusqu'a Timour-Bey ou Tamerlan (The Hague, 1834), vol. 2, pp. 310—14. Cambridge Histories Online © Cambridge University Press, 2008 418 THE REIGN OF KHUBILAI KHAN the inhabitants of the capital that their lives would be spared if they capitulated. With this guarantee in hand, Ta-li opted for surrender. Khubilai kept his word: The population was unharmed; their system of government was scarcely altered; and the Tuan family was allowed to share power with Khubilai's appointee as pacification commissioner (hsiian-fu shih). Khubilai's first military expedition was thus a triumph, for he had accomplished what Mongke had wanted. His forces had suffered few casualties, yet he had extended Mongolian control into a vital region — a base from which to launch an invasion of southern China and a thoroughfare for an expansion of trade with Burma and India. He had proved himself in Mongolian eyes by leading a successful military campaign. Acceptance into the Mongolian nobility required a display of one's skills as a military leader, and Khubilai had now shown his mettle, ensuring him increased stature among his fellow Mongols. KHUBILAI AND C H I N A , 1253-1259 Returning to his appanage after his triumphant campaign in the southwest, Khubilai began to focus his attention on the administration of his lands. With the support and assistance of his Confucian advisers, he had promoted agriculture, issued paper money to encourage commerce, and collected taxes. His domains, in short, became stable and prosperous, allowing Khubilai to concern himself with long-range plans. He sought guidance in his efforts from a Buddhist monk named Liu Ping-chung, a remarkable poet, calligrapher, painter, mathematician, and astronomer whom Hai-yiin had introduced to Khubilai.8 Liu advised Khubilai to nurture, protect, and employ Confucian scholars, whom he portrayed as national treasures. He also urged the Mongolian prince to build schools to train these scholars, to restore the traditional civil service examinations, and to reintroduce the ancient Chinese rituals and musical ceremonies. Finally, he proposed fixed tax and military obligations that did not unduly burden Khubilai's Chinese subjects. Khubilai approved all these proposals except for the revival of the civil service examinations. He did not want to commit himself totally to the use of Chinese, or at least to Chinese-speaking advisers and officials. Perhaps the most significant collaboration between Liu and Khubilai was the construction of a capital within the Mongolian prince's new domains in the sedentary world. In 1256, they selected a location north of the Luan River and thirty-six miles west of the Ch'ing dynasty town of Dolon-nor in 8 On Liu's programs, see Hok-lam Chan, "Liu Ping-chung (1216-1274): A Buddhist-Taoist statesman at the court of Khubilai khan," Toung Pao, 53 (1967), pp. 98—146. Cambridge Histories Online © Cambridge University Press, 2008 KHUBILAI AND CHINA 419 Inner Mongolia. This site was cooler in summer than was north China, was surrounded on all sides by mountains, and had sufficient water for a town of moderate size. It lay about ten days' journey from Peking along the edges of both the Chinese agrarian frontier and the Mongolian pasturelands.9 Traditional Mongols thus could not accuse Khubilai of abandoning his heritage and siding with the Chinese. Yet he had signaled a change in focus to his Chinese subjects. Still another signal for Khubilai's sedentary subjects was that K'ai-p'ing, the name of the new town, was modeled on Chinese capitals of the past. The placement of many of the buildings was based on the prescriptions of the ancient Chinese text entitled the / ching (Book of changes). The town was divided into three sections. The outer city was in the shape of a square surrounded by an earthen wall twelve to eighteen feet high. Most of the population lived in this area in mud or board houses, and several Buddhist temples were also located here. The second section was the Inner City, which contained the residences of Khubilai and his retinue. A brick wall ten to sixteen feet high surrounded the whole Inner City. The Imperial Palace (Chinese: Ta-an ko, Pavilion of great harmony), which was built on an earthen platform, was the principal focus of this section. Inside the palace, "the halls, rooms, and passages all [were} gilded and wonderfully painted, within which pictures and images and birds and trees and flowers and many kinds of things, so well and so cunningly that it [was] a delight and a wonder to see."10 Numerous palaces and government buildings were also scattered around the Inner City. The hunting reserve, composed of meadowlands, woods, and streams and located north of the Outer City, was the last section of K'ai-p'ing. A variety of tame animals, principally deer, were kept in this park for Khubilai's hunts. Also roaming in this park were special breeds of white mares and cows whose milk "no one else in the world dared [drink] . . . except only the great Kaan and his descendants."11 Witnessing all these developments in Khubilai's domains, Mongke must have feared that his younger brother was identifying with his Chinese subjects. Mongke's officials also accused Khubilai of using Chinese laws to administer his lands and of eschewing the traditional Mongolian laws. In 1257, Mongke sent two aides to investigate conditions in Khubilai's lands. Uncovering what they claimed were numerous infractions and transgressions 9 On this capital, see Yoshito Harada, Shang-tu: The summer capital of the Yuan dynasty (Tokyo, 1941); Komai Kazuchikai, "Geno no Joto narabi ni Oaito heimen ni tsuite," Toa rorno, 3 (1940), pp. 129— 39; and Ishida Mikinosuke, "Gen no Joto ni tsuite," in vol. 1 of Nihon daigaku soritsu shichi junen kinen rombun tat, ed. Nihon daigaku (Tokyo, i960), pp. 271—319. 10 Atthut C. Moule and Paul Pelliot, Marco Polo: The description of the world (London, 1938), vol. 1, p. .85. 11 Ibid., p. 187. Cambridge Histories Online © Cambridge University Press, 2008 42O THE REIGN OF KHUBILAI KHAN of the law, they arrested and executed several high-ranking officials.12 But the purge did not reach Khubilai himself. Within a few months, Mongke faced two critical problems, and he felt that Khubilai's assistance would be invaluable in resolving them. The first was a religious conflict between the Buddhists and the Taoists that had escalated into pitched battles and the destruction of temples and monasteries. The second was the conquest of southern China, by far the wealthiest region of the Middle Kingdom. Khubilai and his Chinese advisers could help Mongke resolve the Buddhist—Taoist dispute as well as gain the allegiance of the Chinese in the south. Mongke thus arranged a meeting with his younger brother early in 1258, and the two had a reconciliation. In effect, they both needed each other. Shortly after their meeting, Khubilai convened a conference of three hundred Buddhists, two hundred Taoists, and two hundred Confucian scholars and court officials to debate the issues on which they disagreed. He would judge the conflicting claims of the two religious groups. Both the Taoists and the Buddhists sought supremacy and were willing to appeal to the secular authorities for assistance. Basing their arguments on the so-called bua-hu (conversion of the barbarians) theory, the Taoists asserted that Lao Tzu had left China for the western regions, where he had transformed himself into the Buddha and had started to propagate the teachings of Buddhism. They implied that Buddhism was simply a vulgarized form of Taoism that Lao Tzu developed in order to appeal to the less sophisticated people of India. The Taoist debaters relied on the evidence found in two ancient texts, the Hua-hu ching (Book on the conversion of the barbarians) and the Pa-shih-i hua-t'u (Illustrations of the eighty-one conversions).13 Their Buddhist opponents, however, in particular the Tibetan 'Phags-pa lama, questioned the authenticity of these works. They pointed out that early Chinese writings, including Ssu-ma Ch'ien's great historical account, the Shih chi (Historical records), did not mention the two works. The 'Phags-pa lama suggested that the two texts were later forgeries, and Khubilai agreed with this assessment. The Mongolian prince offered the Taoists one chance to redeem themselves. He invited them to perform the magical feats at which they claimed to be proficient. Unable to meet this challenge, the Taoists were judged to be the losers in the debate. Khubilai ordered all copies of the two Taoist texts burned and all the property confiscated by the Taoists to revert back to their 12 YS, 158, p. 3713. 13 Joseph Thiel, "Der Streit der Buddhisten und Taoisten zur Mongolenzeit," Monumenta Serica, 20 (1961), pp. 1—81; Noritada Kubo, "Prolegomena on the study of the controversies between Buddhists and Taoists in the Yuan period," Memoirs of the Research Department of the Toyo bunko, 25 O967), pp. 39-61. Cambridge Histories Online © Cambridge University Press, 2008 KHUBILAI AND CHINA 42 I Buddhist owners."* He did not proscribe Taoism but merely curbed what he believed to be Taoist excesses. A vindictive purge would have enraged the Taoists, and their many sympathizers would have impeded the Mongols' efforts to govern north China. Khubilai's decision, as well as his moderation in punishing the Taoists, appears to have met with the approval of his Chinese subjects. Having acquitted himself with distinction at the debate, Khubilai received a new assignment. Toward the end of 1258, Mongke devised a plan for the conquest of southern China. He would deploy his forces along four fronts, with the troops under his own personal command first seeking to occupy Szechwan and then marching eastward. Khubilai would then lead his detachment from K'ai-p'ing and cross the Yangtze River at O-chou on the central Yangtze, where he would engage the Sung forces. The other detachments would move from Yunnan and from the Liu-p'an shan area in Shensi toward the Sung stronghold at Hsiang-yang. The Mongols clearly hoped for a quick victory in the west that would induce the Sung to capitulate. But Mongke's own campaign did not fulfill his expectations, for he encountered stiff resistance from the Sung forces. After taking Ch'eng-tu in March 1258, his expedition was bogged down vainly attempting to take the strongly defended city of Ho-chou (modern Ho-ch'uan county, Szechwan) throughout the last half of 1258 and the first seven months of 1259. Then on 11 August, Mongke died while on campaign in the vicinity of Ho-chou. The Mongolian campaigns throughout Eurasia came to a standstill after Mongke's death. His own troops did not make any further advances and did not link up with the three other divisions attacking the Sung. In the Middle East, Mongke's younger brother Hiilegii, who had expanded the lands under Mongol control in the west, hastily headed back toward the Mongolian homeland, leaving only a small detachment to guard his newly conquered domains. This disruption in the Mongolian world was due to the lack of an orderly succession to the khaghanate. The leader with the greatest military power often emerged victorious. The struggle for the throne in 1259 was conducted within the house of Tolui. It was more than a struggle between two men, for it reflected a major division within the Mongolian elite. Khubilai, who was attracted by the civilizations he had helped conquer and who sought advice and aid from the subject populace, represented the Mongols, who were influenced by and tried for an accommodation with the sedentary world. His younger brother Arigh Boke emerged as the defender of traditional Mongolian ways and values. The 14 Edouard Chavannes, "Inscriptions ec pieces de chancellerie chinoises de l'lpoque mongole," Voting Poo, 9 (1908), pp. 3 8 1 - 4 . Cambridge Histories Online © Cambridge University Press, 2008 422 THE REIGN OF KHUBILAI KHAN world of the steppe was, to him, more attractive than was the world of the sown. He mistrusted his older brothers Hiilegii and Khubilai and considered them to be tainted by foreign values and attitudes. The stage was thus set for a fraternal combat concerning the future direction of the Mongolian empire. Yet the struggle was delayed for a few months. In mid-September 1259, Khubilai learned of Mongke's death from a messenger dispatched by his halfbrother, who requested that he return to the Mongolian homeland for the election of the new khaghan. Khubilai had just reached the northern banks of the Yangtze River and was preparing for an invasion of the south. According to the Yuan shih, he told the messenger, "I have received imperial orders to come south. How can I return without having achieved merit?"1' The Persian historian Rashid al-Din substantiates this account, noting that Khubilai responded, "We have come hither with an army like ants or locusts: How can we turn back, our task undone, because of rumors?"16 It seems likely that Khubilai wanted to triumph over the Sung in order to improve his chances in the struggle for the khaghanate. He would enter the contest as a successful military leader. Thus he did not return to the north immediately. KHUBILAI VERSUS ARIGH BOKE Khubilai's troops persisted in their campaign against the Southern Sung through the winter of 1259. They first crossed the Yangtze and then laid siege to the heavily fortified town of O-chou. A victory here would have bolstered Khubilai's prestige in the Mongolian world, but the Sung defenders of the town were determined not to surrender. The Southern Sung chancellor (cb'eng-hsiang) Chia Ssu-tao was, however, willing to compromise. He dispatched an envoy to offer Khubilai an annual payment of silver and textiles in return for a pledge to maintain the Yangtze as their common border. Khubilai's Confucian adviser Chao Pi commented, "Now, after we have already crossed the Yangtze, what use are these words?"'7 Khubilai was intent on victory. The succession crisis saved the Sung. Arigh Boke mobilized his troops right after Mongke's death and began to create alliances with influential Mongolian nobles. Early in 1260, one of his allies marched toward Khubilai's town of K'ai-p'ing. Chabi, who had stayed behind while her husband was on campaign, immediately sent an envoy to inform Khubilai of his younger brother's plans and actions. Khubilai would have to abandon the 15 YS, 4, p. 61. 16 Boyle, The successors of Genghis khan, p. 248. 17 Herbert Franke, "Chia Ssu-tao (1213—1275): A 'bad last minister,' " in Confucian personalities, ed. Arthur F. Wright and Denis C. Twitchett (Stanford, Calif., 1962), p. 227. Cambridge Histories Online © Cambridge University Press, 2008 KHUBILAI AND ARIGH BOKE 423 siege of O-chou and depart for the north to counter Arigh Boke. He withdrew most of his troops from O-chou, leaving only a token force to preserve the territorial gains he had made.18 Capitalizing on Khubilai's sudden withdrawal, Chia Ssu-tao authorized an attack on the small detachment of Mongolian troops, quickly defeating them and reoccupying the Sung territories. Chia deliberately portrayed this minor engagement as a great victory, misleading the Sung court and contributing to its resolve to reject any compromise with the Mongols. Meanwhile Khubilai marched toward and reached K'ai-p'ing in the spring of 1260. The Yuan shih reveals that numerous princes "begged" Khubilai to take the throne. After three ceremonial "refusals," he acceded to their wishes, and on 5 May a hastily convened khuriltai elected him as the great khan. Because much of the Mongolian nobility did not attend the meeting, Khubilai's election could be and was challenged. Within a month, for example, Arigh Boke was proclaimed as the rival great khan in the old Mongolian capital of Khara Khorum. Arigh Boke could count on the support of two of the three remaining principal khanates, the Golden Horde of Russia and the Chaghadai khanate of Central Asia. Khubilai's only supporter was his brother Hiilegii, who himself faced serious threats to his authority in the Middle East. While en route back to Mongolia, Hiilegii had learned that the Mamluk rulers of Egypt had defeated his forces at 'Ain Jalut (in Syria) in September 1260. •» Moreover, the Golden Horde, seeking to dislodge Hiilegii from Azerbaijan along the Russian-Persian border, had declared war on him. Hiilegii's attention was thus diverted elsewhere, and so he could be of little help to Khubilai in the struggle for succession. Khubilai was compelled to rely on the resources of China and on his Chinese subjects for support. He issued a proclamation, which was actually composed by his Confucian adviser Wang O,20 admitting that Mongolian military skills were insufficient to rule China. A sage who cultivated goodness and love and who governed in accordance with the traditions of the ancestors was needed to unite the Chinese, and he implied that he was just the man. He also advocated a reduction of the tax and corvee burdens on the people." A few days after issuing this proclamation Khubilai adopted the Chinese reign title chung-t'ung (literally, "moderate rule"),22 although he did so without adopting a Chinese name for his dynasty. The governmental 18 YS, 4, pp. 6 2 - 3 . 19 Bernard Lewis, "Egypt and Syria," in vol. iA ofThe Cambridge history of Islam, ed. P. M. Holt, Ann K. S. Lambton, and Bernard Lewis (Cambridge, 1970), pp. 2 1 2 - 1 3 . 20 See Chan Hok-Ham, "Wang O (1190—1273)," Papers on far Eastern History, 12 (1975), pp. 43—70. 21 For the text of the proclamation, see YS, 4, pp. 64—5. 22 Also interpreted as "pivotal succession." See Morris Rossabi, Khubilai Khan: His life and times (Berkeley and Los Angeles, 1988), p. 243, n. 12. Cambridge Histories Online © Cambridge University Press, 2008 424 THE REIGN OF KHUBILAI KHAN institutions that he developed, the Chung-shu sheng (Central Secretariat) and the Hsiian-wei ssu (Pacification commissions), resembled the traditional Chinese ones. Khubilai, in effect, wished to signal to all the Chinese that he intended to adopt the trappings and style of a typical Chinese ruler. The Southern Sung Chinese, however, were not receptive to such concessions. They detained Hao Ching, the envoy sent by Khubilai to negotiate a diplomatic settlement of their conflict. Hao remained a prisoner at the Sung court from 1260 until the successful launching of Khubilai's military campaigns against the southern Chinese in the 1270s. Khubilai himself was able to use the resources of north China but sought to deny supplies from the sedentary world to Arigh Boke. Based in Khara Khorum, Arigh Boke needed to import most of his provisions, and Khubilai was determined to sever his younger brother's supply lines. Kansu and northwest China were controlled by one of Khubilai's allies, as were the Uighur lands farther west. Arigh Boke's principal source of support was the Chaghadai khan Alghu, who was based in Central Asia. Alghu at first helped Arigh Boke in his efforts to seize the throne, but disputes over their individual shares of tax revenues and spoils finally drove them apart. After 1262, therefore, Arigh Boke had no dependable allies and no reliable source of supplies. It was only a matter of time before he would have to abandon his struggle for the throne. After several skirmishes, in 1263 he surrendered to Khubilai. Conveniently enough for Khubilai, he died within a few years while still in captivity, giving rise to speculation that he had been poisoned. Despite Arigh Boke's death, Khubilai still faced other threats to his authority. Doubts over the legitimacy of his succession continued to haunt Khubilai in his efforts to portray himself as the khaghan of the Mongolian domains23 (see Map 30). Similarly, Khubilai's acceptance as the emperor of China was tenuous, and rival claimants emerged. His first challenger was Li T'an, the ruler of the I-tu district in Shantung Province. The Chinese sources depict Li as a "rebellious minister" because he eventually turned against Khubilai. Thus he is perceived as a rebel rather than as a loyalist dedicated to the establishment of a Chinese dynasty. Such misrepresentation aside, Li posed a direct threat to Khubilai's claim to be the emperor of China. Earlier, Li had cooperated with Mongke in campaigns against the Sung and had raided several Sung coastal towns. When Khubilai ascended to the throne in China in 1260, therefore, there seemed to be no reason to harbor suspicions about Li's loyalty to the Mongols. Moreover, Li was the son-in-law of Wang Wen-t'ung, whom 23 On his efforts to achieve legitimacy, see Herbert Franke, From tribal chieftain to universal emperor and god: The legitimation of the Yiian dynasty (Munich, 1978), pp. 2 5 - 5 2 . Cambridge Histories Online © Cambridge University Press, 2008 THE GOLDEN HORD (QIPCHAQ KHANAT Tashkent Bukhara. Samarkand. " V o k a n d KHANATE Nishapur v ^ , .Yarkand Khotan TIBET Khanates BURMA Tributary states 1500 km MAP 30. The Mongolian khanates Cambridge Histories Online © Cambridge University Press, 2008 426 THE REIGN OF KHUBILAI KHAN Khubilai had just appointed as chief administrator (p'ing-chang cheng-shih) in the Central Secretariat (Chung-shu sheng), one of the most influential positions in his government. In 1260 and 1261, Khubilai sent gold and silver to Li to cover the costs of campaigns against the Sung. Late in 1261, however, Li prepared to break away from Khubilai and to effect an agreement with the Chinese in the south. Having access to wealth derived from the valuable reserves of salt and copper in Shantung, Li had the resources to mount a major challenge to Mongolian rule. He may have received assurances of support from the Sung and must have determined that trade and other economic relations with the Chinese in the south could offer more benefits than could friendly relations with the Mongols. As an ethnic Chinese, he may, in addition, have felt loyalty to the Sung. Whatever his motivations, on 22 February 1262, he rebelled against those whom he had earlier accepted as his overlords. Khubilai responded immediately by dispatching several of his most trusted military men to deal with the troublesome Chinese leader. Shih T'ien-tse and Shih Ch'u, two of Khubilai's leading generals, together with his Confucian adviser Chao Pi, set forth to crush Li's rebellious forces. Their numerical superiority made itself felt within a few months, and by early August Li had been defeated and captured. The court's troops placed Li in a sack and had him trampled to death by their horses, a method of execution usually reserved for princes. His father-in-law Wang Wen-t'ung was executed shortly thereafter, and Wang's complicity in the rebellion and "treachery" was widely publicized in order to justify his punishment.24 Li T'an's revolt was a turning point in Khubilai's reign, for it made Khubilai increasingly suspicious of the Chinese. A rebellion in an important economic area led by an important Chinese leader with the covert support of a trusted Chinese court official of the highest rank must surely have affected Khubilai. He would, from this time on, naturally hesitate to rely exclusively on his Chinese aides to rule China and would instead seek assistance from non-Chinese advisers. Even before he became the great khan and the emperor of China, Khubilai had recruited advisers from diverse ethnic backgrounds. But Li T'an's rebellion raised even greater doubts about reliance on the Chinese. Khubilai became more keenly aware of his need for non-Chinese advisers and officials. His wife Chabi supported such efforts at governance. She aspired to be the 24 Secondary studies of Li's rebellion include those by Otagi Matsuo, "Ri Dan no hanran to sono seijiteki igi: Moko cho chika ni okeru Kanchi no hokensei to sono shukensei e no tcnkai," TSyoshi ktnkyu, 6 (August-September 1941), pp. 253—78; Sun K'o-k'uan, "Yuan ch'u Li T'an shih pien te fen hsi," Ta lutsachib, 13 (1956), pp. 7—15. Cambridge Histories Online © Cambridge University Press, 2008 KHUBILAI AND ARIGH BORE 427 empress of a powerful state, not simply the wife of a tribal chieftain. 2 ' Her patronage of officials with diverse backgrounds, especially Tibetans, also complemented Khubilai's policies. Yet they both recognized that the bulk of their subjects were Chinese and that some accommodation with Chinese values and institutions was essential. The early administrative system that Khubilai fostered was designed to attract the support of the Chinese and to reflect the concerns of the Mongols. Unlike the preceding Chinese dynasties, however, Khubilai's newly devised government did not institute civil service examinations. These examinations, preparation for which necessitated repeated exposure to and empathy for Confucian doctrines, had since the seventh century provided many of the officials for earlier dynasties in China, and they had been adopted by the Liao and Chin in the north. Khubilai, however, was not anxious to commit himself to a coterie of advisers and officials shaped by a Chinese ideology. Moreover, he wanted the power to appoint his own officials. The institutions he established would, nonetheless, be familiar to his Chinese subjects. The Central Secretariat (Chung-shu sheng), a traditional Chinese governmental agency, took charge of most civilian matters, as it received the reports sent to the throne and drafted the laws. The head of the Central Secretariat (chung-shu ling) consulted with Khubilai on major policy decisions, which would then by implemented by six functional ministries supervised by the prime minister of the left (tso ch'eng-hsiang) and the prime minister of the right (yu ch'eng-hsiang).26 The Privy Council (Shu-mi yuan) was responsible for military affairs, and the Censorate (Yii-shih t'ai) spied on and wrote reports to the emperor about the officials throughout his domains. Although much of this framework of central administration resembled that of earlier Chinese dynasties, the system of local control was different. China was divided into provinces, each of which was administered by a prime minister {ch'eng-hsiang) who was assisted by branch offices (hsing-sheng) of the Secretariat. The emperor also appointed special representatives (darughachi), usually Mongols or Central Asians, to check on the activities of provincial officials as well as those of the local officials of the 180 circuits (/«) into which the provinces were divided. 25 KJ, 114, p. 2871. Francis W. Cleaves, "The biography of the empress Cabi in the Yuan shib," Harvard Ukrainian Studies, 3 - 4 (1979—80), pp. 138—50. 26 The six ministries were (a) Personnel, which oversaw the civilian officials; (b) Revenue, which conducted censuses, collected tax and tribute, and regulated the circulation of money; (c) Rites, which managed the court ceremonies, festivities, music, sacrifices, and entertainments; (d) War, which operated the military commands and colonies and the postal stations, requisitioned military supplies, and trained the army; (e) Justice, which enforced the laws and administered the prisons; and (f) Public Works, which repaired the fortifications, managed the dams and the public lands, and devised the rules for artisans. Cambridge Histories Online © Cambridge University Press, 2008 428 THE REIGN OF KHUBILAI KHAN Khubilai's political system differed from those of earlier Chinese dynasties in significant ways. First, he divided the population into three ethnic groups, with the Mongols occupying the most prominent positions, followed by the so-called se-mu jen, Western and Central Asians. The inhabitants of north China, known as Han jen, at first constituted the lowest-ranked group, though after the conquest of south China, the Chinese of the south, known as the Nan jen, became the lowest group and were excluded from some of the most important civilian positions. Khubilai recognized that the Mongols needed to retain control if they were to avoid being engulfed by the far more populous Chinese (who outnumbered them by at least thirty to one). There was, in general, a much greater emphasis on control than under previous Chinese dynasties. Khubilai was concerned that his officials, many of whom were not Mongols, remain loyal, honest, and incorruptible. Thus "the [Mongol] censorial system . . . was far more pervasive than any preceding one, and its degree of tightly knit centralization was never once had in China's censorial history. "21 Khubilai sought to maintain the officials' loyalty as well as to prevent their abuse of power. Officials who were corrupt or lacked zeal in carrying out their duties or imposed excessive taxes on their subjects were to be severely punished. Khubilai simultaneously needed new regulations to control and dominate the Mongolian leadership. Many Mongolian nobles had been granted appanages (Chinese: fen-ti) since the time of Ogodei, and within their own areas they considered themselves supreme and brooked scant interference. Khubilai had to bring these appanages under the supervision of the central government, insisting that their rulers abide by the laws and regulations devised by his government. Moreover, he anticipated that he, not the appanage holders, would levy taxes and recruit a state army. Recent studies suggest that Khubilai's efforts at control were fruitless. One scholar writes that "with the important exception of the appointment of officials . . . the central government's engagement in empire-wide administration was at best transitory or limited to very restricted activities."38 The Central Secretariat, in this view, functioned effectively only around Khubilai's old domain and his capital; Khubilai's dominance over local affairs was not as pervasive as he wished. Similarly, his control over local officials and appanage holders was limited. Throughout his reign, he issued edicts aimed at corrupt or obstreperous officials, indicating that he was occasionally frustrated in his attempts to impose his own rule. Yet these failures should not be exaggerated, for by the early 1260s Khubilai had established an administra27 Charles Hucker, The censorial system of Ming China (Stanford, Calif., 1966), p. 27. 28 David Farquhar, "Structure and function in the Yuan imperial government," in China under Mongol rule, ed. John D. Langlois, Jr. (Princeton, 1981), p. 51. Cambridge Histories Online © Cambridge University Press, 2008 FOREIGN EXPANSION 429 tton for China that appeared on paper to be workable. It appeared familiar to the Chinese, but it differed sufficiently from previous Chinese systems to accommodate Khubilai's and the Mongols' values and systems and their greater needforcontrolling their subjects. FOREIGN EXPANSION Having established a government in China, Khubilai now turned his attention to foreign relations. Like his Mongolian predecessors, Khubilai knew that he had to persist in territorial expansion. His success as a ruler, in Mongolian eyes, would be measured in part by his ability to incorporate additional wealth, people, and territory into his domain. Similarly, the Chinese believed that a good ruler would induce foreigners to submit and to accept China's supremacy. They would be inexorably attracted to China because of the ruler's virtue and the glory of his state. Both the Mongolian and the Chinese worldviews thus led Khubilai to emphasize expansionism. The manner in which he came to power may also have impelled him to seek foreign conquests. Because he had been challenged by his own brother, there was a real question of his legitimacy as the ruler of the Mongolian world. Khubilai may have tried to quell such doubts by embarking on foreign military campaigns, as additional conquests would bolster his reputation among the Mongols. The conquest of the Sung Khubilai's campaigns against the southern Sung were also prompted by considerations of security. Like any other Chinese dynasty, the Sung aspired to reunify China. Revanchism played an important part in policy debates at the Sung court, and though the Sung military was at that time relatively weak and did not pose any immediate threat to the Mongols, it could be revitalized, and one of its first objectives would be the retrieval of the northern Chinese lands conquered by the Mongols. Khubilai would seek to subjugate the Sung before it could become a more powerful adversary. The Sung's considerable wealth was still another attraction. The land in south China was fertile, a vital consideration for the north, whose population often outstripped its food supplies and which could make good use of grain surpluses from the south. The Sung's seaborne trade with Southeast Asia, India, and the Middle East had enriched its coastal towns, another economic inducement for Khubilai. But the conquest of south China entailed numerous obstacles. Though the Mongolian armies and cavalry had been successful in northern climates and Cambridge Histories Online © Cambridge University Press, 2008 43° THE REIGN OF KHUBILAI KHAN terrain, they were not accustomed to the climate or terrain of the south. They were not prepared for the punishingly high temperatures of the semitropical regions of south China. Neither were they ready for the diseases, the parasites, and the mosquito-infested jungles in the south and southwest. Their horses could not readily adjust to the heat; nor could they forage as easily in the southern farmlands as they could in the steppe. The Mongolian troops, in addition, needed to employ military techniques they had scarcely, if ever, used before. To cope with the Sung navy, for example, they would be required to construct boats, recruit sailors, and become more proficient in sea warfare. On land, they would need to lay siege to populous, well-defended towns and cities. In fact, the Sung had the largest population and the most resources of any of the lands invaded by the Mongols. The subjugation of this great Chinese empire would thus entail enormous expense and effort. The Sung was, on the surface, prosperous. Such lively cities as the capital, Hang-chou, craved and had the resources to pay for luxuries. Hang-chou had splendid restaurants, tea houses, and theaters; "no other town had such a concentration of wealth."2' Southern Sung prosperity derived from both widespread domestic trade and commerce with other countries in Asia and the Middle East. The Sung government, recognizing the potential revenues to be garnered from trade, appointed maritime trade superintendants (t'i-chii shih-po sbih) in the most important ports, employed merchants to supervise the state monopolies and allotted them a higher status in society, and encouraged foreign merchants to trade with China. As the seaborne commerce flourished, the Sung's concern for shipping and, as a result, for naval power grew. The court developed the navy to counter piracy along the coast, and its great ships with their rockets, flamethrowers, and fragmentation bombs became an important branch of the Sung armed forces, posing an obstacle to Mongolian conquest.'0 Despite its commercial prosperity and its naval power, the Sung confronted serious internal political and economic difficulties by the middle of the thirteenth century. Many large landlords had, through good management, oppression of the peasants, or favors from relatives in the bureaucracy, accumulated vast estates and had been granted a tax-exempt status. As more and more land was removed from the tax rolls, the court could not meet its fiscal obligations. Eunuchs and relatives of the empresses played important roles in court deliberations on policy, occasionally overruling high officials. The expendi29 Jacques Gernec, Daily life in China, on the eve of the Mongol invasion, 1250-1276, trans. H. M. Wright (New York, 1962), p. 84. On Hangchow, see also Arthur C. Moule, Quinsai, with other notes on Marco Polo (Cambridge, 1957). 30 Lo Jung-pang, "Maritime commerce and its relation to the Sung navy," Journal of the Economic and Social History of the Orient, 12 (1969), p . 8 1 . Cambridge Histories Online © Cambridge University Press, 2008 FOREIGN EXPANSION 431 tures on the military rose steadily, though corruption and inefficiency hampered its effectiveness. Starting in the early 1260s, Chancellor Chia Ssu-tao attempted to initiate reforms and to curb profiteering eunuchs, imperial relatives, and bureaucrats. He purged some of these men and brought his own allies into prominent positions, thereby polarizing the court and alienating and increasing the size of the opposition. By the time of its confrontation with the Mongols, therefore, the Sung court was bitterly divided. Initially, Khubilai was not belligerent in his dealings with the Sung. In 1260, his envoy Hao Ching had proposed that the Sung acknowledge Khubilai as the Son of Heaven in return for virtual self-rule and an opportunity to enjoy the greater prosperity offered by the Mongols' support of commerce. The Sung detained Hao and did not respond to Khubilai's second embassy of 1261 led by Ts'ui Ming-tao and Li Ch'iian-i. 3 ' Khubilai nonetheless continued to make conciliatory gestures to the Chinese in the south. He released Chinese merchants whom his troops had captured along the border; he was generous in his treatment of defectors from the Sung; and he persisted in offering a peaceful resolution to the hostilities. Because the Sung rejected any diminution of its sovereignty, however, armed conflict was inevitable. Minor clashes occurred from 1260 onward, with a major battle in Szechwan in 1265. The war began in earnest in 1268 and lasted until 1279 (see Map 31). The battle of Hsiang-yang, which was fought from 1268 to 1273, was the longest campaign in the war and proved to be the most critical. Hsiang-yang, located on the banks of the Han River, was a strategic objective of crucial importance, the last stronghold en route to the central basin of the Yangtze. The Chinese had constructed almost impregnable fortifications there, including, as Rashld al-Dln writes, "a strong castle, a stout wall, and a deep moat."' 2 To overcome the defenders' resistance, the Mongolian troops needed naval supremacy on the Han River to prevent the arrival of supplies and reinforcements from the Sung capital. The Mongolian forces also required proficiency in siege warfare and artillery. To provide this expertise, Khubilai selected an international group of commanders, eventually recruiting Mongols, Chinese, Uighurs, and Persians for his army and Koreans and Jurchens for his navy. The siege was begun in the fall of 1268, but a total blockade was not instituted until much later. The Sung thus was able to send supplies and reinforcements to its beleaguered garrison through the first three years of the siege. Meanwhile Khubilai repeatedly authorized reinforcements of his own to place additional pressure on the defenders of Hsiang-yang. From April 31 YS, 5, p. 82. 32 Boyle, The successor! of Genghis khan, p. 290. Cambridge Histories Online © Cambridge University Press, 2008 432 THE REIGN OF KHUBILAI KHAN Hsiang-yang Fellto the Mongolian Army after 6-year siege (1273) »• Right of Sung Emperor, pursued by Yuan forces M A P 3 1 . The conquest of Southern Sung Cambridge Histories Online © Cambridge University Press, 2008 FOREIGN EXPANSION 433 1269 to April 1270, for example, he dispatched about 100,000 men and at least 5,000 boats to his commanders in the area. But Hsiang-yang's defenders remained adamant in their unwillingness to surrender. By early 1272, however, the Sung court encountered more numerous obstacles in breaking the blockade. From that time on, Hsiang-yang was on its own. Yet the Mongolian commanders recognized that they could not storm the castle and the forts without incurring heavy casualties. They could be tied down indefinitely if they chose to avoid such bloodletting. They needed help to break the stalemate. Two Muslim engineers provided the assistance that the Mongols sought. Khubilai's nephew, the Ilkhan Abakha of Persia, responded to the great khan's pleas for help by dispatching Isma'Il and 'Ala al-Dln to China. 33 The two Muslims reached Hsiang-yang late in 1272 and built a mangonel and a catapult capable of hurling huge rocks over a considerable distance. By the end of the year, the Mongolian troops began to use these devices. According to one account, "when the machinery went off the noise shook heaven and earth; everything that [the missile] hit was broken and destroyed. "^ With such strong artillery support, the Mongols could finally storm the town, and the remaining Sung forces, battered by the barrage of rocks and projectiles, could not withstand the enemy troops. By March 1273, the gallant Sung commander Lii Wen-huan surrendered, and a siege of almost five years' duration came to an end. The morale of the Sung court sank, and Chia Ssu-tao was discredited after the fall of Hsiang-yang. Chia tried to bolster his reputation by personally taking charge of the Sung armies resisting the Mongols. He knew that the advancing Mongols would be moving southeastward toward the Sung capital at Hang-chou and decided to make his stand to the northwest, near the city of Yang-chou. With a force of 130,000 men, Chia awaited the enemy. To strengthen his own invasion forces, Khubilai decided to appoint one man as overall leader of the invading Mongols. In the summer of 1273, Khubilai chose Bayan, probably the most gifted military man of his generation, to assume command of his expeditionary forces. Bayan had already distinguished himself in the Mongolian campaigns in Persia and the Middle East under Hiilegu and in the conquest of Ta-li, but he was now granted his most important commission. 3 ' Recognizing the supreme significance of this enterprise, Bayan made elaborate preparations for the expedition. He also encouraged and welcomed Chinese defectors. 33 For their Chinese biographies, see YS 203, pp. 4 5 4 4 - 5 . 34 Moule, Quintal, p. 76. 35 Francis W. Cleaves, "The biography of Bayan of the Barin in the Yiianshih," Harvard Journal of Asiatic Studies, 19(1956), pp. 1 8 5 - 3 0 3 . Cambridge Histories Online © Cambridge University Press, 2008 434 TH E REIGN OF KHUBILAI KHAN Having completed his planning and his training of the troops, Bayan crossed the Yangtze at Han-k'ou in January 1275. Afirecenaval and land battle ensued, but within a short time the Chinese forces were compelled to retreat. In mid-March, Bayan finally met his chief adversary, Chia Ssu-tao, at Ting-chiachou, not far from Yang-chou. The two sides were equally matched except for the Mongols' artillery equipment, which included mangonels and catapults. The artillery spelled the difference; Bayan routed the Sung forces and inflicted heavy casualties. Chia's troops began to desert, and he was forced to regroup and retreat. His political enemies in the Sung capital at Hang-chou now had the opportunity they sought. They stripped him of his offices and had him banished to the southern province of Fukien. En route, his captors killed him. The Sung court was in a state of confusion and chaos. As the Mongols continued to press forward, the Chinese royal family faced other difficulties. The young emperor Tu-tsung died suddenly on 12 August 1274, and his four-year-old son Hsien succeeded him. Hsien's grandmother, the empress dowager Hsieh, acted as regent for her grandson, but she was ill and bereft of good advisers, particularly as more and more influential Chinese defected to the Mongols. Meanwhile, after the battle of Ting-chia-chou, Bayan's troops had besieged Yang-chou and occupied one town after another. In many cases the Sung military and inhabitants simply surrendered. The empress dowager had no choice other than capitulation. At the end of 1275 she sent envoys offering to pay tribute to the Mongols. But Bayan refused these offers, asserting that he would settle for nothing less than unconditional surrender. In late January 1276, the empress dowager finally conceded that the Sung emperor was Khubilai's subject and gave up the dynasty's seal to Bayan.36 Graciously accepting the surrender of the Sung, Bayan escorted the royal family to Khubilai's residence in the north, admonishing his men not to plunder and pillage. Similarly, Khubilai was solicitous of the surrendered royal family. Though he confiscated some of their jewelry and robes, he provided residences, an income, and attendants for the empress dowager and the empress. The young emperor was also accorded the luxuries to which he had grown accustomed, but he was soon exiled to Tibet where he became a serious student of Buddhism and eventually left his palace in 1296 to take monastic vows; in 1323, he was finally forced to commit suicide. Despite the capture of Hang-chou, however, the conquest of the Sung was still not completed. A few Sung loyalists had fled to the south, taking the emperor's two half-brothers with them. On 14 June 1276, they gathered together in Fu-chou to enthrone the older half-brother, a seven-year-old, as Emperor Shih. Faced with such a child emperor, the Sung loyalists needed a 36 YS, 9, p. 176. Cambridge Histories Online © Cambridge University Press, 2008 FOREIGN EXPANSION 435 strong regent to ensure the survival of their cause. Instead, the leading officials differed in their policies, and no single regent was selected. Their lack of unity and constant wrangling weakened the Sung and emboldened the Mongolian-led forces to advance southward as rapidly as possible. A force under a Uighur general, Arigh Khaya, drove southwest through Hunan and modern Kwangsi. By the end of the year, another army under the Mongolian commander So-tu (Sodu) had occupied Fu-chou, compelling the Sung loyalists to retreat farther south to the port of Ch'uan-chou. P'u Shoukeng, the Muslim superintendent of maritime trade at Ch'iian-chou, at first welcomed the fleeing Sung emperor and his entourage, but shortly thereafter disputes flared up, with P'u feeling that the Sung officials were arrogant and overbearing.37 In April 1277, he transferred his loyalty to the Mongols, an important defection for Khubilai, for P'u had a fleet of fine ships at his command. Meanwhile, the Sung loyalists moved south from one port to another throughout the year, starting from Ch'ao-chou and on to Hui-chou and winding up in Kuang-chou at the end of the year. So-tu pursued them constantly, and by February 1278, he had occupied Kuang-chou (modern Canton). The Sung loyalists still did not surrender; they fled yet again. But the pressures, the rugged life, and the continuous changes in climate and environment all took their toll on the child emperor, and on 8 May he died just before his tenth birthday. His death dealt a blow to the Sung loyalists, but their leaders Chang Shihchieh and Lu Hsiu-fu managed to rally them for one last time. They enthroned the deceased Emperor Shih's half-brother Ping and ruled in his name. By this time, they were based in the southeastern extremity of China on the island of Nao-chou off the Lei-chou peninsula. Yet again the Mongols' persistent attacks compelled them to flee, this time to the island of Yai-shan, across the waters from Kuang-chou. The Mongols countered with a blockade of the island. On 19 March 1279, the Sung fleet attempted to break through the blockade, but during the ensuing battle Lu Hsiu-fu, with the child emperor in his arms, drowned. The last Sung emperor had perished at sea, and the Sung dynasty had at last fallen to the Mongols. Three months later Chang Shih-chieh lost his life when a hurricane battered his flotilla. A few Sung loyalists escaped to Champa where they planned to recover their strength and to mount a challenge to Mongolian domination of China. But they were unable to do so. 37 On P'u Shou-keng, see Kuwabara Jitsuzo, "On P'u Shou-keng: A man of the Western Regions who was the superintendent of the Trading Ships' Office in Ch'Uan-chou toward the end of the Sung dynasty, together with a general sketch of trade of the Arabs in China during the T'ang and Sung eras," Memoirs of the Raearth Department of the Toyo bunko, 7 (1935), pp. 1—104; and Maejima Shinji, "Senshu no Perushiyajin to Ho Juko," Shigaku zasshi, 25 (1932), pp. 256—321. Cambridge Histories Online © Cambridge University Press, 2008 436 THE REIGN OF KHUBILA1 KHAN By 1279, Khubilai and the Mongols had crushed the remnants of the Sung dynasty. But Khubilai now faced perhaps a more formidable task, for he needed to gain the allegiance of the Chinese whom he had just subdued. To win their confidence and support, he could not appear to be merely a "barbarian" conqueror interested principally in exploiting the resources of south China. Rather, Khubilai needed to establish a government that served the Mongols and yet did not overly oppress the local people. Continuity in some policies and personnel would also smooth the transition to Mongolian rule. Khubilai thus ordered his military men to permit the Chinese to conduct their economic affairs generally unhindered. He also attempted to recruit Chinese officials for his government, and many talented southerners worked for the Mongols. Some scholars and officials, however, refused to serve the Mongols and devoted themselves to nonpolitical pursuits. A few, such as the renowned scholar-soldier Wen T'ien-hsiang, flaunted their loyalty to the Sung and were either imprisoned or executed by the Mongols.38 Yet no major Sung insurrections are recorded for the remainder of Khubilai's reign, a remarkable indication of his ability to institute Mongolian rule, with as few difficulties as he did, over the most populous country in the world (see Map 32)The subjugation of Korea Khubilai was similarly successful in pacifying Korea. In 1258, his brother Mongke had dispatched an expedition to quell disturbances and to bring Korea under Mongol control. This show of arms led to the submission of Koryd, signaled by the arrival at the Mongolian court of a hostage, Crown Prince Chon. Khubilai and the young Korean got along well, and in the following year when both Mongke and the Korean king died, Khubilai sent Ch6n, accompanied by an escort of Mongolian troops, back to Korea to assume the throne as King Wfinjong. In turn, W6njong quickly demonstrated his loyalty and "virtuousness" by sending the next crown prince as a hostage to Khubilai's court. For the next decade, relations between Korea and the Mongolian dynasty improved steadily. W6njong sent tribute to the Mongols, and Khubilai reciprocated with lavish gifts, permitted Korean merchants to trade in China, and supplied the Koreans with grain and meat in time of economic distress. Khubilai even helped his ally in times of political turbulence. In 1269, a 38 On Wen T'ien-hsiang, see William A. Brown, Wen T'ien-hsiang: A biographical study of a Sung patriot (San Francisco, 1986); on other officials who refused to serve the Mongols, see Frederick W. Mote, "Confucian eremitism in the Yuan period," in The Confucian persuasion, ed. Arthur F. Wright (Stanford, Calif., i960), pp. 202-40. Cambridge Histories Online © Cambridge University Press, 2008 FOREIGN EXPANSION 437 military commander named Im Yon engineered a coup d'etat against W6njong. Within a month after learning of the successful putsch, Khubilai sent a detachment of three thousand troops to oust the forces of the rebel and to reinstate W6njong on the throne. By 1273, the last insurgents had been driven to Cheju Island, off the southern coast of the mainland. To cement his relationship with the Korean royal family, Khubilai had his daughter betrothed to the Korean crown prince, a practice that was continued by the future rulers of the Mongolian dynasty in China. In turn, the Koreans sent annual tribute missions with marmots, otters, silver, falcons, ceramics, and medicines to China until Khubilai's death; they forwarded their population registers to Khubilai's court; and they offered provisions to the Mongolian resident commissioners who were sent to the Korean court. Part of Cheju Island was converted into a grazing area for horses that were offered as tribute or in trade to the Mongols. By the mid-i27os, therefore, Korea had been pacified, and Khubilai had enforced Mongolian military and economic demands on the Koreans.39 The invasion of Japan Perhaps Khubilai's most onerous demand on the Koreans was for assistance in his relations with Japan. Though the Koreans had been plagued by Japanese pirates, known as wako (Korean: waegu), throughout the early thirteenth century, they wished to avoid becoming entangled in Mongolian—Japanese relations. In 1266, for example, they dissuaded an embassy sent by Khubilai to Japan from continuing its journey, by describing the rough seas and turbulent weather around the Japanese islands. Enraged by the Koreans' lack of cooperation, Khubilai sternly reprimanded them, and in 1268 he sent another embassy in which the Koreans did participate. The Japanese shogun and his regent Hojo Masamura, who typified the samurai virtues of pride and patriotism, would not accept the status of vassals of the Mongolian rulers of China. They thus rebuffed the overtures of the embassy and did not even respond to Khubilai's letter, in which he referred to the Japanese emperor as the "king of a little country." Embassies sent in 1271 and 1272 were accorded the same treatment, and both returned to China with descriptions of their discourteous and insulting reception in Japan. Khubilai could not permit the Japanese to defy him indefinitely. In 1274 Khubilai began organizing a punitive expedition to compel the Japanese to accept a tributary status. The mission, composed of 15,000 39 Chong In-ji et al., comps., Koryo-sa (Tokyo, 1908-9), vol. 1, p. 570; YS, 6, p. 122; William E. Henthorn, Korea: The Mongol invasions (Leiden, 1963), pp. 154—60; Louis Hambis, "Notes sur l'histoire de Coree a l'epoque mongole," T'oung Pao, 45 (1957), pp. : 7 9 - 8 3 . Cambridge Histories Online © Cambridge University Press, 2008 Cambridge Histories Online © Cambridge University Press, 2008 438 THE R E I G N OF K H U B I L A I F O R E I G N K H A N 439 E X P A N S I O N C h ' i - w a n g pu Sun-tu-ssu -'"^IVi-wang fu Ch'ien-hu Wai-la Wang f u Kuang-ning Wang pu Ho-ning lu Chi-nan W a n g pu L I N G - P E ...T'ai-ninojuy^ I j K'ai-yuanlu/ L I A O - Ying-ch'ang lu Ning-ch'anglu iHsien-ping _/hia-mi-li Shang-tu lu Y Kuang- Ta-ninglu Te-ningkjy f l-chi-nai lu / V \ ^ "" 9 > : A N G / ,Liao-yanglu J f u l u '•Hsing-ho lu Ta-t'ung lu K A / \ N S y U KORYO ^ \ jr Sha-choulu : Su-chouju/ ; f V W ^^^W±^>^^ / V_ „ •'. Wu-la-hailu '•- ^ A N ^' i Ning-hsialJ^| Kuan-chou JB$& 3 part Chen-ting lu -slgJ 4 part Tung-chang lu m 5 part Te-chou 6 part Ta-ming lu 7 Kao-t'ang chou ^ . / 8 part Kuang-p'mg lu %. 9 Tung-chang lu jjp f f '. S ft (1 * \ •fortress ,1 \Kan-chou lu\ S J/Y / Wu-la-hai / I Under the administration of the HSUAN-CHENG Y U A N i n in iv v 1 Chuang-lang chou 2 Chen-yuan chou 3 P'ing-liang f u 4 Ching-chou part Ho-chienlu port Te-chou part Ts'ao-chou En-chou part Ho-chien lu jJSfijK ^yjp Nan-yang fu H O N A-^N^ ..... -—^rt r^ " v w ~ S 2%E C H W / A N Feng-chouiiT? 1 2 3 4 vCh'ang-te lu • T'ien-linyu Shun-yuan Hsuan-fu ssu Lung-chou.-I Wan-hu fu j \ 1 2 3 4 5 6 / Nan-mien fu Ju-yiianlu Mang-shihlu Lu-cn'uanlu P'ing-mienlu Nan-chien \ . 1 Ch'a-ling chou 1 Chten-ch'ang tu 2 Nan-feng chou 3 part Chten-ch'ang lu 2 Lai-yang chou 3 Ch'ang-ning chou 4 Feng-chou 0 6 M A P 32 . The Yuan empire Cambridge Histories Online © Cambridge University Press, 2008 1£0 l3o 200 200 300 300 miles 400 i o O km Chen-chiang l u Ch"ang-chou l u Chiang-yang lu Ch'ien-shan lu 442 THE REIGN OF KHUBILAI KHAN Mongolian, Chinese, and Jurchen soldiers and 6,000 to 8,ooo Korean troops, guided by 7,000 Korean sailors, departed from Happ'o (near modern Pusan) in Korea and headed for Japan. They took the islands of Tsushima and Iki and then landed at Hakata on the eastern coast of Kyushu. It was not a very large and imposing force, for Khubilai completely underestimated the resistance that the Japanese could muster. Though the Japanese could not match the Mongols' long-range weapons, such as the crossbow and catapaults, and their commanders were not as experienced as the Mongolian leaders were, they had prepared coastal defenses, were fighting on their own lands, and were more familiar with the terrain and the climate. On 19 November the Japanese appeared to be losing the first battle against the Mongols at Hakata, but unexpectedly a gale-force storm struck that night. The Japanese were accustomed to such "unnatural" occurrences and simply took cover in whatever shelter they could find. But the Mongols were terrified and were persuaded by their Korean underlings to return to their ships and head for the open seas until the storm subsided. The results were disastrous: The winds, waves, and rocks shattered a few hundred ships, and about thirteen thousand lives were lost.40 The expedition thus ended in disaster for the Mongols, and the remaining forces sailed for home and reported on the fiasco to Khubilai. Because he was attempting to suppress once and for all the Southern Sung, he could not immediately avenge himself on the Japanese. Instead, in 1275, he sent still another embassy, whose envoys were promptly executed by the cocky Japanese authorities. Although Khubilai could not allow such an outrageous act to go unpunished, some years would elapse before he could send a punitive force against the Japanese. Khubilai and Central Asia Meanwhile, in the Chaghadhai khanate in Central Asia, Khubilai confronted a foe who wished to wrest control from him. Unlike his relations with Korea and Japan, which posed no real threats to his own position, his relations with Central Asia involved a hostile challenger who could lay claim to the Mongolian throne. His principal antagonist, Khaidu, was not only a Mongol but also a member of the royal family, a grandson of the great khan Ogodei. Because Central Asia shared a common border with Khubilai's domains, hostile relations would subject his frontier lands in northwest China to hitand-run attacks, the kinds of raids that the nomads of Inner Asia had perenni40 George Sansom, A history of Japan to 1334 (Stanford, Calif., 1958), pp. 440-4; Ryusaku Tsunoda and L. C. Goodrich, Japan in the Chinese dynastic histories (South Pasadena, Calif., 1951), pp. 74—6. See also Kawazoe Shoji, "Japan and East Asia," in MedievalJapan, ed. Kozo Yamamura, vol. 3 of The Cambridge history of Japan (Cambridge, 1990), pp. 396-446. Cambridge Histories Online © Cambridge University Press, 2008 FOREIGN EXPANSION 443 ally inflicted on Chinese peasants. After these raids, the nomads would simply flee into the vast open spaces of the Central Asian steppes or deserts and evade the pursuing armies of the sedentary peoples. Such attacks imperiled the long-distance caravan trade across Eurasia that Khubilai sought to foster and that required secure towns and oases in Central Asia. If hostile groups controlled these essential halting places, they could disrupt this commerce. Khubilai's cousin Khaidu represented the Mongolian nomadic values that threatened the increasingly sedentary Mongolian dynasty in China. Khaidu enjoyed the life of the nomad and lord of pastoral herdsmen, not the life of the governor and ruler of peasants. His habitat was the open spaces, not a grand palace in a populous capital city. He favored the pastoral nomadic society rather than the sedentary agricultural society ruled by a central government and staffed by a bureaucracy. The Chinese sources unfairly portray him as a plunderer and a treacherous renegade. But he did not intend either to destroy the flourishing towns in the region or to dismantle their commercial bases. In fact, Khaidu actively discouraged the pillaging of the Central Asian oases and must have instructed his underlings not to harass their inhabitants. Instead, he levied taxes on the towns, and the revenues he derived were used to support his forces. Nonetheless, he appeared to be the defender of the Mongolian heritage that, in his view, Khubilai had betrayed.41 The precise date of the onset of hostilities between Khubilai and Khaidu is difficult to establish. As early as 9 July 1266, Khubilai had appointed his son Nomukhan as pet p'ing wang (prince of the pacification of the north), intending that the young man take charge of the military affairs of north China and prevent incursions onto the soil of northwest China. Yet five years elapsed before Khubilai actually assigned his son to the Central Asian outpost of Almalikh (modern Huo-ch'eng, Sinkiang), to protect that region from Khaidu's incursions. Khubilai also erred grievously in sending several cousins of Nomukhan's to accompany him, for they all engaged in bitter disputes that impeded the expedition and finally led to its failure. Nomukhan made little progress in crushing the dissidents in Central Asia. He succeeded in developing supply lines for his troops, but he could not readily engage the enemy. Khaidu's forces organized guerrilla-like operations against his troops but would not fight traditional battles. Whenever they found themselves outnumbered or in distress, they simply fled into the 41 The hostile Chinese views of Khaidu may be found in K'o Shao-min, Hsin Yuan sbib (Tientsin, 1922; author's and revised ed., Peking, 1930); repr. in Erhshih wushih, ed. Erh shin wu shin k'an hsing wei yuan hui (Shanghai, 1935; repr. Taipei, 1962-9; and other recent facsimile reprints), p. 6850; T'u Chi, Meng-um-trh sbib chi (Taipei, 1962), 74, pp. 6a~7b; Shao Yiian-p'ing, comp., Yuan sbib Itipitn (1699; repr. Taipei, 1962, 1968), pp. 1595-7. Cambridge Histories Online © Cambridge University Press, 2008 444 THE REIGN OF KHUBILAI KHAN steppelands or desert, terrain with which they were familiar. Nomukhan's forces were frustrated because they could not easily pursue the highly mobile guerrilla forces and engage them in combat. To break the deadlock, in 1275 Khubilai sent his wife's nephew An-t'ung (1245-93), a capable and prominent figure who was at that time right prime minister (yu Ch'eng-hsiang), to assist Nomukhan. On arriving at Nomukhan's encampment, An-t'ung quickly recognized that factionalism divided the various princes on the expedition and was hampering effective operations. By siding with Nomukhan, however, he too became embroiled in these disputes. In late 1276, the princes who accompanied Nomukhan plotted to sabotage his expedition. The conspirators, who included two of Arigh Boke's sons and one of Mongke's sons, seized Nomukhan and delivered him to the khan of the Golden Horde of Russia, and they turned over An-t'ung to Khaidu.42 Their captors detained the two men for almost a decade but did not harm them. The conspirators were disappointed, however, to discover that Khaidu equivocated on an alliance with them; he did not want them in his domains. Shortly thereafter, they migrated to what they perceived to be a safer location, the Mongolian steppe. Eventually, both the khan of the Golden Horde and Khaidu, unable to raise a ransom for the two captives and seeing no gain from continuing to detain them, released Nomukhan and An-t'ung. Khubilai warmly greeted his son and his nephew when they returned in 1284 and once again granted them their old positions and titles. Khubilai had not been idle during the decade that elapsed between the seizure and release of Nomukhan and An-t'ung. On learning of the capture of Nomukhan, he had dispatched Bayan, his ablest and most renowned general, to retrieve his son. Bayan, who had just returned from his extremely successful campaigns against the Southern Sung dynasty, was repeatedly frustrated in his efforts. Like Nomukhan himself, Bayan was unable to engage the enemy, as Khaidu's forces continually eluded his troops. Nomukhan's leaderless troops also made some valiant attempts to free Khubilai's son, but they too failed to secure his release. Khubilai eventually acknowledged that he could not control Central Asia and was compelled to accept Khaidu as the de facto ruler of the area. Even his most prominent general had been unable to extend Khubilai's suzerainty into Central Asia. Admitting his failure, he reluctantly relinquished his position in the steppelands and oases of the region. He retreated to the more defensible confines of Chinese settlement, allowing Khaidu free rein beyond these villages. Yet he could not prevent Khaidu from raiding these villages, which had been his primary objective. The difficulties he encountered were that the 42 Boyle, The successors of Genghis khan, p. 266. Cambridge Histories Online © Cambridge University Press, 2008 SOCIAL AND ECONOMIC POLICIES 445 supply lines required to maintain his armies and the friendly local inhabitants were long and fragile; that the constant harassment and elusiveness of the nomads irritated and intimidated his soldiers and allies alike; and that the self-sufficiency he sought to achieve for the oases and towns in the region was never realized. In short, Khubilai's foray into Central Asia had achieved nothing. He was more successful in resisting challenges in Mongolia. The conspirators who had betrayed his son Nomukhan had moved to Mongolia and planned an attack against the traditional Mongolian capital at Khara Khorum. Khubilai and his government were much better prepared to cope with such a threat in Mongolia than they were farther west. They had, for some time, encouraged agriculture in the regions around Khara Khorum, by remitting taxes and disseminating knowledge of more advanced farming techniques to the local people.43 They had also founded postal stations to provide a strategic and commercial link with the new Mongolian capital in north China and had sent craftsmen to help the natives develop their own artisan class. Such measures had gained for them the support of the local people, who thus cooperated in the campaigns against the rebellious princes. Early in 1279, Khubilai's troops launched an expedition to crush the dissidents. With the help of the local people, they defeated and captured the princes within a few months. By the end of the year, Mongolia had been preserved as part of Khubilai's domain. SOCIAL AND ECONOMIC POLICIES While engaged in the conquests of southern China and Korea and the pacification of Inner Asia and Mongolia, Khubilai was not unmindful of the difficulties he faced in north China. The China that he sought to rule in 1260 confronted serious problems that he needed to address. It had not truly recovered from the destruction caused by the Mongolian conflicts with the Chin dynasty from 1211 to 1234. Its peasants were uncertain about the intentions of their Mongolian rulers. They needed to be reassured that their lands would neither be confiscated nor would arbitrary and exorbitant taxes be imposed on them. But the succession struggle with Arigh Boke and the conflict with the Southern Sung hampered commerce, and the lack of an accepted legal code caused great confusion. The system of education was in disarray because the civil service examinations, the focus of formal education in China, had been suspended. The formal religions could not gauge 43 John Dardess, "From Mongol empire to Yuan dynasty: Changing forms of imperial rule in Mongolia and Central Asia," lAonumtnta Serica, 30 (1972—3), pp. 143—60. Cambridge Histories Online © Cambridge University Press, 2008 446 THE REIGN OF KHUBILAI KHAN Khubilai's and the Mongols' policies toward them. The Buddhist hierarchy recognized that Khubilai was sympathetic, but the Taoists did not know whether the new government woud discriminate against them. The Confucians worried that the Mongols would abandon the traditional court rituals and would downgrade the status of scholars. Perhaps most important, Khubilai's Chinese subjects were concerned about their position. They would surely be excluded from certain high offices, but would they be discriminated against in other ways? Khubilai needed to respond to such questions and to deal with these problems if he wished to create order in Chinese society. The government agencies that he had formed were valuable mechanisms, but they required direction. Khubilai had to articulate the political, social, and economic policies that he meant them to implement. He had to disclose his plans for ruling rather than merely exploiting China, and his officials might then emulate and assist him in seeking to govern the sedentary civilization. Some scholars have wondered about Khubilai's own involvement in the decisions that charted the course for the early years of his reign. Did he take an active role in devising the policies and programs pursued during this time? Or did his Chinese advisers merely present him with proposals that he, with little reflection, then adopted? To be sure, Khubilai did not originate many of the policies that were subsequently implemented. Yet he did not sit back and simply await proposals. He actively solicited suggestions. One of his officials quotes him as saying: "Those who present memorials to make proposals may present them with envelopes sealed. . . . If the proposals are useful, the Court will liberally promote and reward the persons who make the proposals." 44 Wang Yiin, a contemporary observer, tells us that Khubilai participated in the deliberations at court. In the space of a week in May 1261, for example, Wang had three audiences with Khubilai to discuss governmental affairs. Before embarking on a planned, orderly program to define and clarify his social, political, and economic ideals, Khubilai was required to relieve the misery of the people in north China. The wars preceding the Mongolian takeover caused much destruction and enormous loss of life in north China. 4 ' Such devastation clearly entailed great suffering for the survivors. In the first few years of his reign, Khubilai repeatedly responded to appeals for assistance and tax exemption from numerous regions in his domains. The Chinese 44 Lao Yan-shuan, "The Chung-t'ang shih-chi of Wang Yiin: An annotated translation with an introduction" (Ph.D. diss., Harvard University, 1962), p. 34. 45 Ho Ping-ti, Studies on the population of China, 1368-1953 (Cambridge, Mass., 1959), p. 258; see John Langlois's brief remarks on this demographic problem in the introduction to China under Mongol rule, p. 20. Cambridge Histories Online © Cambridge University Press, 2008 SOCIAL AND ECONOMIC POLICIES 447 sources record that he often waived or reduced taxes on areas that were experiencing economic difficulties. He also provided paper money, grain, and clothing to villages afflicted by natural disasters.46 Together with such immediate measures, however, he also needed to develop a long-range program for China's economic recovery. One of the pivots of such a program was the encouragement of agriculture. In 1261, Khubilai founded the Office for the Stimulation of Agriculture (Ch'iian-nung ssu) which, in turn, selected men knowledgeable in agronomy to help the peasants make the best use of their land. The office submitted to the central government annual reports on agriculture, sericulture, and water control projects. Khubilai eventually organized a large bureaucracy to promote the more efficient and productive use of the land. He ordered the construction of granaries to store surplus grain as insurance against shortages of food in bad times. His concern for his sedentary subjects was demonstrated in an edict of 1262 that prohibited the nomads' animals from roaming in the farmlands. He did not want his own Mongolian people to encroach on and perhaps cause additional damage to the valuable territory of the peasants.47 Khubilai also sought to help the peasants organize themselves for economic recovery. In 1270, he gave official standing to the organizations known as she, composed of about fifty households under the direction ofashechang, or village leader, to stimulate agricultural production and to promote reclamation. Khubilai gave the she the mandate of properly farming, planting trees, opening up barren areas, improving flood control and irrigation, increasing silk production, and stocking the lakes and rivers with fish. Khubilai and his advisers conceived of the she as self-help organizations for the peasants, but they also intended to graft other functions onto them. They hoped to use the she to restore stability in the countryside and to aid in surveillance and in the conduct of censuses.48 Perhaps the government's most innovative objective was to employ the new organization to promote universal education. Each she was entrusted with the task of setting up schools for the children in the villages. The peasant children would attend the schools when little labor was required on the farms. The chronicles of the period make grandiose claims for this educational system. By 1286, there were, according to the dynastic history, 20,166 she schools. But this figure seems inflated, for the leaders of the she, recognizing what was expected of them, probably exaggerated their reports 46 For some of these relief efforts, see YS, 4, pp. 70—1; 5, pp. 83—6; 6, pp. 113—14. 47 Ta Yuan Is'ang k'u chi (Peking, 1936), pp. 1—3; Paul Ratchnevsky, Un code da Yuan (Paris, 1937), vol. 1, pp. 189—90. 48 Inosaki Takaoki, "Gendai shasei noseijiteki kosatsu," Toyoshi kenkyu, 15 (July 1956), pp. 6—10; Yang Ne, "Yuan tai nung ts'un she chih yen chiu," Li shih yen chiu, 4 (1965)4, pp. 117—34. Cambridge Histories Online © Cambridge University Press, 2008 448 THE REIGN OF KHUBILAI KHAN to the central government, conveying an image of a growing number of schools. The vision of a universal education system was certainly never fulfilled; indeed, there is little evidence that she were ever organized throughout China.49 Yet this vision reveals that Khubilai and his advisers embraced the concept of a literate peasantry whose interests the government would serve. No longer would the Mongolian rulers of China concern themselves exclusively with the nomads. The peasants also would receive a fair share of their attention. An even clearer indication was the government's efforts to limit the financial burdens on the peasants. Khubilai devised a fixed, regular system of taxation, dispensed with tax farming, and reduced the power of the appanages. Under the new system, the previously unpredictable payments that the peasants had been compelled to make to the appanages would now be remitted to the government and then divided equally between the appanages and the central government. The peasants would pay annual grain taxes and would not need to be concerned about capricious levies imposed by the appanage holders. They would also be liable to a head tax, as were other segments of the population, including craftsmen and clerics. Their other principal burden was corvee obligations, which could be as onerous as taxes. Khubilai built roads and a capital city, extended the Grand Canal, and organized a postal relay system, all of which required vast investments of labor. Yet he sought, throughout his reign, to limit excessive demands on the peasants and, on occasion, waived other taxes on those called on for corvee. He could not control all his officials, however, and some labor demands on the peasants were unreasonable. It seems clear though that Khubilai, unlike many traditional Mongols, did not seek merely to exploit the Chinese peasants.50 Like his forebears, Khubilai cherished artisans. Unlike the traditional Chinese, he accorded craftsmen a high status. Because the Mongols had few craftsmen of their own, they relied on foreigners for the craft articles they required. Khubilai thereby set up a number of offices in his government both to organize and to safeguard the welfare of artisans. His household agencies, for example, included the Bureau for Imperial Manufactures (Chiang-tso yuan), which was charged with providing jewelry, clothing, and textiles for the court. To win the artisans' allegiance, Khubilai enacted favorable regulations. The government offered them fairly good salaries, rations of food and clothing, and exemptions from corvee labor. It also permitted them to pro49 See Herbert F. Schurmann, Economic structure of the Yuan dynasty: Translation of chapters g$ and 94 of the Yuan shih, Harvard-Yenching Institute Studies, vol. 16 (Cambridge, Mass., 1956; repr. 1967), p. 4750 Aritaka Iwao, "Gendai no nomin seikatsu ni tsuite," Kuwaiara Hakushi ianreii kinen Toyoshi ronso, ed. Kuwabara Hakushi kanreki kinen shukugakai (Kyoto, 1934), pp. 951— !• Cambridge Histories Online © Cambridge University Press, 2008 SOCIAL AND ECONOMIC POLICIES 449 duce articles for sale as long as they fulfilled their annual quotas for the court. In return for these privileges, however, the court mandated a hereditary artisan class. By the late thirteenth century, about 300,000 families were classified as artisan households and could not detach themselve from this designation. Despite this restriction, craftsmen, in general, profited from Mongolian rule in China.' 1 Merchants were perhaps the group that benefited the most from Khubilai's policies. Chinese dynasties had imposed numerous restrictions on merchants because the Confucian scholar-officials disapproved of trade. Khubilai did not share this bias, however, and in fact accorded merchants high status. Ortogh, or merchant associations composed primarily of Muslims, received government support. During the early years of the Mongolian conquests, the ortogh had provided badly needed loans to the Mongolian nobility. In 1268, Khubilai, as a reward, created the "General Administration for the Supervision of the Ortogh" to lend money at low interest to the ortogh. The ortogh used this money principally to finance trade caravans. Taxes on the commercial transactions of the ortogh and of Chinese merchants amounted to the low figure of 3'/ } percent.' 2 To facilitate trade and to promote the welfare of the merchants, Khubilai instituted the use of paper currency on a wider scale than at any other earlier period in Chinese history. In 1260, Khubilai issued three types of currency, but the Chung-t'ung yiian-pao-ch'ao, backed by a silver reserve, prevailed over the others and gained the confidence of the Chinese. The court's willingness to accept taxes in paper money instilled trust in the new currency. The stability of the currency helped foster trade and thus promoted the interests of the merchants. The system operated well until 1276, for the governmennt kept tight controls on the amount of paper notes printed. In 1260, the court printed paper money to the face value of 73,352 ting (silver ingots), and by 1265, the amount had gradually increased to 116,208 ting.™ In 1276, the court, faced with staggering expenses for the military campaigns in southern China and Japan, dramatically expanded the sum it printed to 1,419,665 ting. The revenues derived from the successful conquest of southern China nevertheless permitted the court to keep inflation under control. Khubilai's government also assisted trade and merchants by improvements in the system of transport. The extension of the Grand Canal to the Mongo51 Chii Ch'ing-yiian, "Yuan tai hsi kuan chiang hu yen chiu: Chih jen wei Yuan tai kuan chii chiang hu shih nu li te jen men," Shih huoyueh k'an, i (1935), pp. 367—407; compare the abridged translation by John de Francis and E-tu Zen Sun, "Government artisans of the Yuan dynasty," in Chinese social history: Translations of selected studies (Washington, D.C., 1956), pp. 2 3 4 - 4 6 . 52 Morris Rossabi, "The Muslims in the early Yuan dynasty," in China under Mongol rule, ed. John D. Langlois, Jr. (Princeton, 1981), pp. 282—3. 53 YS. 93. PP- 2 3 7 1 - 2 . Cambridge Histories Online © Cambridge University Press, 2008 45O THE REIGN OF KHUBILAI KHAN lian capital in north China and the construction of roads were two of these notable achievements. Marco Polo was greatly impressed by these roads and wrote: He [Khubilai} has had trees planted there beside the ways on either side two or three paces distant the one from the other . . . the great khan has had this done so that each may see the roads, that the merchants may be able to rest there in the shade, and that they may not lose the way.54 The most remarkable improvement in transport involved the postal relay system. China had had postal stations and relays at least since the Han dynasty, but the Mongolian rulers vastly extended the system. The postal stations were designed for the transmission and delivery of official mail, but they were also available to traveling officials, military men, and foreign state guests, aided in the transport of foreign and domestic tribute, and facilitated trade. They were not intended as hostels for merchants, but they came to be used as such and were vital links in the networks of foreign and domestic commerce. By the end of Khubilai's reign, China had more than 1,400 postal stations, which in turn had at their disposal about 50,000 horses, 1,400 oxen, 6,700 mules, 4,000 carts, almost 6,000 boats, over 200 dogs, and 1,150 sheep.55 The individual stations were anywhere from fifteen to forty miles apart, and the attendants worked in the stations as part of their corvee obligations. In an emergency, the rider-messengers could cover up to 250 miles a day to deliver significant news, a remarkably efficient mail service for the thirteenth, or any other, century. Despite abuses by officials, merchants, and attendants, the postal system operated efficiently, a fact to which numerous foreign travelers, including Marco Polo, attested.'6 All of these developments demonstrated Khubilai's concern for merchants and his efforts to encourage trade. Unlike many Chinese, Khubilai and his Mongolian supporters were not biased against traders. The evidence from contemporary observers indicates that during Khubilai's reign, the Mongols did indeed succeed: Merchants did prosper. Marco Polo, for example, writes: "I believe there is not a place in the world to which so many merchants come and that dearer things and of greater value and more strange come into this town . . . than into any city of the world."57 Other occupational groups fared better under Khubilai than under the Chinese emperors. Physicians were one such group that benefited from Mon54 Moule and Peiliot, Marco Polo, vol. I, pp. 2 4 8 - 9 . 55 Yuan Chi (Kuo-fan), Yuan shihym chiu lun chi (Taipei, 1974), p. 243. For a comprehensive study of the postal relay system, see Peter Olbricht, Dai Postwesen in China unter der Mongolenhenschaft im 13, und 14. Jahrbundert (Wiesbaden, 1934). 56 Moule and Peiliot, Marco Polo, vol. 1, p. 246. 57 Ibid., pp. 2 3 5 - 7 . Cambridge Histories Online © Cambridge University Press, 2008 SOCIAL AND ECONOMIC POLICIES 451 golian rule. As a pragmatist, Khubilai valued medicine and accorded doctors a higher social status. He established branches of the Huang-hui ssu (Imperial Hospitals), staffed primarily by Muslim doctors, in K'ai-p'ing and in north China to care for the court. Mongolian officials consulted Muslim physicians, and thirty-six volumes of Muslim medicinal prescriptions were added to the Imperial Library. Khubilai also set up an Imperial Academy of Medicine (T'ai-i yuan) which specified the criteria for selecting instructors of medicine, supervised the training of physicians and the preparation of medical texts, devised the examinations to certify doctors, and took charge of all physicians and pharmacies. It weeded out unfit doctors as well as ensuring that those candidates who failed the examinations would not practice. As Joseph Needham points out, "There was a general move . . . to raise the intellectual standing of the physicians."' 8 In this effort, the court founded four major medical schools in the capital, all of which were influenced by Persian medical practices. 59 These efforts bore fruit, for medicine attracted members of the Chinese elite in greater numbers than earlier dynasties did. The court did not disdain the medical profession, and the elite began to regard it as a useful and lucrative occupation that offered access to influence through the doctor's patients and conformed to the Confucian emphases on righteousness and altruism. Physicians often received exemptions from corvee labor, another attraction for those considering a medical career. Khubilai also valued scientists and sought to promote their work. He offered them financial support and attempted to elevate their status in Chinese society. Learning that the Persians had made sufficient scientific discoveries, he invited the astronomer Jamal al-Dln to China to expound these innovations. Arriving at the court in 1267, Jamal al-Dln brought along sundials, an astrolabe, a terrestrial globe, a celestial globe, and a new and more accurate calendar, known to the Chinese as the Wan nien li. Four years later, Khubilai founded the Institute of Muslim Astronomy (Hui-hui ssu t'ien chien) to recruit and attract Persian and Arab astronomers to his court. Later the Chinese astronomer Kuo Shou-ching (1231-1316) used the Persian diagrams and calculations to devise another calendar, the Shou shih li, which, with minor revisions, was employed through the Ming dynasty.60 Under Khubilai's patronage, geographers and cartographers prospered. Arab and 58 Joseph Needham, "Medicine and Chinese culture," in his Clerks and craftsmen in China and the West (Cambridge, 1970), p. 263. 59 On these schools, see Jutta Rail, Die vier grossen Medizinschulen der Mongolenzeit: Stand and Entwicklung der chinesischen Medezin in der Chin- und Yiian-Zeit (Wiesbaden, 1970). 60 E. S. Kennedy, "The exact sciences in Iran under the Saljuqs and Mongols," in The Saljuq and Mongol periods, ed. John A. Boyle, vol. 5 of The Cambridge history ofIran (Cambridge, 1968), pp. 6 6 8 - 7 3 ; ^ > 7, PP- 136; 164. PP- Cambridge Histories Online © Cambridge University Press, 2008 452 THE REIGN OF KHUBILAI KHAN Persian travelers and traders brought information about Asia and Europe so that "geography in China [flourished], incorporating data on the non-Chinese world taken from Arab sources."6' Clergy and artists were two other groups favored by Khubilai, and his attitudes toward them will be more clearly revealed in later discussions of religion and art. In sum, Khubilai sought to dispense with the discriminatory practices against occupational groups that had not fared well under the Chinese dynasties. Merchants, physicians, and scientists were accorded more benefits and attracted more court concern, as Khubilai clearly hoped to gain their support in ruling China. He had also pledged not to exploit the peasantry and had, in fact, stimulated improvements in agriculture. The principal group whose interests were undercut by the Mongols was the landed elite from whose ranks derived the bulk of the governing scholar-official class. Khubilai and the Mongols had displaced them as the rulers of the country. Without the civil service examinations, the Chinese elite had few options. Some acquiesced and served the Mongols; others abandoned public life and became recluses or dabbled in the arts; and still others, disgruntled with Mongolian rule, formed a potentially disruptive force. The Chinese elite was the major group that felt excluded from the benefits accorded to other classes and professions by the Mongols. Yet Khubilai and his court tried to appease them by retaining certain government offices, the Han-lin Academy, the Directorate for the Diffusion of Confucian Texts, the Archives, and the Imperial Diarists' Office, in which scholar-officials remained prominent. The military was still another group whose relations with the court needed definition. Khubilai's main concern was that Mongolian control of the military not be jeopardized, and the organization and regulations he developed reflected that objective. In 1263, he reestablished the Privy Council (Shu-mi yuan) to supervise the Imperial Bodyguards, the kesig (the traditional bodyguards for the Mongolian khans), and the myriarchs (those who commanded ten thousand troops). These units were composed mainly of Mongolian cavalry, and the infantry consisted principally of ethnic Chinese. All Mongolian adult males were liable for conscription, and certain Chinese families were designated as hereditary military households. They were exempted from half the taxes imposed on ordinary citizens but, in return, needed to supply themselves, occasionally a heavy financial burden.62 These burdens, together with graft, corruption, and the extortion of funds by officers from their men, eventually led to desertions and a decline in the armed forces, but these problems did not become troublesome until after Khubilai's death. 61 Herbert Franke, "Sino-Western contacts under the Mongol empire," Journal ofthe Royal Asiatic Society: Hong Kong Branch, 6 (1966), p . 59. 62 Hsiao Ch'i-ch'ing, The military establishment of the Yiian dynasty (Cambridge, 1978), pp. 16—25. Cambridge Histories Online © Cambridge University Press, 2008 SOCIAL AND ECONOMIC POLICIES 453 Much more worrisome was the recruitment of Chinese forces. Khubilai could not afford to rely excessively on the Chinese and needed to balance them with Mongolian troops. Accordingly, as bodyguards for himself and for the court, he employed the Mongol kesig. Similarly, in stationing garrisons along the borders, he sensed the need to maintain Mongolian dominance among those troops. Khubilai recognized, too, that Mongolian control of military supplies and equipment was essential. The court, for example, prohibited the Chinese from buying and selling bamboo, which could be used for bows and arrows; bamboo was monopolized by the court.63 Khubilai also sought to guarantee for the court a dependable supply of horses suitable for warfare. As the Mongols in China began to settle in the sedentary world, they faced the same problems as did the Chinese in acquiring horses. To provide the government with the horses it required, Khubilai ordered that one out of every hundred horses owned by Chinese subjects be turned over to the court. He also reserved the right to purchase horses, compelling the owners to sell their animals at official prices. Chinese families who attempted to conceal their steeds or who sold them privately would be severely punished. The government agency known as the Court of the Imperial Stud cared for its horses and managed its pasturelands, which were concentrated in Mongolia, north and northwest China, and Korea. Though the sources occasionally refer to horse smuggling and other abuses, the court, during Khubilai's reign, had access to a sufficient number of horses.64 Another of the court's concerns was the creation of a legal code for its domain. The jasagh, the traditional Mongolian legal regulations, lacked the sophistication necessary to rule a sedentary civilization; rather, it reflected the values of a nomadic society and was unsuited to China. On coming to power, Khubilai retained the law code of the Jurchen Chin dynasty, but by 1262 he ordered Yao Shu and Shih T'ien-tse, two of his most trusted and influential advisers, to devise a new code that was more suitable for his Chinese subjects. These laws began to be implemented in 1271, but Mongolian laws, practices, and customs affected this new code. The Mongols apparently introduced greater leniency into the Chinese legal system. The number of capital crimes amounted to 135, less than one-half the number mandated in the Sung dynasty codes. Criminals could, following Mongolian practice, avoid punishment by paying a sum to the government. Khubilai could grant amnesties, and he did so, even to rebels or political 63 Inosaki Takaoki, "Gendai no take no sembaiken to sono shiko sum igi," Toyoshi kmkyu, 16 (September '957). PP- 2 9 - 4 7 64 Ta Yuan ma cheng chi (Peking, 1937), pp. 1—3; C. R. Bawden and S. Jagchid, "Some notes on the horse policy of the Yiian dynasty," Central Asiatic Journal, 10 (1965), pp. 2 6 1 - 3 . Cambridge Histories Online © Cambridge University Press, 2008 454 T H E REIGN OF KHUBILAI KHAN enemies. Officials of the provincial or central government routinely reviewed local judicial decisions on serious crimes in order to prevent abuses of the rights of the accused. Because there have not been any careful studies of this code in operation, it is difficult to tell whether these statutory reforms translated into a more lenient and flexible system than under the earlier Chinese dynasties. Yet the legal ideals embodied in this code supported by Khubilai and the Mongols did indeed appear less harsh than earlier Chinese ones.6' KHUBILAI AS EMPEROR OF CHINA Though Khubilai wished to be considered as more than the emperor of China, he was unable to coerce the other khans into accepting his authority. As the khaghan of the Mongols, he aspired to universal rule and sought recognition of his status as the undisputed ruler of all the Mongolian domains. The Golden Horde in Russia had supported Arigh Boke's candidacy as the khaghan and were not reconciled to Khubilai's victory. Khaidu, who controlled the Chaghadai khanate of Central Asia, was an implacable foe of Khubilai's. Only Khubilai's brother Hiilegii, the founder of the Ilkhanate of Persia, and his descendants accepted Khubilai as khaghan, but they were essentially self-governing. The Golden Horde and the Ilkhans were entangled in their own conflict over their claims to the pasturelands of Azerbaijan, diverting attention from their relationships with the khaghan. With such limited acceptance of his position as khaghan, Khubilai increasingly became identified with China and sought support as emperor of China. In order to attract the allegiance of the Chinese, he needed to portray himself as and to act like a traditional Chinese emperor. He would have to reinstate some of the Confucian rituals and practices if he hoped to attract the support or at least the acquiescence of the Chinese scholar-officials or the elite. Khubilai remained a Mongol and would not abandon Mongolian values, yet he recognized that he had to make some adjustments to garner such support. Khubilai's clearest signal to his Chinese subjects was his shift of the capital from Mongolia to north China. With the assistance of his adviser Liu Pingchung, he conceived of the idea of moving the capital from Khara Khorum to the modern city of Peking. In 1266, he ordered the construction of a city that came to be known as Ta-tu (great capital) to the Chinese and as Khanbalikh (city of the khans) to the Turks. The Mongols called it Daidu, a transliteration directly from the Chinese. Though a Muslim supervised the 65 Paul Heng-chao Ch'en, Chinese legal tradition under the Mongols: The code of 1291 as reconstructed (Princeton, 1979), p. xix, argues generally that the Yuan code was indeed more lenient and flexible than were earlier Chinese codes. Cambridge Histories Online © Cambridge University Press, 2008 KHUBILAI AS EMPEROR OF CHINA 455 whole project and numerous foreign craftsmen took part in the construction, the city was Chinese in conception and style. The planners followed Chinese models, as Khubilai wanted Ta-tu to serve as a symbol of his efforts to appeal to the traditional Chinese scholars and Confucians. He chose, however, to build the capital on an unconventional site. Unlike earlier Chinese capitals that were, for the most part, situated near the Yellow River or one of its tributaries, Ta-tu was located close to China's northern border (see Map 33). Khubilai selected this site, which had been the site of the Liao and Chin capitals, partly because he perceived that his domains included more than just China and partly because he wished to retain control over his homeland in Mongolia. An administrative center in north China would offer him a listening post and a base from which to assert his authority over his native land. Ta-tu's major deficiency was its inadequate reserves of grain. To cope with this shortage, Khubilai imported vast quantities of food from south China and eventually lengthened the Grand Canal to reach all the way to the capital. The Muslim architect Yeh-hei-tieh-erh and his associates constructed Tatu as a typical Chinese capital, albeit with some Mongolian touches. The city was rectangular and enclosed by a wall of rammed earth. Within this outer wall were two inner walls surrounding the Imperial City and Khubilai's residences and palaces, to which ordinary citizens were denied entry. The city was laid out on symmetrical north—south and east—west axes, with wide avenues stretching in geometric patterns from the eleven gates that permitted access into the city. The avenues were broad enough so that "horsemen can gallop nine abreast." On all the gates were three-story towers that served to warn of impending threats or dangers to the city.66 All of the buildings in the Imperial City, the khan's own quarters and those of his consorts and concubines and the hall for receiving foreign envoys, as well as the lakes, gardens, and bridges, were remarkably similar to those in a typical Chinese capital. Yet Mongolian decor was evident in some of the buildings. In Khubilai's sleeping chambers hung curtains and screens of ermine skins, a tangible reminder of the Mongols' hunting life-style. Mongolian-style tents were set 66 Two fourteenth-century sources, the Nan ts'un ch'o keng lu by T'ao Tsung-i and the Ku kung i lu by Hsiao Hsiin, offer useful descriptions of the layout and the actual buildings of Peking at that time. Nancy Schatzman Steinhardt used these two texts in her "Imperial architecture under Mongolian patronage: Khubilai's imperial city of Daidu" (Ph.D. diss., Harvard University, 1981). See also her article "The plan of Khubilai Khan's imperial city," Artiiui Asiae, 44 (1983), pp. 137-58. Chinese archaeologists have also begun to explore some of the remains of the Mongols' capital of Ta-tu (Daidu). For examples of their recent discoveries, see Yuan Ta-tu k'ao ku tui, "Yuan Ta-tu te k'an ch'a ho fa chiieh," K'ao ku, 1 (1972), pp. 19-28; Yuan Ta-tu k'ao ku tui, "Chi Yuan Ta-tu fa hsien te Pa-ssu-pa tzu wen wu," K'ao ku, 4 (1972), pp. 54—7; Yuan Ta-tu k'ao ku tui, "Pei-ching Hou Ying-fang Yuan tai chii chu i chih," K'ao ku, 6 (1972), pp. 2 - 1 5 ; Chang Ning, "Chi Yuan Ta-tu ch'u tu wen wu," Kao ku, 6 (1972), pp. 2 5 - 3 4 . Cambridge Histories Online © Cambridge University Press, 2008 456 THE REIGN OF KHUBILAI KHAN :ity Surveillance Bureau 'I 'I t r o p o l i t a n Governor] Tsnple of-j City God i- Ast r o n a n i c a l ,' Caimissicn LI-CHENG GATE L Palace Ceremonial Office HEN-MING LHEN-MING GATE Treasury KEY: IMPERIAL PALACE CITY (Inner enclosure) 1. YEN-CH'UN HALL 2. TA- MING PALACE IMPERIAL c m (Outer e n c l o s u r e ) 3. 4 HSING-CHING PALACE OP EMPRESS DOWMER LUHG-FU PALACE OF HEIR APPARENT 5 WAN-SUI ISLAND MAP 33 . The Yiian capital, Ta-tu Cambridge Histories Online © Cambridge University Press, 2008 KHUBILAI AND RELIGION 457 up in the imperial parks, and Khubilai's sons often lived in them rather than in the palaces. When one of Khubilai's wives was in the last stages of pregnancy she moved into one of these tents to give birth.67 Khubilai brought grass and dirt from the Mongolian steppe for his royal altar, another reminder to himself and his fellow Mongols of their heritage. The paramount influence, however, remained Chinese. Possibly the clearest manifestation of the Chinese influence on the city were the temples that Khubilai ordered built near his palaces. The construction of the Great Temple (T'ai miao) illustrated his desire to gain favor with the Confucian elite. Reverence for the ancestors was vital to the Chinese, and the building of the Great Temple indicated that Khubilai intended to maintain the rituals associated with ancestor worship. He also had tablets fashioned for eight of his ancestors, including Chinggis, Ogodei, and Mongke. The same motives doubtless prompted him to build the Altars of the Soil and Grain in the capital. In 1271, he ordered that annual sacrifices be conducted at these altars in the Chinese style, in order to propitiate the gods and ensure bountiful harvests. He even built a shrine for Confucius where court officials made offerings to the Chinese sage and conducted annual ceremonies. Khubilai rarely took part in these ceremonies; instead he sent Chinese advisers to represent him. Khubilai first held court in his new capital in the first month of 1274. As Ta-tu became more of a Chinese-style capital city, K'ai-p'ing or Shang-tu, Khubilai's original summer capital, the "Xanadu" of Samuel Taylor Coleridge, was converted to other uses. It became the principal site for the continuance of the Mongols' traditional shamanistic rituals, which continued to be practiced there throughout Yuan times.68 It would serve more and more as Khubilai's summer retreat and hunting park than as a real capital, a place where he could retain and reaffirm his connections with traditional Mongolian pursuits. Shang-tu, more Mongolian than Chinese in its lack of a massive governmental apparatus, also offered welcome relief for the Mongolian khan from the oppressive style of life obligatory for a Chinese emperor. KHUBILAI AND RELIGION Another of Khubilai's efforts to ingratiate himself with his Chinese subjects involved his policies toward the religions in his realm. Although he himself continued to observe the practices of the Mongols' shamanistic religion, his mother's training had impressed on him the political significance of offering 67 Paul Ratchnevsky, "Uber den mongolischen Kult am Hofe der Grosskhane in China," in Mongolian studies, ed. Louis Ligeti (Amsterdam, 1970), pp. 4 1 7 - 4 3 . 68 See, for example, YS, 77, pp. 1923—7, esp. p. 1924. Cambridge Histories Online © Cambridge University Press, 2008 458 THE REIGN OF KHUBILAI KHAN patronage and support for the leading religions in a newly conquered land. In the 1260s, Khubilai needed to devise a relationship with the various Chinese religions that buttressed his position as ruler of China and hence ensured Mongolian control of the country. Even before he assumed the title of emperor, he had attempted to appeal to Chinese religious dignitaries, but now his efforts took on greater significance and urgency. Khubilai wished, first of all, to cultivate good relations with the Confucians. By 1267, the year the construction of Ta-tu began, he had ordered the building of the Imperial Ancestral Temple (T'ai-miao) and the preparation of ancestral tablets for the practice of dynastic ancestor worship, and he had designated a calendar for the country, a vital task for the ruler of an agricultural society. His selection of a name for his dynasty would be an allimportant signal to Confucians. The adoption of a Chinese name rich in Chinese symbolism would indicate that Khubilai wished to blend with some Chinese traditions. In 1271, Khubilai, at Liu Ping-chung's suggestion, chose the name Ta Yiian from the / ching (Book of changes). Yiian referred to the "origins of the Universe" or "the primal force," but most important, this name for the new dynasty had direct associations with one of the works of the Chinese canonical tradition. ^ In the same year, Khubilai reinstituted at the court the traditional Confucian rituals and their accompanying music and dance. Proper performance of these rituals was essential if the court was to avert imbalances in nature that led to floods, droughts, or earthquakes. Khubilai not only ordered their reintroduction but also had his Confucian advisers teach the ceremonies to a selected group of about two hundred Mongols, another telling indication of his desire to ingratiate himself with the Chinese.70 Further proof of Khubilai's sensitivity to Confucianism and Chinese values may be gleaned from the training and education that he prescribed for his second son, whom he eventually designated as his successor. With the help of the Buddhist monk Hai-yiin, he had chosen a Chinese Buddhist name, Chenchin (True gold), for this son.71 Determined that Chen-chin receive a firstrate Chinese education, he assigned Mao Shu, Tou Mo, and Wang Hsiin, three of his most prized Confucian advisers, to tutor the young man. These learned men introduced Chen-chin to the Chinese classics and presented him with an essay summarizing the political views of some of the emperors and ministers of earlier Chinese dynasties. 69 Maurizia Dinacci Saccheti, "Sull'adozione del nome dinastico Yiian," Annali htituto OrientatediNapoli, 31, n.s. 21 (1971), pp. 553-8. 70 YS, 67, pp. 1665-6; 88, p. 2217. 71 This name is sometimes given the Mongolian reading "Jingim," but strictly speaking, this is incorrect. Cambridge Histories Online © Cambridge University Press, 2008 KHUBILAI AND RELIGION 459 Khubilai also ensured that his young son was exposed to the other religions and cults in his Chinese domain. Chen-chin thus received instruction from a Buddhist monk, the 'Phags-pa lama, who wrote a brief work entitled Ses bya rab-gaal (What one should know), to offer his young student a description of Buddhism.' 2 A leading Taoist master provided him with an introduction to that mystical religion. Pleased with Chen-chin's growing acceptance by the Chinese, Khubilai gave his son ever-increasing responsibilities and repeatedly promoted him, culminating in 1273 with the designation of Chen-chin as heir apparent. In so designating his own successor, Khubilai made a complete break with Mongolian custom, by preempting the normal process of election, and followed the normal procedure of a standard Chinese dynasty. Still another way of attracting the Confucian scholars was to provide tangible support for the propagation of their views. Khubilai promoted, for example, the translation of Chinese works into Mongolian. Such Confucian classics as the Hsiao ching (Book of filial piety) and the Shu-ching (Book of documents), as well as Neo-Confucian writings such as the Ta-hsueh yen-i by Chen Te-hsiu (1178—1235), were translated under Khubilai's patronage. 73 By making these texts available to the Mongolian elite, Khubilai showed the Chinese that he respected Confucian ideas. He also impressed the Chinese scholars by recruiting some prominent literati to teach their own people as well as Mongols and Central Asians. One of the most renowned such recruits was Hsu Heng (1209-81) whom Khubilai appointed chancellor of the Imperial College in 1267. Hsu, universally recognized as one of the greatest scholars of his age, pleased his Mongolian patron because in his teaching he concentrated on practical affairs. He succeeded because he "did not go into speculative, metaphysical matters or 'things on the higher level.' "74 In his advice to Khubilai, he emphasized pragmatic considerations, an attitude certain to gain him favor at the Mongolian court. Khubilai's favorable reaction to suggestions for writing a dynastic history in the traditional Chinese style also met with the Confucians' approval. Confucianism, with its emphasis on the past and on the use of historical models as guides to behavior, provided an impetus to such officially endorsed historiographical projects. In August 1261, the Confucian scholar Wang O (1190—1273) proposed that the historical records of the Liao and Chin dynas72 Constance Hoog, trans., Prince Jin-gim's textbook of Tibetan Buddhism (Leiden, 1983); Herbert Franke, "Tibetans in Yuan China," in China under Mongol rule, ed. John D. Langlois, Jr. (Princeton, 1981), p. 307. 73 Walter Fuchs, "Analecta zur mongolischen Ubersetzungsliceratur der Yiian-Zeit," Monumenta Serica, 11 (1946), pp. 3 3 - 6 4 . 74 Wing-tsit Chan, "Chu Hsi and Yuan Neo-Confucianism," in Yuan thought: Chinese thought and religion under the Mongols, ed. Hok-lam Chan and Wm. Theodore de Bary (New York, 1982), p. 209. Cambridge Histories Online © Cambridge University Press, 2008 460 THE REIGN OF KHUBILAI KHAN ties, as well as those of the early Mongolian rulers, be collected.7' He also suggested that the court establish a History Office under the Han-lin Academy (to be known as the Han-lin kuo-shih yuan) to assemble the records and then compose the histories of the Liao and Chin dynasties. Khubilai, who ostensibly did not share the Chinese enthusiasm for historical writing, nonetheless approved the founding of a History Office, another decision by which he hoped to gain the approval of the Confucians. Khubilai naturally needed to appeal to religions and cults other than Confucianism if he wished to be perceived as the ruler of China. One of the religious groups that he was especially anxious to influence was the Muslims. Islam had reached China as early as the T'ang dynasty, and by Khubilai's time, Muslim merchants, craftsmen, and soldiers - most of them immigrants from Central Asia, but some of them Chinese converts to Islam — were to be found throughout the country, though they tended to be concentrated in the northwest and southeast. Khubilai pursued a benevolent policy toward the Muslims because they were useful to him in governing China. By recruiting Muslims for his government, Khubilai made himself less dependent on Chinese advisers and officials. He thus permitted them to form virtually selfgoverning communities with a shaikh al-lslam (Chinese: hui-hui t'ai-shih) as their leader and a qadi (Chinese: hui-chiao-t'u fa-kuan) as their interpreter of Muslim laws. The Muslim quarters had their own bazaars, hospitals, and mosques, and they were not prevented from using their native languages or following the dictates of Islam. Khubilai, in fact, appointed Muslims to important positions in the financial administration and accorded them special privileges. He exempted them from regular taxation and recruited them as darughachi (commissioners), a position that few Chinese could hold. The Muslims were grateful and responded by loyally serving the court. The most renowned of these Muslims was Saiyid Ajall Shams al-DIn, from Bukhara, who was appointed in 1260 as pacification commissioner of a district in north China and eventually wound up as the governor of the southwestern province of Yunnan.76 The Buddhists were still another group whose support Khubilai wished to cultivate. As early as the 1240s, he himself had received instruction from the Ch'an Buddhist monk Hai-yiin, but he had soon found Ch'an Buddhism too abstruse and unworldly for his purposes. It appeared to lack any concern for practical affairs, as, for example, when a Ch'an master had told Khubilai that 75 Hok-lam Chan, "Wang O (1190-1273)," Papers on Far Eastern History, 12 (1975), pp. 43—70, and "Chinese official historiography at the Yuan court: The composition of the Liao, Chin, and Sung histories," in China under Mongol rule, ed. John D. Langlois, Jr. (Princeton, 1981), pp. 56—106, esp. pp. 64—6. 76 Rossabi, "Muslims in the early Yuan dynasty," pp. 2 5 8 - 9 9 . Cambridge Histories Online © Cambridge University Press, 2008 KHUBILAI AND RELIGION 461 "all things are nothing but the Mind only."77 The Tibetan form of Buddhism, however, provided a much more suitable vehicle for Khubilai's purposes. Tibetan monks had played an active role in secular and political affairs for decades, and they offered greater expertise in practical matters than did the Ch'an monks. The Tibetan monk the 'Phags-pa lama (1235-80) proved to be Khubilai's closest ally among the Buddhists. The 'Phags-pa lama had spent much of his childhood at the Mongolian court and, through his long association with the Mongols, had absorbed many of their values. Yet he was also the nephew of one of the principal leaders of the Sa-skya sect of Tibetan Buddhism, in which Khubilai had been given a consecration in 1253, and as such he was respected, if not revered, by his own people. He could give valuable help to Khubilai in both China and Tibet. The Mongolian monarch thus treated him extraordinarily well. Members of his family were married to members of the Mongolian imperial clan. In 1260, Khubilai appointed the 'Phags-pa lama to the new position of state preceptor (kuo shih) and, early in the following year, placed him in charge of all the Buddhist clergy. By 1264, Khubilai had founded the Supreme Control Commission (Tsung-chih yuan) to administer Tibet and to supervise government relations with the Buddhists, and the 'Phags-pa lama became its first director. The 'Phags-pa lama's authority in Tibet was challenged in a rebellion led by the 'Bri-gung-pa order, a rival Buddhist sect, but in 1267 Khubilai dispatched troops to help restore the young Buddhist cleric to power. After his forces crushed the dissidents in 1268, Khubilai reinstated the 'Phags-pa lama, but he also stationed a Mongol as a pacification official (hsiian wet shih) in Tibet to help keep the country under control. 78 Khubilai expected that the 'Phags-pa lama and his fellow Buddhists would repay him by providing him with the religious sanction he needed. The Tibetan Buddhist did indeed fulfill his side of the bargain, as he delineated the spheres of authority of the church and the state, concluding that "the heads of the religion and of the state are equal." 79 He identified Khubilai with Mafijusri, the Boddhisattva of Wisdom, and also portrayed him as the Universal Emperor (Sanskrit: Cakravartin), in the Buddhist tradition. To heighten the connections between his sect and the emperor, the 'Phags-pa lama pro77 Translated in Yun-hua Jan, "Chinese Buddhism in Ta-tu: The new situation and new problems," in Yuan thought: Chinese thought and religion under the Mongols, ed. Hok-lam Chan and Wrn. Theodore de Bary (New York, 1982), p. 395. 78 For sources on the 'Phags-pa lama, see Miyoko Nakano, A philological study in the 'Phags-pa lama script and the Meng-tu tzu-yun (Canberra, 1971), pp. 152—65; Herbert Franke, "Tibetans in Yuan China," pp. 296—328, esp. pp. 305—11; Luciano Petech, "Tibetan relations with Sung China and with the Mongols," in China among equals: The Middle Kingdom and its neighbors, 10th— 14th centuries, ed. Morris Rossabi (Berkeley and Los Angeles, 1983), pp. 173—203. 79 Franke, From tribal chieftain to universal emperor and god, p. 6 1 . Cambridge Histories Online © Cambridge University Press, 2008 462 THE REIGN OF KHUBILAI KHAN posed the initiation of court rituals associated with Buddhism. Annual processions and parades, which were designed to destroy the "demons" and to protect the state, were organized on thefifteenthday of the second month, and music, rituals, and parades also were scheduled in the first and sixth months of the year. The Buddhist clergy's participation in these ceremonies offered Khubilai greater credibility with the Buddhists in his domains. Khubilai, in turn, rewarded the Buddhists with special privileges and exemptions. Buddhist monks were granted a tax-exempt status for much of his reign; the court supplied funds for the construction of new monasteries and temples as well as for the repair for some that had been damaged during the Buddhist—Taoist disputes; and the government provided artisans and slaves to work in the craftshops and on the lands owned by the monasteries.80 Government support, subsidies, and exemptions enabled the monasteries and temples to become prosperous economic centers and helped ensure Buddhist support for Khubilai's policies. Taoism was still another of the religions from which Khubilai sought sanction and assistance. Khubilai's support for the Buddhists in their debate with the Taoists in 1258 had not endeared him to the Taoist hierarchy. Yet he was entranced by their reputed magical powers and recognized their strong appeal to the lower classes. The court thus provided monies for the construction of Taoist temples and offered them some of the same exemptions and privileges that it had accorded to the Buddhists. A few of the Taoist leaders recognized the need for an accommodation with the Buddhists and the Mongols and sought first to reconcile the three teachings of Confucianism, Buddhism, and Taoism. Later they served Khubilai and his court by performing the sacrifices and ceremonies associated with the Taoist cults, in particular the worship of T'ai-shan, a vital imperial cult. Their willingness to conduct these ceremonies for Khubilai signaled a kind of support that was transmitted to ordinary believers in Taoism, and the Taoists remained relatively quiescent for the first two decades of Khubilai's reign. Khubilai and the Western Christians Khubilai even sought to secure support and assistance from the small Christian population within China, as well as from foreign Christians. Christian emissaries, such as John of Piano Carpini and William of Rubruck, had reached the Mongolian court during the reigns of Khubilai's predecessors, and several craftsmen, such as the renowned artisan Guillaume Boucher, had 80 Nogami Shunjo, "Gcndai dobutsu nikyo no kakushitsu," Otam daigaku kcnkyv nempo, 2 (1943), pp. 230—1; Paul Ratchnevsky, "Die mongolische Grosskhane und die buddhiscische Kirche," in Asiatica: Festschrift Friedrich Wtller zum 6}. Geburtstag, ed. Johannes Schubert (Leipzig, 1954), pp. 489—504. Cambridge Histories Online © Cambridge University Press, 2008 KHUBILAI AND RELIGION 463 served the great khan Mongke.8' But Khubilai made a concerted effort to invite and recruit foreign Christians. Marco Polo was the most renowned Christian in the exchanges between East and West in Khubilai's time.82 The Venetian traveler, whose book was to be the sole source of information for Europeans on China for many years to come, reputedly reached China in 1275.83 Marco tells us that his father and uncle, Nicolo and Maffeo Polo, preceded him to China. The two merchants had left Venice in 1252, had engaged in trade in Constantinople for some years, and had traveled in Russia and Central Asia until they reached Khubilai's court in late 1265 or early 1266. According to Marco, Khubilai "beamed with the greatest kindness" and "received them with great honour and makes them great joy and very great festival."8t After some polite conversation, Khubilai made his request: He asked the Polos to persuade the pope to send one hundred learned Christians with them when they returned to China. They could, he claimed, help convert his subjects to Christianity. Yet his principal motive in making this request was to attract learned men to help him administer his domains in China. With his eclectic attitude toward religion, Khubilai was not particularly eager for conversions to Christianity among his own people. But he needed to persuade the Polos and the Christian hierarchy that he wanted the learned Europeans to help in guiding his peoples to Christianity. When the Polos returned to the Christian world in 1269, they faced disappointments. They soon learned that Pope Clement IV had died in the previous year, impeding their plans to fulfill Khubilai's request and to return 81 See the charming book by Leonardo Olschki, Guillaumt Boucher: A French artist at the court of the khans (Baltimore, 1946). 82 Herbert Franke, "Sino-Western contacts under the Mongol empire," p. 54. The literature on Marco Polo and his book is voluminous. The best translation is the work by Arthur C. Moule and Paul Pelliot, Marco Polo: The description of the world; also useful is Henry Yule's The book ofSer Marco Polo, 3rd ed., rev. by Henri Cordier (London, 1903). This edition was supplemented by Henri Cordier in Ser Marco Polo: Notes and addenda to Sir Henry Yule's edition, containing the results of recent research and discovery (London, 1920). The finest study of Polo's book is Leonardo Olschki's Marco Polo's Asia (Berkeley and Los Angeles, i960). 83 Some scholars have speculated that Marco Polo may never have reached China and that some of the incidents he recounts were derived from conversations that he had with Persian or Arab merchants or travelers. See, for example, John W. Haeger, "Marco Polo in China? Problems with internal evidence," Bulletin of Sung and Yuan Studies, 14 (1978), pp. 22—30. Marco's own words give rise to these doubts. He claims, for example, to have assisted the Mongols in their siege of the Sung stronghold at Hsiang-yang, but that battle ended in 1273, t w 0 years before he allegedly arrived in China. There are other discrepancies as well as obvious exaggerations and some curious omissions in his accounts. These doubts, however, are far from conclusive, as Herbert Franke concluded in his "Sino-Western contacts under the Mongol empire," p. 54. See also Francis W. Cleaves, "A Chinese source bearing on Marco Polo's departure from China and a Persian source on his arrival in Persia," Harvard Journal of Asiatic Studies, 36 (1976), pp. 1 8 1 - 2 0 3 . Recently such doubts were laid permanently to rest by Yang Chihchiu, who in his Yuan shih san lun (Peking, 198;), pp. 97—132, produced conclusive proof of Marco Polo's presence in China during Khubilai's reign. 84 Moule and Pelliot, Marco Polo, vol. 1, p. 77. Cambridge Histories Online © Cambridge University Press, 2008 464 THE REIGN OF KHUBILAI KHAN to China as soon as possible. Just as they resolved to leave without a papal blessing, a new pope was elected, and they received an audience. They could not, however, obtain the requested one hundred learned Christians. In any case, they set forth, in 1271, for the great khan's court. Accompanied by Nicolo's son Marco, they finally arrived in China in 1275. Khubilai must have been dismayed that the one hundred learned men he had sought did not accompany them, but he was evidently impressed with Marco's intelligence and abilities. According to Marco's account, the great khan dispatched him to various locations in China and Southeast Asia to act as the emperor's "eyes and ears" and to bring back reports on the sites he visited.85 Similarly, Marco was struck by Khubilai's abilities. Marco saw the great khan at the height of his powers and described him in a flattering light. Marco assessed Khubilai as "without contradiction the greatest lord that ever was born in the world or that now is."86 He described, in some detail, the court banquets, the celebrations of the New Year, and the hunting and hawking forays led by Khubilai and reported on such curiosities as paper money, coal, and the postal system, all of which would impress Europeans. Khubilai must have been pleased with the young European's identification with and apparent admiration for the Mongols. He recognized that it was in his best interests to treat the young man well, particularly if he wished to entice more Europeans to his court. Khubilai could further ingratiate himself to the Europeans through a tolerant policy toward Christianity. His mother had paved the way by embracing Nestorianism. Khubilai did not become a Christian, but he employed Nestorians at court. He did not restrict the practice of Nestorianism, and Marco commented on the churches he came across in Kan-chou, Su-chou, and other small towns in the northwest. Khubilai also exempted clerics from taxation and conscription. Eventually, he founded a special government agency, the Office for Christian Clergy (Ch'ung fu ssu), to supervise the Nestorian clerics in his domains.87 His dispatch of two Nestorian prelates to the Middle East was still another indication of his efforts to appeal to Christians. In 1275-6, Rabban Sauma and Rabban Markos left Ta-tu to visit the holy sites in Jerusalem; they probably could not have traveled through north China and Central Asia without Khubilai's approval and support. Rabban Sauma met with the Mongolian Il-khan in Persia and would eventually be sent by one of the Il-khans to negotiate an alliance with the Europeans. He had an audience with the pope in Rome, and was granted interviews with 85 This may be another of Marco's exaggerations. See John W. Haeger, "Marco Polo in China?" pp. 2 6 - 7 . 86 Moule and Pelliot, Marco Polo, vol. 1, p. 77. 87 Arthur C. Moule, Christians in China before the year 1550 (London, 1930), pp. 1 3 1 - 2 , 2 2 5 - 8 ; YS, p. 2373- Cambridge Histories Online © Cambridge University Press, 2008 KHUBILAI AND CHINESE CULTURE 465 King Philip the Fair of France in Paris and with King Edward I of England in Bordeaux. These meetings also did not lead to the importation of learned Christians to Khubilai's court, nor did they inspire the formation of an alliance with the Europeans. Yet they indicated that the Yuan court tolerated Christianity, sought Christian officials in government, and welcomed contact with the larger Christian world. 88 KHUBILAI AND CHINESE CULTURE As emperor of China, Khubilai wished to portray himself as a patron of Chinese culture. He could not appear to be a crude, unsophisticated "barbarian" if he wanted to be accepted as the Chinese Son of Heaven. The Mongolian rulers had, from earliest days, prized the works of good craftsmen, a tradition on which Khubilai could capitalize to support the arts, particularly handicrafts, in China. Yet he could not abandon the trappings of Mongolian culture lest he be accused by Mongolian traditionalists of favoring the Chinese. Moreover, as great khan, he had a responsibility to promote the cultural expressions of the diverse ethnic groups within his domains. He could not afford to be associated exclusively with Chinese culture. To maintain all of these various cultural strands in balance required constant vigilance and occasional shifts in policy. Khubilai's policy toward the written language in his domains reveals his characteristic response to cultural questions. The fiscal, military, and welfare responsibilities of his new government necessitated proper written records. Yet the Mongols had had scant experience with collecting and preserving such records. A serviceable written language was essential, and during the reign of Chinggis khan the Mongols had developed a script for their own language based on the Uighur Turkic alphabet. But Khubilai relied initially on Chinese scribes, who normally wrote in classical Chinese. He compelled them, however, to write in colloquial Chinese (pai-hua) both because "the adoption of the classical language would have implied yielding culturally to China"*9 and because the colloquial was more accessible to those Mongols who had studied Chinese. Most court documents were written first in Mongolian, and many were, by a laborious process, translated into colloquial Chinese. Yet the Uighur script was not as precise as it needed to be in transcribing the sounds of the Mongolian language. Moreover, it was unsuitable for an 88 E. A. Wallis Budge offers a complete translation of the Syriac text of the travels of the two prelates in The monks of Khubilai than, emperor of China (London, 1928). See also Morris Rossabi, Voyager from Xanadu: Rabban Sauma and the first journey from China to the West (New York, 1992). 89 Igor de Rachewiltz, "Some remarks on the language problem in Yuan China," Journal of the Oriental Society of Australia, 5 (1967), p. 68. Cambridge Histories Online © Cambridge University Press, 2008 466 THE REIGN OF KHUBILAI KHAN accurate transcription of Chinese and could not fulfill Khubilai's plans for a widely applied official script. Khubilai hoped to use the script that he adopted to help unify his realm and to assert his claim to universal rule. He wanted to go beyond the written languages, the Chinese characters and the Uighur script, available to him at that time. As the ruler of a domain in which various ethnic groups resided and many languages were spoken, Khubilai wanted a script that could be employed to transcribe all of these diverse languages. He aspired, in short, to the development of a universal script. What he failed to recognize, however, was that the imposition of an unfamiliar written language would not readily be accepted. An artificially designed script, no matter how accurate or effective, would encounter insurmountable difficulties with people attached to a tradition-laden written language. Khubilai was nonetheless determined to develop a better and more universal script. He assigned to the Tibetan 'Phags-pa the task of creating a new script, and by 1269 the 'Phags-pa lama had submitted a new alphabet consisting of forty-one letters and based on Tibetan. The 'Phags-pa script, sometimes referred to as the "squate script" because of the square shape of the letters, was more precise than the Uighur script in its rendering of the sounds of colloquial Mongolian. It also more accurately reflected the sounds of other languages, including Chinese, in Khubilai's realm. The 'Phags-pa lama alphabet appeared ideally suited to transcribe all the languages in Khubilai's domain, to serve as a universal script, and to contribute to the unification of the frequently antagonistic peoples under Mongol rule. Khubilai proudly designated it as the Mongolian script {Meng-ku tzu) and eventually referred to it as the state script (kuo tzu). He ordered that court documents be written in the state script and founded academies expressly for the propagation of the new script.'0 Yet Khubilai's expectations remained unfulfilled, for the script was not readily adopted. Even his own officials evaded the regulations mandating its use in court documents. The schools that he had founded in 1269 were also not as effective as he would have wished. A report in 1272 by one of his officials indicated that the children and relatives of his Chinese bureaucrats were not studying the new script. Despite his efforts and repeated admonitions, the 'Phags-pa alphabet never replaced either the Uighur script or Chinese characters. Few examples of'Phags-pa writing have survived. A few inscriptions on seals, coins, paper money, and porcelain and a few edicts and 90 On the new script, see Nicholas Poppe, trans., and John R. Krueger, ed., The Mongolian monuments in 'Pbags-pa script (Wiesbaden, 1957); and Nakano, A phonological study in the 'Phags-pa lama script. Cambridge Histories Online © Cambridge University Press, 2008 KHUBILAI AND CHINESE CULTURE 467 Buddhist texts have been found, but Chinese and the Uighur script remained dominant.91 With the fall of the dynasty, it disappeared. The failure of the 'Phags-pa script should not be attributed to its technical inadequacies; linguists consider it a marvel of phonetic accuracy and flexibility. But it was officially devised and imposed from above, even though it manifests the court's concern for a universal script and a written language that reflected the colloquial language of its time. Khubilai wanted to use the 'Phags-pa script to promote the diffusion of the colloquial language in writing. By accentuating the colloquial, he showed that he did not need to abide by the Chinese scholar-officials' rules or methods of governance, which entailed the use of the classical language and stressed knowledge and application of history to current policy decisions. It should not be surprising therefore that the colloquial also began to be employed outside court writings. The colloquial seeped into Yuan literature, and the vernacular and the popular artsflourishedas never before in Chinese history. The Chinese theater, in particular, blossomed during Khubilai's era and the reign of his immediate successors. The growth of cities in the late Sung and early Yuan periods had offered a fitting environment for the rise of the drama, for it provided the audience as well as the funds needed for performances. Without an urban culture and patronage from the government and private citizens, theater could not prosper. The Yuan cities did indeed support a great theater. At least 160 plays from that time have survived, and over 500 more were performed and written, though they are no longer extant. Districts with dozens of theaters sprouted in many of the cities. Actors and actresses, who had always previously been treated as social pariahs, found themselves in a more enviable position, at least during the early years of Mongolian rule. The Yuan dramas are described as "variety entertainments" because sketches - interspersed with songs, dances, and acrobatics - were characteristic of the genre, making it more accessible and more appealing to a popular audience. Professional playwrights as well as Chinese scholars, who were denied access to official posts because of the abolition of the civil service examinations, wrote the majority of the plays. Both Khubilai and the Mongolian court promoted the theater. They did so first by their lack of interference; playwrights could broach a variety of subjects without fear of government censorship. On a more positive note, Khubilai and his officials ordered the staging of a number of performances at court. They also 91 YS, 7 , p . 142; David Farquhar, "The official seals and ciphers of the Yuan period," Monunmenta Serica, 25 (1966), pp. 362—93; M. Aurel Stein, Innermost Asia (Oxford, 1928), vol. 1, pp. 441—55; Koyama Fujio, "Pasupa moji aru Shina furu toji," Gaieiiu, 1 (1937), pp. 23—31; John Ayers, "Some characteristic wares of the Yuan dynasty," Transactions of the Oriental Ceramic Society, 24 (1957), p p . 69—86. Cambridge Histories Online © Cambridge University Press, 2008 468 THE REIGN OF KHUBILAI KHAN appear to have served as patrons for a number of playwrights, and their espousal of the colloquial facilitated the task of the playwright and contributed to the development of Yuan theater. The Chinese playwrights were responsible for their own artistic creations. Yet Khubilai and his own Mongolian underlings contributed to an environment that encouraged or at the very least did not undercut the Yuan theater. He knew that a good emperor ought, in Chinese eyes, to be a patron of the country's culture and that the theater, as a developing art form in China, ought to be supported.92 Khubilai cannot be given credit for the development of the novel and for the larger number of printed texts available in China. His cultural and literary policies nevertheless offered favorable circumstances for growth. Khubilai's emphasis on the colloquial was a boon to novelists, who often portrayed characters of a lower-class origin. The use of the vernacular permitted novelists to reproduce the patterns of speech of the ordinary person and to present a broader range of characters. The court also promoted the wider diffusion of written texts, as printing under the Yuan sustained the high standards that had been reached under the Sung.55 In 1269, Khubilai founded a special office to print books under official sponsorship, and by 1286, land was assigned to academies, which used the income to print texts. The growth in printing offered more access to books and initiated the rise in literacy characteristic of the Ming and Ch'ing dynasties. Painting was another aspect of culture affected by the court. Khubilai and his fellow Mongols found painting accessible, for they did not have to overcome a formidable language barrier to appreciate it. The khans' own vanity also made them responsive to visual representations. Khubilai had a formal portrait of himself painted and then commissioned the artist, Liu Kuan-tao, to paint him in action during a hunt. He had the Southern Sung's Imperial Painting Collection transported to Ta-tu, where several Chinese connoisseurs cataloged it. The Sung paintings provided the foundation for his own collection, which grew as a result of his patronage of certain artists and his acquisition of some of their paintings. Some art historians have emphasized the negative impact, or the lack of influence of Khubilai and the Mongolian rulers on Chinese painting, but recent interpretations give some credit to the Yuan court.94 92 A good study of theater during this time is that by James I. Crump, Chinese theater in the days ofKublai khan (Tucson, 1980). 93 K. T. Wu, "Chinese printing under four alien dynasties," Harvard Journal of Asiatic Studies, 13 (1950), p. 459; see also Chih-ching (C. T.) Hsia, The classic Chinese novel: A critical introduction (New York, 1968), p. 8. 94 Sherman E. Lee and Wai-kam Ho, Chinese art under the Mongols: The Yuan dynasty (1279-1368) (Cleveland, 1968), p. 1. Cambridge Histories Online © Cambridge University Press, 2008 KHUBILAI AND CHINESE CULTURE 469 Some of the great Chinese painters, it is true, rejected employment or collaboration with the Mongols, but just as many were assisted and cherished by the early Yiian court. Some who refused to accept positions with the foreign conquerors became recluses {yin-shih), whereas others remained loyal to the conquered Sung dynasty and concealed their antipathy for the Mongols by turning to private pursuits. Painting was one such notable pursuit, and a group of so-called amateur painters, as distinct from the official Imperial Academy painters of Sung times, took shape. The literary men's painting (wen-jen hua) that they evolved emphasized the emotional responses of the artist, which naturally allowed the painters to express their hostility toward the Mongols in discreet ways. Cheng Ssu-hsiao (1241—1318), for example, was renowned for his paintings of the Chinese orchid, and when "asked why he depicted it without earth around its roots, he replied that the earth had been stolen by the barbarians."9' Kung K'ai (12221307), Ch'ien Hsiian (ca. 1235-ca. 1301), and other painters also employed their art as a subtle means of social protest. On the other hand, the court employed some of the great Chinese painters of the Yiian in official positions. It appointed Li K'an (1245-1320), the most famous painter of bamboo, to be the minister of personnel; Kao K'o-kung (1248—1310) was granted a position in the Ministry of Works in 1312; and the calligrapher Hsien-yii Shu (1257?1302) was employed in the Censorate and the Court of Imperial Sacrifices. Khubilai subsidized many others through sinecures in his government. Chao Meng-fu (1254-1322) was the most renowned of Khubilai's supporters among the painters. Because Chao was descended from the Sung imperial family, his conversion to the Mongols boosted Khubilai's credibility and legitimacy among the Chinese. To those who criticized him for abandoning the loyalist cause and serving the "barbarians," Chao responded that "each person lives his life in this world according to his own times."96 Despite considerable opposition by Mongols who doubted Chao's loyalty, Khubilai appointed the painter to the Ministry of War, and Chao responded by performing his tasks conscientiously. He proposed reforms in the postal service and reductions in taxes on the Chinese. In art, he found greater freedom under the Mongols than in the previous Sung dynasty. The Sung Imperial Academy had, he argued, established standards that stultified painters, whereas the new Yiian rulers did not interfere with the artistic creations of the painters and also exposed them to new subjects and themes - for example, the depiction of horses. 95 James Cahill, Hills beyond a river: Chinese painting of the Yiian dynasty (New York, 1976), p. 17. 96 Mote, "Confucian eremetism in the Yiian period," p. 236; Herbert Franke, "Dschau Mong-fu: Das Leben eines chinesischen Staatsmannes, Gelehrten und Kiinstlers unter der Mongolenherrschaft," Sinica, 15 (1940), pp. 25—48. Cambridge Histories Online © Cambridge University Press, 2008 47° THE REIGN OF KHUBILAI KHAN Khubiiai and the Mongols' influence on crafts was even more pronounced. When Khubiiai took power in China, hefollowedthe policies of his Mongolian predecessors in ensuring a good supply of artisans who could produce the articles that he and his people prized and needed. He devised organizations under the Ministry of Works to supervise and control the artisans in China, and he offered numerous privileges, including exemptions from most taxes, to craftsmen but made corresponding levies on their time or on their production. No wonder, then, that technical and aesthetic advances were made during his reign. The court promoted, in particular, ceramic production, for it recognized the potential for revenue in this industry. It could obtain the porcelains it needed and could trade the rest with Southeast Asia and the Middle East for handsome profits. The kiln sites of Te-hsing, An-fu, Te-hua, Lung-ch'iian, and Ching-te-chen were located in southeast China and could easily ship porcelains to foreign lands from the great ports in the region. The Yuan potters, who were granted great flexibility by the Mongols and were not inhibited by the Sung canons of taste, could innovate, and such experimentation resulted in the production of beautiful porcelains. The blue-andwhite porcelains originated in the Mongolian period, as did white porcelains and some celadons.97 Khubiiai also had an indirect influence on Chinese architecture. His Tibetan adviser the 'Phags-pa lama was greatly impressed by a newly constructed Buddhist temple in his native land and learned that the Nepalese craftsman A-ni-ko (1244—1306) had designed the building. The 'Phags-pa lama took the Nepalese artisan with him back to China in 1265 and introduced him to Khubiiai, who also formed a favorable impression of the young foreigner. Khubiiai assigned him a number of projects, and A-ni-ko rewarded his patron by designing a Buddhist temple and a pavilion in a park in Ta-tu, an ancestral temple in Cho-chou, and a temple in Shang-tu. Clearly delighted with A-ni-ko, Khubiiai named him, in 1273, the head of the Directorate General for the Management of Artisans, making him the supervisor of all craftsmen in China. Khubilai's wife Chabi, who also was enchanted with the foreign architect, arranged a marriage for A-ni-ko with a high-born Chinese woman, a descendant of the Sung royal family.98 Khubiiai and his family thus recognized a great craftsman and welcomed and rewarded him for his efforts. Neither Khubiiai himself nor the Mongols as a whole made a direct 97 Margaret Medley, Yiian porcelain and stoneware (New York, 1974); Paul Kahle, "Chinese porcelain in the lands of Islam," Transactions of tbe Oriental Ceramic Society, 18 (1940-1), pp. 2 7 - 4 6 ; and John A. Pope, Fourteenth century blue and white: A group of Chinese porcelains in the Topkapu Sarayi Mu'zesi, Istanbul (Washington, D . C . , 1952). 98 Ishida Mikinosuke, "Gendai no kogei Nebaru no ozoku Aniko no ten ni tsuite," Moko gakuho, 2 (1941), pp. 250-1. Cambridge Histories Online © Cambridge University Press, 2008 PRESERVATION OF THE MONGOLIAN HERITAGE 471 contribution to China's arts and crafts. Yet their patronage of the arts is undeniable, and such support promoted advances in the arts. Similarly, by allowing artists and artisans great freedom and flexibility, they inspired innovation and experimentation. As foreigners themselves, they were not averse to introducing non-Chinese forms and ideas into Chinese art. Khubilai's patronage of A-ni-ko, for example, led to the appearance of Tibetan and Nepalese forms in Chinese architecture. Khubilai's sponsorship of Chinese and non-Chinese alike, of course, boosted his claims to universal rule. PRESERVATION OF THE MONGOLIAN HERITAGE Khubilai needed to be accepted as the sovereign of China, but he also had to assert himself as the khan of the Mongols and the ruler of the non-Chinese domains under Mongolian control. An overly strong identification with the Chinese detracted from his image as a ruler of the vast Mongolian territories. Khubilai could not appear to find Chinese civilization more attractive than his own and had to avoid being engulfed by Chinese culture. Eventually he developed policies aimed at preserving the Mongols' uniqueness and identity. He generally attempted to discourage fraternization between the Mongols and the Chinese. Khubilai distinguished between his own governmental practices and those of the Chinese. The civil service examinations were not reestablished until 1315, twenty years after his death, thus dispensing with what had been, for educated Chinese, the most common traditional means of entering the bureaucracy. The financial administration was placed in the hands of nonChinese. Censors were stationed as spies on officials throughout the empire, and there appears to have been a greater emphasis on control than in any earlier dynasty. The military seems also to have been more prominent than under traditional Chinese dynasties. Some scholars have suggested that the Mongols ushered in an era of increased violence and brutality at court, but such accusations are difficult to substantiate.^ It seems unlikely that brutality could be attributed more to the Mongolian heritage than to Chinese traditions. Flogging and brutalization of officials did not suddenly emerge in the Mongolian era. Khubilai took some positive steps to preserve Mongolian rituals and customs. He continued to perform some of the traditional Mongolian ceremonies and to conduct Mongolian-style sacrifices to mountains, rivers, and trees, employing shamans to perform the traditional rituals. Every August, 99 Frederick W. Mote, "The growth of Chinese despotism: A critique of Wittfogel's theory of oriental despotism as applied to China," Orient Extremus, 8 (1961), pp. 17-18. Cambridge Histories Online © Cambridge University Press, 2008 472 THE REIGN OF KHUBILAI KHAN for example, before leaving Shang-tu to spend the fall and winter in Ta-tu, he conducted the ritual of scattering mare's milk, which reputedly ensured a year of good fortune. It consisted of sacrificing a horse and a few sheep, bowing toward Heaven, calling out Chinggis khan's name, and scattering mare's milk from specially bred horses. In this way, Khubilai honored his ancestors and sought their blessings for the following winter. When a member of the imperial family was ill, Khubilai ordered that he or she be placed in a yurt and that two sheep be sacrificed daily until the ailing one had recovered. Before Khubilai went into battle, he poured out a libation of koumiss, invoking the assistance of Heaven against his enemies.100 Khubilai affirmed secular Mongolian practices as well. Unlike Chinese women, Mongolian women did not adopt the practice of foot binding, and Khubilai did not impose the restrictions on them that were enforced on Chinese women. Most Mongols continued to wear their native costume, and elaborate, extravagant feasts, more reminiscent of the nomads' celebrations, with almost uncontrolled eating and drinking, were held on Khubilai's birthday and at the start of the New Year. At these feasts the guests drank enormous quantities of liquor. The alcoholism that was so much a part of early Mongolian history, and indeed of the life-style of all the northern peoples, apparently persisted into Khubilai's reign. The great khan's fascination with the hunt was perhaps the most telling indication of the retention of Mongolian ways. According to Marco Polo, Khubilai was accompanied on hunts by trained lions, leopards, and lynxes who chased and often captured boars, oxen, bears, and wild asses. He also took about five hundred gerfalcons with him to snatch other birds from the sky. An enormous retinue of falconers, hunters, and soldiers accompanied Khubilai on these hunts.101 Khubilai succeeded admirably in charting a cultural policy that affirmed the Mongolian heritage, accepted Chinese practices, and strove for universalism. He wished to appear in different guises to the different audiences he faced. To the Mongols, he seemed to be a staunch defender of his native traditions. He took part in hunts, married Mongolian women, and sought to preserve their rights. To the Chinese, he assumed the role of patron of the arts. He subsidized Chinese painters, potters, and other craftsmen and allowed freedom of expression for Chinese dramatists and novelists. To the rest of his domains, his espousal of a universal script and his encouragement and support for foreign craftsmen in China inspired a cosmopolitanism in Yuan 100 Ratchnevsky, "fiber den mongolischen Kult am Hofe der Grosskhane im China," pp. 426—8, 434— 42101 Moule and Pelliot, Marco Polo, vol. 1, p. 231. Cambridge Histories Online © Cambridge University Press, 2008 ECONOMIC PROBLEMS IN LATER YEARS 473 culture that no doubt added to Khubilai's luster as a ruler of a realm not confined merely to China. ECONOMIC PROBLEMS IN LATER YEARS The year 1279 proved to be a watershed in Khubilai's reign. Until that time, he had scarcely experienced any failures in his undertakings. He had crushed all opposition, including that of his younger brother. He and his advisers had established a government based on Chinese models but not dominated by Chinese ideals and forms. His two capitals at Shang-tu and Ta-tu were well planned, functional, and beautiful. He had, through a carefully considered policy, gained favor with most of the religious leaders in his realm. His armies had conquered the rest of China and had asserted Mongolian control over Korea and Mongolia. He had encouraged the creative arts and had recruited some of the ablest artisans in the land to produce exquisite articles for the court and the elite and for use in foreign trade. His most conspicuous failure had been his abortive invasion ofJapan, but he could rationalize this defeat by blaming it on the natural disaster, the terrible storm, that devastated his forces. All else in the first two decades of his reign seemed to be proceeding smoothly. But appearances were deceptive. Some difficult problems lay beneath the surface. Some Confucian scholars were not reconciled to Mongolian rule, and their dissatisfaction became even more pronounced with the amalgamation of the Southern Sung into the Yuan domains. The scholars in the south had not experienced foreign domination, and quite a few eventually refused to collaborate with the Mongols. Khubilai himself began to slow down after 1279. Now in his late sixties, he was afflicted with health problems. Gout plagued him and made it difficult for him to walk. The most pressing problem that Khubilai faced was finances. His construction projects, his support for public works, and his military expeditions entailed vast expenditures. To obtain the needed funds, Khubilai turned to the Muslim finance minister Ahmad, whom the Yuan dynastic history classifies as one of the "villainous ministers" and who is reviled by both Chinese and Western sources.102 In his defense, we should recognize that Ahmad knew that he would be judged by the amount of revenue he collected for the court. The more funds he raised, the greater his power, prestige, and income would be. He surely profited from his position, but it must be remembered that his accusers (those who wrote the Chinese accounts) were officials unsympathetic to his policies. 102 Herbert Franke, "Ahmed: Ein Beitrag zur Wirtschaftsgeschichte Chinas untet Qubilai," Orient, 1 (1948), pp. 2 2 2 - 3 6 . Cambridge Histories Online © Cambridge University Press, 2008 474 THE REIGN OF KHUBILAI KHAN From 1262, when he was apppointed to the Central Secretariat, until his death in 1282, Ahmad directed the state financial administration. He first increased the number of households on the tax registers from 1,418,499 in 1261 to 1,967,898 in 1274.IO3 He then imposed higher taxes on merchants, instituted state monopolies of new products, and forbade the private production of certain commodities. In short, Ahmad's policies were lucrative for the state treasury. Yet the Chinese sources accuse him of profiteering and cronyism. They assert that he capitalized on the new taxes and monopolies to enrich himself. Moreover, they denounce him for appointing Muslims to prominent positions and for attempting to place his own inexperienced and perhaps unqualified sons in influential posts in the bureaucracy. Viewed from a different perspective, however, the Chinese accusations appear less serious. Bringing like-minded associates and relatives into government was perfectly sensible. If Ahmad were to overcome opposition and implement his policies, he needed to place his supporters in influential positions. He did impose heavy taxes and high prices, but his position at court — not to mention possible promotions and rewards — depended on his ability to satisfy the Mongols' revenue requirements. He was a dedicated and effective agent of the Mongolian court, which had a considerable and pressing need for income. Ahmad's policies, however, aroused the opposition of some leading Chinese at court. Khubilai's Confucian advisers resented Ahmad's power and accused him of profiteering and of having a "sycophantic character and treacherous designs." By the late 1270s, the heir apparent Chen-chin had joined the opposition. Chen-chin objected to the prominent positions accorded to Ahmad's sons and relatives. On 10 April 1282, while Khubilai was in his secondary capital at Shang-tu, a cabal of Chinese conspirators lured Ahmad out of his house and assassinated him.10* Within a few days, Khubilai returned to the capital and executed the conspirators. But his Chinese advisers eventually persuaded him of Ahmad's treachery and corruption. The evidence they used against Ahmad was suspect, but Khubilai was convinced of the Muslim minister's guilt and so had his corpse exhumed and hung in a bazaar; then he allowed his dogs to attack it. Yet the elimination of Ahmad did not resolve Khubilai's financial problems. His revenue requirements became even more pressing after Ahmad's death because he initiated several military expeditions to Japan and Southeast Asia. Simultaneously, by the early 1280s, Khubilai had lost some of his most faithful Chinese advisers, including Hsu Heng, Yao Shu, and Wang O, all of whom had died by that time. Their deaths offered non-Chinese counselors 103 See ibid., p. 232. 104 See Moule, Quinsai, pp. 79—88, for an account of this cabal and its plot to kill Ahmad. Cambridge Histories Online © Cambridge University Press, 2008 ECONOMIC PROBLEMS IN LATER YEARS 475 greater opportunities to influence him. Khubilai's own infirmities compounded these troubles and may, in part, have accounted for his abdicating more and more of his responsibilities as a ruler. The Chinese sources accuse Lu Shih-jung, another of the so-called villainous ministers, of capitalizing on Khubilai's difficulties to increase his own power. After Ahmad's death, Lu became the head of the Ministry of the Left in the Central Secretariat, with jurisdiction over much of the financial administration in China. Like Ahmad, he attempted to increase the government's revenues in order to meet the mounting costs at court. He sought to augment the government's income from monopolies, to impose higher taxes on foreign trade, to issue more paper money (an easier way to pay government debts), and to staff the tax offices with merchants. 10 ' Lu's economic programs engendered the same hostility as did those of Ahmad, his predecessor as financial administrator. The Chinese accused him of profiteering, cronyism, and exploitation of his own people and of persecuting, hounding, and even executing rivals and enemies. The accuracy of many of these charges is subject to doubt because the sources do not reflect Lu's own version of events. Like Ahmad, Lu simply attempted to raise desperately needed revenues, but his efforts earned the enmity of many of his fellow Chinese. Again, Crown Prince Chen-chin led the opposition to Lu, who was arrested by May 1285 and executed by the end of the year. His death may have removed a man that the Chinese perceived to be an exploiter, but it did not alleviate the fiscal problems faced by the court. Aside from fiscal problems, Khubilai also faced difficulties in achieving the economic integration of the Southern Sung into his realm. A truly unified and centralized China was essential if Khubilai wished to fulfill any other economic or political objectives. Khubilai first sought to ingratiate himself with the Chinese in the south by releasing many of the soldiers and civilians whom his armies had captured. Then he issued edicts aimed at the economic recovery of south China, including prohibiting Mongols from ravaging the farmlands and establishing granaries to store surplus grain and to ensure sufficient supplies in times of agricultural distress. The court did not generally confiscate land from the large estates of the southern landowners. Nor did it undermine their power; it simply added another layer — the Mongolian rulers — at the top of the hierarchy. The land taxes it imposed were not onerous and, in times of distress, were waived. Salt, tea, liquor and other commodities were monopolized, but the resulting prices were not burdensome. Khubilai encouraged maritime commerce, one of the bases of 105 Herbert Franke, Geld und Wirtichaft in China unter der Mongolenbemcbaft: Beilrage zur Wirucbaftigeschichte der Yiian-Zeit (Leipzig, 1949), pp- 72—4. Cambridge Histories Online © Cambridge University Press, 2008 476 THE REIGN OF KHUBILAI KHAN the south's prosperity. Self-interest surely was a motivating factor in these policies, for the economic recovery of the south would eventually mean greater profits. Despite his efforts, the hostile feelings of some Chinese in the south did not subside, hampering Khubilai's economic program. There were several uprisings against Mongolian rule. In 1281, Khubilai's troops crushed the first of these, which was led by Ch'en Kuei-lung, by beheading - if we trust the Chinese historians - twenty thousand of the rebel soldiers. One hundred thousand Mongolian troops were required to overwhelm a more serious rebellion in Fukien. Other insurrections continued until the end of the reign. Most of those who resisted the Mongols, however, did not resort to such violent means. A few refused to serve the Mongols, feeling that the "barbarians" were not interested in Chinese civilization and thought. Others founded special academies to pursue their own intellectual interests while simply avoiding involvement with the Mongols. Such opposition deprived Khubilai and the Yuan court of badly needed expertise while the continuing turbulence compelled them to station troops in the south, at great expense. In sum, the south was not totally integrated by the end of Khubilai's reign, and the economic problems, together with political disruptions, continued to plague the Yuan court in this region. His efforts in the south having been partially unsuccessful, Khubilai set about fulfilling the needs of his core territories in the north. Having established his capital in Ta-tu, Khubilai was required to ensure a steady supply of grain for his new city, which compelled him to import grain from the more fertile areas in south China, as the region around Ta-tu did not grow enough food for its needs. At first, Khubilai relied on Chu Ch'ing and Chang Hsiian, two pirates who had cooperated with the Mongolian general Bayan in his military campaigns against the Southern Sung, to transport grain by sea along the eastern coast of China to the north. Their first shipment in 1282 was exceptionally successful, as over 90 percent of the grain reached the north and only six ships were lost in the treacherous waters along the coast. For about four years, Chu and Chang received exclusive and lucrative contracts to transport the desperately needed grain. They became, as a result, "two of the wealthiest and most influential men in southeast China."106 In the mid12808, however, typhoons and unfavorable conditions wreaked havoc with these shipments and convinced the court that an alternative route for transporting the grain was needed (see Map 34). The court decided to lengthen the Grand Canal to a point where grain 106 Lo Jung-pang, "The controversy over grain conveyance during the reign of Qubilai Qagan, 1260— 94," Far Eastern Quarterly, 13 (1952), pp. 262—6. Cambridge Histories Online © Cambridge University Press, 2008 ECONOMIC PROBLEMS IN LATER YEARS 477 Ch'ingjchou Ts'ang-chou Hui-t'ung Canal planned bySangha1288 Opened 1289 jTLin-ch'ing ±#i Finallvcompleted1329J}>*% Dredged and widened 1282-83 under A'urughchiandAbachi. Route abandoned 1287 Chiao-e chou * Feeder canals 1283 proposed 1280 on plan of Yao-yin. built 1283. Abandoned 1285. and again in 1287 \ \ Yang-chou Chen-Chiang ^ ^ Wu-hsi ing-chiarigj Chia-hsing —— ^ 'I'I'I'I Existing canals and canalized rivers Earh/YQan grain transport Sea route used 1282-92 Sea route after 1293 Hang-chou Canals built 1283 xxxxxxx Canal route 1285-87 0 300 kn 200miles MAP 34. The Yiian grain transport system Cambridge Histories Online © Cambridge University Press, 2008 478 THE REIGN OF KHUBILAI KHAN shipments could easily reach Ta-tu. This project entailed the construction of a 135-mile-long canal from Ch'ing-ning to Lin-ch'ing in Shantung Province; from Lin-ch'ing, goods could be transshipped on the Wei River to Chih-ku, a short distance from Ta-tu. Grain could thus be transported from the Yangtze directly to Khubilai's capital. By February 1289, the extension was completed, and the new canal, known as the Hui-t'ung, was opened to boat traffic.IO? The expenses entailed in the extension of the canal were enormous. About three million laborers took part in its construction, for which the government expended vast sums of money. Maintenance was also costly, and the huge expenditures necessitated by the canal no doubt contributed materially to the fiscal problems plaguing the Mongolian court in the late 1280s. THE REGIME OF SANGHA AND ECONOMIC AND RELIGIOUS ABUSES Sangha was the last of the triumvirate of "villainous ministers" who attempted to deal with the court's fiscal problems under Khubilai. Like Ahmad, he was not Chinese, but his ethnic origins are obscure. Historians had assumed that he was.an Uighur, but recent studies suggest that he was a Tibetan. He first came into prominence as a member of the staff of the Buddhist 'Phags-pa lama. Khubilai was impressed with Sangha's capabilities and resourcefulness, and sometime before 1275 he appointed the young Buddhist as head of the Tsung-chih yuan, the office in charge of Tibetan and Buddhist affairs. Here, too, Sangha proved extremely successful, particularly in crushing a revolt in Tibet in 1280 and subsequently stationing garrisons, establishing an effective postal system, and pacifying the various Buddhist sects in the region. After the murder of Ahmad in 1282 and the execution of Lu Shih-jung in 1285, Sangha became the dominant figure in the government. As such, he attracted the same kinds of criticism that his predecessors had. He eventually was accused of corruption, theft of Khubilai's and the state's property, and disgusting carnal desires. Some of the most prominent men of the 1280s, including the renowned painter and official Chao Mengfu, opposed him and warned Khubilai of his nefarious intentions.108 It seems clear, however, that Khubilai prized Sangha's talents, for he continued to promote the Buddhist, naming him the minister of the right in December 1287. 107 YS, 15, p. 319. 108 Herbert Franke, "Sen-ge: Das Leben eines uigurischen Staatsbeamten zur Zeit Chubilai's dargestellt nach Kapitel 205 der Yiian-Annalen," Sinica, 17 (1942), pp. 9 0 - 1 0 0 . Luciano Petech, "Sang-ko, a Tibetan statesman in Yuan China," Ada Orientalia Academiat Scientiarum Hungaricae, 34 (1980), pp. 193—208. Cambridge Histories Online © Cambridge University Press, 2008 SANGHA AND ECONOMIC AND RELIGIOUS ABUSES 479 Which of Sangha's policies attracted the hostility of Chinese officials? One was his active support for foreigners in China. He served as a patron for Uighur scholars and painters; he persuaded Khubilai to halt a governmentsponsored campaign against Muslims; and he sponsored the founding of the National College for the Study of the Muslim Script in 1289. His role as protector of the foreigners did not endear him to the Chinese. Sangha's financial policies also drew hostile reactions. He introduced a higher tax on commerce and increased the price of salt, tea, and liquor. More onerous was his reform of the paper currency, which was threatened by a potentially devastating inflation. In April 1287, Sangha replaced the old notes with a new unit known as Chih-yuan ch'ao, which was named after Khubilai's reign title of Chih-yuan. The old currency would be exchanged on a five-to-one basis for the Chih-yuan notes. Those Chinese who were forced to exchange their less valuable old notes at less-than-satisfactory rates were thus incensed by the decline in their net worth. Sangha's reputation among the Chinese was particularly damaged by his apparent involvement with and support of a Buddhist monk named Yang Lien-chen-chia. Yang, who came from China's western regions and may have been a Tibetan or possibly a Tangut, had been appointed as the supervisor of Buddhist teachings in south China {Chiang-nan tsung-she chang shih-chiao), almost as soon as the Southern Sung had been toppled. IO9 In this office, he served under the jurisdiction of Sangha, who was in charge of Buddhist affairs for all of China. Yang constructed, restored, and renovated numerous temples and monasteries in south China, but he also converted some Confucian and Taoist temples into Buddhist ones. Such conversions generated great hostility among the Chinese. Even more upsetting to the Chinese were the methods that Yang used to raise funds for the construction and repair of the temples and monasteries. In 1285, he broke open the tombs of the Sung royal family and ransacked the valuables buried with emperors and empresses. He plundered 101 tombs and removed 1,700 ounces of gold, 6,800 ounces of silver, 111 jade vessels, 9 jade belts, 152 miscellaneous shells, and 50 ounces of pearls.' 10 Yang used these precious goods to pay for the erection and restoration of the Buddhist temples, and he also converted some of the palace buildings into Buddhist temples. To make matters worse, he employed forced laborers to rebuild and convert these temples and expropriated land from the big landowners to 109 On Yang Lien-chen-chia, see Herbert Franke, "Tibetans in Yuan China," in China under Mongol rule, ed. John D. Langiois, Jr. (Princeton, 1981), pp. 32—5. n o T'ao Hsi-sheng, "Yuan tai Mi-le pai lien chiao hui te pao tung," Shih huoyueh k'an, 1 (1933), pp. 132-3; Yen Chien-pi, "Nan Sung liu ling i shih cheng ming chi chu ts'uan kung h hui nien tai k'ao," Yen-ching hsiieh pao, 30(1946), pp. 28—36. Cambridge Histories Online © Cambridge University Press, 2008 480 THE REIGN OF KHUBILAI KHAN provide revenues for their maintenance. The southern landowners were infuriated by the high-handed expropriation of their lands and by the tax exemptions granted to the temples. The landowners also began to accuse Yang of profiteering, corruption, and womanizing. A more serious transgression of which Yang was accused was the desecration of the corpses of the Sung royal family. The body of one of the last emperors was said to have been exhumed, hung from a tree, and then burned, and as a final indignity the bones were reburied amidst the bones of horses and cows.111 Yang was blamed for this outrage, but official historians are so violently hostile to him that it is difficult to determine how much credence should be given to this account. Why would Yang deliberately and needlessly provoke the wrath of the southern Chinese by acts that violated and grated on Chinese sensibilities? Such a gratuitous deed hardly makes any sense and is scarcely credible. Yang's positive accomplishments can be gleaned only through inference. He was a devout Buddhist who tried to promote the interests of his religion, and Buddhism did indeed flourish in the south during his era. By 1291, there were 213,148 Buddhist monks and 42,318 temples and monasteries in the country, due partly at least to his patronage.112 Yet Yang's abuses rankled the southern Chinese and reflected on his patron Sangha. Both were, from the standpoint of the Chinese, exploitative and oppressive. The Chinese officials reviled them for their financial and personal misdeeds, and finally these accusations compelled action. On 16 March 1291, Khubilai relieved Sangha of his responsibilities and placed him under arrest. By August, the decision had been made to execute him.11* The last of the three villainous ministers now was dead, yet the three men's actions reflected on Khubilai, as the ruler who had appointed them. One minister after another had taken charge, and each had become for a while the virtual ruler of the country. Within a few years, however, each in turn was challenged, accused of serious crimes, and eventually either executed or murdered. Many lower officials doubtless wondered whether there was a guiding figure in China. Was Khubilai really in charge of his realm, and was he aware of the empire's affairs and of his subordinates' actions? He had begun to pursue policies that were on occasion diametrically opposed to those he had earlier upheld. The religious toleration that had been a cornerstone of his policies and had been vital to the Mongols' success appeared to have been abandoned. Problems with the religions of China intensified. 111 Paul Demieville, "Notes d'archeologie chinoise," Bulletin de I'Ecole Frangaise d'Extrem-Orimt, 25 (1925), pp. 458—67; repr. in h\%Cboixdlitudaiinologiquti(ig2i—1970) (Leiden, 1973), pp. 17—26. 112 Ratchnevsky, "Die mongolische Grosskhan und die buddhistische Kirche," p. 497. 113 YS, 16, p. 344. Cambridge Histories Online © Cambridge University Press, 2008 SANGHA AND ECONOMIC AND RELIGIOUS ABUSES 481 It may seem surprising, for example, that Khubilai in the late 1270s and early 1280s began to issue anti-Muslim regulations, for he had earlier recruited numerous Muslims for his government, and until 1282 Ahmad remained his supreme finance minister. But Khubilai may have recognized that the Muslim financial administrators had generated hostility in China. He may also have been concerned about the growing power of the Muslims in government. Whatever the motivation, he issued an edict in January 1280, prohibiting the Muslim method of slaughtering sheep and imposing the death penalty on transgressors."4 Shortly thereafter, he forbade the practice of circumcision. Khubilai's repressive acts were inspired more by political considerations than by a hatred of Islam. He was concerned about an overly powerful Muslim presence in government and about the possibility that their exactions might precipitate a rebellion. His anti-Muslim policy persisted until 1287." 5 By then it had become clear to Khubilai that foreign Muslim merchants would not continue to arrive in China if he persisted in persecuting Muslims. He therefore relented and rescinded his anti-Muslim edicts. The change in policy may also have been prompted by his recognition of the important positions occupied by Muslims. He himself had appointed one Muslim, Saiyid Ajall Shams al-DIn, as the governor of the province of Yunnan, where he had done a remarkable job in sinicizing that remote region without imposing Islam on its inhabitants. The achievements of such remarkable Muslims as Saiyid Ajall must have weighed heavily in Khubilai's decision to curb his anti-Muslim policies. Khubilai's policies toward the Buddhists and Taoists also aroused antagonisms. The Buddhist—Taoist debates of 1258, over which Khubilai had presided, had not concluded the strife between the two religions. Hostilities persisted into Khubilai's reign. In 1280, according to the Chinese accounts, some Taoist monks deliberately set fire to the Ch'ang-ch'un Taoist temple in Ta-tu and sought to place the blame on a Buddhist monk. The court assigned several officials to investigate, and they uncovered the Taoists' scheme. Two of the Taoists were executed; one had his nose and ears chopped off; and six others were exiled."6 This incident offered Khubilai a pretext to weaken the Taoists even further. Late in 1281, he ordered that all the Taoist texts other than Lao tzu's Tao te ching be burned and that the printing blocks be smashed and obliterated. Simultaneously, he forbade the Taoists from selling charms and compelled some of the monks to convert to Buddhism. Taoism survived, but its 114 YS, io, pp. 217—18. 115 Boyle, The successors of Genghis khati, p. 294. 116 Kenneth Ch'en, Buddhism in China: A historical survey (Princeton, 1964), p. 425. Cambridge Histories Online © Cambridge University Press, 2008 482 THE REIGN OF KHUBILAI KHAN political and economic influence was undercut. The Buddhists, who had gained a smashing victory, gloated over the defeat of their religious rivals and became increasingly assertive. Throughout the 1280s, the Buddhists acquired more and more wealth, land, and authority, and the sources are replete with the accounts of abuses by such Buddhists as Sangha and Yang Lien-chen-chia. These Buddhists began to alienate the Chinese, and the Mongols, as foreigners, were also tarnished by their deference toward and support of the Buddhists, particularly those from Tibet and other regions outside of China. DISASTROUS FOREIGN EXPEDITIONS Khubilai's difficulties within China presaged similar catastrophes abroad. A lack of control characterized both domestic and foreign policies. The balanced executive authority that Khubilai had exercised seems to have been dissipated. Ill-considered decisions tended to be the rule rather than the exception. As both the emperor of China and the khan of khans, Khubilai encountered relentless pressure to prove his worth, virtue, and acumen by incorporating additional territory into his domains. He thus initiated several ill-conceived and foolhardy foreign ventures. The second invasion of Japan The most prominent such venture was the new expedition to Japan. After the failure of the first expedition in 1274 and repeated rebuffs by the Japanese shogun of Khubilai's invitations to send tribute embassies to China, Khubilai prepared to launch another invasion of Japan. Seven years elapsed, however, before he could actually dispatch an expedition; only after the pacification of the Southern Sung could he once again turn his attention to Japan. Khubilai chose a multiethnic leadership for the campaign - a Korean as the admiral, Fan Wen-hu as the Chinese general, and Hsin-tu as the Mongolian general. He provided his generals with a massive invasion force: 100,000 troops, 15,000 Korean sailors, and 900 boats." 7 The Yuan military command planned a two-pronged assault on the Japanese islands (see Map 35). Forty thousand troops from north China, transported in Korean ships, were to link up on the island of Iki with forces departing from Ch'iian-chou in Fukien and would then jointly attack the rest of Japan. The soldiers from the north, however, set forth alone in the spring of 1281 because the larger southern force had encountered delays. By June, 117 YS, 11, pp. 226, 228. Cambridge Histories Online © Cambridge University Press, 2008 LIAO-YANG HSING-SHENG 1274 expedition 75 000 Mongolian troops 8000 KoryB troops 900 ships 1281 expedition CHIANG-NAN FLEET 700 000 Chinese troops 3500 ships Koryo court leaves island and returns ' to Kaegyong \110M 1281 expedition EASTERN ROUTE FLEET 30 000 Mongolians 10 000 KoryB troops 900 ships KORYO Mongolian vassal state 1258 Under Mongolian domination 1270 Last resistance in south crushed 1273 ••.1281 EASTERN ROUTE / 1ZA»anj FLEET CHEJU I. Underdirect Mongolian rule Ninqpo (Ch'ing-Yiian)* M A P 3 5 . Khubilai's invasions of Japan Cambridge Histories Online © Cambridge University Press, 2008 484 THE REIGN OF KHUBILAI KHAN the northern forces had occupied Iki, and so they then headed for the main island of Kyushu. Meanwhile, the southern troops had finally completed their preparations and headed for a rendezvous in Kyushu. With such a powerful military machine and with such momentum, victory ought to have been within their grasp. 1 ' 8 Yet the expedition failed miserably. Tensions between the Mongolian commanders and the Chinese leaders weakened the expeditionary force. The Chinese troops, who comprised the vast majority of the soldiers, had scarcely any stake in the combat and so did not exert themselves to put up a strong fight. Moreover, when they landed in Kyushu, they were afforded no protection from either their enemies or the elements. They had to camp out in the open without a castle, stronghold, or town that would provide adequate defense and from which they could make forays against the enemy. The two armies fought for almost two months, but no clear victor emerged. Then a natural catastrophe interceded to dash the hopes of the Mongols. On 15 August 1281, a typhoon struck the coast of Kyushu, and almost one-half of the Mongolian force perished during the storms. The Japanese maintained that the typhoon was sent by the gods to protect their country, that it was a "divine wind" (Japanese: kamikaze) and a guarantee that the gods would never allow Japan's enemies to conquer and occupy their territory. For the Mongols, the defeat was a devastating shock. Their failure shattered their mantle of invincibility, as Khubilai's subjects now realized that the Mongols were vulnerable. One of the principal underpinnings of their power — the psychological edge of terror that they held over their opponents — was badly shaken, if not destroyed. Most damaging of all was the tremendous expense incurred in the expedition, which precipitated some of the revenue problems that compelled Khubilai to turn to such financial administrators as Ahmad and Lu Shih-jung, whom the Chinese hated and despised. Khubilai's efforts once again reveal the same lack of control noticeable in his financial policies of the 1280s. Campaigns in Southeast Asia Similarly, Khubilai's campaigns in Southeast Asia were, for the most part, ill conceived and resulted in some devastating reverses. These expeditions were inspired by the Mongols' traditional expansionist ambitions as well as Khubilai's personal need for new conquests to give him additional claims to legitimacy. Neither he nor the court considered the tropical heat and rain, 118 A useful summary of the Japanese campaign is given in Chen Pang-chan, comp., Yuan sbih chi sbib pea mo (Peking, 1979 ed.), pp. 25—31. Cambridge Histories Online © Cambridge University Press, 2008 DISASTROUS FOREIGN EXPEDITIONS 485 the dense forests, and the diseases of the south, to all of which the Mongols were unaccustomed. Hostilities had first erupted in Southeast Asia as early as the 1270s (see Map 36). In 1273, Khubilai had sent three emissaries to the kingdom of Pagan in Burma to request the dispatch of tribute to China. King Narathihapate (r. 1256-87), a vain despot who described himself as "Supreme commander of 36 million soldiers, swallower of 300 dishes of curry daily" and sexual mate of 3,000 concubines,"9 executed the hapless envoys for daring to suggest that he humble himself to the great khan. In 1277, Khubilai ordered Nasir al-Din, the son of his trusted Muslim retainer Saiyid Ajall, to lead an expedition to Pagan to avenge the murder of the three envoys. In the critical battle, Narathihapate had the advantage of employing elephants. Nasir alDln instructed his archers to take aim at the elephants, which were totally unprotected. The elephants stampeded, and the tide turned to the Mongols' advantage.120 But Narathihapate himself was not captured, and Nasir al-Dln returned to China without having pacified Pagan. Khubilai then turned to Champa (then roughly equivalent to modern South Vietnam), whose king, Jaya Indravarman VI, was hostile to the Mongols. The monarch of Champa refused to accompany a tribute mission to the Yuan court and rebuked several of Khubilai's envoys. In 1281, Khubilai responded by sending Sodu (So-tu), one of his leading officials in Canton, on a punitive expedition against him. Sodu, commanding a force of one hundred ships and five thousand men, landed on the coast of Champa, but the guerrilla warfare that the king employed prevented the Mongols from making any headway. Khubilai then decided to seek the cooperation of the kingdom of Annam (present-day North Vietnam) against its neighbor to the south. Its king, Tran Thanh-ton, although he had sent embassies to Khubilai's court, was not, however, eager to have Mongolian troops cross his land to reach Champa, and so he toofiercelyresisted the Mongolianforcesled by Sodu and Khubilai's son Toghon. Guerrilla warfare, heat, and disease took their toll on the invaders, and in a decisive battle fought at Ssu-ming on the border with Yunnan in the summer of 1285, the Mongols were defeated and Sodu was killed. A second expedition, led by Nasir al-Dln and Khubilai's grandson Esen Temiir, later joined by Toghon, campaigned in Annam in 1286 and 1287 and even reached Hanoi but was forced to withdraw because of the heat and the unfavorable environment, much to Khubilai's anger. The Annamese king and the king of Champa recognized, however, that the Mongols would continue to plague them unless they offered pro forma acquiescence to the 119 Shelley Mydans and Carl Mydans, "A shrine city, golden and white: The seldom-visited Pagan in Burma," Smithsonian Magazine, October 1974, p. 79. 120 Moule and Pelliot, Marco Polo, vol. 1, pp. 2 8 9 - 9 0 . Cambridge Histories Online © Cambridge University Press, 2008 486 THE REIGN OF KHUBILAI KHAN Sodu and Toghon defeated by Annam 1285 Invasion of Annam 1286-89 EsenTemur and Toghon ing-yuan M A P 36. Khubilai's invasions of Southeast Asia Cambridge Histories Online © Cambridge University Press, 2008 DISASTROUS FOREIGN EXPEDITIONS 487 great khan. They therefore began to send tribute, and the Mongolian expeditions against them ended in the late 1280s. In 1287, Khubilai sent one more expedition, led by his grandson Esen Temiir, against the "insolent" king of Pagan. The campaign was successful; the king, Narathihapate, was overthrown; and his successors offered tribute to the Mongolian court.121 The expedition, however, had been costly, and it is difficult to justify the Mongolian expenditures. It is even more difficult to justify the more renowned and spectacular campaigns against Java. In 1289, Khubilai sent an ambassador named Meng Ch'i to seek the submission of Java. The Javanese king, Kertanagara, fearful that the Yiian court wished to take away his control over the spice trade in Southeast Asia, responded by branding the face of the unfortunate envoy. Khubilai used this incident as a pretext to initiate a military expedition against Java. He appears not to have learned his lesson from the failure of his naval expedition against Japan. Again, he outfitted a large seagoing force to punish a ruler who had done violence to one of his envoys. In 1292 the expedition set forth from Ch'iian-chou, and early in the following year most of the troops landed in Java. They were so successful initially that a Javanese rebel took advantage of Kertanagara's weakened position to attack and kill him. Kertanagara's son-in-law Prince Vijaya offered his "submission" to the Mongolian forces and pledged to assist them in crushing the rebels. And he did so but then betrayed the Mongolian expedition. He ambushed a large Mongolian detachment, forcing the rest of the troops tofleeand to set sail for China."3 Yet another of Khubilai's expeditions had resulted in failure, and its vast expense contributed further to the court's growing fiscal problems. Rebellions in Tibet and Manchuria Perhaps a more damaging confirmation of Khubilai's gradual loss of control was the outbreak of rebellion in regions theoretically under his direct jurisdiction. These challenges to his authority were blows to Khubilai's claims as both emperor of China and khan of khans. The first outbreak occurred in Tibet a few years after the death in 1280 of Khubilai's staunchest Tibetan ally, the 'Phags-pa lama. The 'Bri-gung, a rival of the 'Phags-pa lama's Saskya sect, rebelled in 1285, and Khubilai needed to send a punitive force in 1290 to restore Sa-skya and, ultimately, Mongolian authority."3 A second 121 YS, 14, pp. 286, 289, 303; 15, pp. 311, 326—7, 330; 16, p. 333. 122 On the Java campaign, see the dated but still not wholly superseded article by W. P. Groeneveldt, "The expedition of the Mongols against Java in 1293 A . D . , " China Review, 4 (January—February 1876), pp. 2 4 6 - 5 4 . 123 Turrell V. Wylie, "The first Mongol conquest of Tibet reinterpreted," Harvard Journal of Asiatic Studies, 37 (1977), pp. 1 3 1 - 2 . Cambridge Histories Online © Cambridge University Press, 2008 THE REIGN OF KHUBILAI KHAN rebellion erupted in Manchuria, led by its Mongolian commander, Nayan, who allied himself with Khubilai's Central Asian enemy, Khaidu. Khubilai must have believed that the threat posed by Nayan was critical, for he personally led the campaign against the dissident leader. Marco Polo claimed that his forces consisted of 460,000 men, a figure that is certainly greatly inflated.I34 In any case, Khubilai recruited a sizable force, and despite his age and his ailments, he had himself carried to the battlefield in a palanquin mounted on the backs of four elephants. In the autumn of 1287, the two sides faced each other, and for a full day they were evenly matched, but late in the day the tide turned against the Nestorian Christian Nayan. Khubilai captured him and had him executed. KHUBILAI'S LAST YEARS Despite this victory, Khubilai had, in the 1280s, generally encountered frustrations in his dealings with foreign lands. These years also brought him personal tragedy and grief. His favorite wife, Chabi, who had long been a source of support and sound advice, died in 1281, and her son and his own designated successor, Chen-chin, succumbed in 1285. Chen-chin had been carefully groomed to be the next emperor of China and his successor as khan. His early death while still in his forties must have shattered Khubilai and contributed to the gloom at his court. He turned more and more to drink and food for comfort. He drank excessively, which contributed to his health problems. Obesity and gout plagued him for his remaining years. The Chinese sources reveal that by the end of his life, he was extremely dispirited and depressed. He weakened during the winter of 1293—4, anc^ o n x ^ February he died in his palace. The achievements of Khubilai's reign were remarkable. Like the other Mongolian khans, Khubilai continued to embark on military campaigns. His most stunning victory was his conquest of Southern Sung China, a territory with over fifty million people and vast wealth and resources at its command. This campaign, which required more sophisticated planning and logistics than had many earlier Mongolian campaigns, ensured Khubilai's status as a great commander among the Mongols. But his political achievements were probably as impressive. He wished to persuade the Chinese that he was becoming increasingly sinicized while maintaining his Mongolian credentials with his own people. He set up a civilian administration to rule, built a capital within China, supported Chinese religions and culture, and devised 124 See also Paul Pelliot, Notes on Marco Polo (Paris, 1963), vol. 2, pp. 7 8 8 - 9 ; and Jozef L. Mullie, De Mongoolse Prim Nayan (Brussels, 1964). Cambridge Histories Online © Cambridge University Press, 2008 KHUBILAI'S LAST YEARS 489 suitable economic and political institutions for the court. Yet he did not abandon his Mongolian heritage. Khubilai retained numerous Mongolian customs, employed Mongols in key positions in government and the military, and dispensed with the civil service examinations so that he would not be limited to Chinese for government posts. Despite his flaws and the difficulties he faced in the last decade of his reign, Khubilai left his successors a stable and generally prosperous state. Cambridge Histories Online © Cambridge University Press, 2008 CHAPTER 6 MID-YUAN POLITICS INTRODUCTION "Mid-Yuan" refers to the period between the death of the great dynastic founder, Khubilai khan (Emperor Shih-tsu, 1260—94), in 1924 and the accession of the last Yuan monarch, Toghon Temiir khaghan (Emperor Shun-ti, r. 1333-68), in 1333. During these thirty-nine years, nine khaghans ascended the throne in quick succession, resulting in frequent bureaucratic turnovers and reversals of state policies. It was, therefore, a politically volatile and turbulent period, the events of which gradually but steadily emasculated the great and powerful empire left behind by Khubilai and paved the way for the downfall of the dynasty under Toghon Temiir. The mid-Yuan khaghans inherited from Khubilai not only a great empire but also its multifarious problems.1 Khubilai had successfully transferred to China the center of gravity of the previously steppe-based Yeke Mongghol Ulus or "Great Mongolian Nation" and made the Yuan the first dynasty of conquest to rule the whole of China as well as the Inner Asian steppes. He had further provided his far-flung multiracial and multicultural empire with a workable institutional framework by synthesizing Chinese and Inner Asian systems and adorning it with the symbols of legitimacy drawn from Mongolian, Chinese, and Buddhist sources.2 Because of all this, Khubilai was subsequently regarded by his heirs as the most revered dynastic founder and ancestor whose mantle should be strictly preserved. The problems left by Khubilai to his heirs, nevertheless, were as great and complex as the empire itself. 1 The assessment of Khubilai's role and legacy in the following paragraphs is largely based on the following works: Yao Ts'ung-wu, "Yuan Shih-tsu Hu-pi-lieh han: T'a te chia shih t'a te shih tai yii t'a tsai wei ch'i chien chung yao ts'o shih," in vol. 6 of Yao Ts'ung-u/u hsien sheng ch'iian chi, ed. Ch'en Chieh-hsien and Cha-ch'i Ssu-ch'in (Taipei, 1972), pp. 399—416; Chou Liang-hsiao, Hu-pi-lieh (Ch'ang-ch'un, 1986); Zhou Liangxiao (Chou Liang-hsiao), "On Khubilai khan," Social Sciences in China, 2 (1981), pp. 177—94; Psi Kang, "Kuan yii Hu-pi-lieh fu hui Han fa te li shih k'ao ch'a," Chung-kuo shih yen chiu, 4 (1981), pp. 93—107; Otagi Matsuo, Hubirai han (Tokyo: 1941); Morris Rossabi, Khubilai khan: His life and times (Berkeley and Los Angeles, 1988). 2 Herbert Franke, From tribal chieftain to universal emperor and god: The legitimation of the Yuan dynasty (Munich, 1978). 49O Cambridge Histories Online © Cambridge University Press, 2008 INTRODUCTION 491 Fundamentally, the position of the Yuan state in both China and the broader Mongolian empire needed to be carefully reconsidered and defined, and further adjustments made accordingly. Did China constitute the most important part of the empire and its primary object of rule? Or was it merely a part of the universal Mongolian empire to whose maintenance all of China's resources should be devoted? The problems of satisfactorily answering these questions had serious and far-reaching implications for the Yuan rulers. On the one hand, Khubilai's claim to be the khaghan of all Mongols was difficult to substantiate and was seriously challenged from the start by his Inner Asian cousins. The Yuan were still at war with the Ogodei and Chaghadai khanates at the time of his death. On the other, the pulls of the steppe tradition made Khubilai reluctant to strike deeper roots in China by completely basing his dynasty on the traditional Chinese model of government. This slowed down the process of acculturation within the Yuan state and left a number of important problems to his heirs. Politically, the system of government created by Khubilai was the product of a compromise between Mongolian patrimonial feudalism and the traditional Chinese autocratic—bureaucratic system. Under the original Mongolian system, imperial relatives and the meritorious retainers of the imperial family all enjoyed hereditary political, economic, and military privileges. Even though Khubilai did try to curtail the privileges of these aristocrats, he did not go far enough in this direction, for fear of violating one of the fundamental principles of the Mongolian state, the idea that the empire was the joint patrimony of all the descendants of Chinggis khan. Consequently, the aristocrats continued to enjoy much influence and advantages in every sphere of public life, which proved to be a serious source offinancialstrain and political instability under his heirs. In the meantime, Khubilai's attempts to centralize and bureaucratize his government, however imperfect they were, enabled the bureaucrats to acquire a much larger share of power than had their predecessors in the pre-Khubilai era. The bureaucrats were to become an important focus of power, often overshadowing the imperial princes and even competing with the throne for power in the post-Khubilai era. Culturally and socially, the vast gap between the Mongolian and se-mu (Western and Central Asian) elites and their Chinese subjects needed to be bridged. This gap existed partly because the Chinese in general suffered political and social discrimination under an ethnically based system of social stratification. It was especially serious among the educated Chinese elite because Confucian learning and scholars were not given the degree of esteem that they had been accorded previously under native Chinese dynasties. The Yuan government recruited the members of its ruling elite on the basis of ascription rather than achievement. There was, therefore, little incentive for Cambridge Histories Online © Cambridge University Press, 2008 492 MID-YOAN POLITICS the Mongols and the se-mu to absorb Chinese learning. Consequently, although the traditional Chinese elite (the Confucian scholars) were not given their share of power, the Mongolian and the se-mu elites remained, to a significant extent, strangers to the Chinese cultural tradition and socially isolated from the native populace. This dichotomy gave the Yuan regime a strong "colonial" coloration. Diplomatically and militarily, the Yuan foreign policy under Khubilai, as under the previous Mongolian khaghans, had been "imperialistic," always ready to mobilize all available human and material resources under its control in a continuous quest for a greater empire.3 It had become clear by the last years of Khubilai's reign, however, that foreign conquests had reached the point of diminishing returns, for the empire had already attained its maximum geographical limit. Was it now the right time to begin to concentrate all energies on internal consolidation rather than on external expansion? Finally, the Yuan government had been plagued by chronic financial problems from the outset, owing to its excessive military expenditures, lavish imperial grants, and other court expenses and so had been forced to rely on se-mu financial experts to raise the necessary funds by exploitative methods. Consequently, there were recurrent conflicts between these financial experts and the advocates of Confucian ideas of frugality in government and the lightening of the people's tax burden. A drastic curtailment of governmental spending would certainly have been financially desirable but would have been politically difficult to implement, as it would have affected the very basis of the Mongolian state in China. In sum, great as his achievements undoubtedly had been, Khubilai's legacies were not entirely positive. The challenges faced by his mid-Yuan heirs were not merely a question of preserving the mantle of his government but also one of responding creatively to the problems left behind by the great dynastic founder and also to the new stresses that appeared after his death. What was required was to strike deeper roots for the regime in China by reforming it further along traditional Chinese lines and ameliorating some of its colonial characteristics yet at the same time preserving dynastic security and the privileges of the conquering elites. T H E REIGN OF TEMOR KHAGHAN (EMPEROR C H ' E N G TSUNG), 1294-1307 A more specific but still important problem that Khubilai failed to resolve for his heirs was that of the imperial succession, which had always been a 3 Thomas T. Allsen, Mongol imperialism: The policies of the Grand Qan MSngte in China, Russia and Islamic lands, 1251—1239 (Berkeley and Los Angeles, 1987), p. 7. Cambridge Histories Online © Cambridge University Press, 2008 THE REIGN OF TEMOR KHAGHAN 493 source of instability and tension in Mongolian imperial history. Khubilai's inability to clear up this problem was partly the result of the incompatibility of the Mongolian concept of the empire as the joint patrimony of all of Chinggis khan's descendants and the Chinese concept of autocratic rule. It was also partly the result of some specific incidents that occurred during his reign.4 Because Khubilai had not been able, nor had he intended, to recast completely the patrimonial—feudal Mongolian state in the Chinese mold of an autocratic system of government, the traditional claim of the imperial princes to elect a new khaghan in an assembly (khuriltai) could not be denied whenever the matter of succession arose.5 Khubilai's attempt to assert his prerogative as emperor in the Chinese fashion by designating his eldest son by his principal wife as the heir apparent had been aborted when his designated heir apparent, Chen-chin (124385), died in 1285, nine years before Khubilai himself died. 6 Consequently, in the post-Khubilai era, the throne was frequently and often bitterly contested between those who based their claims on the nomination made by the previous khaghan and those who claimed their rights to be "elected." The pattern of contention for the throne in the post-Khubilai era differed from that of earlier times in three ways. First, the scope of contestants was now limited to the descendants of Khubilai, especially, and almost exclusively, to those of Chen-chin. Because it was the common understanding among the Mongols that the Yuan dynasty was the creation of Khubilai, the members of other branches of the imperial clan could take part in the election of a new emperor as the supporters of one or other of the contestants for the throne, but they could not enter the contest as candidates themselves. That the descendants of Chen-chin enjoyed the foremost claim to the throne was based on the fact that Chen-chin was Khubilai's eldest son and his designated heir. 4 For a comprehensive study of the succession crises of the Yuan dynasty as a whole, see Hsiao Kungch'in, "Lun Yuan tai huang wei chi ch'eng wen t'i," Yuanshih chipeifangmin tsushihyenchiu chi k'an, 7 (1983), pp. 2 2 - 3 9 ; see also his study on the succession crises of the period of the Yeke Mongghol Ulus (i.e., the Mongolian empire before Khubilai's accession), "Lun Ta Meng-ku kuo te han wei ch ch'eng wei chi," Yuan shih chi pei fang min tsu shih yen chiu chi k'an, 5 (1981), pp. 4 8 - 5 9 . On the lack of an effective system of succession in nomadic societies in general, see Joseph Fletcher, "Turco-Mongolian monarchic tradition in the Ottoman empire," Harvard Ukrainian Studies, 3 - 4 ( 1 9 7 9 - 8 0 ) , pp. 236—51, and "The Mongols: Ecological and social perspectives," Harvard journal of Asiatic Studies, 46(1986), pp. 11-50, esp. pp. 1 7 - 1 8 , 2 4 - 8 . 5 On the khuriltai's function in "electing" the Mongolian khaghans, see Yanai Wataru, "Moko no kokkai sunawachi 'kurirutai' ni tsuite," in Shigaku zaisbi, 28, (1917), no. 4, pp. 3 2 1 - 4 7 ; no. 5, pp. 4 5 7 - 8 4 ; no. 7, pp. 6 8 8 - 7 2 2 ; repr. in his Moioshi kenkyu, ed. Iwai Hirosato et al. (Tokyo, 1930), pp. 361-447; Chou Liang-hsiao, "Meng-ku hstian han i chih yii Yuan ch'ao huang wei chi ch'eng wen t'i," Yuan shih lun ts'ung, 3 (1986), pp. 31—46; Elizabeth Endicott-West, "Imperial governance in Yuan times," HarvardJournal of Asiatic Studies, 46, (1986), pp. 523—49, esp. pp. 5 2 5 - 4 0 . 6 On Chen-chin's death and its impact, see Huang Shih-chien, "Chen-chin yii Yuan ch'u cheng chih," Yuan shih lun ts'ung, 3 (1986), pp. 193-204. Cambridge Histories Online © Cambridge University Press, 2008 494 MID-YOAN POLITICS Second, the powerful bureaucratic faction in the court played an increasingly important role in creating new khaghans at the expense of the imperial clansmen.~> In the traditional Mongolian system it was the imperial princes and consorts who had decided the "electoral" process. Officials (noyad) had participated in the khuriltai as the retainers of imperial clansmen, but they were never allowed to have an independent voice. The increasingly important role played by the bureaucrats in this regard was clearly the result of Khubilai's effort to centralize and bureaucratize the government and the consequent weakening of the patrimonial—feudal character of the Yuan government. Third, the succession to the throne was often contested, and it alternated between candidates based in China, who normally enjoyed bureaucratic support at the capital, and candidates based in the steppe, who usually commanded a powerful field army. Such alternations between candidates of sharply different interests and backgrounds, coupled with a series of very short reigns, led to wide swings in the government's policies generally and in its cultural orientation particularly, making the politics of the period very volatile. TEMUR'S SUCCESSION The first succession crisis broke out immediately after Khubilai's death in February 1294, for Khubilai had never established his grandson and Chenchin's second son, Temiir (1265-1307), with full credentials as his heir. After the death of Chen-chin in 1285, the aging Khubilai khaghan apparently could not decide on a successor for quite some time. According to the Persian historian Rashld al-Dln, it was Temiir's mother and Chen-chin's principal wife, Kokojin (K'uo-k'uo-chen, also known as Bairan ekechi, or Po-lan yeh-ch'ieh-ch'ih; d. 1300), who was the active promoter of her son's cause.8 However, even though Temiir was given the seal of the heir apparent in July 1293 when he was sent to Mongolia as the princely overseer of all forces in the steppe,» he was not given the panoply of an heir apparent that had previously been accorded to his father. This may have reflected Khubilai's reservations about his grandson's suitability as the future khaghan, for he knew very well that Temiir was a drunkard.10 Thus the question of succession still hung in the balance at the time of Khubilai's death. 7 See Hsiao Kung-ch'in, "Lun Yuan tai huang wei chi ch'eng wen t'i," p. 30. 8 John Andrew Boyle, trans., The successors of Genghis khan: Translated from the Persian ofRasbid al-Din (New York, 1971), pp. 3 0 0 - 1 . According to Chinese sources, it was the Uighur Buddhist-Confucian official Atghun Sali (A-lu-hun Sa-li, 1245-1307) who originally suggested to Khubilai that Temiir would be a suitable heir apparent. See Sung Lien et al., eds., Yiiansbih (Peking, 1976), 130, p. 3177 (hereafter cited as YS). 9 YS, 18, p. 381. 10 Boyle, The successors of Genghis than, p. 301; YS, 154, p. 3261. Cambridge Histories Online © Cambridge University Press, 2008 TEMOR'S SUCCESSION 495 Temiir's rival for the throne was his elder brother Kammala (Kan-ma-la, 1263—1302). As the eldest son of Chen-chin, Kammala was as eligible as Temiir was to be their grandfather's successor. Moreover, he enjoyed as much military backing as did his younger brother, for he had served as commander in Mongolia for many years and had been appointed in 1292 as the prince of Chin to take charge of the "four ordos of Chinggis khan and the troops and land of the Tatar."11 However, Temiir apparently enjoyed an advantage over his elder brother: Not only had he been designated by Khubilai as the heir apparent, but he also had Kokojin's blessing. This was important because she had been favored by Khubilai for her virtue and was widely influential. Although the identity of Kammala's mother remains uncertain, the fact that Kokojin sided with Temiir may suggest that Kammala was not her own son.12 Equally significant was the important bureaucratic support that Temiir commanded. The three executors of his will appointed by Khubilai - Us Temiir (Yii-hsi T'ieh-mu-erh, 1242—95), the grandson of Chinggis khan's distinguished companion Borju (d. 1227) and the manager (chih-yiian) of the Bureau of Military affairs (Shu-mi yuan); Bayan (Po-yen, 1236—94) of the Barin tribe, the conqueror of the Sung and for many years the commander in chief of the armed forces on the northern and northwestern frontiers; and Bukhumu (Pu-hu-mu, 1255—1300), a Turkish Qangli Confucian statesman and manager of governmental affairs (p'ing-chang cheng-shib) of the Secretariat (Chung-shu sheng) - all had established close relationships with Temiir or his mother in one way or another. 1 ' In addition to this trio, Oljei (Wan-tse, 1246-1303), the grand councillor of the right of the Secretariat, had accompanied Temiir to Mongolia as a senior adviser.14 It is not known whether Khubilai's will contained any specific instruction to put Temiir on the throne, but all these powerful bureaucrats apparently worked in that direction. When the khuriltai was convened on 14 April 1294 at Shang-tu, the summer capital, there was no lack of supporters for Kammala's candidacy.1' Nonetheless, they were outpowered and outmaneuvered by Temiir's support11 YS, 29, p. 637; 115, p. 2895. 12 Although Kammala's biography in the Yuan sbib states that he was the son of Kokojin, Kokojin's biography in the same work fails to mention him among her sons. See YS, 115, p. 2893, and 116, pp. 2896—7; see also Chou Liang-hsiao, "Meng-ku hsiian han i chih yii Yuan ch'ao huang wei chi ch'eng wen t'i," p. 43. 13 See YS, 119, p. 2498; 127, p. 3115; 130, pp. 3 1 6 3 - 7 3 ; Su T'ien-chueh, Yuan ch'ao mingcb'tn sbih lu'th (Chien-an, 1335; repr. Peking, 1962), 4, p. 12a; Francis W. Cleaves, "The biography of Bayan of the Barin in the Yuan shib," Harvard Journal of Asiatic Studio, 19 (1956), pp. 185—303, esp. pp. 269—70. 14 YS, 130, pp. 3 1 7 3 - 4 . 15 On Rashid al-DTn's interesting but totally unconvincing account of the way in which the dispute was resolved in the khuriltai, see Boyle, The successors of Genghis khan, p. 321. Cambridge Histories Online © Cambridge University Press, 2008 496 MID-YOAN POLITICS ers. Indeed, Us Temur is said to have persuaded Kammala to lead the way in urging Temiir to accept the crown.l6 In the meantime, Bayan, the great war hero, took a more forceful posture in making Temiir's candidacy acceptable to the dissenters. It is said that standing on the steps of the audience hall with a sword in his hand, Bayan recited Khubilai's will and explained why Temiir should be enthroned. As a result, we are told, "the Princes of the Blood, trembling, hurried to the hall to make obeisance."17 It is clear that the prestige of Bayan and Us Temiir and the powerful military and bureaucratic machinery behind them accounted for much of the influence that they exerted in the khuriltai. The decisive role played by these leading bureaucrats in deciding who should be the new khaghan was unprecedented in the history of the Mongols, portending the many bureaucratic kingmakers of the future. THE CONSERVATOR OF KHUBILAI S ACHIEVEMENTS Temiir's thirteen-year reign began on a conservative note. As the direct inheritor of Khubilai's empire, Temiir khaghan expected, as did his subjects, to be the conservator of his grandfather's mantle. In the edict proclaiming his enthronement issued on 10 May 1294, the new khaghan officially announced his intention to preserve the established pattern (cb'eng-kuei) of his grandfather's reign.18 This was to become the fundamental principle of his government. The conservative nature of Temiir's reign is clearly shown by its continuity with Khubilai's last years, in terms of both its principal government personnel and its major policy lines. In addition to honoring the senior Mongolian statesmen of Khubilai's reign with lofty honorific titles, Temiir khaghan retained in his government most of the ministers of the Secretariat who had been appointed by Khubilai after the downfall of the powerful Tibetan Sangha (Sang-ko) in 1291, to mollify the effects of Sangha'sfiscalpolicies.'9 Oljei remained the grand councillor of the right (yu ch'eng-hsiang) and was the sole grand councillor until Harghasun (Ha-la-ha-sun, 1257-1308) was appointed as the grand councillor of the left (tso ch'eng-hsiang) in 1298. Oljei, the leading figure in Temiir's early administration, was a senior statesman known principally for his caution and leniency rather than for his ability.20 Harghasun, who became the most influential figure in the government after Oljei's death in 1303, was from a distinguished Mongolian 16 17 18 19 Su Tien-chiieh, Kuo ch'ao wen UiiSPTK ed.), 23, p. 7a. Cleaves, "The biography of Bayan of the Barin in the Yuan shih," pp. 270— 1. YS, 18, p. 3 8 1 - 2 . YS, 112, pp. 2 8 0 3 - 1 3 ; T'u Chi, Meng-wu-erb tbih cbi (Wu-ch'm, 1934; repr. Peking, 1958), 157, pp. 13a-18a. 20 On Oljei, see Su T'ien-chiieh, Yiian ch'ao ming cb'en sbib lueb, 4, pp. ia—2a; YS, 130, pp. 3173—4. Cambridge Histories Online © Cambridge University Press, 2008 THE CONSERVATOR OF KHUBILAl'S ACHIEVEMENTS 497 Oronar family and was well known not only for his ability and uprightness but also for his sympathy for Confucianism.21 Another influential figure early in Temiir's reign was the Qangli statesman Bukhumu. 22 Having been brought up in the palace and educated by a leading Neo-Confucian scholar, Hsu Heng (1209-81), in the School for the Sons of the State (Kuo-tzu hsiieh), Bukhumu was one of the best-known non-Chinese Confucian scholar-officials in the entire Yuan period. First serving as the manager of governmental affairs (p'ing-chang cheng-shih) and then as the vice-censor in chief (yu-shih chung-ch'eng), his main role was that of Confucian advocate and remonstrator in Temiir's court. These senior Mongolian and se-mu statesmen were assisted by an array of Chinese administrators and Muslim financiers. The Chinese officials included Ho Jung-tsu, Chang Chiu-ssu (1242-1302), Liang Te-kuei (1259-1304), and Ho Wei (1254-1310), who were of varied ideological persuasions and political backgrounds. Among the Muslim financiers, Majd al-Din (Maichu-ting) and All (A-li) had been in Ahmad's and Sangha's administrations under Khubilai. 2^ The most prominent Muslim, however, was the greatgrandson of the renowned Muslim statesman Saiyid Ajall Shams al-Din (1211—79), Bayan, who is described by Rashid al-DIn as "an extremely great and powerful vizier" in charge of the Sahib divan, which in Persian usually refers to a Ministry of Finance.24 It is apparent that Temiir was continuing Khubilai's policy of maintaining a racially and ideologically balanced cast of central administrators. Under Oljei and Harghasun, the administration adopted policies that were designed to ensure political and social stability. In the fiscal field, the cancellation of tax debts that had accumulated since the beginning of Khubilai's reign was a policy that had already been adopted after Sangha's downfall.25 Such cancellations were continued in Temiir's reign, and exemptions from levies and taxes were granted several times for part or all of the country.26 Temiir khaghan was strongly opposed to imposing any additional fiscal burden on the people in order to increase revenues, and accordingly in 1302 he prohibited the collection of anything beyond the established tax quotas. 27 Ideologically, Temiir's administration showed respect for Confucianism 21 Su T'ien-chiieh, Kuo ch'ao wen lei, 25, pp. la—10a; Su T'ien-chiieh, Yuan ch'ao ming ch'en shih liieh, 4 , pp. 2b—7b. 22 On Bukhumu, see Chao Meng-fu, Chao Meng-fu chi, ed. Jen Tao-pin (Hangchow, 1986), pp. 158—62; YS, 130, pp. 3 1 6 3 - 7 7 3 . 23 See Yang Chih-chiu, "Yuan tai Hui-hui jen te cheng chih ti wei," in his Yiian shih san fan (Peking, 1985), pp. 2 4 5 - 8 2 , esp. pp. 2 5 4 - 9 . 24 Boyle, The successors of Genghis khan, p. 322. 25 YS, 18, p. 383. 26 YS, 96, pp. 2 4 7 0 - 6 . 27 YS, 20, p. 440. Cambridge Histories Online © Cambridge University Press, 2008 498 MID-YtiAN POLITICS and Confucian scholars. Shortly after his accession, Temvir issued an edict to revere Confucius.28 Mainly under the instigation of Harghasun, a new temple of Confucius was built in Ta-tu, where the School for the Sons of the State was relocated, and it increased its enrollment of students.29 Harghasun was particularly close to Confucian scholars and is said to have always sought their advice on important policy decisions.'0 SIGNS OF DECLINE In the Yuan shih, Temiir's reign is coupled with that of Khubilai as the peak of the Yuan dynasty. 31 But there already were signs showing that his reign marked the beginning of the decline of the Yuan government. The khaghan himself lacked his grandfather's kind of intellectual and physical vigor that was necessary to provide adequate imperial leadership. Nor was his minister Oljei strong enough to provide substitute leadership. The government seems to have lost both administrative vigor and fiscal health under the two men's excessively indulgent and procrastinating administration. There was a huge expansion of government personnel. Even though the combined quota for officials at the court and the capital was set at 2,6oo,32 the Censorate reported in 1294 that the number of officials on the payroll in the capital amounted to 10,000 and that there were even more in the provinces.33 In fact, the situation became so serious that the Secretariat was ordered in 1303 to weed out all supernumeraries.34 The huge increase in the number of official supernumeraries, however, was not matched by any improvement in administrative efficiency. As Cheng Chieh-fu pointed out in the T'ai-p'ing ts'e (Treatises for the great peace) submitted to the government in 1303, even though the Chih-yiian hsin-ko (Chin-yuan new code) promulgated by Khubilai in 1291 had required officials to settle ordinary cases in five days, cases of medium importance in seven days, and cases of major importance within ten days, it usually took an official half a year to settle an unimportant matter and a whole year for an important case.3' The khaghan was so exasper28 29 30 31 32 YS, 18, p. 386. YS, 20, p. 4 4 1 ; 2 1 , pp. 467, 471. YS, 136, p. 3293. YS, 9 3 , p. 2352. Ta Yuan sheng cheng kuo ch'ao tien cbang (rev. and expanded 1322 ed.; repr. Taipei, 1972), 7, p. 26a (hereafter cited as YTC). 33 YS, 18, p. 383. 34 YS, 2 1 , p. 447. 35 For Cheng's memorial, see Shao Yiian-p'ing, Yuan shih Id pirn, ed. Jen-ho Shao shih (1699; repr. in Sung Liao Chin Yuan pith shih, comp. Hsi Shih-ch'en, Ch'ang-sha, 1797), 25, pp, 1 ia—22a. For the relevant passage in the 1291 code, see Paul Heng-chao Chen, Chinese legal tradition under the Mongols: The code of 1291 as reconstructed (Princeton, 1979), pp. 108—9. Cambridge Histories Online © Cambridge University Press, 2008 SIGNS OF DECLINE 499 ated by the widespread problem of bureaucratic procrastination that in 1294 he reprimanded the ministers of his Secretariat and even expressed his nostalgia for the administrative efficiency that had existed under the infamous Sangha.36 The government was further plagued by endemic corruption, and the reign saw a succession of such scandals. The greatest corruption scandal at the highest level of government occurred in 1303. In the previous year Chu Ch'ing (1237-1303) and Chang Hsiian (d. 1303) were accused of serious offenses.37 They had been pirates and salt smugglers before they had defected from the Sung to the Yuan with a large number of ships at the time of the conquest of Sung. Subsequently they had made legendary fortunes from managing the maritime transportation of grain for the Yuan government and by operating overseas trade on their own account. Their trial resulted in the imprisonment of all their families and the confiscation of their properties. The investigation also led to the discovery that their two families had paid bribes to several high-ranking officials in the central government. Saiyid Ajall Bayan, Liang Te-kuei, Tuan Chen, and Arghun Sali, all managers of governmental affairs, and four other ministers of the Secretariat were cashiered on the same day for accepting bribes.' 8 Even Oljei was impeached for the same offense.39 To weed out corruption, the enraged khaghan promulgated twelve articles on corruption, and in the late spring of 1303 he sent officials to make investigations in seven regions of the empire. 40 Their investigations led to the conviction of 18,473 officials and clerks, and the illegal profits they uncovered amounted to 45,865 ting. Even this figure was apparently an underreporting. 4 ' The khaghan, however, did not persist with his effort to suppress corruption. Not only did Oljei come out of the corruption scandal totally unscathed, but most of the ministers who had been dismissed from office for accepting bribes also were reinstated within the next two years.42 36 YS, 18, p. 388. 37 On Chu's and Chang's case, see Uemacsu Tadashi, "Gendai Konan no gomin Shu Sei Cho Sen ni csuite: sono chusatsu to zaisan kambotsu wo megutte," Toyoshi kenkyu, 27 (1968), pp. 292-317; T'ien Juk'ang "Mongol rulers and Chinese pirates," History Today, 33 (1983), pp. 3 3 - 8 . 38 YS, 21, p. 449. Rashid al-DIn states that eleven ministers were arrested but were saved by Tan-pa (1230-1303), the imperial preceptor. The reason he gives for the arrest is that they had accepted commissions from some merchants for purchasing jewelry for the court. See Boyle, The successors of Genghis khan, p. 330. On Tan-pa's intervention, see Herbert Franke, "Tan-pa: A Tibetan lama at the court of the great khans," in Orientalia Venetiana, vol. 1, ed. Mario Sabattini (Florence, 1984), pp. 157-180. 39 YS, 21, p. 448. 40 YS, 21, p. 449. For the text of the twelve articles, see YTC, 46, pp. ib—2a. 4 : YS, 21, pp. 449, 456; for the report submitted by Liu Min-chung, who was sent to investigate, see his Chung an chi (SKCS ed.), 7, pp. 1 ib—21b. 42 YS, 21, pp. 4 6 0 - 5 . Cambridge Histories Online © Cambridge University Press, 2008 500 MID-YOAN POLITICS The problems of official supernumeraries and bureaucratic corruption only partly explain the government's increasingly seriousfinancialdifficulties that were to plague the Yuan throughout the rest of the dynasty. Another major reason for financial stress was the constant giving of lavish grants to members of the nobility, especially to imperial relatives.4^ Khubilai had been prudent in making such grants and was known for being "miserly" in treating the members of his clan. Beginning with Temiir, however, the Yuan khaghans made much more generous grants to imperial relatives and meritorious officials, in addition to the quota of regular annual grants (sui-tz'u) that had been fixed by Khubilai. This generosity was politically necessitated by the need to reward these relatives and officials and to retain their support as the rivalry for the throne became increasingly bitter. On Temiir's enthronement in 1294 the imperial relatives were given 400 percent more gold and 200 percent more silver than the sums they had previously received as annual gifts.44 In addition, Temiir frequently made enormous special grants to individual imperial relatives for one reason or another. Grants given to three imperial sonsin-law in 1294, for example, amounted to more than 120,000 taels of silver.45 These excessive grants soon exhausted the imperial treasury.46 The Secretariat reported in 1294, two months after Temiir's enthronement, that the treasury had left only 270,000 ting of paper currency after paying the imperial relatives their gifts for attending the ceremony.47 Early the next year the Secretariat again reported that nearly all the wealth accumulated in Khubilai's reign had been used up to pay grants to imperial relatives and meritorious officials.48 In short, as a result of the special nature of Mongolian politics in the post-Khubilai era, political bribes in the form of imperial grants became one of the ways that the Yuan government became financially exhausted. Because Temiir's administration was opposed to increasing the tax quotas, there was no way that it could cover its deficits other than by using its silver monetary reserves. The government thus ordered in 1294 that out of 936,950 taels of silver kept as a monetary reserve by the Stabilization Bureaus (P'ing43 On the imperial grants in various forms, see Shih Wei-min, "Yuan sui tz'u k'ao shih," Yuan shih lun ts'ung, 2 (1986), pp. 144—53; Iwamura Shinobu, Mongoru shakai keizaishi no kenkyu (Kyoto, 1964), pp. 4 2 1 - 3 2 44 YS, 18, p. 382. 45 YS, 18, p. 382. 46 The revenues collected by the Yiian government in 1298 were 19,000 taels of gold, 60,000 taels of silver, and 3.6 million ting in paper currency. See YS, 19, p. 417. On the relations between the grants and the income of the government, see Tayama Shigeru, "Gendai zaiseishi ni kansuru oboegaki - shushi no gaku o chushin to shite," in Toyo no seiji ieizai (Hiroshima, 1949), pp. 191—266. 47 YS, 18, p. 384. 48 YS, 19, p. 402. Cambridge Histories Online © Cambridge University Press, 2008 THE TRANSITION TO PEACE 5OI chun k'u) of the various routes, only 192,450 taels should be kept for the original purpose, and the rest was to be sent to the capital — apparently to pay for current expenses. Four years later, another 200,000 ting were taken out of the reserves. By 1299 the government's financial situation had deteriorated so much that half of its expenditures had to be covered by the monetary reserves.49 This draining of the monetary reserves in turn greatly weakened the credibility of the paper currency system and touched off an inflationary spiral that was even more harmful than the one that had taken place in the I28OS.5° THE TRANSITION TO PEACE Temiir's reign was significant as the transition between a period of continuing conquests and one of general peace. All the foreign conquests that Khubilai had launched late in his reign had ended in failure. Apparently admitting that further conquests would be fruitless and costly, Temiir khaghan reversed the expansionist policy of his ancestors. Immediately after his enthronement he canceled the campaign against Annam planned by Khubilai in the last year of his reign and instead readmitted Annam as a tributary state. In 1298 Temiir rejected a recommendation to renew the invasion of Japan and instead in the following year sent a monk to that island country on a peace mission.' 1 The only foreign campaigns launched in Temiir's reign were those against Burma in 1300-1 and against Pa-pai hsi-fu, a small state in present northern Thailand and eastern Burma, in 1301—3. Although both these campaigns ended in failure, they were not attempts at foreign conquest. The campaign against Burma was launched to punish the Burmese for dethroning a king who had recognized Yuan suzerainty, and that against Pa-pai hsi-fu was a police action against a border state that was expanding rapidly at the expense of the Yuan.' 2 The greatest political and military achievement of Temiir khaghan was the successful conclusion of the protracted and costly war with Khaidu (ca. 1235-1301) and Du'a (r. 1282-1307), the khans of the Ogodei and Chaghadai khanates, who had begun to challenge Khubilai's authority as 49 YS, 18, pp. 387; 19, p. 417; 20, p. 426. 50 P'eng Hsin-wei, Chung-kuo huopi shih (Shanghai, 1958), pp. 409-10; Maeda Naonori, "Gencho jidai ni okeru shihei no kachi hendo," in his Gencho shi no kmkyu (Tokyo, 1973), pp. 107-43. 51 YS, 208, p. 4630; 209, p. 4650. 52 YS, 211, p. 4639; Su T'ien-chiieh, Kuo ch'ao wen lei, 42, pp. 32a-33a; Yuan cb'ao cheng mien lit (.TSCC ed.). This work has been translated into French; see Edouard Huber, "Etudes indochinoises: V. — La Fin de la dynastic de Pagan," Bulletin de I'kolt franfaisectextremt-orient, 9(1909), pp. 6 3 3 - 8 0 , esp. pp. 662-80. Cambridge Histories Online © Cambridge University Press, 2008 502 MID-YOAN POLITICS the khaghan in 1268 or i26c>.5} This achievement, however, did not come easily. Despite Khubilai's continual efforts for a quarter of a century to strangle his Central Asian enemies by denying to them the food-producing sedentary areas, Khaidu and Du'a still remained resilient at the time of Khubilai's death, extending their control over Uighuristan and frequently invading Mongolia. Throughout the greater part of his reign Temiir was obliged to devote large resources to the defense of Inner Asia and actually suffered a serious setback in Mongolia early in his reign. On one winter night in 1298, Du'a launched a surprise attack at Ho-erh-ha-t'u against the main forces of the Yuan army in Mongolia under the command of Prince Kokochii (K'uo-k'uoch'u, d. 1313), Temiir's uncle who had replaced Temiir as the overseer of all Yuan forces in the steppes. The ensuing battle resulted in the capture of Temiir's son-in-law, the prince of Kao-t'ang, Korgiiz (K'uo-li-chi-ssu, d. 1298), and the decimation of the Yuan forces in Mongolia.54 The Yuan position in Mongolia was restored only after its defense system had been reorganized under Khaishan (Hai-shan, 1281-1311). Khaishan was the eighteen-year-old son of Temiir's elder brother Darmabala (1264-92) who was appointed in 1299 to replace the feckless Kokochii.« In addition to the Mongolian troops under various princes and the Qipchaq troops under the veteran general *Chong 'ur (Ch'uang-wu-erh, 1260-1322) who had originally been stationed in Mongolia, Khaishan was also given Chinese troops from the imperial guards (wet) as well as fresh Mongolian troops from China. The restrengthened Yuan army turned to the offensive the next fall and defeated Khaidu south of the Altai Mountains. A more historically decisive battle took place at T'ieh-chien-ku and Khara Khada (Ho-la ho-ta) east of the Altai Mountains in September 1301 when Khaidu and Du'a launched their last offensive. The encounter itself was perhaps inconclusive, as the Chinese and Persian sources contradict each other about its results.56 What was 53 On the war against Khaidu and Du'a, see Wilhelm Barthold, Turkestan down to the Mongol invasion, trans. T. Minorsky, 3rd ed. (London, 1968), pp. 492—4; Paul Pelliot, Notes on Marco Polo (Paris, 1959), vol. i , pp. 124—9; Et an i Toshiyuki, "Kaido no ran ni kansuru ichi kosatsu," in Tamura Hakushi shiju Tbyoshi ronso, ed. Tamura Hakushi taikan kinen jigyokai (Kyoto, 1968), pp. 89-104; John D . Dardess, "From Mongol empire to Yuan dynasty: Changing forms of imperial rule in Mongolia and Central Asia," Monumenta Serica, 30 (1972—3), pp. 117—65; Hsiao Ch'i-ch'ing, The military establishment of the Yuan dynasty (Cambridge, Mass., 1978), pp. 56—8. 54 On this defeat see Boyle, The successors of Genghis khan, pp. 326—7; Liu Ying-sheng, "Yuan ch'ao yii Ch'a-ho-t'ai han kuo te kuan hsi," Yuan shih lun ts'ung, 3 (1986), pp. 56—81; esp. pp. 76—7. 55 YS, 22, p. 477; Matsuda Koichi, "Kaishan no seihoku Mongoria shussei," Tohogaku, 64 (1982), pp. 73—87. Matsuda, however, erroneously gives the date of Khaishan's commission as 1298. 56 YS, 22, p. 477; Boyle, The successors of Genghis khan, p. 329; Liu Ying-sheng, "Yuan ch'ao yii Ch'a-hot'ai han kuo te kuan hsi," pp. 7 8 - 9 . Cambridge Histories Online © Cambridge University Press, 2008 THE TRANSITION TO PEACE 5O3 important was its indirect result, for Du'a was wounded in the battle and Khaidu died soon thereafter, perhaps of a wound he had suffered.57 With the death of Khaidu, the main stumbling block to peace among the Mongolian khanates was removed. It was Du'a, Khaidu's ally, who took the initiative for peace. Tired of his hopeless challenges to the authority of the khaghan and more concerned with establishing his own supremacy in Central Asia, Du'a first manipulated to have Chapar (Ch'a-pa-erh), Khaidu's son, succeed Khaidu as the ruler of the Ogodei khanate in the summer of 1303. Then in the fall of the same year he persuaded Chapar to join him in proposing to Temiir that they cease their hostilities, making known their willingness to recognize Temiir's authority as the khaghan of all Mongols. Temiir responded to this proposal quickly and favorably. Though a khuriltai for peace, as was proposed by Du'a, was never held, an agreement on the cessation of war was nevertheless achieved. A mission sent jointly by Temiir, Du'a, and Chapar arrived at the court of Il-khan Oljeitii (r. 1304-16) in early 1304, seeking the latter's agreement to the peace proposal and the reestablishment of unity among all Mongols.' 8 The peace agreement meant much to Oljeitii as well as to the rulers of the other Mongolian khanates. In a letter to the king of France, Philip IV, in 1305, Oljeitii proudly announced the peace that had been achieved in the Mongolian world and urged the European princes to follow their example.' 9 Though the peace achieved in 1303 was short-lived, it not only reestablished the nominal suzerainty of the Yiian over other Mongolian khanates, but it also gave the Yiian the opportunity to remove permanently the threat from Central Asia by separating the heirs of Chaghadai from those of Ogodei. As soon as the external pressures decreased, however, Du'a and Chapar, who had been allies, soon clashed with each other over the question of territory. In this conflict Temiir backed Du'a, and in the fall of 1306 Temiir sent an army commanded by Khaishan across the southern Altai to his aid. 57 On the contradictory accounts on Khaidu's death, see Wilhelm Barthold, Four studies on the history of Central Asia, trans. V. Minorsky and T. Minorsky (Leiden, 1956), vol. 1, pp. 128-30; Liu Yingsheng, "Shih chi Wo-k'uo-t'ai han kuo mo nien chi shih pu cheng," Yiian shih chi pei fang min tsu shih yen child chi fan, to (1986), pp. 4 8 - 5 9 , esp. pp. 4 9 - 5 0 . 58 The most detailed account of the peace of 1303 is given by the Persian historian Qashani in Tarikh-i Oljaytu, ed. Mahin Hambly (Tehran, 1969), pp. 3 2 - 5 . Chinese accounts may be found in Su T'ienchiieh, Kuo ch'ao wen lei, 26, pp. 13b—14a; YS, 21, pp. 454, 456. For some studies, see Wladyslaw Kotwicz, "Les Mongols, promoteurs de l'idee de paix universelle au d£but du XH-e [sic] siecle," Rocznik Orientalistyczny, 16 (1953), pp. 428—34; Saguchi Toru, "Juyon seiki ni okeru Gencho daikan to seiho san-oke to no rentasei ni tsuite," Kita Ajia gakuho, 1 (1942), pp. 151-214; Liu Ying-sheng, "Yuan tai Meng-ku chu han kuo chien te yiieh ho chi Wo-k'uo-t'ai han kuo te mieh wang," Hsinchiang ta bsiieh hsiieh pao, 2 (1985), pp. 3 1 - 4 3 ; Liu Ying-sheng, "Shih chi Wo-k'uo-t'ai han kuo mo nien chi shih pu cheng," pp. 50—1. 59 Antoine Mostaert and Francis W. Cleaves, Les lettrts de 1289 et 1305 des Man Arghun et Oljeitii a Philippe le Bel (Cambridge, Mass., 1962), pp. 5 5 - 8 5 . Cambridge Histories Online © Cambridge University Press, 2008 504 MID-YOAN POLITICS Attacking Chapar's forces from the rear, Khaishan captured several members of Chapar's family and advanced as far as the Irtysh River. Chapar had no choice but to surrender to Du'a.60 Chapar was later dethroned by Du'a, probably in 1307, and his younger brother Yangichar (Yang-chi-ch'a-erh) was established by Du'a as the puppet khan of the Ogodei ulus. Forced by circumstances, Chapar surrendered to the Yuan in 1310 during Khaishan's reign, thus marking the end of the Ogodei ulus that had challenged the Yuan on the battlefield for nearly forty years. Du'a and his successors recognized Yuan suzerainty most of the time, repeatedly sending tribute missions to Ta-tu. Peace was subsequently maintained between the Yuan and the Chaghadai ulus with the exception of a short period from 1316 to 1320.6l Thus by the end of his reign Temiir had established nominal Yuan suzerainty over the whole Mongolian world, thereby succeeding where his grandfather had failed. THE DOMINANCE OF EMPRESS BULUKHAN Though he was able to restore peace in the Mongolian world, Temiir khaghan was unable to ensure harmony in his own family and court or to arrange a smooth succession to himself. Because of heavy drinking in his youth, he suffered from chronic illness in his later years.62 Taking advantage of this situation, Empress *Bulukhan played an active part in the court as well as in family affairs. From a distinguished Mongolian noble family of the Baya'ud clan, Bulukhan was made empress in 1299 after the khaghan's principal consort, Shirindari (Shih-lin-ta-li), had died earlier that year. The death of the empress dowager Kokojin the next year undoubtedly made Bulukhan even more politically influential. Bulukhan was apparently a woman of ability, for her biography in the Yuan shih, even though basically hostile in tone, concedes that the governmental policies enacted under her influence were for the most part "just and appropriate."63 The prosecution of Chu Ch'ing and Chang Hsiian for corruption in 1302—3 was engineered by her, and the enormous amount of property subsequently confiscated from their families went to her Household Service (Chung-cheng yuan) and immeasurably strengthened her ownfinances.64In 60 YS, 22, p. 477; Su T'ien-chiieh, Kuo ctiao wen lei, 23, p. 13b; Liu Ying-sheng, "Shih chi Wo-k'uo-t'ai han kuo mo nien chi shih pu cheng," pp. 51—3. 61 Thomas T. Allsen, "The Yuan dynasty and the Uighurs of Turfan in the 13th century," in China among equals: The Middle Kingdom and its neighbors, 10th—14th centuries, ed. Morris Rossabi (Berkeley and Los Angeles, 1983), pp. 2811—310; esp. p. 259. 62 Boyle, The successors of Genghis khan, pp. 301—2. 63 YS, 114, p. 2873. 64 Wang Feng, Wu ch'i chi (TSCC ed.), 4 hsia, p. 205; YS, 21, p. 459; see also Uematsu Tadashi, "Gendai Konan no gomin Su Sei Cho Sen ni tsuite," pp. 4 8 - 5 3 . Cambridge Histories Online © Cambridge University Press, 2008 THE REIGN OF KHAISHAN 505 June 1305 she made her son, Te-shou, the heir apparent. To ensure the boy's smooth succession in the future, all the potential rivals were removed from the court. Darmabala's son Ayurbarwada (Ai-yu-li-pa-li-pa-ta, 1285—1320) was sent in August to Huai-chou (Ch'in-yang, Honan) as the prince of Huaining. But unexpectedly, the heir apparent Te-shou died in January 1306. 65 Thus when Temiir khaghan died without an heir on 2 February 1307 at the age of forty-one, the question of his succession remained unresolved. THE REIGN OF K H A I S H A N ( E M P E R O R W U - T S U N G ) , 1307- I3II Temiir's successor, Khaishan, came to the throne after a violent conflict that involved many imperial family members and top bureaucrats. When Temiir died without an heir, alliances were immediately formed on the basis of existing aristocratic and bureaucratic factions, each supporting its own candidate for the throne. 66 One alliance was led by Empress Bulukhan herself and was supported by some senior officials of the Secretariat under the leadership of Akhutai (A-hu-t'ai, d. 1307), its grand councillor of the left, and the officials of the Household Service for the empress. They intended first to make Bulukhan the regent and eventually to put Ananda (A-nan-ta, d. 1307), the prince of An-hsi, on the throne. 67 The other alliance was led by the grand councillor of the right, Harghasun, and was supported by Prince Tura (T'u-la, d. 1309), a great-great-grandson of Chaghadai, and Yakhudu (Ya-hu-tu, d. ca. 1310), a descendant of Tolui. They fought for the candidacy of Darmabala's sons, Khaishan and Ayurbarwada.68 Each of the two factions had advantages. The assumption of the regency by Bulukhan after her husband's death was legitimate in terms of the Mongolian imperial tradition. Her protege, Ananda, as the son of Chen-chin's younger brother Manggala (Mang-ko-la, d. 1278) and thus the late khaghan's cousin, 65 YS, 2 t , pp. 465, 467. The identification ofTe-shou as Bulukhan's son is tentative. Although YS 114, p. 2873, indicates that Empress Shirindari was Te-shou's mother, Rashid 'al-Din, Yang Yii, and T'ao Tsung-i all hold that Te-shou was Bulukhan's own son. See Boyle, The successors of Genghis khan, p. 319; Yang Yu, Shan chit hsin htia (Chih pu tsu tsai ts'ung shu ed.), p. 13b; and Tao Tsung-i, Nan ts'un ch'o keng lu (Peking, 1959), 5, p. 76. The Persian chronicle Mu'izz 'l-Ansab says that Temiir khaghan had four sons, of whom the name of the first is given as "Qung-taill," which apparently means huang t'ai tzu and refers to Te-shou. But it fails to identify this Qung taiSTs mother. See Louis Hambis, he chapitre cvii du Yuan che. Les genealogies imperials mongoles dans Vhistoire chinoise offuielle de la dynastie mongole, T'oung Pao Supplement no. 38 (Leiden, 1945), p. 136. For a discussion of both Te-shou's relationship with Bulukhan and his death, see Han Ju-lin, "Hsi pei ti li cha chi," in his Ch'iung lu chi (Shanghai, 1982), pp. 69—89, esp. pp. 83—4. 66 On the succession crisis of 1307, see John W. Dardess, Conquerors and Confucians: Aspects of political change in late Yuan China (New York, 1973), pp. 12-17; Hsiao Kung-ch'in, "Lun Yuan tai huang wei chi ch'eng wen t'i," pp. 26—30. 67 YS, 22, pp. 4 7 7 - 9 ; 24, pp. 5 3 5 - 6 ; 114, PP- 2 8 7 3 - 4 . 68 YS, 117, pp. 2 9 0 7 - 1 0 . Cambridge Histories Online © Cambridge University Press, 2008 506 MID-YOAN POLITICS had strong claims to the throne by way of generational seniority.69 Ananda's position, however, was weakened when the issue of lineage was brought into question. A strong argument was put forward by Khaishan and by Ayurbarwada's partisans that "collateral sons were not eligible for the succession," for Ananda was the only contender for the throne in the post-Khubilai era who was not a descendant of Chen-chin. Moreover, although Ananda had under his command the Tangut territory and a large army and had played an important role in the war against Khaidu and Du'a, he was a newcomer at the capital and lacked any sizable local army to back up his candidacy. Equally damaging was the fact that he was a pious Muslim, a clear political disadvantage at a time when the Yuan imperial family had become increasingly Buddhist. 7° The other faction was disadvantaged by the fact that neither of its candidates was in the capital at the time of Temiir khaghan's death. Khaishan was still in western Mongolia and Ayurbarwada in his fief at Huai-chou. Nevertheless, these two brothers had a stronger claim to the throne than Ananda did, as they were the grandsons of Chen-chin. Moreover, both had sufficient credentials for their candidacies. Khaishan had been the hero of the war in Inner Asia and still commanded the most powerful military machine in the empire. His younger brother Ayurbarwada, on the other hand, had a reputation as a paragon of Confucian virtue and was popular among the Chinese.71 But their seizure of the throne could not have succeeded without the support of Harghasun, who not only controlled the administrative machinery of the state but who also had taken charge of the imperial guards after Temiir khaghan fell ill. As the grand councillor of the right, he was able to use delaying tactics by refusing to countersign the orders issued by Empress Bulukhan and denying to her faction the use of all seals and treasuries, while at the same time he sent messengers to hasten the arrival at court of Khaishan and Ayurbarwada.72 When Ayurwarbada arrived in Ta-tu, he stormed the palace on 4 April with the troops given to him by Harghasun, in the process killing Akhutai and arresting Prince Ananda and Empress Bulukhan. With the opposition thus eliminated, the choice between the two brothers remained a difficult issue. Although Ayurbarwada enjoyed the advantage of having the capital under his control, Khaishan was not only senior in age but also enjoyed military superiority over his younger brother. With their mother, Targi (Ta69 On Ananda and his family, see Matsuda Koichi, "Gencho chi no bumposei — Anseid no jirei wo chushin to shite," Shigaku zasshi, 88 (1979), pp. 1249-86. 70 Boyle, The successors of Genghis khan, pp. 323—8; see also Wen Yu-ch'eng, "Yuan An-hsi wang yii tsung chiao," K'ao ku yii wen wu, 4 (1984), pp. 9 5 - 7 . 71 Dardess, Conquerors and Confucians, pp. 13—15. 72 Liu Min-chung, Chung an chi, 15, pp. 12a—b. Cambridge Histories Online © Cambridge University Press, 2008 ADMINISTRATIVE ANOMALIES 507 chi, d. 1322), as the arbitrator, a mutually acceptable agreement was reached whereby Ayurbarwada would dissolve the regency that he had established after the palace coup d'etat. In turn, Khaishan would make his younger brother the heir apparent after his own enthronement. The subsequent enthronement of Khaishan, who had arrived from Mongolia with thirty thousand soldiers, at Shang-tu on 21 June 1307, was performed properly in the manner of a khuriltai.1^ It is clear, however, that the succession crisis of 1307 was not resolved here. Rather, it was the bureaucratic support at the capital lent by Harghasun and, to a lesser extent, his own powerful field army from Mongolia that enabled Khaishan to seize the throne. The khuriltai was no more than a ritual performed after the event to provide the necessary aura of legitimacy to the forceful seizure of the throne by Khaishan's faction. ADMINISTRATIVE ANOMALIES Khaishan khaghan, who was to reign for only three and a half years, was the exact opposite of his predecessor in terms of political style. Whereas Temiir khaghan had striven to preserve Khubilai's pattern of rule, Khaishan held little regard for it. In a memorial submitted in 1310 the Chinese censor Chang Yang-hao (1270-1329) bluntly pointed out that every policy of the khaghan differed from those of Khubilai, and he asked whether the khaghan believed that by disregarding Khubilai's model of government he could create one of his own. ™ Khaishan's political orientation can be partly understood from his early life. 7 ' Although for a short while he had shared with Ayurbarwada the tutorship of the Confucian scholar Li Meng (1265-1321), he apparently was little affected by Confucian culture. He had spent all his early adulthood as the supreme commander of the Yuan forces in the steppe and had always been engaged in field action. Khaishan exemplified the typical impetuosity and simplicity of a nomadic warrior and was impatient with the bureaucratic rules and administrative precedents so painstakingly instituted by his greatgrandfather. Lacking trust in the existing bureaucratic establishment at Ta-tu, only two months after his enthronement, he transferred Harghasun to Mongolia as the grand councillor of the left of the newly established Branch Secretariat of Ling-pei, despite the great contribution that Harghasun had made to his own seizure of the throne.' 6 Thereafter Khaishan relied mostly on the personal retainers and commanders he had brought with him from Mongolia. 73 74 75 76 Hsiao Kung-ch'in, "Lun Yuan tai huang wei chi ch'eng wen t'i," pp. 28—30. Chang Yang-hao, Kuei t'ieti lei kao (SKCS ed.), 2, p. 36. YS, 22, pp. 4 7 7 - 8 . YS, 136, p. 3294. Cambridge Histories Online © Cambridge University Press, 2008 508 MID-YOAN POLITICS Having little regard for the established system, Khaishan freely gave away inflated noble and official titles and filled the government with supernumeraries who had few or no proper qualifications. Apparently in order to court support for his authority, he granted nineteen princely titles in 1307 and 1308 alone, fourteen of which were princes of the first rank (i-tz'u wang). Of the fourteen recipients, only two were the sons of the khaghans, thus totally disregarding Khubilai's unwritten law that only the sons of khaghans could be made princes of the first rank.77 Khaishan gave out honorary titles and ranking bureaucratic appointments even more indiscriminately. As Chang Yang-hao pointed out, there were even cases of actors, butchers, and Buddhist and Taoist clergymen being given the titles of ministers of the Secretariat, and artisans who were given the title of dukes (kuo-kung) and councillors (ch'eng-hsiang). This statement is by no means an exaggeration, as it can be corroborated from our sources.78 In addition to this sort of arbitrariness by the khaghan, another practice that contributed to the swelling of the bureaucracy as well as to the deterioration of the administration in Khaishan's reign was the issuance by court attendants (chin-shih) of orders dealing with personnel, financial, and legal matters for their own gain, without going through the proper channels of government. The "court attendants" in these cases clearly refer to palace officials and members of the imperial guard, kesig, who were close to the khaghan and the empress dowager Targi.7? Within two months of the khaghan's enthronement, as many as 880 requests for appointments through palace edicts (net chiang-chih) had been made, 300 of which were granted. Despite repeated protests from the Secretariat, this practice continued, for such orders were very likely issued with the full knowledge of Empress Dowager Targi, who often exerted a strong hand in governmental affairs, if not with the compliance of the khaghan himself.80 The rapid expansion of the bureaucracy can be clearly seen at the top levels. It was reported in 1307 that there were no fewer than fourteen ministers in the Secretariat and four censors in chief.8' One year later the Bureau of Military Affairs reported in a memorial that the number of the bureau's chief officials had jumped from six in Khubilai's time to thirty-two in 1308.82 In order to save the cost of supporting this tremendously swollen 77 Noguchi Shuichi, "Gendai Mushu cho no ogo juyo ni csuite: gcnshi Shoohyo ni kansuru ichi kosacsu," in Ajia shominzoku ni okeru shakai to bunka: Okamoto Keiji lensei taikan kinen ronshu, ed. Okamoto Keiji sensei taikan kinen ronshu kankokai (Tokyo, 1984), pp. 271—305; Li Tse-fen, Yuan sblh hsin chiang (Taipei, 1978), vol. 3, pp. 3 4 9 - 5 1 . 78 Chang Yang-hao, Kuei t'ien lei kao, 2, pp. ioa-b; YS, 22, pp. 481, 484, 501; 23, p. 524. 79 On the influence of the kesig, see Ch'i-ch'ing Hsiao, The military establishment of the Yuan dynasty, p. 41. 80 YS, 22, pp. 4 8 5 , 487, 4 9 2 , 497; 23, pp. 509, 516. 81 YS, 22, p. 4 8 1 . 82 YS, 22, p. 501. Cambridge Histories Online © Cambridge University Press, 2008 ADMINISTRATIVE ANOMALIES 509 bureaucracy, the khaghan issued an order in 1307 to dismiss the supernumeraries and to bring the total number of officials in line with the quota that had been set by Temiir khaghan in the previous year. The order apparently produced no practical results, however, as the khaghan himself admitted two years later that there were just as many supernumeraries as in the past. 83 The increase in the size of the bureaucracy naturally entailed great administrative costs. The financial strain this caused was made worse by Khaishan's personal spending, which was even more lavish than Temiir's. Khaishan made the grants to the princes and officials who attended his enthronement ceremony in accordance with the amounts set by his predecessor. But because of the lack of funds, only 1.7 million of the 3.5 million ting needed had actually been disbursed for this purpose by the fall of that year.84 Huge amounts, moreover, were spent on the construction of Buddhist temples at Ta-tu and Shang-tu and on the sacred mountain Wu-t'ai, on the building of residences for court officials, and on the purchase of jewelry.8' Even more costly was the building of a new palace city at Ongghochatu called Chung-tu (modern P'ai-ch'eng-tzu, in Chang-pei, Hopei) to provide a comfortable stopover for the Khaghan during his annual progresses between the two capitals. 86 Because of the lack of reserves inherited from his predecessor and his own reckless spending, Khaishan encountered financial difficulties quite early in his reign. Only four months after his accession, the Secretariat summed up the government's financial situation thus: Of the 4 million ting collected as state revenue, only 2.8 million had actually reached the capital as revenue for the central government. However, 4.2 million ting had already been spent since the accession; thus there was a deficit of 1 million ting that could not be paid. 87 The total government expenditure for the year 1307 was 10 million ting of paper notes and 3 million shih of grain. 88 Facing such enormous deficits, Khaishan dug further into the monetary reserves for his expenditures. By the fall of 1310, 10,603,100 ting had been borrowed from the monetary reserves for current expenditures. 89 In addition, salt licenses (yenyin) also were frequently sold in advance to help make up the deficit.90 83 84 85 86 87 88 89 90 YS, 22, p. 504; 23, p. 527. YS, 22, pp. 481, 486. YS, 22, pp. 486, 488, 489, 496, 497, 504, 509; 23, p. 517. On the construction of Chung-tu, see YS, 22, pp. 493, 495, 498. See also Yanai Wataru, "Gendai no higashi Moko," in his Mokoshi ientyii (Tokyo, 1930), pp. 585-661, esp. pp. 640-3; Pelliot, Notes on Marco Polo, vol. I, p. 322; and Cha-ch'i Ssu-ch'in (Sechin Jagchid), "Yuan tai Chung-tu k'ao," Pirn cbengym chiu so nimpao, 18 (1987), pp. 31-41. YS, 22, p. 488. YS, 23, p. 510. YS, 23, p. 516. YS, 22 pp. 491, 495. Cambridge Histories Online © Cambridge University Press, 2008 510 MID-YOAN POLITICS Having exhausted all existing resources, Khaishan's government was therefore facing afiscalproblem of crisis proportions. THE NEW DEALS Khaishan khaghan's response to the mounting fiscal crisis was not to adopt a policy of retrenchment by cutting down expenses and dismissing supernumerary personnel, as he was repeatedly urged to do by his officials. Instead, he made a bold attempt to increase state revenues by resurrecting the Department of State Affairs (Shang-shu sheng).91 This department had been established twice in Khubilai's reign, under the leadership of the two se-mu financiers, Ahmad and Sangha, in order to increase state revenue. But it was abolished in both cases because of the widespread resentment caused by its highly exploitative policies. Shortly after his accession Khaishan approved a recommendation to reestablish the department from the commissioner for Buddhist and Tibetan affairs (hsiian-cheng yuan shih), Toghto (T'o-hu-t'o, d. 1311), a Uighur and an old retainer of the khaghan. The idea, however, was dropped as a result of protests from the Censorate.92 But as the fiscal health of the government deteriorated further, the office was finally reestablished in September 1309 when both Yiieh Shih (d. 1311) and Pao-pa (d. 1311) urged the khaghan to restore the department in order to carry outfinancialreforms.9' Although the Secretariat retained its jurisdiction over most routine governmental matters, the Department of State Affairs was supposed to deal only with policies relating to fiscal reform. But the new Department of State Affairs soon took over most of the important functions of the Secretariat in finance, personnel, and legal matters and was given a large measure of discretionary power. All Branch Secretariats were renamed "Branch Departments of State Affairs," and through them the tentacles of the Department of State Affairs were extended to all parts of the empire. The moving spirits of the new department were Toghto, the grand councillor of the left, San-pao nu and Yueh Shih, both managers of governmental affairs, and Pao-pa, the assistant administrator of the right. Both Toghto and San-pao nu were old retainers of Khaishan.94 Yiieh Shih is known to have been an official in 91 On the history of the Department of State Affairs, see Aoyama Koryo, Gencho sboihosho ko (Tokyo, 92 YS, 22, pp. 488-9. 93 YS, 23, p. 513. 94 On Toghto, see K'o Shao-min, Him Yuanihib (Tientsin, 1922; author's 2nd rev. ed.; Peking, 1930), 199, pp. r ib—13a (hereafter cited as HYS). On San-pao nu, see Ch'eng Chii-fu, Hsmh-lou chi (SKCS ed.), 2, pp. n b - i 2 a . Cambridge Histories Online © Cambridge University Press, 2008 THE NEW DEALS 5II Khubilai's reign who had been dismissed for corruption. 9 ' Nothing is known about Pao-pa before his appointment; presumably he was appointed for his expertise in state finance. Central to Khaishan's "new deals" were monetary reforms. New bills called the "silver notes of the Chih-ta era" (Chih-ta yin-ch'ao) were issued to replace the existing Chung-t'ung and Chih-yuan paper notes. An exchange rate of five to one was established between the new notes and those of the Chih-yuan currency issued in 1287 which, in turn, were five times greater in value than were the Chung-t'ung notes issued in 1260 — a vivid indication of the rate of inflation during the previous half-century. To strengthen the value of the new paper notes, gold and silver were demonetarized, and for the first time in the Yuan two kinds of copper coins were minted, called, respectively, "circulating treasure of the great Yuan" (Ta-yiian t'ung-pao) and "circulating treasure of the Chih-ta era" {Chih-ta t'ung-pao). These monetary reforms were designed to absorb the mounting inflation and to make up for the ever-increasing deficits in the budget. The new bills issued in 1310 alone amounted to 1.45 million ting, which was the equivalent of 36.3 million ting of the old Chungt'ung bills, more than seven times the amount issued in any of the three preceding years and more than three and a half times the amount issued in 1302, the previous peak in the issue of paper money under the Yiian. 96 Other measures were also taken to augment the state revenues: The selling price of salt licenses issued under the state monopoly was raised by 35 percent over the price at the end of Temiir's reign.^ The ban on liquor production was lifted, and liquor tax control bureaus were established to collect taxes on liquor. The tax debts, which had been canceled in Temiir's reign, were collected again. A grain tax surcharge of 2 percent was imposed on the wealthy families of Chiang-nan with annual incomes over fifty thousand shih. To make the tax collection more effective, the merits of tax collectors were evaluated on the basis of the percentage increase in the taxes they collected over the tax quota of 1307. To stabilize food prices, ever-normal granaries (ch'ang-p'ing ts'ang) were established in various localities, and the quota for the maritime shipment of grain from the Yangtze valley was drastically 95 YS, 15, p. 319; 17, p. 366. For a summary of the epitaph for Yiieh Shih written by Yii Chi, see Ch'ien Ta-hsin, Cb'ien yen fang chin shih wen pa wet (Changsha, n.d.), 19, pp. 29a-b. 96 On the monetary reforms, see YS, 23, pp. 515—20; 93, pp. 2 3 7 0 - 1 ; Herbert F. Schurmann, Economic structure of the Yiian dynasty: Translation of chapters 93 and 94 of the Yiian shih (Cambridge, Mass., 1956; repr. 1957), pp. 139—40; Herbert Franke, Geld und Wirtscbaft in China unter der Mongolenherrshaft: Beitrdge zur Wirlschaftsgeschichte der Yiian-Zeit (Leipzig, 1949), pp. 5 7 - 9 ; N. Shatzman Steinhardt, "Currency Issues in Yuan China," Bulletin of Sung—Yiian Studies 16(1980), pp. 59—81, esp. pp. 68—9. 97 YS, 23, p. 320. See also Maeda Naonori, "Gencho jidai ni okeru shihei no kachi hendo." in his Gencho shi no kenkyu (Tokyo, 1973), pp. 118—20. For a discussion of the salt prices in the Yiian, see Ch'en Kaohua, "Yiian tai yen cheng chi ch'i she hui ying hsiang," Li shih lun-ts'ung, 1 (1964), pp. 1 7 5 - 2 1 7 . Cambridge Histories Online © Cambridge University Press, 2008 512 MID-YOAN POLITICS increased, reaching 2.93 million shih in 1310.98 Only a few measures were adopted to reduce the government's expenditures, including a modest reduction in the number of chief officials in the Secretariat, the Censorate, the Bureau of Military Affairs, and the Bureau of Transmission (T'ung-cheng yuan) and the dismissal of supernumeraries in various offices." In retrospect, the Department of State Affairs and the reforms it was designed to implement seem to have been doomed from the start. The earlier attempts to set up the department in Khubilai's reign had already proved its unpopularity beyond doubt and had left an image in the public mind of a highly oppressive and exploitative institution. The department, as a result, encountered strong opposition from the outset. It was established against the advice of Tas Bukha (T'a-ssu Pu-hua), the grand councillor of the right of the Secretariat, and of the officials of the Censorate.100 Its policies were strongly criticized by Inal Toghto (T'o-t'o), an older retainer of Khaishan and grand councillor of the right; by the Chinese scholar-officials Ching Yen, Chang Yang-hao, and Kao Fang (1264-1328); and by the officials of the Censorate.'01 Several Chinese officials appointed to the department refused to take up their posts.102 Moreover, even though the new policies may have succeeded in increasing state revenue, the excessive issues of paper notes and the sharp rise in the prices of salt licenses certainly created greater inflationary pressures and caused prices to soar even higher.IO' Any effective attack on the budget deficits and inflation would have entailed drastic cuts in the expenditures of the court and government. This, however, would not only have been contrary to Khaishan's general policies but would also have been extremely unpopular with the aristocracy and the bureaucracy and hence politically undesirable. Khaishan opted for an easy way of solving his fiscal difficulties. But the inflationary methods he adopted actually aggravated the problems facing his government. The reforms, in any case, were not given the opportunity to run their natural course, for they all were nullified only sixteen months after their inception, following Khaishan's death in January 1311 and Ayurbarwada's succession to the throne in the early spring. 98 Schurmann, Economic structure of the Yuan dynasty, p. 124. On the maritime transportation of grain, see also Wu Chi-hua, "Yuan ch'aoyii Ming ch'u hai yiin," Cbung-yang yen chiu yuan, Li shih yii yen yen chiu so chi k'an 28 (1956), pp. 363—80. 99 YS, 23, pp. 511, 520, 522. 100 YS, 22, p. 488; 23, p. 513. 101 YS, 23, p. 518; 175, pp. 4091, 4094; HYS, 201, p. 156. 102 MaTsu-ch'ang, Ma Shih-t'ien win chi (Ming [1368—1644} ed.); repr. in Yuan jen wen chi chenpen ts'ung k'an, ed. Wang Te-i (Taipei, 1985), 12, p. 146; YS, 176, p. 4168. 103 Ch'ang-ku-chen-i, Nung t'ien yiihua (TSCC ed.), shang, p. 6b; Ch'uan Han-sheng, "Yuan tai te chih pi," Chung-yang yen chiu yuan. Li shih yii yen yen chiu so chi k'an 15 (1948), pp. 10—48; repr. in vol. I of his Chung-kuo ching chi shih Inn ts'ung (Hong Kong, 1972), pp. 3 6 9 - 4 1 6 , esp. p. 403. Cambridge Histories Online © Cambridge University Press, 2008 T H E REIGN O F A Y U R B A R W A D A K H A G H A N 513 T H E REIGN O F A Y U R B A R W A D A K H A G H A N (EMPEROR JENT S U N G ) , 13 I I - I 320 Early orientations Ayurbarwada's succession to his elder brother Khaishan's throne in April 1311 was the first peaceful and smooth transition in Yuan imperial history. This was made possible by the fact that Khaishan had designated his younger brother as the heir apparent in June 1307, in accordance with their earlier agreement, and had subsequently appointed him as the titular head of the top central administrative organs, just as Khubilai had done when grooming Chen-chin to be his successor.104 In view of the fraternal love between Khaishan and Ayurbawada and the peaceful way in which one succeeded the other, one might expect a general continuity in policy and personnel between the two reigns. But what was to happen early in Ayurbarwada's reign was actually the opposite: a political purge of Khaishan's chief ministers and a reversal of most of his policies. These reversals of policies can be traced to Ayurbarwada's cultural and ideological orientation and his uneasy political relationship with his late brother. Since his early teens, Ayurbarwada had been tutored by the Confucian scholar-official Li Meng, who injected into his impressionable mind Confucian ethical and political precepts that would strongly affect his future political attitudes. 10 ' First as a prince in Huai-chou and then as Khaishan's designated heir apparent, Ayurbarwada had further surrounded himself with such Chinese scholars as Ch'enHao( 1264—1339), Wang I, Wang Yiieh (1252—1333), Chao Meng-fu (1254-1322), Wang Chieh (1275-1336), Chang Yang-hao, Shang Yeh (1244-1319), Yao Sui (1201-78), and Hsiao K'u (1241-1318); the artists Shang Ch'i and Wang Chen-p'eng; Chaghan (Ch'a-han), a se-mu scholar from Balkh who was to become a great translator; and Sewinch Khaya (Hsiaoyiin-shih Hai-ya, 1286-1324), the great Uighur san-ch'ii lyricist. Consequently, Ayurbarwada was not only able to read and write Chinese and appreciate Chinese paintings and calligraphy, but he also had a thorough grasp of Confucian learning and Chinese history. Io6 Strongly influenced by Confucian political ethics as he was, Ayurbarwada was naturally opposed to the exploitative policies carried out by the Department of State Affairs under Khaishan. 104 YS, 22, p. 480; 24, p. 536. 105 YS, 175, p. 4084. 106 On the influence of Chinese culture on Ayurbarwada, see YS, 24, pp. 535—6; Yoshikawa Kojiro, "Gen no shotei no bungaku," in vol. 15 of Yosbikawa Kojiro zenshu (Tokyo, 1969), pp. 232—303, esp. pp. 2 3 5 - 4 0 . Herbert Franke, "Could the Mongol emperors read and write Chinese?" Asia Major, 3 (1952), pp. 28—41, esp. pp. 3 1 - 3 ; Dardess, Conquerors and Confucians, pp. 13-15; and Lo Hsien-yu, "Yuan ch'aochu ti Han hua shu-i," Min tsu yen chiu, 5 (1987), pp. 67—74, e s P- PP- 7 ' ~ 7 2 - Cambridge Histories Online © Cambridge University Press, 2008 514 MID-YOAN POLITICS Ayurbarwada's relations with his elder brother seem to have been as politically delicate as they were fraternally warm. Because he was the one who had actually seized the throne for his brother, Khaishan always seems to have suspected Ayurbarwada of himself harboring imperial designs. Ayurbarwada's mentor, Li Meng, had left the government immediately after Khaishan's accession because he had been accused of having advised Ayurbarwada to keep the throne for himself. Ayurbarwada found it impolitic to speak out in Li Meng's defense.107 In view of Ayurbarwada's position, Wang Yiieh, deputy administrator (chan-shih ch'eng) of the heir apparent's establishment, repeatedly urged him to keep a low political profile.108 Even so, San-pao nu, the manager of governmental affairs of the Department of State Affairs, and Li Pang-ning, the chief eunuch, had suggested replacing Ayurbarwada as the heir apparent with Khaishan's own son, Khoshila (i 300-29). IO9 Ayurbarwada seemingly exerted little influence over his elder brother's policies, but his disagreement with them remained concealed until his own enthronement. The purge Disagreeing with his elder brother's policies and aspiring to make the Yuan government more Confucian, Ayurbarwada carried out a bloody purge and reversed most of his elder brother's policies immediately before and after his accession. On 30 January 1311, only three days after Khaishan's death, Ayurbarwada abolished the Department of State Affairs and had its chief ministers - Toghtd, San-pao nu, Yiieh Shih, Pao-pa, and Wang P'i arrested and executed.110 In the following months the Chih-ta paper notes and coins were abolished and recalled, and the Chung-t'ung and Chih-yuan notes were restored as the only official currency. A general retrenchment entailed trimming the bureaucracy to the 1293 level and reducing those offices that had been raised in rank to the original status they had had in Khubilai's time. The various public building projects initiated by Khaishan also were halted.111 Ayurbarwada enhanced the importance of Confucian scholar-officials in the government. In addition to appointing Mongols and se-mu as grand councillors, he successively appointed his mentor, Li Meng, and Chang Kuei (1264—1327), a scholar-general from one of the most distinguished northern 107 108 109 no in YS, 174, p. 4087. YS, 178, p. 4140. YS, 138, p. 3324; 204, p. 4551. YS, 24, p. 537. YS, 24, pp. 545-6, 549, 552. Cambridge Histories Online © Cambridge University Press, 2008 THE REIGN OF AYURBARWADA KHAGHAN 515 Chinese military families, as managers of governmental affairs of the Secretariat, giving them a large share of power in running the government. Among his earliest acts, Ayurbarwada summoned to the capital sixteen older officials who had served in Khubilai's court, including the famous scholars Li Ch'ien (1234—1312), Hao T'ien-t'ing (1261 —1317), Ch'eng Chii-fu (1249—1318), and Liu Min-chung (1243-1318), many of whom were subsequently appointed to high offices, and the others served as councillors." 2 Ayurbarwada also repeatedly ordered the selection of learned men to be members of the Han-lin Academy and the Academy of Scholarly Worthies (Chi-hsien yuan)."' His appointment of so many Confucian officials to important posts was by no means merely decorative. With their guidance and support, early in his reign the new khaghan was to take some important steps to reform the Yiian state further along Chinese and Confucian lines. The revival of the civil service examinations The most important aspect of Ayurbarwada's effort to reform the Yiian state was his vigorous attempts to make the bureaucracy more Confucian in its composition. Confucian scholars had not before played an important role in the Yiian government mainly because their type of learning had never in the earlier reigns been regarded as an appropriate basis for bureaucratic recruitment. Although most of the high-ranking offices were filled in accordance with ascriptive criteria - through either heredity or the exercise of the yin privilege - that is, appointment to a lower office by virtue of one's father's nomination - most of the officials of the middle and lower ranks came into service by way of service as clerks. Consequently, most of the officials were not educated in Confucianism nor were they Confucian in their political orientation. To improve the educational standards of the officials appointed through the yin privilege, it was stipulated in 1311 that the Chinese candidates for official appointment through the yin privilege would be required to pass a test in one of the classics and a historical work and that successful candidates in this test would be given substantive appointments without going through a probationary period. "* The test was to be optional for the Mongolian and se-mu candidates, but they too would be appointed to offices one grade higher than that to which they were originally entitled if they had passed the test. And perhaps to reduce the competition for the scholar-officials, the highest possible rank attainable by any official who had started as a clerical officer in 112 YS, 24, p. 537. 113 YS, 24, pp. 545-8. 114 YS, 83, p. 2061. Cambridge Histories Online © Cambridge University Press, 2008 516 MID-YOAN POLITICS local government was lowered from the fourth to the fifth rank."5 The most significant institutional change made by Ayurbarwada, however, was to revive the civil service examination system. The revival of this examination system, which had became increasingly important as the major channel of elite recruitment under the native Chinese dynasties, had repeatedly been debated but never put into effect during Khubilai's reign. Its revival was put off indefinitely for two reasons."6 First, because the Yuan government relied mainly on ascriptive criteria for elite recruitment, the adoption of the examination system would undermine the hereditary and yin privileges of the Mongolian, the se-mu, and, to a lesser extent, Chinese elite families. As such it was objectionable. Second, there was sharp disagreement even among Chinese scholar-officials about the value of the examination as an effective system of elite recruitment and about which curriculum should be adopted. One school favored adoption of the system practiced under the Sung and Chin dynasties, which had tested candidates' literary skills as well as their knowledge of Confucian classics. Opposing this view, Neo-Confucian scholars, heavily influenced by the great Sung master Chu Hsi's ideas about the examinations, urged the exclusion of literary composition from the curriculum and an exclusive emphasis on understanding the classics and current affairs. Because this debate was inconclusive, no action was taken to revive the examinations until Ayurbarwada's reign. By then the situation had changed. The growing need to improve the standard of the civil service and the expressed desire by the khaghan himself to identify his administration more closely with Confucianism gave rise to repeated pleas to revive the examinations.117 By this time, because the ascendancy of Neo-Confucianism had been clearly established in Yuan official circles, the position formerly propounded by the Neo-Confucianists was now accepted without much argument by those involved in the deliberations. The new examination system, promulgated in 1313 and put into practice for the first time in the following two years, clearly favored the NeoConfucian position. In its curriculum, classical learning was emphasized at 115 YS, 183, p. 4220; see also Hsii Fan, Yuan tai li chih yen cbiu (Peking, 1987), pp. 5 3 - 4 . 116 On the reasons for the long delay in reviving the examination system under the Yuan, see Abe Takeo, "Gendai chishikijin to kakyo," in his Genda'tshi no kenkyu (Tokyo, 1972), pp. 3 - 3 3 ; Ting K'unchien, "Yuan tai te k'o chii chih tu," Hua hsiiehyiieh k'an, 124 (1982), pp. 4 6 - 5 7 ; Yao Ta-li, "Yiian tai k'o chii chih tu te hsing fei chi ch'i she hui pei ching," Yuan shib chi pet fang min tsu shib yen cbiu cbi k'an, 6 (1982), pp. 2 6 - 5 9 , esp. pp. 3 3 - 8 ; Wm. Theodore de Bary, Neo-Confucian orthodoxy and the teaming of the mind-and-heart (New York, 1981), pp. 53—4; Yuan-chu Lam, "On the Yiian examination system: The role of northern Ch'eng-Chu pioneering scholars," Journal of Turkish Studies (Festschrift for Francis W. Cleaves), 9 (1985), pp. 1 5 - 2 0 . 117 YS, 175, p. 4089; 178, p. 4142; and Huang Chin, Chin-hua Huang hsien sheng wen chi (SPTK ed.), 4 3 . P- 5b- Cambridge Histories Online © Cambridge University Press, 2008 THE REIGN OF AYURBARWADA KHAGHAN 517 the expense of literary refinement. It further prescribed Chu Hsi's version of the Four Books as the authoritative texts for all candidates and the commentaries by Chu Hsi and other Sung masters on the Five Classics in an additional test for Chinese candidates. This innovation, which was generally followed in the Ming and Ch'ing dynasties, had a historical significance that went beyond the Yuan itself and helped consolidate Neo-Confucianism as the state's orthodoxy." 8 Besides this, the new examination system had some special features that reflected the peculiar multiracial society under the Yuan. Under the new system, Mongolian and se-mu candidates not only were given easier tests than those required of their northern and southern Chinese counterparts; they also were assured "equal representation" under a racial quota system that prescribed an equal number of successful candidates at the provincial level - seventyfive - to each of the four ethnic and geographical groups. Moreover, in order not to undermine the privileged positions of the old elite families and to upset the existing system of recruitment, the quota for the successful candidates at the metropolitan level was purposely kept low, to a maximum of ioo per examination. The sixteen examinations later administered under the Yuan produced a total of only 1,139 successful metropolitan candidates (chin-shih). The number of officials so recruited came to only slightly more than 4 percent of all ranked civil officials in that period. " 9 The examination system therefore by no means posed a serious threat to the interests of the members of the elite families who enjoyed hereditary or yin rights to official appointment. The system instituted by Ayurbarwada should be considered a compromise between the Chinese ideal of equal opportunities in public service and the sociopolitical realities of the Yuan as a dynasty of conquest. Nevertheless, his revival of the examination system still had cultural as well as sociopolitical significance. By making Confucian learning a basis for elite recruitment, it gave Chinese scholars a legitimate path to officialdom. Southern Chinese scholars especially benefited, as hitherto they had been largely excluded from the government under the Yuan. Moreover, the examination system encouraged the Mongols and the se-mu, especially those who did not belong to elite families, to study Chinese thought, thus accelerating the sinicization of the alien conquerors.120 118 De Bary, Neo-Confucian orthodoxy, pp. 57—66. 119 Yao Ta-ti, "Yuan tai k'o chii chih cu te hsing fei chi ch'i she hui pei ching," pp. 47—8. On the civil examination system set up by Ayurbarwada and its subsequent development, see also Miyazaki Ichisada, "Gencho chika no Mokoteki kanshoku wo meguru Mo Kan kankei - kakyo fukko no igi no saikento," Tayoshi kmkyu, 23 (1965), pp. 428—91; Yang Shu-fan, "Yuan tai k'o chii chih tu," Kuo li chevg chih ta hsiith hsiith pao, 17 (1968), pp. 99—120. 120 Hsiao Ch'i-ch'ing, "Yuan tai k'o chii yii ching ying liu tung - i Yiian-t'ung yuan nien chin shih wei chung hsin," Han hsiith yen chiu, 5 (1987), pp. 129—60. Cambridge Histories Online © Cambridge University Press, 2008 518 MID-YOAN POLITICS Codification Codification of the law was another area in which Ayurbarwada's efforts to reform the Yuan state produced the desired results. Perhaps because of the unsurmountable difficulties in legislating a uniform legal code for a multicultural society and perhaps also because of the reluctance by the Mongolian ruling elites to have their power circumscribed by such a code, the Yuan state had never enacted a national statutory code. Its absence caused the Chinese officials great anxiety, and various efforts were made to compile laws and subsidiary legislation to serve as references for use in conducting their duties. The earliest of those remedial measures, the Chih-yiian hsin-ko (Chih-yuan new code) promulgated in 1291 had been largely a collection of noncriminal ordinances.121 Later efforts at codification in the reigns of Temiir and Khaishan produced few results. Ayurbarwada quickly took action to remedy this situation. In the same month that he was enthroned in 1311, he instructed the Secretariat to systematize the codes and regulations promulgated since the beginning of Khubilai's reign. This compilation and editing was completed in 1316.122 The process of reviewing the collection, however, took much longer than expected. It was not until 1323, more than two years after Shidebala's succession to the throne, that the code was formally promulgated under the title Ta Yuan t'ung chih (Comprehensive institutions of the great Yuan). The new code consisted of more than 2,400 legal documents promulgated since the beginning of the dynasty and was divided into four categories: tuan-li (decided precedents), t'iao-ko (articles and codes), chao-she (decrees and amnesties), and ling-lei (miscellaneous ordinances).123 Though still short of being a comprehensive statutory code, the Ta Yuan t'ung chih, in the view of a modern legal historian, "marked a mature stage in Yuan legal history because of its substantial content and its adoption of the structure of the Chinese traditional code as represented in the T'ai bo lit," the code of the Chin dynasty, which, in turn, had been modeled on the T'ang code.I24 However, as the code of a dynasty of conquest, the Ta Yuan t'ung chih did not accept completely the previous Chinese codes. In many ways it reflected Mongolian customs and the institutional features peculiar to the 121 See Ch'en, Chinese legal tradition under the Mongols, pp. 14—23; and Uematsu Tadashi, "Ishu Shigen shinhaku narabini kaisetsu," Toyoshi kenkyu, 30 (1972), pp. 1-29. 122 See Ch'en, Chinese legal tradition under the Mongols, pp. 24—6; Po-chu-lu Ch'ung, "Ta Yuan t'ung chih hsii," in Kuo ch'ao wen lei, 36, p. 7a. 123 A part of the t'iao-ko section of the Ta Yuan t'ung chih has survived and was reprinted in 1930 by the National Peking Library under the title T'ung chih t'iao ko (Code of comprehensive institutions). 124 Ch'en, Chinese legal tradition under the Mongols, p. 29. Cambridge Histories Online © Cambridge University Press, 2008 THE REIGN OF AYURBARWADA KHAGHAN 519 dynasty.12' The Ta Yuan t'ung chih and the Yuan tien chang (Institutions of the Yuan), also compiled in Ayurbawada's reign by either the local government of Chiang-hsi or private authors, were two milestones in Yuan legal history, reflecting the Yiian's growing sophistication as a conquest dynasty. Translation and publication of books Ayurbarwada khaghan's fondness for Chinese culture and his and his officials' (especially the Mongols and the se-mu) desire to benefit from Confucian political wisdom and Chinese historical experience can be seen in the number and nature of the books translated or published with Ayurbarwada's authorization. 126 The Chinese works translated into Mongolian included the Confucian classic Shang shu (Book of history);127 the Sung compendium compiled by Chen Te-hsiu (1178-1235), Ta-hsueh yen-i (Extended meaning of the Great Learning); two works concerned with the great T'ang monarch T'aitsung (r. 627-49), the Chen kuan cheng yao (Essentials of the government of the Chen-kuan period) by Wu Ching (670-749) and the Ti-fan (Plan for an emperor) written by T'ai-tsung himself for his heir apparent; and the great history the Tzu chih t'ung chien (Comprehensive mirror for aid in government) by Ssu-ma Kuang (1019-86). Chinese works published with Ayurbarwada's endorsement include the Confucian classic the Hsiao ching (Book of filial piety); the Lieh nti chuan (Biographies of women) by Liu Hsiang (ca. 7 7 - 7 6 B.C.); studies of the Spring and Autumn Annals by the T'ang scholar Lu Ch'un, as well as the Yuan official agricultural work Nung sang chi yao (Essentials of agriculture and sericulture). I28 125 See Huang Shih-chien, "Ta Yuan t'ung chih k'ao pien," Cbang-kuo she hui k'ohsiieh, 2 (1987), pp. 157-71; see also Abe Takeo, "Daigen tsusei no kaisetsu," in his Gendaishi no kenkyu (Tokyo, 1972), PP- 2 5 3 - 3 I 9 126 On the translation of Chinese works into Mongolian during the Yuan dynasty, see Walter Fuchs, "Analecta zur mongolischen Ubersetzungsliteratur der Yiian-Zeit," Monumenta Strica, 11 (1946), pp. 33—46; Herbert Franke, "Chinese historiography under Mongol rule: The role of history in acculturation," Mongolian Studies, 1 (1974), pp. 15—26. For some of the studies and annotated translations of the Hsiao ching, the only surviving Mongolian translation of the Chinese works done in the Yuan dynasty, see F. W. Cleaves, "The first chapter of an early Mongolian version of the Hsiao Ching," Acta Orientalia Academiat Scientarum Hugaricae, 36 (1982), pp. 6 9 - 8 8 ; "The second chapter of a Mongolian version of the Hsiao ching," in K. Sagaster and M. Weiers, eds., Documenta Barbarorum. Festschrift fur Walter Heissig zum 70. Geburtstag (Wiesbaden, 1983), pp. 272—81; "The eighteenth chapter of an early Mongolian version of the Hsiao ching," Harvard Journal of Asiatic Studies, 45 (1985), pp 2 2 5 - 5 4 . See also Igor de Rachewiltz, "The preclassical Mongolian version of the Hsiao-ching," Zentralasiatischen Studien, 16 (1982), pp. 7—109; "More about the preclassical Mongolian version of the Hsiaoching," Zeatralasiatischen Studien, 19(1986), pp. 27—37. 127 On the following translated works, see YS, 24, pp. 536, 544; 25, p. 565; 26, p. 578; 137, p. 3311; 181, p. 4172. 128 See YS, 24, p. 536; 26, p. 587. Cambridge Histories Online © Cambridge University Press, 2008 520 MID-YOAN POLITICS Although the list of the Chinese works published under Ayurbarwada's auspices reflects his duty as the Son of Heaven in promoting popular morality and material welfare, the selection of works for translation also reveals his pragmatic aims. When he ordered the translation of the Chen kuan chengyao, he gave as his reason the benefits that the state might gain from the work and expressed his wish that the Mongols and the se-mu would study the translated text. "9 It is apparent that the Mongolian monarch wished that the Mongolian and the se-mu elites, including himself, would learn from Confucian political philosophy and Chinese historical experience, especially that of the glorious Tang T'ai-tsung's reign, in order to govern the Yuan state better. Failure in curtailing aristocratic privileges Ayurbarwada's attempt to reform the Yuan state along traditional Chinese lines could not go very far, as he was unable to centralize the administration by diminishing the administrative and judicial powers and economic privileges of the Mongolian princes. Despite Khubilai's centralization policies, the Mongolian princes had been left with considerable administrative, military, fiscal, and judicial powers in those areas under their appanage administrations (t'ou-hsia).l*° Further reduction of their powers was politically hazardous, for it would have entailed a frontal challenge to one of the most fundamental principles of the Mongolian-Yuan empire. Ayurbarwada's attempts to reduce the powers of the princes were aborted precisely for this reason. In the winter of 1311 the khaghan ordered the abolition of the jarghuchi (judges, Chinese: tuan-shih kuan) of the various princely establishments and placed all Mongolian violators of the law under the jurisdiction of the battalions (ch'ien-hu) to which they directly belonged.131 Because the jarghuchi, as representatives of their respective princes, had been in charge of judicial matters concerning the Mongols under the appanage administration, their abolition apparently meant the removal of the princes' direct jurisdiction over the Mongols under them.1'2 The khaghan's order seems to have been 129 YS, 24, p. 544. 130 The literature on the appanages known as the t'ou-hsia under the Yuan is vast. See, for example, Murakami Masastsugu, "Gencho ni okeru toka no igi," Moko gakuho, i (1940), pp. 169—21;; Iwamura Shinobu, Mongoru shakai keizaishi no kenkyu (Kyoto, 1964), pp. 401—69; Paul Ratchnevsky, "Zum Ausdruck Youhsia' in der Mongolenzeit," in Collectanea Mongolica: Festschrift fiir Professor Dr. Rintcben zum 60. Ceburtstag (Wiesbaden, 1966), pp. 173-91; and Chou Liang-hsiao, "Yuan tai t'ou hsia fen feng chih tu ch'u t'an," Yuan shih tun ts'ung, 2 (1983), pp. 53—76; Hung Chin-fu, "Ts'ung t'ou hsia fen feng chih tu k'an Yuan ch'ao cheng ch'iian te hsing chih," Cbung-yang yen yuan lisbihyii yen yen cbi so chi-k'an, 58 (1987), pp. 843—907. 131 YS, 24, p. 547. 132 On jargucbi, see Tamura Jitsuzo, Cbugoku seifuku ochono kenkyu (Kyoto, 1971), vol. 2, pp. 4 4 4 - 6 3 ; Cha-ch'i Ssu-ch'in (Jagchid Sechin), "Shuo Yuan shih chung te cha-lu-hu-ch'ih ping chien lun Yuan ch'u te Shang shu sheng," in vol. 1 of his Meng-iu shih lun ts'ung (Taipei, 1980), pp. 2 4 1 - 3 6 3 . Cambridge Histories Online © Cambridge University Press, 2008 THE REIGN OF AYURBARWADA KHAGHAN 52I enforced only for a short time, for it is known that both the prince of Chin, Yesiin Temiir (Yen-sun Tieh-mu-erh; the future emperor T'ai-ting, r. 1 3 2 3 8), and the prince of Chou (Khoshila; Ho-shih-la), Khaishan's son and the future emperor Ming-tsung (r. 1329), were allowed in 1316 to have several jarghuchi appointed under them. 133 The khagan's attempt to undercut the princes" administrative power provoked still more powerful opposition and also ended in failure. Urged by Temiider (T'ieh-mu-tieh-erh, d. 1322), the grand councillor of the right, in 1315 the khaghan ordered that the right of the princes to appoint darughachi (ta-lu-hua-chih), "overseers,"134 in their appanages would be taken over by the Secretariat, and the enfeoffed princes would be allowed to appoint only deputy darughachi.1^ One year later, even the right of the princes to appoint their own deputy darughachi was abrogated. Because the darughachi had been the chief official of the administration of a territorial division under an appanage and the post was usually filled by a personal retainer of the prince concerned, these moves provoked strong criticisms from certain imperial princes and from the Censorate, which charged that the administrations had departed drastically from both Chinggis khan's agreement with his brothers and the system set up by Khubilai. In the face of these accusations, the court was forced to withdraw its reform and in 1317 again allowed the appanage holders to appoint their own darughachi.Ii6 Little effort was made in Ayurbarwada's reign to curtail the various imperial grants that had been so lavishly granted to the princes by Khaishan. In the month of Khaishan's death, Ayurbarwada disbursed 39,650 taels of gold; 1,849,050 taels of silver, 223,279 ting of paper notes; and 472,488 lengths of silk to the princes who had come to attend his enthronement.' 37 Both the annual grants and those for special purposes continued to be as generous as in Khaishan's reign.' 38 The khaghan could not make any great reduction. Such imperial grants were a part of the Mongolian system, and he needed to ensure '33 Y$i 25, pp. 572—3; YTC, 9, p. 7a See also the discussion by Elizabeth Endicott-West in her Mongolian rule in China: Local administration in the Yiian dynasty (Cambridge, Mass., 1989), pp. 96—7. 134 Darughachi is a Mongolian term that literally means the "one who presses," in the sense of affixing a seal, hence the chief official of an office. Under the Yiian system, the darughachi were placed above titular officials of many central and local government offices. Their function was mainly supervisory rather than executive. With few exceptions, only Mongols and the se-mu were qualified to serve as darughachi. See Francis W. Cleaves, "Darugha and gerege," Harvard Journal of Asiatic Studies, 16 (1953), pp. 237—59; Cha-ch'i Ssu-ch'in, "Shuo Yiian shih chung te ta lu hua ch'ih," pp. 465—631. Elizabeth Endicott-West's Mongolian rule in China is the most systematic study of the darughachi. 135 YS, 25, p. 569. 136 YTC, 9, pp. 9 a - i o a ; YS, 25, pp. 5 7 3 - 4 ; 26, p. 579. See also Elizabeth Endicott-West's comment in "Imperial governance in Yiian times," Harvard Journal of Asiatic Studies, 46 (1986), p. 545, and her Mongolian rule in China, pp. 9 7 - 1 0 1 . 137 YS, 24, p. 538. 138 Shih Wei-min, "Yuan sui tz'u k'ao shih," p. 148; see also HYS, 78, pp. 7b-9a. Cambridge Histories Online © Cambridge University Press, 2008 522 MID-YOAN POLITICS the continued support of the princes, initially for his own enthronement and later for the appointment of his son Shidebala as the heir apparent, in violation of the agreement earlier made with his late brother. Moreover, such grants were necessary because the princes had been impoverished and urgently needed financial subsidy from the court in the form of grants. According to the statistics of 1319, the total number of "five-household silk households" {wu-hu-ssu hu) on which the members of nobility depended for their income had sunk to approximately one-fourth of their original quota set in 1236 when the system had been first established by Ogodei khagan (r. 1229— 41).139 The khaghan could not decrease their income further without seriously undermining the nobility system, which was an important cornerstone of the dynasty's political structure. Economic and fiscal policies Ayurbarwada's failure to curtail the imperial grants was just one of the reasons fot his inability to restore the government's fiscal health. The main reason was that the khaghan did not have any effective economic-fiscal policy, nor was he persistent enough in carrying out his other policies. His basic economic concept was typically Confucian: The way to alleviate the sufferings of the people was to be sparing in punishments and to lighten taxes so that they could lead their lives in their own way. I4° Consequently, in addition to abolishing Khaishan's exploitative policies, Ayurbarwada's administration attempted to reduce government expenditures by stopping the public building projects begun by Khaishan, cutting the numbers of supernumeraries, and imposing a tighter but still moderate rein on the imperial grants. If they had been carried out systematically, these policies could have cut governmental spending. But they were not: The efforts to abolish supernumerary offices and to curtail imperial grants were not continued."*1 The khaghan had no other plan to increase state revenue other than by encouraging agriculture early in his reign.'42 More drastic measures were taken in 1314 and 1315 when Temiider assumed his second term as grand councillor. It is not certain how many of these new measures, which were reminiscent of those of Khaishan's time, can be attributed to the khaghan himself. As we shall see shortly, because Temiider was Empress Dowager Targi's protege, the khaghan could exert little control over him, and Temiider's biography in the Yuan shih attri139 140 141 142 Iwamura, Mongoru sbakai'• keizaishi noktnkyu, pp. 458—61. YS, 26, p. 577. Li, Yuan sbih hs'm cbiang, vol. 3, pp. 3 7 7 - 8 . YS, 24, pp. 538, 552, 556, 558. Cambridge Histories Online © Cambridge University Press, 2008 THE REIGN OF AYURBARWADA KHAGHAN 523 butes all these measures to him.' 4 ' The measures that Temiider took to increase state revenue included the reestablishment of the state monopoly over foreign trade under the reinstituted Maritime Trade Supervisorate (Shih-po t'i-chii ssu) and the advance sale of salt licenses and iron products made by governmental foundries. I44 But the most important of Temiider's programs was renewing the cadastral survey (ching-li) carried out previously under Sangha. 14 ' This cadastral survey was first proposed by Chang-lii, a manager of governmental affairs left over from Khubilai's administration, and aimed at increasing the total land tax by discovering landholdings that had been fraudulently omitted from the tax registers.' 46 The program required landowners in the three provinces of Chiang-che, Chiang-hsi, and Ho-nan to report the size of their actual holdings to the authorities within forty days; failure to comply would result in severe punishment. If implemented properly, this survey would not only have greatly increased the state revenues, but it would also have helped establish a more equitable tax structure. Indeed, in the former Sung territories the rich landowners often did not pay taxes on their estates, but the poor still had to pay taxes even on land that they had already sold. When the survey was put into effect in the winter of 1314, ineffective implementation by overzealous or corrupt officials caused widespread hardship and resentment, especially among the rich landowners in southern Chiang-hsi. As a result, a serious revolt broke out at Ning-tu in Chiang-hsi in the fall of 1315, and its leader, Ts'ai Wu-chiu (r. 1315), proclaimed himself the king of Ts'ai. I47 Although the revolt was suppressed within two months, the government had to scrap the survey program completely and in 1316 even exempted the taxes on those lands that had been voluntarily declared.148 Consequently, Temiider's move to enhance state revenue, which was itself a deviation from Ayurbarwada's main policy line, ended dramatically and abruptly. There is little indication of any further positive move by the government to strengthen its financial position. 143 YS, 205, pp. 4 5 7 7 - 8 . 144 YS, 94, p. 2402; 205, p. 4578; Schurmann, Economic structun of the Yuan dynasty, pp. 224, 233. 145 On Sangha's cadastral survey, see Tadashi Uematsu, "The control of Chiang-nan in early Yuan," Ada Asiatica, 45 (1983), pp. 49—68, esp. pp. 5 7 - 6 0 . 146 YS, 25, pp. 466, 467, 571; 94, p. 2353; Schurmann, Economic structure 0/the Yuan dynasty, pp. 3 1 2; and Yang Yii-mei, "Yuan tai Chiang-nan t'ien fu sui chih k'ao," Chung-kuo It shih hsu'eh hui shih hstieh chi k'an, 21 (1989), pp. 143—70, esp. pp. 155-7. 147 On Ts'ai Wu-chiu's revolt, see Ch'en Kao-hua, "Yuan tai ch'ien ch'i ho chung ch'i ko tsu jen min te ch'i i tou cheng," in vol. 2 of Chung-kuo nung min chan cbeng shih lun ts'ung, ed. Lu Shu-ch'ing (Honan, 1980), pp. 2 8 6 - 3 2 0 , esp. pp. 3 0 6 - 8 ; Yang Ne and Ch'en Kao-hua, comps, Yuan tai nung min chan cheng shih liao hui pien (Peking, 1985), vol. 1, pp. 151—9. 148 YS, 93, p. 2353; Schurmann, Economic structure of the Yuan dynasty, p. 38. Cambridge Histories Online © Cambridge University Press, 2008 524 MID-YOAN POLITICS Factional struggles Ayurbarwada khaghan failed to achieve his ambition in reforming the Yuan state not only because of resistance from the imperial princes but also because his court was constantly plagued by bitter factional struggles. The khaghan was never the complete master of either his own house or his court, for his power was always greatly circumscribed by his mother, Empress Dowager Targi, and her henchmen. Targi, originally from the hereditary imperial consort clan, the Onggirad, was a woman of domineering character and questionable morality.'49 The khaghan never took any effective action to check her dominance. Consequently, under her patronage, her favorites in the Household Service of the empress dowager (Hui-cheng yuan) and in the Palace Provisions Commission (Hsiian-hui yuan) came to form a center of power rivaling the Secretariat. Among her favorites, Temiider was the dominant figure in Ayurbarwada's reign and in the early part of Shidebala's reign, and the bitter power struggle around him immobilized Ayurbarwada's government in the last two years of its existence. Though from a distinguished Mongolian family, Temiider owed his power and survival completely to the patronage and protection of the empress dowager.1'0 His steppingstone to power was the Palace Provisions Commission, which was in charge of the provisions and cuisine of the palace. As head of the commission since the beginning of Khaishan's reign, Temiider won the favor of the empress dowager and forged a close relationship with her that remained his political trump card throughout the rest of his life. The empress dowager appointed Temiider the grand councillor of the right, perhaps against Ayurbarwada's wishes, in February 1311, two months before the khagan's actual accession. Temiider served in that top administrative post for two years before his departure from the Secretariat for the first time in the spring of 1313.151 After being forced out of the Secretariat he is said to have strengthened his ties with Shiremiin (Shih-lieh-men), the commissioner of the Household Service of the empress dowager, and with other of her favorites.152 With her renewed support Temiider was triumphantly reappointed to the Secretariat in the fall of 149 For Targi's biography, see YS, 116, pp. 2900—3. On che importance of the Household Administration of the empress dowager (Hui-cheng yuan) as a center of power, see Fang Kuang-ch'ang, "Yiianshih k'ao cheng liang p'ien," Wen shih, 1 (1988), pp. 2 2 9 - 5 3 , es P- 2 3 I - 3 150 Temiider belonged to the Siikener, a subclan of the Barin. His grandfather, Biirilgidei (Pu-lien-chitai), was a distinguished general during Mongke's reign. His uncle, Khulu Bukha (Hu-lu Pu-hua), was the grand councillor of the left at the beginning of Khubilai's reign. See T'u, Meng-wu-erb sbih chi, 122, p. ib. 151 According to his biography in the Yuan Mb, Temiider resigned in 1313 because of ill health (YS, 205, pp. 4 5 7 6 - 8 1 ) . We learn elsewhere, however, that he was actually dismissed for his crimes. See Yii Chi, Tao yuan hsiieh ku lu (SPTK ed.), 16, p. ib. 152 Yii Chi, Tao yuan hsiieb ku lu, 16, p. ib; YS, 175, p. 4075. 153 YS, 25, p. 566. Cambridge Histories Online © Cambridge University Press, 2008 THE REIGN OF AYURBARWADA KHAGHAN 525 It is said that during his second term as the grand councillor of the right, Temiider became even more corrupt and despotic, "promoting and demoting officials according to his whims."1'4 Two antagonistic factions soon emerged: One was centered on Temiider, and the other was composed of Li Meng, Chang Kuei, the Khitan Hsiao Baiju (Pai-chu, d. 1320), a manager of governmental affairs of the Secretariat, and the censors headed by the Tangut Yang Dorji (Tuo-erh-chih, 1279-1320) and the Onggiid Chao Shih-yen (1260-1336). l » The clash came to a peak in the summer of 1317 when more than forty censors impeached Temiider on the grounds of corruption, extortion, tyranny, and factionalism. So overwhelming was the evidence that the khaghan ordered Temiider's arrest. However, because Temiider still enjoyed the empress dowager's protection, the khaghan could do nothing more than finally dismiss him from the grand councillorship. '' 6 Another round of clashes came in the early summer of 1319, when more than forty censors under the leadership of Chao Shih-yen, the vice-censor in chief, protested to the throne regarding Temiider's appointment to the prestigious post of grand preceptor of the heir apparent (t'ai-t'zu t'ai-shih) and cited a list of more than ten crimes that he had committed. Notwithstanding, Temiider's appointment was maintained at the insistence of the empress dowager. Chang Kuei, the manager of governmental affairs, who had also protested Temiider's appointment, was flogged, on the empress dowager's orders, for his opposition. By this time, all of Temiider's main detractors had been forced into retirement or exile. As the grand preceptor of the heir apparent, Temiider thus was able to control the government during the remaining six months of Ayurbarwada's reign. 1 " On the surface, the conflict between Temuder and his detractors appears to have been a typical case of an evil and corrupt prime minister pitted against loyal ministers and conscientious censors. Examined more deeply, however, the conflict was a bitter power struggle at the highest level of the government, with significant political and ideological implications. On the one hand, behind Temuder and his allies in the palace stood Empress Dowager Targi. Unlike the later powerful ministers of the reigns of Tugh Temiir (r. 1328-32) and Toghon Temiir (r. 1333-70), El Temiir (d. 1333) and the Merkid Bayan (d. 1340), who overshadowed their sovereigns with their own personal power, Temuder remained to a large extent the creature of the empress dowager and depended on her continued support for his survival. 154 YS, 176, p. 4112. 155 YS, 175, p. 4073; 179, pp. 4153-4; 180, pp. 4164-5; 205, pp. 4578-9. 156 YS, 26, p. 579; 205, p. 4579; Yii Chi, Taoyuan hsiieb kit /», 16 pp. ib—2a; 18, pp. 3b—4a; Huang Chin, Chin-hua Huang hsitn sheng wen chi, 4 3 , p. 4b. 157 YS, 26, p. 589; 205, pp. 4579—89; Hsu Yu-jen, Chih chmg chi (Liao-ch'eng, Shantung, 1911); repr. in vol. 7 of Yuan Jen wen chi then pen ts'ung k'an, comp. Wang Te-i (Taipei, 1985), 76, pp. 19b—20a; Yu Chi, Too yuan hsueh ku lu, 18, pp. l o b - n a . Cambridge Histories Online © Cambridge University Press, 2008 526 MID-YOAN POLITICS Temiider's self-enhancement meant the enhancement of the empress dowager's role in the government. In opposing her, it is likely that behind the impressive alignment of Temiider's detractors lay the encouragement, if not the active support, of Ayurbarwada khagan himself. Indeed, the khaghan must have resented his mother's ceaseless meddling in his government and her repeated intervention to put her favorites into powerful positions against his own wishes. Though not ideologically oriented, Temiider's faction, which consisted exclusively of Mongolian and the se-mu officials, was primarily interested in preserving the status quo, in which they had a great stake. They were, therefore, antipathetic toward the khaghan's reforms. Though not exactly a financier in the manner of Ahmad and Sangha,1'8 Temiider was not reluctant to adopt fiscal policies deemed by his detractors to be exploitative. His detractors, although a racially mixed group, were mostly Confucian scholarofficials united by their sympathy for the khaghan's reforms. They looked on Temiider and his partisans as a stumbling block to the achievement of their political goals. The reason for the Confucian detractors' failure to bring down Temiider was twofold: First was the khaghan's weakness in relation to his mother. The khaghan is known to have been an exemplary filial son "who never changed his attitude in serving the empress dowager."159 Filial piety was not only inherent in Ayurbarwada's nature, but it was also a cornerstone principle in the Confucian government over which he hoped to preside. Unwilling to oppose or offend his mother, the khaghan could therefore not eliminate Temiider. Second, the khaghan's Confucian policy was both politically and ideologically unacceptable to the establishment. The policy threatened to undermine the traditional political and economic privileges of the imperial princes and the Mongolian and se-mu officials. Consequently, the struggle of the Confucian faction against Temiider evoked little support from the Mongolian and the se-mu elite. Ayurbarwada khaghan died on 1 March 1320 at the age of thirty-five. His reign had begun with great hope and determined action and is still characterized by some historians as "the Confucian rule of the Yen-yu era" (Yen-yu juchih).l6° But in fact, although it added more Chinese features to the Yuan 158 John Dardess refers to Temiider as one of "the successors of the financial managers of Khubilai's time" (Conquerors and Confucians, p. 37). However, unlike Ahmad and Sangha, who were originally outsiders in the Mongolian imperial establishment and employed by Khubilai exclusively for their fiscal expertise, Temiider was always a part of the establishment, and his rise to power had nothing to do with financial matters. 159 YS, 26, p. 594. 160 Sun K'o-k'uan, "Chiang-nan fang hsien yii Yen-yu ju chih," in his Yuan tai Han wen hua le huo lung (Taipei, 1968), pp. 3 4 5 - 6 3 . Cambridge Histories Online © Cambridge University Press, 2008 THE REIGN OF SHIDEBALA KHAGHAN 527 state, it failed to curb the vested interests of the Mongolian and the se-mu elites and hence did not fundamentally affect the "constitution" of the Mongolian—Yuan state. THE REIGN OF SHIDEBALA KHAGHAN (EMPEROR YINGTSUNG), 1320-1323 Peaceful succession Ayurbarwada was succeeded by Shidebala, his eighteen-year-old son, on 19 April 1320. This was the only succession in the Yuan dynasty that took place peacefully according to the Chinese principle of primogeniture. The succession was peaceful mainly because Ayurbarwada had already taken care of possible disputes. Some sources claim that it was a part of the original agreement between Khaishan and Ayurbarwada that the throne would go to one of Khaishan's sons after Ayurbarwada's death. l6 ' The authenticity of these sources, however, remains dubious, as the records concerning this matter must have been revised in favor of Khaishan's sons after they had regained the throne in 1328. The idea of designating Shidebala, instead of Khoshila (1300-29), Khaishan's eldest son, is attributed variously in our sources to Empress Dowager Targi, to Temiider, and to Ayurbarwada khaghan himself.l62 Perhaps it was the general opinion in Ayurbarwada's court that the line of succession should be kept among Ayurbarwada's sons rather than revert to his elder brother's, for Ayurbarwada had carried out a bloody purge against Khaishan's ministers and had reversed all Khaishan's policies immediately after his death. In any event, Shidebala had been designated as heir apparent in 1316 and was made the nominal head of both the Secretariat and the Bureau of Military Affairs one year later.' 63 To ensure Shidebala's succession, Khoshila, his potential rival, was awarded the title of king of Chou and sent away from the capital in I 3 i 5 . l 6 t At one time Ayurbarwada had even toyed with the idea of abdicating the throne in favor of his son.' 6 ' As a result of all these arrangements, Shidebala succeeded to his father's throne without any opposition, three months after Ayurbarwada's death. 161 162 163 164 165 YS, 31, p. 639; 138, p. 3324. YS, 27, p. 579; 31, p. 693; 116, p. 2902; 138, p. 3325. YS, 25, p. 575. YS 25, p. 572; 31, p. 693. Wei Su, Wei T'ai-p'u wen hsii chi (Wu-hsing, 1914); repr. in vol. 7 of Yuan jen wen cbi chen pen ts'ung fan, comp. Wang Te-i (Taipei, 1985), 7, pp. 17b—18a. Cambridge Histories Online © Cambridge University Press, 2008 528 MID-YtiAN POLITICS Temiider's reign of terror Between Ayurbarwada's death in March 1320 and his own death in October 1322, Temiider attained even greater power than he had enjoyed under Ayurbarwada. He was reappointed as the grand councillor of the right for the third time by a decree of Targi, now the grand empress dowager, only three days after Ayurbarwada's death.'66 Because he enjoyed the full support of the grand empress dowager and because the khagan was an inexperienced young man whom he had supervised closely in his role as the grand preceptor of the heir apparent during the previous two years, Temiider was able to consolidate his power with ease, by placing his relatives and partisans in strategically important posts and by taking revenge against those who had attacked him or had proved uncooperative in the past. Among his partisans, Hei-lii and Mai-lii, the sons of his close ally *Ireshiba (I-lieh-shih-pa), and Chao Shihjung were successively made managers of governmental affairs of the Secretariat. Mubarak (Mu-pa-la) and Chang Ssu-ming (1260—1337) were appointed, respectively, assistant administrators of the right and the left of the Secretariat.'67 In addition, Tegshi (T'ieh-shih, d. 1323) was appointed censor in chief, and through him Temiider was able to control the traditional stronghold of his enemies, the Censorate.'68 Temiider also placed several of his own sons in important posts.'69 To root out the opposition, Temiider staged a reign of terror against his political enemies. Yang Dorji, Hsiao Baiju, and Ho Sheng (Bayan) (1264— 1320), the regent (liu-shou) of Shang-tu, who had exposed Temiider's corruption in one way or another, all were executed on various fabricated charges.'70 Li Meng, the late khagan's mentor, was demoted to academician expositor in waiting (shih-cbiang hsiieh-shih) of the Academy of Scholarly Worthies, and to underline his disgrace the tombstones of his ancestors were demolished.'7' Chao Shih-yen, who had led the censorial attack against Temiider in 1319; Wang I, the manager of the governmental affairs; Kao Fang (1264-1328), the assistant administrator of the right of the Secretariat; and Han Jou-yii (12601333), a councillor of the Secretariat all were framed by Temiider and were saved from death only by the khaghan's personal intervention.'72 Thus, early in Shidebala's reign, Temiider's power nearly amounted to a dictatorship. 166 167 168 169 170 YS, YS, YS, YS, YS, 18, 171 YS, 172 YS, 27, p. 598. 27, p. 598; 112, pp. 2282-5. 207, p. 4600. 27, pp. 623, 626. 205, p. 4580; 179, pp. 4154, 4151, 4157; Yii Chi, Too yuan bsiieh ku lu, 13, p. 6a; 16, p. 2a; p. 4a. 175, p. 4089. 27, p. 605; 136, p. 3303; 176, p. 4112; 205, pp. 4580—1. Cambridge Histories Online © Cambridge University Press, 2008 THE REIGN OF SHIDEBALA KHAGHAN 529 Shidebala's self-assertion Shidebala, the young khaghan, however, did not sit with folded hands. The throne soon became the focus of loyalty for the Confucian scholar-officials in their, struggle against the powerful Temiider. The khaghan was prepared for such a role, for he had been as well educated in Chinese as his father had been. When he was the heir apparent, court officials had repeatedly urged his father to appoint reputable scholars to educate him. 173 Among his tutors were the Chinese scholars Wang Chi and Chou Ying-chi; K'o Chiu-ssu (1290—1343), who was to become an eminent painter, calligrapher, and connoisseur; and Sewinch Khaya, the great Uighur lyricist.174 Deeply affected by Confucianism as well as by Buddhism, Shidebala could cite T'ang poems from memory and also was a creditable calligrapher. 1T> From the beginning of his reign, Shidebala showed a political independence and resolution beyond his years. He had dared to defy his grandmother's instructions even before his accession when she requested the replacement of Ayurbarwada's court officials with her own proteges. Indeed, the grand empress dowager was so enraged by his independence that she is quoted as saying, "We should not have raised this boy!"' 76 In a masterly move to counter the influence of the grand empress dowager and Temiider, in the summer of 1320, Shidebala appointed the twenty-two-year-old Baiju (Pai-chu, 1298—1323) as the grand councillor of the left. This appointment brought the young emperor two political advantages. First, Baiju's illustrious family background helped him rally the support of the old Mongolian aristocracy. The influence of Baiju's family was unparalleled among the Mongols, for he was a descendant of the great Mukhali (1170—1223), Chinggis khan's companion and general, and the grandson of An-t'ung (1245—93), the popular grand councillor of Khubilai. 177 Second, Baiju was uniquely qualified among the Mongols to rally support from the Confucian scholar-officials. An-t'ung had become a legendary figure for his brave defense of the Confucian principles of the government against Khubi173 YS, 26, p. 585; 174, p. 4061. 174 YS, 187, p. 4269; Yii Chi, Too yuan hsiieh ku lu, 19, p. 10b; Ou-yang Hsiian, Kuei chai chi (SPTK ed.), 9, p. 21a; Yang Lien, Kuan Yiin-shih p'ing chuan (Urumchi, 1983), p. 187; Tsung Tien, K'o Chiu-ssu nimp'u (Shanghai, 1963), p. 187. On Sewinch Khaya, see Richard J. Lynn, Kuan Yiin-sbih (Boston, 1980). 175 T'ao Tsung-i, Shu shih hui you (Hung-wu [1368-98} ed.; repr. Wu-chin, 1929; repr. Shanghai, 1984), 7, p. la; Hsu Yu-jen, Chih cheng chi, 73, p. 61b; Yoshikawa Kojiro, "Gen no shotei no bungaku," pp. 240—5; Herbert Franke, "Could the Mongol emperors read and white Chinese?" p. 73176 YS, 27, p. 599; 116, p. 2902. 177 On the influence of Baiju's family, see Hsiao Ch'i-ch'ing, "Yuan tai ssu ta Meng-ku chia tsu," in his Yuan lai shih hiin t'an (Taipei, 1983), pp. 141-230. Cambridge Histories Online © Cambridge University Press, 2008 53° MID-YOAN POLITICS lai's se-mu financiers.178 Baiju himself had received a good Confucian education and, as the commissioner for ritual observances (t'ai-ch'ang li-i shih) since 1315, had established close relationships with many Confucian scholars.179 Thus the khaghan and Baiju formed a young and vigorous team, able to circumscribe Temiider's influence in various ways and to protect Confucian scholar-officials from Temiider's persecution. The conflict between the two parties reached a climax only two months after the khaghan's enthronement, when a conspiracy to dethrone the khaghan and, perhaps, to put Udus Bukha (Wu-tu-ssu Pu-hua), his younger brother, on the throne in his place was discovered. l8° Because all the alleged conspirators were well-known favorites of the grand empress dowager and partisans of Temiider, the khaghan was uncertain how to deal with them. It was Baiju who encouraged the khaghan to take swift action and to execute them before the grand empress dowager and Temiider could intervene. Temiider himself, however, escaped unscathed and was even given a part of the properties confiscated from the conspirators. 18 ' Biology, nevertheless, favored the young khaghan. The aging Temiider's health was failing, and so the khaghan was able to give more power to Baiju. After Temiider's death in October 1322 and that of the grand empress dowager a month later, the khaghan was able to dismantle his faction.182 In the early half of the next year the late Temiider and his partisans were attacked for misappropriating public funds and for accepting bribes, and a thorough excoriation of Temiider was carried out at the instigation of the Censorate. l83 The faction that had exerted nearly unchallenged influence at the court for more than a decade was greatly weakened, but it was not yet finished. The reform of the Chih-chih era With both the grand empress dowager and Temiider out of his way, Shidebala khaghan was now completely his own master. Among his first moves after Temiider's death he appointed in December 1322 Baiju as grand councillor of the right. As the sole grand councillor throughout the rest of Shidebala's reign, Baiju combined in his person the roles of prime minister, 178 On An-t'ung, see Igor de Rachewiltz, "Muqali, B6I, Tas and An-t'ung," Papers on Far Eastern History, 15 (1978), PP- 4 5 - 6 2 179 On Baiju, see Huang Chin, Cbm-hua Huang hsiensheng wen chi, 24, pp. ia—8a; YS, 136, pp. 3300—6; K'uang Yii-ch'e, "Pai-chu chi ch'i hsin cheng," Nei Meng-ku she hui k'o hsiieh, 5 (1984), pp. 5 9 - 6 2 . 180 YS, 27, p. 602; YTC, hsin chi, chao-ling (decrees), p. 5a. Yang Chih-chiu contends that the alleged conspiracy was fabricated by Shidebala in order to isolate the grand empress dowager; see Yang, "Yuan tai Hui-hui jen te cheng chih ti wei," pp. 2 6 2 - 3 . 181 YS, 27, p. 603; 136, p. 3301; 175, p. 4075. 182 YS, 205, p. 4580. 183 YS, 28, pp. 626, 630—1; 124, p. 3046; 136, p. 3304; 205, p. 4581. Cambridge Histories Online © Cambridge University Press, 2008 THE REIGN OF SHIDEBALA KHAGHAN 531 moral preceptor, and remonstrator. He recruited for the government a great number of Chinese scholar-officials, many of whom had resigned when Temiider was in power. Heading this list, Chang Kuei, a veteran administrator, was reappointed manager of governmental affairs and became Baiju's chief partner in carrying out reforms. l84 Among other former senior officials, both Wu Yiian-kuei (1251-1323) and Wang Yiieh were appointed grand academicians of the Academy of Scholarly Worthies (Chi-hsien yuan), and Han Ts'ung-i was appointed grand academician of the Institute for the Glorification of Literature (Chao-wen kuan). These three elderly scholars were concurrently appointed as councillors to the Secretariat. Several famous scholars were appointed to the Han-lin Academy: Chao Chii-hsin; Po-chu-lu Ch'ung (1279-1333), the famous essayist; and Wu Ch'eng, now in his seventies.185 The Confucian scholar-officials thus regained a large measure of respect and influence in Shidebala's court. With the guidance and assistance of Baiju and these scholar-officials, the khaghan carried out a number of reforms, some of which were merely revivals or extensions of his father's policies.186 He revived his father's policy of retrenchment and attempted to impose stricter discipline on the bureaucracy. He proclaimed again his father's principle that all supernumerary posts created after Khubilai's reign be abolished and then proceeded to eliminate many offices subordinate to the personal establishments of the empress dowager and the empress. l87 A set of rules (chen-chu t'ai-kang) was promulgated in February 1323 that required censors to expose various kinds of official malfeasance.'88 It was approximately at this same time that the Ta Yuan t'ung chih was revised in order to rationalize the administration and facilitate the dispensation of justice.' 89 In the field of finance, a corvee-assistance scheme (chu-i fa) was adopted by May 1323 to alleviate the people's onerous burden of performing corvee.' 90 The government also stipulated that landowners set aside a certain proportion of the lands registered under their ownership from which revenues could be collected to cover corvee expenses. The apparent purpose of this system was to relieve the corvee burdens of small landowners, as they were not required to contribute to the scheme.' 9 ' 184 YS, 175. P- 4°74185 YS, 26, pp. 6 2 6 - 7 ; 183, p. 4220. 186 On Shidebala's reforms, see Hsiao Kung-ch'in, "Ying-tsung hsin cheng yii Nan-p'o chih pien," Yiian shih chi pel fang min liu shih yen chiu ch k'an, 4 (1980), pp. 36—46. 187 YS, 26, pp. 625; 175, p. 4079. 188 YS, 28, pp. 6 2 8 - 9 ; Nan t'aipeiyao, 1, pp. I 4 a - i j b , in Yung-le ta tien (repr. Peking, i960), chiian 2610-11. 189 YS, 28, pp. 628—9. See also Po-chu-lu Ch'ung's preface to the Ta Yiian t'ung chih in Kuo ch'ao wen lei, 36, pp. 6a~9a. 190 YS, 28, p. 630; see Ch'en Kao-hua, "Yiian-tai i fa chien lun," Wen chih, 11 (1981), pp. 157—73. 191 Huang Chin, Chin-hua Huang hum sheng wen chi, 27, pp. 9b—10a; 10, pp. 1 ib—12b. Cambridge Histories Online © Cambridge University Press, 2008 532 MID-YOAN POLITICS Though Shidebala was determined to continue his father's reforms, he was neither emotionally mature nor ideologically consistent. He was, after all, still only twenty years old at the time of his death in 1323. Though he made bold attempts to curtail governmental expenditures, he still enjoyed pomp and extravagance. Most extravagant was his patronage of Buddhism. Shidebala was as ardently devoted to Buddhism as he was to Confucianism, and he even once asked Baiju whether that religion could be relied on to rule the empire.'^2 He personally visited Wu-t'ai, the sacred mountain in Shansi, sent monks abroad for scriptures, and repeatedly funded the copying of scriptures in golden letters. Moreover, he ordered a temple greater in size than its Confucian counterpart to be built in every prefecture, in honor of 'Phags-pa (1235-80), the Tibetan lama who had been honored as the imperial preceptor (ti-shih) in Khubilai's reign. Most costly of all was the construction of the Ta chao-hsiao temple on the Shou-an Mountain west of Ta-tu, which took three years and the labor of tens of thousands of soldiers. So obsessed with the project was the khaghan that he executed or exiled four censors who remonstrated against it.*93 Perhaps also because of Shidebala's fanatical devotion to Buddhism, Islam suffered particularly severe discrimination during his reign.'« The mosque in Shang-tu was demolished to make room for a 'Phags-pa temple. The Huihui Directorate of Education (Hui-hui kuo-tzu chien), which was in charge of teaching the Persian language, was abolished.195 And Muslims other than the clergy were required to pay two taels of silver as pao-yin, the household tax payment in silver, from which they had previously been exempted. THE COUP D'feTAT AT NAN-P'O Regardless of the merits of Shidebala's reign, it came to an abrupt and tragic end on 4 September 1323 when the imperial party had encamped at Nan-p'o, thirty // south of the summer capital, on its way back from Shang-tu to Ta-tu. Late that night, Tegshi, the censor in chief, assisted by the Asud Guard (A-su wei) under his command, stormed the imperial camp, killing both the khaghan and Baiju.l?6 Directly participating in this coup d'etat were, among others, Esen Temiir (Yeh-hsien T'ieh-mu-erh), the 192 Huang Chin, Chin-bua Huang bsia-sheng wen cbi, 24, p. 5a. 193 The Grand Chao-hsiao Temple is presently known as the Wo-fu Temple on the Western Hills. See Chih cbeng chi, 47, pp. 69a—71b; Ch'en Kao-hua, Yuan Ta-tu (Peking, 1982), p. 74. 194 On Shidebala's anti-Muslim policies, see Yang Chih-chiu, "Yuan tai Hui-hui jen te cheng chih ti wei," pp. 263—4. 19; On the Hui-hui Directorate of Education, see Huang Shijian, "The Persian language in China during the Yuan dynasty," Papers on Far Eastern History, 34 (1986), pp. 83—95. 196 On the Nan-p'o incident, see YS, 28, pp. 632—3; 29, pp. 637—8; 136, p. 3305; 207, p. 4600; and Hsiao Kung-ch'in, "Ying-tsung hsin cheng ytt Nan-p'o chih pien," pp. 43—6. Cambridge Histories Online © Cambridge University Press, 2008 THE COUP D'ETAT AT NAN-P'O 533 manager of the Bureau of Military Affairs; Shiktiir (Shih-t'u-erh), the commissioner of the Grand Agricultural Administration (Ta ssu nung); Chigin Temiir (Ch'ih-chin T'ieh-mu-erh), the former manager of governmental affairs of the Secretariat; Oljei (Wan-che), the former manager of governmental affairs of the Branch Secretariat of Yunnan; *Sodnam, Temiider's son and a former secretarial censor; and another *Sodnam, Tegshi's younger brother and the commissioner of the palace provisions. Besides these highranking officials, five princes were involved: Altan Bukha (An-t'an[t'i] Puhua), the younger brother of the prince of An-hsi, Ananda; Bolod (Po-lo), the prince of Chi and grandson of Arigh Boke, Orliig Temiir (Yiieh-lu T'ieh-mu-erh), the son of Ananda and the recently enfeoffed prince of Anhsi; Kiiliid Bukka (unidentified); and Ulus Bukha (Wu-lu-ssu Pu-hua), a descendant of Mongke khaghan. After killing the khaghan, the conspirators rushed back to Ta-tu and seized control of the government. At the same time, envoys were sent to Yesiin Temiir, the prince of Chin, in Mongolia, requesting him to become the new khaghan. The list of conspirators reflects the nature of the conflict that had culminated in the assassinations. Tegshi, the chief conspirator, was both an imperial relative and a former protege1 of Temiider. He was from the Ikires, the hereditary imperial consort clan, and the son of Princess *Ilig Khaya (I-li Hai-ya), Temiir khaghan's daughter.' 97 Even more significant, *Sugabala (Su-ko-pa-la, d. 1327), his sister, was the empress of the young khaghan, Shidebala.'98 Tegshi himself, however, had become the foster son and protege of Temiir. After serving as commissioner of palace provisions in Ayurbarwada's reign, Tegshi emerged as one of the most powerful officials early in Shidebala's reign, as the censor in chief and the chief military commissioner of the imperial guards, the Chung-i and the Asud. Though he had been implicated in a major corruption scandal during the purge directed against the Temiider faction, Tegshi had nevertheless been pardoned by imperial decree, apparently because of his status as imperial brother-in-law. Like Tegshi, most of the other conspirators had been among the Mongolian and se-mu allies of Temiider. As a result they had been dismissed from office or were on the verge of dismissal. Their fear of punishment was aggravated as the purge against Temiider's partisans deepened in 1323. They, therefore, joined the conspiracy against the khaghan in order to save themselves. Significantly, five out of the sixteen named conspirators were princes. 197 For Tegshi's biographies, see YS, 207, pp. 4599—4600; T'u Mmg-wu-erb shih cbi, 122, pp. 4 6 - 5 3 . 198 YS, 114, p. 2876. Cambridge Histories Online © Cambridge University Press, 2008 534 MID-YOAN POLITICS Their actual participation in the conspiracy seems, however, to have been more widespread than the list of conspirators suggests. Kiimeijil (Hsii-maichieh, d. 1325), the grand councillor of the right of Yesiin Temiir, the new khaghan, told the khaghan shortly after his accession that Mai-nu was the only one among the imperial clansmen (tsung-ch'i) who had taken part in the conspiracy and had remained loyal to the court. 1 " The relationship of the princes with Temiider remains unknown. The princes as a group seem to have had ample reason for resenting Shidebala khaghan. Twice in his short reign, he had canceled the annual grants to the princes because of financial stringency — an act unprecedented in the history of the dynasty.200 Moreover, in order to increase imperial authority, Shidebala seems to have been stricter than his predecessors were in enfeoffing princes201 and to have attempted to discipline them more stringently.202 He had carried out these actions in disregard of the traditional privileges of the imperial clansmen, thereby prompting them to revolt. In brief, the tragedy at Nan-p'o was staged by the persecuted remnants of Temiider's faction in alliance with the discontented Mongolian princes. Viewed from a broader perspective, the assassination of Shidebala khaghan seems to have been the culmination of the conflict between two political groups that had existed at least since the beginning of Ayurbarwada's reign.203 Ayurbarwada and Shidebala, aided by their sinicized Mongolian ministers and Chinese scholar-officials, had made vigorous efforts to transform further the Yuan state along traditional Chinese lines, which meant increased centralization and bureaucratization. On the other hand, the faction led by Empress Dowager Targi and her proteges Temiider and Tegshi seems not only to have fought for their own interests but also to have represented a much wider constituency among the Mongolian and se-mu aristocracy and officials who were naturally disposed to oppose the reforms that had threatened to diminish their hereditary political andfinancialprivileges. Though Shidebala enjoyed a brief period of triumph after the deaths of Targi and Temiider, his further actions prompted the conspiracy that led to his own tragic death. 199 YS, 29, p. 642. 200 YS, 27, p. 606; 28, p. 621. 201 During Shidebala's three-and-a-half-year reign, only seven imperial relatives weere enfeoffed as princes, whereas fifteen were enfeoffed during Temur's reign, twenty-eight during Khaishan's reign, and thirty during Ayurbarwada's reign. See Noguchi Shuichi, "Gendai kohanki no ogo juyo ni tsuite," Shigaku 56, no. 2 (1986), pp. 169-99. 202 See, for example, YS, 28, p. 632. 203 Hsiao Kung-ch'in, "Ying-tsung hsin cheng yii Nan-p'o chih pien," pp. 41—3. Cambridge Histories Online © Cambridge University Press, 2008 THE REIGN OF YESON TEMOR KHAGHAN 535 THE REIGN OF YESON TEMOR KHAGHAN (EMPEROR T*AITING), 1 3 2 3 - 1 3 2 8 Shidebala's successor, Yesiin Temur, 2°4 was not merely the principal beneficiary of the conspiracy that had led to the murder of Shidebala but was also very likely a participant in it. From his family background and early history it can easily be imagined that Yesiin Temiir might have harbored imperial ambitions. As we have already seen, his father, Kammala, the eldest son of Chen-chin, had been a serious contender for the throne in 1294. Inheriting his father's mantle as the prince of Chin and the guardian of the four ordo of Chinggis in 1302, Yesiin Temiir was the most senior in line among Chenchin's grandsons and hence had been as eligible as Khaishan and Ayurbarwada had been for the throne in 1307. In the following reigns, Yesiin Temiir, with a large fief and powerful army in Mongolia, not only became one of the princes most respected by the court, being showered with various imperial favors, but he also emerged as the undisputed leader of the princes in the steppe. He undoubtedly shared with the other princes the discontent with the measures that Shidebala had taken against them as a group. Without the tacit agreement, if not the active encouragement, of Yesiin Temiir, Tegshi and the other conspirators probably would not have dared to commit regicide, the first such incident in Mongolian imperial history. It is known that Daula-shah (Tao-la-sha, d. 1328), the administrator {nei-shib) of Yesiin Temiir's princely establishment, had established close contact with the conspirators and that the latter had informed Yesiin Temiir of their plot two days before the actual murder, proposing to elect him as the new khaghan 204 YS, 29, pp. 637—8. Both Yesiin Temiir's age at death and the date of his bitth remain uncertain because the relevant passages in the basic annals of his reign in the Yiian shih are contradictory. The year of his birth is given as the "thirteenth year of [the] Chih-yuan [reign]" (1276), and his age at the time of his death in 1328 is given as 36™/ (YS, 29, p. 637; 30, p. 687). In a recent note Kao Wente argued that although 1276 is correct for his date of birth, his age at death should be changed to 33 sui. See Kao Wen-te, "Yiian T'ai-ting ti shou nien cheng wu," in vol. 1 of Min-tsu shih lun Is'ung, ed. Chung-kuo she hui k'o hsiieh yiian min tsu yen chiu so min tsu shih yen chiu shih (Peking, 1987), p. 38. I do not accept Kao's argument, and share with A. C. Moule and Li Tse-fen the view that the "thirteenth year," given in the Yuan shih, is a mistake for the "thirtieth year" of the Chih-yuan reign (1293). See Arthur C. Moule, The rulers of China (London, 1957), p. 103; Li Tse-fen, Yiian shih hsin chiang, vol. 3, p. 481. My reason is threefold: First, Kammala, Yesun Temiir's father, was born in 1263 (YS, 115, p. 2893). It would seem biologically impossible for him to have had a son when he was only thirteen years old. Second, Yesiin Temiir is said to have been born in the "residence of the prince of Chin" (YS, 29, p. 637). Kammala, however, became the prince of Chin only in 1292. Third, in the decree proclaiming his enthronement, Yesun Temiir referred to both Khaishan and Ayurbarwada as "elder brothers" (ko-ko) (YS, 29, p. 638). Yesiin Temiir is also referred to as "younger uncle" (shu-fu) by Tugh Temiir, Khaishan's son, in one of his decrees (YS, 32, p. 709). Because Khaishan was born in 1281 and Ayurbarwada in 128;, Yesiin Temur, as their younger cousin, could not have been born in 1276. It is likely that he was born in 1293 and died in 1328 at the age of thirty-five. Cambridge Histories Online © Cambridge University Press, 2008 536 MID-YOAN POLITICS should the planned assassination be successful. The basic annals of Yesiin Temiir's reign in the Yuan sbih state that Yesiin Temiir did try to warn Shidebala khaghan about the plot as soon as he had learned about it but that the assassination had taken place before his messenger reached Shang-tu. This version of the story, however, is probably a whitewash concocted by Yesiin Temiir's partisans after his enthronement.20' Tugh Temiir (r. 1328 and I 2 3 9~3 2 )» w ho seized the throne from Yesiin Temiir's son in 1328, later alleged that Yesiin Temiir had actually colluded with Tegshi in the murder, and his accusation was perhaps not unfounded.206 The purge Whatever his relationship with the conspirators, as soon as he had received the imperial seal sent by the conspirators, Yesiin Temiir enthroned himself on the bank of the Keriilen River in Mongolia on 4 October, exactly one month after Shidebala's murder. Even though his accession was the result of a bloody coup d'etat, in the edict issued to proclaim his enthronement Yesiin Temiir defended his legitimacy as the new khaghan on the basis of his seniority among the grandsons of Chen-chin and also the consensus reached among the princes and ministers in Mongolia and China.207 Perhaps in accordance with his original agreement with the conspirators, on the day of his accession Esen Temiir was made the grand councillor of the right, and Tegshi, the manager of the Bureau of Military Affairs. This agreement, however, was not to be honored for long, for one month later, once his party had established initial control over the two capitals, the new khaghan launched a bloody purge against his erstwhile allies, simultaneously in Ta-tu and Shang-tu. Esen Temiir, Tegshi, and other bureaucratic participants in the coup d'etat were executed, and the five princes who had been directly involved were exiled to distant places.208 The purge of the conspirators was Yesiin Temiir's masterstroke to boost his legitimacy. That is, he had 205 YS, 29, pp. 6 3 7 - 8 ; 136, p. 3305; 207, p. 4600. 206 YS, 32, p. 709. 207 YS, 29, pp. 638—9. Of all the edicts proclaiming the enthronement of Yuan khaghans, T'ai-ting's is the only one written in the vernacular Chinese language of the time and clearly reflects that it was translated from a Mongolian original. This shows that there were no Chinese literati accompanying Yesiin Temiir at the time of his enthronement. The text in classical Chinese now found in the Ch'ienlung editions of the Yuan shib is a later rendition. See Chang Yiian-chi, Chiao sbih sui pi (Taipei, '967), pp. H 2 a - i 3 b . 208 YS, 29, pp. 639—41. Yesiin Temiir may have reached a secret agreement with the officials in Ta-tu who remained loyal to the late Shidebala khaghan. Under the leadership of Chang Kuei, this group of officials sent a letter to Yesiin Temiir urging him to accept the throne and to eliminate the conspirators. See Chang Kuei's epitaph by Yii Chi in Su T'ien-chiieh, Kuoch'aoum Ui, 52, pp. 17b-18a. The relevant passage is missing in Yii Chi, Too yuan bsiitb ku lu, 18, p. 12a; see also pp. 12b—13a for the early reactions of the officials in Ta-tu to the murders. Cambridge Histories Online © Cambridge University Press, 2008 THE REIGN OF YESON TEMOR KHAGHAN 537 to draw a strict line between himself and the act of regicide, which was inexcusable from the viewpoints of both Mongolian and Chinese political ethics. In a spirit of revenge, the Chinese officials repeatedly urged the new khaghan to extend the purge to all former allies of Temiider and Tegshi and their families.209 But the khaghan refused because it was not in his interest to rely completely on one political faction in the bureaucratic establishment by eliminating another. 210 Principal members of the court In regard to the principal numbers of his court as well as the khaghan himself, Yesiin Temiir's was in all likelihood the most "un-Chinese" of all the administrations in the post-Khubilai era. Born in Mongolia in 1293, Yesiin Temiir came to the throne as a mature man at the age of thirty-one. He had come from a steppe background and is not known to have had any Chinese education. His chief ministers were mostly senior members of his princely administration brought from Mongolia, including Kiimeijil and Tas Temiir (T'a-shih T'ieh-mu-erh, d. ca. 1335), who served successively as grand councillors of the right; Daula-shah, who served as the manager of governmental affairs of the Secretariat, then as censor in chief, and finally as grand councillor of the left and was apparently the moving spirit behind the administration; and Andachu (An-ta-ch'u), the manager of the Bureau of Military Affairs.2'1 Perhaps because of the influence of Daula-shah, who was himself a Muslim, the Muslims gained unprecedented importance in this administration. 212 In the Secretariat there were two Muslims who served as managers of governmental affairs: Ubaidallah (Wu-po-tu-la, d. 1328), who had twice served in the same capacity in Ayurbarwada's and Shidebala's reigns but had been dismissed by Temiider, and Bayanchar (Po-yen-ch'a-erh), the brother of Saiyid Ajall Bayan. In the Bureau of Military Affairs, both Mahmiid-sha (Ma-mu-sha), Daula-shah's elder brother, and Hasan Khoja (A-san Ho-che) were its managers. Even Hasan (A-san), who had collaborated in Shidebala's murder, was kept on as the vice-censor in chief. None of Yesiin Temiir's Mongolian and Muslim ministers is known to have had a deep understanding of Chinese culture. In contrast with the Muslims, the Chinese exerted little influence on the administration. As the manager of governmental affairs, Chang Kuei was the 209 YS, 29, pp. 641, 646, 648; 175, p. 4075; Hsii Yu-jen, Cbih chengcbi, 76, pp. 2 o b - 2 i b . 210 Subsequently, even the confiscated properties of the executed conspirators were returned to their families; see YS, 29, pp. 6 4 9 - 5 0 . 211 YS, 29, p. 639; T'u, Matg-wu-erb shih cbi, 157, pp. 26a—28a. 212 Yang, "Yuan tai Hui-hui jen te cheng chih ti wei," pp. 2 6 4 - 6 . Cambridge Histories Online © Cambridge University Press, 2008 538 MID-YOAN POLITICS only high-ranking minister remaining from the previous administration and the only Chinese of any importance in the new government. His influence, however, was limited, as his recommendations often went unheeded. Aging and ailing, he retired from the Secretariat in 1325 to become the honored but politically unimportant Han-lin academician recipient of edicts.213 The other Chinese officials in the Secretariat, such as Yang T'ing-yii, Hsu Shih-ching (Hsu Heng's son), Shih Wei-liang (1273—1347), and Wang Shih-hsi, all were junior members.2'4 Thus Yesiin Temiir's administration marked a sharp break with the previous administrations and established the ascendancy of men of both steppe and Muslim backgrounds. Reconciliatory policies Given his own background and that of his chief ministers, Yesiin Temiir could not, of course, have been expected to continue the reforms of Ayurbarwada and Shidebala. But Yesiin Temiir's administration by no means represented a complete reversion to the steppe tradition. As the emperor of a dynasty in China, he could not turn back the clock of history. And as a ruler who had seized the throne by intrigue and violence, he needed to rally the widest possible support. The main tone of his administration, therefore, was reconciliatory, as he tried to win support from all key political and religious groups. To win the support of the bureaucratic establishment, the khaghan compensated for the injustices done by Temuder to many officials. The names of officials who had been killed, such as Yang Dorji, Hsiao Baiju, and Ho Sheng, were posthumously cleared. Li Ch'ien-heng and Ch'eng Kuei, the censors who had been exiled, and Wang I and Kao Fang, the members of the Secretariat who had been dismissed, all were recalled and reappointed within a few months of Yesiin Temiir's enthronement.21' To compensate for the murder of Baiju, his young son Darmashiri (Ta-erh-ma-shih-li) was appointed chief military commissioner of the Tsung-jen Guard formerly commanded by Baiju.216 In order to strengthen his legitimacy as khaghan of all the Mongols and to halt the revolt of the princes, Yesiin Temiir made strenuous efforts to win the goodwill of the princes of all branches of the imperial clan. Special care was taken to reconcile with the descendants of Darmabala, from whose hands he had acquired the throne. In 1324, he recalled two sons of Khaishan, Tugh Temiir, who had been exiled by Shidebala to Hainan Island, and Amugha 213 214 215 216 YS, 175, pp. 4074-83. YS, 112, pp. 2826-8. YS, 29, p. 640. YS, 29, p. 643. Cambridge Histories Online © Cambridge University Press, 2008 THE REIGN OF YESt)N TEMOR KHAGHAN 539 (A-mu-ko), who had been exiled to Ta-t'ung in Shansi. 2 ' 7 Tribute was secured in 1327 from Khoshila, Tugh Temiir's elder brother, who had fled to Chaghadai territories after his unsuccessful revolt against Ayurbarwada. Indeed, the exchange of gifts and tributes with the Chaghadai khans, Kebek (r. 1320-7) and Eljigidei (r. 1327-30), with the Il-khan Abu Said (r. 131735), and with the Qipchaq khan Uz-beg (r. 1313-41) became more frequent than ever before. 2 ' 8 To consolidate further the support of the imperial clansmen, Yesiin Temiir enfeoffed or reconfirmed twenty-four princes, compared with only seven enfeoffments made in Shidebala's reign. 2 ' 9 He also reversed the trend toward making the empire's military structure less feudal, by assigning many princes to military commands in China proper as well as in the steppe. 220 The annual and special grants that Shidebala had either suspended or curtailed were restored. In fact, the khaghan was so indulgent to the princes that he permitted some of them who had committed serious crimes to go unpunished. Toghto (T'o-t'o), the prince of Liao, a descendant of Temiige Odchigin, Chinggis's younger brother, is a case in point. Toghto was known to have killed more than one hundred persons of his own household. But the khaghan took no action against him, despite repeated impeachments by the Censorate.221 Thus in order to win the support of the princes, Yesiin Temiir not only reversed Ayurbarwada's and Shidebala's policy of curtailing the patrimonial—feudal privileges of the princes, but he also relaxed his control over them. In the cultural and religious realms, too, Yesiin Temiir tried to win the widest possible support. As the emperor of China, Yesiin Temiir duly showed his respect for the Confucian tradition. He dispatched officials to offer sacrifices at Ch'ii-fu, the sage's birthplace, shortly after his own accession. He also rejected proposals to scrap the civil service examinations and to reduce the number of sacrifices at the Imperial Shrine (T'ai-miao) from once a quarter to once a year.222 Most significant and rather surprising was Yesiin Temiir's revival of the classics colloquium (ching-yeri). The classics colloquium, an old Confucian institution, was a kind of imperial seminar during which distinguished scholars expounded to the 217 YS, 29, p. 656; 30, p. 680. 218 Saguchi Tom, "Jushi seiki ni okeru Gencho daikan to seiho san-oke to no rentasei ni tsuite," pp. 173-8. 219 See Hsiao Kung-ch'in, "Ying-tsung hsin cheng yii Nan-p'o chih pien." 220 YS, 29, pp. 646, 647, 649, 651; 30, pp. 669, 670, 672, 677, 678, 669; 117, p. 2910. 221 YS, 29, pp. 644, 646; 175, p. 4076; Hsu Yu-jen, Chih chmg chi, 76, pp. 22a-b. See also Horie Masaaki, "Temuge Otchigin to sono shison," Toyo shim, 24-5 (1986), pp. 225—70, esp. 240—50; and Yeh Hsin-min, "Ou-ch'ih-chin chia tsu yii Meng Yuan han t'ing te kuan hsi," Net Meng-ku ta hsiith hsiieh pao (Che-she-pan), 2 (1988), pp. 14-26. 222 YS, 29, pp. 640, 641; 172, p. 4027. Cambridge Histories Online © Cambridge University Press, 2008 54O MID-YOAN POLITICS emperor the tenets of the classics and their relevance to the problems of the day. Though it had been informally held from time to time since Khubilai's time, the colloquium had never been formally revived under the Yuan. Because Yesiin Temiir was not proficient in the Chinese language, the classics colloquium was conducted through translation after its revival in 1324. The lecturers included such eminent Chinese scholars as Wang Chieh (12751336); Chao Chien; Wu Ch'eng; Yii Chi (1272-1348); Ts'ao Yiian-yung (d. 1329); Teng Wen-yuan (1259-1328); Chang Ch'i-yen (1285-1352); Khutulugh Tormish, the great Uighur translator; and Arukhui (A-lu-wei), the famous Mongolian lyricist. The original scope of the lectures was restricted to previously translated texts, but several other works were compiled and translated specially for the purpose.22' The actual function of the lectures was perhaps no more than to educate and acquaint the Mongolian sovereign with Chinese political ideas and the dynasty's history. But the lectures apparently exerted little influence on the actual political orientation of the government, as Chao Chien, the grand academician of scholarly worthies who had originally proposed the establishment of the colloquium, expressed his disappointment in 1327 that no single policy had originated from the lectures in the colloquium."* Yesiin Temiir did not revere Confucianism exclusively. As a typical Mongolian ruler, he also favored Islam and, in particular, Buddhism. Apparently due to the influence of the Muslim ministers in his government, Yesiin Temiir showered Islam with favors. Public funds were used in 1304 to build the mosques in Shang-tu and Ta-t'ung in Shansi."' And in the same year Dashman (Ta-shih-man) and Erke'iin (Yeh-li-k'o-wen), Muslim and Christian clergymen, respectively, were exempted from corvee.226 Muslim merchants, especially, reaped great profits from the government. Although the debts owed by the ortogh (Muslim merchant associations) to the government were canceled in 1324, the court paid more than 400,000 ting in the same year and another 102,000 ting three years later to Muslim merchants to settle the debts accumulated through the purchase of curios."? The khaghan's patronage of Confucianism and Islam was in both cases politically motivated, but his patronage of Buddhism was a matter of personal faith. He was as much devoted to the religion as Shidebala had been. In addition to spending enormous sums on the construction of Buddhist tem223 YS, 29, p. 644; Yii Chi, Too yuan lei kao (Yuan Fu-chou lu ju hsiieh, 1343); repc. in vols. 4 and 5 of Yiian jen wen chi chat pen ts'ung k'an, comp. Wang Te-i (Taipei, 1985), 33, pp. 16b—18a. 224 Yii Chi, Too yiian lei kao, 33, p. 17b. 225 YS, 29, p. 648. 226 YS, 29, p. 652. 227 YS, 30, p. 678; 175, p. 4077; Elizabeth Endicott-West, "Merchant associations in Yiian China: The ortogh," Asia Major, 1 (1989), pp. 1 2 7 - 5 4 , esp. pp. 1 4 9 - 5 1 . Cambridge Histories Online © Cambridge University Press, 2008 THE REIGN OF TUGH TEMUR 541 pies and pagodas and on performing sacrifices,228 Yesiin Temiir asked Kun Dgah blog ros pa, the imperial preceptor, to perform repeatedly the ceremony of initiation for himself and his family. Indeed, the favor enjoyed by the imperial preceptor was so great that in 1326 Bsod-nams btsang-pa (Sounan tsang-pu), his younger brother, was appointed the prince of Pai-lan (sa ten dbang) and married to an imperial princess, and when his elder brother came to the capital, the officials of the Secretariat were ordered to honor him with a feast on the way.229 Not only were the members of the family of the imperial preceptor greatly honored; many Lamaist monks also were granted honorific titles and given gold or jade seals. 2 ' 0 Yesiin Temiir was, indeed, an upholder of the Mongolian tradition of treating all religions with impartial honors. But this was done, of course, at great financial and other cost. THE REIGN OF TUGH TEMUR (EMPEROR WEN-TSUNG), 1328-1332 The war of restoration The death of Yesiin Temiir khaghan at the age of thirty-five in Shang-tu on 15 August 1328 immediately touched off the bloodiest and most destructive succession struggle of all Yuan history, a struggle that resulted in the imperial line's shifting back to the Khaishan house for the rest of the dynasty. 23" After Yesiin Temiir's death, his young son, *Aragibag (A-la-chi-pa, d. 1328), also known by his reign title as Emperor T'ien-shun, who had been designated as heir apparent four years earlier, was duly enthroned at Shang-tu in October by Daula-shah, the grand councillor of the right; Toghto, the prince of Liao; and Ongchan (Wang-ch'an d. 1328), the nephew of Yesiin Temiir and the recently invested prince of Liang. Despite this, a coup d'etat had already taken place in Ta-tu, with the objective of restoring the sons of the late Khaishan khaghan as the legitimate successors to the throne. 232 The attempt to restore Khaishan's imperial line had been a political undercurrent ever since 1316 when Ayurbarwada desig228 Cha-ch'i Ssu-ch'in (Sechin Jagchid), Meng-kuyti Hsi-tsang it shih kuan hsi chihyen chiu (Taipei, 1978), pp. 289, 2 9 4 - 5 . 229 YS, 30, p. 669. On Bsod-nam btsang-po, see Louis Hambis, he chapitre cviii du Yuan cbe: La Fiefs attribute aux membres de la famille imptriale et aux ministres de la cour mongolt d'apr'a I'histoire cbinoise offuielle de la dyaastie mongolt, Monographies du T'oung Poo, vol. 3 (Leiden, 1954), p- 137. See also Ch'en Ch'ing-ying, "Yiian-ch'ao tsai Hsi-tsang so feng P'ai-lan wang," in Hsi-tsang yen chiu, 4 (1983), pp. 29—32, for a discussion of the title of the prince of Pai-lan. 230 YS, 202, p. 4521. 231 On the restoration of 1328, see Dardess, Conquerors and Confucians, pp. 3 1 - 5 2 ; see also Hsiao Kungch'in, "Lun Yuan tai huang wei chi ch'eng wen ti," pp. 3 2 - 6 . 232 YS, 31, p. 694; 32, p. 704; 138, p. 3326. Cambridge Histories Online © Cambridge University Press, 2008 542 MID-Y(JAN POLITICS nated Shidebala, his own son, instead of Khoshila, Khaishan's eldest son, as the heir apparent. The death of Yesiin Temiir merely gave this undercurrent an opportunity to surface. The restoration movement involved three key figures. Although Khoshila and Tugh Temiir, his half-brother, emerged as the dual candidates for the throne, El Temiir, an old follower of Khaishan, was the actual moving spirit behind the movement. Both Khoshila and Tugh Temiir had suffered political persecution during Ayurbarwada's and Shidebala's reigns. Khoshila, invested as the prince of Chou, had been sent away from the capital to remote Yunnan in 1316 in order to clear the way for Shidebala's designation as heir apparent. In protest, Khoshila staged a pointless and abortive revolt in Shensi and subsequently, for the next twelve years, became a political refugee in the Tarbagatai region in Central Asia under the protection of the Chaghadai khans. 2 " Shidebala had exiled Tugh Temiir to the tropical island of Hainan in 1321. After being recalled by Yesiin Temiir three years later, Tugh Temiir was then sent to Chien-k'ang (Nanking) and Chiang-ling (Chiang-ling, Hupeh), with the title of the prince of Huai.234 Significantly, as the persecuted sons of a former khaghan, the two brothers still enjoyed a measure of sympathy among the imperial princes and, more importantly, the lingering loyalty of some of their father's followers who had survived various political vicissitudes. But it was mainly due to El Temiir's political ingenuity and military power that Khaishan's line was restored. El Temiir was from a Turkish Qipchaq family that had achieved great distinction in the steppe wars against the rebel princes Nayan (Nai-yen d. 1287), Khaidu, and Du'a.235 Both Chong'ur (Ch'uang-wu-erh, 1260— 1322), El Temiir's father, and El Temiir himself had been generals loyal to Khaishan in the steppe wars and his supporters in his seizure of the throne. Consequently, the family's fortune reached its zenith during Khaishan's reign. It had, however, suffered heavy losses in the following two reigns. At the time of Yesiin Temiir khaghan's death, El Temiir was holding the relatively modest but pivotal post of assistant manager (ch'ien-yuan) of the Bureau of Military Affairs. It was probably the combination of partisan loyalty to Khaishan's house and his personal ambition to restore his family glory that motivated El Temiir to conspire against Yesun Temiir's heir. Not everyone who participated in the movement had ties as close as El Temiir's to Khaishan's house or had the restoration of Khaishan's imperial line 233 YS, 31, pp. 693—4; see also the study on Khoshila by FujishimaTateki, "Gen noMinso no shogai," Otani sbigaku, 12 (1970), pp. 14-28. 234 YS, 35, p. 387; 32, p. 703. 235 On El Temiir, see YS, 138, pp. 3 3 2 6 - 3 4 ; Ma Tsu-ch'ang, Ma Shih-t'ienwm chi, 14, pp. 6 b - n a ; see also Dardess, Conquerors and Confucians, pp. 1 0 - 1 1 , 3 9 - 5 0 . Cambridge Histories Online © Cambridge University Press, 2008 THE REIGN OF TUGH TEMCR 543 as his main motive for joining the movement. Some may have been motivated by the desire to avenge Shidebala's murder, 2 ' 6 whereas others joined the movement because they disliked or felt threatened by Daula-shah and Toghto and Ongchan. 2 ' 7 In its initial stage, the movement should therefore be considered an attempt by Darmabala's descendants to restore rule. But because by this time there were no surviving heirs of Ayurbarwada, restoring the sons of Khaishan to the throne became the movement's rallying cry. 2 ' 8 El Temiir organized the movement as soon as Yesiin Temiir khaghan had fallen ill in the late spring of 1328, planning to stage a coup d'etat simultaneously in the two capitals. Though relatively junior in rank, El Temiir enjoyed the advantage of being in control of the imperial guards in Ta-tu during the khaghan's absence, for it was normal Yuan practice for the chief officials of the Bureau of Military Affairs and of the Secretariat to accompany the khaghan on his annual trip to the summer capital. Collaborating with Aratnashiri (A-la-te-na-shih-li), the prince of Hsi-an who had been assigned to take charge of the capital during the khaghan's absence,239 El Temiir successfully stormed the palace at dawn on 8 September, rounding up Ubaidallah and the remaining officials of the administration in Ta-tu. In Shang-tu, however, eighteen of El Temiir's collaborators were discovered and executed. Thus the two capitals fell into the hands of opposing camps. After seizing control of Ta-tu, El Temiir immediately organized a temporary government and notified Tugh Temiir in Chiang-ling and Bayan of the Merkid tribe in Ho-nan. 24 ° Bayan, a former junior aide to Khaishan in the steppe wars, emerged as another important figure in the restoration. 241 As the manager of governmental affairs of the Ho-nan Branch Secretariat, Bayan was able to seize control of that strategically located province, raise considerable armies and funds, and personally escort Tugh Temiir to Ta-tu, where Tugh Temiir was enthroned on 16 October, promising to abdicate in favor of Khoshila, his elder brother, when he arrived from Central Asia. In the ensuing conflict, the loyalists at Shang-tu had a more just cause, for they were supporting the designated heir of the late khaghan. This moral 236 Jen Su-ko, one of Shidebala's former confidants, is said to have initiated the idea of a revolt to avenge Shidebala's murder. He discussed this idea with Su-su, a Uighur and former assistant director of the right of the Hu-kuang Branch Secretariat, and persuaded El Temiir to lead the movement even before Yesun Temiir's death. See YS, 184, pp. 4 2 3 5 - 7 . 237 YS, 32, p. 704. Daula-shah is said to have arrogated power to himself and to have protected Muslim merchants and corrupt officials. See YS, 32, p. 707; 182, p. 4194; 176, p. 4112. 238 On the heirs of Ayurbarwada, see Hambis, Le chapitrt cvii du Yuan cht, p. 138. 239 Aratnashiri was the son of Tura, the prince of Yueh; see Hambis, Le chapitrt cvii du Yuanche, pp. 57— 8, 6 1 - 2 . 240 YS, 31, pp. 694-5; 3 2 . PP- 704-5; 138, PP- 3326-7. 241 On the Merkid Bayan, see YS, 138, pp. 3335-9; Ma Tsu-ch'ang, Ma Shih-t'itn wen chi, 14, pp. ia— 5a; Louis Hambis, "Notes prlliminaires a une biographic de Bayan le Markit,"Journal Aiiatiquc, 241 (1955), pp. 215-48; Dardess, Conquerors and Confucians, pp. 53—74. Cambridge Histories Online © Cambridge University Press, 2008 544 MID-YOAN POLITICS advantage, however, was not enough; rather, it was the military power and the political considerations of the participants, not any considerations of legitimacy, that were the decisive factors in the post-Khubilai succession struggles. In contrast, the Ta-tu restorationists enjoyed enormous geographical and economic advantages over the loyalists. They had at their disposal the extensive human and material resources of Chung-shu, Ho-nan, Chiang-che, Chiang-hsi, and Hu-kuang provinces, whereas the loyalists at Shang-tu enjoyed the support of only Ling-pei, Liao-yang, Shan-hsi, Ssu-ch'uan, and Yunnan provinces, all of which were either economically poor or geographically peripheral.242 Equally important was the fact that no one on the loyalist side could match the leadership of El Temiir, who constantly moved with singleminded determination and superb political acumen and military prowess. Initially, the loyalist forces had the upper hand, being able to break through the Great Wall at several points and penetrate as far as the outskirts of Ta-tu. Personally leading the fighting on several fronts, El Temiir, however, was able to turn the tide quickly in the restorationists' favor. What proved fatal to the loyalists was a surprise attack launched by the restorationists from Manchuria and eastern Mongolia. Perhaps under the influence of Bukha Temiir (Pu-hua T'ieh-mu-erh), El Temiir's uncle and the head of the Eastern Mongolian Chief Military Command (Tungrlu Meng-ku chiin tu yiian-shuai fu), many of the eastern Mongolian princes supported the restorationist cause. Their army, under the command of Bukha Temiir and Orliig Temiir (Yiieh-lu T'ieh-mu-erh), the prince of Ch'i and a descendant of Chinggis khan's younger brother, surrounded the city of Shang-tu on 14 November, at a time when most of the loyalist forces were involved on the Great Wall front. Thunderstruck, the Shang-tu court was compelled to surrender on the very next day. Daula-shah and most of the leading loyalists were taken prisoner and later executed. Aragibag, the young khaghan, was reported to be missing.24' With the surrender of Shang-tu, the way to restoring Khaishan's imperial line was cleared. The capitulation of the loyalists at Shang-tu did not mean complete victory for the restorationists. Loyalists elsewhere carried on fighting for much longer. Indeed, the loyalists in Shansi did not lay down their arms until December 1328,244 and their counterparts in Szechwan surrendered only in the following May.245 With the support of the province's aboriginal tribes, 242 243 244 245 Dardcss, Conqueron and Confucians, pp. 39-42. YS, 32, pp. 605—715; 138, pp. 3326—31; Ma Tsu-ch'ang, Ma shib-i'ien wen chi, 14, pp. 6a-l 11a. On the Shansi loyalists, see YS, 32, pp. 712—19; 137, pp. 3314-17. On the Szechwan loyalists, see YS, 3 2 - 3 . See also Feng Ch'eng-chun, "Yuan tai te chi ko Nan-chiat'ai," in his Hsi-yii Nan-hai shih li k'ao cheng tun cbu hut chi (Kowloon, 1976), pp. 200-16, esp. pp. 213-16. Cambridge Histories Online © Cambridge University Press, 2008 THE REIGN OF TUGH TEMOR 545 the loyalists in Yunnan under the leadership of Prince Tiigel (T'u-chien), a follower of Ongchan, fought doggedly and persistently for more than four years and gave up their cause only in March 1332.2+6 All in all, the war of restoration and its repercussions made the entire reign of Tugh Temiir a wartorn era. Regicide at Ongghochatu The war with the loyalists, however, constituted only a part of the restoration. The most dramatic and tragic episode of this entire process took place between Khaishan's two sons. Although neither Tugh Temiir nor El Temiir was prepared to turn over the throne to Khoshila, who had so far played no part in the restoration effort, due deference had to be shown to him, as he was the eldest son and legitimate heir of Khaishan and the restoration had been made in his name. Consequently, after the seizure of Shang-tu, the Ta-tu court duly notified Khoshila, urging him to come and assume the throne. Accompanied by the Chaghadai khan, Eljigidei, Khoshila in response enthroned himself on 27 February 1329 north of Khara Khorum on his way to China from the Tarbagatai region and appointed Tugh Temiir as his heir apparent, in accordance with the precedent set by their father and uncle. Mistaking Tugh Temiir's and El Temiir's polite gesture as a sincere offer, Khoshila had proceeded to appoint his own loyal followers to important posts in the Secretariat, the Bureau of Military Affairs, and the Censorate, thus threatening to undermine the political machinery so painstakingly created by Tugh Temiir and El Temiir in China. But this only strengthened El Temiir's determination to get rid of him.2*? The two royal brothers met on 26 August in a seemingly joyful reunion at Ongghachatu, where Khaishan khaghan had built the city of Chung-tu. Four days later, however, Khoshila was found dead. On 8 September Tugh Temiir assumed the throne at Shang-tu for the second time. The death of Khoshila was apparently the result of a conspiracy masterminded by El Temiir, probably in collaboration with Tugh Temiir. The basic annals of Khoshila's reign in the Yuan shih state that the luckless khaghan died of violence.248 Ch'iian 246 On the Yunnan loyalists, see YS, 33—6; Francis W. Cleaves, "The Lingji of Amy of 1340," Harvard Journal of Asiatic Studies, 25 (1964—65), pp. 3 1 - 7 9 ; Tu Vii-t'ing, Yuan tai Lo-lo-ssu shih liao chi k'ao (Chengtu, 1979), pp. 3 0 - 4 3 . 247 YS, 31, pp. 696—7; Fujishima Tateki, "Gen no Minso no shogai," p. 22. Khoshila and his followers apparently were not fully aware of the circumstances under which they lived and consequently handled the situation badly. According to an oral tradition still current in the early Ming dynasty, Khoshila's retainers treated El Temiir discourteously when he went to Khoshila's camp in Mongolia to present the imperial seal, thus making him both angry and fearful. See Hu Ts'ui-chung, Yuan shih hsiipirn (SKCS ed.), 10, p. 15a. 248 YS, 31, p. 701. Cambridge Histories Online © Cambridge University Press, 2008 546 MID-YOAN POLITICS Heng, a private historian, states explicitly that he was poisoned and that El Temiir directed his murder.249 Late in 1340 Toghon Temiir khaghan (Emperor Shun-ti, r. 1333-70), Khoshila's son, blamed Tugh Temiir for his father's death and, in revenge, posthumously expelled the latter from the Imperial Shrine.2'0 Khoshila's failure to win the throne as a "steppe candidate," as his father had done twenty years earlier, is seen by some historians as reflecting the final passing of the steppe frontier as a factor in Yuan politics and the triumph of the Yuan dynasty over the Mongolian empire.2'' Certainly the importance of the princes in the steppe region seems to have decreased after the restoration of intra-Mongolian peace in 1303 and the establishment of the bureaucratic Branch Secretariat of Ling-pei in Mongolia during Khaishan's reign. This fact, however, had not prevented the seizure of the throne by Yesiin Temiir in 1323 as a "steppe candidate" in close collaboration with the conspirators in Shidebala's court. More important were the vast differences between Khoshila and his father in regard to the bureaucratic support they enjoyed in the capital and the military power at their disposal. As the princely overseers of all forces in the steppe region, Khaishan had always been a part of the Yuan imperial establishment and had kept in close contact with the court. He thus was the favored candidate of the bureaucratic establishment in the capital under the leadership of Harghasun when the succession crisis had broken out, and although at first Ayurbarwada gained control of the court, he dared only to assume the title of regent, not the imperial title. By contrast, Khoshila had been a political refugee in distant Central Asia for twelve years, and by the time of the regicide, Tugh Temiir and El Temiir had already created their own powerful political machinery in China, to which Khoshila was a complete stranger. Moreover, as the supreme commander of the most powerful field army in the empire, Khaishan had taken thirty thousand men with him when he set out to contend for the throne, whereas Khoshila, as an erstwhile refugee, brought with him to Ongghochatu only eighteen hundred guardsmen and so stood no chance whatever of overpowering his younger brother by force.2'2 Khoshila's failure to capture the throne, therefore, should be attributed to his personal status as a political 249 Ch'iian Heng, Keng sben wai shih, in Poo yen t'angpi cbi, ed. Ch'en Chi-ju (1906; repr. Taipei, 1965), p. ib; Helmut Shulte-Uffelage, trans, and ed., Das Keng-shen wai-sbih: Eim Quelle zur spdten Mongolenzeil, Ostasiatische Forschungen, Sonderreihe Monographien no. 2 (Berlin, 1963), p. 27. 250 YS, 4 0 , p. 856. 231 For example, to Dardess, "Khaishan's accession was the product of an as-yet unstabilized frontier," but his effective integration of Mongolia into an imperial system whose controlling authority lay in China and not in Mongolia made it "impossible for his eldest son to follow in his footsteps." See Dardess, Conquerors and Confucians, p. 30. 252 Hsiao Kung-ch'in, "Lun Yuan tai huang wei chi ch'eng wen t'i," p. 33. Cambridge Histories Online © Cambridge University Press, 2008 THE REIGN OF TUGH TEMtJR 547 refugee and his lack of political and military support rather than to the declining importance of the steppe region in Yuan imperial politics. Explaining the forces that enabled the restoration of Khaishan's imperial line in 1328—9, it has been claimed that the motivation behind the coup d'etat of 1328 was not based on any single issue but, rather, from a tangle of personal and group interests that linked the restorationist camp together. 2 " These included the desire to restore the imperial line of Khaishan, the attempts of the se-mu — particularly the Turkish elements in the bureaucracy — to seize power, and the attempts of the Confucian scholar-officials to expand their influence within the bureaucracy. The importance of the restoration camp's racial and ideological bonds should not, however, be overemphasized. Feng Ch'eng-chiin (1885-1946) advanced the theory that the restoration was a struggle of the Turkish Qipchaqs and Asuds against the incumbent Muslims and Mongols.254 But although the leaders of the loyalist and restorationist camps were Muslims and Turks, respectively, on both sides Mongolian princes and officials played as important a role as did their Muslim and Turkish colleagues. The relationship between the Confucian scholar-officials and the restoration camp also seems tenuous. 2 " The leaders of the group certainly were not themselves Confucian in orientation. Nor did any prominent Confucian scholar-official, with the exception of the Onggiid Chao Shih-yen, actively participate in the restoration. Neither race nor ideology, therefore, was an important factor in the restoration. Rather, it was the idea of restoring Khaishan's imperial line that brought together a group of princes and bureaucrats who had been estranged from Yesiin Temiir's rule, and this idea was used effectively as a rallying cry to advance their partisan interests and personal ambitions. The joint dictatorship of El Temiir and Bayan The four-year reign of Tugh Temiir was dominated by the Merkid Bayan and especially by El Temiir. As the persons who had been chiefly responsible for making the restoration possible, El Temiir and Bayan acquired a measure of power and honor that had never before been attained by any official in Mongolian imperial history. El Temiir was given the titles of prince of T'aiping; darkhan, a prestigious Mongolian title; and grand preceptor, the highest Chinese honorific title. 2 ' 6 He was the grand councillor of the right and 253 254 255 256 Dardess, Conquerors and Confucians, p . 4 2 . Feng Ch'eng-chun, "Yuan tai te chi ko Nan-chia-t'ai," p . 206. Dardess, Conquerors and Confucians, p . 32. YS, 138, pp. 3 3 3 2 - 3 ; Dardess, Conquerors and Confucians, pp. 46—50. Cambridge Histories Online © Cambridge University Press, 2008 548 MID-YOAN POLITICS the only councillor for most of the reign, and he also held successively the posts of manager of the Bureau of Military Affairs, censor in chief, chief administrator of the heir apparent's establishment (Kung-hsiang fu), and grand academician of the Pavilion of the Star of Literature (K'uei-chang ko ta hsiieh-shih). As such, El Temiir's power covered all the political, military, censorial, and cultural activities of the government. To safeguard his power, in 1329 El Temiir established the Chief Military Commission (Ta tu-tu fu), which enabled him to control directly six of the imperial guards, including three composed of fellow Qipchaqs. The imperial favor was also expressed in familial terms. El Temiir was allowed to take as his wives one of the late Yesiin Temiir khaghan's consorts as well as forty women of the imperial clan. The power that El Temiir enjoyed can also be seen in the status of his relatives. His uncle, Bukha Temiir; Sadun (Sa-tun) and Darindari (Ta-lin-ta-li), his younger brothers; and *Tangkish (T'ang-chissu, d. 1335), Sadun's son, all were appointed to important posts, and three of El Temiir's four sisters were married into the imperial house. The Merkid Bayan's power and honor were second only to that of El Temiir.2'7 As the only other person who was allowed to hold three or more concurrent posts, Bayan was successively given the honorary titles of defender in chief (t'ai-wei), grand guardian (t'ai-pao), and grand mentor (t'ai-fu), and he served as censor in chief, grand councillor of the left, and for more of the time, as manager of the Bureau of Military Affairs, among many other concurrent assignments. Like El Temiir, Bayan was given direct command over two elite military units, the Chung-i Guard and the Hsiian-i Brigade (Wan-hu), and he was invested as the prince of Chung-ning. To cap all these honors, he was given a great granddaughter of Khubilai as his consort. Again, as in El Temiir's case, Bayan's relatives shared his power. Despite his great power, Bayan seems to have been content with playing second fiddle to El Temiir, and a modus vivendi was apparently maintained between the two, perhaps on the basis of their past comradeship during the steppe war under Khaishan and in the recent restoration. Together, they undoubtedly represented a new type of powerful minister in MongolianYuan imperial history, quite different from Bayan of the Barin, Us Temiir, Harghasun, or Temiider in the past. Although Bayan of the Barin, Us Temur, and Harghasun each played an important role in placing their khaghans on the throne, they remained obedient and loyal once their khaghans had been enthroned, as a Mongolian noyan was supposed to be toward his khaghan. Temiider was powerful and scheming, but he depended on the support of Empress Dowager Targi and had no independent 257 YS, 138, p. 3337. Cambridge Histories Online © Cambridge University Press, 2008 THE REIGN OF TUGH TEMl)R 549 power base of his own. By contrast, El Temiir and Bayan were not only kingmakers in the true sense of the word, but they also built their own power bases in the bureaucracy and the military. Throughout his short reign, they overshadowed Tugh Temiir, whose role was mainly restricted to that of nominal emperor, symbol of legitimacy, and dispenser of titles and honors. Political development and fiscal policies Coming to power through a coup d'etat in the name of restoring Khaishan's imperial line, Tugh Temiir's administration carried out a bloody purge against its enemies and made some moves to justify the name of the restoration. The purge against the supporters of Yesiin Temiir's heir was carried out thoroughly and mercilessly after the surrender of Shang-tu in November 1328. Not only were the leading loyalists killed or exiled, but their properties also were confiscated.2'8 This vengeful spirit was so pervasive in the court that there was even a suggestion to kill all the officials who had followed Yesiin Temiir on his annual trip to Shang-tu. 2 ' 9 In addition, to make Yesiin Temiir's line illegitimate, not only was the khaghan denied a posthumous temple name, but the chamber in the Imperial Shrine where the tablet of Kammala, Yesiin Temiir's father, had been placed also was destroyed. Furthermore, the purge was extended to Khoshila's followers. Khoshila's three senior supporters who had survived their lord's murder were either executed or dismissed from office in 1330 on one ground or another.260 To justify the restoration, *Baidasha (Po-ta-sha, d. 1332), the commissioner of palace provisions in Khaishan's administration, and Yeh-erh-chi-ni, Khaishan's manager of the Bureau of Military Affairs, were made honorific grand mentor and defender in chief, respectively.2&l Moreover, the honors of San-pao nu and Toghto, the senior minister in Khaishan's Department of State Affairs who had been executed and vilified by Ayurbarwada khaghan, were restored and their confiscated properties returned to their families. Apart from the old followers of Khaishan, the participants in the restoration were given most of the positions of importance in the administration. Significantly, because Yesiin Temur's administration had been marked by the dominance of Muslims, not one single Muslim held a position in Tugh Temur's central government, and only a few held posts in the provincial 258 259 260 261 YS, 32, pp. 716-24. YS, 175, p. 4096. YS, 34, pp. 759, 761, 766. YS, n o , p. 3792; 124, p. 3058; T'u, Meng-wu-erh ihihchi, 156, pp. nb-i2a. Cambridge Histories Online © Cambridge University Press, 2008 55° MID-YOAN POLITICS administrations, a serious blow to the Muslims and one from which they were never able to recover under the Yiian.262 Equally importantly, the Chinese Confucian scholar-officials and their sinicized se-mu colleagues were also unable to exert much influence on the administration. Of the few Chinese se-mu scholar-officials who served as managers of governmental affairs, the highest possible position that could be attained by a Chinese, Ching Yen and Wang I held the post only for one and eight months, respectively.263 The Onggiid Chao Shih-yen served in that capacity for one year, but Chao had by this time become senile and tended to ingratiate himself with El Temiir. He was retired from the Secretariat in 1330 after being impeached by the Censorate.264 Because Tugh Temiir's accession was so transparently illegitimate, it was more important for his regime than for any previous reign to rely on liberal enfeoffments and generous awards to rally support from the nobility and officialdom. During his four-year reign, twenty-four princely titles were handed out, nine of which were of the first rank. Of these nine first-rank princes, seven were not even Khubilai's descendants.26' Not only were the imperial grants restored in 1329, but all the properties confiscated from the Shang-tu loyalists also were given to princes and officials who had made contributions to the restoration; in all, 125 individual properties are estimated to have changed hands.266 Action was also taken to win recognition from the other Mongolian khanates. In 1329, apparently after the murder of Khoshila, Naimantai (Naiman-t'ai, d. 1348), a descendant of Mukhali, was sent to the court of Eljigidei, the Chaghadai khan and a chief supporter of Khoshila's candidacy, taking with him as a gift an old seal originally given by Ogodei khaghan to Chaghadai one century earlier, apparently to mollify the Changhadai khan's anger over Khoshila's murder.267 A more general diplomatic offensive was made in the following year when three princes were sent as envoys to the Chaghadai, Qipchaq, and Ilkhanate. The three western khanates responded favorably to this gesture. In the remaining three years of the reign, the Qipchaq sent two tribute missions to Tugh Temiir's court; the Chaghadai sent four; and the Ilkhanate sent as many as eight.268 Thus Tugh Temiir was able to reestablish suzerainty over the Mongolian world for himself and to 262 263 264 265 266 267 268 Yang, "Yuan tai Hui-hui jen te cheng chih ti wei," pp. 266—9, 2 8 l . YS, 175, p. 4069; T'u, Meng-wu-trh shib cbi, 157, pp. 29a—b. YS, 34, p. 762. Noguchi Shuichi, "Gendai kohanki no 6g6 juyo ni tsuite," pp. 6 5 - 7 . YS, 32, pp. 716—24; Dardess, Conquerors and Confucians, p. 51. YS, 139. P- 3352Saguchi Tom, "Jushi seiki ni okeru Gencho daikan to seiho san-oke to no rentaisei ni tsuite," pp. 174-8. Cambridge Histories Online © Cambridge University Press, 2008 THE REIGN OF TUGH TEMOR 551 maintain a close relationship with the three western khanates. It therefore is difficult to regard the restoration of 1328 as a turning point in the relations of the Yuan with the broader Mongolian world. 269 Despite these efforts, discontent with Tugh Temiir's illegitimacy lingered among the nobility and bureaucracy and constituted an important source of political instability. There were at least eight plots against the imperial court during his reign, involving several princes and many of the highest-ranking officials in the government. 270 Of these plots, only that in 1330 led by Koncheg-beg (K'uan-ch'e po), the manager of the Bureau of Military Affairs, is known to have been made in the name of Khoshila's heirs. 271 The motivations for the other conspirators remain uncertain. But these plots clearly reflect the discontent among the ruling class with Tugh Temiir's regime. Political stability was further undermined by the increasing incidence of natural calamities and by uprisings of ethnic minorities. The provinces of Shan-hsi, Chung-shu, Ho-nan, Hu-kuang, and Chiang-che all were hit seriously and frequently by either droughts or floods, and millions of people lost their homes. 272 Such widespread calamities were immensely costly and also politically damaging. The frequent uprisings of ethnic minorities in southern China and especially in the southwest also concerned the government. Before Yesiin Temiir's reign, China had been relatively free from popular rebellions since Khubilai's successful suppression of numerous uprisings in the former Sung territories in the immediate postconquest era. Yuan control, however, began to break down in those regions inhabited by ethnic minorities. According to a study of these popular uprisings, between 1295 and 1332 only 24 took place in China proper, whereas 131 occurred in regions inhabited by ethnic minorities. 2 " Of these 131 uprisings, 65 broke out during Yesiin Temiir's reign and 21 in Tugh Temiir's. Of the 86 rebellions in these two reigns, no fewer than 50 were in Hu-kuang and 28 in Yiinnan. Some of them were quite large and required the government to use enormous resources to suppress them. The revolt of so many indigenous groups was due not simply to the misgovernment of these two reigns but also to the cumulative grievances of these groups against the Yiian government's exploitation and harsh control of them. The occurrence of so many revolts of ethnic minorities during these two reigns aggravated the financial difficulties of the Yiian government and also reflected the progressive weakening of its local control over these border 269 270 271 272 273 This is the contention of Dardess, Conquerors and Confucians, pp. 7—8. YS, 32, p. 740; 34, pp. 759, 761; 35, pp. 776. 778, 789, 793; 36. P- 803. YS, 38, p. 816. Li, Yuan shih hsin chiang, vol. 3, p. 543. Ch'en Shih-sung, "Shih lun Yiian tai chung ch'i te shao shu min tsu ch'i i," in Yiian shih lun chi, ed. Nan-ching ta hsiieh li shih hsi Yiian shih yen chiu shih (Peking, 1984), pp. 565—82. Cambridge Histories Online © Cambridge University Press, 2008 55 2 MID-YOAN POLITICS regions, foreshadowing the outbreak of large-scale rebellions in China proper in the following reign of Toghon Temiir. Although the government wasfinanciallyas hard pressed as it had been in the previous reigns, there was no return to Khaishan's "new deal." In addition to the traditional causes of financial strain, the added costs of the war against the loyalists, the suppression of the revolts by the ethnic minorities and the relief needed by victims of natural calamities heavily taxed the resources of Tugh Temiir's government. The war against Shang-tu in 1323 alone, as a censor pointed out, cost the government a sum several times greater than its annual income.274 The campaigns against the loyalists of Yiin-nan are known to have cost at least 630,000 ting of paper currency in 1330 alone.275 For relief of the refugees devastated by natural calamities, the government spent 1,349,600 ting of paper currency and 251,700 sbih of grain in the single year of 1329.276 Because of these and other expenses, the government is known to have had a budget deficit of 2,390,000 ting in 1330.277 The government did not attempt to boost its income by any drastic means. The only measure adopted to increase revenue was the continuation of Yesiin Temiir's policy of selling offices.278 The government did try to curtail its spending on such items as imperial grants, Buddhist sacrifices, and palace expenses. For example, the princes and officials who were entitled to grants of more than five ting of gold or silver for attending the enthronement were given only two-thirds of the rate originally fixed by Khaishan.2™ The number of Buddhist sacrifices performed annually in Shang-tu was also reduced in 1330 from 165 to 104.280 And in a general retrenchment of the imperial guards and palace services in the same year, the number of guardsmen, falconers, cooks, and other persons engaged in various services in the palaces was cut by more than 10,000.28' By these and other measures the government was able to keep its deficit within manageable figures. In the meantime the government had sufficient grain reserves at its disposal, as shipments from the Yangtze region to the capital had been pushed up to an all-time high, reaching 3,340,306 shih in 1329.282 With a manageable deficit and large quantities of grain in reserve, the government did not have to resort to printing paper currency to solve its financial problems. Partly because of 274 275 276 277 278 279 280 281 282 YS, 3 1 , p. 700. Li, Yuan shih fain chiang, vol. 3, p. 527. YS, 34, p. 755. YS, 184, p. 4238. YS, 96, pp. 2476-7; 139, p. 3352. VS, 33, p. 739YS, 34, p. 763. YS, 34, p. 765. YS, 9 3 , p. 2369; Schurmann, Economic structure of the Yuan dynasty, p. 125. Cambridge Histories Online © Cambridge University Press, 2008 THE REIGN OF TUGH TEMOR 553 this, general commodity prices were kept at the level attained in Ayurbarwada's reign. 28 ' Patronage of Chinese arts and letters With his actual power greatly circumscribed by El Temiir and Bayan, Tugh Temiir khaghan devoted much of his time and energy to creating a sinicized, Confucian image for his court. His actions in this regard were partly necessitated by the need to improve his own prestige and legitimacy in the eyes of his Chinese subjects, but they partly reflected his personal interests. 284 Probably the most erudite and versatile of all the Yuan khaghans, Tugh Temiir showed a wide range of scholarly and artistic interests quite early in his life and had surrounded himself with many distinguished Chinese literati and artists when he served as the prince of Huai in Chien-k'ang in 1 3 2 5 - 8 . j 8 ' He is known to have had a good knowledge of the Chinese language and history and also to have been a creditable poet, calligrapher, and painter. Also a collector and connoisseur of Chinese painting and calligraphy, Tugh Temiir has been compared with the two versatile emperors Sung Hui-tsung (r. 1101-25) and Chin Chang-tsung (r. 1189-1208). It has even been suggested that he may have taken Hui-tsung as his model. Posing as a cultivated Chinese sovereign, Tugh Temiir adopted many measures honoring Confucianism and promoting Chinese cultural values. For example, the khaghan sent an official to Ch'ii-fu in 1329 to offer sacrifices to Confucius on his own behalf,286 and in the following year he awarded laudatory titles to several past Confucian sages and masters.287 The khaghan himself performed the suburban offerings (cbiao-ssu) to Heaven in 1330, thereby becoming the first Yuan monarch to perform in person this important traditional Chinese state observance.288 To promote Confucian morality, the court each year honored many men and women who were known for their filial piety and chastity.28* To prevent the Chinese from following Mongolian and hence un-Confucian 283 Maeda, "Gencho jidai ni okeru shihei no kachi hendo," pp. 139—40. 284 John D. Langlois, Jr., "Yu Chi and his Mongol sovereign: The scholar as apologise, "Journal of Asian Studio, 38(1978), pp. 99—116. 285 On Tugh Temiir's understanding of Chinese culture, see Kanda Kiichiro, "Gen no Bunso no furyo ni rsuite," in Haneda Hakushi shoju kinen Tdyoshi ronso, ed. Haneda Hakushi kanrcki kinenkai (Kyoto, '95°). PP- 4 7 7 - 8 8 ; Yoshikawa Kojiro, "Gen no shotei no bungaku," pp. 2 4 3 - 7 6 ; Lo Hsien-yu, "Yuan-ch'ao chu ti Han hua shu i," pp. 72—3. 286 YS, 37, p. 730; 172, p. 4028. 287 YS, 34, pp. 763, 770. 288 YS, 34, p. 768; 72, pp. 1 7 9 1 - 2 . See also Franke, From tribal chieftain to universal emperor and god, pp. 32-3289 YS, 3 4 - 6 , passim. Cambridge Histories Online © Cambridge University Press, 2008 554 MID-YOAN POLITICS 'customs, the government decreed in 1330 that men who took their widowed stepmothers or sister-in-law as wives, in violation of their own community's customs, would be punished. In the following year, a law was promulgated that explicitly stated that northern and southern Chinese were forbidden to make such marriages.290 In the meantime, to encourage the Mongols and the se-mu to follow the Chinese customs, the officials of these two ethnic groups were allowed in 1329 to observe the Chinese custom of three years of mourning for deceased parents. This was the reversal of a decision made by Yesiin Temiir khaghan in the previous year to strip Mongolian and se-mu officials of their jobs if they followed the Chinese custom by taking leave for mourning. This reversal also paved the way for the imposition, by Toghon Temiir's court five years later, of the Chinese mourning custom on all the Mongols and the se-mu.2S>I The most concrete of the khaghan's efforts to patronize Chinese learning was his founding of the Academy of the Pavilion of the Star of Literature (K'uei-chang-ko hsiieh-shih yiian).292 First established in the spring of 1329, this institution was designed to undertake "a number of tasks relating to the transmission of Confucian high culture to the Mongolian imperial establishment. "393 These tasks included the elucidation of the Confucian classics and Chinese history to the khaghan; the education of the scions of high-ranking notables and the younger members of the kesig; the collection, collation, and compilation of books; and the appraisal and classification of the paintings and calligraphic works in the imperial collection. Of the 113 officials successively serving in the academy, there were many distinguished Chinese literati, including such scholar-writers as Yii Chi (1271 —1348), Hsu Yu-jen (1287— 1364), Chieh Hsi-ssu (1274-1344), Sung Pen (1281-1334), Ou-yang Hsiian (1283-1357), Su T'ien-chiieh (1294-1352), and K'o Chiu-ssu.294 In addition, there were the best Mongolian and se-mu scholars of Chinese learning of the time: Khutulugh Tormish, a great Uighur translator; Chao Shihyen, an Onggiid scholar-statesman; A-yung (d. 1335), a Kereyid poet; Naonao (1295-1345), a great Qangli calligrapher; Temiir Tash (T'ieh-mu-erh 290 YS, 34, p. 766; 103, pp. 2643-4; see also Henry Sertuys, "Remains of Mongol customs in China during the early Ming," Monumenta Serica, 16 (1957), pp. 137—90, esp. pp. 174-6. 291 YS, 30, p. 686; 38, p. 823; 83, p. 2086. On the influence of Chinese mourning customs on the semu, see Ch'en Yiian, Western and Central Asians in China under the Mongols: Their transformation into Chinese, trans. Ch'ien Hsing-hai and L. Carrington Goodrich, Monumenta Serica Monograph no. 15 (Los Angeles, 1966), pp. 241-52. 292 On the Academy of the Pavilion of the Star of Literature, see the following studies: Chiang 1-han, Yuan tai Kuei-chang ko chi Kuei-cbang jen urn (Taipei, 1981); Fu Shen, Yuan tai buang sbib sbu bua shou ts'angshih liieh (Taipei, 1981); Langlois, "Yii Chi and his Mongo.1 sovereign," pp. 106-8. 293 Langlois, "Yu Chi and his Mongol sovereign," p. 108. 294 On the personnel of the academy, see Chiang I-han, Yiian tai Kuei-cbang ko chi Kuei-chang jen urn, pp. 77-178. Cambridge Histories Online © Cambridge University Press, 2008 THE REIGN OF TUGH TEMOR 555 Ta-shih, 1302—47), a Qangli Neo-Confucian; and Tai Bukha (T'ai Pu-hua, 1304-52), a young Baya'ud chin-shih who was to become one of the most versatile of the Mongolian literati. This list undoubtedly represented the largest concentration of scholarly and artistic talents from all ethnic groups to serve together in a single governmental body in the Yuan dynasty. Concentrating so many talents in one governmental organ to perform various literary, artistic, and educational activities was unprecedented not only in the Yuan dynasty but also in Chinese history. The nearest comparison was perhaps the Academy of Scholarly Worthies under T'ang Hsiian-tsung. The academy's activities had considerable political significance. They were a part of the campaign to improve the khaghan's image by applying "a carefully prepared veneer of'civilization' to his court." 295 The cultivation of the arts and literature, nevertheless, was also an official extension of the khaghan's personal interests. He spent much of his leisure time, of which he had plenty, in the academy, practicing calligraphy and viewing its art collection. Yii Chi, the mastermind behind the academy, and K'o Chiu-ssu, the painter, enjoyed his particular favor. Yii Chi employed his substantial talents in the khaghan's cause by drafting edicts justifying the khaghan's legitimacy and by flattering the khaghan in various essays and poems, and K'o Chiu-ssu pleased the khaghan by authenticating and appreciating the art collection with him. 2 ' 6 The academy was responsible for compiling and publishing a number of books.297 But its most important achievement and a hallmark of Tugh Temiir's reign was its compilation of the Ching shih ta tien (Grand canon for governing the world), a vast institutional compendium. 298 The purpose of bringing together and systematizing all important Yuan official documents and laws in this work according to the pattern of the Hui-yao (Comprehensive essentials of institutions) of the T'ang and Sung dynasties was apparently to demonstrate that Yuan rule was as perfect as that of early Chinese dynasties. This ambitious project was placed under El Temiir's overall supervision, but it was Yii Chi who assumed the main responsibility. Started in May 1330, the project was completed in thirteen months. The end product was a work of eight hundred chiian grouped in ten sections. The first four sections dealt with matters concerning the khaghans and the imperial house. For the 295 Langlois, "Yii Chi and his Mongol sovereign," p. 106. 296 On K'o Chiu-ssu's role, see F. W. Cleaves, "The 'Fifteen "Palace poems" ' by K'o Chiu-ssu," Harvard Journal of Asiatic Studies, 20(1957), pp. 391-479. 297 See Chiang I-han, Yuan tai Kuei-chang ko chi Kuei-chang jen wu, p. 75. 298 The most comprehensive study of the Ching shih ta tien is Su Chen-shen's Yuan chtng shu "Ching shih ta tien" chih yen chiu (Taipei, 1984). See also Herbert Franke, Geld und Wirtschaft in China unter der Mongolenherrschaft, pp. 25—31; Franz Schurmann, Economic structure of the Yiian dynasty, pp. ix—xiv; Hsiao Ch'i-ch'ing, The military establishment of the Yuan dynasty, pp. 6 7 - 9 ; Langlois, "Yii Chi and his Mongol sovereign," pp. 108—10. Cambridge Histories Online © Cambridge University Press, 2008 556 MID-YOAN POLITICS compilation of these sections, many Mongolian documents that had hitherto been inaccessible to the Chinese were translated into Chinese. The last six sections, dealing with governmental matters, were arranged in the manner of the Confucian classic the Cbou li (Rites of Chou), and the Hut yao. As the repository of official Yuan documents, the Ching shih ta tien was the pride of the khaghan and also of great benefit to later historians. It provided the basis for the various treatises (cbib) of the Yuan shih, which was compiled at the beginning of the Ming dynasty. Though the compendium itself was lost sometime between 1509 and 1605, many parts survived as a consequence of being incorporated into the Ming encyclopedia Yung-lo ta tien (Great canon of the Yung-lo era).2^9 These surviving parts still constitute important sources for the institutional history of the Yiian dynasty. Despite its remarkable achievements, the academy cannot be considered to have been successful in making the government Confucian. Because the bureaucracy was dominated by El Temiir and the Merkid Bayan, who did not share the khaghan's sinophile proclivities, the academy's influence was effectively contained within the palace grounds. Perhaps because of the Confucian implications of the academy and the excessive imperial attention it enjoyed, its officials were repeatedly impeached by censors on a variety of grounds. Compelled by circumstances, therefore, its chief officials offered their joint resignation in 1330.3°° Yii Chi, for one, believed that his work in the academy had no impact on the government's policies.'01 Finally, El Temiir seized control of the academy in early 1332, just six months before the khaghan's death. El Temiir's taking control of the academy was in part intended to restrict its members' independent access to the khaghan. *03 After Tugh Temiir's death, the academy'sfloweringsoon came to an end. Although the academy did help the khaghan add a facade of Confucianism to his court, its impact on the government as a whole, however, was limited. THE FAILURE OF THE SUCCESSION ARRANGEMENTS Throughout his brief reign, Tugh Temiir seems to have been obsessed by the problem of his own illegitimacy as emperor and by the question of his own successor. Both he and his principal consort, Empress Budashiri (Puta-shi-li, d. ca. 1340), originally intended to pass the throne to Aratnadara (A-la-t'e-ta-la, d. 1331), their eldest son. Anatnadara was invested in 299 On the dissemination of the Ching shih ta tien and its surviving texts, see Su, Yuan chetig shu "Ching shih ta lien" chih yen chiu, pp. 13-18, 33-78. 300 Chiang I-han, Yiian tai Kuei-cbang ko chi Kuei-chang Jen wu, p. 20. 301 YS, 181, p. 4178. 302 Dardess, Conquerors and Confucians, p. 48. Cambridge Histories Online © Cambridge University Press, 2008 THE PERIOD IN RETROSPECT 557 March 1330 as the prince of Yen, a prestigious title previously held only by Chen-chin, and Aratnadara was then designated as heir apparent in January I 33 I - 3 ° 3 By this time, preparatory measures to ensure Aratnadara's smooth succession had already been taken: the murder of Empress Babusha (Pa-pusha), Khoshila's principal wife, and the exile to Korea in May 1330 of Toghon Temiir.5°4 These proved unnecessary, however, as Aratnadara died about one month after his designation as heir.*°' This, the premature death of his son, completely upset Tugh Temiir's plan for succession. Moreover, he also seems to have begun to be haunted by the fear of retribution for what he had done to his own elder brother. He therefore asked El Temiir to take care of *Gunadara (Ku-na-ta-la), his second son, and changed the latter's name to El Tegus (Yen T'ieh-ku-ssu), meaning "perfect harmony."' 06 Thus when the khaghan died on 2 September 1332, at the age of twenty-eight, the issue of succession had not yet been decided. It is said that on his deathbed the khaghan expressed remorse for what he had done to his elder brother and his intention to pass the throne to Toghon Temur, Khoshila's eldest son, instead of El Tegus, his own son.*0? Fully aware of his own role in Khoshila's death, El Temiir was quick to realize the harm that the enthronement of any of Khoshila's sons could do to himself, and he consequently proposed to place El Tegus on the throne.3°8 When Empress Dowager Budashiri, who perhaps shared her husband's fear of retribution, rejected this plan, Khoshila's second son, the six-year-old Irinjibal (Emperor Ning-tsung), was chosen as a compromise candidate, and he was enthroned in Ta-tu on 13 October 1332. Again, the child khaghan died unexpectedly after only fifty-three days on the throne. 309 The death of Irinjibal marked the end of the mid-Yiian period and thus paved the way for the thirty-five-year reign of Toghon Temiir, his elder brother. THE PERIOD IN RETROSPECT This chapter shows that the mid-Yiian rulers missed the opportunity afforded by the general peace in the country to make further constructive changes on the basis of what Khubilai had left them. Throughout most of this period, the country was free from foreign wars, campaigns of conquest, and popular 303 YS, 34, pp. 754—70. For a translation of the Empress Budashiri's biography in the YS, see F. C. Cleaves, "The Sino-Mongolian inscription of 1335 in memory of Chang Ying-jui," Harvard Journal of Asiatic Studies, 13 (1950), pp. 1—131, esp. pp. 35—6, n. 35. 304 YS, 34, p. 756; 38, p. 815; 114, p. 2877. 305 YS, 34, p. 774. 306 YS, 35, p. 790; 36, p. 802. 307 Ch'iian Heng, Keng shen wai shih, pp. la-b; Schulte-Uffelage, Das Kmg-shen wai-shih, pp. 27-8. 308 YS, 38, p. 815. 309 On Irinjibal, see YS, 37. Cambridge Histories Online © Cambridge University Press, 2008 558 MID-YOAN POLITICS revolts, as the khaghans had inherited a powerful and peaceful empire from Khubilai and they had abandoned the expansionist, imperialist policy pursued by their ancestors. Moreover, they succeeded in restoring peace in the Mongolian world and maintaining their suzerainty over other Mongolian khanates. They were unable to consolidate their gains, however, mainly because of the chronic and treacherous conflicts within the ruling class itself. Bitter struggles to control the throne made the politics of the period both brutal and volatile. During these 39 years, nine khaghans ascended the throne, reigning on average for only 4.3 years. Six of these nine became khaghan only after heated disputes or even armed struggles. Two of the nine were killed, and another was reported missing after being overthrown.'10 Not only did the khaghans fall victim to these recurrent struggles; there also were frequent massive turnovers of personnel at the highest level of the government, as bloody purges always followed in the wake of a succession struggle. Making the politics even more volatile, these disputes often involved rival candidates, with bases in China and the steppe. Though it was clear that the steppe candidates could not succeed in gaining the throne without the support of leading bureaucratic factions in China, the alternations between the khaghans of sharply different backgrounds inevitably resulted in drastic changes in governmental policies and even affected its cultural orientation. The aggravation of succession disputes in the post-Khubilai era can thus be attributed to the contradictions between the Mongolian concept of the empire as the joint patrimony of all descendants of Chinggis khan and the Chinese concept of autocratic rule, to the continuing tension between the steppe and the sedentary parts of the empire, and to the rise of powerful ministers and bureaucratic factions. This rise of powerful and dominant ministers in the mid-Yuan was unprecedented in Mongolian-Yuan history. It was the result of Khubilai's policy of centralization and bureaucratization, which, though far from being complete, enabled the bureaucrats to acquire more power at the expense of the imperial princes, who had shared power to some degree with the khaghan. The power of the bureaucrats was further enhanced at the expense of the throne as the succession disputes worsened, as bureaucratic support now became more important than the backing of the imperial princes for a candidate hoping to win the throne. As a result, not only was more power concentrated in the hands of the leaders of bureaucratic factions, but the relationship of bureaucratic strong men to the monarch also was changed. Bayan of the Barin, Us Temiir, and Harghasun all played important roles 310 The average life span of the six khaghans who were not killed or overthrown was only 29.3 years; this sharply contrasts with the life spans of the first five Mongolian khaghans, which averaged 58.2 years, and may reflect the biological as well as the political degeneration of the Mongolian royal house. Cambridge Histories Online © Cambridge University Press, 2008 THE PERIOD IN RETROSPECT 559 in putting their candidates on the throne, but they still belonged to the old school of Mongolian noyad and remained absolutely loyal and obedient to the throne once the new khaghan was in place. Temiider exemplified the transitional type. Although still lacking a solid independent power base and utterly dependent on Empress Dowager Targi's support for his political survival, he was able to defy the authority of both Ayurbarwada and Shidebala and, early in Shidebala's reign, to establish his own reign of terror against his detractors. El Temiir and the Merkid Bayan represented the ultimate power of bureaucratic strong men. They were "kingmakers" in the true sense of the word, and they also kept a tight grip on the governmental activities in all spheres. With the rise of such men, the throne as the focus of all loyalty and the source of all authority was therefore weakened. Factional conflict was the corollary of the succession crises and the rise of bureaucratic strongmen. Factions were formed among the bureaucrats, usually in alliance with groups of imperial princes, to support or fight against a powerful minister and to control the throne. Apart from power, factional struggles also sometimes had ideological implications. They often were fought to control the direction in which the state was moving, whether it should be further sinicized and centralized or whether its Mongolian character and the patrimonial-feudal privileges of the Mongolian and se-mu elites should be retained. The bitterness of factionalism and its ideological implications can be seen most clearly in the prolonged and bitter conflict during Ayurbarwada's and Shidebala's reigns between Temiider's faction and their Confucian opponents. Temiider's faction stood for the status quo and strongly resisted those of the khaghan's policies that were supported by their Confucian enemies. This conflict paralyzed the government in the later part of Ayurbarwada's reign and led to Shidebala's assassination. In the reigns of Khaishan and Yesiin Temiir, the old Mongolian school apparently held the upper hand over the Confucian scholar-officials, whose voices were muted. Similarly, in Tugh Temiir's reign, El Temiir and the Merkid Bayan dominated the court. Even though imperial princes and high-ranking bureaucrats repeatedly plotted against them or the khaghan, such conspiracies were easily suppressed. Although the khaghan trusted the Confucian scholar-officials, they were unable to exert much influence on the government's policies, which were tightly controlled by El Temiir and the Merkid Bayan. The recurrence of succession crises and the rise of bureaucratic strongmen and factionalism added to the political turbulence and volatility of the period and weakened the Yuan government. But the government was also plagued by such problems as a swollen but increasingly inefficient bureaucracy, financial shortages, and inflation. The government was further enervated by the Cambridge Histories Online © Cambridge University Press, 2008 560 MID-YOAN POLITICS debilitation of its armed forces. Although these problems are usually associated with dynastic decline, the fact that the Yuan was a dynasty of conquest and the special political situation of the mid-Yuan undoubtedly made them worse. The mid-Yuan government did make some improvements. It tried to incorporate the Mongolian state more closely into China by further reforming it along Chinese lines, especially during the reigns of Ayurbarwada, Shidebala, and Toghon Temiir. Ayurbarwada enhanced the importance of Confucian scholar-officials in the government, reinstituted the civil service examination system in order to make the bureaucracy more Confucian, compiled a new code of laws to rationalize the administration and facilitate the dispensation of justice, and sponsored the Mongolian translation of Chinese books in order to acquaint the Mongolian and se-mu elites with the Chinese historical experience and philosophical wisdom. To centralize the administration, the government attempted to curtail the aristocrats' privileges. Shidebala continued his father's reforms or expanded them. Tugh Temiir made even greater efforts to promote Confucian values and patronize Chinese arts and letters. Even during the reigns of Khaishan and Yesiin Temiir - the two khaghans with the strongest steppe backgrounds — there was no complete reversion to the Mongolian steppe tradition. All of these changes did not, however, add up to a fundamental transformation of the Yiian state. Even the innovating monarchs did not change the Yuan state much, for fear that a radical departure from their forefathers' pattern of rule would provoke strong conservative opposition. Ayurbarwada's attempts to curb the elite's privileges failed precisely because of conservative opposition. And the civil service examinations was never expanded to the point that it would seriously undermine the elite families. Other innovative measures meant nothing more than adding a sinicized and Confucian veneer to a sociopolitical structure that remained essentially unchanged. Thus the Yiian government remained partly patrimonial-feudal and partly bureaucratic in structure. Confucianism was still one of many "religions" that competed with one another for imperial attention. The vast political and social gaps between the Mongolian and se-mu elites and the Chinese subjects, although somewhat narrowed, remained unbridged. What was left by the mid-Yuan monarchs to Toghon Temiir, the last Yiian khaghan, was a state that had been greatly weakened by the constant and violent conflicts within the ruling class itself and a state whose roots in Chinese society, though deep, were still not deep enough. Cambridge Histories Online © Cambridge University Press, 2008 CHAPTER 7 SHUN-TI AND THE END OF YUAN RULE IN CHINA YOAN CHINA AT T H E ACCESSION OF T O G H O N TEMOR (SHUN-Tl) When Toghon Temiir, barely thirteen years of age, was brought to the summer capital, Shang-tu, and installed there as the tenth emperor of the Yuan dynasty in July 1333, the realm he nominally headed had long been under stress, owing in part to complex, endemic tensions among its ruling elites and in part to troubles of long standing in China itself. Although there was no sign of imminent collapse, it is still a little ironic that of all the Yuan emperors, it was his reign, the last in China, that turned out to be the longest. It is not so ironic that as passive a ruler as he turned out to be, the whole quality of political life during his reign came to constitute a powerful negative example to the builders of the next dynasty, the Ming. Toghon Temiir (often referred to by his temple name, Shun-ti) reigned in China, or at least Ta-tu, until 1368.' He died in Mongolia in 1370. The history of his reign raises the grand question, Why did the Yuan dynasty fall as and when it did? To that question there appears to be no simple answer. There was certainly no lack of energetic effort on the part of Mongols and Chinese alike to save it. The younger emperor inherited a system of government whose size, complexity, and costliness stemmed both from its need to provide offices and employment for a large minortiy population of privileged Mongols and foreigners and from the fact that although the state was mainly a public bureaucracy, it was in part also the private patrimony of the imperial house and of certain high-ranking elites. A short description of these special features, as they had evolved by 1333, will serve to give some background to the march of events during Toghon Temiir's troubled reign. There was not one imperial capital, but two. The principal capital was Ta-tu (on the site of modern Peking), and the summer capital was Shang-tu, some 1 Toghon Temiir was given the posthumous name Shun-ci by the Ming court. For a brief biography, see Herbert Franke, "Toghon Temiir," in the Dictionary of Ming biography, ed. L. Carrington Goodrich and Chaoying Fang (New York, 1976), pp. 1289-93. 561 Cambridge Histories Online © Cambridge University Press, 2008 562 SHUN-TI AND THE END OF YOAN RULE IN CHINA two hundred miles north in the steppes of Inner Mongolia. Toghon Temiir was enthroned in July 1333 in Shang-tu and returned to Ta-tu in September or October. Subsequently he faithfully removed his court to Shang-tu every summer, as a remnant of the nomadic life-style of his forebears. Each year, until the destruction of Shang-tu by Chinese rebels in January 1359 put an end to the custom, in the fourth lunar month Toghon Temiir moved north and in the eighth returned to Ta-tu. The trip north was time-consuming: In 1347 it took twenty-three days.2 Thus the emperor spent about a month and a half on the move each year, traveling at the leisurely pace of about ten miles per day. Each year he took with him a large retinue of officials, who worked from "branch" offices in Shang-tu during the summer months. These annual circuits were accomplished at a cost no one has yet tried to calculate, and they involved a great array of logistical support systems, transport and courier services, and a host of special traveling agencies. In the 1330s and 1340s, two southern Chinese literati-officials, Huang Chin and Hu Chu, wrote enthusiastic, poetic descriptions of these scenic journeys and of the summer capital. Hu Chu's, the earlier such description, elicited a great deal of favorable comment.3 The main capital, too, impressed the literati from south China, and it is to them we owe several late descriptions and accounts, ranging from the young Hu Chu's awed reactions to its magnificence and opulence set amid a large population of very poor people (it was an expensive place to live), to T'ao Tsung-i's detailed notes on it, and to Hsiao Hsiin's careful inventory, written in 1368, just before the new Ming government deliberately demolished the city.4 The outer walls measured 28.6 kilometers around. The walled imperial compound toward the city's southern edge, with its central court, palaces, and lake, took up approximately a twelfth of the entire intramural urban area (see Map 33). That the Yuan regime still encompassed a patrimonial dimension is evident in the existence of what amounted to a sort of semipublic, superficially bureaucratized business empire with holdings in such fields as farming, palace and temple construction, and manufacturing. Manufacturing took in everything from the procurement of raw materials to the shipment, storage, and distribution of an astounding range of items, mainly luxury goods. Some three hundred workshops and other enterprises, mostly in north China, drew on a labor pool of registered hereditary artisan households to turn out fabrics of all kinds, foodstuffs, beverages, jewelry, carriages, ironware, felt, tiles, 2 Huang Chin ( 1 2 7 7 - 1 3 5 7 ) , Huang Wtn-hsim cbi (TSCC ed.), 7, pp. 7 i b - 7 2 b . 3 Hu Chu (1276—ca. 1353), Ch'un-pai chat lei kao (TSCC ed.), 2, pp. 5a ff. 4 Hou Jen-chih and Chin T'ao, Pei-ching shih hua (Shanghai, 1980), pp. 61—95. Cambridge Histories Online © Cambridge University Press, 2008 Y O A N C H I N A AT T H E A C C E S S I O N O F T O G H O N T E M O R 563 leather goods, and much more.' Their output was consumed directly by the upper echelons of the conquest elite; some of it was also put on sale in the stores of the capital. The owners of these businesses were the members of the imperial family, their consorts, and the highest court officials; but the emperor could and did reassign these assets, and political overturn at the top ensured their periodic confiscation and redistribution. Counting patrimonial as well as civilian and military posts and their authorized support staff, the Yuan bureaucracy numbered some 33,000 at the beginning of Toghon Temiir's reign. It was a multinational body, as roughly 30 percent of all positions were occupied by non-Chinese. Statutory specifications reserved certain positions for members of one or another ethnic group, but these specifications were often quietly circumvented and on occasion were openly disregarded or revised. Most government offices regularly employed a mixture of people of different ethnic origins, and modes of accommodation had to be worked out in order that the work of government might proceed smoothly.6 Official procedure required uneasy compromises between the Chinese sense of hierarchy and personal authority and responsibility and the Mongol conciliar tradition. Functions of state were performed in at least four written languages: standard literary Chinese; a strangely colloquialized chancery Chinese; Mongolian; and to judge from the regular assignment of "Muslim" (Hui-hui) clerks to important government bodies, probably also Persian. 1 Translators and interpreters were regular employees of bureaucratic agencies. It all worked better than one might suppose. Deliberate efforts were constantly made to foster ethnic cohesion in officialdom in the only way that it could be done: by encouraging loyalities to specific institutions. For example, late in the Yuan, a large number of stone inscriptions (t'i ming chi) were carved and displayed to advertise the duties and the signal importance of various government agencies, high and low, and to list the names of all the current incumbents of their offices, thus to build a sense of common purpose among all the different kinds of people who held the positions. There still remains a complete history of one Yuan office, the Directorate of the Palace Library, which was compiled as late as the 1360s, and it shows how persistent a sense of institutional identity could be for the Chinese, Mongols, Muslims, and other ethnic elements that made up its large staff.8 Politically the most 5 Li Kan, Yuan tai she hui thing chi shih kao (Wuhan, 198;), pp. 229—43 (tables); and Oshima Ritsuko, "The chiang-hu in the Yuan," Acta Asiatica, 45 (1983), pp. 69—95. 6 Elizabeth Endicott-West, "Imperial governance in Yuan times," Harvard Journal of Asiatic Studies, 46, (1986), pp. 5 2 3 - 4 9 , touches on this problem. 7 Huang Shijian, "The Persian language in China during the Yuan dynasty," Papers on Far Eastern History, 34 (1986), pp. 8 3 - 9 5 . 8 Wang Shih-tien (fl. 14th c ) , ed., Pi sbu cbien cbib (SKCS ed.). Cambridge Histories Online © Cambridge University Press, 2008 564 SHUN-TI AND THE END OF YOAN RULE IN CHINA important manifestation of interethnic institutional cohesion in Toghon Temiir's reign was to be found in the Censorate, whose members insisted on their right of remonstrance and who therefore played a pivotal role in highlevel political processes right to the end of the Yuan dynasty. Foreign participation in government helped bloat its size. To take one example, the high court of justice, whose jurisdiction was limited to cases involving Mongols or aliens in the two capitals, employed a staff of ninetysix, headed by forty-two Mongolian judges (jarghuchi), each of whom held the very high civil service rank of iB and enjoyed the accompanying high salary and perquisites.9 The problem of overstaffing {Jung kuan) was voiced occasionally at court in Toghon Temiir's time, but for political reasons the Yuan ascendancy was never able to go too far for too long in cutting supernumerary positions, despite the shortages of revenue. There were several recruitment paths for the bureaucracy in the late Yuan. Most lower officials served an apprenticeship as clerks or Confucian schoolteachers. Young members of the conquest elite (and some Chinese) were first groomed and informally scrutinized as kesigden, hereditary guards or housemen to the monarch and his family. The descendants of three warriors who had been Chinggis Khan's closest comrades headed this corps, which came to number some thirteen thousand young men, and it has been characterized as a "cradle of officialdom" and the "citadel of the Yuan ruling class."10 The kesigden enjoyed what was known in Chinese as ken chiao, literally, "roots and feet," an invidious reference to their having an aristocratic leg up on everyone else. The triennial Confucian examination system, instituted in 1315, was statistically a minor port of entry into the lower bureaucracy (there was a total of only some 550 cbinshih degree holders by 1333, at most 2 percent of the officials), but as a matter of politics it was crucial. The system was heavily weighted in favor of Mongols (a small pool of candidates and an easy examination) and against southern Chinese (a huge pool of candidates and a demanding examination), yet despite these statutory inequalities, the system did create a common outlook and interethnic cohesion among the degree holders. The examination round of 13 3 3, which had begun in the spring, ended with thefinalpalace examination in the ninth lunar month of that year, shortly after the young emperor's return to Ta-tu from Shang-tu." This was the first examination under the Yuan to produce its full quota for all ethnic groups, 9 Sung Lien et al., eds., Yiian shib (Peking, 1976), 87, pp. 2 1 8 7 - 8 (hereafter cited as YS). 10 Hsiao Ch'i-ch'ing, The military establishment of the Yiian dynasty (Cambridge, Mass., 1978), pp. 39—44; see also his Yiian tai sbih hi in t'an (Taipei, 1983), pp. 141-230. 11 Sung Chiung ( 1 2 9 4 - 1 3 4 6 ) , Yen-sbib chi (SKCS ed.), 15, p. 13a. Cambridge Histories Online © Cambridge University Press, 2008 YOAN CHINA AT THE ACCESSION OF TOGHON TEMCR 565 clear evidence of growing interest in the examinations and the spread of higher education. The graduation list from that year is still extant." It shows, among other things, that the fifty Chinese were, as a group, a bit older than the fifty non-Chinese, with median ages (sui) of thirty-one and twenty-eight, respectively. It also shows that 92 percent of the Chinese were married, as against only 74 percent for the others. Interestingly, the Mongols and members of other minorities were heavily intermarried with Chinese: In all, 58 percent of them had Chinese mothers, and of those married, nearly 70 percent had Chinese wives. Everyone, irrespective of ethnic origin, tended to receive similar kinds of first appointments as local officials. The second name on the nonChinese list is Yii Ch'iieh, a Tangut of undistinguished ancestry from Honan Province and, in that sense, typical of the obscure, non-Chinese youth for whom the examinations were the likeliest road to fame and fortune. Yii Ch'iieh will be met later as an outstanding local official and reformer. He also became a talented poet and litterateur in Chinese; his works still survive.'3 The chin-shih class list of 1333 thus captures for a certain moment in time some of the intricate cleavages and blendings that characterized the Yuan ascendancy: institutionalized ethnic preference alongside widespread intermarriage; a conquest society well on the way to assimilating some of the moral values and literary and administrative traditions of its colonial inferiors; and a Chinese elite that had come actively and fully to participate in the foreign dynasty that ruled them. Yuan government in 1333 still weighed unevenly on the two principal regions of China, the north and the south. Especially striking was the maintenance of so dense a network of local administrative units in north China, a region that had borne the brunt of the wars of the Mongol conquest and had then suffered further depopulation by continual emigration to the south or to the cities, repeatedly visited as it was by floods, earthquakes, droughts, locust infestations, epidemics, and famine.14 Yang Wei-chen did not exaggerate when stating (in 1348) that the population of an entire northern county {hsien) was often smaller than that of a single southern Chinese lineage (tsu).1* South China, much richer and more heavily settled, was, as a matter of deliberate policy, more lightly administered. In proportion to its population, its local government officials were only one-fifth as numerous as those in the north, and its population paid nothing approaching the north's rates of 12 Fully annotated in Hsiao Ch'i-ch'ing, "Yiian-t'ung yuan nien chin shih lu chiao chu," Shih huo yueb k'an, 13 (1983), pp. 7 2 - 9 0 , 147-62. 13 Yii Ch'fieh (1303-58), Ch'ing-yang hsien ihtng wen chi (SPTK ed.). 14 Contemporary references to these conditions are quoted in Aritaka Iwao, "Gendai no nomin seikatsu ni tsuite,"inKuwabaraHakushikanrtkikinenToySshironso(Kyoto, 1934), pp. 979ff.;and WuHan, "Yuan ti Icuo chih peng k'uei yii Ming chih chien kuo," Ch'ing-hua hsueh pan, 11 (1936), pp. 359—423. 15 Yang Wei-chen (1296—1370), Tung-wei-tzu wen chi (SPTK ed.), 4, pp. 9b—10b. Cambridge Histories Online © Cambridge University Press, 2008 566 SHUN-TI AND THE END OF YOAN RULE IN CHINA taxation. By way of tacit compensation for this disparity, the northern Chinese were officially favored over the southern, especially in the matter of eligibility for bureaucratic positions. Finally, in 1333, the Yuan dynasty was funding itself from a wide variety of sources. Besides taxes on land and commerce, there were state cash investments in many commercial enterprises, and outright state control of farms and especially of saltworks. By 1333, income from the national salt monopoly had steadily risen to provide a maximum of some 7.6 million ingots (ting; each nominally equivalent to 50 ounces of silver) in paper cash yearly, enough to meet about 80 percent of central revenue needs. A sea transport system moved vast quantities of grain from the Yangtze delta up to Ta-tu, where it fed the whole conquest establishment, plus the large numbers of poor people and dependents of all kinds who inhabited the city, and provided grain for the Mongolian people living in the steppes to the north. These vital shipments peaked in 1329. Soon thereafter the amounts began to slide, at first gradually (there was a 25 percent shortfall by 1341) and then catastrophically. The Yuan dynasty had almost exhausted further possibilities for raising income and in fact would soon be caught between diminishing revenues and rising costs.'6 So much for a short tour d'horizon of late Yuan China at Toghon Temiir's accession, two decades before things began to fall apart. It is important to bear in mind how close all these events were to the age of Khubilai (r. 126094), even though eight emperors had come and gone in the meantime. Several senior officials in their sixties in 1333 had come of age and begun their careers in the time of the dynasty's founder. Khubilai was still very much a living memory, as Bayan's chancellorship would show. TOGHON TEMUR'S ENTHRONEMENT AND BAYAN'S CHANCELLORSHIP, 1 3 3 3 - 1 3 4 0 It is not absolutely certain who Toghon Temiir really was. In 1340, he officially and publicly asserted that he was indeed a legitimate descendant of Khubilai, in the sixth generation, the elder son of the assassinated emperor Khoshila (posthumous temple name, Ming-tsung; r. 1329) and a Qarluq (Turkish) consort. However, Tugh Temiir (posthumous temple name, Wentsung; r. 1328, 1329-32) had earlier issued an edict (drafted by Yii Chi, one of the most respected and influential of the Chinese literati of his time) that 16 The principal work on Yuan finance is still that by Herbert Frankc, Gild und Wirtscbaft in China timer der Mongoltnherrschaft: Beitrage zur Wirlshaftsgescbichte der Yiian-Zeit (Leipzig, 1949). See also Herbert F. Schurmann, Economic structure of the Yuan dynasty: Translation of chapters 9} and 94 of the Yuan shih (Cambridge, Mass., 1956; repr. 1967). Cambridge Histories Online © Cambridge University Press, 2008 T O G H O N T E M O R AND BAYAN'S C H A N C E L L O R S H I P 567 declared, apparently on the authority of the husband of ToghSn Temiir's wet nurse, that ToghSn Temiir was not really Khoshila's son.' 7 There was also a story, widely believed, that Toghon Temiir was actually the son of a Chinese father (a descendant of the Sung emperors) and a Muslim mother and that Khoshila had adopted him. l 8 Accordingly, Toghon Temiir had been exiled at the age of ten to an island off the coast of Korea and, at the age of twelve, removed to what is now Kuei-lin in Kwangsi Province, where he reportedly spent the year before his recall and enthronement taking lessons from a Buddhist monk in the Analects and the Hsiao ching (Book of filial piety) and making friends with a troop of monkeys, symbols of his birth year (1320). Upon Tugh Temiir (Wen-tsung)'s death in September 1332, ToghSn Temiir was bypassed in favor of his younger half-brother Irinjibal, a child of six, but that child died in December 1332, after reigning for only two months. There followed an interregnum for the next seven months, during which time the kingmakers of Ta-tu maneuvered on behalf of their favorite candidates. The politically dominant clique, that of El Temiir and his kinsmen, backed the candidacy of Tugh Temiir's young heir, El Tegiis. El Tegiis's mother, Budashiri, argued that El Tegiis was still too young and recommended that he instead be made the heir apparent to Toghon Temiir. She eventually got her way, in part by allowing El Temiir to marry one of his daughters to Toghon Temiir, but most of all because she won the support of Bayan, a Mongol of the Merkid tribe, who was on the verge of becoming the dominant political personality of his time. He was already a senior official with concurrent leading positions in the Bureau of Military Affairs (or Privy Council, Shu mi yuan), various praetorian guard units, and inner-palace agencies. As a result of his assistance in securing Toghon Temiir's succession, in 1333 Bayan finally obtained the highest civil appointment of all, the chancellorship of the right (grand councillor of the right, yu ch'eng hsiang). In 1335, he became chancellor {ch'eng hsiang) pure and simple, a position that he held from 1333 until his overthrow in 1340. Clearly, Toghon Temiir's intended role was that of a temporary figurehead. He was a thirteen-year-old child, untrained, still officially illegitimate, and easily disposable; indeed, he was advised to stay in the shadows and take no direct part in ruling the empire. He later stated, surely with some truth, that he had spent his early years as emperor in constant fear. It was a very shaky beginning for what turned out to be a long, if somewhat passive, reign. In the summer of 1335, Chancellor Bayan carried out a bloody coup 17 John D. Langlois, Jr., "Yii Chi and his Mongol sovereign: The scholar as apologist," Journal of Asian Studio, 38(1978), p. 111; YS, 181, p. 4180. 18 The case in favor of this story has been pressed by Wan Ssu-t'ung( 1638-1702) in Keng shen chiin i shih, repr. in vol. 4 of Shih liao u'ungpitn (Taipei, 1968). Cambridge Histories Online © Cambridge University Press, 2008 568 SHUN-TI AND THE END OF Y0AN RULE IN CHINA against the surviving members of the family of his former colleague El Temiir, all of whom he charged with treason. He then set in motion an extraordinary program of reaction, designed, as the Yuan shih (Official history of the Yuan) has it, to "impose the old regulations."'9 Toghon Temiir's reign title (nien bad) was changed to Chih-yuan, exactly the same as the title Khubilai had used from 1264 to 1294. With all possible literalism, Bayan seized the calendar, turned its pages back to the era of the Yuan founding, and restarted it. What could that have possibly meant? Why did he do it? Bayan's personal experiences provide a few clues. He and his ancestors had been kesigden, hereditary housemen to the khaghan and his family. As a youth, Bayan had served as warrior-houseman to Prince Khaishan, and for bravery in action in the last of the steppe wars (1300-6), he was awarded the title ba'atur (valiant) at a traditional princely convention, or khuriltai. When Khaishan became emperor (posthumous temple name, Wu-tsung; r. 130711), Bayan assumed a series of high central posts and military commands in China. After 1311, he served effectively in high-level provincial positions, during which time he donated large tracts of farmland (given to him by the emperor) to the kesig and to the Tibetan Buddhist hierarch who was chaplain to the Yuan court. In 1328 he had been a major power behind the succession of Khaishan's son Tugh Temiir (Wen-tsung). He was, as his biography says, "resolute, serious, intelligent, and decisive,"20 at least in his earlier years. All this appears to draw a portrait of an aristocrat of long and wide experience, with deep personal, ethnic, and institutional loyalties to the more specifically Mongolian side of that Sino-Mongolian hybrid, the Yuan dynasty. It is clear that Bayan believed that things had drifted in undesirable directions in the forty years since Khubilai's death and that he wanted to restore the status quo ante. But whatever that might mean, it was in no way clarified by the imperial edict that announced the readoption of the Chihyuan reign title. The edict spoke vaguely of portents of celestial warning. It said that although the realm was enjoying prosperity and peace, the astrologers had noted irregularities and that these called for improving government by way of reviving the old norms (cbiu tien) that had worked so well under Khubilai.21 In other words, major and perhaps painful changes were promised, at a time when things were officially acknowledged to be fairly quiet in the realm. Because there was no general sense of crisis, because there was no broad spectrum of political and moral support for what he was trying to do, and because his program lacked clear definition, Bayan was soon obliged to use some very forceful means to impose his will. 19 YS, 138, p. 3338. 20 YS, 138, p. 3335. 21 YS, 38, p. 830. Cambridge Histories Online © Cambridge University Press, 2008 TOGHON TEMOR AND BAYAN'S CHANCELLORSHIP 569 In all that follows, it is important to try to establish what Bayan himself had in mind and to distinguish that from what his many opponents, Chinese and non-Chinese alike, believed that he had in mind. There were two main parts to Bayan's reforms. One part, often overlooked, stemmed from his considerable administrative knowledge and experience in China, and it aimed at alleviating distress and improving the general welfare of the realm. Palace expenditures were cut (with the expenditure levels of Khubilai's era sometimes used as a target); the salt monopoly quotas were reduced; and conscientious and continuing efforts were made to provide timely and appropriate relief to areas throughout the country that had been stricken by famine and other disasters. So far, so good. It was the other part of his effort, the attempt to reestablish the ruling system as Bayan thought Khubilai had originally designed it, that in the end proved impossible to carry through. What this mainly entailed was reimposing in both the military and political spheres the sharp lines of ethnic separation that, in Bayan's view, constituted the absolute bedrock of Yuan rule in China. To attempt this amounted to putting half a century of history into reverse, and evidently Bayan had no idea at the outset just how difficult that would be. Cultural and social relationships among the Mongols, other foreigners, and the native Chinese elites of north and south had long obscured the oncesimple ethnic distinctions, which were now shot through with ambiguities and complex cross-shadings. Many ambitious Chinese had accommodated themselves to Mongol ways by adopting Mongolian names (a practice Bayan frowned upon), learning the Mongolian language, intermarrying, and, in a variety of ways, insinuating themselves into the Mongolian power structure. On the other side of the coin, many Mongols and other foreigners found China and the Chinese congenial. Take, for example, A-jung (d. ca. 1335), a Mongol of the Kereyid tribe, a kesigdei to the same Khaishan whom Bayan had served, and a competent civil and military administrator - in short, someone to all appearances much like Bayan. But A-jung went culturally in a wholly different direction from Bayan: He liked gambling, hunting, and playing ball, and he was also a student of Chinese history and an admirer of south China's lush landscapes. He owned a Hunan retreat called the Plum Moon Estate where he planted several hundred plum trees22 and enjoyed his cultivated pleasures in the close company of Chinese literati. Bayan's reimposition of ethnic separatism had the effect of placing in doubt interethnic personal relationships such as these, as well as patronage and career expectations. Therefore, Bayan's plans had no appeal what22 Sung Chiung, Yen-shihehi, 8, pp. 13a—15a; YS, 143, pp. 3420—1. Cambridge Histories Online © Cambridge University Press, 2008 570 SHUN-TI AND THE END OF YtJAN RULE IN CHINA soever to A-jung (who mournfully predicted to Yii Chi that Bayan would soon abolish the Confucian examination system) or to the many others like him. At the elite level, foreigners and Chinese could no longer be cleanly separated. This was at the heart of the political troubles encountered by Bayan's program. Specifically, Bayan's reassertion of the foreign ascendancy in China certainly included the disarming of all Chinese (and Koreans) and the confiscation of their horses. He forbade the Chinese henceforth from learning Mongolian or foreign scripts, although the measure was soon rescinded, and he reserved a range of leading positions in the bureaucracy exclusively for the Mongols and foreigners in China. It may have been zealous local officials, rather than Bayan himself, who ordered the confiscation of all iron agricultural tools and outlawed Chinese opera and storytelling. It was no part of Bayan's actual policies, but it was indeed indicative of the mass psychology of the time and the apprehension that his policies aroused among the Chinese, that the chancellor was widely believed to have ordered a nationwide roundup of all unmarried Chinese children and the extermination of all Chinese bearing the five most common surnames. One of Bayan's actual measures that provoked intense opposition at the elite level was his cancellation late in 1335 of the examination system. This measure affected all ethnic categories: northern Chinese, southern Chinese, Mongols, and other foreigners. In particular it dashed not only the career hopes of educated Chinese, but also those of many young Mongols and foreigners who lived in the provinces of China, without access to the kesig or to influential people in the capital. Much of their lives and energies had been focused on studious preparation for the examinations. What, then, was the point of abolishing the system and arousing their opposition? There is no simple answer. The decree canceling the examinations did not venture to offer a rationale. When challenged later, Bayan was unable to offer a satisfactory defense of the measure. Still, he stood firm, and the examinations scheduled for 1336 and 1339 were never held. The most articulate proponent of abolishing the examinations was not Bayan, but Cherig Temiir of the Turkic Arghun tribe. Like Bayan, he had served in the kesig as a youth and had then had a distinguished career as a highlevel administrator and military commander. No extremist, Cherig Temiir was knowledgeable and competent and particularly experienced in handling famine relief. What distressed him about the examinations was the expense they entailed: He had witnessed the commandeering of state transport and the outlay of scarce resources to make the arrangements for the examinations at the provincial level. He also urged that the school subsidies earmarked for examination candidates be used instead to support the kesigden.^ 23 YS, 142, pp. 3403—6. Cambridge Histories Online © Cambridge University Press, 2008 T O G H O N T E M O R AND BAYAN'S C H A N C E L L O R S H I P 571 Yet the issue went much further than that; it was not solely a matter of reducing expenses and reallocating resources. Khubilai had never held examinations. Since their institution in 1315, newly minted chin shih had enjoyed something of an unfair advantage in competing with young clerks and translators for entry-level positions in the regular bureaucracy, and they had, or were thought to have, better chances for subsequent promotion. Holding a degree was beginning to challenge service in the kesig for prestige. But a chin shih degree was no guarantee of administrative competence: Bayan himself and many others stood witness to that. Rather, abolishing the examinations was a way to underscore the value of aristocratic hereditary privilege and of demonstrated practical experience. In preparation for government service, the study of the Confucian canon, though certainly not proscribed, was relegated to a secondary position. So much for the reforms. Following aristocratic custom, after his rise to the chancellorship Bayan began to accumulate vast personal assets. Chinese sources hostile to Bayan are careful to list these in detail. He also came to hold a first-class princely fief and an awesome range of concurrent official positions over central and palace organs, capital guard units, and patrimonial business agencies. So crowded were the streets with officials when Bayan's birthday was celebrated in 1338 that one elderly Chinese official of high rank was nearly trampled to death. 24 The chancellor's own progresses through the capital featured a large and impressive military escort that reportedly made the emperor's entourage look modest by comparison. Perhaps Bayan, his political aims at least superficially achieved, was beginning to grow complacent and corrupt. But his position remained insecure. Bayan soon found he had enemies among a higher stratum of aristocracy than his own. Four enemies whom he prosecuted were princes of the blood. One of them, Chechegtii (prince of T'an), posted in command of troops in Mongolia, was arrested in 1338, brought to Ta-tu, and publicly executed outside one of the east gates. That was perceived as a grossly unjust act. The prince was held in popular esteem in the capital, and his standing was also high in Confucian circles in south China, because he had kept a southern scholar at his court for a time and had respected his advice. 2 ' And far down the social scale, out in the provinces of China, ominous commotions came to Bayan's attention. There were several small-scale millenarian uprisings in Honan, Kwangtung, and Kiangsi, which were quickly 24 Yang Yii ( 1 2 8 5 - 1 3 6 : ) , Shan chii hsin him (TSCC ed.), p. 17a. This work has been translated by Herbert Franke; see his Beitrage zur Kulturgacbichte Chinas unter der MongoUnbtmchafi: Das Sban-ku sinhua dts Yang Yii (Wiesbaden, 1956). 25 Wei Su (1303-72), Wei Tai-p'u chi (Wu-hsing, 1913), 8, pp. 8a-9b. Cambridge Histories Online © Cambridge University Press, 2008 572 SHUN-TI AND THE END OF Yt)AN RULE IN CHINA suppressed, but Bayan was quick to suspect that much wider conspiracies lay behind them. Late in 1339, a frustrated and unhappy Chinese clerk named Fan Meng carried out a mass murder of officials and seized Pien-liang, the provincial capital of Honan. He was soon caught and killed, but Bayan insisted that a huge Chinese conspiracy lay behind this incident, too, and demanded the most vigorous investigation and prosecution of everyone even remotely implicated, as well as a general purge of ethnic Chinese from certain sensitive bureaucratic positions. It was all too much. There were too many presumed enemies. Bayan's chancellorship had come to something of a dead end. The principal engineer of Bayan's ouster was none other than his nephew Toghto, a young man to whom he had accorded every kind of preferential treatment. In March 1340, in a very carefully arranged coup, Bayan was removed from all his offices and banished, first to Honan and then to the far south. He died in April on the way to banishment. TOGHTO AND HIS OPPOSITION, 1340-1355 The coup against Bayan represented to some degree a revolt of a younger generation of Mongols against an older one, a generation more acclimatized to China than its predecessors had been. With the demise of Bayan, the effort to restore the era of Khubilai was dead as a political ideal. Political conflict shifted to wholly new ideological ground. Toghto was about twenty-six years of age and well on the way to becoming the most brilliant figure of his time. He is described as having been tall and strong and a superior bowman. He had had ten years of grooming in the kesig, had served in leadership positions in the palace, and had developed impressive skills in backstairs intrigue. He had a modicum of Confucian learning as well, absorbed from his southern Chinese family tutor, Wu Chih-fang. As things turned out, Toghto would have two terms as chancellor of the right: from 1340 to 1344 and from 1349 to 1355. As things also turned out, late Yuan political history at the central level came to assume a certain periodic rhythm, with different administrations succeeding each other, each using a different set of guiding ideas, on average about everyfiveyears. Thus after Bayan had dominated the scene from 1333 to 1340, there followed Toghto, from 1340 to 1344. An interlude dominated mainly by Berke Bukha lasted from 1344 to 1349, and this was followed by Toghto's return to power, from 1349 to 1355. After 1355, the rhythm was broken, as the breakup of the Yuan realm was by then well under way, and the beleaguered central government no longer had effective political control of the country, Cambridge Histories Online © Cambridge University Press, 2008 T O G H T O AND HIS O P P O S I T I O N 573 although it still remained a focus of loyalty and a source of legitimacy even in some of the parts of China that it no longer directly governed. At first glance, the study of these post-Bayan administrations seems unconnected to an understanding of the imminent collapse of the dynasty, because none of them lacked for ideas or elan and all were willing to take on and remedy the major problems of the time and to make changes. Hardly effete and no longer reactionary, high administration after the fall of Bayan came into the hands of vigorous men of insight and capability. Under Toghto, the approach to political solutions was predominantly centralist. Under Berke Bukha, the approach was just the reverse - to give as wide a latitude as possible to provincial and local initiative. It seems highly problematical where responsibility for the Yuan collapse should be assigned: to the leaders personally, to systemic flaws, or to crisis conditions in China too overwhelming for any government to handle? Toghto's first administration certainly exhibited fresh new spirit. The young leader was quick to distinguish his regime as something wholly different from Bayan's. A new reign title was decreed to show this. Bayan's purges were called off. Positions that Bayan had closed to the Chinese were reopened to them. Many of the great Chinese literati came back to the capital from voluntary retirement or from administrative exile. The examination system was restored. Bayan's old adherents were dismissed. Bayan himself died in exile only a month after his removal. Toghto also gave a few early signs of a new and positive direction in central government. He spearheaded an effort - ultimately unsuccessful for technical reasons - to provide an all-water transport route from the sea through the capital to the Shansi foothills.26 Following that came a project, this time successful, to find funds for and finish the long-stalled official histories of the Liao, Chin, and Sung dynasties.27 The administration also permitted a new visibility to the emperor Toghon Temiir. Now nineteen, he had been allowed a minor role in the planning of the anti-Bayan coup. In July 1340, the young emperor issued a posthumous denunciation of his uncle Tugh Temiir and rid himself of his palace handlers and rivals. He sent his aunt Budashiri, the grand empress dowager, into exile and also arranged the exile, and possibly also the subsequent murder, of his cousin and designated successor, El Tegiis. His own infant son, Ayushiridara, he entrusted to Toghto's household for feeding and care. The ruler now began 26 See John W. Dardess, Conquerors and Confucians: Aspects of political change in late Yuan China (New York, 1973). PP- 7 9 - 8 o . 27 On the latter, see Hok-lam Chan, "Chinese official historiography at the Yuan court: The composition of the Liao, Chin, and Sung histories," in China under Mongol rule, ed. John D. Langlois, Jr. (Princeton, 1981), pp. 56—106. Cambridge Histories Online © Cambridge University Press, 2008 574 SHUN-TI AND THE END OF YtJAN RULE IN CHINA to appear in person to conduct Confucian state sacrifices, to listen to Confucian lectures, and to host state banquets. In June 1344, after a series of local rebellions had broken out in widely scattered areas of China, the emperor accepted Toghto's unusual request to resign his office.28 The several short-lived administrations that followed from 1344 and 1349 developed an agenda very different from Toghto's, for some compelling reasons. The cumulative effects of years of natural disasters throughout China, together with the continued spread of banditry and other signs of civil disorder, demanded that central government give special attention to improving provincial and local administration so as to handle these problems. Mainly this required two things: ensuring the appointment of men of quality and ability to local positions and giving those people discretionary authority in handling relief and other problems - in effect, decentralizing national relief efforts. One of the principal figures of this period, Berke Bukha, had been an effective provincial official and had discovered at first hand (when dealing with the aftermath of the great Hangchow fire of 1341) that one had to violate central regulations in order to issue relief before it was too late.2* Similarly, local military garrisons needed blanket authorization in order to combat roving bandits. And in handling disasters or civil disturbances, local officials needed to dictate less and to do more to gain the cooperation of the local people. In 1345, the administration sent out twelve investigation teams, each headed jointly by a Chinese and a non-Chinese official, to visit every part of the realm, correct abuses, and do whatever was needed to "create benefits and remove harm" for the people. Also, the boundaries of some local jurisdictions in Honan were carefully redrawn in order to coordinate more efficiently the antibandit efforts.30 Toghon Temiir was encouraged to participate in certain aspects of this new program. In person he exhorted newly appointed local officials to achieve good results, and he took part in rewarding and promoting those who had done outstanding work at the local level. Yet, far from abating under this new administrative approach, troubles mounted in Yuan China in the 1340s. The troubles appeared to be of such a nature, or on such a scale, that piecemeal initiatives by local officials, or local conciliation, were inadequate to deal with them. The central government was now also faced with chronic revenue shortfalls. Maritime grain shipments 28 For the details, see Dardcss, Conquerors and Confucians, pp. 8 0 - 1 . 29 Sung Lien ( 1 3 1 0 - 8 1 ) , Sung Wen-bsien kung ch'iian cbi (SPPY ed.), 49, pp. 6 b - n a ; Yang Yu, Shanchii bsinbua, pp. 35a~36b; YS, 51, p. 1100, and 138, p. 3366. 30 For the latter, see SuT'ien-chiieh (1294—1352), Tzu hsi wen kao; repr. in vol. 3 ofYuan tai then pen wen cbi bui k'an (Taipei, 1970), pp. 12a— 15a; Yang Wei-chen, Tung-wei-tzu wen chi, 4, pp. 9 b - i o b . Cambridge Histories Online © Cambridge University Press, 2008 TOGHTO AND HIS OPPOSITION 575 had not only seriously declined, from a peak of 3.34 million bushels in 1329 to 2.6 million in 1342; but beginning in 1348, they continued only at the pleasure of a major piratical operation led by Fang Kuo-chen and his brothers, whom the authorities had failed to suppress and therefore tried to conciliate.31 In addition to that, the Yellow River, swelled by prolonged rains, kept breaching its dikes and finally began shifting its course, creating widespread havoc and ruin.32 A different administrative strategy seemed to be in order, and after some complex struggles within the bureaucracy, in August 1349 the emperor recalled Toghto to office as chancellor of the right. Radical solutions were now sought and implemented. Enthusiasm ran high among Toghto and his partisans. It reportedly was said at the time that the earlier [Yuan] prime ministers lacked renown, and nothing in the way of ceremonial, literary, or institutional achievement is remembered of them. But [Toghto} wants to undertake great acts and dazzle the world; he wants to surpass the old methods of the ancestors, and leave behind an immortal name in the historical records.'3 It may be argued that these words accurately capture the euphoria of the moment. Surely none of what followed would even have been attempted without there having been a powerful belief in government circles that all crises were soluble, that the world could be remade overnight on orders from the top. All of Toghto's new plans entailed central direction and control.34 New ideas from below were welcomed, but once accepted, it was the central government that implemented them. Regional or local initiative was now heavily restricted. To take Toghto's major projects in their chronological order, the first answered the need to find more revenue immediately. It was not thought feasible to try to increase rates on the traditional tax sources (land, salt, commerce, etc.), most of which were continuing to shrink. Consequently the decision was taken late in 1350 to print a new issue of (inadequately backed) paper currency - two million ingots' worth in the first run in 1351 - and to circulate it by way of government payments for materials and labor.35 31 See Dardess, Conquerors and Confucians, pp. 88—9. 32 See ibid., p. 87. 33 Ch"iian Heng (fl. 14th c ) , Keng shen wai sbih; repr. as vol. 3 of Shib liao ts'ungpien (Taipei, 1968), pp. I9a-b. This translation differs slightly from both that in Dardess, Conquerors and Confucians, pp. 96— 97, and that in Helmut Schulte-Uffelage, trans, and ed., Das Keng-shen wai-sbih: EineQuelle zurspalen Mongolenzeit (Berlin, 1963), p. 56. 34 For a detailed account of Toghto's second regime, see Dardess, Conquerors and Confucians, pp. 95—118. 3 ; Yuan money is the subject of a large bibliography. The standard Western-language authority is still Herbert Franke's Geld und Wirtscbaft in China unter der Mongotenherrschaft. A recent contribution is Nancy S. Steinhardt's "Currency issues of Yuan China," Bulletin of Sung and Yuan Studies, 16 (1980), pp. 5 9 - 8 1 . Cambridge Histories Online © Cambridge University Press, 2008 576 SHUN-TI AND THE END OF YUAN RULE IN CHINA Following the instant solution of the financial question, the government next announced, in April 1351, its plan to tame the Yellow River, by rechanneling it along its lower course, so that it would againflowinto the sea south of the Shantung peninsula. There was much opposition to this idea, and Toghto and his people were well aware that nothing like it had ever been done before. Yet, funded by the new currency issue, the Yellow Riverrerouting project was begun in May and was brought to a successful conclusion in December of the same year, 1351; 170,000 troops and civilian workers took part in the work. An official commemorative inscription by Ou-yang Hsiian celebrates this triumph of hydraulic engineering in great technical detail.' 6 Meanwhile, in the summer of 1351, there broke out in the Huai valley region, at some distance from the river project, first one, then another, and then still another popular uprising in what rapidly spread to become a nationwide social explosion of extraordinary proportions, as bandits, religious sectarians, and other dissidents ran amok, moving from place to place capturing administrative cities, pillaging stores, killing officials, and settling local scores. These rebellions are described at great length elsewhere.37 Suffice it to relate here that far from reeling in shock from this catastrophic outbreak of disorder, Toghto and his people acknowledged it, quickly accepted the challenge of dealing with it, and over the next three years (by 1354) definitely gained the upper hand. In a sense, central authority came to view national pacification as one more large-scale problem that it was already equipped and competent to deal with. It has been asserted that owing to corruption, mismanagement, and the like, the regular Yuan armies had grown too decadent to defend the dynasty from these upheavals.38 There may be much truth in this assertion, but the fact remains that no matter what their condition or state of training, there were never at any time enough regular Yuan troops to handle the nationwide breakdown of order that set China aflame in the early 1350s. Some Mongolian units were undisciplined and poorly prepared at the outset, but if they lost some early encounters with the insurrectionary forces, they soon won a good many more. Yuan military units of every description and national origin were pressed into service, and especially in south China, local defense militias, Chinese armies recruited for the occasion, played a major part in 36 YS, 66, pp. 1646-54. This remains untranslated, as far as I know, but is briefly discussed in Joseph Needham et al., Science and civilization in China, vol. 4 , Physics and physical technology, pt. 3, Civil engineering and nautics (Cambridge, 1971), p p . 3 2 3 , 344; and Yang Lien-sheng, Excursions in sinology (Cambridge, Mass., 1969), p p . 222—3. 37 See the chapters by Frederick W. Mote and Edward L. Dreyer in F. W. Mote and D. C. Twitchett, eds., The Ming dynasty, 1368—1644, pt. i . v o l . 7 ofThe Cambridge history ofChina (Cambridge, 1988). 38 For example, Hsiao, The military establishment of the Yiian dynasty, p. 6 3 . Cambridge Histories Online © Cambridge University Press, 2008 TOGHTO AND HIS OPPOSITION 577 turning back the rebel tide. The rebellions could not have been handled otherwise.39 It is remarkable that Toghto managed not only to create a nationwide apparatus of pacification but also to keep effective control over it. By always assembling and disassembling the larger military groups, by constantly transferring commanders from place to place, by not allowing the heads of any single government agency or princely establishment to dominate any major operation, and by carefully keeping supply organizationally distinct from other military operations, Toghto was able to prevent a downward slippage of military power into regional hands and to prevent military commanders from establishing provincial power bases. The largest, or at least the symbolically most important, campaigns Toghto chose to command in person: He led a successful expedition in 1352 to recapture Hsii-chou (in present-day northwest Kiangsu Province), an especially critical administrative center, and to restore order in the Huai valley area. While all these military operations were going on, the maritime grain shipments from south China were ended. Rather than negotiate with the pirate Fang Kuo-chen, who was still active, with a view to resuming the shipments to the capital, Toghto decided instead to make a rice basket of the greater metropolitan area itself. This decision was consistent with Toghto's whole centralist approach to administration. It was a colossal and extremely costly undertaking, being allocated five million ingots in new paper currency — two and a half times more than the cost of rerouting the Yellow River. North China was unusually wet in the Yuan period, and earlier trials had shown that rice could indeed be grown there. In 1353, a supervisory branch office of agriculture (fen ssu nung ssu) was set up there, which recruited two thousand dike builders and paddy farmers from the south for a year's paid service as technical advisers to the native farmers, who were unfamiliar with rice cultivation, and other farmers were brought in from Shantung to augment the local population. Other state farm systems were established in Honan Province and in southern Manchuria. It is not wholly clear how successful these new state-run farms actually became, because of Toghto's imminent political downfall and the continuation of the civil wars, but as a bold plan to revolutionize north China's age-old farming traditions overnight, it is an example of Toghto's vision of the limitless possibilities of inspired government. Late in 1354, Toghto took personal command of a second grand military expedition, this time against the city of Kao-yu (just north of Yang-chou, on 39 Details are given in Dardess, Conqutnn and Confucians, pp. 104 fT. Cambridge Histories Online © Cambridge University Press, 2008 578 SHUN-TI AND THE END OF Yt)AN RULE IN CHINA the Grand Canal), which was occupied by a former salt smuggler, Chang Shih-ch'eng. This expedition, as it turned out, was Toghto's last official act. Had the siege of Kao-yu in fact forced Chang Shih-ch'eng's surrender, which was all but inevitable, the back of the nationwide rebellions would have been resoundingly broken. Even as it was, those rebels still active had been chased into hiding or were barely surviving. But while the siege was still in progress, in a stunning act of misjudgment and bad political timing, Toghon Temiir suddenly issued Toghto an order of dismissal and banishment. Toghto, unfortunately for the dynasty, obeyed. The siege thereupon disintegrated. The Yuan lost the military and political initiative. The rebellions, all but quashed, took on a new form and new life. It is from this crucible that the successor dynasty, the Ming, emerged and thirteen years later reunified China. Why was Toghto dismissed? There seem to be several considerations, probably one of which was a well-establishedfive-yearcycle of administrative turnover. Over a period of five years or so, corruption, favoritism, and personal ill feelings seem to have reached a critical point within each successive late Yuan ruling faction; this was certainly the case in Toghto's circle, in that it was Khama, one of his own disaffected adherents, who took a leading role in the palace intrigues that led to his downfall. Moreover, Toghto's ambitious programs had pretty much run their course and achieved their purposes. The Yellow River was tamed, and the various insurrectionary movements were all but destroyed. There was nothing left of Toghto's national agenda but details. It was time for a change. In this connection, one must reckon with those in government opposed to Toghto not so much on personal grounds but because they held different beliefs about the appropriate goals and operating procedures of government. Toghto's ambitious activism demanded discipline and centralization. The Censorate was held in check; provincial and local officials were given very little initiative; and military commanders were given the least possible freedom of action. There is clear evidence that at least some of the officials who had served the administrations of 1344-9 disliked Toghto's tight controls and resented the protection he gave his own loyal partisans. These men now demanded a return of the decentralized mode of governance and a greater scope for the institutional, regional, and personal initiative that characterized it. This philosophical conflict reminded some contemporaries of the struggle between the reform and conservative factions in the late Northern Sung. Togh5n Temiir also had his own reasons for removing Toghto, who, like his uncle Bayan before him, had grown very powerful. Perhaps inadvertently, he had shown this power to the emperor by delaying his agreement to the formal investiture of Ayushiridara as heir apparent. Toghon Temiir was anCambridge Histories Online © Cambridge University Press, 2008 TOGHTO AND HIS OPPOSITION 579 gered and upset. What was it that had poisoned relationships between the chancellor and the imperial house? By all the signs, Toghon Temiir, now thirty-four years of age, had withdrawn into a kind of semiretirement. 4 ° He was regularly participating with a select circle of adepts, and an all-female dance ensemble and orchestra, in the sexual rituals of Tibetan Buddhism. On at least one occasion, he sponsored a holy circumambulation of the imperial palace grounds by a group of 108 monks. He was also having built a huge pleasure boat (the model for it was of his own making) for sailing on the lake in the imperial palace, and he himself also had a major hand in the design and fashioning of a large, technically elaborate clepsydra, or water clock.•»' Perhaps in order to accommodate the new boat, a costly project was later undertaken to dredge the palace waterways. Meanwhile, Toghon Temiir authorized a series of steps to allow his oldest son, Ayushiridara, to learn about and assume some responsibilities for government. In 1354 Ayushiridara was about fifteen years old. At the age of nine he had been ordered to learn Uighur writing (Wei-wu-erh wen tzu), and at the age often, Chinese. Shortly after this, in 1349, with elaborate and solemn ceremony, a special Chinese-style academy was set up in the palace for Ayushiridara, with a tutorial staff of nine men. A senior tutor, Li Hao-wen, compiled four textbooks on Chinese subjects especially for his young pupil. Four years later a Household Administration for the heir apparent (Chanshih yuan) was established, with a staff of no fewer than eighty-three officials and clerks. Other, independent offices for the heir apparent controlled a ceremonial guard plus two regular guard units, and in addition there were assigned to Ayushiridara 250 housemen {ayimagh kesigden) of his own. The housemen received 27,500 ingots in cash grants, and Ayushiridara's consort, 100,000 more. An unused palace was selected and repaired for Ayushiridara to live in. He was given the privilege of appointing his own officials, and late in 1354 (while Toghto was out on campaign) he was given the power of review (ch'i) over all official business sent up to his father, the emperor. In sum, Toghon Temiir created by degrees a new, large, and costly center of political influence within the palace, headed by his designated heir. All that remained was to conduct the final official ceremonies of Ayushiridara's installation as heir apparent. The obvious inference is that Toghto delayed those ceremonies because he considered the rise of Ayushiridara a threat to himself. Yet Ayushiridara and Toghto were actually very close. Ayushiridara's childhood was spent in 40 In one view, the emperor's interest in government peaked during Toghto's first incumbency (1341 —4). See Fujishima Tateki, "Gen no Juntei to sono jidai," Qtani gakuho, 49 (1970), pp. 50—65. 41 Joseph Needham et al., Heavenly clockwork: The great astronomical clocks of medieval China, 2nd ed. (Cambridge, 1986), p. 140. Cambridge Histories Online © Cambridge University Press, 2008 580 SHUN-TI AND THE END OF YUAN RULE IN CHINA Toghto's house, and when at age ten he was given his earliest tuition in Chinese books (the Hsiao ching), his instructor was Toghto's family tutor, Cheng Shen.42 Moreover, Toghto personally contributed 122,000 ingots toward the building of a Buddhist temple outside one of the north gates of Ta-tu, so that prayers might continually be offered for Ayushiridara's well-being. Another possible reason for the delay may have been that because Ayushiridara was not a son by a principal consort, Toghto considered him unqualified. Ayushiridara's mother was a former palace maid and tea server of Korean nationality, of whom Toghon Temiir was very fond. He had made her second empress (//' erh huang hou) in 1340, an act that many people had opposed, in view of the low place that the Mongols had assigned to the Koreans as a race and the effect of her new status on the complex issue of Yiian relations with Korea.43 Toghon Temiir's principal empress, a selfeffacing Mongolian woman, bore only one male child, who died in infancy. The dates of his short life are not known, but if he were alive in 1353-4, then Toghto's hesitancy on the investiture question becomes understandable. But all this is speculation: Toghto may simply have been too busy dealing with the empire's crises to schedule the event. All that really matters is that the emperor suspected the delay to have been deliberate. In April 1355, after Toghto's removal, the solemn investiture ceremony was at last carried out. THE DISINTEGRATION OF THE YUAN The emperor may have considered the dismissal of Toghto a fairly routine measure, something he had done on a number of occasions in the past, without the integrity of the dynasty having been compromised as a result. But the dismissal of Toghto in fact put an end to the Yiian as an integrated political system, largely for the reason that unlike the period of decentralization from 1344 to 1349, various new military and administrative structures had been created in many parts of China in response to the insurrections after 1351 that had both the manpower and resources to act independently and soon did so. Official appointments at the regional level and below could no longer be made by the authorities in Ta-tu as a matter of central bureaucratic prerogative, as had normally been the case earlier. The provinces of China fell into the hands of autonomous regional figures (warlords is perhaps the right word), some of whom had earlier been rebels and others commanders of antirebel forces. From 1355 to 1368, Yiian central government did the best 42 Sung Lien, Sung Wm-hsien kung ctiiian chi, 49, pp. 6b—1 la. 43 For the Korean side of this story, see Hsiao, Yiian tat shih hsin t'an, pp. 2 3 1 - 6 2 . Cambridge Histories Online © Cambridge University Press, 2008 THE DISINTEGRATION OF THE YOAN 581 it could to retain at least the token loyalty of its self-supporting commanders in the field and to negotiate diplomatically the nominal surrender of former rebels, men like Fang Kuo-chen and Chang Shih-ch'eng. Meanwhile, the leaders of other rebel movements (e.g., Ch'en Yu-liang and Chu Yuan-chang) emerged after Toghto's downfall to build new and stronger regional governments and, on the whole, to maintain consistently hostile attitudes toward the Yuan. Thus the Yuan central government itself became in many respects no more than a regional government, controlling in China only the capital and its outlying areas, although it retained to the end a residual legitimacy as the government of all of China (see Chapter 9). The most significant development of Toghon Temiir's last years in China (1355-68) was the rise of the Ming dynastic movement based in Nanking and, in particular, the alliance it struck in 1360 with the leaders of what was in fact a fundamentalist, Confucian moral-political revolution. This revolution took shape in the general conditions of late Yuan rule, in the insurrections of 1351 to 1354, and in the dynastic disintegration that followed Toghto's dismissal. The problem of the origin of this moral revolution deserves more than the few paragraphs available here, because it involves such subtle filaments as the history of society and family, taxation and resource mobilization, legal history, and the development of Confucian thought.44 Suffice it to say that this revolution gathered its force at the grass roots, principally in southeast China. One early manifestation of it can be traced to 1342 and the successful demonstration of a pilot project in tax and service reform in Shao-hsing Prefecture, a negligible producer of revenue. But in this small crack in the giant structure, as it were, Confucianminded activists in both office and private life effected in the face of many obstacles an equitable reapportionment of fiscal obligations, by combining the reform with a campaign of moral reawakening. The local community was deliberately polarized into good and evil through the revival of the ancient village drinking ceremony (hsiangyin chiu It), in which the selfish and recalcitrant were publicly exposed and disgraced and the virtuous were honored. This project was repeated in 1350 in Chin-hua and Ch'u-chou prefectures; here Yii Ch'iieh, the Tangut encountered earlier on the non-Chinese portion of the chin-shih list of 1333, was a guiding official hand. Again, the work was exhausting; the revenue benefit to the Yuan government was almost nil; and 44 Recent work in this direction includes Paul Heng-chao Ch'en, Chinee legal tradition under the Mongols: The code of 1291 as reconstructed (Princeton, 1979); Jennifer Holmgren, "Observations on marriage and inheritance practices in early Mongol and Yuan society with particular reference to the levirate," Journal of Asian History, 20 (1986), pp. 127—92; and John D. Langlois, Jr., "Political thought in Chin-hua under Mongol rule," in China under Mongol rule, ed. John D. Langlois, Jr. (Princeton, 1981), pp. 137-85. Cambridge Histories Online © Cambridge University Press, 2008 582 SHUN-TI AND THE END OF YUAN RULE IN CHINA the effort was made at so low a level in the official administrative hierarchy (and so irrelevant to the thrust of Toghto's central planning) that it quite escaped the notice of the court. However, it was celebrated as a great moral triumph in local Confucian writings, and these radiated disproportionately wide influences. If the Yuan authorities in Ta-tu took little note of these influences, the Ming founders soon would.45 It is also in the writings of local Confucians that one discovers "public" reactions to the mass insurrections and the breakdown of Yuan central authority after 1355. There was a general consensus that the insurrections were an inevitable popular response to the size, cost, and corruption of the Yuan government and to the heavy and iniquitous revenue demands that this produced. To remedy these abuses it was generally agreed that the costs of government must be cut severely and that all corruption be resolutely purged from bureaucracy. (The Yuan state was never in a position to proceed very far along these lines, and eventually it was the Ming state that made frugality a guiding policy.) As to the regionalization of Yuan rule after 1355, Confucian opinion was divided, and there was indeed a dilemma in it. Some contemporary Confucian writers argued that the appointment of autonomous regional warlords like Fang Kuo-chen or Chang Shih-ch'eng as Yuan provincial officials was congruent with the ancient Chou pattern of "feudal", decentralization and hence legitimate, provided that the regional leaders obeyed the court and conducted themselves in accordance with the strictest ethical standards. It was believed they would do this if the Confucians could exert enough moral pressure on them. This line of argument was totally unrealistic, but it promised a reduction of armed conflict and was attractive enough to help keep Yuan loyalism alive in many parts of south China into the 1360s. A small but powerful minority of Confucians argued the opposite, and this was a voice of moral revolution. In their view, regionalization was no good at all; it simply masked the perpetuation of the iniquity, the acquisitiveness, the self-aggrandizement, and the chronic favoritism and corruption that had brought the Yuan to its present sorry state in the first place. A dynasty that rewarded criminals with the highest of its offices and honors must forfeit all claim to the moral leadership of society. One of the most forceful and articulate partisans of this line was Liu Chi, the thirty-eighth name (of fifty) on the Chinese part of the chin shih list of 1333. Liu Chi had held several low-ranking local and provincial posts through the early 1350s, in which he had ample opportunity to nourish his 43 John W. Dardess, "Confucianism, local reform, and centralization in late Yuan Chekiang, 1342-39," in Yuan thought: Chinese thought and religion under the Mongols, ed. Hok-lam Chan and Win, Theodore de Bary (New York, 1982), pp. 327-74. Cambridge Histories Online © Cambridge University Press, 2008 THE DISINTEGRATION OF THE YOAN 583 outrage at the malfeasance and cover-ups in which he found many of his official superiors engaging. He was early on so implacable an enemy of Fang Kuo-chen that on one occasion he was imprisoned after the court decided, in the interest of preserving the peace, to try to accommodate the pirate rather than suppress him. Later, in 1356, Liu was appointed to the modest post of registrar in the Chiang-Che Branch Bureau of Military Affairs that had just been established in Hangchow (one of many examples of the proliferation of Yuan bureaucracy in the crisis years of the 1350s). As registrar, Liu Chi was at once sent to inland Chekiang, where a superior, Shih-mo I-sun, administrative assistant in the same agency, had charge of several prefectures. What follows is a singular and crucial episode in late Yuan history. It is important to try to reconstruct this scene because of its direct implications for the future of China. Far away in Ta-tu, the Yuan central government coped as best it could with a very confused national situation. Provincial officials now enjoyed plenipotentiary powers. As of March 1356, all ethnic qualifications were dropped for local official appointees. Rebels abandoned the war-wrecked Huai region of central China to set up base areas elsewhere: Chu Yuan-chang, leading what was still to some extent a religious sectarian movement, crossed the Yangtze River and took Nanking; Chang Shih-ch'eng, saved at Kao-yu by the emperor's cashiering of Toghto, came south and took Soochow. The Chiang-Che provincial governor Dash Temiir, a Confucian-trained Khangli aristocrat, desperate for expedients, tried setting the warlords against one another, cooperating first with one and then with another. In 1356 he induced Fang Kuo-chen to help in a fight against Chang Shih-ch'eng. The Chiang-nan Branch Censorate liked this strategy. But the Chiang-Che Branch Bureau of Military Affairs resolutely opposed it, in part because the governor himself had regularly appointed to that body men who hated Fang Kuo-chen. Each agency had its own army. In addition, there existed independent militia forces, which compounded the confusion by their penchant for unpredictably switching sides. This is how badly fragmented Yuan authority had become by 1356. Liu Chi and Shih-mo I-sun and their colleagues occupied one small splinter near the bottom of this shattering system; yet it was they, and they alone, who possessed the clarity of moral vision to advertise openly their thought and action in the provincial interior as the only possible working model of the kind of effort needed to bring about a true and lasting restoration of Yuan rule in China. In the provincial interior, good and evil completely disentangled themselves and migrated to the opposite ends of an almost Manichaean polarity. All good was altruistic and centralizing. All evil was self-interested and particularistic or regionalizing. One would achieve local pacification by placing all resources and leadership at the head of the forces Cambridge Histories Online © Cambridge University Press, 2008 584 SHUN-TI AND THE END OF YUAN RULE IN CHINA of good and directing them passionately and relentlessly against the forces of evil (i.e., landlords and "bandits" believed to be in league with Fang Kuochen). Shih-mo I-sun was infinitely more than an administrative assistant in the Branch Bureau of Military Affairs in an obscure corner of southeast China. He was, according to Liu Chi, a world-saving hero whose true destiny was to rectify and restore the Yuan dynasty, just as two thousand years earlier, Duke Huan of Ch'i had rescued the failing Chou. In 1357, the Yuan court agreed to give minor promotions to Shih-mo I-sun and Liu Chi and their circle, but it refused to raise them to a more influential level in the Yuan bureaucracy. The policy of regionalization would continue. Liu Chi thereupon resigned in despair and, as did several other Confucian theoreticians from the provincial interior at this time, wrote out his thoughts on morals and politics. In 1360, these ideologues joined the future Ming founder, Chu Yuan-chang, who had just invaded their territory. And so it was the first Ming emperor, rather than Toghon Temiir, who made effective use of the techniques of tax equalization and popular mobilization that had been recognized in Chiang-Che Province since 1342, who adopted the cause of moral absolutism the Yuan had refused in 1357, and who eventually fastened upon China a centralization of an unprecedentedly harsh kind.*6 CONCLUSION: WHY DID THE YUAN DYNASTY FALL? For some years after its retreat to the steppe in 1368, the Yuan court maintained its claim to China, but it never seems to have gone beyond selfserving fantasy to make any sober-minded effort to understand the reasons for the catastrophic collapse of its rule there. By contrast, that collapse was so fresh and vivid in the minds of the Ming founders that to an appreciable extent, they explicitly fashioned the Ming to counteract what they seriously thought to have been the causes of the Yuan breakdown. Put simply, their prescription was that if bureaucracy were cut in size and unflinchingly purged of any sign of selfish and corrupt behavior, and if the emperor himself took direct charge of things and acted without fear or favor (as Toghon Temiir had most lamentably failed to do), the ruling system would truly reinforce fundamental Confucian moral principles, and a millennium of peace and plenty would ensue. Were these Ming policies based on a correct or fair assessment of the Yiian's shortcomings? Did the Yuan collapse because Toghon Temiir ne46 John W. Dardess, Confucianism and autocracy: Professional elites in the founding of the Ming dynasty (Berkeley and Los Angeles, 1983). Cambridge Histories Online © Cambridge University Press, 2008 WHY DID THE YOAN DYNASTY FALL? 585 glected his duties, allowing factional struggle in the bureaucracy to intensify and corruption to grow unchecked? Perhaps. At least, it does seem possible to interpret broadly the events of the 1350s and 1360s as an example of what may happen when a large, privileged ruling elite is faced with worsening deficits in its national resource base: A division of the realm into autonomous parts does afford regional elites closer access to resources. It would be hard indeed to build a persuasive case on behalf of Toghon Temiir's ability to provide inspiring leadership. Toghto tackled energetically the problem of maintaining continued central control over national resources and personnel, in part, of course, through the desperate expedient of printing more and more unbacked paper money, an inflationary measure that could not have continued indefinitely (in fact, paper money had already become worthless and ceased to circulate in 1356). It might be argued that the integrity of the Yuan realm was sustained as long as it was only by the extraordinary personal talents of Toghto and that, inopportune as his dismissal may have been, centralized Yuan rule would have not long survived him anyway. The lessons that the Ming founders drew from the chaotic events of the late Yuan did have some rational basis. But it may be worth recalling that the fourteenth century was calamitous everywhere. Within and beyond the various Mongol empires, from Iceland and England at one end of Eurasia to Japan at the other, societies were suffering plagues, famines, agricultural decline, depopulation, and civil upheaval. Few societies were spared at least some of the symptoms. China was spared none of them. No fewer than thirty-six years in the fourteenth century had exceptionally severe winters, more than in any other century on record.47 In the greater Yellow River region, major floods and droughts seem to have occurred with unprecedented frequency in the fourteenth century.48 Serious epidemics broke out in the 1340s and 1350s. 49 Famines were recorded for almost every year of Toghon Temiir's reign, leading to great mortality and costing the government vast sums in relief.'0 These natural disasters created huge numbers of uprooted and impoverished people, fodder for the revolts that wracked the realm in the 1350s. The various late Yuan regimes all tried seriously to alleviate these disasters. None ignored them. Yuan medical and food relief efforts, by all appearances, were both conscientious and sophisticated. Indeed, the history of Toghon Temiir's reign raises the question whether any regime could have 47 See H. H . Lamb, Climate: Present, past and future (London, 1977), vol. 2, p. 4 4 7 . For details on the extremely severe cold during Toghon Temiir's reign, see YS, 5 1 , pp. 1 0 9 7 - 8 . 48 Chao Kang, Man and land in Chinese history: An economic analysis (Stanford, 1986), p. 203. 49 YS, 51, p. 1111. 50 YS, 51, pp. 1109-10. Cambridge Histories Online © Cambridge University Press, 2008 586 SHUN-TI AND THE END OF Y0AN RULE IN CHINA coped with these repeated large-scale disasters any better than the Yuan did. It might well be the case that the long-term cumulative effects of such repeated natural calamities were too great for any government to handle and that if normal conditions had prevailed in China, the Yuan dynasty might have lasted much longer than it did. The principal factors and the chains of causation in the fall of the Yuan will certainly be studied and debated for a long time to come. Yet we should recollect that the Yuan was not in any direct sense a victim of the blind forces of history. In 1368 it was forcibly driven out of China by a consciously extremist, morally revolutionary movement that, by the most diligent efforts, through sheer determination and hard struggle, took every advantage of the palpable weaknesses of the Yuan government of the 1350s and 1360s to impose its own vision of peace and order on China. The Yuan government itself had been given an opportunity to endorse that same vision, and it had understandably declined it. In the direct sense, the Yuan fell at last because the Ming founders willed it. Cambridge Histories Online © Cambridge University Press, 2008 CHAPTER 8 THE YUAN GOVERNMENT AND SOCIETY GOVERNMENT The structure of Yuan government took shape during the reign of Khubilai (Emperor Shih-tsu, r. 1260—94). Although institutional functions and forms shifted throughout the Yuan period, the essential components of the government bureaucracy formulated under Khubilai remained intact until the end of the dynasty in 1368. Khubilai himself supplied many of the innovations in government. He heeded the advice of Chinese, Khitans, Jurchens, Turkic Uighurs, Tibetan Buddhists, and Mongols in order to create a bureaucratic system that reflected all these various cultures. The official terminology of Yuan bureaucracy might lead to the conclusion that Khubilai merely adopted an almost purely Chinese structure of government, but in fact, the Yuan bureaucracy consisted of a mixture of different political and cultural elements. Even the "purely Chinese" elements of Yuan bureaucracy can be traced to Khitan Liao, Jurchen Chin, and Chinese T'ang-Sung governments. The strongest Chinese influence at Khubilai's early court came from Liu Ping-chung (1216—74), a Ch'an Buddhist and confidant of the Mongolian emperor. Under the direction of Liu Ping-chung and a small group of Chinese advisers, including Wang O (1190-1273), Yao Shu (1202-79), and Hsu Heng (1209-81), a central government administration was established within the first decade of Khubilai's reign.' The traditional Chinese tripartite division of authority among civil, military, and censorial offices was kept intact (at least on paper) by the establishment of the Central Secretariat (Chung-shu sheng) to manage civil affairs, the Privy Council (Shu-mi yuan) to manage military affairs, and the Censorate (Yii-shih t'ai) to conduct internal surveillance and inspection.2 The actual functioning of both central and 1 See Hok-Iam Chan, "Liu Ping-chung (1216-74): A Buddhist-Taoist statesman at the court of Khubilai Khan," T'ouag Pao, 53 (1967), pp. 98—146. Liu Ping-chung's biography is in Sung Lien et al., eds., Yuan shih (Peking, 1976), 157, pp. 3687-95 (hereafter cited as YS). See also Hok-lam Chan, "Wang O (1190-1273)," Papers on Far Eastern History, 12 (1975), pp. 4 3 - 7 0 . 2 The following on Yuan central government relies on David M. Farquhar's "Structure and function in the Yuan imperial government," in China under Mongol rule, ed. John D. Langlois, Jr. (Princeton, 587 Cambridge Histories Online © Cambridge University Press, 2008 588 THE YUAN GOVERNMENT AND SOCIETY local government agencies reveals a great deal of overlap between civil and military jurisdictions. Such overlap, for which the Mongols were much criticized by Chinese scholar-officials, derived from the Mongols' traditional reliance on military institutions and offices as the core of governance. In spite of the Mongols' inclination to entrust power to the military, a selfcontained civilian bureaucracy was created in China. At the very top of the Yuan civilian government, as it took shape under Khubilai, was the Central Secretariat, first established under Ogodei (T'ai-tsung) in 1231 but, as with most Yuan institutions, clearly defined only under Khubilai. The Central Secretariat was the nerve center of the entire civilian bureaucracy. Most other agencies in the Yuan structure of communication and control were ultimately responsible to it. All memorials to the emperor, with the exception of those written by high-ranking military and censorial officials, for example, passed through the Central Secretariat. In turn, that office was empowered to make recommendations, draft regulations, and make responses subject to imperial approval. In addition to its role as communication center, the Central Secretariat controlled official appointments to virtually all civilian offices in the empire. However, appointments to the military, the Censorate, the Bureau of Imperial Etiquette (Hsiian-hui yuan), and the Bureau of Buddhist and Tibetan Affairs (Hsiian-cheng yuan), and to certain offices in the hereditary appanages were handled through their own channels. At various times in the Yuan, a Secretariat of State Affairs (Shang-shu sheng) was established, only to be abolished again. The Yuan, however, never employed all three of the traditional three secretariats (or three departments, san-sheng) that had existed in T'ang times, that is, the Shang-shu sheng, the Chung-shu sheng, and the Men-hsia sheng (Chancellery).3 In its reliance on only one such central government secretariat rather than on three, the Yuan most resembled the Jurchen Chin dynasty, which in 1156 had abolished two of its three secretariats, leaving only the Secretariat of State Affairs, to which the six ministries were subordinate in Chin times. It is doubtful whether this reliance on one rather than three secretariats signified a centralization of government in Yuan times. Many other factors conspired against centralization: notably, the tendency of the military bureaucracy to encroach on the civilian sphere, the existence of semi-independent appanages outside regular government control, and the wide latitude taken by regional and local officials despite the pyramidal structure of communication and control. 1981), pp. 25—55; P a u ' Ratchnevsky, Un code da Yuan, vol. i (Paris, 1937); and the "Monograph on officials" (Pai kuan cbih), in YS, chaps. 85-92. 3 YS, 85, p. 2121. See Charles O. Hucker, A dictionary of official titles in imperial China (Stanford, 1985), pp. 2 8 - 3 1 , 4 0 - 4 , 5 5 - 7 . Cambridge Histories Online © Cambridge University Press, 2008 GOVERNMENT 589 The top official in the Central Secretariat was the chung-shu ling, in Khubilai's reign a post assumed by the heir apparent. Because the chung-shu ling was most often left vacant throughout the Yuan, the next two subordinate officials, the councillor of the right (yu ch'eng-hsiang) and the councillor of the left (tso ch'eng-hsiang), were in effect the highest civil officials in the empire. They in turn had direct control over the six ministries, the ministries of Personnel (Li-pu), Revenue (Hu-pu), Rites (Li-pu), War (Ping-pu), Punishments (Hsing-pu), and Works (Kung-pu.) Of the six ministries, all formally established under Khubilai, the Ministry of Personnel was arguably the most influential, by virtue of its power to appoint civilian officials throughout the empire. Regional and local officials, the only civilian officials with whom commoners might have had direct contact, were regularly evaluated by the Ministry of Personnel for promotion, demotion, and transfer once in office. Such appointed officials were supposed to serve terms of either thirty lunar months (if they served in the capital) or three years (if they served in the provinces), but in reality, Yiian regulations refer frequently to cases of excessively long tenure in office. The Ministry of Revenue was charged with overseeing population censuses, taxation records, state treasuries, currency, and government manufacturing. One of this ministry's most important duties was enforcing the numerous and elaborate Yiian regulations concerning paper currency. Because the Yiian government was committed to the empirewide circulation of paper notes, the procedures necessary for printing and administering paper currency were extensive. The government's deep concern is suggested by the fact that counterfeiting paper money was punishable by death. 4 In terms of political and economic authority, the jurisdiction of the Ministry of Rites was far more narrowly defined than that of either the Ministry of Personnel or the Ministry of Revenue. Court ceremonies, music, assemblies, and sacrifices came under its aegis, as did such matters as granting posthumous titles, provisioning the imperial kitchen, and manufacturing the imperial seals. The authority of the Ministry of Rites did, however, extend beyond the limited sphere of court etiquette into the realm of sumptuary regulations, marriage rites, mourning rites, and burial rites, all of which affected commoners to a certain degree. In addition, the ministry upheld the rights of the different ethnic groups in Yiian China to practice their own particular rituals and not to have to conform to Chinese standards. Uighurs, for example, were 4 See the regulations on counterfeiting in Ta Yiian sheng cheng kuo ch'ao tien chang (facsimile repr. of rev. and expanded 1303 ed., Taipei, 1972), chap. 20 (hereafter cited as YTC). On Yuan paper currency, see Lien-sheng Yang, Money and credit in China: A short history (Cambridge, 1952; repr. 1971), pp. 62—6; a more extensive treatment appears in Herbert Franke, Geld und Wirtschaft in China unter der Mongolenherrschaft: Beitrage zur Wirtschaftsgeschichte der Yiiati-Zeit (Leipzig, 1949), pp. 34—106. Cambridge Histories Online © Cambridge University Press, 2008 590 T H E YOAN GOVERNMENT AND SOCIETY directed to conduct their mourning in accordance with their own regulations; if they were to ignore their own mourning customs and follow instead Chinese practices, they would be subject to confiscation of their property.5 The Ministry of Rites was also charged with administering the state schools and regulating religious establishments.6 Of the six ministries, the Ministry of War was the least significant, as the real military authority in Yuan times resided in the Privy Council (Shu-mi yuan). The Privy Council, established in 1263, was at the pinnacle of a separate military bureaucracy, whereas the Ministry of War was subordinate to the Central Secretariat within the civilian bureaucracy. The insignificance of the Ministry of War is demonstrated by the fact that the Ping chih (Monograph on the military) in the Yuan shih (Official history of the Yuan) does not even mention the Ministry of War in its description of the structure of the Yuan military, instead stating that "the Privy Council was set up to take overall charge."7 All military offices, including the imperial guard (su-wei), were ultimately responsible to the Privy Council in the military chain of command. The main duties of the Ministry of War were to manage the population rosters of military colonies and postal personnel, manage the requisitioning of animals for military purposes, and oversee the provisioning of postal relay stations. By 1320, however, the Ministry of War had relinquished its jurisdiction over the postal relay stations to the Bureau of Transmission (T'ungcheng yuan), which had been created in 1276 separate from the military bureaucracy to supervise the postal relay system. All in all, the powerlessness of the Ministry of War reflects the Mongols' discomfort at having their military establishment subordinate to a civil branch of government. By investing power in the Privy Council the Yuan rulers were able to keep military affairs separate and secret from the civilian bureaucracy. In fact, the description of the Privy Council in the Yuan shih opens with the observation that it was "charged with responsibility over military armaments and secret [military] affairs throughout the empire."8 The Ministry of Punishment's duties were drafting criminal laws, reviewing cases involving capital punishment, and registering criminals' dependents and confiscated goods. In comparison with that of earlier dynasties, the Yuan Ministry of Punishments gained in importance because it took over the 5 YTC, 2 9 - 3 0 . For regulations governing Uighur mourning, see YTC 30, pp. 8a-b. This passage is translated by Francis W. Cleaves in "Uighuric mourning regulations," Journal of Turkish Studies, 1 (1977), pp. 65—93. Th e extent of the ministry's confiscation was said to be half the property. 6 See YTC, 31—2, for its regulations on state schools; for regulations on religion, see YTC, 33. 7 YS, 9 8 , p. 2508. Chapter 98 of the YS was translated by Ch'i-ch'ing Hsiao in The military establishment of the Yuan dynasty (Cambridge, 1978). 8 YS, 86, p. 2155; Ratchnevsky, Vn code da Yuan, vol. 1, p. 140. Cambridge Histories Online © Cambridge University Press, 2008 GOVERNMENT 591 responsibilities for judicial review that in previous times had rested with the Grand Court of Judicial Review (Ta-li ssu). The Grand Court originated in Northern Ch'i and Sui times and functioned as the highest legal agency in the Chinese empire, but it did not exist as such in Yuan times. For a brief time, from 1283 to 1285, a Grand Court existed in name only as a temporary redesignation of the Court of Justice for Uighurs (Tu-hu fu).9 Thus, by not having a Grand Court of Judicial Review, the Ministry of Punishments resolved and implemented legal decisions, which were subject only to an occasional revision by the Central Secretariat or the emperor himself. As powerful as the Ministry of Punishments was in the Yiian legal system, its authority did not extend to legal cases involving Mongols and Uighurs. The former were adjudicated by the Grand Court of the Imperial Clan (Ta tsung-cheng fu), and cases involving Uighurs and other Western and Central Asians fell under the jurisdiction of the Court of Justice for Uighurs. Members of different ethnic groups were to be tried and punished in accordance with their own laws and customs. 10 Thus, Mongolian judges known as jarghucbi (transcribed as cha-lu-huo-ch'ih and translated into Chinese as tuanshih kuari), under the aegis of the Grand Court of the Imperial Clan, resolved legal disputes involving Mongols. Cases involving members of more than one ethnic group were decided by a mixed board consisting of Chinese and Mongols. After 1328, such cases were handled by the Court of the Imperial Clan. The sixth ministry, the Ministry of Works, supervised government workshops, the repair of fortifications, the assignment and labor of government artisans, the evaluation of artisan officials, and the conscription of laborers for government projects. The very existence of the six ministries, which had been part of the traditional Chinese central government since Sui and T'ang times, tends to create the impression of a thoroughly sinicized bureaucratic administration in Yiian China. The actual functioning of these ministries, however, reflects how Mongolian priorities and policies could reshape and redirect those institutions. The hollowness of the Ministry of War and the limits imposed on the jurisdictions of the Ministry of Punishments, for instance, are examples of Mongolian adaptation. Mongolian regional and local governance reflects further divergence from standard Chinese notions of regional and local governance. The Mongols 9 On the Grand Court in pre-Yiian times, see Hucker, A dictionary of official titles in imperial China, p. 468. On the Yiian Ministry of Punishments, see Paul Heng-chao Ch'en, Chinese legal tradition under the Mongols: The code of 1291 as reconstructed (Princeton, 1979), pp. 78—9. 10 For examples of differing application of laws according to ethnic groups, see Ch'en, Chinese legal tradition under the Mongols, pp. 82—4. The name of the Ta tsung-cheng fu (Grand Court of the Imperial Clan) was abbreviated to Tsung-cheng fu during the reign of Emperor Jen-tsung (r. 1311 —20). Cambridge Histories Online © Cambridge University Press, 2008 59 2 THE Yt)AN GOVERNMENT AND SOCIETY extensively used branch, or regional, administrations, attaching the prefix hsing to such central government agencies as the Central Secretariat and the Privy Council to create Branch Secretariats (Hsing chung-shu sheng or hsingsheng) and Branch Privy Councils (Hsing shu-mi yuan). The Mongols were not the first Inner Asian people to rely on branch administrations to govern the empire. The term hsing t'ai, or regional administration, was first employed by the Wei dynasty of the Three Kingdoms period (A.D. 220-64) t o refer to predominantly military local administrations established on a temporary basis." Most important as a precedent for the Yuan regional administrations were the Branch Secretariats of State Affairs (Hsing t'ai shang-shu sheng), which the Jurchen Chin dynasty set up to oversee both military and civil affairs in the provinces." In Yuan times, the branch administrations were of two types: permanent and temporary.'3 The Central Secretariat and the Censorate exerted bureaucratic authority down to the local level through permanently established Branch Secretariats and Branch Censorates. Only in the event of a military emergency would the Privy Council establish Branch Privy Councils, all of which would be abolished once the emergency had passed. A few other examples of temporary branch administrations can be found in the early Yuan period when military and civil authorities were still preoccupied with consolidating their power. In 1263, for example, an imperial decree ordered a Branch Ministry of Finance (Hsing hu-pu) to be established in Shan-hsi to assess and collect that region's taxes.14 In 1276, another Branch Ministry of Finance was established in Ta-ming Prefecture in north China to print paper money for circulation in Chiang-nan, presumably after the conquest of South China had been completed.1' In 1274, a Branch Ministry of Works (Hsing kung-pu) was given jurisdiction over two thousand slaves who had deserted their masters.16 Such temporarily established regional ministries were the exception, however; Yuan sources do not indicate that the ministries of Personnel, Rites, War, or Punishments ever exerted their authority through branch or regional administrations. The permanently established Branch Secretariats (Hsing chung-shu sheng or 11 On the regional administration from the third to seventh centuries, see Aoyama Koryo, "Rekidai kodai ko," Taiboku teikoku daigaku bunsei gakubu ihigaku ia keniyu nembo, 2 (1933), pp. 143—66. 12 Jing-shen Tao, "The influence of Jurchen rule on Chinese political institutions," Journal of Asian Studies, 30 (1970), pp. 121—30; and Jing-shen Tao, The Jurchen in twelfth-century China: A study of sinidzation (Seattle, 1977), pp. 35-6, 4 3 - 4 . 13 The following description relies on YS, 91; Ratchnevsky, (Jn code des Yuan, vol. 1; and Elizabeth Endicott-West, Mongolian rule in China: Local administration in the Yiian dynasty (Cambridge, Mass., 1989)14 YS, 5, p. 90. 15 YS, 9, p. 183; 157, p. 3697. 16 YS, 8, p. 158. Cambridge Histories Online © Cambridge University Press, 2008 GOVERNMENT 593 hsing-sheng), which numbered eleven in all, were formally established during Khubilai's reign in order to manage the affairs of lesser territorialadministrative units, to pacify frontier areas, to manage the transport of grain, and to take overall charge of military and civil affairs at the regional level. They were directly answerable to the Central Secretariat in terms of the structure of communication and control. Initially, in the decades before Khubilai's reign, when the Mongols were engaged in pacifying north China, persons charged with both civil and military responsibilities were often designated as hsing-sbeng, but their actual duties had not yet been fixed by statute. In the pre-Khubilai period, such official designations as hsing-sheng were at times used interchangeably with darughachi (Ta-lu-hua-ch'ih) and liushou. Despite the consolidation of separate civil and military bureaucracies under Khubilai, civil and military jurisdictions were united at the level of the Branch Secretariats. The Branch Secretariats held authority over most garrison troops stationed throughout the empire, except in dire emergencies when Branch Privy Councils were temporarily established. At least on two occasions (in 1287 and 1309), the Hsing chung-shu sheng (Branch Secretariats) were briefly redesignated Hsing shang-shu sheng (Branch Secretariats of State Affairs), later to be reinstated with their original names. This sort of redesignation is typical of Yiian institutional history. As far as we can determine, such changes in office names were usually not accompanied by changes in actual functions. Some historians have stressed the independence and autonomy of the Branch Secretariats vis-a-vis the Central Secretariat, but there seems to be little in Yiian sources to substantiate that view. In its description of the duties and offices of the Branch Secretariats, the Yiian shih alludes to their authority to appoint their own clerks, interpreters, seal keepers, messengers, and other lowly subbureaucratic personnel "according to need," but there is no evidence that hsing-sheng had any wider latitude in their power of appointment at the higher or more important levels of administrative authority. ' 7 Yiian civilian government departed from earlier patterns and precedents in Chinese governmental history in the multiplicity of its levels of submetropolitan government and in the sheer number of civilian officials staffing those units of government. Thus, the levels of government subordinate to the Branch Secretariats were (in descending order): circuit (tao), route (/»), prefecture (san-fu or fu), subprefecture {chou), county (hsien), and special districts under the jurisdiction of lu or fu called lu-shih ssu. Not every unit was necessarily present on every level in the administrative hierarchy. In 17 David Farquhar described the Branch Secretariats as "separate vassal states." See his "Structure and function in the Yiian imperial government," p. 52. Cambridge Histories Online © Cambridge University Press, 2008 594 T H E YOAN GOVERNMENT AND SOCIETY other words, eight of the eleven Branch Secretariats directly administered prefectures that were not subordinate to an intermediate route. In addition to administering seven routes, the Branch Secretariat of Kan-su (Kan-su tengch'u hsing chung-shu sheng) also directly administered two subprefectures. The units of government subordinate to the Branch Secretariat of Korea (Kory6), the so-called Invade the East Branch Secretariat (Cheng tung tengch'u hsing chung-shu sheng), were somewhat different because the king of Korea in Yuan times served as the Secretariat's chief councillor and was authorized to select his own subordinates.18 For a brief time in the year 1281, the same year as their second unsuccessful invasion of Japan, the Mongols established a Branch Secretariat of Japan (Jih-pen hsing chung-shu sheng).'» Needless to say, this ephemeral secretariat had no staff and no authority based in Japan; its establishment reflected only the Mongols' optimism concerning eastward expansion, and their intention to use Korea as a base for invading Japan. The Branch Secretariat of Ling-pei, which governed the Mongols' homeland, also was organized differently. Despite Ling-pei's great size, it lacked a differentiated substructure. The Ling-pei Branch Secretariat governed one route, the Ho-ning lu, but no additional administrative units came under its aegis. Below the Branch Secretariat, the next unit of government was the circuit (fad). Those circuits administered by Pacification Offices (Hsiian-wei ssu) were particularly important as coordinators of civil and military affairs at the regional level. The Pacification Offices themselves handled military affairs in frontier areas and supervised troop movement and provisioning at the local level. The circuit as a territorial unit of government was also a part of the Censorate's field operations. Surveillance Bureaus (Su-cheng Lien-fang ssu) of the Censorate were established in the circuit, whose number over the years increased from eight in 1277 to twenty-two after 1299. It is not entirely clear whether the geographical boundaries of those circuits administered by the Branch Secretariats and the Pacification Offices were identical to the geographical boundaries of circuits as units of the Censorate's Surveillance Bureaus.20 The next administrative level below the circuit was the route (lu). Routes 18 See YS, n , pp. 2 3 1 , 236. 19 On Korea in Yuan times, see William £. Henthorn, Korea: The Mongol invasions (Leiden, 1963). On the "Invade the East" Branch Secretariat, see Ikeuchi Hiroshi, Man—Sen sbi kenkyu: chusei, vol. 3 (Tokyo, 1963), pp. 119—74. Ikeuchi points out that it is also called the "Invade Japan" Branch Secretariat (Cheng Jih-pen hsing-sheng) in the Yuan shih. 20 On the Pacification Offices and circuits (too), see YS, 9 1 , p. 2308; Ratchnevsky, Un coaedes Yuan, vol. i , p. 9 3 , n. 1; and Yeh Tzu-ch'i, Ts'ao mu tzu (Peking, 1959; repr. 1984) 3, p. 64. On the Surveillance Bureaus and the circuits, see YS, 86, pp. 2 1 8 0 - 2 ; Ratchnevsky, Un coaedes Yuan, vol. 1, pp. 169—70, 179; and Farquhar, "Structure and function in the Yuan imperial government," p. 34. Cambridge Histories Online © Cambridge University Press, 2008 GOVERNMENT 595 were classified as either "upper" or "lower" depending on population or strategic position. In similar fashion, subprefectures and counties were classified as upper, middle, or lower, depending on population, whereas prefectures were not differentiated. The Yuan shih notes that the prefectures' duties of encouraging agriculture and supervising the ao-lu (a'urugh), or military households, were the same as those of the routes. A similar feature existed at all levels below the Branch Secretariat. For each, a darughachi was appointed at the same rank and with the same salary and amount of office land as the other principal official. For example, the head official of a county, the magistrate (hsien-yin), was allotted the same salary and amount of office land as was the county darughachi, and the two officials were equal in rank. Although the initial impulse behind this system of dual staffing may have been inspired by an occupation mentality, the darughachi of the civilian bureaucracy from Khubilai's reign onward enjoyed few exclusive privileges or prerogatives. The Yuan regulations concerning carrying weapons contain one of the few indications that civil darughachi were given preferential treatment over their civilian government counterparts. An imperial decree issued in 1263, early in Khubilai's reign, specifically grants permission to Mongols, Uighurs, Muslims, ortogh merchants, hunting households, police, and darughachi to bear arms (bows and arrows).21 Yet a Yuan scholar-official, Wang Yiin (1227— 1304), in his impassioned essay on Yuan arms control, singles out only military households, ortogh merchants, police, hunters, and Muslims as groups allowed to carry bows and arrows. Moreover, he states that provincial civil officials of the third rank and below, who were prohibited from carrying weapons, were defenseless in the face of armed mounted highwaymen.22 Darughachi are notably absent from Wang Yiin's list of privileged armsbearing groups. The highest rank that a civilian darughachi could obtain was Rank 3a (at the level of an upper route), thus reinforcing Wang Yiin's observation that all nonmetropolitan officials below Rank 2b were subject to the same regulations regarding carrying arms. In other Yuan regulations concerning weapons, civilian darughachi, along with Western and Central Asian officials, are given permission — but only if there were no available Mongolian military officials - to enter military arsenals for supervisory purposes. 2 ' Like their colleagues' duties, the daily duties of the darughachi consisted mainly of supervising the work of other bureaucrats, rather than engaging in such tasks as tax collection, which would have brought them in direct contact with local populations. Even in terms of ethnic categories, the 21 YTC, 35, pp. 2b—3a. 22 Wang Yiin, Ch'iu-chim hsien sbeng la ch'iian wen chi (SPTK ed.), 84, pp. 6b—7a. 23 YTC, 35, pp. 3a-b. Cambridge Histories Online © Cambridge University Press, 2008 596 THE YOAN GOVERNMENT AND SOCIETY darughachi could not always be distinguished from other chief officials of regional-local government. Khubilai's imperial decrees reserving the office of darughachi for Mongols (or Western and Central Asians if there were no Mongols) proved difficult to enforce, and the Mongolian office of darughachi was sometimes filled by Chinese. Because of their limited population, Mongols with the training and skills to serve as local darughachi were in short supply. The local gazetteers of the Yuan period attest the wide range of nationalities and religions among darughachi - Mongols, Muslims, Uighurs, Nestorian Christians, Chinese, Jurchens, Qipchaq, Qangli, Tanguts, and others. In a very real sense, the office of darughachi epitomizes the evolution of Mongolian rule in China over the course of the Yuan dynasty. Though forced by the constraints of ruling a sedentary society to alter their institutions, the Mongols nonetheless preserved those same institutions in one form or another until the end of the dynasty. Under the Yuan government, the number and power of subbureaucratic personnel grew tremendously. Because civil service examinations, instituted only in 1313, did not provide a significant path for career advancement in the Yuan, many people with scholarly training, and some without, worked their way up through unranked or low-ranking positions in the clerical subbureaucracy to attain, at a relatively advanced age, a ranked, salaried position as a local official.24 This process was different from the Ch'ing dynasty process, in which private secretaries (mu-yu), often with advanced examination degrees, would remain as secretaries for long periods while awaiting a suitable opening in the Ch'ing bureaucracy. Unlike the Ch'ing mu-yu, Yuan clerks did not have examination degrees, and their advancement was subject to regularly scheduled evaluations at intervals of thirty or forty months. For instance, after serving for thirty months as a county clerk (hsien ssu-li) and passing an evaluation, a person could be promoted to clerk in a prefecture or a subprefecture for another thirty-month term. If he passed that evaluation, he could be appointed clerk at the route level. A route clerk had to weather three more evaluations, or a total of ninety months, in that office before being considered for a position in the regular bureaucracy. Altogether, on the way to reaching a low-ranked position in the local government bureaucracy (i.e., a position not appreciably higher than the highest clerkship), a person who worked his way up would have spent twelve and a half years as a clerk. However, there was no guarantee of promotion after a thirty-month 24 The following relies on chap. 5 of Endicott-West, Mongolian rule in China; and Makino Shuji, Gendai kotokan no taikeiuki kenkyu (Tokyo, 1979), pp. 5 - 7 3 , esp. pp. 6 5 - 6 . Cambridge Histories Online © Cambridge University Press, 2008 GOVERNMENT 597 term; clerks were often transferred from one specialized domain (e.g., a granary clerk or jail clerk) to another at the same level. Contemporary critics of the Yuan government noted that clerks were transferred from local to central government posts and then back again without any great concern for continuity in such posts. 2 ' Becoming even a county clerk was not automatic. At the age of fourteen or fifteen, a person would most likely begin as a lowly writer's assistant (t'iehshu) and could remain an unsalaried writer's assistant for ten years. Thus this person might attain the position of county clerk in his mid-twenties and, if he were lucky, the position of route clerk by the age of forty. Yuan clerks were much vilified by Chinese scholars, who were quick to pin the blame for the bureaucracy's multifarious malfunctioning on them. Though they certainly were literate, Yuan clerks usually had had limited classical textual learning and were adept at substituting legal professionalism for ethical considerations. Therefore they were contemptible in the eyes of Chinese scholars, whose Confucian training was not readily converted into an official position. The new legal professionalism, which distinguished Yuan from earlier clerks, undoubtedly better prepared them for the demands of holding office. Conversely, however, the legal knowledge and skills for which the clerks were renowned allowed them to exert far greater control over government affairs on a day-to-day level than was ever before possible. Many exhortations urged local government officials to reestablish control over their own clerks. 26 Specialized government agencies The Grand Bureau of Agriculture (Ta Ssu-nung ssu), the Directorate of Waterways (Tu-shui chien), and the maritime trade and ortogh merchant-control administrations were fiscal agencies outside the streamlined structure of civilian government, but they were equally important and more specialized. Even before a formal agriculture bureau had been created, Khubilai in the first year of his reign (1260) ordered pacification commissioners {hsiian-fu shihi) throughout China to select people skilled in agriculture to serve as "agriculture-encouragement officials."27 The Yuan shih praises Khubilai for 25 Hu Chih-yu, Tzu-shan ta cb'iian chi (SKCS ed.), 22, p. 31a. 26 See Ch'en, Chinese legal tradition under the Mongols, pp. 8 8 - 9 8 ; and John D. Langlois, Jr., "Political thought in Chin-hua under Mongol rule," in China under Mongol rule, ed. John D. Langlois, Jr., (Princeton, 1981), pp. 184—5. For example, see Chang Yang-hao (1270-1329), Mu minchungkao, in Wei cheng chung kao 'SPTK ed.), pp. i o b - n b , 13a. 27 For agricultural agencies, see YS, 87, pp. 2 1 8 8 - 9 ; 93. PP- 2 3 5 4 - 7 ; see also Ratchnevsky, Uncodedes Yuan, vol. 1, pp. 188—91; and Herbert Franz Schurmann, Economic structure of the Yuan dynasty: Translation of chapters 93 and 94 of the Yiian shih (Cambridge, 1956; repr. 1967), pp. 4 3 - 6 4 . Cambridge Histories Online © Cambridge University Press, 2008 598 T H E YOAN GOVERNMENT AND SOCIETY promptly turning his attention to China's economic foundation, asking rhetorically, "How indeed can the Liao and Chin [rulers] be compared [with Khubilai]?" The eastern Mongols' commitment to remaining in China is reflected not only in Khubilai's shifting the capital from Mongolia to China in 1260 but also in the numerous efforts from 1260 onward to make China's agricultural economy prosper. In 1261 the first bureau to supervise agriculture, the Agriculture-Encouragement Bureau (Ch'iian-nung ssu), was created, to be superseded in 1270 by the Bureau of Agriculture (Ssu-nung ssu). The Bureau of Agriculture was charged with managing all affairs pertaining to agriculture, sericulture, and irrigation. Special agriculture and irrigation experts were dispatched to investigate and report back on local officials' successes and failures in promoting agriculture. In fact, records of such activities were taken into account at their end-of-term evaluations. Later, in 1270-1, the Bureau of Agriculture was renamed the Grand Bureau of Agriculture (Ta ssu-nung ssu), and in spite of three more name changes (Nung-cheng yuan, Wu-nung ssu, and Ssu-nung ssu), it remained the Grand Bureau of Agriculture from 1286 on.28 A decade after the conquest of the Southern Sung, Regional Branches of the Grand Bureau of Agriculture (Hsing Ta ssu-nung ssu) were established in Chiang-nan (south China) in 1288. In 1295 under Temiir Khaghan (Ch'engtsung), however, they were abolished, but this does not necessarily mean that post-Khubilai Yuan emperors deemphasized agriculture. Local officials, especially darugbachi, continued to be held responsible for promoting and protecting agriculture in local society. One Yuan source attributes the initial creation of the Regional Branches of the Grand Bureau of Agriculture in south China to an attempt to investigate powerful families who were hiding their agricultural assets from taxation. According to this source, because such cases of concealment were found not to be numerous, the regional branches were abolished in 1295.29 There is no question that the government's involvement in, and supervision of, agriculture began under Khubilai. Not only did Khubilai's reign see the creation of agriculture-supervising agencies, but his Bureau of Agriculture also printed and distributed an agricultural handbook called Nung sang chi yao, which an early Yuan encyclopedia states was printed in order that 28 Schurmann, Economic structure ofthi Yuan dynasty, p. 48, writes that the Grand Bureau of Agriculture was abolished in 1290 and never again reestablished. I see no direct evidence that it was abolished then. In fact, YS, 87, p. 2188, has several post-1290 references to the bureau's officials; YS, 17, p. 372, refers to the Chiang-nan Branch Grand Bureau of Agriculture in 1293; and YS, 43, p. 908, refers to the appointment of T'o-t'o (Toghto) as head of the Grand Bureau of Agriculture in 1353. 29 Ta Yuan kuan chih tsa chi, in vol. 9 of Hsiieb shu ts'ungpitn, ed. Chi-Fo-t'o (Taipei, 1971), pp. 5a-b. See also Farquhar, "Structure and function in the Yiian imperial government," pp. 4 1 - 2 . Cambridge Histories Online © Cambridge University Press, 2008 GOVERNMENT 599 "the people of the empire would read its contents and acquire its skills." 30 The handbook's original preface by Wang P'an, a Han-lin academician, admits that the Nung sang chi yao was compiled from the contents of previous agricultural handbooks; in other words, the book's information about agricultural techniques was not new. Rather, the purpose of the handbook, according to Wang P'an, was to educate government agricultural officials.3' In addition to its publishing effort, the Yuan government in the early 1270s established agricultural communities {she), utilizing an earlier form of local social organization. The unsalaried appointees of the she were to encourage agricultural production, maintain charity granaries (i-ts'ang), supervise taxation and corvee obligations, and set the proper moral tone in their units. The she unit itself was to consist of fifty households, but whether these units were in fact superimposed on all of China's villages in the thirteenth century is open to question. The creation of a state institution in local society is, however, entirely consistent with Khubilai's thirty-four-year effort to revitalize the Chinese economy after decades of watfare. The Directorate of Waterways, like the Grand Bureau of Agriculture, was a specialized civilian agency, charged with overseeing canals, dikes, irrigation systems, bridges, and locks.32 First instituted in 1291, the Directorate of Waterways was given jurisdiction over the Waterways Supervisorate (Ho tao t'i-chii ssu) in 1292, thus centralizing all matters concerning interior waterways. Not surprisingly, the government established Branch (Hsing-) Directorates of Waterways to manage local problems and natural disasters. Similar to the Grand Bureau of Agriculture and the Directorate of Waterways were those agencies set up to regulate maritime trade and, in particular, to regulate the activities of the ortogh (Muslim partnership) merchants. 33 Before the Mongols conquered south China, no such supervisory agencies had been established. It was only during Khubilai's reign that the Yuan government, in order to encourage foreign trade and thereby profit from it through maritime trade taxes, reestablished the Maritime Trade Bureau (Shih-po ssu). In doing so, the Yuan followed Sung precedent. The Sung in 1087 had set up a Maritime Trade Bureau in Ch'iian-chou, an important port city on the Fukien coast, and the Yuan established its first bureau there in 1277. The number of Maritime Trade Bureaus rose to seven by 1293, as Khubilai's 3 0 Ch'en Yiian-ching, c o m p . , Shih lin kuang chi (Chien-an, 1330—3; repr. Peking, 1963), 12, p . ia. 31 Nung sang chi yao(SPPY ed.), p. ia. 32 Y5, 90, pp. 2295—6; Ratchnevsky, Un code des Yuan, vol. 1, p. 267, no. 2; Hucker, A dictionary of official titles in imperial China, p. 542; and Farquhar, "Structure and function in the Yuan imperial government," pp. 4 2 - 3 . 33 On maritime trade, see Schurmann, Economic structure of the Yuan dynasty, pp. 2 2 2 - 3 6 . Sources on the ortogh merchants and attempts to regulate their activities are cited in Elizabeth Endicott-West, "Merchant associations in Yuan China: The ortogh," Asia Major, 3rd. series, vol. 2, pt. 2 (1989), pp. 127-54- Cambridge Histories Online © Cambridge University Press, 2008 600 T H E YOAN GOVERNMENT AND SOCIETY financial advisers at the court sought to fill the government treasury through percentage levies on cargoes and trade taxes. Jurisdiction over the Maritime Trade Bureaus shifted: At one time they were put under the local salt administration offices, and at another point they were placed under the main mercantile agency, called the Supervising Money Bureau (Ch'iian-fu ssu). The history of overseas trade in Yuan times is closely tied to the history of those foreign, mainly Muslim, merchants whose trade partnerships with the Mongolian imperial family and government officials were referred to as wo-t'o (Mongolian: o[r]to[gh], from the Turkic ortaq; also Persian iirtaq, derived from the Turkic ortaq). There even was an unsuccessful attempt in 1286 by one of Khubilai's advisers, Lu Shih-jung, to ban all private foreign trade in order to let the government and the ortogh merchants monopolize overseas trade.34 The Supervising Money Bureau was not the first Yuan agency set up to regulate ortogh merchants' activities. In 1268 the General Administration for the Supervision of the Ortogh (Wo-t'o tsung-kuan fu) was first established. Regional offices (Wo-t'o so or Wo-t'o chii) were established as early as 1269. In 1280 the General Administration was replaced by the Supervising Money Bureau, which oversaw the entrusting of imperial funds to ortogh merchants. Such funds were used by the 'merchants for moneylending, for financing overland trade caravans to Western Asia, and for maritime trade. Thus, the fact that the Supervising Money Bureau was given jurisdiction over the Maritime Trade Bureaus in 1286 signifies the overwhelmingly important role that ortogh merchants played in maritime trade, as contrasted with the lesser role of private merchants. Ortogh merchants accordingly developed an unpopular image among Chinese literati. The military Scholars have long debated the degree of militarization in Yuan government and society. It is clear that separate civil and military bureaucracies existed, though there also is evidence that military officials did not always refrain from meddling in civilian affairs, and vice versa. Although it may be tempting to describe an essentially Chinese civilian bureaucracy and an essentially Mongolian military, our discussion of Mongolian innovations in, and alterations of, civilian bureaucracy makes such a clear-cut dichotomy less probable.35 Another issue is whether the military establishment was purely Mongolian. 34 YTC, 22, p. 47a. 35 Herberc F. Schurmann wrote about "the duality of Chinese bureaucracy and Mongol military" in his article "Problems of political organization during the Yuan dynasty," Trudy XXV Mezhdunarodnogo kongrtssa vostokovedov, 5 (Moscow, 1963), pp. 26—30; esp. p. 27. Cambridge Histories Online © Cambridge University Press, 2008 GOVERNMENT 6oi Even in the decades of the conquest of north China — that is, before Khubilai's reign — the Mongols, by virtue of their limited numbers, found it necessary to absorb foreign troops into their ranks. Central Asian Turks, Jurchen defectors from the faltering Chin dynasty, Khitans eager to oppose the Chin, and Chinese conscripts all contributed to the defeat of the Chin. Ethnic differentiation was the rule in Yuan military organization. The so-called Han armies (Han chiin) consisted of northern Chinese, Khitans, Jurchens, and Koreans, whereas the Mongolian army (Meng-ku chiin) and the tammachi (t'an-ma-ch'ih chiin) consisted of Mongols. Inhabitants of the conquered Southern Sung were organized into the newly adhered armies (hsin-fu chiin). The elite imperial guard (su-wei), organized under Khubilai to augment the kesig (the Mongolian aristocratic guard first instituted by Chinggis), was, however, ethnically mixed. Half of its units consisted of Western and Central Asians and Mongols, and half were Chinese guardsmen. By the early fourteenth century, the ethnic composition of the imperial guard had already been affected by its very appeal. Lured by financial and status rewards, a great number of Chinese commoners entered the kesig, in spite of the Yuan government's attempts to maintain the kesig as a reservoir of Inner Asian military strength. In regard to administrative organization, the units of the imperial guard were under the jurisdiction of the Privy Council (Shu-mi yuan), which was at the apex of the separate military bureaucracy. The council did not, however, exert direct control over garrison troops stationed in the Branch Secretariats outside the metropolitan province. The myriarchies (wan-hu fu, M: ttimeri) from which the garrison troops were drawn were answerable to the Branch Secretariats, which of course were territorial administrations in the civilian bureaucracy. This meshing of civil and military authority at the regional level was apparently aimed at facilitating cooperation between the two. Nevertheless, as we mentioned earlier, in dire military emergencies, as in the case of insurrections against the dynasty, a temporary Branch Privy Council would be established until the emergency had passed. Civil and military jurisdictions overlapped in their administration of the agricultural colonies that supported troops (t'un-t'ien). Some of these lands and their households were administered directly by the Privy Council; some were run by the Grand Bureau of Agriculture; some were directed by the Court of Imperial Etiquette (Hsiian-hui yuan), to provide special products for the palace and other government offices; others were under the Central Secretariat; and still others were run by the Branch Secretariats.36 36 See YS, 100, pp. 2 5 5 8 - 7 9 ; YS, 87, p. 2204; Farquhar, "Structure and function in the Yuan imperial government," p. 50; and Hsiao, The military establishment of the Yuan dynasty, p. 177. Cambridge Histories Online © Cambridge University Press, 2008 602 THE YOAN GOVERNMENT AND SOCIETY The civil and military bureaucracies also shared jurisdiction over the a'urugh (ao-lu), or military households. Administration of the a'urugh households would naturally seem to be part of the military bureaucracy: Conscripting soldiers and collecting provisions were military affairs. Nonetheless, in 1268 the jurisdiction over these military families was handed over to local civil officials, and the special military offices administering the a'urugh were abolished. Civil officials then proceeded to ignore the a'urugh households' special exemptions and levied taxes on those living within their jurisdictions. The Privy Council, in what must be seen as a civil-military jurisdictional dispute, thereupon suggested that those local civil officials with authority over a'urugh households be made subordinate to the Council and even subject to its civilian ratings for demotion and promotion.37 Both the civilian and military bureaucracies had in common the Mongolian emphasis on inheriting offices. One of the main avenues to civilian office in Yuan times was hereditary privilege (yin-pu), that is, the privilege granted to high-ranking officials by which they could nominate their sons and grandsons for civilian office.38 The sons of Yuan military officers - mainly Mongols and Western and Central Asians — could inherit their fathers' offices upon their death, retirement, or even promotion. The traditional Mongolian emphasis on loyalty as the primary virtue made inheritance of military office in the eyes of the Mongolian rulers a desirable method of ensuring the officers' reliability. Of course, this emphasis on loyalty eliminated any systematic search for competence and talent, as evidenced by the deterioration of the officer corps as early as the turn of the fourteenth century. Thus, by the 1260s, even though members of both the civil and military bureaucracies benefited from fixed salaries, traditional Mongolian values such as loyalty and the belief in hereditary principles determined the evolution of those bureaucracies in China. The Censorate Members of both the civil and military bureaucracies were subject to the scrutiny of the Censorate, the third of the three major divisions in Yuan government.39 The Yuan Censorate differed from earlier dynasties' censorial 37 See Hsiao, The military establishment of the Ytian dynasty, pp. 8 i , 193, 135-6; and Endicott-West, Mongolian rule in China, chap. 2. 38 On the use of the yin privilege in Yuan China, see Elizabeth Endicott-West, "Hereditary privilege in the Yuan dynasty,"Journal of Turkish Studies (Festschrift for Francis W. Cleaves), 9 (1986), pp. 1 5 - 2 0 ; on inheritance of office in the military bureaucracy, see Hsiao, The military establishment of the Yuan dynasty, pp. 2 5 - 7 . 39 The following section on the Yuan Censorate is based on YS, 86, pp. 2177-82; Charles O. Hucker, "The Yuan contribution to censorial history," Bulletin of the Institute of History and Philology, extra vol. 4 ( i 9 6 0 ) , pp. 219—27; Charles 0 . Hucker, The censorial system of Ming China (Stanford, 1966), pp. Cambridge Histories Online © Cambridge University Press, 2008 GOVERNMENT 603 establishments in its greater scope of activity and greater number of officials. The increase in numbers of officials and their more extensive activities are related to the proliferation of organizational levels within the Censorate itself. The Censorate simply mirrored the overall trend in Yuan regionallocal government toward the multiplication of jurisdictional levels and the consequent rise in the number of officials necessary to staff the more elaborate structure. The Mongolian Censorate in Ta-tu had a staff of thirty-two investigating censors (chien-ch'a yu-shih), and the two permanently established Branch Censorates (Hsing Yu-shih t'ai) in Chiang-nan and Shan-hsi had twenty-eight and twenty investigating censors, respectively. Also, two temporary Branch Censorates were established in Ho-hsi (from the late 1270s to 1283) and Yunnan from 1290 to 1297. Although the Branch Censorates were granted surveillance jurisdiction over the eleven Branch Secretariats, up to twenty-four Surveillance Offices (at first called T'i-hsing an-ch'a ssu and later renamed Su-cheng lien-fang ssu) maintained surveillance over the lower levels of civil government. The Censorate's involvement in the government's daily activities is clearly indicated by Yuan governmental regulations that refer to bureaucratic revision (cbao-shua) or record checking (shua-chiian). Censorial officials were enlisted in the tedious activities of examining, correcting, or verifying dates, signatures, seals, and calculations in the daily records.4° The records of all government agencies from the Central Secretariat down to the county were regularly checked by censorial officials. Only those records pertaining to militarily sensitive matters were exempted from such checking; for. instance, information concerning numbers of troops and numbers of horses was considered top secret. The Yuan Censorate's range of duties was also expanded to include remonstration, an activity that had traditionally been the prerogative of specific offices of remonstration outside the Censorate. From Yuan times on, the different activities of surveillance and remonstration coexisted in the Censorate. The Yuan Censorate was a politically active institution and so was not above taking part in court intrigue and factional fighting. A politically motivated memorial of impeachment by a censor, for example, led to the dismissal of the powerful chancellor of the right, Toghto, late in 1354, whose 25—8; Hucker, A dictionary of official titles in imperial China, p. 61; and Hung Chin-fii, Yuan tai chien ch'a chih tu yen chin (Taipei, 1972). 40 See YTC, 6, pp. 14a—18b; Ratchnevsky, Un code da Yuan, vol. 1, pp. 40—1; Hucker, "The Yuan contribution to censorial history," p. 221; and Hung Chin-fu, Yuan tai chien ch'a chih t'u, Yen chiu, vol. 2, pp. 74a-77b. Cambridge Histories Online © Cambridge University Press, 2008 604 T H E YOAN GOVERNMENT AND SOCIETY fall is generally seen by historians as a disastrous turning point in the dynasty's fortunes. Although the top censorial offices in the capital were, with few exceptions, filled only by Mongols or Western and Central Asians, the lower-ranking (Rank 7a) offices of the investigating censors (chien-ch'a yii-shih) were open to Chinese (Han-jen). In fact, the first twelve posts for investigating censors in the Metropolitan Censorate were filled by Chinese in 1268. Although Mongols came to outnumber Chinese in 1282, it is still noteworthy that in 1285 two southern scholars (nan-ju) also were appointed. In the Chiang-nan Branch Censorate in 1286, fourteen new positions as investigating censor were designated for Mongols, and only four for Chinese. Yet the fact that a few Chinese literati, some with either chin-sbih degrees (after 1315) or earlier careers as government school instructors, became investigating censors demonstrates that the Censorate was far from being an exclusive preserve for Mongols and Western and Central Asians.4I Imperial agencies and bureaus Outside the tripartite framework of civilian, military, and censorial offices were a number of agencies whose duties revolved around the emperor's own establishment and those of other members of the imperial family. In fact, some categories of Yuan households were administered directly by imperial agencies and bureaus, such as the various bureaus in the imperial establishment set up to supervise artisans, and hunting and falconry households.42 Among the most important of those agencies in charge of ensuring the imperial family's well-being was the Bureau of Imperial Etiquette (Hsiianhui yuan). This bureau, whose staff was very large in Yuan times, was a Chinese-style agency with precedents in the T'ang, Sung, Liao, and Chin. First instituted under Khubilai in 1278, the Bureau of Imperial Etiquette took over the management of imperial household duties, such as provisioning the imperial kitchens, duties that in the pre-Khubilai era had been performed exclusively by the kesig, the imperial guard.4' The kesig, however, was not entirely displaced by the Chinese-style bureau, because members of the imperial guard actually served in agencies subordinate to it, such as the Court of Imperial Attendants (Shih-cheng fu). 41 For examples ofsuch Chinese literati, see Li Tse-fen, Yiian shihbsinchiang(Taipei, 1978), vol. 4 , p. 439. 42 In his article "Structure and function in the Yiian imperial government," David Farquhar separates those agencies concerned with the emperor's well-being from those concerned with other members of the imperial household, which differs from the categorization used here. 43 On the Bureau of Imperial Etiquette and its many subordinate agencies, see YS, 87, pp. 2 2 0 0 - 1 3 ; Ratchnevsky, Un code da Yuan, vol. 1, pp. 1 4 3 - 6 ; and Hsiao, The military establishment of the Yiian dynasty, pp. 39—40. Cambridge Histories Online © Cambridge University Press, 2008 GOVERNMENT 605 This mixture of Mongolian and Chinese elements was a persistent feature of Yuan bureaucracy. We can also identify a purely Mongolian agency with no Chinese precedents, the Agency of Men and Things Gone Astray, or Lan-i chien, under the aegis of the Bureau of Imperial Etiquette. This agency was charged with overseeing displaced goods, animals, and people (mainly slaves), finding their owners, and, if that proved impossible, handing them over to the imperial household. The duties of the Agency of Men and Things Gone Astray clearly reflected Mongolian notions of property management and of what belonged within the sphere of the imperial family; on the other hand, one is not surprised to find this agency subordinate to the Bureau of Imperial Etiquette, a thoroughly Chinese institution. In addition to the many directorates that oversaw the material well-being of the imperial household, the emperor's ritual and intellectual activities were served by the Han-lin and National History Academy (Han-lin chien kuo-shih yuan) and the Mongolian Han-lin Academy (Meng-ku Han-lin yuan), among others. The joining together of the Han-lin Academy and the National History Office into one joint academy was an institutional innovation undertaken by Khubilai in 1261 on the advice of the senior Han-lin academician Wang O.44 This apparently stemmed from Wang's attempt to convince Khubilai of the need to begin compiling standard histories of the Liao and Chin as well as the records of the pre-Khubilai Mongolian rulers. In 1264, with the removal of the capital to Ta-tu (modern Peking), the Han-lin and National History Academy was formally established, and the foundations for the composition of the Liao and Chin dynastic histories were laid. In addition to the Han-lin and National History Academy, there was also a Mongolian Han-lin Academy, which drafted imperial decrees in Mongolian and translated state documents from Mongolian into Chinese and other languages, and vice versa.4' A copy of every document was made in the 'Phags-pa script, an alphabetic script based on Tibetan and written vertically, as was the Uighur script, which had been adopted by the Mongols in 1204. Uighur writing continued to be used in the Yuan along with the newer 'Phags-pa script, although an imperial decree of March 1269 ordered the 'Phags-pa script to be used in all government documents from then on. 44 See Hok-Iam Chan, "Chinese official historiography at the Yuan court: The composition of the Liao, Chin, and Sung histories," in China undtr Mongol rule, ed. John D. Langlois, Jr. (Princeton, 1981) pp. 6 2 - 4 ; Chan, "Wang O (1190—1273)," pp. 5 4 - 5 7 ; YS, 87, p. 2189; and Ratchnevsky, Vn code da Yuan, vol. 1, pp. 148—9. 45 On this office, as well as on the Mongolian schools and the use of Mongolian in Yuan official documents, see YS, 87, pp. 2190— 1; Ratchnevsky, Uncodedes Yuan, vol. 1, pp. 149—51; Lien-sheng Yang, "Marginalia to the Yuan ticn-chang," in his Excursions in sinology (Cambridge, 1969), pp. 126—8; and Endicott-West, Mongolian rule in China, chap. 3. Cambridge Histories Online © Cambridge University Press, 2008 606 T H E YOAN GOVERNMENT AND SOCIETY Officials with the right to present memorials to the throne were required to use the Mongolian script. The Mongolian Han-lin Academy was thus of central importance in the network of official communications to and from the throne, as Mongolian, not Chinese, remained the official language in Yuan China. The Mongolian Han-lin Academy also administered the Mongolian Directorate of Education (Meng-ku kuo-tzu chien) and the Mongolian National University {Meng-ku kuo-tzu bsu'eh), which educated the sons and younger brothers of both Mongolian and Chinese nobles and officials. Similarly, Mongolian language schools {Meng-ku tzu-hsiieh), established in the routes in 1269, accepted a set number of sons, grandsons, younger brothers, and nephews of all officials without reference to ethnic background. The ethnic composition of classes at the Mongolian National University in 1315 was 50 percent Mongolian, but unfortunately, we have no evidence for the routelevel Mongolian language schools. Anecdotal evidence shows, however, that Mongolian language instruction was available to Chinese seeking advancement in the Yuan bureaucracy. The Chinese equivalents of the Mongolian Directorate of Education and the Mongolian National University were administered by the Academy of Scholarly Worthies (Chi-hsien yuan).*6 The National College {Kuo-tzu chien) admitted the sons and grandsons of court officials of the seventh rank and above, whether Chinese, Mongolian, or other. Court officials of the third rank and above could recommend exceptionally gifted commoners to attend the National College as "free auditors." The Academy of Scholarly Worthies itself was detached from the Han-lin Academy in 1285 when it was assigned the additional responsibility of watching over Taoist affairs, through stateappointed Taoist administrators, at all Taoist monasteries. In its capacity as watchdog of the Taoist establishment in Yuan China, the Academy of Scholarly Worthies was not significantly different from the Bureau of Tibetan and Buddhist Affairs (Hsiian-cheng yuan), which directly supervised the Buddhist clergy in China, as well as exerting nominal authority over Tibet. Another parallel may also be drawn: The academy's supervision of Taoism by appointing Taoist "officials" in the temples was akin to the Yuan government's regulation of both the so-called public or state schools 46 The following discussion of the Academy of Scholarly Worthies, the Bureau of Tibetan and Buddhist Affairs, and the Yuan schools is based on YS, 87, pp. 2 1 9 2 - 3 ; Ratchnevsky, Un code des Yuan (Paris, 1972), vol. 2, pp. 25—6; Yan-shuan Lao, "Southern Chinese scholars and educational institutions in early Yuan: Some preliminary remarks," in China under Mongol rule, ed. John D. Langlois, Jr. (Princeton, 1981), pp. 107—33; K'o-k'uan Sun, "Yii Chi and southern Taosim during the Yuan period," in China under Mongol rule, ed. John D. Langlois, Jr. (Princeton, 1981), pp. 2 1 2 - 5 3 , es Ppp. 223—4; Herbert Franke, "Tibetans in Yuan China," in China under Mongol rule, ed. John D. Langlois, Jr. (Princeton, 1981), pp. 2 9 6 - 3 2 8 , esp. pp. 311—15; and Ruby Lam, "The role of shuyiian in Yuan China" (unpublished paper). Cambridge Histories Online © Cambridge University Press, 2008 GOVERNMENT 6oj and the private academies (shu-yuan). Public school instructors (chiao-shou), as well as headmasters {shan-chang) of privately funded academies, were appointed to, or confirmed in, their educational posts by the government. If the Yuan government had intended to co-opt religious and educational establishments by incorporating them structurally into the official bureaucracy, then it had at best a mixed success. Many of the top officials of the Bureau of Tibetan and Buddhist Affairs were Buddhist monks or laymen, which may explain why the bureau granted such generous tax exemptions to Buddhists (and, in fact, even to Taoists, Muslims, and others). Such tax exemptions caused great friction between the bureau and provincial authorities. Along the same lines, the surge in numbers of newly established private academies during the last three decades of the Yuan was undoubtedly connected to the fact that people who donated land to private academies were exempt from corv£e labor, and in the late Yuan, corvee obligations had grown increasingly burdensome. There were other specialized agencies serving the imperial family, the names of which clearly reflected their functions: for example, the Academy of Calendrical Studies (T'ai-shih yuan), the Institute of Astronomy (Ssu-t'ien chien), the Institute of Muslim Astronomy (Hui-hui Ssu-t'ien chien), and the Imperial Academy of Medicine (T'ai-i yuan). Some agencies managed the artisan households assigned to the heir apparent and the empress. Some of these artisans and their households were classified as civilian (min-chiang) and others as ch'ieh-lien-k'ou (colloquial Mongolian: ger-in k'e'ii; literary Mongolian: ger-iin kobegiid), literally, "sons of the yurt," or slave artisans. The assignment of designated hereditary craftsmen households to members of the imperial family and nobles in the Yuan was a remnant of the early Mongols' custom of transporting whole populations of skilled artisans and laborers from one locale to another for construction projects. In addition to grants of artisan households, the Mongolian rulers also granted territories (including their populations), especially in north China, to Mongolian princes, nobles, and meritorious officials. The terminology describing such grants varied, but the most significant units were the t'ou hsia, which may be translated loosely as "appanages," and the fen ti, or "apportioned territories. "•" Recipients of appanages were generally permitted to appoint, subject to imperial approval, the darughachi and the judges {jarghuchi or tuan-shih kuan) of their territories, but lesser officials were appointed through regular government channels. The struggle between the 47 The secondary literature on the appanages (t'ou hsia) is extensive. See, for example, Huang Ch'inglien, Yiian lai hu chi chih tuyen chiu (Taipei, 1977), pp. 4'—7; 209; and Endicott-West, Mongolian rule in China, chap. 4. See also Isenbike Togan, "The chapter on annual grants in the Yiian shih" (Ph.D. diss., Harvard University, 1973)- Cambridge Histories Online © Cambridge University Press, 2008 608 T H E YOAN GOVERNMENT AND SOCIETY imperial court and the imperial princes for fiscal and political control over the appanages continued throughout the Yuan and contributed to the constant instability at the very apex of the government. The dynastic history of the Yuan, the Yuan shih, devotes a chapter to annual grants (sui-tz'u), which were grants in silver and silk in north China and paper currency in south China. Most of the grant recipients were members of the imperial family or of Mongolian noble families, and the source of the grants was the Chinese population in the apportioned territories. Households in such territories generally paid taxes to both the grantees and the Yiian government in Ta-tu. We should point out that the annual grants chapter in the Yiian shih was unprecedented. No previous dynastic history included such a chapter because the practice was itself purely Inner Asian and derived from the early Mongolian custom of distributing conquered peoples or tribes as shares to members of the ruling family. As is the case with so many Yiian government institutions and practices, the Chinese official nomenclature readily peels away to reveal the Inner Asian core. SOCIETY The separation of state and society into two different realms of inquiry is a modern construct on which historians habitually rely when researching and writing about China. The thirteenth-century Mongols, however, would have recognized no such separation. Originating as a tribal, military society, the pastoral nomadic Mongols of the early and mid-thirteenth century evidenced little social stratification. Extremes in wealth are far less radical and far less easy to measure in pastoral nomadic societies than in sedentary, agricultural societies.48 For instance, the members of Chinggis khan's early retinue, the so-called nokod (plural of niikor, companion), were treated by Chinggis as his equals. As Chinggis khan's power grew, the nokod became privileged bodyguards and household attendants. In fact, they became the nucleus of the kesig or imperial guard, while at the same time they were treated by Chinggis as younger brothers. In Chinggis khan's own lifetime, the line between state and society, each mobilized for warfare, would have been difficult to draw. Mongolian society naturally had its full complement of rituals and taboos, all well documented by the Franciscans and Dominicans who visited Khara Khorum. In comparison with traditional Chinese expectations regarding the use of court ritual to define privilege and social status, however, the Mongols at the court in Ta-tu must have appeared excessively informal and, indeed, 48 For an interesting discussion, see Ernest Gellner, "Anomalies of no fixed abode," Times Literary Supplement, 13 March 1981, p. 273. Cambridge Histories Online © Cambridge University Press, 2008 SOCIETY 609 lax in attending to ritual activities. It is thus fitting to begin a discussion of Yuan society at the very top of the social scale, with a few insights into the nature of Mongolian court life. The informal atmosphere at the Mongolian court in Ta-tu is apparent in the fact that the Mongolian rulers did not observe the use of taboo names at court. 49 Chinese observers also noted the alien (in their eyes) table manners and cuisine that the Mongols preferred to Chinese cuisine. 50 For instance, the Mongols' use of small steel daggers to carve meat at mealtime was not calculated to elevate their image in the eyes of Chinese literati-officials. This is not to say, however, that the Mongols ignored all Chinese rituals. Under persistent pressure from Chinese literati, the Mongolian rulers acquiesced in carrying out Confucian rites, such as the suburban offerings to heaven, but they did not usually attend them in person. One may surmise that the imperial family felt more comfortable with the Buddhist rituals introduced by the 'Phags-pa Lama. The imperial family attended in person the Buddhist celebratory processions and plays in the first month of the New Year at which Chinese, Muslim, and Tangut musicians entertained. 5 ' The ruling Mongols also continued to practice shamanist rituals and apparently saw no conflict of interest in deriving legitimacy from more than one ideological—religious tradition. Even the way in which the Mongols lived in their imperial city of Ta-tu bespeaks a continuing adherence to the customs of the steppe. There can be no doubt that Ta-tu as an imperial city used models of Chinese architecture, yet the fact that well into the fourteenth century a number of Mongolian rulers and other members of the imperial family chose to live in tents set up in the city's imperial park, instead of in the palaces, is not without significance. The tents were placed in steppe grass that Khubilai had imported from Mongolia and planted in the imperial parks. One of the imperial parks even sported a felt pavilion, and animal skins were a part of the wall decor inside the palace. The other imperial capital, Shang-tu, was by all accounts used mainly as a game reserve and hunting park for the imperial family once the imperial city of Ta-tu had been completed.52 All these details concerning table etiquette, ceremonial, housing, and hunting suggest that the imperial family was, for the most part, less than 49 Yeh Tzu-ch'i, Tj'ao mu tzu, 3, p. 59; YTC, 28, pp. fib—7a. 50 Frederick W. Mote, "Yuan and Ming," in Food in Chinese culture: Anthropological and historical perspectives, ed. K. C. Chang (New Haven, 1977), pp. 195—257, esp. pp. 2 0 4 - 8 . 51 The best work on the role of ritual in the legitimation of the Yuan is that by Herbert Franke, From tribal chieftain to universal emperor and god: The legitimation of the Yuan dynasty (Munich, 1978), esp. pp. 32—5, 6 0 - 1 . 52 For a thorough description of Yuan Ta-tu, see Nancy Riva Shatzman Steinhardt, "Imperial architecture under Mongolian patronage: Khubilai's imperial city of Daidu" (Ph.D. diss., Harvard University, 1981). Cambridge Histories Online © Cambridge University Press, 2008 6lO THE YtJAN GOVERNMENT AND SOCIETY enthusiastic about imitating Chinese material life and was more or less indifferent to Chinese culture. There were, of course, exceptions to this. The Yuan emperor Wen-tsung (Tugh Temiir; r. 1328-32) supported scholarship and the arts, in one instance by founding the Pavilion of the Star of Literature (K'uei-chang ko) in the imperial city." The ruling Mongols also evinced great interest in the Hsiao ching (Book of filial piety), ordering it to be translated into Mongolian (the 'Phags-pa script), printed, and distributed among the imperial princes.'* Outside the imperial family itself, a small but significant number of Mongols, most with elite rather than commoner backgrounds, studied and even excelled in Chinese scholarship, literary pursuits, and the arts.5' Some of these accomplished Mongolian literati were the products of mixed marriages (Chinese mothers and Mongolian fathers) and were obviously raised in a Han Chinese cultural milieu. Even though the number of Mongols learned in Chinese culture represented only a tiny percentage of all Mongols in China, that tiny percentage was on the increase in late Yuan times. Chinese culture made little overall impact on the Mongols as a people, and conversely, Mongolian court life found little reflection in China at large. Given this peculiar situation of a self-contained nation within a nation, one is still left with the question of how Chinese society functioned and fared under foreign, nonsinicized rulership. An earlier stereotype of Chinese society in Yuan times is that the Mongols imposed a strictly enforced class system on all of society. ' 6 The four ethnic classes, in order of descending privilege, consisted of the Mongols themselves; the se-mu jen, or Western and Central Asians; the Han jen, or the various peoples in the former Jurchen Chin territory in northern China; and finally the southern Chinese {nan jen) inhabitants of the territory of the fallen Southern Sung dynasty. Various scholars over the past half-century have dispelled the notion that a sort of caste system was at work in Yuan China. It is quite easy, in fact, to find examples of Chinese holding official positions 53 See John D. Langlois, Jr., "Yii Chi and his Mongol sovereign: The scholar as apologist," Journal of Asian Studies, 38 (1978), pp. 9 9 - 1 1 6 ; and Steinhardt, "Imperial architecture under Mongolian patronage," p. 38. 54 See Francis W. Cleaves's comments on the appeal of the Hsiao ding to the Mongols in his "The first chapter of an early Mongolian version of the Hsiao ching" Acta Oritntalia Academiae Scientiarum Hungaricae, 36 (1982), pp. 6 9 - 8 8 , esp. p. 70. See also Herbert Franke, "Chinese historiography under Mongol rule: The role of history in acculturation," Mongolian Studies, 1 (1974), pp. 15—26, esp. pp. 2 2 - 4 . 55 This relies on Hsiao Ch'i-ch'ing, "Yuan tai Meng-ku jen te Han hsiieh," in Kuo chi Chung-kuo pin chiang hsiieb shu hui i lun wen chi (Proceedings of the international conference on China border area studies'), ed. Lin En-hsien (Taipei, 1985), pp. 369—428. 56 The following pages on Yuan society rely mainly on Meng Ssu-ming, Yuan tai she hui chieh chi Mb tu (Peking, 1938; repr. Hong Kong, 1967); and Hon-ming Yip, "The class system of Yuan society: A critique of Meng Siming's Yuandai sbehui jieji zhidu," Journal of Asian Culture, 4 (1980), pp. 82—106. Cambridge Histories Online © Cambridge University Press, 2008 SOCIETY 6ll (such as darughachi) that according to government regulation they were not entitled to hold. The frequent exceptions to the Yuan regulations, which attempted to set aside certain offices for specific ethnic groups, indicate a fair degree of political mobility, although the paths to office were nonetheless not seen as traditional by Chinese scholars. This is not to say, however, that the ruling Mongols did not attempt to construct a differentiated society, in which they themselves and their preconquest allies — the Western and Central Asians — would benefit most. Although demographic realities dictated their employment of Chinese at virtually all levels of government, the Mongols nonetheless withheld certain privileges for non-Han ethnic groups. In regard to ethnic elites, certainly the Turks were the most relied-on and privileged group after the Mongols themselves in thirteenth- and fourteenthcentury China. This privileged position is easily explained by the special relationship forged between the various Turkic tribes and the Mongols in the early thirteenth century.57 By 1225, the Uighur, Qarluq, Qangli, Qipchaq, Onggiid, Kereyid, and Naiman (whether the last two were Turkic or Mongolian is debatable) all had submitted to Chinggis khan. The Mongols adapted the Turkic Uighur vertical script to their spoken language, and the Turks, many of them Uighur, served as military advisers, secretaries, imperial tutors, and civil administrators to the Mongols in the pre-Khubilai period. In this period, the lingua franca on the eastern steppe was Turkic, not Persian. During Khubilai's reign, a large number of Turks continued to be employed at the court as imperial advisers, tutors to the princes, translators, and military officers. Khubilai's mother, Sorghaghtani Beki, was, after all, a Kereyid princess, and Khubilai had grown up surrounded by Turkic advisers. In post-Khubilai Yuan China, Turks excelled in political intrigue, the most prominent political personage among them being the Qipchaq El Temiir (Yen T'ieh-mu-erh, d. 1333). Among the various groups of Turks, the Uighurs were most responsible for bridging the gap between Chinese and Mongolian culture. Many Uighurs were translators and interpreters of Chinese and other languages into Mongolian, and several were unquestionably sinicized, as can be seen in their literary and scholarly achievements in Chinese. The majority of Western and Central Asians who held chin-shih degrees were Uighurs. Yet it would be inaccurate to define elites in Yuan China only in terms of their ethnic origins. Some elite groups seem to have been favored by the 57 See Igor de Rachewiicz, "Turks in China under the Mongols: A preliminary investigation of TurcoMongol relations in the 13th and 14th centuries," in China among equals: The Middle Kingdom and its neighbors, loth—14th centuries, ed. Morris Rossabi (Berkeley and Los Angeles, 1983), pp. 2 8 1 - 3 1 0 . Cambridge Histories Online © Cambridge University Press, 2008 6l2 THE YtJAN GOVERNMENT AND SOCIETY Mongols more on account of their religion or their commercial utility. An example is the Muslim ortogh. The members of these merchant associations formed partnerships {ortogh) with the Mongolian imperial family. According to the Chinese sources they were Muslims, but specific information on their ethnic background is scanty. The Yiian regulations dealing with taxation and corvee attempt to clarify the taxation category of ortogh merchants: They were not exempted from household taxes just because they were Muslims or merchants. An exemption was extended only to Muslim military households or Muslim clergy households that "reside within Muslim mosques and, moreover, are without livelihoods." The ortogh merchants, despite their privileged business relationship with the imperial family, were subject to the same tax and corvee obligations as were commoners.'8 But the ortogh merchants were perceived by Chinese literati as a highly favored commercial—religious elite and were accused of tax evasion and usurious moneylending practices. Undoubtedly, some ortogh merchants were guilty of these charges, but it is also likely that the Mongols were not at all displeased that Muslims should bear the brunt of such hostility, as it was then deflected from the Mongols themselves. Some Chinese merchants fared well under Mongolian patronage. Two former pirates who had surrendered their fleets to the Mongols in the 1270s, Chang Hsiian and Chu Ch'ing, amassed fortunes from their central role in transporting grain from south China to Ta-tu. During their lifetimes, Chang and Chu not only were allowed to print currency but also were granted military titles usually reserved for Mongols and Western and Central Asians, and they were exempted from corvee obligations. Although Chang Hsiian was eventually executed and Chu Ch'ing died in disgrace, the successes of these two Chinese merchants showed that the Mongols in China, as in other parts of Eurasia under their control, heavily rewarded commercial agents who served the imperial court. In addition to ethnic, religious, and commercial elites, the Mongols also favorably treated the old landholding elite left over from Southern Sung China. The conquest of north China had caused far more economic dislocation among the Jurchens and Chinese in the north than the conquest of south China caused for southern Chinese landlords. In fact, the importance of south China as a revenue base prevented the Mongols from tampering with that area's socioeconomic order. Thus hereditary appanages and apportioned territories were carved out mostly in the north. The Mongols' full cognizance of the importance of southern China to the overall economy is reflected in their 58 T'ung chih t'iao ko (Peking, 1930), chap. 2, pp. 14b-15a. On Muslims in general in Yiian China, see Morris Rossabi, "The Muslims in the early Yuan dynasty," in China under Mongol rule, ed. John D. Langlois, Jr. (Princeton, 1981), pp. 2 5 7 - 9 5 . Cambridge Histories Online © Cambridge University Press, 2008 SOCIETY 613 completion under Khubilai of the Grand Canal, the all-important economic and political link between the Yangtze region and the capital, Ta-tu. South China's landowners benefited greatly under the Yiian's economic policies, except when the Tibetan Buddhist cleric Sangha (Sang-ko), as imperial adviser under Khubilai in the late 1280s to 1291, initiated a campaign to call in unpaid taxes. Once Sangha had been executed and his unpopular financial policies reversed, south China was no longer subject to such onerous tax collection. Thus, the southern Chinese landowning population can be seen as an economic elite, largely left to its own devices in the Yiian period. 59 The Mongols devised household categories based mainly on occupation to describe and keep track of both the elite and nonelite population of Yiian China.6° Nonelite households that produced and manufactured goods, such as peasant, artisan, and mining households, were mainly composed of Han and southern Chinese, whereas Mongolian households were classified typically as military households, hunting households, and postal households. Western and Central Asians were generally categorized as military households, ortogh households, merchant households (not all foreign merchants were ortogh), and religious households. Most of the categories were hereditary, and from the Mongols' point of view, each served the state. The Mongols granted tax and corvee exemptions and other privileges according to both ethnic criteria and the relative importance of the household's occupation to the state economy. The granting of government stipends and exemptions from corvee and military obligations to scholarly households (ju-hu), however, would seem to contradict these criteria. The Mongolian emperors acquiesced to memorialists' requests for favorable status for the scholarly households, most likely in order to appease that small but important segment of the population. In 1276 the ju-hu numbered only 3,890, a small enough group for the Mongols to waive certain of their state obligations. The number of scholarly households remained relatively low, in large part because this category of household was not hereditary; an incompetent scholar could lose his status. At the very bottom of Yiian society were various types of slaves. The Yiian period vis-a-vis earlier periods in China's history saw an increase in the number of slaves. Historians have looked to the internal dynamics of preconquest Mongolian society to explain this phenomenon. Although the trend among scholars in the People's Republic of China has been to describe the early thirteenth-century Mongols as passing through a slave-holding 59 See Uematsu Tadashi, "The control of Chiang-nan in early Yiian," Ada Aliatica, 45(1983), pp. 4 9 - 6 8 . 60 See Huang Ch'ing-lien, Yiian lai hu chi chih tu yen chiu; Oshima Ritsuko, "The Chiang-hu in the Yiian," Ada Asialica, 45 (1983), pp. 69—95; Hsiao Ch'i-ch'ing, "Yiian tai te ju hu: Ju shih ti wei yen chin shih shang te i chang," in his Yiian tai shib hsin t'an (Taipei, 1983), pp. 1—58. Cambridge Histories Online © Cambridge University Press, 2008 614 T H E YOAN GOVERNMENT AND SOCIETY stage on the way to an early stage of feudalism in the linear Marxist scheme of historically determined socioeconomic stages of development through which all peoples pass, Soviet and Mongolian People's Republic scholars resolutely hold that the Mongols skipped the slave-holding stage, thus moving directly from a clan to a feudal society.6l Although these debates are not of direct interest to us here, they do exemplify the difficulty that historians experience in describing the role of slaves in early Mongolian society. Although the thirteenth-century Mongols did indeed have slaves — usually non-Mongolian war captives rather than indigenous slaves — it would not be correct to describe slave holding as a fundamental characteristic of the Mongols' tribal and clan-based pastoral nomadic society and economy. Slavery was of particular importance to the economy of Mongolian soldiers in Yuan China.62 The Mongols kept captives from military campaigns, and many of these captive slaves (ch'ii-k'ou) and their families were allocated to soldiers for use in cultivating their lands, as Mongolian soldiers were loath to till the soil themselves. Many of the captive slaves were Chinese, and by the turn of the fourteenth century, so many of these captive slaves had run away that Mongolian military households became impoverished, and ironically, Mongolian men and women themselves began to be exported as slaves to India and Islamic countries, starting as early as the late thirteenth century. Although most slaves in Yuan China were thirteenth-century prisoners of war, there also is evidence of the continuing enslavement, as well as the buying and selling, of slaves throughout the Yuan period. Some were captives taken during internal rebellions, but others were apparently just arbitrary victims enslaved by officials and soldiers. Contemporary Yuan observers deplored the existence of slave markets in Ta-tu, remarking that people were being treated like cattle. To the Mongols, however, the category of slave was indeed connected conceptually with ownership of animate and inanimate objects. This is demonstrated by the existence of the so-called Agency of Men and Things Gone Astray (Lan-i-chien), in which the disposition of runaway slaves, lost material goods, and lost cattle was not differentiated. Yuan government and society reflect both continuities and breaks with the Chinese past. Yuan political institutions and styles of governance were based on Mongolian, Inner Asian, and Chinese precedents, often difficult to disen61 See, for instance, Kao Wen-te, Meng-ku tin li chih yen chiu (Koke Khota, 1980); and Lu Ming-hui, "San shih nien lai Chung-kuo Meng-ku shih yen chiu kai k'uang," in Meng-ku shih yen chiu lun wen chi, ed. Lu Ming-hui et al. (Peking, 1984), pp. 240—3. On the treatment of Mongolian social development and Yuan history in the Soviet Union and the Mongolian People's Republic, see Elizabeth EndicottWest, "The Yuan," in Soviet studies of premodem China, ed. Gilbert Rozman (Ann Arbor, 1984), pp. 97-110. 62 See H s i a o , The military establishment of the Yiian dynasty, p p . 2 1 , 29—30; and Tetsuo Ebisawa, "Bondservants in the Yiian," Ada Asiatica, 45 (1983), pp. 27—48. The Japanese scholarship on slaves in Yiian China is extensive. Cambridge Histories Online © Cambridge University Press, 2008 SOCIETY 615 tangle from one another. The Mongols often used Chinese means to achieve specifically Mongolian goals (e.g., the use of the Chinese hereditary yin privilege to maintain ethnic elites) and again used Mongolian means to achieve goals that any dynasty on Chinese soil inevitably strove to attain (as in the establishment of the Mongolian office of darughachi to oversee local government). The particular needs of the Mongolian ruling elite led to governing measures that probably would not have arisen indigenously. Historians of Yuan China continue to assess the unique elements in Mongolian rule, or more specifically, to define what constituted "un-Chinese" (actual or perceived) ways of governance. Identifying, explaining, and assessing Yuan governmental institutions and social practices reconfirm for historians the distinctive character of the era of Mongolian rule. Cambridge Histories Online © Cambridge University Press, 2008 CHAPTER 9 CHINESE SOCIETY UNDER MONGOL RULE, 1215-1368 THE MONGOLIAN PERIOD IN CHINESE HISTORY The great khan (more properly, khaghan) Khubilai, who had taken that title denoting supreme rule of the Mongolian empire in 1260, took a further step at the end of the year 1271: He proclaimed that starting with the New Year, his government in China would be called the Great Yuan dynasty. He was acting on the advice of Chinese and sinified non-Chinese counselors, and his proclamation employed allusive wordings from Chinese tradition supplied by them. They devised the terminology to place the alien conquest dynasty within the traditions of Chinese statecraft, to express for him benevolentsounding objectives vis-a-vis his Chinese subjects and their cultural traditions.1 That gave an appropriate mask to, but did not conceal, the fact that the Mongols had come into China to enrich themselves and sustain their military empire beyond China. They were under pressure to maintain their military and political superiority in China in order to exploit the resources of the world's largest and richest nation. They altered their approaches to that problem successively throughout the 150 years from Chinggis khan's early campaigns against the Jurchen Chin dynasty in 1215 until the Mongols were driven out of China in 1368. Khubilai khan's ceremonious adoption of Chinese dynastic forms in 1272 began the period of greatest Mongolian adaptation to Chinese influences on the patterns of government. Khubilai's long and illustrious reign also marked the fullest regularization of Yuan governing procedures. But we must note that in not all of these matters did he accept Chinese "guidance" designed to detach Mongolian rule from its origins in the other body of historical experience on which the Mongols drew: the traditions of the steppe and the norms of the Mongolian empire. The Chinese at that time and since have nonetheless accepted the period of Mongolian rule as a legitimate dynasty in their political tradition. And despite cogent reasons today for seeing that century and a half, as the Chinese always have, as an era in Chinese social history, we must not let this obscure 1 For a translation of the edict proclaiming the dynasty and a discussion of its significance, see the introduction to John D. Langlois, Jr., ed., China under Mongol rule (Princeton, 1981), pp. 3-21. 6l6 THE MONGOLIAN PERIOD IN CHINESE HISTORY 617 for us the fact of extraordinary changes in the management of Chinese society. We must note the consequences of those for Yuan social history and must attempt to assess their consequences for post-Yuan history as well. In the long-range view, the continuities nonetheless dominate. We see no fundamental displacement and redirection of the course of national history of the kind that historians describe for Russia, induced by the Mongolian destruction of Kiev in 1240 and the Golden Horde's subsequent domination of the Russian state until 1480.2 In East Asia the Mongolian conquests terminated some histories, transformed others, and created new nations, most notably their own. The early years of Mongolian dominance witnessed, from 1215 to 1234, the destruction of the Jurchen and Tangut states that lay largely within the northern borders of China, and the dislocation or virtual extinction of their peoples. There was nothing comparable to that in the case of the Chinese nation. It was spared the direct onslaught of the early campaigns of conquest, and in any event its massive size may have buffered it against such thoroughgoing dislocations. Mongolian patterns of conquest changed somewhat after the 1240s. The later Mongolian conquerors, Mongke (r. 12509) and Khubilai (r. 1260-94), under whom China was brought into the Mongolian empire, dealt with their sedentary subjects more purposively and more effectively in the interests of the Mongolian state than had their formidable warrior predecessors. Their policies also better served the interests of their conquered subjects; in some measure, a congruence of interests was worked out. This is not to ignore the troubling departures from the normal Chinese order of things that ensued. Eventually, however, the Chinese felt that they had survived and triumphed over an unprecedented disaster in the Mongolian conquest of their venerable civilization. Apart from these externally imposed crises and the Chinese adaptation to them, there also are cogent arguments to support the idea that the Yuan dynasty coincided with a watershed in Chinese historical development. Some of the net changes in Chinese civilization evident by the end of the period, particularly in the realms of government and statecraft thought, can be seen as culminations of trends long present in China, though enhanced by the special conditions of Mongolian rule. On the other hand, we must also take 2 Charles J. Halperin, Russia and the Golden Horde: The Mongol impact on medieval Russian history (Bloomington, 198;), though not contradicting what has here been called the "fundamental displacement and redirection of the course of history" in the case of Russia, nonetheless emphasizes cultural continuities and sees incidental benefits to Russia arising out of the "Mongol impact." Halperin aims to correct Russian historians' consistent ignoring of all good consequences of Mongolian overlordship in Russia. In contrast, the Chinese, though critical, have tended to deemphasize the destructive aspects of alien rule in China and to emphasize cultural continuity, nonetheless imputing the "good" to Chinese cultural superiority, not to the alien presence. 6l8 CHINESE SOCIETY UNDER MONGOL RULE account of the disruptive change and the varied Chinese responses to all the new factors introduced by the alien presence. The point of view adopted here is that the latter, the set of circumstances directly attributable to the Mongolian overlordship, accounts for much in both political and social history. It is more difficult to marshal supporting evidence from the life of the society at large than it is from the political sphere. This chapter will attempt to suggest the kinds of issues in social history that give the Yuan dynasty its interest and importance in the minds of historians today. THE POPULATION OF YUAN CHINA Some of the most basic facts about Yuan society remain subject to speculation and debate. The most glaring example is the uncertainty about the size and distribution of the Chinese population. A later section of this chapter will show that the Yuan government went beyond all precedent in its effort to classify and register its subjects according to status and occupation, in order to serve its social management objectives. The actual counting of households and individuals, nonetheless, was less directly relevant to the Yuan fiscal system than it had been under Chinese dynasties, and the administrative machinery responsible for the census, taxation, and land registration was not notably efficient. The quality of these data is thus more than usually suspect. Historical demographers point to the figures from a census in 1290, late in the reign of Khubilai khan, as the most reliable of the Yuan period. As reported in the Yuan shih,* it registered 13.19 million households containing 58,834,711 persons (see Map 37). The historians note, however, that no figures are given for the newly conquered province of Yunnan, for a number of widely scattered prefectural and county level administrative units, for people "dwelling in mountains and marshes" and other remote places, or for several large categories of residents such as monks and priests, the military, and households in bondage to appanages. The only other nationwide figures from the Yuan following the conquest of the Southern Sung are supposedly from a new census dated 1330; they show an insignificant increase, and there is reason to believe that they are largely figures carried over from 1290, not the result of a new census. The number of individuals per household in the 1290 censusfiguresis about 4.5, low but not impossible. It seems plausible that the population of Yuan China shortly after the conquest of Southern Sung in the 1270s was in the range of 65 million persons. The 1290 figures closely match those provided by the 3 Sung Lien et al., eds.. Yuan shih(Peking, 1976), 5 8 , p. 1346 (hereafter cited as YS). THE POPULATION OF YUAN CHINA 1 dot represents 5000 households MAP 37. Registered population distribution of Yiian China 619 62O CHINESE SOCIETY UNDER MONGOL RULE best early Ming census, that of 1393, which registered 10,652,789 households and 60,545,812 individuals. There the mean household size is 5.68, but the total of 60.5 million individuals is very close to the Yiianfiguresof a century earlier. One long-standard work expresses the opinion that the actual population in 1393 was larger, and it shows that fiscal considerations took precedence in the census taking, allowing the undercounting of nontaxpaying young children, widows, and the infirm (although that would raise still further the ratio of individuals to households).4 The Yiianfiguresof 1290 thus seem to be corroborated by the early Ming figures of 1393. Our confidence in those figures is challenged by the fact that the population of China had been substantially greater in Sung times. The Northern Sung government in 1109 registered 20 million households (implying a total population of more than 100 million), and the combined registered population of Chin and Southern Sung (i.e., more or less the same area) in about the year 1200 has been calculated to be over 100 million.' It is difficult to believe that the population of China was reduced by one-half during the thirteenth century and was still that low at the end of the fourteenth century after a quarter-century of recovery from the Yiian period. Yet if we assume that administrative laxity - meaning the inability to conduct a thorough census - or the intentional omission of some groups (e.g., the captive households granted in fief to Mongolian leaders' appanages) offers an adequate explanation for the low 1290 figures,6 it would be reasonable to assume that the more comprehensive early Ming figures from 1393, when administrative rigor was in force, would show a marked increase. At least, one could assume that the household figure was closer to reality, even if the counting of individuals was skewed byfiscalconsiderations. Instead, they corroborate the 1290 figures. Although none of these census registrations should be interpreted as the result of a thorough attempt to count all the individuals in China, as their purpose wasfiscalmanagement and not demographic research per se, it is probably true that they indicate the general contours of population increase and decrease and of distribution. Thus one must assume that there was a catastrophic reduction in China's population between 1200 and 1400, the most extreme in the history of China. A closer look at some of thefiguresstrengthens that assumption. The Chin dynasty's registration of 1207 gives a total population - essentially that of China north of the Huai River - of 8.4 million households containing 53.5 4 Ho Ping-ti, Studies on the population ofChina, 1368-1953 (Cambridge, Mass., 1959), pp. 10—12. 5 Ho Ping-ti, "An estimate of the total population of Sung-Chin China," in Eludes Song in memoriam ttienne Balazs, 1st series, no. i , ed. Franchise Aubin (Paris, 1970), pp. 3 3 - 5 3 . 6 For a discussion of census-taking procedures in the Yiian, see Huang Ch'ing-lien, Yiian tai hu chi Mb tu yen chiu (Taipei, 1977), pp. 128-35. THE POPULATION OF YtJAN CHINA 621 TABLE 8 Numbers of households registered in Hopei and Shantung 1207 1290 1393 3 . 3 million 1.3 m i l l i o n 1.1 m i l l i o n million persons (6.36 persons per household). Southern Sung China in 1223 registered 12.6 million households, which, at an estimated ratio of five persons per household (Sung census procedures did not include all persons in counts of individuals; in any case, the estimated ratio of five persons per household is low), gives a figure of 63 million persons. Those two registrations taken together provide the basis for estimating the total population of China in the year 1215 — the year that the Mongols commenced their incursions into north China — in the range of 110 million to 120 million. When we look at the distribution of the Yuan registered population in 1290, we see that the greatest losses occurred in the North China plain, the region that suffered most severely from Mongolian-Jurchen warfare between 1215 and 1234, from loose management leading to constant disorder throughout the decades from 1235 until Khubilai's accession in 1260, and finally, from the revolts of Hopei and Shantung warlords early in Khubilai's reign. The 1290 census shows figures for the population of the administrative subdivisions that are more or less the same as those for Hopei and Shantung and that can be compared with the census registrations for those same areas in 12077 and 1393, as in Table 8. The household figure of 3.3 million in 1207 implies a total population of 17 million to 20 million persons. Roughly one-third that number were registered in 1290 in those two provinces; decreases on that order are typical for all north China, including Honan, Shansi, and Shensi. The region probably did not recover the late-twelfth- and early-thirteenth-century level of population until the late sixteenth century. Yet we are not certain whether these reductions in the registered population were caused by the failures of administrative machinery to record people or by the actual loss of people. If the latter, we do not know whether people fled disorders in a great series of internal migrations, were killed in warfare, died as a consequence of famine and epidemics attendant on the disorders, or were lost through forestalled births because of hardships and dislocations. 7 The figure for 1207 is derived by extrapolating from T'o-t'o et al., eds., Chin shih (Peking, 1975), 24, pp. 5 7 2 - 8 ; 25, pp. 599—616; 26, pp. 627—9 (hereafter cited asCT), omitting three chou later included in Kiangsu and seven fu and chou later included in Honan. The figure for 1290 is based on YS. 58, pp. 1347-83. The figure for 1393 is from Ho Ping-ti, Studies on the population of China, Table 3, p. 10. 622 CHINESE SOCIETY UNDER MONGOL RULE That the thirteenth and fourteenth centuries were a long period of repeated disasters throughout much of north China is strongly supported by descriptive information, yet the precise facts of population history cannot be derived from that. If migration from one locality in China to another were a significant factor, we might expect the migrants' descendants to show up in the 1393 census, but they do not. One might assume a combination of these factors, but the net loss from warfare and calamity plus the forestalled replacement attributable to long decades of hardship presents itself as the inescapable conclusion. It is troubling to face a large riddle of this kind: If modern historians cannot know with more precision about the size and distribution of the population and the causes of its fluctuations, what can they say with assurance about the social history of the period?8 Although the quantitative data have so far failed to resolve the riddles of the Yuan period's population history, fortunately the qualitative information allows historians to come to more satisfying, though by no means undisputed, conclusions about the life of Chinese society under Mongolian rule. SOCIAL-PSYCHOLOGICAL FACTORS The Chinese people had experienced alien rule at several points in their history, but never before over their entire nation. After a number of probing campaigns into north China throughout the decade following acceptance of Chinggis khan's leadership by his Mongolian and allied compatriots in 1206, Mongolian armies first conquered the two other alien states then occupying northern parts of China: The Tangut Hsi Hsia state in the northwest fell in 1227, and the Jurchen Chin dynasty, whose conquest had required twenty years, finally collapsed in 1234. At that stage in their history — while their armies simultaneously drove westward across Asia and into Europe — the Mongols' goal was to defeat any nation or fortified city foolish enough to resist them, but not to occupy and govern it. North China was repeatedly crossed by armies and ravaged; local military leaders were often left in loose 8 A recent and thorough exploration of historical materials relevant to the Yuan population problem is offered by Ch'iu Shu-shen and Wang T'ing in "Yuan tai hu k'ou wen t'i ch'u i," Yuan sbih lun ts'ung, 2 (1983), pp. 111—24. This study accepts the Yuan period registration figures just cited, proposing that they represent a 20 percent underreporting. It then estimates that a maximum figure for the Yuan period was reached some decades after the registration of 1290, in about 1340 and that it can be calculated to have reached 19.9 million households and 90 million persons, figures that were reduced again by the late Yuan warfare to 13 million households and over 60 million persons by the end of the dynasty in 1368. This solution compounds the dilemma by proposing wide fluctuations both up and down, without analyzing the annual rates of increase between 1290 and 1340 that would have been necessary to produce the high estimate for 1340 or without explaining why such rates would not have been present again in stable years after 1368. Also, it requires two catastrophic reductions in the actual population, ranging from 33 to 50 percent, one following the peak of around 1215 and another following that of 1340. Nonetheless, this study merits careful consideration. SOCIAL-PSYCHOLOGICAL FACTORS 623 control, only to fight among themselves; and thus ad hoc arrangements to maintain peace and order in some places emerged from the local society. This general instability in north China began to change in the 1250s when Khubilai, grandson of Chinggis and younger brother of the great khan Mongke, was delegated to deal with problems on the China rim of the Mongolian world empire. Khubilai campaigned in southwest China in 1253—4 anc^ conquered Yunnan, and in the late 1250s he became deeply involved in north China, not as conqueror but as governor, utilizing Chinese and other assistance. Then, after the death of Mongke and succession to his title as great khan in 1260, Khubilai became the actual ruler of China north of the Huai River boundary with the Southern Sung, and the claimant to universal rule over the Mongolian world. This claim he never substantiated, but his increasing involvement in China, including the completion of the Mongolian conquest of the Southern Sung in the 1270s, led to longer-range planning and more responsible governing in China. Ruling over that vast sedentary nation - then as always the largest society in the world - was to prove a new kind of challenge for the remarkable Mongolian conquerors. Successive phases of the conquest brought differing Chinese responses. This alien conquest meant conflicting things to the Chinese, especially to the literati who understood it best. On the one hand, China was again unified for the first time since the tenth century. After the fall of the Southern Sung to Khubilai's generals, some Chinese joyously observed that people in the long-separated southern and central provinces could again travel to such venerated sites of Chinese antiquity in the north as the tomb of Confucius in Shantung and the capitals of Han and T'ang in Shensi. The reunification of all Chinese under one regime bearing the Mandate of Heaven was a fact of momentous psychological import. At the same time, however, the fallen Sung dynasty was revered as an age of civil virtues and cultural florescence. In all-too-stark a contrast, the Mongols were aliens whose military prowess, not cultural achievement, was what most impressed the conquered. Whether these steppe warriors would in time succumb to civilizing, that is, to Chinese ways, appeared at best uncertain. Some Chinese of the time rejected Mongolian rule, declined to serve the new regime, and remained resentfully loyal to the fallen Sung in ways supported by Chinese tradition. Fear of the Mongols' military might and the initial rejection of them as unworthy rulers over the center of civilization undoubtedly intensified such reactions through the end of the thirteenth century and perhaps thereafter. Nevertheless, Chinese civilization knew only one criterion for legitimating a new dynasty, and the Mongols could be seen as meeting that test. The Mandate of Heaven theory did not demand that China's rulers be Chinese, 624 CHINESE SOCIETY UNDER MONGOL RULE only that they accept the conceptual framework on which the Chinese imperial institution rested {cheng) and that they bring all the Chinese under one unified rule (t'ung).9 The mandate theory implied a common ground of humane ethical and social values, the adherence to venerable ritual norms, and a well-worked-out pattern of civilian rule through a bureaucracy of merit as measured by the standards of Confucian cultivation. Khubilai claimed this mandate for his Mongolian imperial house, though he might otherwise have ignored its demands and ruled simply by force. He claimed the mandate formally in 1272 when the new Yuan dynasty was proclaimed, and his claim was validated when he succeeded a few years later in ending the Sung dynasty by conquest. At that juncture, Khubilai demanded the formal abdication of the last Sung ruler, to whom he then granted a minor title and income. The Mongols possessed clear-cut military superiority; nonetheless, they were punctilious in meeting these Chinese formalities. They may have calculated, correctly, that this would weaken further Chinese military and psychological resistance, by demonstrating that the Sung ruler had acknowledged the transfer of the mandate. It also paved the way for Chinese to serve in the new civil government. The manner in which all this was done, contrasting with earlier Mongolian practice, shows how far Khubilai and his generation of Mongolian leaders then on the scene in China had adapted to conditions there. Nonetheless, the Mongolian government thereafter failed to measure up to Chinese ideal norms. Khubilai was the one Mongolian ruler who best understood China, and he made brilliant beginnings in devising the accommodations that might meet both Mongolian imperial needs and Chinese expectations. Nonetheless, despite serious efforts by some outstanding Mongolian administrators, a corps of Inner Asian and Western Asian assistants (the se-mu Jen), and large numbers of officials drawn from the Chinese elite, all striving to bring reasoned and orderly governing to the Chinese nation - albeit under the necessity to maintain Mongolian prerogatives - the Yuan dynasty never became a normal age in the long political history of China. There was an unresolved incompatibility between the military structure of power that directly served Mongolian imperial interests, and the gradually more fully established Chinese forms of civil government. Another problem was that after Khubilai's long and politically percipient reign, the succeeding Mongolian emperors to the end of the dynasty in 1368 were mostly short-lived and ineffectual, often puppets of competing factions. 9 Richard L. Davis, "Historiography as politics and Yang Wei-chen's 'Polemic on legitimate succession'," T'oung Poo, 59 (1983), pp. 33—72. For background on the "legitimate succession" issue, see Hok-lam Chan, Legitimation in imperial China: Discussions under the Junhtn-Chin dynasty ( u 13—1234) (Seattle, 1984)- SOCIAL-PSYCHOLOGICAL FACTORS 625 Mongolian power was thus weakened by being used in an increasingly disorderly fashion. The quality of civil government declined. The glaring discrepancies between nominal standards and actual practices induced cynicism and despair among many of the governed and governors alike. Only a few Chinese took this as grounds for expressing their doubt that the Mongols had gained the mandate legitimately; many more began to foresee their early loss of it. To say that the Chinese elite accepted the legitimacy of the Mongolian overlordship may be going too far, although no doubt many merely acquiesced in it, in an effort to make the best of a bad situation. That circumstance was not unknown under Chinese dynasties. In any event, Confucian ideas about the importance of serving the legitimate rulers and of loyalty to ruler and dynasty from which one has accepted appointment to office continued to provide binding norms of behavior. The anomalies in Chinese responses to Mongolian rule are intricately convoluted.10 Modern historians, who must see the Yiian dynasty as an extraordinary period, get less direct assistance on this point than they might have expected from contemporary observers and traditional Chinese historiographers. In claiming the Yiian dynasty as an era in their own history, the Chinese have always overlooked the fact that Yiian China was an appendage of a larger world empire. They have been oblivious to the fact that Mongolian history had its own integrity quite apart from the movement of Chinese history, even when the two histories overlapped to the great extent that they did during the Yiian dynasty." Today we must acknowledge that both the Chinacentered and the Mongolian empire—centered histories of the period are valid and that we should attempt to transcend the limitations of both. Nonetheless, it is the Chinese record that provides the principal documentation for both. And that record must be used perceptively in order for us to understand the relationship of the two societies and the way in which their juxtaposition generated special conditions in China. We cannot expect the Chinese of the past, even those of the recent past, to have analyzed the Chinese—Mongolian relationship in ways that we will find completely appropriate today. For example, the Chinese record often expresses dismay at Mongolian social delinquencies and political shortcomings. At the same time, it is committed to upholding Chinese cultural faith in the role played by Chinese institutions and ideals and to expressing confidence that what formed the familiarly Chinese component of Yiian period govern10 Frederick W. Mote, "Confucian eremicism in the Yiian period," in The Confucian persuasion, ed. Arthur F. Wright (Stanford, i960), pp. 2 0 2 - 4 0 . 11 The contemporary Chinese scholar Li Tse-fen called this the first of the three "defects" (pi) of Chinese historiography on the Yiian period; see the preface to his Yuan shih hsin chiang (Taipei, 1978), vol. 1, p. 2. Contemporary scholarship in Taiwan and mainland China shows increasing awareness of this shortcoming but has not overcome it. 626 CHINESE SOCIETY UNDER MONGOL RULE ment constituted its essential element. We, however, can see that it counted for far less than they believed and that the special elements stemming from Mongolian norms also must be given their true weight. Most Chinese professed to believe that the Mandate of Heaven had been justly bestowed on the Mongolian great khans, enabling them to win the Chinese throne. We today perceive that Mongolian military effectiveness gave them their unaided victory, and we are apt to see the Chinese view of that as post facto rationalization, if not naivete. The Chinese believed, or hoped, that the conquerors undertook in some measure to govern in traditional modes because they acknowledged the superiority of humankind's single civilization. We, on the other hand, perceive that the Mongols in China, as in the steppe, in Central Asia and Persia, and in Russia, adapted fluidly to various civilizations in all the realms they conquered. What the Chinese have observed as obeisance to their universal culture was in fact pragmatic decision about how best to serve Mongolian interests in all places and times.12 Thirteenth- and fourreenth-century Chinese writers and officials often professed to see childlike virtues of simplicity and honesty, generosity and trust, in the Mongols: They were to some degree "noble savages" while at the same time they were backsliders on the road to civilization. But we can perceive that the Mongols, whatever their endearing or awesome qualities, had a shrewd understanding of how to best exploit their conquered populations. Khubilai khan led in identifying Mongolian long-range interests with the China base of their international power, but not to the abandonment of the Mongolia-centered concerns. Subsequently, some degree of adaptation to China at the expense of Mongolian mastery of the steppe may have been under way by mid- and late-Yuan times; modern historians disagree about how far the process may have gone or what the eventual outcome might have been had the Mongols continued to rule from Peking for another century. •» That question, however intriguing, is historically irrelevant. The Mongols were driven out from China in 1368 at a time when they still were, essentially, aliens — people of the steppe. That largely explains the failings of their governing in China, as well as their subsequent existence as a powerful nation 12 Thomas T. Allsen, Mongol imperialism: The policies of the Grand Qan Mongke in China, Russia, and the Islamic lands, 1251—1259 (Berkeley and Los Angeles, 1987), pp. 221—5. 13 In a pioneering study of Mongols during the Yuan dynasty who acquired some measure of Chinese learning and cultural skills, Hsiao Ch'i-ch'ing studied seventy-eight individuals whose attainments are attested to in Yuan period writings. He concluded that toward the end of the Yuan dynasty, significant numbers of Mongols from the elite strata had become "Confucianized" and that had the dynasty not fallen, the trend toward cultural adaptation would have grown stronger. See his conference paper "Yuan tai Meng-ku jen te Han hsiieh," in Kuo chi Chung-kuo pien chiang hsiieh shu hui i tun wen chi (Proceedings of the international conference on China border area studies'), ed. Lin En-hsien (Taipei, 1985). PP- 369-428. SOCIAL CLASSES: TRADITIONAL AND NEW ELITES 627 in Mongolia. Their failings in governing China, however, were the most telling of the special conditions of life for their Chinese subjects. The Chinese adaptations to those special conditions will be the focus of what follows. SOCIAL CLASSES: TRADITIONAL AND NEW ELITES It has long been customary to see in the Sung dynasty era (960-1279) a significant lessening of all lingering barriers to upward and downward social mobility. That is, in Sung society, elite status was individually achieved through merit as measured by the civil service examinations or by other displays of personal energy and ability. The older pattern by which great family wealth and advantage ensured de facto elite status and political position through successive generations, by the Sung period gave way to a more Confucian ideal of an open society in which a new kind of elite of merit was recruited from a broad social base to rise into personally achieved, nonhereditary official rank and so to dominate society. '* Recent scholarship has modified this view somewhat but has not invalidated it. The open society ideal had a lively impact on social consciousness, and to greater extent than in other premodernized societies in both East and West. China before the Mongolian conquest had become a society having no aristocracy by ascription, no legally privileged or legally disfranchised (in the sense of being denied access to upward mobility or official status) closed classes, and no segment of its population bound to place or to occupation. There are, to be sure, some small exceptions to these generalizations, at both the top and the bottom of the social scale, but they are insignificant. The awareness of living in an open society was pervasive; it contributed much to the social expectations and behavior of all, rich and poor, urban and rural, official and commoner. Enjoying the greatest measure of social prestige and favor from the state that the scholar-official elite had ever known in history, the Sung period literati could not doubt that they were the proper leaders of society and government. They had benefited from the revived Confucianism, or NeoConfucianism, and for three centuries had triumphed over previously competitive Buddhism and Taoism, gaining great self-assurance concerning the Tightness of Confucian norms. Scholars and officials - expectant, actual, or retired - all were designated ju (pronounced "rue" and meaning, in this period of Chinese history, men of Confucian education or just men of cultivation); to be a ju was the most 14 That view of the Sung period is best summarized in Saeki Tomi and Chikusa Masaaki, So no shin bunka, vol. 6 of Toyo no reikisbi, ed. Saeki Tomi (Kyoto, 1967). 628 CHINESE SOCIETY UNDER MONGOL RULE gratifying and most highly rewarded condition of the individual in Sung society. Even beyond its expected application to the scholar-official elite group, the most honored painters became ju artists; the most learned physicians became ju doctors; and even military leaders strove to merit the appellation ju general (with, however, no noticeable improvement in the military strength of the Sung state). The state recognized and encouraged this trend by means of various legal, economic, and institutional adaptations. This group of natural social leaders thus had the most at stake in the Sung period status quo. They had seen their ideals strongly supported even by the alien Chin or Jurchen dynasty of conquest in north China. Despite the loss of the north to those precursors of the Mongols, self-confidence marked the Southern Sung elite mind-set; they had the most to lose, the most to be bitter about, from the social changes that the Mongolian conquest effected.15 Nomadic societies of the steppe were organized quite differently. Although in actual practice the fluid demands for able military leadership kept the elite status relatively open to men of exceptional ability, these tribal military societies were nonetheless organized on the principles and ideals of closed social classes, of hereditary privilege even when individually earned, of inherited occupations and statuses, and of differentiations among people based on real or mythical blood lines. The Mongols brought such a social system with them into China and struggled to make it work to serve their interests as conquerors. Among them, too, ideal and actual patterns diverged in some degree, but the contrast between their social expectations and those of the Chinese was stark. For the Chinese at that stage in their social development to have come under the rule of a conquering nation bringing with it such different experience in organizing society was a greater shock than it might have been at an earlier time in Chinese history. The social history of the Yuan turns on matters such as how, through successive modifications, the Mongols applied their ideas of rule, to what extent those ideas were realized when applied to Chinese society, how the Chinese resisted and adapted to Mongolian governing norms, and to what extent the unprecedented circumstances brought lasting changes to the Chinese social base. The Yuan dynasty was an interval of alien rule between the native Chinese Sung dynasty and the native restorationist regime of the Ming dynasty. It is clear that the tone and modes of politics in early Ming contrasted sharply with those of the Sung, and one can ascribe this in no small part to the cumulative effects of Mongolian rule. It is much less clear that the structure 15 See Hsiao Ch'i-ch'ing, "Yuan tai te ju hu: Ju shih ti wei yen chin shih shang te i chang," Tung fang wen hua, 16 (1978), pp. 151—78; repr. in his Yuan tai sbih bsin fan (Taipei, 1983), pp. 1-58. SOCIAL CLASSES: TRADITIONAL AND NEW ELITES 629 or spirit of Chinese society underwent profound or enduring alteration, despite some observable influences for change (not all of which were dysfunctional or counterproductive). To be sure, many Chinese of the literate elite deplored the interruptions that Yiian rule brought to venerated patterns of social and political life, but at the same time they were strongly inclined, in their focus on ideal forms and on continuities, to overlook or discount the cumulative consequences of such interruptions. To repeat a point made earlier, we must read their reflections on their times with care and insight. Soon after the Mongols' conquest of the Chin in north China between 1215 and 1234, Mongolian governors in that part of China, under the distant leadership of the great khans Ogodei (r. 1229-41) and Mongke (r. 1250—9), began to devise systems for maintaining the distinction between conquerors and conquered and for exploiting the latter to support their Eurasia-wide military campaigns. "Census taking was the key to Mongolian efforts to mobilize the human and financial resources of the sedentary regions of the empire. The object of the registration was to facilitate the assessment of taxes, to identify skilled craftsmen and technicians, and to recruit military personnel."16 The idea of census registration for fiscal management was Chinese; the Mongols probably first encountered it in their contacts with the Khara-khitan (Western Liao) state founded in the early twelfth century by sinicized Khitans fleeing the Jurchen conquest of the Liao empire, or among the Uighurs of Chinese Turkestan who provided administrative experts to both the Khara-khitan and Chinggis khan early in the thirteenth century.I7 In China the Mongols made adaptations of their own governing methods and their borrowed administrative systems. In the earlier decades of their conquests in China, they applied principles common to their conquests everywhere: They regarded captive peoples and conquered nations as booty, as mere property, and so they enslaved them. Seizure and plunder on the field of battle were rights of Mongolian soldiers. Real estate and the people, animals, houses, and goods on it were granted as rewards to leading military commanders and associates of the ruler. At the highest level those formed appanages of the principal military associates, often enfeofTed as princes. Within the conquering hordes, the Mongols of course held the ruling positions, but they rapidly came to accept many associates in their empire building. Already by the death of Chinggis khan in 1227 the Mongolian war machine included many Inner and Western Asian peoples from non16 Thomas T. Allsen, "The Yiian dynasty and the Uighuts of Turfan in the 13th century," in China among equals: The Middle Kingdom and its neighbors, 10th—14th centuries, ed. Morris Rossabi (Berkeley and Los Angeles, 1983), p. 262. 17 Allsen, "The Yiian dynasty and the Uighurs of Turfan," pp. 246—8 et passim. See also Huang Ch'inglien, Yiian tai hu chi chih tu yen chin, pp. 1 2 8 - 3 5 . 63O CHINESE SOCIETY UNDER MONGOL RULE Mongolian nomadic steppe groups, from the great oasis communities of sedentary peoples, and from other states. Some had been conquered and forcibly incorporated, and others had voluntarily associated themselves with the Mongols. Many of those were granted privileged status, along with duties and obligations, as companions in the great tasks of conquest. In the Chinese records they are referred to as the peoples of the various nations (chukuojen) or people of varied categories (se-mu jen). Those terms are often loosely translated as "Western Asians," for in fact most of them were natives of the regions to the west of China. Of unusual importance among these Western Asians, numerically and functionally, were the Uighurs, who had gained a special relationship with the Mongols in 1209 by voluntarily declaring themselves vassals, before Chinggis khan subjected them to any form of coercion.18 Previously the Uighurs had created one of the great nomadic states, in the eighth and ninth centuries, when they were in close but hostile relationship to T'ang China. In the middle of the ninth century they were driven west into Turkestan (modern Sinkiang), where they underwent several kinds of transformations. They became sedentary peoples, farming and trading in the rich oases. They also began to give up their ardent Manicheism for Buddhism and Nestorian Christianity, and by the thirteenth century some were beginning to convert to Islam. The Uighurs brought with them considerable knowledge of Chinese administrative practices, becoming specialists in trade, finance, and civil governing and serving in the partially sinicized Khara-khitan (Western Liao) state as well as in their own territories. They also had become literate, having adapted to their Turkic language an alphabetic script derived from the Syriac. The Uighurs' close association with the Mongols as trusted allies and able warriors as well as masters of technical services demanding literacy meant that the Mongols were predisposed to rely on them for aid in governing China. Among the twenty or more groups covered by the classification se-mu jen, the Uighurs were preeminent. The se-mu jen, Western Asians, came to constitute a legally acknowledged second class in the Mongolian state. As that state absorbed, first, north China and then all of China, two more legally established class distinctions were created. Under the Mongols the conqueror—conquered differentiation was given a breadth of meaning and a declaration of permanence that it had not received under earlier conquerors. Within the twenty years after the Chin 18 Allsen, "The Yuan dynasty and the Uighurs of Turfan"; and Igor de Rachewiltz, "Turks in China under the Mongols: A preliminary investigation of Turco-Mongol relations in the 13th and 14th centuries," in China among equals: The Middle Kingdom and its neighbors, 10th— 14th centuries, ed. Morris Rossabi (Berkeley and Los Angeles, 1983), pp. 281—310. Uighur leaders had received the Mongols' favorable attention as early as 1204. SOCIAL CLASSES: TRADITIONAL AND NEW ELITES 631 state was conquered in the years 1215—34, twenty million or more sedentary people — primarily the Chinese of north China but also the Tanguts (Hsihsia), Po-hai (Parhae), Koreans, and others resident in the region - were brought under Mongolian rule. 19 These newly conquered subjects were designated, for administrative purposes, as the Han jen. To the Chinese that term meant (as it does today) themselves, the entire Chinese cultural or ethnic community, and of course most of the Mongols' subjects given this legal designation at this time were Chinese. But the term also was used to categorize all others who had been subjects of the Chin state in north China, whether Chinese or not and whether sedentary or not. 20 With the further conquest of the Southern Sung in 1275-9, a fourth classification was instituted - nan jen, southerners - to accommodate the fifty million or more Chinese who had been subjects of the Southern Sung. 21 This was the notorious system of four legal classes by which the Mongols, in the earlier years of their rule in China, tried to create by the promulgation of laws, a social order that ran counter to all the features of Chinese social structure and social ideology. No systematic proclamation of the fourfold, ethnically defined stratification - Mongols, Western Asians, Han jen, and Nan jen - seems ever to have been issued. Yet even before the conquest of the fourth and largest category, the tables of organization and the regulations for the conduct of civil governing worked out early in Khubilai's reign fully incorporated those distinctions and gave them legal force. Indeed, they held legal force until the end of the Yuan dynasty a century later. These regulations were applied with discriminatory effect in all matters regulating the lives of people relative to the state: They affected taxation, determined qualification for appointment to office, differentiated rights and privileges in adjudicating civil and criminal disputes at law and in determining penalties, established exemptions from liabilities, and were the basis for granting many kinds of privilege. Some of the advantages offered by this system to the favored two upper categories matched the privileges and favors that civil service official status had formerly given to scholar-officials in the Sung dynasty, but civil service official status was won through merit examinations. The Yuan system was imposed on all without regard to merit; it was hereditary; and in principle it allowed the individual or family no escape from an assigned status. 19 Earlier in this chapter, a figure of more than 50 million was given for the registered population of the Chin state in 1207; less than half that number can be accounted for in early Yuan census registrations. 20 Ch'ien Ta-hsin, Shib chia chai yang bsin lu (Pref. 1799; repr. Shanghai, 1935, 1957), 9, pp- 205—6, "Han jen pa chung," identifies eight ethnic groups included under the category Han jen. 21 Yao Ts'ung-wu, "Hu-pi-lieh p'ing Sung i hou te nan jen wen t'i," in vol. 7 of Yao Ts'ung-wu hsien sheng ch'iian chi, ed. Ch'en Chieh-hsien and Cha-ch'i Ssu-ch'in (Taipei, 1982), pp. 1—86. 632 CHINESE SOCIETY UNDER MONGOL RULE From the ruler's point of view, the system of four classes possessed its own rationality. It was an expedient way of maintaining the rulers' interests, of gaining assistance from and giving reward to reliable associates so that they would perform basic military and civil tasks, and of keeping the conquered in safely subordinate roles. This system appears to have come about through a series of ad hoc measures for dealing with the ever-larger set of pressing governing tasks while the Mongols continued their focus on conquests far and wide. It thus represents an advance over the stage of conquering for immediate plunder, as it displays a perceptive sense for giving to constituent groups in society clearly defined functions and obligations, rewards and duties. This system of four classes must be seen in relation to the highly complex household registration system put in place by the Mongols.22 That device both complemented and extended the fourfold system of ethnic social classes. It granted certain functional (occupational) household classifications to Mongols and all their subjects. Most important, it classified the conquered populations, especially the Chinese, in either the Han jen or the nan jen classes, according to their productive significance to the rulers. One scholar tabulated eighty-three household classifications.23 These classifications mainly designated specialized skills and productive functions and were designed to ensure that successive generations of the families so categorized would continue in those occupations. This was without precedent in the history of Chinese institutions. Here we see that a nomadic, wholly militarized tribal society, selectively using many kinds of advice and experience, was able to work out and supervise an elaborate system of social statuses and productive functions for the management of an entirely different kind of society. Even though the fourfold social class system and the interlocking one of occupationally differentiated household registration classifications were short range in social vision and ad hoc in their evolution, they represent a not-inconsiderable achievement. That should be assessed from the ground of Mongolian history as well as within the context of Chinese social history.24 One must, in any event, conclude that the Yuan design of imposed social classes and hereditary occupational statuses did not work well, probably could not have been made to work well, and was not rigorously maintained because its anomalies were obvious. But it was not abandoned. It frustrated the best administrators, irritated the governed, stood in the way of needed 22 Huang Ch'ing-lien, "Yuan tai hu chi chih tu yen chiu," pp. 13-18. 23 Ibid., pp. 197-216. 24 Meng Ssu-ming's Yuan tai she hut chieh chi chih tu (Peking, 1938) is now fifty years old and has not been superseded. It leaves many issues unresolved. See the comment by Hsiao Ch'i-ch'ing in his Yuan tai shih hsin fan (Taipei, 1983), pp. 43-4, n. 25. SOCIAL CLASSES: TRADITIONAL AND NEW ELITES 633 change, yet lingered to the end of the dynasty, and even influenced institutional developments in the succeeding Ming period. Our interest here, however, lies in understanding how all the special circumstances of Yiian rule affected the lives of China's people. The imposition by force of a two-tiered privileged elite composed of Mongols and Western Asians who monopolized the benefits of status and authority struck directly at the existence of the old Chinese elite of learning and cultivation and their traditions of political and social leadership. Their responses ran the gamut from bitter complaint and defiance to hesitant acceptance or passivity. Eventually, however, most of them accepted the Yiian rule, many going so far as to accept posts in government under the somewhat demeaning circumstances of being quite junior associates of their cultural inferiors. An interesting phenomenon in this Chinese elite adaptation is that the presence of an alien elite holding the real power did not eliminate the prestige accruing to literati in Chinese society, nor did it wholly dismantle the economic foundation of elite status among the conquered. That is, Chinese scholar-official types, even though they were cut off from holding higher office, continued to be accepted by the ordinary people as the leaders of local society. As far as one can see in contemporary writings, commoners' attitudes toward the "proper" claimants to social superiority were not greatly affected by the conqueror's new order. And rich Chinese whose wealth came from landholding or commerce — although subject to confiscation in the early decades of Mongolian rule in the north — by Khubilai's time in many cases found that their wealth was not notably less secure than it had been under the Sung. In fact, then, the new conditions of Yiian rule as systematized under Khubilai in the last decades of the thirteenth century permitted the coexistence of two elites, one de jure and one de facto. At first the former had many advantages, and the latter suffered great psychological and varying degrees of material deprivation. Yet the old Chinese elite - even in the more exposed north of pre-Khubilai times and still even less so in the newly conquered Sung south — were not subjected to any intent on the part of the conquerors to eliminate them and revolutionize the social order. The preexisting Chinese elite were forced to adapt to rude and uncomfortable circumstances; most did so, and some prospered. Yet there also were many who, as individuals and not in politically organized conspiracies, nonetheless derided the conquest and repudiated the Yiian dynasty and many more who chose the path of passive resistance and withdrawal, to become in some cases the subject of legend and myth. 25 A strong spirit of resistance was engendered in the first 25 See Mote, "Confucian eremitism in the Yiian period," pp. 2 0 3 - 4 0 . 634 CHINESE SOCIETY UNDER MONGOL RULE generation following the conquest in the north, and a still more marked display of loyalist sentiment is evident in the Sung territories that fell to Khubilai's armies in the 1270s. The most persistent but nonetheless unfounded myth is that of the "ten ranks" supposedly imposed on Chinese society at the time of Khubilai's conquest. According to this bit of folklore, Mongols and Western Asians were, of course, outside the system and superior to the Chinese. With officials and clerks assigned to the top two ranks, the scale descended to prostitutes in the eighth rank; Confucian scholars were in the ninth; and beggars were granted the tenth and lowest ranking in the society. Half a century ago it was clearly demonstrated that the de facto elite of the Chinese social order, while beset by economic uncertainties and psychological rebuffs, in fact maintained themselves as a superior stratum in society.26 They no longer had access to the civil service system of examinations for gaining status, office, and wealth. They were forced to cooperate with the de jure elite, displaying a variety of noble and less than noble motives. But the Chinese elite survived the Mongolian experience with their culture intact, if not wholly unmodified by the stresses and opportunities of the age. It also must be pointed out that even though the fourfold system of social classes did not eliminate the preexisting Chinese elite or attempt to reduce all Chinese to one debased economic level, neither did it ensure superior economic status for all Mongols and Western Asians. Despite their legally guaranteed privileges of many kinds, these two privileged classes displayed a full range of economic statuses, with many slipping into poverty, even into indebtedness to Chinese. Their richer members were quite naturally drawn into association with their Chinese counterparts, as their poorer ones suffered the same poverty as did the Chinese lower strata. A principal consequence of Mongolian discriminatory regulations, although not the intended consequence, was to induce a general condition for the Chinese elite of idleness or underemployment, without destroying its group consciousness and cohesion. The Mongols meant to keep the Chinese out of the highest positions in government and to keep the civil service from becoming an arena of uncontrolled Chinese action. Whether the de facto Chinese elite, denied those crucial supports to their status, would survive as an elite group in society was simply not of concern in Mongolian political strategy. Throughout the dynasty no more than a handful of Chinese ever achieved the higher substantive offices in administration. Eventually, fewer and fewer tried. The majority of those qualified by Chinese standards for holding office 26 Meng, Yuan tai she hui chieb cbi cbih tu, passim. CONFUCIAN HOUSEHOLDS 635 were discouraged by the milieu from devoting their lives to the civil service career. Dual staffing of offices gave superior rank and authority to Mongols and Central Asians, who were not required to demonstrate the objective qualifications that their Chinese inferiors normally possessed. The tone of public life under these circumstances did as much as restrictive regulations did to induce apathy in the Chinese elite. On the one hand, it distorted (from the traditional point of view) the makeup of officialdom, and on the other, it so altered the career prospects for self-identifying Confucian literati that many of them betook themselves into varied and unusual careers. Those who could afford to live in idleness often devoted their lives to art or scholarship, to good works, or to frivolity. Many more, of course, could not afford to live such lives. They had to find work. Their literacy could be employed in humble occupations as clerks, teachers, medical practitioners, fortune-tellers, or lesser callings. Some entered the Buddhist or Taoist clergy, if not often as ordained monks, then as lay associates of the local religious institutions. With the single career ideal of holding office "temporarily" suspended and its normative influences thereby weakened, the would-be scholar-official of Yiian times often found himself drawn into curious byways. CONFUCIAN HOUSEHOLDS Some of the elite of learning encountered the novel fate of being classified under "Confucian households" in the household registration system. The social duty assigned to that category of persons was solely "to study so as to be available for appointment." A somewhat anomalous phenomenon, this aspect of the Yiian dynasty's social system merits further discussion.27 Confucian households (ju-hu), in this technical Yiian period sense of the term, designated the households of individuals who were expected to constitute a functional category of mid-level service personnel, that is, members of a hereditary occupation obligated to provide specific services generation after generation. Such a conception of "Confucian household" was previously unknown. Dating back to the early decades of the Mongols' rule over north China, this classification was analogous to special categories already decreed for the Buddhist, Taoist, Nestorian, and Muslim clergy, as well as those for geomancers, prognosticators, shamans, and other religion-defined household classifications. The Mongols encouraged all the professional religious with favors, grants, and exemptions. When one sect gained a favor, others disputatiously claimed and usually won parallel treatment. But the Confucians did not regard themselves as members of a religious sect and tended to 27 The following discussion draws on Hsiao Ch'i-ch'ing, "Yiian tai te ju hu." 636 CHINESE SOCIETY UNDER MONGOL RULE disdain the professional practitioners of religions. Nonetheless, in the Chinese usage of that time, religions were called teachings {chiao), as were Confucian doctrines. To the Mongols this analogy was undoubtedly exact enough. And so this functional classification was bestowed on the Confucians on the advice given to the great khan Ogodei, under quite special conditions. As the Mongols completed their destruction of the Chin state in north China in the 1230s, a cruel fate descended on the population. Commanders victorious in the field, as we have seen, were allowed to seize property, livestock, and people, making slaves or bond servants of the latter, without regard to their social position. In 1238, observing the social loss in the failure to make more purposive use of human talent, the highly sinicized Khitan adviser at Ogodei's court, Yeh-lii Ch'u-ts'ai,28 proposed the addition of the "Confucian household" category to the household registration system. He had previously gained exemptions from taxes and services for them, also on analogy to the favored status of the Buddhist and Taoist clergy. The new household registration category confirmed those gains for them, an important tactical consideration at the time. It remained in force to the end of the dynasty 130 years later. At first a relief measure to protect men of learning at a time of great distress, this household registration category became a fixed part of the dynasty's institutional system. Yeh-lii's intention had been to use the special household classification as a first step and then to draw systematically on the literati that it protected to staff the administration throughout the entire government. He was not successful in realizing that larger objective. His influence at court waned, and after his death in 1243 the Confucian literati were for a time without a sympathetic spokesman of great influence. During the reign of the great khan Mongke (1250—9), Khubilai — who was Mongke's younger brother and future great khan — served as his brother's viceroy on the China front, with responsibility for governing north China. He was susceptible to Confucian persuading in this matter. In the early 1250s he issued regulations offering greater protection to Confucian households, and he sought to locate and release other men of cultivation who had languished in slavery since the Mongolian conquest of the Chin. The government's offer of significant exemptions from kinds of taxation and other privileges brought a rush of applicants for Confucian household status; rich families with no particular pretensions to learning bribed local officials, causing criticism among Mongolian officials and leading to the imposition of 28 Yeh-lii had previously served Chinggis khan but acquired greater influence under Ogodei; see Igor de Rachewiltz, "Yeh-lii Ch'u-ts'ai (i 189-1243): Buddhist idealist and Confucian statesman," in Confucian personalities, ed. Arthur F. Wright and Denis C. Twitchett (Stanford, 1962), pp. 189-216. CONFUCIAN HOUSEHOLDS 637 qualifying examinations. By that means almost four thousand households in north China were certified in 1276 for that status. By the time of the conquest of the Southern Sung in the years 1275-9, attitudes toward the Confucian literati were more favorable than they had been forty years earlier in the north. Orders were issued forbidding military commanders in the field from capturing and enslaving them. Local authorities were ordered to judge which households truly merited the special status, and the military authorities were required to follow their decisions. A decade or more passed during which further additions to the Confucian household category were made, as part of the process of absorbing the Chinese population of the Sung south. Finally, it was declared that all Confucian households so registered in the census of 1290 would remain in that hereditary status thereafter. Rough estimates indicate that as many as 100,000 such households were registered in the south, plus the 4,000 in the north, together considerably less than 1 percent of all registered households. Those are the outlines of the system as it developed through the earlier years of the Yuan dynasty. Confucian household registration presented many anomalies. Traditionally, Confucian scholars were proud to belong to families marked by "the fragrance of books generation after generation." Thus they could accept the idea of hereditary service, but they could find no Confucian principles for defending the inclusion of those not individually qualified or for excluding talented men not born to that status. Moreover, to be held comparable, in the eyes of the state, to monks and geomancers - to be looked upon as the clergy of an organized religion - demeaned them by association. The principal defect in the system, in the Confucians' eyes, was that it did not establish them as an elite with ensured access to high office. Although the system might work to certify certain individuals as possessing qualifications in some portion of classical learning so as "to be available for appointment," it did not provide that those who excelled in studies should automatically go on to serve in office. In short, it did not function as a substitute for the lapsed examination system. That was because throughout the dynasty, the important offices were largely reserved for Mongols and Western Asians, who either inherited their offices or enjoyed preference in appointments. The government needed a large number of Chinese skilled in the traditional paperwork of administration, but Chinese who were employed in those posts had limited career opportunities. Most of the educated sons sent to serve from those specially designated Confucian households had to bow their heads and take subofficial postings as clerks in local government or low official posts as local education officials, in either case having dim prospects for advancement. 638 CHINESE SOCIETY UNDER MONGOL RULE On the one hand, then, the Confucian household status carried with it certain economic advantages, although the amounts of stipend and types of tax and service exemptions varied unreliably. Those meager advantages could not be compared with the prospects for economic security that civil service careers had offered in Sung and Chin times. Confucian households were far from the bottom of the social ladder, but they also were far from the top. Their economic benefits were far from satisfying, given their self-esteem and sense of proprieties, and the psychological rewards were pitifully insignificant. Therein lay the root of so much elite bitterness. When the examinations were finally reinstituted in 1315 (after being announced in 1313), the literati, whether registered in Confucian households or not, were euphoric. A way out for them, and the proper way at that, would at last be provided. But the civil service examinations during the remaining years of the Yuan, held triennially on sixteen occasions between 1315 and 1366, awarded only 1,139 chin-shih degrees (although the quota of 100 per year would have allowed 1,600). According to the regulation, half of those were to be awarded to Mongolian and Western Asian candidates who took simpler examinations judged by lower standards. Including those, the averge production oi chin-shih degrees was only about 23 new ones per year, a small fraction of the average for Sung and Chin times. Moreover, the Yuan examinations were so tainted by collusion and cheating that self-respecting scholars tended to avoid them. In short, the restitution of the examination system, however important it was to bringing ambitious aliens into the Chinese cultural modes, and despite the cheer it initially brought to Chinese literati ever hopeful that civilization was triumphing, did not materially alter the gloomy prospects of Confucian scholars. It is not surprising that many men of literary and scholarly talent sought careers elsewhere and often pursued ways of life that would have been most irregular in other ages. DIFFUSION OF ELITE ROLES Much of the cultivated talent in Chinese society, for the reasons that have been indicated, was forced to find avenues for less-than-ideal expression, mostly apart from government and at humble levels of society. This group may have formed a quite large talent pool. One might calculate that they numbered perhaps 500,000 educated men, on the assumption that, with their families, they may have constituted 5 percent of the total population.29 29 For purposes of a loose calculation, one may use the low figure of 60 million Chinese in the reign of Khubilai. One might calculate the total number of persons in the families of the cultivated elite as follows: If 5 percent of the total population, or 3 million people, belonged to elite families averaging 6 persons per household and if one of those was an adult male, that would produce a figure of 300,000 adult male members of the elite. This is merely a plausible estimate, unsupported by any verifiable data. DIFFUSION OF ELITE ROLES 639 The civil service examinations had produced their last Chin dynasty chin-shih in 1233 in the north, and in 1274 in the Sung south. The chin-shih examinations were not resumed until 1315, and then on a much smaller scale. The civil service examinations both recruited talent for official employment and held out the goal guiding career preparation to all would-be elite members of society. Those who passed the examinations and gained official status thereby were, in all times, only a select number from a larger group of similarly prepared men. After coming under Mongolian rule in the midthirteenth century, leftover chin-shih, as well as tens of thousands of qualified expectants, had little access to appropriate employment. Yet the talent pool to which their high education gave them entry continued to grow. The high value their society placed on that kind of education plus the confidence that it would someday again be utilized continued to draw aspirants for literatus status into the time-honored path of scholarly preparation. This condition was unchanged by the conquest and its aftermath. Members of the de facto elite continued to be produced; what else was there for young men in such families to do? The state's lack of concern for traditional ways of establishing social leadership, if anything, intensified the Confucian commitment to education. Much private energy of otherwise frustrated "scholars of high purpose" was devoted to teaching. 30 A self-perpetuating elite of education, despite dwindling fortunes, political apathy, and the need to be ingenious in devising survival techniques, nonetheless was on the scene and in a position to make unusual contributions to society that, under the abnormal Yuan conditions, provided novel outlets for their talents. The impact of the traditionally defined elite on the life of the entire society thus was different from that of other ages. First, because they were not drawn off into isolating high positions in government, they could live more closely with their larger kinship groups, and they shared more closely the life of the community, whether they resided in rural or urban settings. If, for example, for reasons of economic self-interest or considerations of physical protection, the Confucian literati sought some kind of relationship with the local Buddhist temple, they became involved with the religious life of the community. If they were forced to teach elementary students for a living, or did so out of a sense of duty when otherwise they would have had to remain idle, they became involved in the lives of those ordinary families who could give a son such schooling. If they had to assume the tasks of lowly clerks in local offices of government, they saw government from an angle of vision usually denied 30 Yan-shuan Lao, "Southern Chinese scholars and educational institutions in early Yiian: Some preliminary remarks," in China under Mongol rule, ed. John D. Langlois, Jr. (Princeton, 1981), pp. 1 0 7 - 3 3 . 640 CHINESE SOCIETY UNDER MONGOL RULE to persons qualified for elite status in society, and they brought their kind of historical and philosophical knowledge to bear on jobs that normally were performed without access to such intellectual resources. If these Confucians had skills as painters or calligraphers but now had to sell their art to earn their living, they came to recognize the conditions of the marketplace and the uses that non-aesthetes had for art. If management skills that could not find their outlet in administration had to be turned to commerce, the world of commerce stood to grow in certain kinds of sophistication, as its elite entrepreneurs grew in practical knowledge. In short, the shocks that the abnormal conditions of Yuan life had on the elite surely did bring suffering to many individuals and certain kinds of social loss to the larger society, but they also created a great potential for new growth and change. The difficult tasks of reconstructing these aspects of Yuan social history are only now beginning to attract scholarly attention. THE MEANING OF THE YUAN DRAMA IN YUAN PERIOD SOCIAL HISTORY The history of Yiian drama (tsa-chii) offers a revealing sidelight on the interaction of elite and popular culture and on roles new to Yiian times that the Chinese elite of cultivation could play in fostering that interaction. The history of Yiian drama has long been misinterpreted, used to sustain myths about the extent to which the de facto elite of Yiian times suffered humiliation and deprivation. As early as the fourteenth century, Chinese writers began to embroider the notion that the drama suddenly emerged full blown from the minds of impoverished scholars who, unable to take their proper roles in society, were forced to earn their living by writing vulgar entertainments for their crude Mongolian and Western Asian masters. According to this interpretation, the unprecedented concentration of displaced and economically desperate geniuses in an activity in which their profound cultural resources were perforce turned to inventing a vehicle explains the sudden burst of masterpieces for the popular theater. The partial truths in that explanation make it diabolically appealing, but it is essentially wrong. It is true that drama as a literary art form developed rather late in Chinese history and then suddenlyfloweredwithin a rather short space of time. All of the foremost masterpieces of the tsa-chii form of drama (loosely known as the Yuan drama) were produced by a small group of talents within the space of about two generations.31 Moreover, the dramas do indeed present the dual 31 Chung-wen Shih, The golden age of Chinese drama: Yiian tsa-chii (Princeton, 1976), pp. 3-19; James I. Crump, Chinese theater in the days of Kublai khan (Tucson, 1980), pp. 3-30. THE MEANING OF DRAMA IN Y0AN SOCIAL HISTORY 641 and, to Chinese traditional minds, the mutually exclusive elements of high literary value and popular entertainment. This phenomenon demanded special explanation, and the traditional explanation, exaggerated through the centuries, was accepted uncritically until early in the twentieth century. Then, a period of reevaluation of China's literary heritage plus greater Chinese and Japanese interest in the place of the Mongols in East Asian and world history led to a reassessment of Yuan drama. This phase of reinvestigation culminated in Yoshikawa Kojiro's Gen zatsugeki kenkyii (Research on Yuan drama), which in addition to providing strictly literary analysis, also examines the issues of authorship and audience as elements in Chinese social history.32 Yoshikawa stressed the interactions among writers, actors, and audience, including alien and Chinese patrons, in bringing Yuan drama to a sudden flowering. The history of Yuan drama's hitherto-ignored antecedents has also been more fully revealed, in part because archaeological evidence has demanded a closer study of documentary records. It is now clear that the process by which Yuan drama flowered in the thirteenth century had been one of longer standing and that the art form was ready to flourish with or without the Mongolian conquest.a Reassessing that history, a writer who, like Yoshikawa, is interested both in literary and in social history, sees in the special circumstances of the Yuan period the explanation for much of Yuan drama's character and the causes of its further development: There has been, in the past two decades, enough archaeological evidence unearthed to show that a long popular dramatic tradition has existed in China since at least the eleventh century. This evidence refutes the earlier statements that the Mongols were responsible for the rise of drama through their dispossession of the literati. It also refutes the assumption that it was the literati who brought the dramatic form to its height. What seems more plausible is that drama, as a self-sustaining and self-developing tradition, had by the thirteenth century simply matured to the point that it offered a suitable and attractive vehicle for literary expression. Under the Mongols, the elite writer of the north found himself adrift in an unfamiliar world, cut off from social and political success and denied the peer respect that the traditional forms of literature would have provided. Given the time and the opportunity to consort with actors on an intensive and prolonged basis, he began to participate in dramatic activities, writing scripts and perhaps even performing. Thus, the elite writer from this transitional period onward played a major part in developing drama from a purely performing art to a written literature.34 In social terms, the most significant consequence of Mongolian overlordship may have been the temporary diffusion and dispersion of the Chinese elite 32 See Yoshikawa Kojiro, Gen zatsugeki kenkyu (Tokyo, 1948, rev. ed., 1954), pp. 72-241; and Cheng Ch'ing-mao, trans., Yiiati tsa chii yen chiu (Taipei, i960), pp. 44-162. 33 Idema, Wilt and Stephen H. West, Cbinae theater 1100—1450: A source book (Wiesbaden, 1982), pp. 1—94, et passim. 34 Stephen H. West, "Mongol influence on the development of northern drama," in China under Mongol rule, ed. John D. Langlois, Jr. (Princeton, 1981), pp. 434—65. 642 CHINESE SOCIETY UNDER MONGOL RULE social roles. The social history of the Yuan drama offers one illustration of that. 35 To extend this argument, one can see parallel developments in all the elite spheres of activity. Contradictory developments affected two polar groups within the old elite. Those lacking independent means often were forced into broader interaction with the lives of common people and into creative roles in mass culture in an age that was conducive to innovation. This was true not only in drama but also in thought and popular religion, in government at the lowest levels in which men of elite qualifications would not normally have served, in extending applied technology (as in agriculture and medicine), and no doubt in other ways that scholars have not yet identified and studied. By contrast, however, some of the old elite having secure wealth - especially true in the less disturbed south China - and disdaining the trends of the times and vulgar associations, became perhaps less relevant to the daily life of their own society. They could not serve in government or would not, and so they lacked the practical Confucian impulse to lead responsible public lives. These are the hypothetical extremes; many more examples of the former kind than of the latter come readily to mind, but both were present. In their two different ways, these two types of displaced Yuan literati elite, and the entire spectrum of elite society, produced an enormous harvest of art and literature. The special conditions produced by Mongolian overlordship in some areas of cultural development exerted a constructive force. During the long reign of Khubilai khan in the last four decades of the thirteenth century, the Mongolian capital at Ta-tu (modern Peking) was built on a vast scale, exhibiting dazzling features of city layout and architecture, supported by hydraulic engineering works of a high order. The alien elite of Mongols and Western Asians included Persians and Arabs versed in astronomy and mathematics and other fields of knowledge from Middle Eastern and Levantine centers. Some of that knowledge was exchanged with Chinese counterparts in the Yuan capital, where the worldview of the rulers was not confined to the Chinese past. Of still greater import in the minds of Chinese then and since has been the fact that Mongolian rule reunited the long-separated northern and southern parts of the Chinese cultural sphere. Trends in thought, classical scholarship, statecraft, and science as well as those in calligraphy, painting, and poetry had diverged and acquired distinctive qualities. These divergent trends now were drawn into the matrix of the cosmopolitan Mongolian capital and spilled over into the great cities of the provinces. Yuan China provided a far broader spectrum of elite — and ex35 Shao Hsiin-cheng, "Yuan tai te wen hsiieh yii she hui," T'u shu yiieh k'an, 3 (1943); repr. in Yuan shih lun ts'ung, 1 (1982), pp. 221—4. Shao's views anticipate those of more recent writers. CULTURAL DIVERSITY 643 elite — social patterns than was normal for China. At the same time, it provided an arena for kinds of stimulus and interaction that the Chinese elite had not experienced since the great days of the T'ang. Much of that, as we have noted, was not valued by the elite of the time and was denigrated by succeeding ages, but may be assessed anew in our time. For social history, the Yuan as an interval between the Sung period and the Ming, despite the shocks it delivered to the traditional elite, demonstrates the tenacity of China's social structure and of its normative ideals. The "temporary" diffusion of elite roles continued to be regarded as merely temporary, as improperly irregular, for more than a century. A persistent, virtually unshakable view of what society should be helped sustain it through a century of makeshift compromises. All the pressures, both the intended and the incidental, of the Yuan era failed to deflect China from the course of social development on which it had long been set. In ways not yet fully appreciated, they also induced fruitful responses to unusual conditions. CULTURAL DIVERSITY Cultural historians cannot but be intrigued by the remarkably diverse elements of Eurasian cultures that the pax Mongolka allowed to assemble in China in the thirteenth century. The Mongolian emperors welcomed trade and traders (as Chinese rulers did not), sought out craftsmen in all the regions they traversed, patronized all the religions they encountered, and eagerly employed assistants in governing in their multiethnic, pan-Eurasian civil service. Much of that diversity lingered on in China into the fourteenth century, after the division of the Mongolian world into four autonomous khanates and the waning of the brief general peace that the early great khans had imposed on their vast realms. By the fourteenth century, travelers from the farthest western extensions of the empire to the easternmost domains in China became far fewer, but communities of non-Chinese remained behind in the Chinese capital and throughout the provinces. The religious communities included Arab and Persian Muslims as well as the more recently Islamicized groups from east of the Pamirs; both Nestorian and Roman Christians from Inner Asia all the way to the Levant; Manicheans; Jews; Buddhists of diverse non-Chinese backgrounds; and various kinds of Siberian and East Asian shamanists. Representatives of virtually all the peoples then residing east of the Pamirs had communities in China at that time, and several from west of the Pamirs were also known. Many alphabets as well as three variants on the Chinese ideographic script (Khitan, Jurchen, and Tangut) were in use, and texts written in them 644 CHINESE SOCIETY UNDER MONGOL RULE were read by people residing in or traveling through China. Spoken languages then in use in China included almost all the Sino-Tibetan and Altaic languages that historians think then existed, as well as the important western Asian and some European tongues. Such variety of costumes, manners, rituals, foods, arts and crafts, techniques, philosophies, and doctrines as might be encountered in China of Khubilai's time had not been seen there since the great days of T'ang Ch'ang-an in the seventh and eighth centuries. What were the cultural and social consequences of that contact with diversity? Specific consequences such as distinct new currents in taste, new styles in art, music, literature, or architecture or new movements in native religion or thought appear to have been marginal and are at best vaguely suggested. The Chinese elite of this period cannot be said to have been truly cosmopolitan in the sense of taking an interest in alien peoples, ideas, and things for their own sake. There is not much writing from the period that explores the exotic and displays delight in knowing about it. On the contrary, much of the writing carelessly confuses one distant place with another, mixes up Manicheism with Nestorianism or with Buddhism, or cites a foreign word then in current use but says that there is no way of knowing what it really means. Solutions to such problems were not at all beyond the reach of truly inquisitive minds. Perhaps if the rulers of China at that time had been Chinese whose military exploits had brought all the forms of diversity into China, men of learning would have reached out to understand them. Instead, with few evident exceptions, it was all part of the phenomenon of alien rule, something to be borne, not delighted in. Whatever the reasons, we see little impact of the Yuan period's alien cultural resources on Chinese life of that time or of subsequent ages. THE WESTERN ASIANS' CHANGING RELATIONSHIP TO THE CHINESE ELITE This discussion of elite life in Yuan times cannot conclude without first taking note of the fate of those Western Asians (se-mu jen) who held de jure elite status in Yuan society. If the Chinese were not much interested in whatever contributions those interesting aliens may have made to their civilization, they were deeply interested in the sinicization of the foreigners. That interest has continued to the present and with little difference in attitude. An eminent modern scholar wrote: From the time when the Liao, Chin, and Sung observed only a fitful peace, communications between north and south China were interrupted for three hundred years. Beginning with the Yuan, the doors were thrown open, and tens of thousands of square WESTERN ASIANS A N D THE CHINESE ELITE 645 miles of territory on our northwest were added to the domain. Se-mu people also lived in Chinese territory without restriction. As the culture of our land was open to all, and as the people from the western regions admired it, they took to it unconsciously.36 That statement, written in the 1920s, could have been written at any time from the Yuan period onward; it is totally sinocentric. Although its author criticized those Chinese of the past who failed to appreciate the cultural achievements of sinicized foreigners under Yuan rule, he himself was interested in the impact of China on the aliens, and not in the reverse of that. The nonetheless valuable work of historical scholarship from which those lines are quoted studies the sinicized lives and accomplishments of 133 men and women of se-mu background who lived in Yuan China. What interests the author particularly is the persuasive force of Confucian (to a lesser extent of Chinese Taoist and Buddhist) ideas per se, but this does not fully represent all the dimensions of those persons' acculturation. That is, they were capable of becoming culturally dual or plural, able to live in the Mongolian- or Turkic- or Persian-speaking worlds of their rulers and associates, and, eventually, also able to function in cultivated Chinese society. That they were attracted to Chinese civilization and, in some cases, fully mastered its content and its forms has always seemed to demonstrate the superiority of China. Quite naturally, this has been a satisfying observation for Chinese of the Yuan period and since. Moreover, there is a large measure of objective truth in it. But modern historians can also see other kinds of meaning in that record. There is no doubt that the pull of China's civilization on the peoples of the steppe was very strong. Throughout the two centuries of their Liao dynasty (906-1125), the Khitans managed to have their tribal aristocracy become in large part culturally dual while their common people remained outside that cultural process. Yet despite their intellectual sophistication and mastery of Chinese high culture, the Khitans remained "people of the steppe," not significantly detached from the values of their nomadic life. The Jurchens who conquered them, as well as most of north China, at the beginning imitated the cultural duality of the Khitans, but only very briefly. The Khitans had only a foothold on the edge of sedentary life; the Jurchens ruled almost half of China. The Jurchens quickly transformed their way of life, adapting enthusiastically to Chinese norms and in the process losing their own language and essentially all of their original culture, even though retaining some Jurchen institutional features and their warrior ethos. 36 Ch'en Yuan, Western and Central Asians in China under the Mongols: Their transformation into Chinese, trans. Ch'ien Hsing-hai and L. Carrington Goodrich, Monumenca Scrica Monograph no. 15 (Los Angeles, 1966), p. 287. 646 CHINESE SOCIETY UNDER MONGOL RULE Of all the northern neighbors and conquerors of the Chinese, the Mongols were the least affected by sedentary civilizations, whether they encountered sedentary peoples in China or elsewhere in Eurasia. They displayed in China little elite acculturation on the Khitan pattern," much less the wholesale assimilation of the Jurchens. This kept them in some ways strong, yet it also made them more dependent on se-mu assistance to mediate between them and their Chinese subjects and to perform the routine tasks of governing. For almost a century after Chinggis khan, few Mongols in official ranks learned to speak Chinese, and still fewer learned to read and write, although that number was increasing late in the dynasty. To the end of the dynasty there were Mongolian and se-mu officials serving in China who were illiterate in Chinese. Three of four of the last Mongolian emperors from Jen-tsung (r: 1312-21) onward were in some measure literate in Chinese,'8 but they did not, however, establish that as a model for their people. Mongolian cultural self-confidence and a deep attachment to the values of steppe life, not a lack of the capacity to learn, kept them apart from the seductions of China's civilization. The se-mu jen, Western Asians, were given high status and great leeway in administering; the Mongols could trust them, knowing they were fully dependent on their masters for their positions in China. Ample opportunity to acquire wealth did not give the se-mu jen independence. Their situation was difficult. On the one hand, they had to accomplish assigned tasks in order to satisfy their masters' wishes; on the other hand, they had to learn how to be effective in Chinese society in order to accomplish those tasks. They were true middlemen, of necessity in constant communication with their superiors and with their inferiors, alternately on the cultural ground of both. Polylinguality was common among them, rare among both Mongolian overlords and Chinese subjects. As a hypothetical example, a Uighur official in the Ministry of Revenue had to use Mongolian at court, perhaps Persian or a second Altaic language among his fellow se-mu colleagues both at the office and socially, and he would find it most useful if he could read Chinese administrative documents or at least converse with his Chinese clerks directly. The latter condition may not have been common in the thirteenth century, but it did become the norm in the fourteenth century. Ch'en Yiian's study just quoted shows that 130 or more Western Asians went much further than merely becoming functionally literate in Chinese; 37 But note the study of incipient Mongolian Confucianization by Hsiao Ch'i-ch'ing, "Yuan tai Meng-ku jen te Han hsiieh" (cf. n. 13). 38 That has been demonstrated independently by both Herbert Franke and Yoshikawa Kojiro; see Herbert Franke, "Could the Mongol emperors read and write Chinese?" Asia Major, 3 (1952), pp. 2 8 41; and Yoshikawa Kojiro, "Gen no shotei no bungaku — genshi sosetsu no ichi," in vol. 15 of Yoshikawa Kojiro zenshu (Tokyo, 1969), pp. 232—303. WESTERN ASIANS AND THE CHINESE ELITE 647 they acquired an outstanding command of traditional Chinese cultivation. Their numbers included painters and calligraphers, poets and dramatists, classical scholars and model scholar-bureaucrats. They must be seen as the most noteworthy examples in that respect among the Western Asian sector of society resident in China; less noteworthy examples whose achievements have not been recorded but who were able to function on Chinese cultural ground must have numbered in the thousands. One can see that as the decades went by the intermediary group drew closer to the de facto Chinese elite in sympathies, special skills, and cultural ideals. By the mid-fourteenth century the Mongolian court obviously had become an unreliable guarantor of the Western Asian people's futures. Return to their Inner Asian or Western Asian homelands must have been for many a less attractive alternative, and so they wholeheartedly adapted to the idea of becoming Chinese. To their credit, the Chinese were generally open-minded about that, even when they chose to interpret it as proof of the superiority of Chinese civilization. To many of the aliens caught in this dilemma, it may have represented the better choice among less than perfect alternatives, a practical-minded but not an ideal solution. That the children and grandchildren of immigrants, especially of economically favored immigrants, should excel in their adopted homeland does not demand special explanation from late-twentieth-century observers of our world. That the Chinese of past times have found it intrinsically fascinating, as well as a strong confirmation of their cultural pride, also is quite understandable. In 1368 the Mongols accompanying the court departed before the advancing Ming armies. A smaller number remained behind permanently, most of those in specially approved garrison communities where they could exist as unassimilated Mongols in military groups under their own commanders, subsidiary to the new Ming armies.39 But by far the greater number of those Mongols who departed from China to rejoin their compatriots in the steppe did so with their way of life essentially unchanged by their experience in China. We do not know much about what happened to the Western Asians in China at that time. Some retreated with the Mongolian court to Khara Khorum, but many, perhaps most, remained in China. Merely by adopting Chinese names, they could set their families on the way to being wholly absorbed, for in their physical appearance they were not strikingly distinct. A few were prominent in the period of the Yuan-Ming transition, earning 39 How the Mongols departed to reconstitute their society in Mongolia is not clear; there was no orderly withdrawal en masse, but the presumption must be that most of them eventually returned to Mongolia. See Henry Serruys, The Mongols in China during the Hung-wu period, Melanges chinois et bouddhiques no. 11 (Bruges, 1959), chaps. 5, 6 et passim. 648 CHINESE SOCIETY UNDER MONGOL RULE commendations from the Ming founder for exemplary Confucian behavior in loyal service to the doomed Yuan dynasty. They were held up as models to shame Chinese who performed less well in that stressful period. What does not emerge from the record is to what degree these Western Asians may have retained elements of their original cultures, using those to enrich the Chinese life around them. Yet it is clear that these two elite groups - the de jure elite of the se-mu peoples and the Chinese de facto elite - originally mutually suspicious if not openly hostile and originally on different cultural footings, had gradually drawn closer together, eventually merging into one elite sector of society. We must assume that a similar process was simultaneously under way among the common folk of both societies. SOCIAL CLASSES: SUB-ELITE AND COMMONER Military households The effort of the Mongols to alter fundamentally the Chinese social order, through the division into four legally denned classes and through the workings of the household registration system fixing hereditary occupational classifications, has been discussed. Particularly noteworthy was their effort to create eighty or more hereditary occupational groups encompassing the entire society, which were in effect intended to be closed occupational classes, set apart from one another and each held in its own special relationship to the government. Two groups were of particular strategic concern to the Mongolian state: One was the military households, and the other was the group of household classifications for artisans and skilled craftsmen. The military profession and the artisans merit further discussion, as they offer a glimpse into the lives of ordinary people, and the experiments with these strategically important social groups reveal the problems of social management that faced the Mongols in China. Long before the Yuan period the principle that the military arm of China's civilian government should be constituted of civilian conscriptees drawn from society at large was firmly established. Yet no reliably workable and consistent system for accomplishing that was ever developed — in sharp contrast with the increasingly effective system of training and recruitment for staffing the bureaucratic civil government. The Sung dynasty, ideologically antimilitary yet beset by the most demanding military needs, was one of unending experimentation with professional guard troops, conscript armies, border garrisons occasionally using non-Chinese soldiers, and village-based militia forces. To the end of the dynasty the Sung had problems with conscription systems in particular and was plagued by widespread evasions, deser- SOCIAL CLASSES: SUB-ELITE AND COMMONER 649 tions, and administrative corruption. Thus although we can say that there existed a long-accepted principle of the civilian-soldier who served in the armed forces under civilian, nonprofessional control, the Sung experience displayed all too clearly that there was a striking dearth in the Chinese tradition of successful models for its operations. The succession of steppe invaders who constituted the ultimately fatal threat against which the Sung struggled for three centuries were precisely the opposite: They were from militarized societies with weak, almost nonexistent civil institutions. Military status for their warriors, virtually all the males between fourteen and fifty-nine, was hereditary. Among China's historic enemies the Mongols in particular had realized the highest degree of militarization ever seen in human history, with the entire society mobile and deployed for war and every human and animal prepared to contribute directly to military goals. They recognized few professions other than that of fighter, and they knew themselves to be the world's best fighters. Thus it is not surprising that they would believe their model of the hereditary military profession to be successful and would attempt to extend it universally to other societies under their control. This feature of Mongolian military strength matched an institutionally insoluble Chinese weakness. Perhaps it is thus not surprising that after living for a century under the Mongolian attempt to duplicate their military system within the society of China, the Chinese who took over the reins after expelling the Mongols in 1368 would be inclined to retain this system. The Mongolian system provided four categories for military households, or chun hu.*° The first was labeled simply "Mongolian," and the second one was for so-called tammachi military households, that is, Mongols and associated steppe nomads from the southern rim of Mongolia.41 The other two categories were for the Han-chun (essentially the Chinese who came over to the Mongolian side in the pre-1250 phase of the conquest) and the hsin-fu-chun (newly adhered armies) absorbed at the time of the conquest of Southern Sung in the 1270s. Their status and treatment were different. During the century of militarily imposed rule over China, the Mongols retained the hereditary military categories comprising the vast bulk of their own society, and they extended similar categories to the bulk of the Inner and Western Asian populations absorbed during their conquests, many of whom were brought into China. They also drew heavily on the Chinese population 40 For a detailed analysis, see Ch'en Kao-hua, "Lun Yuan tai te chun hu," Yuan shih lun u'ung, I (1982), pp. 7 2 - 9 0 41 There has been considerable scholarly investigation of the tammacbi armies. Recent publications include Yang Chih-chiu, Yuan ihih san lun (Peking, 1985), pp. 1-66; and Chia Ching-yen, Tan-mach'ih chiin k'ao," Yuan ihib lun ti'ung, 2 (1983), pp. 23—42. 65O CHINESE SOCIETY UNDER MONGOL RULE for army recruits and organized analogous military sectors of the Chinese population to accommodate them.42 As early as 1232, Ogodei's field commanders in north China were authorized to organize local Chinese warlord armies into the Mongolian command, designating them chiliarchies or myriarchies, thus organizationally analogous to their Mongolian counterparts, and establishing the distinction between civilian (min) and military (chiiri) households for the conquered population. According to official statistics for 1241, the military households in north China at that time constituted one-seventh of the total population, and in some districts the proportion was as high as one-third.43 Obviously, Chinese men at arms organized into the so-called Han-chun units were a significant component of the Mongolian army in China, from the initial conquest of north China onward.44 Some exceptional Chinese in the north who had served in the Chin cavalry or who had experienced the frontier warfare against steppe armies were in many ways equal to the Mongols; similarly socialized by analogous experience, they could fight alongside them as peers. But they were not typical of Chinese soldiery. Most were village farmer conscriptees. The surrendered Sung armies in the 1270s, the "newly adhered armies," were absorbed into the Mongolian military machine, provided with Mongolian or Western Asian commanders, and deployed to the various fronts in East Asia on which Khubilai campaigned. The Mongols would have been unable to meet their manpower needs without them. Khubilai's navies, essential to the conquest of Southern Sung in the 1270s and to the attempted invasions of Japan and Indonesia, were made up entirely of surrendered Chinese and Korean forces. In some regions local militia also were taken into the military system and either retained in their regions for policing purposes or added to major field campaigns. All of these types of Chinese-manned armed forces were classified under the two categories of Chinese military household; each household was required to supply one fighting man on active service at all times, generation after generation. Military households were favored. They received certain exemptions from taxes and service obligations and drew stipends and grants of money and food. Among commoners in the Chinese population, these households were a relatively privileged group. At the same time however, the household was required to equip and support the soldier on duty and to replace him if he deserted, was killed, or died; this could be a heavy burden. 42 Hsiao Ch'i-ch'ing, The military establishment of the Yuan dynasty (Cambridge, Mass., 1978) is the standard work on the military organization of the Yuan dynasty; the following discussion is indebted to it at many points. 43 YS, 98, pp. 2 5 0 7 - 2 2 , esp. p. 2510, as quoted by Ch'en Kao-hua in "Yuan tai techiinhu,"p. 73, n. 9. 44 Sun K'o-k'uan, Meng-ku Han chin chi Han wen hua yen chiu (Taipei, 1958), pp. 1—5. SOCIAL CLASSES: SUB-ELITE AND COMMONER 651 The central features of this military system were hereditary status, separation from the society at large in their relation to the government, and direct subordination to the Bureau of Military Affairs (Shu-mi yuan) through regional and local military officials. There is administrative logic in this way of solving the military needs of the time if one overlooks the nature of Chinese society. The system, once in place, was kept in force through the century of Yuan history. Its defects, obvious as time wore on, were frequently criticized, but no essential changes were made. This institutional rigidity was characteristic of many aspects of Yuan history after Khubilai's reign. From the beginning, this military system ran counter to the ideological and organizational tendencies of Chinese society. Chinese men seldom wanted to be soldiers, and when restricted to that disdained career, they often sought ways to escape. The favors accruing to their professional status did not outweigh the liabilities. Initially, and through the conquest of the Southern Sung, the system served to supply large numbers of soldiers at minimal cost in money and management. Thereafter, under the sharply declining administrative effectiveness after Khubilai's reign, it also became clear that the Yuan military machine, although it continued to function more successfully where Mongolian units were involved, was by and large undermanned, badly trained, undisciplined, and lacking in proper equipment to the point of being useless. Late in the dynasty, local warlord armies, some nominally loyal to the regime and others in open rebellion against it, repeatedly demonstrated that the Mongolian government no longer commanded a reliable force with which to coerce them or the population at large. General disorder and civil war ended the dynasty. It had been founded by means of military superiority, and it fell because of military weakness. A number of factors contributed to that failure. One was the Mongols' uncertainty about how far to accept non-Mongolian soldiers or generals as full partners. In the civilian sphere Chinese officials could bear the main burdens of daily work but were not admitted to the highest ranks of decision-making officialdom. The infiltration of Chinese into the military, even as lowerranking officers and soldiers, was a still more sensitive issue. The Mongols did not wish to create among the Chinese population any military force that might encourage a centrifugal military tradition, one whose interests could diverge from Mongolian priorities. Before Khubilai's reign, when the problems of governing a sedentary Chinese population had not yet assumed high importance, Chinese warlords were given full acceptance as able associates; they enjoyed high rank and trust as long as they demonstrated military abilities. During Khubilai's reign, the emphasis was on institutional regularization. The most powerful northern Chinese warlords and their myriarchies were brought under closer control. 652 CHINESE SOCIETY UNDER MONGOL RULE Some, to be sure, retained their hereditary positions to the end of the dynasty, but their freedom of action was greatly curtailed. After Khubilai's reign, suspicion of Chinese and discrimination against them in high places increased, and accordingly, Chinese military units and leaders were granted less acceptance.1" During the reign of Shun-ti, the last emperor, who ruled from 1333 until the end of the dynasty, the case of Wang Pao-pao well illustrates the court's suspicions of Chinese military men. Wang was the son of a Chinese father and a Turco-Mongolian mother, a daughter of Chaghan Temiir who had made himself the warlord of Honan Province.46 Wang Pao-pao became Chaghan's adopted son and heir, eventually succeeding to command the powerful Honan regional army. The emperor granted Wang Pao-pao the Mongolian name Toghto in recognition of his loyalty and important service to the throne. Wang as Toghto went all the way in assuming Mongolian identity and in displaying unalterable loyalty to the Yuan cause while remaining independent of court politics. In the decade of military struggles leading to the founding of the Ming dynasty, Chinese rivals could not buy this man nor could they defeat him. Indeed, the Ming founder honored Toghto posthumously as his most worthy opponent, although some Chinese literati scorned him as a cultural turncoat. Although Toghto was the principal military supporter of the Yuan in its last phase, the Mongols' suspicion of him as a Chinese, an outsider, dominated the court's deliberations on how to use his indispensable military support, keeping him at a distance and failing to use him properly. There was, in fact, no secure place for most Chinese military associates of the Mongols, not even for a Wang Pao-pao who was half steppe warrior by blood and wholly Mongolian by self-identification. The Chinese subsequently were more successful in accepting and utilizing Mongolian military groups in early Ming society, although those groups were in most cases, whether by choice or design, communally separate for some generations.47 The Yuan system for establishing hereditary military households in Chinese society failed, in part, because the Mongols felt that they must maintain a distance separating themselves from their conquered sedentary populations 45 In his analysis of the military households, Ch'en Kao-hua wrote: "One basic principle of Yuan government was to promote racial discrimination and to create separation and contradiction between the ethnic groups. The Mongol army units, the tammachi armies, and the Han [Chinese] armies all served as instruments of the state's control; nonetheless, both in their levels of compensation and in their deployment, there were wide discrepancies in the generosity shown to them and in their closeness [to the rulers]. The Mongol government exerted its fullest efforts to preserve the special status of the Mongol and tammachi armies." See "Yuan tai te chiin hu," p. 78. 46 John W. Dardess, Conquerors and Confucians: Aspects of political change in late Yuan China (New York, 1973). PP- i 3 2 - 4 6 . 47 Henry Serruys, The Mongols in China during the Hung-wu period, passim. SOCIAL CLASSES: SUB-ELITE AND COMMONER 653 and thus would not effectively integrate them into their military machine. That was possible because the Chinese military units were never the sole, or even the main, component of the Mongolian army. The Mongols benefited from a mystique of invincibility that lingered long after it was deserved. But the major part of the explanation is that their system, modeled on Mongolian institutions, was imposed on Chinese society without regard for congruence. On the one hand, the alien rulers could not create on command the structural conditions that would have been necessary to make it work in the way it did in their own nomadic society. And on the other hand, Chinese society, unless restructured, could not absorb that model of military organization into its dominant and enduring cultural patterns. A third element in its failure reflects a general condition hastening the rapid decline of the Yiian dynasty in the fourteenth century: In both the Mongolian and the Chinese armed forces, the management became utterly corrupted and debased, subjecting common soldiers and their families to unbearable exploitation. They responded by deserting in large numbers or by refusing to fight. They no longer were soldiers. Artisans Artisan households presented somewhat different problems. The Yiian household registration system provided a number of different artisan (kung, chiang) classifications. Some designated specific specializations in production, such as ceramics, smelting, weaving and dyeing, wine and oil making, and salt production. In the Chinese language the designation of artisan was not extended to skilled workers in all productive activities, for example, fishing, milling, food processing, and some others, although the registration system also established separate household classifications for those. In short, the system did not grow naturally out of Chinese concepts or linguistic usages; it was conceptually Mongolian, reflecting their sense of social functions. This is particularly apparent in the case of three classes of artisans {chiang) that were specially favored in the Yiian system and given differential status to bring them under close government supervision. The three of greatest importance to the Yiian rulers were the categories of artisans for construction or engineering works, for those who produced items for military use, and for those who supplied government offices and produced luxuries for elite consumption. They were designated as government artisans, military artisans, and civilian artisans.48 All were specially favored by exemptions and grants of rations, salaries, and other forms of material support. 48 Huang Ch'ing-lien, "Yiian tai hu chi chih tu yen chiu," pp. 81—3. 654 CHINESE SOCIETY UNDER MONGOL RULE These artisan households were comparable to the military households in the degree to which their favored treatment set them apart from other commoner households. Many people sought artisan household status; indeed, their numbers had to be restricted, and proof of competence occasionally had to be demonstrated. It was well known that in times of warfare, Mongolian army commanders in the field were under rigorous orders to identify and protect all kinds of artisans when populations otherwise were subjected to slaughter or enslavement. Accordingly, in such circumstances many ordinary people claimed to be artisans to save their lives. The numbers of artisans taken in battle thus would appear to be inflated. One historian calculated that the number of artisans employed in government workplaces midway through Khubilai's reign was 400,000.49 Western Asians as well as Chinese were registered in artisan households. Artisans were taken captive in the early Mongolian campaigns in Central Asia, when great cities famed for their crafts, such as Bukhara and Balkh, were captured, and some of those craftsmen were brought eastward. Most of those in China, however, had been specially sought out in the Chinese population. The bureaucratic management of skilled workers was carried out on a vast scale, and they became one of the more important components of Yuan government.50 The so-called government artisans were attached to various government agencies, to produce their products or construct buildings, bridges, fortifications, and the like, as ordered. The military artisans were attached to the military registers and so were not under civilian control. They produced mainly weapons, armor, wagons, siege machinery, and so forth. Civilian artisans were under jurisdiction of the civil authorities, mostly under the Ministry of Works; it appears that their numbers included workmen skilled in many fields, those of military significance as well as others. All three groups were required to perform assigned tasks for a certain number of months each year, or they were assigned quotas of work to be performed. After that they could hire themselves out for pay or make products for sale. Many details of the system remain unclear, though further research could add considerably to our knowledge. The management of artisans under the Yuan dynasty system clearly reflected the place of artisans in Mongolian society before their conquests of any sedentary peoples. The few skilled artisans native to Mongolian society were held in the highest regard. The outstanding example is the blacksmith, whose skills were crucial to the maintenance of the armies. Blacksmiths typically 49 Chii Ch'ing-yiian, "Government artisans of the Yuan dynasty," in Chinese social history: Translations of selected studies, trans, and ed. John de Francis and E-tu Zen Sun (Washington, D.C., 1956), pp. 234-46. 50 Sun K'o-k'uan, Meng-ku Han chiin chi Han wen hua yen chiu, pp. 66—74. SOCIAL CLASSES: SUB-ELITE AND COMMONER 655 traveled with armies and also served as fighters. Early in the period of Mongolian rule in China, artisan households from Western Asia and the Chinese population were required to supply a quota of men who could also serve as soldiers.' 1 Within the societies they conquered, the Mongols displayed little regard for farmers or scholars; they honored the professional religious of all kinds, but in particular they valued the artisan classes for their strategic relationship to the conduct of war. Skilled craftsmen also were the source of less essential but eagerly sought luxury goods: furs, textiles, jewelry, leather products, fine steel objects, and the like. Thus the Mongols were in the habit of classing artisans as prize booty even before they entered China. In China the system under which artisans were organized invited corruption and evasion and other forms of misuse. One example from the base level of society illustrates this. The family of the future Ming founding emperor, Chu Yuan-chang, had been registered as a "gold-panning household," one of the artisan household classifications. They then lived in southern Kiangsu, near modern Nanking, where in the fourteenth century there was little gold production. They appear, rather, to have been professional gold panners who washed gold from ore-bearing sands, and they were obligated to produce and present to the government a certain quantity of gold dust each year. Failing to locate sources of such gold but unable to escape the household classification and its obligations, they had taken in desperation to tenant farming in order to exchange agricultural products for gold in the marketplace. But merchants with official protection manipulated the price of gold in order to extort more money. The head of the Chu household, like many in similar circumstances, found it necessary to flee with his family into the Huai River region north of the Yangtze. There, as migrants liable to capture and prosecution, they were all the more susceptible to exploitation by landlords who took in the family as sharecroppers; they had to work vacant land desolated by war and disaster. The history of this family through the two generations before the birth of Chu Yuan-chang himself in 1328, in a destitute household forced constantly to move about in search of livelihood, exemplifies the evils of the system. This story of a desperately poor family can be reconstructed only because it produced a future emperor. Otherwise we know very little about the impact of the household registration on the lives of ordinary artisans or, indeed, on the lives of the farmers who constituted the vast bulk of the population. Nonetheless, the little that we do know strongly suggests that the system did not fit the needs of society, that it was neither economically rational nor socially equitable. If those qualities are not prominent in the social histories 51 Ibid., pp. 66—7. 656 CHINESE SOCIETY UNDER MONGOL RULE of most countries in the thirteenth and fourteenth centuries, they nonetheless are more apparent in Chinese social history in both pre- and post-Yuan times. Curiously, the emperor of the succeeding Ming dynasty who had experienced the inequity of the system in his own family and who made discerning efforts to improve the lot of China's poorest farmers nonetheless retained the Yuan concept of hereditary military and artisan households, attempting to keep them separately registered and tied to hereditary occupational status. Otherwise the Ming did not maintain the elaborate differentiation of households imposed in the Yuan; instead, they established essentially four classifications: official, commoner, military, and artisan.52 Only the latter two were hereditary, and in the case of each, the system broke down, soon failing to provide the necessary manpower for the army or for the government workplaces. The Ming founder's judgment in this was faulty. One must assume that the special circumstances of breakdown and disorder in the late Yuan had deprived him of experience with Chinese society under normal conditions, thereby distorting his judgment. During the Yuan period, however, the systems for administering artisan households functioned quite differently from that for military households, and so the two must be evaluated separately. The Yuan government had nothing to fear or to lose from sponsoring the widest development of the artisanate; it did not threaten any Mongolian prerogatives. The lives of artisans saved in warfare repaid the conquerors and the entire society, many times over. The broadly conceived organization and the large numbers of skilled workers deployed contributed directly to the regime's readiness for war and to its capacity to build and rebuild cities, restore public works, and provide lavishly for its elite. This organization undoubtedly helped preserve and diffuse skills, maintain traditions of craftsmanship, and enrich society. To be sure, these men's skills might have been better preserved and diffused, and society made richer still, if there had been no alien conquests. But given such incursions as historical realities, one must be grateful for whatever amelioration of disaster the Mongolian focus on craftsmen provided. It clearly strengthened the Mongolian regime, and it also helped Chinese society survive. 52 Salt-producing households (tsao hu) also were a hereditary occupational group whose specialized productive function was highly important for revenues and the Ming defense system. See Ray Huang, Taxation and governmentalfinancein sixteenth-century Ming China (Cambridge, 1974), pp. 189—224. It should be noted, however, that Wang Yuquan, in a forthcoming study, shows that the early Ming employed an elaborate system of more than 80 household categories, showing the influence of the Yuan model. YOAN DYNASTY SOCIAL HISTORY 657 OTHER ASPECTS OF YOAN DYNASTY SOCIAL HISTORY Cities North and south China underwent distinctly different patterns of social change following the Chin conquest of the Northern Sung in the early twelfth century. Differences between north and south China stemming from the ecology have always been present. During periods of national unity, the impact of those differences on patterns of social life and cultural expression has tended to diminish. From the tenth century onward the differences in the natural setting were coupled with those caused by the alien incursions, the dislocations of warfare, altered networks of trade, and changes in the modes of government reflecting the interests of alien rulers. The Jurchen Chin conquest of the entire north in the 1120s was the culmination of two centuries of such disturbances; it deepened the division of north and south China that was finally overcome by the Mongolian reunification in the 1270s. When the Mongolian dynasty at that time reunified China by force of arms, there were striking differences between the economically declining north and the flourishing south - by which is meant the "historical south" of the Yangtze drainage area in central China and the southeast coastal provinces. Population shifts are the most obvious indicator of those differences. The north, which had included the demographic center of China well into T'ang times, now held at best one-third of the population. It supported much less regional trade, was served by its poorer and much more expensive land transport routes, and now contained far fewer cities of large size. The Mongolian ruler, to be sure, had built a magnificent capital at Ta-tu (modern Peking), and one designed according to classical Chinese imperial plan," but it probably held at most 500,000 inhabitants. Although that put it in the first rank of world cities at that time, it was not as large as the Sung or earlier capitals in Chinese history. The former Southern Sung capital at Hang-chou remained at that time the greatest city in China, with a population close to 1 million.54 Other than Ta-tu, no cities in Yuan north China had populations in excess of 100,000. The former Northern Sung capital at K'ai-feng had been in the range of 1 million in the eleventh century but had dwindled after its abandonment by the Sung court early in the twelfth century. In the late thirteenth century the entire prefecture in which it was located had a registered population of only 185,000 persons. 53 NancyS. Steinhardt, "TheplanofKhubilai khan'simperiaJ city," Ar»;A/u Afi<K, 44(1983), pp. 137—58. 54 Gilben Rozman, Urban networks in Ch'ing China and Toiugawa Japan (Princeton, 1973), pp. 30—6. This work offers a systematic survey of urbanism in Chinese and Japanese history. 658 CHINESE SOCIETY UNDER MONGOL RULE Below the Huai River line that separated north and south China from Chin dynasty times until the Mongolian reunification, there were many great Yuan period cities with populations from 250,000 to 500,000 and some that were larger. Separate population figures for cities apart from the county or prefecture by which they were administered usually do not exist, but the 1290 census figures for the prefectures in which major cities were located tell the story: Yangchow's prefecture had a registeted population of 1.5 million; nearby Ch'ang-chou Prefecture had 1 million; Chia-hsing (adjacent to Hangchou in northern Chekiang) had 2.25 million; several other Chekiang and Kiangsu prefectures had from 1 million to 2.5 million registered inhabitants; and for Jao-chou Prefecture in northern Kiangsi, which included the great ceramics center at Ching-te-chen, we find a (scarcely credible) figure of more than 4.5 million inhabitants. The population densities indicated by such figures not only indicate the probable size of urban concentrations, but they also can be taken as an indicator of labor available for crafts and industry, of specialized agricultural production for commercial distribution, and of welldeveloped water transport by inland waterways and coastal shipping to maintain the dense populations. The core region of central—south China appears to have remained economically well integrated, as we know it had been in the Southern Sung period preceding the Mongolian conquest." The Japanese social historian Shiba Yoshinobu calls attention to the "medieval Chinese commercial revolution" that elsewhere has been described as a "fundamental transformation" of Chinese society between A.D. 750 and 1000, in which the striking feature is the "enormous expansion of trade," accompanied by "progressive urbanization and the emergence of an identifiable urban class with its own sub-culture."'6 The development that had transformed China in the immediate pre-Mongolian centuries was not sustained in north China after 1125, but it continued apace in the Sung south. The questions before us with regard to cities, urban culture, trade, and the commercialization of the agrarian sector of the economy are to what extent the reunification of China helped the north regain momentum during the Yuan and whether Yuan rule depressed the economy in the south. The Mongols favored the ortogh, merchant associations of Western Asians that were given both trading privileges and fiscal duties in service to the state. Ortogh merchants operated throughout the breadth of the Yuan empire from Persia and the Mediterranean to China and Korea; they may have 55 Shiba Yoshinobu, "Sung foreign trade: Its scope and organization," in China among equals: The Middle Kingdom and its neighbors, loth-iqth centuries, ed. Morris Rossabi (Berkeley and Los Angeles, 1983), pp. 89—11;. See also Shiba Yoshinobu, Sodai shogyo shi tenkyu (Tokyo, 1968) or the abridged translation by Mark Elvin, Commerce and society in Sung China (Ann Arbor, 1970). 56 The quotations are from Denis C. Twitchett, "Merchant, trade and government in late T'ang," Asia Major, 14 (1968), p. 6 3 . YOAN DYNASTY SOCIAL HISTORY 659 contributed to a more international character in China's long-range trade at this time. Did ortogh privileges disrupt trading patterns in China, or did they contribute to the overall growth of commerce and add to China's wealth? It is not possible to give full answers at this time.57 The reputation of the ortogh in Yuan China is one of base collusion with the Mongolian overlords whose capital, wrung from the exploited Chinese population, was then lent to the ortogh tofinancetheir, at best, shady operations that harmed government and people. The Mongols and Western Asians in high positions clearly did lend money to the ortogh merchants, who in turn lent it at usurious rates to units of local government that could not otherwise meet tax payments or to individuals facing similar financial needs, and then they relied on their special relation with the governors to collect their debts. The merchants' reputation for unbridled avarice may be exaggerated, but ordinary persons seem to have regarded them as the cause of much general suffering. Descriptive comments of the time often note that the Western Asian ortogh merchants "understood the ways of cities," where commerce was conducted, and ruthlessly used those skills to fatten their own purses and those of their Mongolian masters. There are some descriptions of life in great cities such as Ta-tu and Hangchou in Yuan times. Hang-chou greatly attracted Mongolian and Western Asian officials, religious personnel, and merchants to its scenic beauties and mild climate, its luxury and entertainment. The high levels of urban sophistication for which it was famous in the last decades of the Sung survived the conquest.'8 Marco Polo came to know the city well during his years in China, 1275 to 1291. He called it "the most noble city and the best that is in the world." But that great city suffered destructive fires in the mid-fourteenth century and exchanged hands several times in the civil wars of the late Yuan. That it underwent a decline throughout the dynasty seems clear. No other cities give evidence of having experienced notable expansion. The reopening of the Grand Canal undoubtedly had a beneficial impact on the domestic economy. Operations had not been affected in its southern portion, from Hang-chou to the Yangtze, but the canal had to be reconstructed in northern Kiangsu, and it was extended north from the Yellow River to reach Ta-tu. It was damaged by floods and intercepted by warfare from the 1340s onward. The economic integration of north and south was at first aided by the canal's reconstruction but then failed to gain full and lasting benefit. Large trading 57 Two new studies of the ortogh appeared as this chapter was being prepared. See Thomas T. Allsen, "Mongolian princes and their merchant partners, 1200—1260," Asia Major, 2 (1989), pt. 2, pp. 83— 126; and Elizabeth Endicott-West, "Merchant associations in Yuan China: The ortogh" Asia Major, 2 (1989), pt. 2, pp. 1 2 7 - 5 4 . 58 The city is magnificently described in Jacques Gemet, Daily life in China on the eve of the Mongol invasion, 7230-1276, trans. H. M. Wright (New York, 1962). This work first appeared in French as La vie quotiditnne en Chine, a la vielle de I'invasion Mongole, I2}o—i2j6 (Paris, 1959). 660 CHINESE SOCIETY UNDER MONGOL RULE cities along its northern portion revived but did not flourish. No sector of the economy appears to have genuinely flourished during Yiian times, but there was a wide variation in regional conditions. Some of the most direct and colorful information about the life of urban China in Yiian times comes from the Yiian dramas (tsa-chii), whose social history we have already discussed. Many of those depict the lives of ordinary persons in familiar settings, and although their plots may be fanciful, the dramas nonetheless reveal a wide range of urban pursuits, show the interaction of urban and rural life, expose us to the cycle of festivals and religious observances that punctuated the year, and depict the attitudes toward rich and poor, toward officials, and occasionally toward aliens then resident in China. Some of the dramas have been well translated, but their analysis as a resource for the study of Yiian society is as yet only beginning.59 The same is true of the great compendium of case law and precedents, the Yiian tien chang, and other kinds of documentation of Yiian social history.60 Rural life If our knowledge of urban society is inadequate, the situation with regard to the lives of the vast majority of all Chinese, the village-dwelling farmers and small craftsmen and traders who constituted the rural sector of society, is even less satisfactory. The dramas, informal notes and sketches (pi-chi), and other unofficial and informal kinds of historical source materials are less frequently informative about rural life than about urban life, or about the elite in various settings. Nonetheless, the possibility exists for developing a much more detailed understanding of some aspects of rural society than has yet emerged. The base level of government was built on two systems of local management derived from earlier Chinese practice. At the beginning of Khubilai's reign, his government ordered that the various categories of households be brought under local supervision. In cities and towns and their immediate suburbs, as well as in rural areas, the households were organized into what we might call wards, or precincts, for which a variety of Chinese names were used, differently designated in different parts of the country. Their heads, nominally chosen by acclamation, appear in fact to have been selected with county government concurrence from among the largest taxpayers. Their 59 See George A. Hayden, Crime and punhbmtnl in medieval Chinese drama: Three Judge Pao plays, Harvard East Asian Monograph no. 82 (Cambridge, Mass., 1978). This work uses courtroom dramas to explore concepts of social justice. Many of the Yiian dramas have now been translated. 60 For a comment on studies of the Yuan tien chang, see my "Note on traditional sources for Yiian history" in this volume. HOUSEHOLDS IN BONDAGE 66l duties were to supervise tax collection, labor duties, and other levies on the population, as well as to be responsible for order. Early in his reign, Khubilai ordered that rural households of all occupational categories, but principally farming households, be organized into she communes, nominally of fifty households each. The she system was extended to south China after the conquest of the Southern Sung, but bits and pieces of evidence indicate that the she varied widely in size and might include several hundred households. At the beginning, nominally at least, the she units were responsible for supervising and promoting agriculture and rural rehabilitation. Eventually, the she units tended to be amalgamated with the wards and precincts of the tax-collecting and peacekeeping base-level organization. There is suspicion that neither system was universally applied or effectively maintained. Yet both show that the patterns of communal organization more or less natural to the society were acknowledged and utilized for social management purposes, much as they had been used under previous dynasties. 6 ' To most Chinese farmers, petty tradesmen, or craftsmen in rural areas, the hereditary household categories and the fourfold system of social classes probably were less demanding impositions on their daily lives than they were for urban and suburban populations. HOUSEHOLDS IN BONDAGE The aspect of Yuan rule that most directly affected rural life was the attachment of total local populations — sometimes in numbers running to thousands or even tens of thousands of households — along with their farming lands, to conquering army leaders. The Mongolian practice was for conquered peoples to be included among the kinds of war booty that could be seized during military operations or to be granted to the appanages of imperial relatives and meritorious officers, removed from the regular population registers, and placed in a status like that of serfdom or slavery. The term for that category of persons was ch'ii-k'ou (meaning something like "captives").62 this practice was most widespread in north China early in the conquest, but it continued in some degree through the reign of Khubilai and was extended into the southern provinces. 63 In 1235, immediately following the Mongolian conquest of the Chin dynasty, an attempt was made to count and register households. The motive 61 See Elizabeth Endicott-West's chapter on government in this volume and also her study Mongolian rule in China: Local administration in the Yuan dynasty (Cambridge, Mass., 1989) pp. 119—22 et passim. 62 For a collection of impressionistic materials on ch'ii-k'ou and persons held in other kinds of bondage, see Li Kan, Yuan tai she hut ching chi shih kao (Wuhan, 1985), pp. 38—58. 63 Examples from Khubilai khan's reign are given in Li Tse-fen, "Yuan tai te she hui," in vol. 5 of his Yuan shih hsin chiang (Taipei, 1978), pp. 348—528, esp. pp. 506—11. 662 CHINESE SOCIETY UNDER MONGOL RULE was to determine the number of households claimed infiefby the appanages of the military leaders and to fix uniform tax and service schedules for the entire population. The number of households then held by appanages was found to be in excess of 760,000, out of a total registered population of roughly 2 million households.64 The Yuan court attempted to place the administration of the appanages under officials of its appointing, in order to share in the revenues drawn from the "captive" households, butfiefholders tended to regard those as their private property, to conceal actual numbers, and to exploit their productive capacities to the fullest extent. Some of the north China appanages were held by Chinese military leaders {Han-chiin) who held their fiefs in hereditary right, just as the Mongols and Western Asians held theirs. The central government's interests conflicted with those of their military nobles in the exploitation of the conquered Chinese population, a conflict that continued throughout the dynasty. The ordinary people had no way of protecting themselves in this conflict of interests, except tofleetheir homes and become vagrant households or wandering individuals. A large portion of the reduction in north China's registered population between 1215 and 1260 thus must be explained by such large-scale internal migrations. Khubilai khan and his advisers saw the seriousness of both the competition with nobles over the control of captive populations and the disruption of rural life caused by the excessive exploitation of them. They attempted to implement broad measures to curb excesses and to stabilize the agricultural villages. Through the she system of promoting agriculture and the wards and precincts for tax collection and security, just described, Khubilai's government attempted to reduce and to administer directly the rural households assigned to appanages and to make their tax and service obligations similar to those of free farming households. Yet Khubilai khan could not abandon the Mongolian institution of the appanage with its attached serf and craftsmen households. The compromise was to appoint supervising administrators to the appanages and to grant annual payments from the revenues to the fief holders. That did not fully succeed in eliminating evasion and exploitation and other points of conflict with the central government. During Khubilai's long reign, about 15 percent of the population still was registered as attached households, and the concealed numbers may have been equally high. Abuses of the ultimate sources of revenue, the producing farmers and craftsmen, continued through the dynasty. In managing its rural population, the Yuan government retained forms of taxation and service obligations established before 1260 in the north and 64 See the table in Han Ju-lin et al., Yiianch'aoshih(Peking, ing discussion. 1986), vol. i , p . 2 2 2 , and the accompany- HOUSEHOLDS IN BONDAGE 663 applied a different system, the two-tax system inherited from the Sung dynasty, in the south. This is evidence of both the government's realistic flexibility and its administrative weakness. That weakness, obvious to all, encouraged the abuses of landlordism and truancy throughout China, but particularly in the richer southern provinces. The economic circumstances of rural China remained poor. Under Khubilai some gains were made in agricultural rehabilitation but were again lost under his weaker successors. This is attested in various ways, particularly by the many references in contemporary documents to large numbers of migrant people (//» miri).6' One other window on this aspect of Yiian history is opened by the extensive accounts of popular religion and the associated rebellious movements during the late Yiian. Messianic doctrines, remotely Manichean in derivation, predicted the imminent triumph of the Buddha Maitreya who would appear at the moment the world reached its darkest and its people were the most desperate. Millions of rural Chinese, mostly from north of the Yangtze River, were persuaded to believe that the world of mid-fourteenth-century Yiian China was on the verge of descending to that nadir and that therefore it soon would be transformed into golden, blissful radiance once again. Paradise would be realized on earth. One implication of such doctrines' success is that rural life was indeed very harsh, lending credibility to the idea that the worst had been attained and that its sudden reversal was at hand. This idea made masses of the ordinary people susceptible to a leadership claiming some affinity with the Buddha Maitreya, and many such leaders so portrayed themselves. Another implication is that the traditional Confucian-minded elite, even though their prestige in the eyes of society remained intact, could nonetheless exert only a weakened local leadership, so that large numbers of ordinary people were less subject to elite normative influences. One feature of the late Yiian rebellions is that the elite most commonly sided with the government in its ineffectual suppression of the sectarian revolts. Other purely secular rebellions of the desperate farming population also arose in great number, many of them starting as no more than movements of community self-defense against banditry that the government was unable to suppress. Study of the fourteenthcentury rebellions has progressed greatly in recent years, a synthesis that will reveal the contours of rural society more fully and describe the actual conditions of daily life may soon be at hand. 66 65 Ch'iu Shu-shen and Wang Ting, "Yiian tai hu k'ou wen t'i ch'u i," pp. 116—18. 66 See Chapter 7 in this volume and also my discussion of the popular rebellions leading to the founding of the Ming dynasty in "The rise of the Ming dynasty, 1330—1350," in The Ming dynasty, 1368—1644, pt. 1, vol. 7 of The Cambridge history of China, ed. Frederick W. Mote and Denis C. Twitchett (Cambridge, 1988), pp. 1 1 - 5 7 , *sp- PP- I 2 ~ 4 3 - 664 CHINESE SOCIETY UNDER MONGOL RULE We must return to the point with which this chapter opened: We face the dilemma of being unable to state how many people - within the largest tolerable margins of error - inhabited China in Yuan times or of knowing how to account for the large decreases indicated by the difficult-to-interpret figures available to us. As long as the answers to a question of such basic importance are still not within our grasp, we can at best raise other questions about the entire range of historical issues, survey the recent scholarship to which it is relevant, and propose summary statements of limited, temporary usefulness. Yuan history intrigues us because there seems to have existed in that important segment of time a conjunction of influences, of currents making for social change, that might be expected to possess great explanatory value for our understanding of Chinese history. BIBLIOGRAPHICAL ESSAYS I . THE LIAO Traditional sources The Liao is a particularly poorly documented period. Its government, like all Chinese bureaucracies, produced a mass of paperwork, written in both Chinese and Khitan. But little of this documentation survived the fall of the dynasty, and nothing remains today. The peculiar nature of the Liao government was an important factor in the poverty of the historical record. Although it supported court diarists and a historiographical office, its historians never achieved the smooth routines and professional competence of their successors under the Chin, let alone of their Sung contemporaries. One reason for this was the fact that even until the end of the dynasty the Liao never had a static capital, palace, court, and government on the Chinese model. The Khitan court remained peripatetic, the emperors never abandoning their annual tours around their empire and their annual visits to the four seasonal camps (na-po). This style of government was not conducive to keeping orderly state archives. Nor was the personal arbitrary style of government at every level, and the Liao's fragmented administrative structure, divided into northern (tribal) and southern (Chinese) bureaus, the former keeping some of their records in Khitan and the latter exclusively employing Chinese. The Liao court employed court diarists1 who, as at a Chinese court, kept court diaries {ch'i-chii chu) recording the emperor's acts. These diaries were supposed to be kept secret, but emperors sometimes interfered, demanded to see what the diarists were recording about their administration, and punished them if they refused.2 By the end of the tenth century the Liao also maintained a Historiographical Office (Kuo-shih yuan) on the T'ang model, with a director and three compilers, some of them Khitan and some Chinese.' They regularly recompiled the court diaries into a daily record. In the reign of Sheng-tsung (982—1031) they also began compiling "veritable records" (shih-lu) for preceding reigns; thefirstwas presented to the throne in 991.«In the reign 1 T'o-t'o et al., eds., Liao sbih (Peking, 1974) (hereafter cited as IS) 47, p. 776. 2 LS, 23, p. 278. 3 A Khitan named Yeh-lii Lu-pu-ku held the title of director of the national history in T'ai-tsung's reign, but it is doubtful if an office of historiography existed so early. See LS, 47, p. 781. 4 LS, 13, p. 141; LS, 79, p. 1272. 665 Cambridge Histories Online © Cambridge University Press, 2008 666 BIBLIOGRAPHICAL ESSAYS ofTao-tsung (1055—IIOI) the compilation of a "national history" was commissioned, and in 1085 the Historiographical Office presented to the throne veritable records for all the reigns down to that of Hsing-tsung.' These together provided a chronological history of the dynasty from its founding to 1055, and this compilation probably presented the first systematic account of the early reigns. After Tao-tsung's death, his successor T'ien-tso (1101—25) commissioned a continuation extending the veritable records down to his own accession in 1101. This was completed in 1103 under the direction of Yeh-lii Yen and contained seventy chapters.6 This work survived until early Ming times. No veritable record was ever compiled for T'ien-tso's reign. The Liao emperors also commissioned at least two histories of the pre-dynastic Khitan. The first was a record of the dynastic ancestor Ch'i-shou khaghan, which was completed in 941 and may have been written in Khitan.' The second was a twentychapter history of the imperial line written in Chinese in 1044 by Hsiao Han-chia-nu, the director of the national history.8 Hsiao Han-chia-nu was anxious to give both his ruler and the members of the Khitan elite a sense of history in the Chinese fashion, and so he translated a group of Chinese historical works into Khitan: the Tung li, a general history by the ninth-century author Ma Tsung; the (Chili) Wu-tai shih of Hsiieh Chiicheng, then the standard history of the Five Dynasties; and the Chen-kuan cheng-yao of Wu Ching, a somewhat idealized account of the discussions between the T'ang emperor T'ai-tsung and his ministers.9 This last work seems to have appealed greatly to non-Chinese rulers from the north, who could readily empathize with the personal, collegial style of government it described. It was later translated into Tangut, Jurchen, Mongolian, and Manchu. The historians and scholars at the Liao court were also aware of historical writing in Sung China, in spite of strict Sung restrictions on the export of books.10 In 1096 a Chinese courtier at the Liao court protested to Tao-tsung about the insulting treatment given to the Khitan in Ou-yang Hsiu's Hsin Wu-tai shih, in which they are simply included among the barbarian peoples, despite the Sung court's having recognized the Liao as their equals. He suggested that the Sung be given the same treatment in the Liao's own dynastic record." The conquest of the Liao by the Jurchen Chin, who were later far more thoroughly subjected to Chinese cultural influences, made it natural that they should undertake 5 6 7 8 9 10 LS, 24, p. 290. LS, 27, p. 320; 98, p. 1416. LS, 4, p. 49. LS, 103, p. 1450. LS, 103, p. 1450. After 1006 the Sung government embargoed all sales of books at the border markets, apart from the Confucian canonical works and their, commentaries. Individuals who broke the ban were to be punished, and the books were confiscated by the authorities. See T'o-t'o et al., eds., Sung shib (Peking, 1977), 186, p. 4 ) 6 2 ; Li T'ao et al., comps.,Hsu Tzucbiht'ungchiench'angpien (Peking, 1979), 64, p. 1423. The Liao had an equally strict ban on the export of books to Sung. Offenders were liable for the the death penalty. See Shen Kua, lAmg-ch'i pi-t'an (Hsin cbiao-cbeng Meng ch'ipi I'an) (Peking, 1958), 15, p. 160. See also Karl A. Wittfogel and Feng Chia-sheng, History of Chinese society, Liao ( 9 0 7 1125), Transactions of the American Philosophical Society, n.s., vol. 36 (Philadelphia, 1949) (hereafter cited as Wittfogel and Feng), p. 302, n. 4 1 . 11 LS, 104, pp. 1455—6. Cambridge Histories Online © Cambridge University Press, 2008 BIBLIOGRAPHICAL ESSAYS 667 the compilation of a standard dynastic history for their Liao predecessors. But the question was complicated by disputes over whether the Liao could claim to have been a legitimate dynasty. Although the Liao had acquired all the outward forms and status of a Chinese-style imperial dynasty and had been recognized as their equals by the Sung court, it never controlled more than a small margin of northern China. Even after two centuries of continuous contact with Sung China and the absorption of large numbers of Chinese subjects, its rulers remained only partially sinified. It was a difficult question for the historiographers whether Liao should be treated as a marginal frontier regime and thus be dealt with in a supplement to the histories of the Five Dynasties and the Sung, or whether it should be considered as a full-fledged legitimate northern dynasty, separate from the Sung and deserving a dynastic history of its own, as were the Northern Wei and their sixth-century successors. This question remained unresolved until the fourteenth century and even then was settled only by the arbitrary decision of a Mongolian chief minister, who overruled the ideological prevarications of his orthodox Chinese historians. The first attempt to write a Liao history came under the Chin emperor Hsi-tsung (1135-50) who ordered Yeh-lii Ku, a descendant of the Khitan ruling family, to write a history of Liao, a task that he turned over to a disciple and fellow Khitan, Hsiao Yung-ch'i. This history, in seventy chapters, was completed in 1148." It was not, however, found to be satisfactory, and from 1189 until 1206 the Chin Historiographical Office was engaged in a full-scale official project to compile a Liao history. The work was constantly interrupted by political bickering, and the history, still incomplete, was eventually presented to the throne by Ch'en Ta-jen in 1207. •' Under the Yuan, emotional discussions about the legitimate status of the Liao again delayed the undertaking of a dynastic history, until finally in 1343 the chief minister Toghto (T'o-t'o) was placed at the head of a commission entrusted with preparing dynastic histories of the Liao, Chin, and Sung, and he put a stop to these interminable arguments. The Liao history that the commission produced, the existing Liaoshih, was completed in little more than a year in 1344 and printed in 1345. ' 4 Because it remains our main source for the period, a few points about it need to be made. Unusually for a standard dynastic history, the Liao shih was written more than two centuries after the Liao had come to an end, after two highly disruptive changes of dynasty and at a time when the official archives of the Liao government had long been destroyed. It is most unlikely that the compilers had access to any of the Khitan documents from the period, although there were some scholars at the Mongolian court who could still read the Khitan script. Nor does it seem that they had access to any Liao archival materials in Chinese. The Liao shih was mostly based on three already existing works: the extended "veritable records" compiled under Yeh-lii Yen in 1103, the incomplete Liao history 12 T'o-t'o et al., eds., Chin shih (Peking, 1975) (hereafter cited as CS), 4, p. 84; 89, p. 1988; 125, p. 2720. 13 CS, 12, p. 282; 125, p. 27271. 14 Hok-lam Chan, "Chinese official historiography at the Yuan court: The composition of the Liao, Chin, and Sung histories," in China under Mongol rule, ed. John D. Langlois, Jr. (Princeton, 1981). Cambridge Histories Online © Cambridge University Press, 2008 668 BIBLIOGRAPHICAL ESSAYS edited under Ch'en Ta-jen and presented to the Chin court in 1206, and the Ch'i-tan kuo chih. •' The last was an account of the Liao in twenty-seven chapters completed by a Southern Sung scholar named Yeh Lung-li at the imperial command, probably in 1247.l6 It gives a picture of the Liao entirely based on Sung written sources and very much from a Chinese viewpoint. Still a valuable work, it is the only one of the Liao sbih's main sources that survives. The other two histories were lost during the Ming. There is an integral translation of the Ch'i-tan kuo-chih into Russian, by V. S. Taskina." Ch'ing period critical scholarship The Liao shih was carelessly compiled. It is not only full of internal contradictions and historical inconsistencies, but it is also frequently at variance with material in the Chin shih and the Sung shih (which were compiled simultaneously by the same team of historians), with the two standard histories of the Wu-tai period, with Sung histories such as the Hsu Tzu chih t'ung chien ch'angpien, and with other contemporary literary sources. The text-critic historians of early Ch'ing times had a field day with the errors and inconsistencies in the text and with the history of the period, and their painstaking critical work is an essential source for modern historians. The first to undertake a systematic study of the Liao was Li E (1692—1752) who compiled a Liao shih shih i'8 in which he brought together additional material culled from almost four hundred Sung and Yuan works. His book was supplemented by Yang Fu-Chi (1747—1820) in his Liao shih shi i pu.'9 Both Ch'ien Ta-hsin (1782— 1804) and Chao I (1727—1814) wrote extensive notes on the Liao period. In the late nineteenth century Li Yu-t'ang (1843-1902) rearranged the materials collected by Li E and Yang Fu-chi under a scheme of topical headings, in his Liao shih chi shih pen mo.20 In this century, textual research on the Liao shih has been continued by such scholars as Feng Chia-sheng and Lo Chi-tsu. Their principal works are conveniently available in Chao T'ieh-han, ed., Liao shih chiao k'an chi. This includes Feng Chiasheng, Liao shih ch'u chiao; Lo Hsi-tsu, Liao shih chiao k'an chi; Feng Chia-sheng, Liao shih yii Chin-shih, Hsin, Chiu Wu-tai shih hu cheng ch'u lieh; and Feng Chia-sheng, Liao shih yuan liu k'ao.21 Other historians have examined the epigraphical evidence remain15 There is clear evidence of this in the preface to the monographs on ritual. See IS, 49, p. 834. 16 The preface is dated 1180. However, this date must be wrong, as the author received the chin shih degree in 1247. The regnal title in the preface is probably miscopied. See the preface to the index Ch'itan kuo chih t'ung chien published by the Centre franco-chinois d'lrudes sinologiques (Peking, 1949; repr. Taipei, 1968). See also the extensive collection of critical postfaces and notes on the book appended to the 1 9 8 ; Shang-hai ku-chi ch'u-pan she edition, p. 259. 17 Facsimile reproduction of Yeh Lung-li (E-lun O's Ch'i-tan kuo-chih, with Russian translation, introduction, commentary, and indexes by V. S. Taskina, Utoriia gosudarstva Kidanei (tsidan' go chzhi): Pernods kitalskogo, wedenie, kommtntaril i prilozheniia (Moscow, 1979). 18 Li E, Liao ihih shih-i, in Liao shih hui pien, ed. Yang Chia-lo (Taipei, 1973), vol. 3, no. 9. 19 Yang Fu-chi, Liao sbib shi-i pu, in Liao shih hui pien, ed. Yang Chia-lo, vol. 3, no. 10. 20 Li Yu-t'ang, Liao shih chi shih pen mo, 3 vols. (Peking, 1980). 21 Chao T'ieh-han, ed., Liao shih chiao k'an chi, in Sung Liao Chin Yuan ssu shih tzu liao ts'ung k'an (Taipei, Cambridge Histories Online © Cambridge University Press, 2008 BIBLIOGRAPHICAL ESSAYS 669 ing from the period, in both Khitan and Chinese. The labors of these scholars have cleared up many difficult points, but much remains obscure. The Liao will always remain one of the less well-documented periods of Chinese history. A convenient resource for modern historians of the Liao is the ten-volume Liao shih hui pien" compiled under the direction of Yang Chia-lo. This brings together all the major sources, the textual and critical works just mentioned and many others, together with a large selection of modern scholarship down to the 1960s (the tenth volume is a reprint of Wittfogel and Feng's History of Chinese society: Liao [907—1125]). A supplementary volume, Liao shih hui pien pu,2> was published in 1974. Two compendia of Liao writings published in the last decade are important resources. Ch'en Shu's Ch'iian Liao wen provides an exhaustive collection of Liao writings, including inscriptions, in Chinese.34 For a similar complete collection of writings in the Khitan language, see the collaborative volume by Ch'ing-ko-erh-t'ai (Chinggeltei) and others, Ch'i-tan hsiao tzuyen chiu,^ which also includes material on the current state of decipherment of the Khitan script and language. Modern scholarship Early Western historians of China paid almost no attention to the Liao. The first to devote more than a few lines to the subject was V. P. Vasil'jev, a Russian-language study of the history of eastern Central Asia, published in 1859. The first book on the subject was by H. C. von der Gabelenz, Geschichte der grvssen Liao,*6 a translation of a Manchu abridgment of the Liao shih published in 1877. There was an almost contemporary study by H. H. Howorth, "The northern frontagers of China, part v: The Khitai or Khitans,"*7 and a long section was devoted to the Khitan in E. H. Parker's A thousand years of the Tartars.** But serious modern scholarship on the subject began with the study by Edouard Chavannes, "Voyageurs chinois chez les Khitans et les Joutchen"2' in Journal Asiatique, which translates travel accounts concerning the Khitan by the Chinese envoys Hu Ch'iao, Wang I, Fu Cheng, and Sung Huan. This, however, did not stimulate further study of the subject, and apart from the archaeological descriptions by Joseph Mullie, the next Western contribution of note was Rolf Stein's lengthy study and translation from the Liao chib,*° the Shuo fu abridgment of the Chi-tan kuo-chih. This deals imaginatively and in great detail with 22 23 24 25 26 27 Yang Chia-lo, Liao shih hui pien, 10 vols. (Taipei, 1973). Yang Chia-lo, Liao shih hui pien pu (Taipei, 1974). Ch'en Shu, Ch'iian Liao wen (Peking, 1982). Ch'ing-ko-erh-t'ai et al., Ch'i-tan hsiao tzu yen chiu (Peking, 1985). H. C. von der Gabelenz, Ceschichte der grossen Liao aus dan Mandschu ubersetzt (St. Petersburg, 1877). H. H. Howorth, "The northern frontagers of China, part v: The Khitai or Khitans," in Journal of the Royal Asiatic Society, n.s. 13 (1881), pp. 1 2 1 - 8 2 . 28 E. H. Parker's A thousand years of the Tartars (London and Shanghai, 189;). 29 Edouard Chavannes, "Voyageurs chinois chez les Khitans et les Joutchen," Journal Asiatique, 9th series, vol. 9 (1897), pp. 3 7 7 - 4 2 2 ; 9th series, vol. 11 (1898), pp. 361-439. 30 Rolf Stein, "Leao-Tche," T'oung Poo, 35 (1939), pp. 1—154. Cambridge Histories Online © Cambridge University Press, 2008 670 BIBLIOGRAPHICAL ESSAYS questions relating to the kinship system and social structure and with the customs and observances of the Khitan. Karl August Wittfogel and Feng Chia-sheng's massive volume on the Liao, History of Chinese society: Liao (907—x 125),31 which was published in 1949, is unquestionably the most important single contribution to Liao historical studies yet published in any language. It not only provides a detailed and systematic analysis of every aspect of Liao social organization, economic life, administration, and institutions, but it also provides copious translations from the original sources and an exhaustive bibliography of secondary scholarship in all languages down to its date of publication. It does not, however, attempt to give an integrated chronological history of events, and because of its rigid compartmentalized structure it is difficult to derive from it an overall picture of secular developments. Largely because its awkward structure makes it hard to read, it has not received the general recognition it deserves. It is essential reading for any scholar interested in the period. Perhaps because Wittfogel and Feng's work was so comprehensive and broke so much new ground, in the years since its completion, little of consequence has been written on the Liao in Western languages. The one exception has been the field of foreign relations. The main Chinese contribution to Sung—Liao diplomatic relations was Nieh Ch'ung-ch'i's major essay "Sung Liao chiao p'ing k'ao,"3* which first appeared in 1940 and is reprinted in his Sung shih ts'ung k'ao. Fu Le-huan (1913—66) also wrote extensively on the same general field. See his volume of collected essays entitled Liao shih ts'ung k'ao.» Recent Western scholarship has reacted against the negative traditional Chinese view of the Khitan as inferior "barbarian" neighbors of the Sung and has concentrated on the emergence during the period of a real multistate system. Several articles in the excellent symposium volume edited by Morris Rossabi, China among equals: The Middle Kingdom and its neighbors, ioth-i4th centuries:** Those by Wang Gung-wu, Tao Jing-shen, and Michael Rogers'' are pertinent to this question and establish a picture of the emergence in northeast Asia in the tenth and eleventh centuries of a multistate system, in which the Khitan and Liao participated as an important power. The monograph by C. Schwartz-Schilling, Der Friede von Shan-yuan (1005 n.Chr.),*6 deals with events leading to the Sung—Liao treaty of 1005. Klaus Tietze, "The Liao-Sung border conflict of 1074—76,"" gives a detailed account of one subsequent crisis in interstate relations that was settled 31 Wittfogel and Feng. 32 Nieh Ch'ung-ch'i, "Sung Liao chiao p'ing k'ao," Yen-ching hsiieh poo, 27 (1940), pp. i - 5 i ; r e p r . in his Sung shih ts'ung k'ao, 2 vols. (Peking, 1980), vol. 2, pp. 283—387. 33 Fu Le-huan, Liao shih ts'ung k'ao (Peking, 1984), pp. 174—286. 34 Morris Rossabi, ed., China among equals: The Middle Kingdom and its neighbors, loth-i^th centuries (Berkeley and Los Angeles, 1983). 3 ; Wang Gung-wu, "The rhetoric of a lesser empire: Early Sung relations with its neighbours"; Tao Jingshen, "Barbarians or northerners: Northern Sung images of the Khitan"; and Michael Rogers, "National consciousness in medieval Korea: The impact of Liao and Chin on Koryd," in Rossabi, China among equals. 36 C. Schwartz-Schilling, Dtr Friede von Shan-yuan (1005 n.Chr.): Ein Beitrag zur Geschichte der chinesisehen Diplomatic (Wiesbaden, 1939). 37 Klaus Tietze, "The Liao—Sung border conflict of 1074—76," in Studia Sino-Mongolica: Festschrift fur Herbert Franie, ed. Wolfgang Bauer (Wiesbaden, 1979), pp. 1 2 7 - 5 1 . Cambridge Histories Online © Cambridge University Press, 2008 BIBLIOGRAPHICAL ESSAYS 671 without hostilities. Dagmar Thiele, Der Abschluss eines Vertrages: Diplomatic zwischen Sung undChin Dynastie 1117—1123,1s gives a detailed account of the last years of the dynasty, though its primary focus is on Sung-Chin relations rather than the Liao. T'ao Ching-shen (T'ao Chin-sheng) presented a full account of Sung-Liao relations in his Sung—Liao kuan hsi shihyen chiu.i9 His English-language study of the same theme, Two Sons of Heaven,*0 appeared after this chapter was completed. Two recent Ph.D. theses, as yet unpublished, also deal with Liao foreign relations: Janet Novey, "Yii Ching, a Northern Sung statesman, and his treatise on the Ch'i-tan bureaucracy,"'1' and Melvin Thlick-len Ang, "Sung—Liao diplomacy in eleventh- and twelfth-century China: A study of the social and political determinants of foreign policy."42 The latter is almost exclusively a study of the Sung officials appointed as diplomatic envoys to Liao. A number of significant works on Liao history have appeared in Chinese. The rich and imaginative studies by Fu Le-huan, whose early death in 1966 was a major loss to Liao studies, were reprinted in his posthumous Liao shih ts'ung k'ao.Ai The most notable and prolific Liao historian in the People's Republic has been Ch'en Shu, whose Ch'i-tan she hui ching chi shih ka&u is perhaps the most ambitious attempt by a Chinese historian at analyzing Liao society and its economy. Its analysis differs from that of Wittfogel and Feng and is at times somewhat more systematic than the thin and fragmented evidence would seem to justify. It is also static and gives little sense of secular change. Ch'en Shu's most recent book, Ch'itan cheng chih shih kao," contains a series of interesting and sometimes original essays on aspects of Khitan social and governmental institutions and on various political incidents and issues. A general account written for a nonscholarly audience is by Ch'en Shu, Liao-tai shih hua.<6 Also of interest is Chang Chen-ming's Ch'i-tan shih lu'eh.*1 A great part of this also deals with socioeconomic and institutional matters, but it is more integrative and gives more of a sense of historical development. A much more detailed and systematic treatment of Liao history is that by Shu Fen, Liao shih kao,*8 which was not available to me until after the completion of this chapter. Perhaps the best general outline of Liao political history is contained in a set of detailed lecture notes by Yao Ts'ung-wu, posthumously published after his death in 38 Dagmar Thiele, Der Abschluss eines Vertrages: Diplomatie zwischen Sung-und Chin Dynastie n 17—1123 (Wiesbaden, 1971). 39 T'ao Chin-sheng (Tao Jing-shen), Sung Liao kuan hsi shih yen chiu (Taipei, 1984). 40 T'ao Chin-sheng (Tao Jing-shen), Two sons of heaven (Tucson, 1988). 41 Janet Novey, "Yii Ching, a Northern Sung statesman, and his treatise on the Ch'i-tan bureaucracy," Ph.D. diss., Indiana University, 1983. 42 Melvin Thlick-len Ang, "Sung-Liao diplomacy in eleventh- and twelfth-century China: A study of the social and political determinants of foreign policy," Ph.D. diss., University of Pennsylvania, 1983. 43 Fu Le-huan, Liao shih ts'ung t'ao (Peking, 1984). 44 Ch'en Shu, Ch'i-tan she hui ching chi shih kao (Peking, Shang-hai, 1963; repr. 1978). 45 Ch'en Shu, Ch'i-tan cheng chih shih kao (Peking, 1986). 46 Ch'en Shu, Liao tai shih hua (Cheng-chou, 1981). 47 Chang Cheng-ming, Ch'i-tan shih liieh (Peking, 1979). According to the author's postface, this book was actually completed around 1963. 48 Shu Fen, Liao shih kao (Wu-han, 1984). Cambridge Histories Online © Cambridge University Press, 2008 672 BIBLIOGRAPHICAL ESSAYS 1971 and edited for publication with his collected writings by T'ao Chin-sheng.*9 Professor Yao wrote several excellent studies of Liao political and institutional history, which are notable for both their imaginative understanding of complex events and their scrupulous scholarship. The most important are included in his Tung-pei shih lun ts-ungS° and they also are reprinted in Liao shih hut pien and its supplement. Since the 1930s much of the best work on Liao history, and especially on its society and legal institutions, has come from Japanese scholars. The most important contributions are the series of exhaustive and systematic studies by Shimada Masao: on the Liao law codes (with Takigawa Masajiro),5' Liao society,52 the Liao government institutions," Liao society and culture,'« the Liao bureaucratic system," and various studies of Liao history.'6 These studies are particularly thorough and detailed in regard to legal and administrative institutions. Their conclusions on law are summarized and further developed in a forthcoming study by Herbert Franke, "Chinese law in a multinational society: The case of the Liao (907—1125),"" a paper presented at the History of Chinese Medieval Law conference in Bellagio, Italy, in 1981. The same author has also published an integral translation with full commentary of the monograph on law from the Liao shih.^ Two interesting contributions to the study of Liao kinship were published by Jennifer Holmgren: "Marriage, kinship and succession under the Ch'i-tan rulers of the Liao dynasty (907—ii25)"'9 and "Yeh-lii, Yao-lien and Ta-ho: Views of the hereditary prerogative in early Khitan leadership."60 On Liao Buddhism, the best account is still that in Nogami Shunjo, Ryo Kin no Bukkyo.6l On Liao material culture, the first important evidence came from archaeology, especially from the excavation of the Liao imperial tombs. Joseph Mullie, "Les anciennes villes de l'empire des grands Leao au Royaume mongol de Barin"6* and "Les sepultures de K'ing des Leao,"6' first drew the attention of Western readers to 49 Yao Ts'ung-wu, "Liao Chin Yuan shih chiang-i (chia): Liao ch'ao shih," in Yao Ts'ung-wu bsien-sheng ch'iian-chi, comp. T'ao Chin-sheng (Taipei, 1972), vol. 2. 50 Yao Ts'ung-wu, Tung-pei shib lun Is'ung, 2 vols. (Taipei, 1959); repr. in Yang Chia-lo, Liao sbih huipien (Taipei, 1973). 51 Takigawa Masajiro and Shimada Masao, Ryoritsu no kenkyu (Tokyo, 1943). 52 Shimada Masao, Ryodaishakaishikmkyi (Kyoto, 1952). 53 Shimada Masao, Ry osei no kenkyu (Tokyo, 1954; repr. Tokyo, 1973). 54 Shimada Masao, Ryo no shakai to bunka (Tokyo, 1956). 55 Shimada Masao, Ryocho kanset no kenkyu (Tokyo, 1978). 56 Shimada Masao, Ryochoshi no kenkyu (Tokyo, 1979). 57 Herbert Franke, "Chinese law in a multinational society: The case of the Liao (907—1125)," paper presented at the History of Chinese Medieval Law Conference, Bellagio, Italy, 1981. 58 Herbert Franke, "The "Treatise on punishments' in the Liao history," Central Asiatic Journal (1983), pp. 9 - 3 8 . 59 Jennifer Holmgren, "Marriage, kinship and succession under the Ch'i-tan rulers of the Liao dynasty (907—1125)," T'oung Pao, 72 (1986), pp. 44—91. 60 Jennifer Holmgren, "Yeh-lii, Yao-lien and Ta-ho: Views of the hereditary prerogative in early Khitan leadership," Papers on Far Eastern History, 34 (Canberra, 1986), pp. 37—81. 61 Nogami Shunjo, Ryo Kin no Bukkyo (Kyoto, 1953). 62 Joseph Mullie, "Les anciennes villes de l'empire des grands Leao au Royaume mongol de Barin," T'oung Pao, 21 (1922), pp. 105-231. 63 Joseph Mullie, "Les sepultures de K'ing des Leao," T'oung Pao, 30 (1933), PP- I - 2 5 - Cambridge Histories Online © Cambridge University Press, 2008 BIBLIOGRAPHICAL ESSAYS 673 the physical remains of Liao. The field was developed by Japanese archaeologists, beginning with the anthropologist-archaeologist Torii Ryuzo (1870-1953). His series of articles in the early 1930s, "Ryodai no hekiga ni tsuite," 64 in Kokka, introduced the startling wall paintings discovered in the royal tombs. They were followed by the publication of his Kokogaku jo yori mitaru Ryo no bunka,6* which presents a large collection of plates; the text seems never to have been published. See also his Ryo no bunka wo saguru.66 On the early Japanese excavations, see also Tamura Jitsuzo and Kobayashi Yukio's study of the Ch'ing-ling, the mausoleum of Shengtsung, 67 and Shimada Masao's report on the site of Tsu-chou.68 An interesting and more up-to-date account of the Ch'ing-ling tomb and its paintings is that by Tamura Jitsuzo, Keiryo no hekiga.69 Since the 1950s other tombs containing paintings and a wide variety of artifacts have been discovered by Chinese archaeologists. Three Western studies imaginatively used this material to illustrate aspects of Khitan life: Ellen Johnston Laing, "Patterns and problems in later Chinese tomb decoration";70 Linda Cooke Johnson, "The wedding ceremony for an imperial Liao princess: Wall paintings from a Liao dynasty tomb in Jilin"; 71 and Albright Rorex, "Some Liao tomb murals and images of nomads in Chinese paintings of the Wen-chi story."72 On Liao ceramics and kiln sites, see William Watson, T'ang and Liao ceramics,™ and Lin Wen-hsin and Chu Tzu-tang's descriptive work on Liao ceramics in the Liaoning Provincial Museum. 74 See also the exhibition catalogue by Mino Yutaka, Ceramics in the Liao dynasty, published in 1973, which includes an excellent bibliography to that date. 7 ' One field in which the Liao produced unquestioned masterpieces was architecture. Over thirty major buildings, constructed of both stone and timber, survive, some incorporating important technical innovations. The classic descriptive work on Liao and Chin architecture is the collection of fine photographs published by Sekino Tadashi and Takejima Takuichi in 1934 and the accompanying study by Takejima that appeared ten years later.76 The most beautiful of all Liao buildings, the great wooden pagoda at Ying-hsien in northeastern Shansi, is the subject of a monograph 64 Torii Ryuzo, "Ryddai no hekiga ni tsuite," Kokka 490, pp. 272-80; 491, pp. 283-9; 492, pp. 313— I7M93. PP- 343-5° ( ' 9 3 0 . 65 Torii Ryuzo, Kokogaku jo yori mitaru Ryo no bunka: Zufu, 4 vols. (Tokyo, 1936). 66 Torii Ryuzo, Ryo no bunka wo saguru (Tokyo, 1937). 67 Tamura Jitsuzo and Kobayashi Yukio, Keiryo (Kyoto, 1953). 68 Shimada Masao, So-sbu jo (Tokyo, 19;;). 69 Tamura Jitsuzo, Keiryo no hekiga (Kyoto, 1977). 70 Ellen Johnston Laing, "Patterns and problems in later Chinese tomb decoration," Journal of Oriental Studies, 16(1978), pp. 3—20. 71 Linda Cooke Johnson, "The wedding ceremony for an imperial Liao princess: Wall paintings from a Liao dynasty tomb in Jilin," Artibus Asiae, 44 (1983), pp. 107—36. 72 Albright Rorex, "Some Liao tomb murals and images of nomads in Chinese paintings of the Wen-chi story," Artibus Asiae, 45 (1984), pp. 174—98. 73 William Watson, T'ang and Liao ceramics (New York, 1984). 74 Li Wen-hsin and Chu Tzu-fang, Liao-ning sheng po-wu-yiian tsang Liao tz'u hsiian-chi (Peking, 1961). 75 Mino Yutaka, Ceramics in the Liao dynasty: North and south of the Great Wall (New York, 1973). 76 See Sekino Tadashi and Takejima Takuichi, Ryo Kin jidai no kenchiku to sono Butsuzo, 2 vols. of plates (Tokyo, 1934); and Takejima Takuichi, Ryo Kin jidai no kenchiku to sono Butsuzo (Tokyo, 1944). Cambridge Histories Online © Cambridge University Press, 2008 674 BIBLIOGRAPHICAL ESSAYS by Ch'en Min-ta, published in 1980.77 A useful summary can be found in Nancy Schatzman Steinhardc's Chinese traditional architecture, published in 1984.78 Finally, on the related subject of city planning, Nancy Schatzman Steinhardt's Chinese imperial city planning includes plans of the Superior Capital, the Southern Capital, and the Central Capital." 2 . THE HSI HSIA The history of the Hsi Hsia was little studied until this century, surprising when one considers its considerable intrinsic importance. Theirs was a dynasty that for two centuries played a major role in the international politics of north China and Inner Asia, which had its own individual and complex system of institutions and a highly sophisticated culture. However, the Hsi Hsia was never deemed a legitimate dynasty by orthodox Chinese historians, perhaps because it occupied territories that, even in T'ang times, had been only on the borders of the Chinese world. As a result, no standard history was ever compiled for the Hsi Hsia: When the histories of Liao and Chin — the other non-Chinese dynasties of the period — were finally compiled under the last Yiian emperor in 1344—5, ar * er decades of acrimonious dispute over the legitimacy of the two dynasties, there was never any suggestion that the Hsi Hsia should receive similar treatment. In regard to Chinese official historiography, the record of the Hsia was therefore relegated to the lieh-chuan (biography section) devoted to their state in the three dynastic histories of Sung, Liao, and Chin" that were compiled together in the 1340s. These chapters, like the accounts of other "foreign" states, focused not on the internal events and institutions of the Tangut state, but on its relations with the other powers in the Chinese world. Moreover, the source material from which they were compiled was almost exclusively Chinese. The Hsia, like the Liao and the Chin, had their own historiographers, but their books, together with the Hsi Hsia state documents written in Tangut, were destroyed during the Mongolian conquest in 1227. With the lapse of time, knowledge of the Tangut script and language gradually disappeared, and so what monuments written in Tangut script remained were unintelligible until recent decades. Many documents relating to the Hsi Hsia survive in Sung histories and literary works, beyond what is included in the dynastic histories. The names of a few specialized Sung works concerning the Hsia, most of them works on border defense and strategy, survive in bibliographies, but the books themselves have long been lost.1 The only exception, the Hsi Hsia shih liieh by Wang Ch'eng, which has existed as a separate title since the thirteenth century, is in fact, as the Ssu-k'u editors noted in the eighteenth century, merely the chapter on Hsia from Wang Ch'eng's Tung tu 77 78 79 1 See Ch'en Min-ta, Yittg-hsien ma t'a (Peking, 1980). See Nancy Schatzman Steinhardt, Chinese traditional architecture (New York, 1984), pp. 109—19. See Nancy Schatzman Steinhardt, Chinese imperial city planning (Honolulu, 1990), pp. 122—8. T'o-t'o et al., eds., Sung shih (Peking, 1977) 485—6, pp. 13981—14033; T'o-t'oet al., eds., Liao shih (Peking, 1974) 105, pp. 1523-30; T'o-t'o et al., eds., Chin ihih(Peking, 1975) 134, pp. 2865-79. 2 See Wu T'ien-ch'ih, Hsi Hsia shih kao (Ch'eng-tu, 1983), pp. 3 3 8 - 9 , for details. Cambridge Histories Online © Cambridge University Press, 2008 BIBLIOGRAPHICAL ESSAYS 675 shih liieh (of 1186) published as a separate book and given a new title by a later publisher.3 Attempts to synthesize a history of the Hsia from all surviving Chinese sources began in the eighteenth century. Among the scholars who wrote on the Hsia were Hung Liang-chi, whose Hsi Hsia kuo chih was never completed, and Ch'in En-fu, whose Hsi Hsia shu was destroyed soon after its completion. Two major nineteenthcentury works were printed and still survive: Wu Kuang-ch'eng's chronological history Hsi Hsia sbu shih in forty-two chapters, completed in 1825—7, a n d Chang Chien's topical history, Hsi Hsia chi shih pen mo in thirty-six chapters, completed in 1884. Both remain important to modern historians. Most of the source material for the Hsi Hsia chi shih pen mo can readily be identified in other works. The Hsi Hsia shu shih, on the other hand, contains a great deal of important information that cannot be found elsewhere. However, because the provenance of much of Wu Kuang-ch'eng's material is unclear and sometimes dubious, his book, valuable as it is, needs to be used with caution. The most recent attempt at this type of traditional historical synthesis was the Hsi Hsia chi completed by Tai Hsi-chang in 1927. Unlike his predecessors, the author of this carefully compiled work identifies his sources and lists the works he consulted. The only major Sung source that he did not use is the Sung huiyao, which was not yet published in his time. A short but still useful study of these early histories was published in 1943 by the prolific scholar of Chinese historiography Chu Hsi-tsu.4 Discovery ofTangut documents and the decipherment of the language Sometime before Tai Hsi-chang's work was published, the study of the Tanguts and the Hsi Hsia was greatly advanced with the discovery of a great body of original documentation in the Hsi Hsia language and script. Some Hsi Hsia documents were found by Sir Aurel Stein and Paul Pelliot at Tun-huang. In 1908 and 1909 a Russian expedition led by Major P. K. Kozlov discovered and excavated the remains of a Hsi Hsia city at Edzina (Khara-Khoto). Among their discoveries was the tomb of a noble Buddhist lady containing thousands of books and documents, both printed and manuscript. The same area was later investigated by Sir Aurel Stein, Langdon Warner, and the expedition led by Sven Hedin, and although their discoveries were not on such a grand scale as Kozlov's, more documents came to light. Still other documents were discovered by Chinese archaeologists in recent decades. This new documentation, although it could not yet be deciphered, provided a large body of historical evidence from the Tangut themselves, not from their Chinese neighbors. Some tentative studies about the Tangut language had already been published before these discoveries, based on a few bilingual stone inscriptions, notably by M. G. Deveria at the very end of the nineteenth century, but to all intents and purposes the language was still unreadable. The serious study of the Tangut language was 3 See Chi Yiin et al., comps., Ssu k'u ch'iian shu tsung mu t'iyao, 66, p. 1443. 4 Chu Hsi-tsu, "Hsi Hsia shih chi k'ao," in Shuo wen yiieb k'an, 3 (1943), pp. 2 5 - 3 0 . Cambridge Histories Online © Cambridge University Press, 2008 676 BIBLIOGRAPHICAL ESSAYS begun in 1909 by A. I. Ivanov, who later compiled a Tangut dictionary, largely based on a printed Tangut—Chinese glossary that was included among the Kozlov finds, and worked on some of the other materials. In 1922 he was sent with the Soviet diplomatic mission to Peking, where he made contact with various Chinese scholars, notably the brothers Lo Fu-ch'eng and Lo Fu-ch'ang, sons of Lo Chen-yii, and Wang Ching-ju, and he encouraged them also to pursue their study of Tangut. Ivanov also influenced another Russian scholar named N. A. Nevskii, who later began making a systematic catalogue of the Kozlov collection in Leningrad. Unfortunately for Tangut studies, in 1937 both Ivanov and Nevskii fell victim to Stalin's purges, and their manuscripts were confiscated. World War II also brought an end to serious studies of Tangut in China. For a while, almost the only serious work in the field was that by the Japanese scholars Ishihama Juntaro and his pupil Nishida Tatsuo, who worked mainly on Tangut Buddhist texts. In the late 1950s and 1960s, Tangut studies were revived in the Soviet Union, where a group of young scholars at last achieved a level of skill in reading Tangut that enabled them to interpret with confidence new types of material, some of it of immense historical value. From that time is dated the first modern Western summary of Tangut history, E. I. Kychanov's Ocherk istorii tangutskogo gosudarstva,* a work still necessarily largely based on Chinese sources but written with a knowledge of the Tangut material. Shortly afterward, there appeared two thorough studies of the early history of the Tanguts based exclusively on Chinese sources: Paul Friedland's dissertation, "A reconstruction of early Tangut history,"6 and a detailed Japanese study on the same topic by Okazaki Seiro.7 The only comprehensive study of Hsia in English is the Ph.D. dissertation by Ruth Dunnell,8 which incorporates the results of most Soviet and Chinese scholarship down to 1983. These studies appeared at much the same time as the general revival among Western sinologists of interest in China's neighboring peoples, which began in the late 1960s and has flourished during the last two decades. We can now look more deeply at the Tanguts' relations with China and also at those with their Khitan, Uighur, and Tibetan neighbors and see the complex world of north China and its bordering peoples in all its variety. Recent Chinese scholarship In China, too, there has been a renaissance of Hsi Hsia and Tangut studies since the late 1970s, which has produced a lively periodical literature on both linguistic and historical subjects. The most important Chinese study to date is Wu T'ien-ch'ih's 5 E. I. Kychanov, Ocherk istorii tangutikogo gosudarstoa (Moscow, 1968). 6 Paul Fried land, "A reconstruction of early Tangut history" (Ph.D. diss., University of Washington, 1969)7 Okazaki Seiro, Tanguto kodaishi kenkyu (Kyoto, 1972). 8 Ruth W. Dunnell, "Tanguts and the Tangut state of Ta Hsia" (Ph.D. diss., Princeton University, 1983). Cambridge Histories Online © Cambridge University Press, 2008 BIBLIOGRAPHICAL ESSAYS 677 draft history of Hsi Hsia, first published in 1980 and since republished in an expanded and revised edition.9 This is the best synthesis of Tangut history currently available. The range of issues addressed by Chinese scholars can be gauged from the anthology of previously published articles edited in 1984 by Pai Pin.'° Other scholars who have published collections of articles on Hsia history and culture include Li Fanwen," Ch'en Ping-ying,12 and Li Wei.1' Of these authors, Ch'en Ping-ying attempts to incorporate Tangut-language materials, and Li Wei confines himself to Chinese sources. Li Fan-wen is an experienced Tangut linguist who has published a facsimile and Chinese translation of the Tangut dictionary of homophones entitled T'ung-yin.'« Although some of Li's phonological reconstructions have been contested, his work remains an indispensable resource, along with the facsimile and Chinese translation of another Tangut dictionary, the Wen-hat, by Shih Chin-po, Pai Pin, and Huang Chen-hua.15 Shih Chin-po has also published an important monograph on Tangut Buddhism.'6 More recently the Hsi Hsia wen wu,11 an illustrated catalogue of Tangut artifacts, with introductory essays, edited by Shih Chin-po, Pai Pin, and Wu Feng-yun, presents a large portion of the extant relics of Hsia culture and their sites in excellent photographic reproductions (including many color plates), together with copious documentation. The most important aspect of this new Chinese scholarship is the integration into Hsi Hsia studies of newly readable Tangut texts and also a wide range of new archaeological discoveries that provide invaluable and completely new evidence on the Tanguts' material culture, life, and art. Russian scholarship, too, continues to provide important new work. The linguist Ksana Kepping has published many skillful translations and analyses of Tangut texts from the Leningrad archives.'8 The most significant contribution based on Tangutlanguage material is undoubtedly E. I. Kychanov's monumental study and translation of the almost-complete Hsia law code. •' Its now completed publication will enable historians to achieve a completely new level of understanding of Hsi Hsia social institutions, government, and the process by which the Tanguts adapted Chinese written law to the needs of their own, different world. More recently, a 9 10 11 12 13 14 15 16 17 18 Wu T'ien-ch'ih, Hsi Hsia sbih kao (Chengtu, 1980; 2nd rev. ed., Chengtu, 1983). Pai Pin ed., Hsi Hsia shih lun wen chi (Yin-ch'uan, 1984). Li Fan-wen, Hsi Hsia yen chiu lun chi (Yin-ch'uan, 1983). Ch'en Ping-ying, Hsi Hsia wen wu yen chiu (Yin-ch'uan, 1985). Li Wei, Hsi Hsia Shih yen chiu (Yin-ch'uan, 1989). Li Fan-wen, T'ung-yin yen chiu (Yin-ch'uan, 1986). Shih Chin-po, Pai Pin, and Huang Cheri-hua, Wen-hai yen chiu (Peking, 1983). Shih Chin-po, Hsi Hsia Fo chiao shih liieh (Yin-ch'uan, 1988). Shih Chin-po, Pai Pin, and Wu Feng-yun, Hsi Hsia wen wu (Peking, 1988). See the review article by Ruth Dunnell, "Soviet Scholarship on Medieval China, 1982-1987," Bulletin of Sung—Yuan Studies, 20(1988), pp. 137—42. 19 E. I. Kychanov, Hmennyi i zanwo utverzhdennyi koaeks devisa tsarstvovaniia mbesnoe protsvetanie ( 1 1 4 9 1169), vol. 1 (Moscow, 1988); vol. 2 (Moscow, 1987); vol. 3 (Moscow, 1989); vol. 4 (Moscow, 1989). Cambridge Histories Online © Cambridge University Press, 2008 678 BIBLIOGRAPHICAL ESSAYS translation and study of the Tangut code of military law was written by E. I. Kychanov and Herbert Franke,20 which clarifies many issues concerning military organization under the Hsia and compares parallel institutions under the Sung. 3. THE CHIN DYNASTY The basic source for the Chin dynasty is the history of Chin (Chin shih)1 compiled under the Yuan dynasty in 1344—5 by a committee headed by the Chinese scholar Ou-yang Hsiian (1283—1357) and sponsored by the chancellor Toghto (1314-55). The original printed edition of 1345 has been partly preserved. A later Yuan edition printed in 1348 still exists and was republished in 1958 in a photolithographed edition in Shanghai. The edition most widely used until recently was the Po-na pen edition2 (1931), which is a composite facsimile made up of two incomplete Yuan editions. This text is reliable apart from some typographical and copying errors. Altogether, over thirty different printed editions and manuscripts of the Chin shih are known today. The Po-na pen edition has been superseded by the eight-volume punctuated and annotated Chung-hua shu-chii edition published in Peking in 1975, which is cited in our footnotes. Another punctuated edition in two volumes was published in Taipei in 1970;3 it is much inferior to the Chung-hua shu-chii edition but still useful because much additional material on the Chin dynasty, such as a bibliography of Chin literature in Chinese, has been included. The contents of the Chin shih are arranged according to the traditional model used for dynastic histories. It is composed of imperial annals (ti-chi), treatises (chih), chronological tables (piao), and biographies (lieh-chuan). The last category, as is customary, also includes monographs on relations with other states (Hsi-hsia and Korea). The Chin shih is, however, distinguished from other histories of Chinese dynasties by two features. A chronological table lists embassies, treaties, and other relevant events in foreign relations, which is a convenient guide to Chin's relationships with its neigbhors (Sung, Hsi-hsia, Korea, and Liao). The other peculiarity is to be found at the end of the work. It is a list of foreign, that is, non-Chinese, names and terms occurring in the Chin shih, with a Chinese translation of these Jurchen words. There is also a list of Jurchen clan names with their Chinese equivalents, some of which can be shown to be translations of the Jurchen names. These materials are of great importance for a study of the Jurchen language of the twelfth and early thirteenth centuries. However, the attempt by the scholars of the Ch'ien-lung reign (1736—95) to supply Manchu etymologies for the foreign words occurring in the Chin shih is largely based on fantasy. For this reason, the Chin shih yii chieh (Explanation of words in the Chin history),4 first published in 1781, should be disregarded. 20 E . L. K y i a n o v a n d H e r b e r t Franke, Tangutischeandchinaische i}.Jahrhundtrts, Philosophisch-Wstorisch Klass Abhandlungtn: 1 2 3 4 Quellen zur Militdrgaetzgebung des 11. bis Neue Folge, Heft 104. Bayerischen Akade- mie der Wissenschaften, Munich, 1990. Chin shih (Peking, 1975)Po-na pen edition of the Chin shih, published in 1931. Chin shih (Taipei, 1970). Ch'in ling Chin shih yii chieh (Peking, 1781). Cambridge Histories Online © Cambridge University Press, 2008 BIBLIOGRAPHICAL ESSAYS 679 Also, those editions of the Chin shih and other contemporary sources that have suffered the "improvements" of the Ch'ien-lung scholars cannot be used today. The Chin shih was based chiefly on primary Chin materials such as the shih lu (veritable records) of the individual Chin rulers, biographies included in literary works, and occasionally other contemporary sources. The corpus of primary sources is relatively homogeneous, and so there are fewer internal discrepancies in the Chin shih than in the other two histories compiled at the same time in the 1340s, the Liao shih and the Sung shih, the histories of Liao and Sung. The best modern study of the compilation and sources of the Chin shih and the most comprehensive survey of editions before 1970 is that by Hok-lam Chan, The historiography of the Chin dynasty: Three studies.' The use of the Chin shih has been greatly facilitated by several important modern reference tools. A first guide to collections of biographies, including those of Chin shih, was number 35 of the Harvard-Yenching Institute Sinological Index Series, Combined indices to thirty collections of Liao, Chin and Yiian biographies.6 A complete index to all personal names was prepared by Ts'ui Wen-yin, Chin shih jen ming so yin (Index to personal names in the Chin shih).1 Even more useful is the very detailed index, amounting to almost a concordance to the Chin shih, compiled under the direction of Onogawa Hidemi, Kinshi gokai shusei (A concordance to the Chin shih), in three volumes.8 This comprehensively lists not only personal and geographical names but also all important terms occurring in the work. It is based on the Po-na-pen edition, whereas the index by Ts'ui Wen-yin refers to the new punctuated Peking edition of 1975. Another contemporary source of some importance is the Ta Chin kuo chih (Records of the great Chin state).9 This work, which was based on Chin and Sung primary sources, has been ascribed to Yii-wen Mou-chao, but the date of its compilation remains uncertain (perhaps mid-thirteenth century). It is organized according to the model of the standard dynastic histories, beginning with the imperial annals. These are followed by biographies of thirteen statesmen of the formative period of the state and of some thirty Chinese literati active under the Chin. Special chapters are devoted to the puppet states of Ch'u and Ch'i. The treatises are very brief and much less comprehensive than are those in the Chin shih. The Ta Chin kuo chih is, nevertheless, a source not to be neglected because it contains some details not found in the Chin shih and includes the full texts of several political documents, such as treaties between Sung and Chin. Using the materials in the Ta Chin kuo chih is easy because there exists a rather full index by Wu Hsiao-ling et al., Ta Chin kuo chih t'ung chien (Index du Ta Kin kouo-tche).IO Another collection of documents is the Ta Chin tiao fa lu (Records of submission Chan Hok-lam, The historiography of the Chin dynasty: Thru studio (Wiesbaden, 1970). Combined indices to thirty collections of Liao, Chin and Yuan biographies (Peking, 1940). Ts'ui Wen-yin, Chin shih jen ming so yin (Peking, 1980). Onogawa Hidemi, Kinshi gokai shusei (Kyoto, 1960-2). Yii-wen Mou-chao, Ta Chin kuo chih. A modern annotated edition is Ta Chin kuo chih chiao cheng, ed. Ts'ui Wen-yin (Peking, 1986). 10 Wu Hsiao-ling et al., Ta Chin kuo chih t'ung chien (Peking, 1949). 5 6 7 8 9 Cambridge Histories Online © Cambridge University Press, 2008 680 BIBLIOGRAPHICAL ESSAYS and attack under the great Chin)." It contains mainly diplomatic correspondence, treaties, and edicts concerning the relations between Chin and Sung in the crucial years 1123 to 1127. Unfortunately, the most accessible editions are disfigured by the "reforms" of orthography carried out under Ch'ien-lung. Nevertheless, the work remains a mine of information on the military and political situation during the collapse of the Northern Sung state. The last months of the Chin state in 1233—4 a r e t n e subject of a diary written by a former Chinese official of the Chin, Wang E (1190-1273), who after the fall of the Chin state joined the victorious Mongols. His Ju nan i shih (Neglected facts from Junan)" describes in great detail the desperate attempts of the court of the last Chin ruler to survive the combined attacks of the Sung and Mongolian armies. The text was studied by Hok-lam Chan in "Prolegomena to theJu-nan i shih: A memoir of the last Chin court under the Mongol siege of 1234."'3 Hok-lam Chan also wrote a valuable piece on another important late Chin text, the Kuei ch'ien chih (Records written in retirement), by Liu Ch'i (1203-50), in his book The historiography of the Chin dynasty.I4 The Kuei ch'ien cbihl> contains many biographies of late Chin personalities, anecdotes based mostly on personal experience, and other details of life in K'aifeng during the last years of the Chin state. It also includes an eyewitness account of the conquest of K'ai-feng by the Mongolian armies, which has been translated by Erich Haenisch in his Zum Untergang Zweier Reiche. "6 Among the Chin sources we should also mention the Chin author Chang Wei's forty-chapter Ta Chin chi It (Collected rites of Chin).17 It is a voluminous text describing the Chinese rituals and court ceremonials under the Chin dynasty in much greater detail than do the corresponding chapters of the Chin shih. It goes without saying that among the contemporary sources, the collected works of Chinese authors writing under the Chin are of primary importance. They include many biographies that have been conveniently indexed in two recent works, by Umehara Kaoru and Kinugawa Tsutomu, Ryo—Kin—Genjin denki sakuin (Index to biographies of Liao, Chin, and Yiian persons),18 and Igor de Rachewiltz et al., Index to biographical material in Chin and Yiian literary works.19 The latter is particularly useful because it also includes biographical materials from Taoist works of the Ch'iian-chen sect contained in the Taoist canon (Tao-tsang),2° as well as biographical information taken from Buddhist historiography. 11 Ta Chin tiaofa lit (Shanghai, 1959, TSCC ed.). 12 Wang E, Ju-nan i shih (Shanghai, 1959, TSCC ed.). 13 Chan Hok-lam, "Prolegomena to the./* nan i shih: A memoir of the last Chin court under the Mongol siege of 1234," Sung Studies Newsletter, 2, supp. 1, pp. 2—19. 14 Chan, The historiography of the Chin dynasty, pp. 121—66. 15 Kuei ch'ien chih (Peking, 1983). 16 Erich Haenisch, Zum Untergang zweier Reiche: Berichte von Augenzeugen aus dtn Jahren 1232—33 und 1268—70 (Wiesbaden, 1969), pp. 7—26. 17 Chang Wei, Ta Chin chi It (Shanghai, 1936, TSCC ed.). 18 Umehara Kaoru and Kinugawa Tsutomu, Ryo—Kin—Genjin denki sakuin (Kyoto, 1972). 19 Igor de Rachewiltz et al., Index to biographical material in Chin and Yiian literary works, 3 vols. (Canberra, 1970, 1972, 1979). 20 Tao-tsang (repr. Taipei, 1962). Cambridge Histories Online © Cambridge University Press, 2008 BIBLIOGRAPHICAL ESSAYS 68l Also valuable as historical sources are two anthologies of Chin literature containing much material that cannot be found in the surviving collected works of Chin authors. These are by Chuang Chung-fang (1780-1857), Chin wen ya (Florilegium of Chin literature),21 and Chang Chin-wu, Chin wen tsui (Complete collection of Chin literature).22 Both anthologies are now easily accessible in modern reprints. The Chin wen tsui, in particular, contains much important information, including the texts of inscriptions on stelae and other epigraphical material. The two works are arranged according to literary genres, so that it is relatively easy to locate a source even without an index. The Chung chou chi,** an anthology of Chin poetry compiled between 1233 and 1249 by Yuan Hao-wen (1190—1257), is also a good biographical source because it provides a short biography of each writer represented in the anthology. This work was analyzed by Hok-lam Chan.24 Our description has centered on Chin writings. But of almost equal importance are the Sung sources. Without them no history of the Chin dynasty or of the long struggles between Sung and Chin could be written. The amount of material to be gathered from Sung sources is staggering. It is perhaps not an overstatement to say that it is difficult to find works of a political nature written under the Southern Sung that do not mention the Jurchen state of Chin at some place or other. This is above all true of the annalistic sources of Southern Sung date. Among these the most voluminous and informative is the San ch'ao pei meng hui pien (Collected documents on the treaties with the north during three reigns), compiled around 1196 by Hsu Menghsin (1126-1207). It covers events from 1117 to 1162 and also contains valuable details on the pre-dynastic Jurchen. The 250 chapters of this compilation include a great number of original documents such as letters, instructions, memorials, and reports from ambassadors and other officials. The overwhelming mass of information in the San ch'ao pei meng hui pien makes it possible to write an almost day-to-day account of crucial events, for example, the fall of the Northern Sung in 1126—7. The current edition is that published by Yuan Tsu-an in 1878, which was recently reprinted.2' Other important sources are the embassy reports that have preserved some eyewitness accounts of conditions in the Jurchen state. A bibliographical study of the extant embassy diaries was made by Herbert Franke, "A Sung embassy diary o f i 2 i i —1212: The Shih Chin lu of Ch'eng Cho. "*6 Other accounts of embassies were translated nearly a century ago by Edouard Chavannes - who excelled in so many otherfieldsin Chinese studies - in his articles "Voyageurs chinois chez les Khitans et les Joutchen"27 and "Pei 21 22 23 24 25 Chuang Chung-fang, Chin wen ya (Taipei, 1967). Chang Chin-wu, Chin wen tsui (Taipei, 1967). Yuan Hao-wen, comp., Chung chou chi (SPTK ed.). Chan, The historiography of the Chin dynasty, pp. 67—119. Hsu Meng-hsin, comp., San ch'ao pei meng hui pirn (Yuan Tsu-an, 1878; repr. 4 vols., Taipei, 1966). 26 Herbert Franke, "A Sung embassy diary of 1211—1212: The Shih Chin lu of Ch'eng Cho," Bulletin de I'icole FranfaiseifExtreme-Orient, 69(1981), pp. 171-207. 27 Edouard Chavannes, "Voyageurs chinois chez les Khitans et les Joutchen," Journal Asiatique (1897 and 1898). Cambridge Histories Online © Cambridge University Press, 2008 682 BIBLIOGRAPHICAL ESSAYS Yuan Lou, recit d'un voyage dans le nord par Tcheou Chan."28 Another text based on personal experience among the Jurchen is Sung mo chi wen (Personal experiences in the northern deserts), 2 ' written by Hung Hao (1088— 1155), who was sent as an ambassador to Chin in 1129 and detained by the Chin who kept him interned until 1143 when he was released. The most complete edition of this work, which also includes sections omitted in the transmitted text, is that in the Yii-chang ts'ung shu. Much of the material on the early period of Chin in the San ch'aopei meng huipien was taken from the Sung mo chi wen, which is a fascinating firsthand account of life in the north in the 1130s. It should be evident that the sources for the Liao and Yuan dynasties are also of some importance in studying the formative stage and the collapse of Chin. The dynastic histories of Liao (Liao shih) and Yuan (Yuan shih) must therefore be consulted for the period before 1125 and for the early thirteenth century, respectively. A completely independent source of great importance for the relations between the Jurchen and the Korean state of Kory6 is the Koryd-sa>° (History of Kory6) by Ch&ng In-chi, which covers the period from 918 to 1392. It can be used for checking the Chin and Sung sources with regard to events on the northern borders of Kory6. Secondary scholarship on the Chin The scholarly study of Chin history was practically nonexistent in Yuan and Ming China. Only after the Manchus had conquered China in the seventeenth century did interest in the Chin state and in the Jurchen as the ancestors of the Manchus revive. In 1646, only two years after the formal establishment of the Manchu Ch'ing dynasty, an adaptation of the annalistic parts of the Chin history was translated into the Manchu language. This book, Aisin gurun-i suduri bithe^' (Annals of the "gold" state), was translated into French by Charles J. de Harlez in 1887. The first Chinese scholar to apply textual criticism to the Chin shih seems to have been Shih Kuo-ch'i (fl. 1790—1820). His Chin shih hsiang chiao>2 (Comparative details concerning the Chin history) is a valuable contribution to a better understanding of the Chin shih, above all in pointing out inconsistent orthographies and discrepancies among various chapters of the Chin shih. A large part of Shih Kuo-ch'i's emendations were incorporated into the 1975 Chung-hua shu-chii edition. The first attempts to compile an anthology of Chin literature by Chang Chin-wu and Chang Chung-fang date from the first half of the nineteenth century. But these activities were isolated, and Chin studies in China resumed only in the early twentieth century, for example, in the works of Wang Kuo-wei (1877-1927), who studied the border fortifications built in the 1190s by the Chin against the Mongols.'3 a8 Edouard Chavannes, "Pei Yuan Lou, Recit d'un voyage dans le Nord par Tcheou Chan," Young Pao (1904), pp. 162-92. 29 Hung Hao, Sung mo chi wen, in Yii-cbang ts'ung sbu (Nan-ch'ang, 1915). 30 Ch6ng In-chi et al., comps., KoryH-sa, 3 vols. (Tokyo, 1908—9). 31 Aisin gurun-i suduri bitbe, translated in Charles J. de Harlez, Histoire de Vempire Kin ou Empire d"Or, traduit de I'Aisin Gurun (Louvain, 1887). 32 Shih Kuo-ch'i, Cbin shih hsiang chiao (Peking, 1975). 33 Wang Kuo-wei, "Chin chieh hao k'ao," Yencbing hsiiehpao, 1 (1927), pp. 1 —14. Cambridge Histories Online © Cambridge University Press, 2008 BIBLIOGRAPHICAL ESSAYS 683 Until very recently scholarly interest in the Chin dynasty was much greater in Japan than in China. This interest coincided with political developments in Japan, which regarded Manchuria as a Japanese sphere of influence. Already before World War I, Japanese scholars had made considerable contributions to the history of Manchuria in Chin times, particularly in the field of historical geography. The studies by scholars like Inaba Iwakichi and Ikeuchi Hiroshi supplied a solid foundation for later generations of scholars. Some of their authoritative studies were translated into German in 1912 and 1914 under the direction of Shiratori Kurakichi.M These volumes also have an excellent index and are most useful to modern students who can read German but not enough Japanese to consult the original Japanese version.35 Even since World War II the flow of scholarly Japanese publications on the Chin and the Jurchen has continued without interruption. Two large-scale standard works deserve particular mention. Mikami Tsugio's monumental Kinshi kenkyu (Research on Chin history)36 is a revised collection of his many previous articles on problems in Chin history. The first volume, Kindai Joshin shakai no kenkyu (Research on the society of the Jurchen of the Chin period) (1972), is a meticulous study of the pre-dynastic and early dynastic Jurchen people and gives a detailed account of the rise of the Jurchen to dynastic power. The second volume, Kindai seiji seido no kenkyzu (Research on the political system of the Chin dynasty) (1970), is concerned with government and institutional structures from early dynastic times on. Volume 3, Kindai seiji shakai no kenkyu (Research on government and society of the Chin dynasty) (1973), concerns the problems of Jurchen control over the subject populations in the Chin state, the unification of the Jurchen tribes, and the preservation of Jurchen culture in a Chinese environment, and it also includes a chapter on Chin—Kory6 relations. All three volumes have indices. Another Japanese standard work is by Toyama Gunji, Kinchoshi kenkyiP1 (Research on the history of the Chin dynasty). Like Mikami Tsugio's work, it is mainly a collection of earlier studies. Some of these deal with national minorities in the Chin state, others with Chin-Sung relations and the role of the Mongols in the twelfth century. An important contribution from Toyama Gunji is his study of the role of Yellow River floods in the eventual decline of Chin economic power. In the West, practically no comprehensive works on the Chin existed until recently, apart from the pioneering studies by Edouard Chavannes. The chapters on the Chin in volumes 4 and 5 of Otto Franke's Geschichte des chinesischen Reiches^8 were for many years the most comprehensive account of Chin history in a Western language but must now be considered somewhat outdated. They are, however, still useful for their accounts of Sung-Chin relations and of military events. A mine of information is Lucien Gibert's Dictionnaire historique et geographique de la Mandchourie.39 Although no references to sources are given in this alphabetically arranged dictionary, it 34 Shiratori Kurakichi, Beitrdge zur historischen Geographic der Mandschurei, 2 vols. (Tokyo, 1912—14). 3 ; Shiratori Kurakichi, Yanai Wataru, Matsui Hitoshi, and Inaba Iwakichi, Manshu rekiihi chin, 2 vols. (Tokyo, 1913; repr. Tokyo, 1940). 36 Mikami Tsugio, Kinshi kenkyu, 3 vols. (Tokyo, 1970—3). 37 Toyama Gunji, Kinchoshi kenkyu, Toyoshi kenkyu sokan no. 13 (Kyoto, 1964). 38 Otto Franke, Geschichte des chinesischen Reiches, 5 vols. (Berlin, 1930—54). 39 Lucien Gibert, Dictionnaire historique et glographique de la Mandchourie (Hong Kong, 1934). Cambridge Histories Online © Cambridge University Press, 2008 684 BIBLIOGRAPHICAL ESSAYS provides many relevant data on Chin personalities and the peoples of ancient Manchuria. As a first source of information it remains distinctly valuable. The best comprehensive Western-language history of the Jurchen and the Chin state is by a Russian scholar, M. V. Vorob'ev, Chzhurchzheni i gosudarstvo Tszin'*° (The Jurchens and the state of Chin). Published in 1975, it brilliantly describes and analyzes events and socioeconomic structures. (For a review in a Western European language, see Herbert Franke's in Monumenta Serica.*') Another important comprehensive study is by Jing-shen Tao, The Jurchen in twelfth-century China: A study of sinicization,*1 published at almost the same time as Vorob'ev's book. It is much more than a study of sinicization, as the title would suggest; it describes government institutions, the recruitment of personnel, and also the economic situation of the Jurchen. A brief study of Chin society and economics, including the fiscal system, is that by Herbert Franke, Nordcbina am Vorabend der mongolischen Eroberungen: Wirtschaft undGesellschaft unter der Chin-Dynastie 1115-1234.V Among the contributions of contemporary Chinese scholars, mention must be made of Ch'en Shu's Chin shih shih pu wu chung** (Five supplements to the Chin history). This work is indispensable for the study of the history of the Jurchen clans, Jurchen personal names, and Jurchen genealogy. A particularly useful feature of Ch'en Shu's book is the list of variant readings for transcriptions of Jurchen names in different sources. Generally speaking, a linguistic study of Jurchen personal names in light of comparative Tungusic studies is long overdue, and Ch'en Shu's book would be an excellent basis for Jurchen onomastics. For socioeconomic conditions under the Chin we have now several detailed studies. Ho Ping-ti's "An estimate of the total population in Sung-Chin China"43 was the first and is easily the best study of population figures under the Chin. The foreign trade of Chin was examined by Kato Shigeshi, "So to Kinkoku to no Boeki ni tsuite" (Trade between the Sung and Chin states), and "Sokin boeki no okeru cha sen oyobi kinu ni tsuite" (On tea, copper coins, and textiles in Sung—Chin trade), first published in 1937 and 1935, respectively, and reprinted in the author's Shina keizaishi kosho.*6 The flow of currency between Chin and its neighbors was covered by Sogabe Shizuo in Nissokin kahei koryushiv (History of currency exchange among Japan, Sung, and Chin). Illicit trade between Sung and Chin was considered by Ch'iian Han-sheng in "Sung—Chin chien ti tsou ssu mao i"4* (Smuggling trade between Sung and Chin). 40 41 42 43 44 43 46 47 48 M. V. Vorob'ev, Chzhunhzbeni i gosudarstvo Tszin'(Moscow, 1975). Herbert Franke, Monumtnta Serica, 32 (1978), pp. 4 0 4 - 8 . Tao Jing-shen, The Jurchen in twelfth-century China: A study of sinicization (Seattle, 1976). Herbert Franke, Nordchina am Vorabend der mongolischen Eroberungen: Wirtscbaft und Gesellschaft unter der Chin-Dynastie in)—1234 (Opladen, 1978). Ch'en Shu, Chin shih shih pu wu (hung (Peking, i960). Ping-ti Ho, "An estimate of the total population in Sung—Chin China," Etudes Song in memoriam ttienne Baldzs, 1, (Paris, 1970). Kato Shigeshi, "So to Kinkoku to no boeki ni tsuite," 1937, "So Kin beki no okeru cha sen oyobi kinu ni tsuite," 1935; repr. in Shina keizaishi kosbo (Tokyo, 1953), vol. 2, pp. 247—304. Sogabe Shizuo, Nissokin kahei koryushi (Tokyo, 1949). Ch'iian Han-sheng, "Sung-Chin chien ti tsou ssu mao i," Bulletin of the Institute of History and Philology, Academia Sinica, 11 (1944), pp. 423—47. Cambridge Histories Online © Cambridge University Press, 2008 BIBLIOGRAPHICAL ESSAYS 685 A brief but succinct description of the Chin economy is that by Chang Po-ch'iian, Cbin-tai ching-chi shih-liieh*9 (Outline of economic history of the Chin dynasty). The tea monopoly of the Chin was competently described by Hok-lam Chan in "Tea production and tea trade under the Jurchen-Chin,"'° and the rather inefficient wine monopoly of the Chin was discussed summarily by Herbert Franke in "A note on wine."5' There are several modern studies in Western languages concerning the diplomatic and military relations between Sung and Chin. Dagmar Thiele, Der Abschluss eines Vertraga: Diplomatic zwischen Sung-undChin-Dynastie 1117-1123,1* describes in great detail the attempts of the Sung government to conclude a treaty with the emerging state of Chin in order to overthrow the already-toppling Liao dynasty, and the eventual break between Chin and Sung that led to the annihilation of the Northern Sung state. A textual study of the treaties between the two powers is that by Herbert Franke, "Treaties between Sung and Chin."" This covers the period up to the peace treaty of 1141 and also gives some information on later agreements. An episode of the war between the two states in 1205—8 is the subject of Corinna Hana's Bericht iiber die Verteidigung der Stadt Te-an wahrend der Periode K'ai-hsi 1205 bis 1208.™ This book includes not only a profusely annotated translation of the siege diary written by an eyewitness of the battles for the strategic town of Te-an but describes also the various embassies between Sung and Chin that led to a renewal of earlier treaties and the resumption of a precarious coexistence between Sung and Chin. Another study dealing with the Jurchen defector P'u-hsien Wan-nu, who established an ephemeral state of his own in Manchuria, is that by Iwai Hirosato, "The source and meaning of Ta-chen, the dynastic title of P'u-hsien Wan-nu."» The relations between the Chin state and Kory6 have been much studied by Michael.C. Rogers, "Studies in Korean history II: Kory6's military dictatorship and its relations with Chin," and "The regularization of Kory6—Chin relations (1116—1131)."56 These articles are based on both Chinese and Korean sources. The civilization of the early Jurchen has been studied through translations of relevant texts; see Herbert Franke, "Chinese texts on the Jurchen: Translation of the 49 Chang Po-ch'iian, Chin-lai ching chi shih liieh (Shen-yang, 1981). 50 Chan Hok-lam, "Tea production and tea trade under the Jurchen—Chin dynasty," ed. Wolfgang Bauer, Studia Sino-Mongolica, Festschrift fiir Herbert Franke (Wiesbaden, 1979), pp. 104-25. 51 Herbert Franke, "A note on wine," Zentralasiatische Sludien, 8 (1974), pp. 2 4 1 - 5 . my—112) 52 Dagmar Thiele, Der Abschluss eines Vertraga: Diplomatic zwischen Sung undChin Dynastie (Wiesbaden, 1971). 53 Herbert Franke, "Treaties between Sung and Chin," in ttudes Song in memoriam ttienne Balazs, 1, (1970), pp. 5 5 - 8 4 . 54 Corinna Hana, Bericht iiber die Verteidigung der Stadt Te-an wdhrend der Periode K'ai-hsi 1203 bis 1208 (Wiesbaden, 1970). 55 Hirosato Iwai, "The source and meaning of Ta-chen, the dynastic title of P'u-hsien Wan-nu," Memoirs 0/the Research Department of the Toyo bunko, 9 (1937), pp. 1 1 1 - 6 1 . 56 Michael C. Rogers, "Studies in Korean history, II: Koryd's military dictatorship and its relations with Chin," T'oung Poo, 47 (1959), pp. 4 2 - 6 2 ; Michael C. Rogers, "The regularization of Kory6-Chin relations (1116—1131)," Central Asiatic Journal, 6 ( 1 9 6 1 ) , p p . 5 1 - 8 4 . Cambridge Histories Online © Cambridge University Press, 2008 686 BIBLIOGRAPHICAL ESSAYS Jurchen monograph in the San ch'ao pet meng hut pien,"'>1 and "Chinese texts on the Jurchen II: A translation of chapter one of the Chin shih."*8 The material culture of the Jurchen, particularly in their pre-dynastic period, is now relatively well known, thanks to excavations by Soviet archaeologists. The remnants of Mo-ho (the ancestors of the Jurchen) civilization are the subject of E. I. Derevianko's Mokheskiepamiatniki srednogo Amur** (Mo-ho monuments of the middle Amur). Jurchen tombs in the Soviet coastal province have yielded much interesting material; see V. E. Medvedev, Kul'tura Amurskich Chzhurchzhenei Konets X—XI vek (po materialam gruntovykh mongil'nikovf° (The civilization of the Amur Jurchen in the late tenth and eleventh centuries). Another comprehensive study of Jurchen remains is by V. D. Len'kov, Metallugiia i metalloobrabotka u Chzhurchzhenei v XII veka (po materialam issledovanii Shaiginskogo gorodishcha)61 (Metallurgy and metal work among the Jurchen of the twelfth century based on materials from the village of Shaiginsk). A convenient summary of the results of Soviet archaeology is the article by A. P. Okladnikov and V. E. Medvedev, "Chzhurchzheni Priamur'ia po dannym arkheologii"62 (The Amur region Jurchen according to archaeological data). Unfortunately, there has not yet been any attempt to present the findings of the Soviet archaeologists in a Western European language. We are in a better position in regard to Chinese art and archaeology in the Chin state. Many recent Chinese excavations in Manchuria have been published in archaeological journals such as Wen wu and K'ao ku. The article by Susan H. Bush, "Literati culture under the Chin (i 122—1234),"6j broke new ground in establishing Chinese painting under the Chin as a separate phenomenon independent of Southern Sung developments. Important material for the architectual history of Peking in the Chin period is available in G. N. Kates, "A new date for the origins of the forbidden city."6* A comprehensive description of Buddhist art under the Chin is that by Sekino Tadashi and Takejima Takuichi, Ryokin jidai no kenchiku to sono butsuzb** (Architecture and Buddhist sculpture under the Liao and Chin dynasties). A fully documented art history of the Chin has yet, however, to be written. Chinese literature under the Chin, which was seminal for some popular literary 57 Herbert Franke, "Chinese texts on the Jurchen: A translation of the Jurchen monograph in the San ch'aopei menghuipirn," ZentralasiatischeStudien, 9 ( 1 9 7 5 ) , pp. 119—86. 58 Herbert Franke, "Chinese texts on the Jurchen, II: A translation of chapter one of the Chin shih," Zentralasiatische Studien, 12 (1978), pp. 413—52. 59 E. I. Derevianko, Mokheskie pamiatniki Srednogo Amur (Novosibirsk, 1975). 60 V. E. Medvedev, Kul'tura Amurskich Chzhurchzhenei Konets X-XI vek (po materialam gruntovykh mogil'nikov) (Novosibirsk, 1977). 61 V. D . Len'kov, Metallurgiia i metalloobrabotka u Chzhurchzhenei v XII veka (po materialam issledovanii Shaiginskogo gorodishcha) (Novosibirsk, 1974). 62 A. P. Okladnikov and V. E. Medvedev, "Chzhurchzheni Priamur'ia po dannym arkheologii'o," Problemy Dal'nego Vostoka, 4 (1974), pp. 1 1 8 - 2 8 . 63 Susan H. Bush, "Literati culture under the Chin (1122—1234)," Oriental Art, n.s., 15 (1969), pp. 103-12. 64 G. N . Kates, "A new date for the origins of the forbidden city," Harvard Journal of Asiatic Studies, 2 (1943), pp. 180—202. 65 Sekino Tadashi and Takejima Takuichi, Ryokin jidai no kenchiku to sono Butsuzo, 2 vols. of plates (Tokyo, 1934). Cambridge Histories Online © Cambridge University Press, 2008 BIBLIOGRAPHICAL ESSAYS 687 genres, has been repeatedly studied in the West, for example, in J. I. Crump, "Yuanpen, Yuan drama's rowdy ancestor,"66 and Stephen H. West, Vaudeville and narrative: Aspects of Chin theater.6'1 The most famous of the Chin chantefables was translated and studied in M. Dolezelova-Velingerova andJ. I. Crump, Liu Chih-yiian chu kung tiao: Ballad of the hidden dragon.6* Although Buddhism was favored by the Chin court, the most important religious innovation of this period was the emergence of the Taoist Ch'uan-chen school. The first author to acquaint the West with these developments was Arthur Waley, in his Travels of an alchemist,69 who used Taoist literature for his book. An authoritative study of the Ch'iian-chen school is that by Ch'en Yuan, Nan-Sung ch'u Ho-pei hsin taochiao k'ao10 (A study of renewed Taoism in Hopei at the beginning of Southern Sung). His interpretation of the Ch'uan-chen school as a Chinese anti-Jurchen movement seems not quite to the point, but as a source book, Ch'en Yiian's work remains unrivaled. The best study of the law of the Chin dynasty is by Yeh Ch'ien-chao, Chin lit chih yen chiu11 (A study of Chin law). The author not only analyzes the textual evidence for customary law of the Jurchen but also compares the remnants of the Chin codes with T'ang law. A brief study of the interplay of Jurchen customary law and the codified Chinese law under the Chin is that by Herbert Franke, "Jurchen customary law and the Chinese law of the Chin dynasty. "72 Finally, some bibliographical information must be given on the study of the Jurchen language and script. There is a survey of extant documents in Jurchen scripts by Osada Natsuki, "Joshin moji to genson shiryo"73 (The Jurchen script and its remaining documents), but this is now outdated because of new discoveries in Manchuria and elsewhere. Our modern knowledge of the Jurchen script goes back to a glossary with appended documents compiled under the Ming in the early sixteenth century. The glossary is arranged according to categories. For each word the original Jurchen characters, the phonetic transcription in Chinese, and the meaning in Chinese are supplied. The language of this invaluable document is late Jurchen. Another late Jurchen vocabulary, but without script, was published by Ishida Mikinosuke, "Jurcica."'1' The pioneer study of the glossary with script was made by Wilhelm Grube, Die Sprache und Schrift derjulen,1^ who deciphered the 871 Jurchen 66 J. I. Crump, "Yuan-pen, Yuan drama's rowdy ancestor," East and West, 14 (1970), pp. 4 7 3 - 9 1 . 67 Stephen H. West, Vaudeville and narrative: Aspects ofChin theater (Wiesbaden, 1977). 68 M. Dolefelovd-Velingerovi and J. I. Crump, trans., Liu Chih-yiian chu kung tiao: Ballad of the hidden dragon (Oxford, 1971). 69 Arthur Waley, Travels of an alchemist: The journey of the Taoist Ch'ang-ch'un from China to the Hindukush at the summons ofChingiz Khan. Recorded by his disciple Li Chih-ch'ang (London, 1931). 70 Ch'en Yuan, Nan-Sung ch'u Ho-pei hsin tao chiao k'ao (Peking, 1941; repr. Peking, 1962). 71 Yeh Ch'ien-chao, Chin lit chih yen chiu (Taipei, 1972). 72 Herbert Franke, "Jurchen customary law and the Chinese law of the Chin dynasty," in State and law in East Asia: Festschrift Karl Biinger, ed. Dieter Eikemeier and Herbert Franke (Wiesbaden, 1981), pp. 215-33. 73 Osada Natsuki, "Joshin moji to genson shiryo," Rekisbi kyoiku, 18 (1970), pp. 2 5 - 3 1 . 74 Ishida Mikinosuke, "Jurfica," in Ikeuchi hakushi kanreki kinen Toyoshi ronso (Tokyo, 1940), pp. 39—57; repr. in his Toa bunkashi soko (Tokyo, 1973), pp. 71—86. 75 Wilhelm Grube, Die Sprache und Schrift derjulen (Leipzig, 1896). Cambridge Histories Online © Cambridge University Press, 2008 688 BIBLIOGRAPHICAL ESSAYS words contained in the original manuscript. After Wilhelm Grube, many more scholars in Japan and Europe have studied the Jurchen language and script. Louis Ligeti contributed important improvements to previous studies, especially of Jurchen phonology, in his articles: "Note pr£liminaire sur le dechifrrement des petits caracteres Jou-tchen"76 and "Les inscriptions djurtchen de Tyr. La formule om mani padme hum."" Whereas Grube had only one single manuscript of the glossary and the accompanying documents at his disposal, a new study that might be regarded as definitive used additional manuscripts preserved in Japan. This is the book by Gisaburo N. Kiyose, A study of the Jurchen language and script: Reconstruction and decipherment. ' 8 The phonology of the Jurchen language studied by scholars like Grube and Kiyose is that of a period around 1500 and therefore is different from the old Jurchen of the twelfth and thirteenth centuries. The script recorded in the Ming manuscripts of the glossaries is, however, similar to and to a great extent identical with the old Jurchen script found in stone inscriptions. These range in date from 1185 to 1413. It therefore has been possible to decipher these inscriptions to a certain extent, insofar as they used characters already known from the Ming bilingual glossary. The best study of the surviving inscriptions in Jurchen script is now that by Chin Kuang-p'ing and Chin Ch'i-tsung, Nii-chen yii yen wen tzu yen chiu19 (Study of the Jurchen language and script). The authors succeeded in proposing plausible readings and translations for the greatest part of the extant inscriptions. This epigraphic material is not perhaps very informative in content, and for historians the surviving Chinese inscriptions from the Chin period are much more important. The old Jurchen texts are, however, invaluable as evidence of the multilingual culture of the Chin state. Almost all known texts in Jurchen script are on stone or metal (such as seals), but a single piece ofJurchen writing on paper has survived. It is kept in the Leningrad Oriental Institute and was published some years ago, see D. Kara et al., "Pervaia nakhodka chzhurchzhen'skikh rukopisnykh tekstov na bumage" (The first find of Jurchen manuscript texts on paper).80 The paper manuscript is still undeciphered. The best modern study of the Sino-Jurchen glossary without script is that by D. A. Kane, The Sino-Jurchen vocabulary of the Bureau of Interpreters .tl Its language probably reflects an even later stage than does the language of the glossaries with script. 76 Louis Ligeti, "Note priliminaire sur le dechiffrement des petits caracteres Jou-tchen," Ada Onentalia Academiae Scientiarum Hungarian, 3 (1933), pp- 221—8. 77 Louis Ligeti, "Les inscriptions djurtchen de Tyr. La formule om mani padmt hum," Ada Onentalia Academiae Scientiarum Hungaricae, 12 (1961), pp. 5 - 2 6 . 78 Gisaburo N. Kiyose, A study of tht Jurchen language and script: Reconstruction and decipherment (Kyoto, 1977)79 Chin Kuang-p'ing and Chin Ch'i-tsung, Nii-chen yu yen wen tzu yen chi (Peking, 1980). This work was the basis for the dictionary of Jurchen characters by Chin Ch'i-tsung, Nii-chen wen tz'u lien (Peking, 1984). 80 D. Kara et al., "Pervaia nakhodka chzhurchzhen'skikh rukopisnykh tekstov na bumage," Pis'mennye Pamiatniki Vostoka (1969), pp. 2 2 3 - 3 8 . 81 D. A. Kane, The Sino-Jurchen vocabulary of the Bureau of Interpreters (Ph.D. diss., Australian National University, 197;). A revised and enlarged edition is that by Daniel Kane, The Sino-Jurchen vocabulary of the Bureau of Interpreters (Bloomington, 1989). Cambridge Histories Online © Cambridge University Press, 2008 BIBLIOGRAPHICAL ESSAYS 689 It is obvious that this brief bibliographical essay cannot do justice to all the scholarly achievements in the field of Chin historical studies. The selection of titles has been, to a certain extent at least, subjective. It will, however, have demonstrated that our present state of knowledge is due to the individual and fragmented efforts of scholars in many different countries. There have been few attempts at a broad synthesis. It is striking that in many cases little if any attention was paid to Japanese scholarship in China, and vice versa. And in both countries, even the leading scholars blithely ignore the achievements of Western scholarship. The result is some unnecessary duplication of effort because the available sources are more or less the same. In any event, a comprehensive bibliography of Chin studies remains a definite desideratum because it could help bridge the existing gaps among scholars in this field from all countries. A NOTE ON TRADITIONAL SOURCES FOR YUAN HISTORY The official history of the Yuan The modern historian's essential resource for the study of the Yuan period is the Yuan shih (Official history of the Yuan) in 210 chiian, compiled during 1369 and 1370. The printing blocks for the first edition were engraved late in 1370; it has been frequently edited and reprinted. The standard edition of the work now is the punctuated, collated edition (fien chiao peri) in fifteen volumes,1 published by Chung-hua shu-chii in 1976, which is also available in various facsimile reprintings. The Po-na edition of the twenty-four dynastic histories2 includes a facsimile of the 1370 edition, slightly marred by the editors' efforts to restore unclear or missing characters from the copy used for photo duplication.' It has often been noted that this large history, composed of 47 chiian of basic annals (pen-chi), 8 chiian of tables (piao), 58 chiian of treatises (chih), and 97 chiian of biographies {lieh-chuan), was compiled in greater haste than the other dynastic histories were, and in consequence it suffers from inadequate preparation and careless editing. Although it is true that errors abound,4 the work's defects by the standards of traditional Chinese historiography are somewhat offset in the minds of modern scholars by its other virtues. That is, the work consists in significant part of undigested and verbose raw materials hastily copied into it at length, reflecting all too little of the historian's analysis and judgment. Thus, it lacks the internal coherence •By F. W. Mote. 1 Sung Lien et al., eds., Yuan shih (Official history of the Yuan) (Peking, 1976), in 210 chiian. 2 The Po-na edition of the twenty-four dynastic histories (erh shih ssu shih) (Shanghai: Commercial Press, 1935)3 See the discussion of editions and editorial procedures in the preface to the 1976 edition. 4 See the examples cited inChaol (1727—1814), Nien erh shih cha chi cbi pu-pien, Tu Wei-yiin, ed. (Taipei, 1975) cb. 29, pp. 6 4 2 - 7 8 ; and Ch'ien Ta-hsin (1728-1804), Shih chia thai yang hsin lu (Shanghai, 1935; repr. 1957), ch. 9, pp. 195-225- Cambridge Histories Online © Cambridge University Press, 2008 690 BIBLIOGRAPHICAL ESSAYS and succinct exposition of the meticulously polished Mingshih (Official history of the Ming), on which a succession of commissions toiled on and off over a span of ninety years. Yet precisely because the Yuan shih was not refined in that way, it rather fully preserves the materials on which it drew, many of which otherwise no longer exist. Moreover, the two commissions of scholars who worked on the Yuan shih in Nanking, the new Ming capital, under the general direction of Sung Lien (1310—81) and Wang Wei (1322—72) — a total of sixteen scholars who worked for 188 days in 1369 and fourteen scholars who worked for 143 days in 1370 (with, of course, a staff of scribes and assistants) — all were men who had lived, and many of whom had served, under the Yuan government and thus had direct knowledge of the period. Nonetheless, it is astonishing that such a work could be put together in a mere 331 days. It might be doubted that any of its compilers could have read it through before it was submitted to the throne. The first of the two historical commissions, working in 1369, brought the work down to the year 1333, the beginning of the reign of Togh5n Temiir, the emperor Shun (r. 1333—68, died in Mongolia in 1370). Those compilers utilized what are referred to as the "veritable records" (shih-lu) of thirteen reigns; "thirteen reigns" refers to all the great khaghans and succeeding emperors from Chinggis khaghan to Irinjibal, the emperor Ning-tsung, who died in the spring of 1332 after reigning for only fifty-three days. The thirteen veritable records and other archival materials were saved from probable destruction by the heroic actions of a few Chinese scholarofficials serving the Mongolian government in September 1368 as the Ming armies entered the Yuan capital at Ta-tu (modern Peking).5 Those archives were promptly transported to Nanking and in the following year were made available to the Yuan History Commission. As the responsible officials stated in their "Memorial for presenting the Official history of the Yuan," they felt compelled to terminate their account in 1333 because of the lack of a veritable record covering the last reign and thus produced a truncated work in 159 chiian.6 But such an incomplete history was considered intolerable, and so a second Yuan History Commission was established in the following year, 1370, to bring the political account to 1368. Of course, no committee was established to compile a veritable record for the reign of Toghon Temiir; that would not be done by a succeeding dynasty. Nonetheless, more materials were sought, biographies were added, and other portions of the work were supplemented, producing a total of 53 new chiian. The first commission's 159 chiian work plus the second commission's 53 new chiian should have produced a work with a total of 212 chiian instead of the 210 chiian found in the final product. The discrepancy has been explained variously: Some of the new chiian probably were combined with others among the first 159 chiian. It is impossible to know much about the veritable records of thirteen reigns, as no 5 See the biography of Wei Su by Hok-tam Chan in L. Carrington Goodrich and Chaoying Fang, eds., Dictionary of Ming biography (New York and London, 1976) (hereafter cited as DMB), p. 1465. The preservation of Yuan archives at Peking in 1368 is also discussed by Francis Cleaves in the introduction cited in n. 20. 6 See Francis Woodman Cleaves, "The memorial for presenting the Yuan sbib," Asia Major, 3rd series, 1 (1988), pp. 5 9 - 6 9 - Cambridge Histories Online © Cambridge University Press, 2008 BIBLIOGRAPHICAL ESSAYS 691 portion of them is known to exist now, nor were they described by historians of historiography in the past who might have examined them. They were not, in any event, the products of daily record keeping at the court, as were the veritable records of Chinese historiographical tradition.7 Nonetheless, they probably came into being as a consequence of the urgings of history-minded Chinese scholars at the Yuan court. In 1260 the scholar-official Wang O, who had served the former Chin dynasty (before its annihilation in 1234) and who had joined Khubilai's entourage in the 1250s, urged Khubilai khaghan to establish a Historical Bureau to compile, somewhat after the fact, the veritable records for all the preceding reigns before the historical knowledge was lost. Accordingly, in 1262 a decree was handed down, commanding Wang O and others to give special attention to historical issues, leading them to seek out records and oral traditions, although the Mongolian rulers were opposed to allowing their Chinese subjects full access to their historical records and oral traditions.8 The resulting book, (Huang Yuan) Sheng wu ch'in ch'eng lu,9 which was compiled shortly thereafter, has sometimes been credited to Wang O, but that is not generally held to be true today. It is broadly informed and more accurate than the subsequently written-down Secret history, especially in its coverage of the political narrative.10 Similarities of wording in the basic annals (which are derived from the veritable records) show that it became a principal source used in compiling the veritable records of the reigns of Chinggis khan (posthumously designated Emperor T'ai-tsu), Ogodei khaghan (T'ai-tsung), Giiyiig khaghan (Ting-tsung), and Mongke khaghan (Hsien-tsung), all of whom are granted basic annals (chiian 1-3) in the Yuan shih. These preceded the basic annals of Khubilai himself, imposing a Chinese pattern of imperial reigns on the exploits of the first four Mongolian great khans, even though none of the four proclaimed himself, or was in his own time regarded as having been, an emperor of China. Those chapters in the Yuan shih are, however, fragmentary at best, showing that the efforts to compile veritable records for their lives had not been fully successful in recovering the necessary historical information. It is not known to what extent historians in the 1260s or, for that matter, thereafter to the end of the Yuan dynasty, may have had access to Mongolian sources, oral or written. From the 1260s onward, however, the veritable records for the nine reigns from Khubilai to Ning-tsung were prepared at the end of each reign, although 7 See L. S. Yang, "The organization of Chinese official historiography: Principles and methods of the standard histories from the Tang through the Ming dynasty," in W. G. Beasley and E. G. Pulleyblanlc, eds., Historians of China and Japan (London, 1961), pp. 4 4 - 3 9 . The organization of the Official history of the Yuan Commission receives special notice. 8 Hok-lam Chan provides an excellent discussion of Wang O's influence on historiography at the early Yiian court; see his "Wang O (1190-1273)," Papers on Far Eastern History, 12 (September 1975), pp. 43—70, esp. pp. 54—7. This is a provisional version of a biography to be included in the Yiian Biographical Project, under the direction of Igor de Rachewiltz. 9 Sheng wu ch'in cheng lu chiao chu. In Meng-ku shih liao ssu chung, ed. Wang Kuo-wei (Peking, 1926; repr. Taipei 1962, 1975). 10 There is a partial translation begun by Paul Pelliot that was continued and published by his student L. Hambis: Histoire da campagnts de Gengis khan (Leiden, 1951). Cambridge Histories Online © Cambridge University Press, 2008 692 BIBLIOGRAPHICAL ESSAYS there is no evidence that a regular system for recording the daily words and deeds of the rulers was established to produce the ch'i chii chu (Diary of action and repose) on the model of previous official Chinese historiography to serve as the basis for compiling veritable records upon the demise of a ruler. Nonetheless, the basic annals of most reigns in the Yiian shih from the reign of Khubilai khagan (Emperor Shih-tsu) onward, however they may have been compiled, are unusually detailed and verbose, though poorly edited. The treatises (chih) in 53 chiian appear to have been largely compiled from the Ching shih ta tien," a massive compilation in 880 (or, as sometimes reported, in 800) chiian, prepared by scholars of the K'uei-chang ko (Pavilion of the star of literature), an imperial library and art collection founded by Tugh Temiir, the emperor Wentsung (r. 1328—32), to serve as an imperial academy." Compilation of the Ching shih ta tien was principally under the supervision of Yu Chi (1272—1348), a leading literary official during the later Yuan period.'3 The vast work was never printed. All manuscript copies seem to have disappeared before the end of the Ming dynasty. Less than 5 percent of it remains today, principally as items copied into the Yung-lo ta tien in the first decade of the fifteenth century, a work that itself has survived only in small part. Because the manuscript copy of the Ching shih ta tien was at their disposal, however, it has long been surmised that the compilers of the Yuan shih were able to compile the treatises by drawing directly from it. Indirect evidence for this has been noted; for example, the Ti-li chih (Treatise on administrative geography)14 chapters 58-63 includes changes in administrative geography through the year 1331 but none thereafter. Scholars have long echoed the observation of Ku Yen-wu (1613— 82) that the texts of the treatises read like working documents from government offices, preserving terminology and references peculiar to such materials but not expected in historical writings.'5 There is no i-wen chih (Treatise on literature), a serious deficiency. That led Ch'ien Ta-hsin (1728—1804) to compile the Pu Yuan shih i wen chih (A treatise on literature to supplement the Official history of the Yiian),16 among his other detailed studies of Yiian history (including his notable table of tribal lineages, the Yiian shih shih tsu piao, ' 7 which may have been intended for a new Yiian shih that he failed to complete). The six tables (piao) in eight chiian of the Yiian shih are for imperial consorts, the imperial house lineage, princedoms, imperial princesses and their consorts, the three dukes (san kung), and chief ministers (tsai-hsiang). They are somewhat incomplete, and they introduce the names of many high-ranking officials for whom there are no 11 See Su Chen-shen, Yiian chtng shu Ching shih ta tien chih yen-cbiu (Taipei, 19S4), p. 270, for a useful study of the Ching shih ta tien. Su is critical of the way that the work was used by the Yiian shih compilers. 12 See Chiang I-han, Yuan tai K'uei chang ko chi K'uei changjen wu (Taipei, 1981). 13 John D. Langlois, Jr., "Yu Chi and his Mongol sovereign: The scholar as apologist,"Journal 0/Asian Studies, 38(1978), pp. 99—116, esp. pp. 108-10. [4 Yiian shih, chaps. 58—63. 15 Ku Yen-wu, Jih chih lu, ch. 26, "Yiian shih." 16 Ch'ien Ta-hsin, Pu Yiian shih i wen chih (Kiangsu, 1874). Preface dated 1791. 17 Ch'ien Ta-hsin, Yiian shih shih tsu piao (Kiangsu, 1874). Preface dated 1791. Cambridge Histories Online © Cambridge University Press, 2008 BIBLIOGRAPHICAL ESSAYS 693 biographies in the biographical chapters of the work, a serious anomaly by Chinese historiographical standards. The biographies (lieh-chuan) in ninety-seven chiian are the most frequently criticized portion of the Yuan shih. In no fewer than nine cases, for example, one individual (all non-Han persons) appears as the subject of two biographies, although his name is written differently, as if two different persons were being recorded. In many cases the name of one person is written differently in different chapters. Many other faults in the biographies have been pointed out, especially during the Ch'ing dynasty when several important attempts were made to supplement or supersede the Yuan shih. Of particular concern then was the inadequacy of all kinds of information on the pre-1260 portions of Mongolian history. That concern remains; in this century it has until recently been addressed primarily by Japanese and Western historians. A high point of interest in the Yuan period by Chinese scholars was reached in the late eighteenth and nineteenth centuries,'8 but probably no Han Chinese historians during all of Ming and Ch'ing times ever acquired a sound working knowledge of Mongolian. Their emphasis was mainly on China in the period of alien rule, not on the Mongols themselves. The Secret history The Mongols' most important record of their own early history is the so-called Secret history of the Mongols, usually called in Chinese Yuan ch'ao pi shih (Secret history of the Yuan dynasty). Its name in Mongolian as reconstructed from the Chinese-character transliteration is Mongghol-un ni'ucha tobchiyan, which is more correctly translated into Chinese as Meng-ku pi shih, with the same meaning as the English title. Both Chinese names are now used. The existing translation into Chinese accompanied by a transliteration using Chinese characters as (highly imperfect) phonetic symbols is the only surviving written version and has been the basis for all modern reconstructions of the original Mongolian text and for its translations into modern Chinese, Japanese, and Western languages.I9 Scholars do not agree on the date when the Secret history was composed, the date when it was first written down in Mongolian (using Uighur script), the date when it was first translated into Chinese, or the date when it was first published. The 18 See Cheng Ho-sheng, "Ch'ing ju tui yii Yuan shih hsiieh chih yen chiu," Shih ti hsiiehpao, 3, no. 4 (December 1924), pp. 1-23; and 3, no. 5 (March 1925), pp. 1-22. Also Tu Wei-yiin, Ch'ing taishih hsiieh yii shih chia (Taipei, 1984). 19 See the introduction of Francis Woodman Cleaves, The secret history of the Mongols, for the first time done into English out of the original tongue and provided with an exegetical commentary, pt. I (translation) (set in type, 1957; Cambridge, Mass., 1982), pp. xvii—lxv. For references to other scholarship on the Secret history, see Thomas T. Allsen, The Mongols in East Asia, twelfth—fourteenth centuries: A preliminary bibliography of books and articles in Western languages. Sung Studies Newsletter (Philadelphia, 1976), esp. pp. 7—10; and Sechin Jagchid (Cha-ch'i Ssu-ch'in), Meng-ku pi shih hsin i ping chu shih (A new translation of the Secret history of the Mongols with commentary), with an introduction by Yao Ts'ung-wu (Taipei, 1979). The latest translation of the Secret history into English is that by Igor de Rachewiltz, published serially in Papers on Far Eastern History, nos. 4, 5, 10, 13, 16, 18, 21, 23, 26, 30, and 31 (Canberra, 1971—85). Cambridge Histories Online © Cambridge University Press, 2008 694 BIBLIOGRAPHICAL ESSAYS complex story of its transmission, first as an oral work and then as a written text in Mongolian and in Chinese transliteration and translation, was extensively examined by Francis Woodman Cleaves in the introduction to his translation. Cleaves argues that the text was first written down between 1228 and 1264. Some scholars, however, place this date as late as 1340. He accepts as highly.likely the date of 1368-70 for the translation into Chinese with accompanying transliteration, yet he agrees that there is no evidence that it was used by the compilers of the Yuan shih. It is also unclear whether the translation-with-transliteration, after being made on command of the Ming court in 1382, was published (in 1387?) along with the Hua i i yii, an imperially ordered glossary of Mongolian terms with Chinese equivalents, or whether a manuscript version of it was merely consulted by its two Uighur compilers. Fragments of a Hung-wu period printed version of the latter still exist, but whether existing fragments of a printed version of the Secret history were printed at the same time is the subject of scholarly disagreement. No matter how many questions about the provenance of the Secret history are eventually resolved, all scholars agree that it is of first-rank importance for the study of Yuan history and Mongolian civilization. It first narrates in storyteller fashion the legendary origins of the Mongolian tribe and then offers a more detailed historical account, starting with the early life of Chinggis khaghan and running to late in the reign of his son and successor Ogodei khaghan, that is, to about 1240. The contemporary scholar Yao Ts'ung-wu characterized the Secret history as follows: "When one goes beyond the Chinese standard histories and Chinese historical documentation, it is the only large-scale historical work written in Mongolian and from a Mongolian point of view that directly reports on the life of peoples of China's border regions beyond the Great Wall." He values it for reliably reporting aspects of history ignored by the Chinese histories.20 This book's form is somewhat more literary than historiographical, causing it to be criticized by some for its obvious inaccuracies and its mythic narrative style, while being praised by others for the richness of its social history content. The Secret history lay unnoticed by Chinese scholars through most of Ming and Ch'ing times, to be rediscovered in the nineteenth century. First the Chinese scholar Ku Kuang-ch'i (1776—1835) found a copy that he collated and made known to scholars in 1805; then a Russian priest stationed in Peking, known as the Archimandrite Palladius, produced his translation into Russian in 1877, basing that on the accompanying abridged Chinese translation (not on a reconstructed Mongolian text), using a manuscript derived from the Yung-lo ta tien. That translation, however imperfect, made the work known in the West and initiated a century of interest in the book beyond China's borders. It has become a specialized field of scholarship bearing on Yuan history. 20 Yao's essay on the study of the Secret history is reprinted as a preface to Sechin Jagchid's Chinese translation of the work, cited in n. 19. Jagchid's preface (pp. 2 1 - 4 ) adds further information on studies of the work up to 1978. Cambridge Histories Online © Cambridge University Press, 2008 BIBLIOGRAPHICAL ESSAYS 695 Chinese knowledge of Mongolian history beyond China Some knowledge of Mongolian oral traditions and of the then-still-not voluminous writings in Mongolian, existed among Chinese scholars in Yuan times and is reflected in Chinese writings available to the two historical commissions, but in early Ming times the extent of such knowledge still appears to have suffered from general Chinese indifference as well as from the Yuan tradition of withholding it from Chinese scholars.21 This is unfortunate. The Yuan shih had to function for the Chinese not only as their history of the Yuan dynasty in China after 1260, the year of Khubilai's accession (more properly, after his proclamation of the founding of the dynasty and the adoption of the name Yuan in 1272) but also as their primary record of the entire Mongolian national history in the thirteenth and fourteenth centuries. Somewhat anomalously, the Chinese took pride in the vast extension of the empire created by the force of Mongolian arms yet did not seek to know the historical details. Then as now, that empire (by reason of its alien rulers having been accepted as legitimate emperors of China) was seen as a triumph for China, not as that of a Mongolian world empire to which the Chinese were subject. The Yuan shih, however, virtually ignores the Mongolian empire beyond China and Eastern Asia, causing a serious gap (but one not seriously regarded, with a few notable exceptions) in Chinese historians' knowledge. This knowledge gap did not begin to be filled until the late nineteenth century. It was especially the diplomat Hung Chun (1840-93) who, while stationed in Berlin and St. Petersburg, discovered European and Western Asian sources and current scholarship, thereby bringing into his awareness startling new information that could greatly supplement Mongolian and Chinese history.22 Much of this new material, via Hung's translations of it, was incorporated into the Hsin Yuan shih (New official history of the Yuan) written by K'o Shao-min (1850-1933) and published in 1922.2' Yet in general, Chinese awareness of the Mongols as major players on the stage of world history quite apart from their place in Chinese history or, alternatively stated, apart from China's place in Mongolian history, or even the acceptance by the Chinese of Mongolian history and civilization as proper subjects of study have had to await the broadening of the Chinese worldview in the twentieth century. Scholars in China, including Chinese as well as Mongolian, Manchu, and other Inner Asian scholars, have made important contributions to that broadening of Chinese horizons during the past one hundred years. They have built on a tradition of historiographical scholarship that for several centuries was focused on revising and supplementing, or superseding, the Yiian shih. 21 See Langlois, "Yii Chi and his Mongol sovereign," p. 109. 22 See Hung's biography by Tu Lien-che in Arthur W. Hummel, ed., Eminent Chinese of the Ch'ing period (Washington, D.C., 1943—4), pp. 3 6 0 - 1 . Hung translated into Chinese some medieval Western Asian materials as well as some then current European scholarship relevant to Western Asian sources. 23 K'o Shao-min, Hsin Yuan shih (New official history of the Yiian) (Tientsin, privately published, 1922). Cambridge Histories Online © Cambridge University Press, 2008 696 BIBLIOGRAPHICAL ESSAYS Ming and Ch'ing scholarship on the official history of the Yuan Awareness of the need to improve the faulty Yuan shih dates back virtually to its first publication in 1370. Hsieh Chin (1369—1415) appears to have been commanded by the first Ming emperor, late in the Hung-wu reign period (1368—98), to prepare a corrected version, but nothing came of that.2* Little further study of Yuan history followed during Ming times,2' the important exception being the Yuan shih chi shih pen mo'6 {.Official history of the Yuan retold in topical format), in twenty-seven chiian, completed in 1606 by Ch'en Pang-chan (d. 1636). That substantial retelling of Yuan history, by one of the scholars who had just completed a similar reworking of Sung history, is highly regarded in Chinese historiographical tradition. For a number of reasons, the Ch'ing period, in contrast with the Ming, was one of renewed interest in the Yiian period. A succession of works produced then importantly supplement and correct the Yuan shih but do not supplant it. The more important of those are the following:27 1. Shao Yiian-p'ing (ca. 1664), Yiian shih lei-pien,28 42 chiian, completed in 1693. A not-entirely satisfactory attempt to supersede the Yiian shih, arranging the contents on the pattern of the Sung dynasty institutional history, Tung chih, by Cheng Ch'iao (1102—60).29 2. Ch'ien Ta-hsin (1728-1804). In addition to producing the two "supplements" intended for the Yiian shih, Ch'ien devoted much of his long and brilliant scholarly career to textual and critical studies that are said to have been intended to produce a new Yiian shih. Manuscript portions of such a work are reported to have been seen by nineteenth-century scholars, but it is unlikely that more than a few studies toward such an undertaking were in fact produced. Ch'ien's various writings contain a large number of critical studies of topics relevant to the Yiian period. Note in particular Ch'ien's Shih chia chai yang hsin lu, 20 chiian, also his Nien erh shih k'ao i,*° 100 chiian, and his collected works, the Ch'ien yen t'ang wen chi,il 70 chiian. Ch'ien's specialized studies of Yiian history were discussed in the context of Ch'ing period historiography by Cheng Ho-sheng and Tu Wei-yun (cf. note 11). 3. Wang Hui-tsu (1731-1807), Yiian shih pen cheng, 50 chiian.*2 The title stresses "verification of facts" (pen cheng); this is Ch'ing critical scholarship of a 24 See Hsieh Chin's biography in Ming sbih (Peking, 1974), 147, p. 4120. 25 Li Ssu-ch'un, Yiian shih bsiieb (Shanghai, 1926), pp. 5 8 - 6 1 , notes three or four other works relevant to Yiian history produced during Ming times. 26 Ch'en Pang-chan, Yiian shih chi shib pen mo (Peking, 1955, repr. 1979). 27 Li Ssu-ch'un's Yuan shih bsiieb, though more than fifty years old, still provides a useful discussion of Ch'ing period scholarship on the Yiian, esp. on pp. 6 1 - 7 4 . 28 Shao Yiian-p'ing, Yuan sbih lei-pim. Original printed edition, 1699. 29 See the brief reference to this work in Hok-lam Chan, "Chinese official historiography at the Yiian court: The composition of the Liao, Chin, and Sung histories," in China under Mongol rule, ed. John D. Langlois, Jr. (Princeton, 1981), p. 103 and n. 140. 30 Ch'ien Ta-hsin, Him erh shih k'ao i (Peking, 1935; repr. 1958). 31 Ch'ien Ta-hsin, Ch'ien yen t'ang wen chi (Shanghai, 1929). 32 Wang Hui-tsu, Yuan shih pen cbeng, 2 vols. (Peking, 1984). Cambridge Histories Online © Cambridge University Press, 2008 BIBLIOGRAPHICAL ESSAYS 4. 5. 6. 7. 697 high order, focusing on the correction of errors, omissions, and misrendered names. Wei Yuan (1794-1856), Yiian sbih hsinpien, 95 cbiian, completed in 1853.5' The first full-scale rewriting of Yiian history in the format of a standard history, using the Secret history and Yiian period writings of many kinds. Tseng Lien (b. i860), Yiian shu,* 102 chiian. A reworking of Wei Yiian's work just cited; of limited usefulness today but further evidence of the importance of Yiian history in later Ch'ing critical scholarship. Hung Chun (1840—93), Yiian shih i wen chengpu," 30 cbiian. T'u Chi (1856—1921), Meng wu erh shih chi. This is a notable late work of traditional scholarship. As an official stationed in Manchuria in the 1890s, T'u became interested in the geography of China's northern frontiers and spent the last twenty years of his life studying the history of the Mongolian nation, calling his work The historical record of the Mongols. That choice of title is to indicate that the work is a history of the Mongolian nation, not of the Yiian dynasty in Chinese history. He was influenced by the new knowledge introduced by Hung Chiin. In many ways, The historical record surpasses all other Ming and Ch'ing works in the field, including K'o Shao-min's Hsin Yiian shih (1919—30),36 in breadth of materials used and in accuracy. It is a work of high reference value for scholars today.37 The Yiian tien chang and other specialized topics in Yiian history The Yiian tien chang, sixty chiian, is a compilation of codes and regulations issued from about 1270 to about 1320. As a legal scholar described it: The text consists of a huge collection of codes, ordinances, precedents, cases, and bureaucratic notes, thus reflecting the rich variety of the legal and social life of the Yuan dynasty. Many items in the Yiian tien chang were written in the style of Chinese bureaucratic documents by governmental clerks and bureaucrats who were familiar with administrative and legal matters during the Yiian dynasty. Colloquial language was also prevalent in the text. In addition, many items were translated directly from documents issued in the Mongolian language. Owing to these special features the Yiian lien chang was often downgraded by traditional Confucian scholars.'8 The Yiian tien chang is a virtual encyclopedia of Yiian period social history and government. Because, however, so much of the text is in the peculiar style of Yiian period colloquial Chinese and further reflects, in many cases, the diction and grammar of Mongolian documents that underlie the Chinese texts, it has been difficult to 33 34 35 36 Wei Yiian, Yiian shih hsin pien (Kiangsu, 1905). Tseng Lien, Yiian shu (Shao-yang, 1911). Hung Chiin, Yiian shih i wen chengpit (Kiangsu, 1897). K'o Shao-min, Hsin Yiian shih (Tientsin, 1922; 2nd rev. ed. Peking, 1930); repr. in Erhshih wu shih, ed. Erh shih wu shih k'an rising wei yiian hui (Shanghai, 1935; repr. Taipei, 1962—9). 37 For a critical discussion of T'u Chi's worth in relation to other traditional histories of the Yiian period, see Yii Ta-chiin, "Lun T'u Chi te Meng-wu-erh shih chi," Yiian shih tun ts'ung, 3 (1986), pp. 219—30. 38 Paul Heng-chao Ch'en, Chinese legal tradition under the Mongols: The code of 1291 as reconstructed (Princeton, 1979), pp. 31—2. Cambridge Histories Online © Cambridge University Press, 2008 698 BIBLIOGRAPHICAL ESSAYS read and also is offensive to cultivated Chinese because of the crudeness of its language. Since its publication in 1908 by the noted legal scholar Shen Chia-pen, the Yuan tien chang has drawn the attention of scholars in China and Japan and has become an important special field of Yuan history studies, involving historical linguists, legal scholars, institutional historians, and many others. The Shen Chia-pen edition, based on faulty manuscripts circulated over several centuries, was standard for all scholars working on the text until 1972 when the Palace Museum (Taipei) published a facsimile of the 1320 edition that is far superior.'» Japanese scholarship on the text and related historical problems has been particularly noteworthy. *° The special problems of the style of colloquial Chinese used in official contexts in Yuan times extend beyond the Yuan tien chang to other kinds of materials. Research on the texts of stele inscriptions, the so-called pat hua pei, has become a special field of study in recent times, as has that of the language of Yuan drama. These are, of course, interrelated problems, both as linguistic and as social study (see Chapter 9). All three subjects (Yuan tien chang, pai hua pet, and Yuan drama) have drawn the attention of scholars only in the twentieth century and have quickly developed as scholarly undertakings, in which researchers from many countries have made significant contributions to a common fund of knowledge. In general, twentieth-century scholarship on the Yuan period has become an international effort, perhaps more so than that on any other period of pre-modern Chinese history. In no small measure, this is because the Mongolian empire of the thirteenth and fourteenth centuries directly impinged on the national histories of western Asian and eastern European states. But it also reflects the strategic significance of Mongolian studies for Japan, Russia, Great Britain, and other powers in the nineteenth and twentieth centuries. Sinological traditions bearing on Yuan history have remained strong in China, and at the same time new scholarly approaches have been introduced from outside. For an example of the latter, in the years between the two world wars a number of Chinese scholars went to Japan and Europe to study Mongolian and other Altaic and Western Asian languages; they returned to China to initiate such scudies in the Chinese universities and research institutions. They also were responsible for introducing and stimulating extensive translations of modern Japanese and Western scholarship. Because of continuing interactions between Chinese and Mongolian scholars, on the one hand, and centers of scholarship abroad, on 39 Yuan tien chang, published under the full title: Ta Yuan sheng cheng kuo ch'ao tien chang (Taipei: National Palace Museum, 1972), 60 chiian, in 16 ts'e. Appended to the final volume is an important bibliographic essay by Ch'ang Pi-te of the museum staff. 40 Representative examples are the products of the Yuan tien chang Research Group (Gen tensho ienkyuhan) of Kyoto University in the 1940s and 1950s, which produced Gin tensho sakuin ko, a topical index to the Yiian tien chang in 19)7 (repr. Taipei, 1973), and a special issue of Toho gakuho (Kyoto), 24(1934), devoted to issues in Yuan tien chang research. More recent is the chronological index to the work compiled by Uematsu Tadashi, Gen tensho nendai sakuin (Tokyo, 1980). A careful editing of chapters 39 through 37, those relevant to the Ministry of Justice — that is, the chapters defining offenses and their punishments - was published under the title (Kotei bon) Gen tensho keibu by Iwamura Shinobu and Tanaka Kenji, 2 vols. (Kyoto, 1964, 1972). For a fuller review of relevant Japanese scholarship, see Uematsu Tadashi, "Institutions of the Yuan Dynasty and Yuan Society," Gat Library Journal, 5 (Spring, 1992). Cambridge Histories Online © Cambridge University Press, 2008 BIBLIOGRAPHICAL ESSAYS 69 9 the other, one can almost speak today of "national schools" of Yiian dynasty and Mongolian studies. 4 ' That has had a continuing impact on Yiian history studies in China as well. The scope of recent publications in the field is far too vast to be reviewed here but is well indicated by the bibliographic citations in the various chapters of this volume. 4 . THE RISE OF THE MONGOLIAN EMPIRE AND MONGOLIAN RULE IN NORTH CHINA The rise of the Mongolian empire is well documented in a variety of contemporary and near-contemporary sources, of which the Mongolian, Chinese, and Persian are the most important. Though sometimes undervalued because of its mythic features, chronological vagueness, and the ongoing debate over its authorship, date of compilation, and textual history, the Secret history of the Mongols remains a key source for the reigns of Chinggis khan and Ogodei. Completed sometime in the middle decades of the thirteenth century, the Secret history (more properly the Chinggis Khaghan-u huja'ur, The origin of Chinggis khan) supplies a unique perspective on the emergence of their empire, revealing, as no other source can, Mongolian motives and goals. Of equal importance, it contains a vivid picture of the development of Mongolian institutions in their formative stage. We now have the translation by Francis W. Cleaves of The secret history of the Mongols,' and a second English rendering, by Igor de Rachewiltz, is in the offing. For the reigns of the first four emperors, the Yiian shih (Official history of the Yiian), compiled in 1369, is somewhat disappointing, owing mainly to the loss and destruction of the early Mongolian records. The monograph sections, for example, rarely give extensive coverage of the period before Khubilai. The basic annals are also brief, and many key figures, for example, Mahmud Yalavach, lack biographies. The inadequacies of the Yiian shih can in part be compensated for by the biographical data found in the extant literary collections (wen chi) of the period. The "spirit-way stelae" of Yeh-lii Ch'u-ts'ai, found in the Yiian wen lei' of Su T'ien-chiieh (1294—1352), includes important data edited out of his Yiian shih biography. Information on other important personages in the early empire (e.g., Chinkhai, Siibetei, etc.) can be located in Igor de Rachewiltz et al., Index to biographical material in Chin and Yiian literary works.* Ambassadorial reports and travel accounts are another vital source for this period. These include the Meng ta pei lu by the Sung envoy Chao Hung, who journeyed to 41 For a survey of one of these, see Elizabeth Endicott-West, "The Yuan," in Soviet studies of pre-modern China, ed. Gilbert Rozman (Ann Arbor, 1984), pp. 9 7 - 1 1 0 . Current surveys of Yiian studies in other countries are not presently available. 1 Francis W. Cleaves, The secret history of the Mongols: For the first time done into English out of the original tongue and provided with an exegetical commentary (Cambridge, Mass.; London, 1982). 2 Su T'ien-chiieh, Kuo cb'ao wen lei (SPTK ed.). 3 Igor de Rachewiltz et al., Index to biographical material in Chin and Yiian literary works, 3 vols. (Canberra, 1970—9). Cambridge Histories Online © Cambridge University Press, 2008 7OO BIBLIOGRAPHICAL ESSAYS north China in 1221, and the Hei ta shih liieh, a joint work by P'eng Ta-ya and Hsu Ting, who traveled to Ogodei's court on diplomatic missions in 1234 and 1235, respectively. The Hsiyu chi by Li Chih-chang, which records the journey of the Taoist master Ch'ang Ch'un to Central Asia from 1221 to 1223 for an interview with Chinggis khan, contains the only eyewitness description of the empire's founder. There is an English translation of this by Arthur Waley, The travels of an alchemist: The journey of the Taoist Ch'ang Ch'unfrom China to the Hindukush at the summons ofChinghiz khan.4 In a different category is the Sheng wu kh'in cheng lti> (Record of the personal campaigns of the holy warrior), a work of unknown authorship, which gives a chronological account of the reigns of Chinggis khan and Ogodei. Though often laconic, this text is of great interest, for it goes back to a Mongolian original no longer extant. The latter was translated into Chinese sometime before 1285 and was subsequently used by the compilers of the Yiian shih as one of their sources for the basic annals of the first two emperors. The Chinese texts of these four works are available in Wang Kuo-wei, ed., Meng ku shih liao ssu chung.6 Documentary materials from the period before Khubilai are, on the whole, rather limited. The major administrative manuals of the Yiian were compiled in the early fourteenth century, and the documents they contain date from the reigns of Khubilai and his successors. There are, however, references in these later decrees and memorials to the policies of Ogodei and Mongke, and in some instances the actual contents of earlier documents, usually imperial decrees, are quoted or summarized. The preparation of a full guide to such references in the Yiian tien chang of 1322 and the Tung chih t'iao ko of 1321 would greatly facilitate research into the history of the early empire. For the present, the index to their contents by date, compiled by Uematsu Tadashi, Gendensho nendai sakuin,1 is a very useful aid. The Persian sources for the reigns of the first four khans contain many data not found elsewhere. Of these the most comprehensive is the J'ami 'al-Tavarikh (Collected chronicles), a work completed during the reign of the Il-khan Oljeitii (1304-16), by Rashld al-DIn, a high functionary at the Mongolian court of Iran. Because of his official status and court sponsorship of the project, Rashld al-DIn was given access to now-lost Mongolian records and narratives to prepare his account of Chinggis khan and his immediate successors. One of his sources is clearly the same Mongolian chronicle that was translated into Chinese under the title Sheng wu ch'in cheng lu. The only complete, but by no means satisfactory, edition of the text is by B. Karlml.8 The sections on Ogodei, Giiyiig, and Mongke were translated by John A. Boyle, The successors of Ghengis khan."> Another of Rashld al-Dln's works, the Shu 'ab-i panj- ganai>,'° forms a genealogical supplement to his "Collected chronicles" and contains, 4 Arthur Waley, The travels of an alchemist: The journey of the Taoist Ch'ang Ch'un from China to the Hindukush at the summons ofCbinghiz Khan (London, 1931). 5 Sheng wu ch'in cbeng lu, in Meng-ku shih liao ssu chung, ed. Wang Kuo-wei (Taipei, 1970). 6 Wang Kuo-wei, ed., Meng-ku shih liao ssu chung (Taipei, 1970). 7 Uematsu Tadashi, Gendensho nendai sakuin (Kyoto, 1980). 8 Rashld al-Din, Jami 'al-Tavartkh, 2 vols., ed. B. Karlml (Teheran, 1959). 9 Rashld al-Drn, The successors of Ghengis Khan, trans. John A. Boyle (New York, 1971). 10 Rashld al-Din, Shu 'ab-ipanjgdnah, ms., Topkapi Sarayi Museum, cat. no. 2932. Cambridge Histories Online © Cambridge University Press, 2008 BIBLIOGRAPHICAL ESSAYS 7OI as well, many data on the titles and ethnic background of officials who served at the courts of Chinggis khan, Ogodei, and Mongke. Unfortunately, there is as yet no printed edition of this unique and important manuscript. In preparing his history of the reigns of Ogodei, Guyiig, and Mongke, Rashld alDln frequently used Juvayni's Tarktkh-i jahdn-gushd" (History of the world conqueror), finished in around 1260, of which there is an excellent edition by Muhammad Qazvlnl and a fine translation by John A. Boyle. Juvaynl, a mid-level official in the Mongolian administration in Iran, was intimately familiar with the politics and personalities of the 1240s and 1250s and, although generally pro-Mongolian and always pro-Toluid, his narrative is crucial to this period, especially in regard to matters of taxation and administration. Another Persian historian who chronicles the rise of the Mongolian empire from its inception to 1259 is Juzjani, whose work, the Tabaqdt-i-nasjrP* (The tables of Nasjri), is a useful corrective to Juvayni's rather proMongolian attitude. Juzjani, as an outsider hostile to the invaders, presents their conquest of the Islamic lands in a much different light. There is a serviceable but rather eccentric translation of Juzjani by H. G. Raverty, "Xabaqat-i-nasjri.^ In looking at the period of the early empire, these three Persian historians should always be consulted jointly and their data carefully compared. European accounts of the early Mongols are not numerous, but the travelogues of Carpini, who journeyed to Mongolia in 1245—7, and Rubruck, who made the trip in 1253-5, a r e fascinating descriptions of the empire at its height. Rubruck, in particular, was an attentive and critical observer of the Mongolian scene, and his account provides much information on the political relations among the Chinggisid princes, Mongolian policy in conquered lands, and an unsurpassed picture of Khara Khorum, the imperial camp. These two works can be found in Christopher Dawson, ed., The Mongol mission.1* The state of the field of early Mongolian history is revealed in the fact that the first scholarly biography of the empire's founder did not appear until 1970, when Li Tsefen published his fine study, Ch'eng-chi-ssu han hsin-ch'uanl"> (A new biography of Chinggis khan). More recently, Paul Ratchnevsky produced an equally solid biography: Chinggis khan, sein Leben und Wirken.l6 Although popular in format, the succinct account of Louis Hambis, Gengis-khan,'7 remains an excellent introduction to the subject. On the reign of Ogodei, see N. Ts. Munkuev, Kitaiskii istochnik 0 pervykh mongol'skikh khanakh'8 (A Chinese source on the first Mongolian khans), which concentrates on the reforms of Yeh-lii Ch'u-ts'ai. This topic is also explored by Igor de Rachewiltz, "Yeh-lii Ch'u-ts'ai (1189-1243): Buddhist idealist and Confucian 11 'Ata Malik Juvaynl, Tarkikh-i jahan-gusha, 3 vols., ed. Mirza Muhammad Qazvlnl (London, 1 9 1 2 37); and 'Ata Malik Juvaynl, The history of the world conqueror, 2 vols. trans. John A. Boyle (Cambridge, Mass., 1958). 12 Minhaj al-DIn Juzjani, Jabaqat-i-nafiri, ed. W. Nassau Lees (Calcutta, 1864). 13 Minhaj al-DIn Juzjani, fabaqat-i-naslri, trans. H. G. Raverty, 2 vols. (New Delhi, 1970). 14 Christopher Dawson, ed., The Mongol mission (New York, 1955). 15 Li Tse-fen, Ch'eng-chi-ssu han hsin-ch'uan (Taipei, 1970). 16 Paul Ratchnevsky, Chinggis khan, sein leben und wirken (Wiesbaden, 1983). 17 Louis Hambis, Gengis-khan (Paris, 1973). 18 N. Ts. Munkuev, Kitaiskii istochnik 0 pervykh mongol'skikh khanakh (Moscow, 1965). Cambridge Histories Online © Cambridge University Press, 2008 702 BIBLIOGRAPHICAL ESSAYS statesman."1' Paul Buell's dissertation, "Tribe, Qan and Ulus in early Mongol China: Some prolegomena to early Yuan history,"20 is an important contribution to the study of Ogodei's reign and to Mongolian institutions in general. For Mongke's tenure as khaghan, see Thomas T. Allsen, Mongol imperialism: The policies of the grand Qan Mongke in China, Russia and the Islamic lands 1251—59." The best study of the Mongolian campaigns in Central Asia is still to be found in W. Barthold's now-classic Turkestan down to the Mongol invasion.'1'' For the assault on the Chin dynasty, a subject deserving of a fresh study, see H. Desmond Martin's rather outdated The rise of Chingis khan and his conquest of north China.2* The political side of the conquest is discussed by Igor de Rachewiltz in his important essay "Personnel and personalities in north China in the early Mongolian period."2'' See also Yuan personalities,2* which includes biographies of numerous officials, Chinese and foreign, who served the early Mongolian rulers in north China. On the fate of the Confucian literati in north China, see Makino Shuji, "Transformation of the shih-jen in the late Chin and early Yiian."26 The fiscal policies of the early khans are discussed in detail by H. Franz Schurmann in "Mongol tributary practices of the thirteenth century,"27 and John Masson Smith, "Mongol and nomadic taxation."28 On their governmental style, see Paul Buell, "Sino-Khitan administration in Mongol Bukhara,"29 and Thomas T. Allsen, "Guard and government in the reign of the grand Qan Mongke, 1251-59."3° One of the most difficult problems facing the student of this period is the plethora of Mongolian and Turkic technical terms found in the sources. In searching for explanations, one should first consult Gerhard Doerfer, Turkische und Mongolische elemente im Neupersischen.*1 The field as a whole is greatly indebted to Paul Pelliot and Francis W. Cleaves, who carriedout much of the philological spadework on which later historical studies are based. Moreover,. their technical studies often contain valuable historical discussions. Pelliot's entry on Chinggis khan in his Notes on Marco 19 Igor de Rachewiltz, "Yeh-lii Ch'u-cs'ai (i 189—1243): Buddhist idealist and Confucian statesman," in Arthur F. Wright and Denis C. Twitchett, eds., Confucian personalities (Stanford, 1962), pp. 189-216. 20 Paul Buell, "Tribe, Qan and Ulus in early Mongol China: Some prolegomena to early Yiian history" (Ph.D. diss., University of Washington, 1977). 21 Thomas T. Allsen, Mongol imperialism: The policies of the grand Qan Mongke in China, Russia and the Islamic lands 1251-59 (Berkeley and Los Angeles, 1987). 22 W. Barthold, Turkestan down to the Mongol invasion, 3rd ed. (London, 1968). 23 H. Desmond Martin, The rise of Chingis khan and his conquest of north China (Baltimore, 1950). 24 Igor de Rachewiltz, "Personnel and personalities in north China in the early Mongolian period," Journalof the Economic and Social History of the Orient, 9(1966), pp. 8 8 - 1 4 4 . 2 ; Igor de Rachewiltz and Hok-lam Chan, eds., Yuan personalities, vol. 1 (Wiesbaden, forthcoming). 26 Makino Shuji, "Transformation of the shih-jen in the late Chin and early Yiian," Ada Asiatica, 4 ; (1983), pp. 1-26. 27 H. Franz Schurmann, "Mongol tributary practices of the thirteenth century," Harvard Journal of Asiatic Studies, 19 (1956), pp. 304—89. 28 John Masson Smith, "Mongol and nomadic taxation," Harvard Journal of Asiatic Studies, 30 (1970), pp. 4 8 - 8 5 . 29 Paul Buell, "Sino-Khitan administration in Mongol Bukhara," Journal of Asian History, 13 (1979), pp. 121-51. 30 Thomas T. Allsen, "Guard and government in the reign of the grand Qan Mongke, 1251-59," Harvard Journal of Asiatic Studies, 46 (December 1986), pp. 495—521. 31 Gerhard Doerfer, Turkische und Mongolische elemente im Neupersischen, 4 vols. (Wiesbaden, 1963-75). Cambridge Histories Online © Cambridge University Press, 2008 BIBLIOGRAPHICAL ESSAYS 703 Polo*2 addresses and solves some vexing chronological problems, and Cleaves's "The Sino-Mongolian inscription of 1240"" provides, unexpectedly, important data on the Mongols' postal relay system. In the absence of a Doerfer-like reference work on the Mongolian and Turkic words in the Chinese sources, a guide or index to the numerous terms discussed in the monographs and articles of Pelliot and Cleaves would be a most welcome and useful contribution to the field. For bibliographical works on the Mongolian empire, see Denis Sinor, Introduction a I'etude de I'Eurasie centrale* and Thomas T. Allsen, The Mongols in East Asia, twelfthfourteenth centuries: A preliminary bibliography of books and articles in Western languages." More extensive and more up to date is Henry G. Schwarz's Bibliotheca Mongolica,i6 which covers works in English, French, and German. 5 . THE REIGN OF KHUBILAI KHAN The Chinese primary sources for the reign of Khubilai khan are fragmentary and diverse. The pen chi (basic annals) of Khubilai's reign in the Yuan shih Ate a useful chronological record and should be read along with the biographies of notable figures found in the same history. The later Hsin Yuan shih by K'o Shao-min, the Yuan shih lei pien by Shao Yuan-p'ing, and the Meng-wu-erh shih chi by T'u Chi supplement the biographies of the Yuan shih and provide valuable details not found in the Yuan dynastic history. Essays in Ch'en Pang-chan's Sung shih chi shih pen mo and Yuan shih chi shih pen mo offer useful thematic treatments, not a simple chronology, of important events, particularly of military campaigns. Fiscal and government regulations and problems during Khubilai's reign are covered extensively in the Yuan tien chang, the principal administrative handbook of the Yuan period published in the 1330s. The histories of the two dynasties conquered by the Mongols, the Chin shih and the Sung shih, provide not only an antidote to the Mongolian viewpoint but also a glimpse of the Chin and Sung responses to the Mongols. Charles A. Peterson used these two sources, among others, in two important articles on Sung reactions to the Mongol invasion of northern China.' Other primary Chinese sources focus on specific aspects of Khubilai's career and reign. The Pien wei lu2 and Nien-ch'ang's Fo tsu li tai t'ung tsafi give the principal accounts of the Buddhist—Taoist debates of 1258 and 1281. The Nan chao yeh shih 32 Paul Pelliot, Notes on Marco Polo (Paris, 1959), vol. 1, pp. 2 8 1 - 3 6 3 . 33 Francis Woodman Cleaves, "The Sino-Mongolian inscription of 1240," Harvard Journal of Asiatic Studies, 23 (1960—1), pp. 62—75. 34 Denis Sinor, Introduction a I'itude de I'Eurasie centrale (Wiesbaden, 1963), pp. 294—319. 35 Thomas T. Allsen, The Mongols in East Asia, twelfth—fourteenth centuries: A preliminary bibliography of books and articles in Western languages (Sung studies research aids, I; Philadelphia: Sung Studies Newsletter, 1976). 36 Henry G. Schwarz, Bibliotheca Mo»go/«» (Bellingham, 1978). 1 CharlesA. Peterson, "Old illusions and new realities: Sung foreign policy, 1217—1234," inChinaamong equals: The Middle kingdom and its neighbors, toth-i^th centuries, ed. Morris Rossabi (Berkeley and Los Angeles, 1983), pp. 204—39; and "First Sung reactions to the Mongol invasion of the north 1211—17," in Crisis and prosperity in Sung China, ed. John Winthrop Haeger (Tucson, 1975), pp. 215—52. 2 Hsiang-mai, Pienweilu, TaishoTripitaka, vol. 5 2 , pp. 751—81. 3 Nien-ch'ang, Fou tsu li tai t'ung tsai, Taisho Tripitaka, vol. 49, pp. 477—735. Cambridge Histories Online © Cambridge University Press, 2008 704 BIBLIOGRAPHICAL ESSAYS (translated by Camille Sainson)* describes Khubilai's campaign in Yunnan before he assumed leadership of the Mongolian world. Specialized texts deal with the Yiian dynasty's important relations with Korea (Yiian Kao-li chi shih),* its grain tax policies (Ta Yiian ts'ang k'u chi),6 its horse administration (Ta Yiian ma cheng chi),1 and its innovations in sea transport (Ta Yiian hai yiin chi).9 The writings of such prominent Yiian officials and cultural figures as Chao Meng-fu yield insights into and anecdotes about the personalities, both Chinese and Mongolian, who played a vital role during Khubilai's reign. Because the Yiian was part of a larger Mongolian world, the writings of foreign historians and travelers are invaluable sources. The contemporary observations of the Persian historians JuvainI, Rashld al-Dln, and JuzjanI (the first two translated into English by John A. Boyle and the third by H. G. Raverty)9 offer perceptive and unique accounts of Khubilai's court, and the Korean chronicle, Chong In-ji's Korydsa completed in 1451, provides an indispensable record of the Yiian's relations with a land that had frequent and extended contacts with China. The works of Christian envoys to the Mongolian domains are conveniently translated in Christopher Dawson's The Mongol mission'0 and by William Rockhill in The journey of William of Rubruck to the eastern parts of the world.11 Marco Polo, the most renowned European traveler to China in this era, offers an unsurpassed description of Khubilai's reign at its height. The translation of Marco's work by A. C. Moule and Paul Pelliot, Marco Polo: The description of the world," supplemented by their extensive (separately published) notes1' and by Leonardo Olschki's Marco Polo's Asia, •« supply a remarkable portrait of Khubilai, his court, and the China of his time. The Mongolian chronicle, the Chaghan teiike,'* offers what purports to be a combination of shamanist and Buddhist views of Khubilai and his ancestors. Later Mongolian chronicles such as the Allan Tobchi,16 are so imbued with Buddhism that they cannot be accepted as unbiased, accurate accounts. The Tibetan account, the Blue annals, •' the Syriac record of Bar Hebraeus,18 and the Armenian description of King Het'um's travels'9 are useful, though not as significant as the other more detailed foreign chronicles. 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 Camille Sainson, Nan-tcbao ye-che, bistoinparticuliirt de Nan-ubao (Paris, 1904). Anon., Yuan Kao-li chi shih, in Shih liao u'ungpicn, ssu-pien (Taipei, 1972). Anon., Ta Yiian Wang k'u chi, in Shih liao u'ung pirn, ssu-pien (Taipei, 1972). Anon., Ta Yuan ma cheng chi, in Shih liao ts'ungpien, ssu-pien (Taipei, 1972). Anon., Ta Yiian hai yiin chi, in Shih liao ts'ungpien, ssu-pien, ed. Hu Ching (Taipei, 1972). See "Bibliographic Note for Chapter 4," nn. 8—13. Christopher Dawson, ed., The Mongol mission: Narratives and letters of the Franciscan missionaries in Mongolia and China in the thirteenth and fourteenth centuries (New York, 1935). William Rockhill, The journey of William of Rubruck to the eastern parts of the world (London, 1900). A. C. Moule and Paul Pelliot, Marco Polo: The description of the world (London, 1938). A. C. Moule, Quinsai, with other notes on Marco Polo (Cambridge, 1957); and Paul Pelliot, Notes on Marco Polo, 3 vols. (Paris, 1958-73). Leonardo Olschki, Marco Polo's Asia (Berkeley, i960). Klaus Sagaster, trans., Die tutisse Geschichte(Wiesbaden, 1976). Charles R. Bawden, trans., The Mongol chronicle Allan Tobci (Wiesbaden, 1955). George Roerich, trans., The blue annals, 2nd ed. (Delhi, 1976). Wallis Budge, trans., The chronography of Gregory Abu 'I Faraj the son of Aaron, the Hebrew physician commonly known as Bar Hebraeus, 2 vols. (London, 1932). The description was explicated by John Andrew Boyle in "The journey of Het'um, king of Little Armenia, to the court of the great khan Mb'ngke," Central Asiatic Journal, 9 (1964), pp. 175-89. Cambridge Histories Online © Cambridge University Press, 2008 BIBLIOGRAPHICAL ESSAYS 705 Modem secondary scholarship The only Western-language biography of Khubilai is Morris Rossabi's Khubilai khan: His life and times,2° which also contains a comprehensive bibliography of Asian- and Western-language sources. Other works by the same author that deal with Khubilai and his family include "Khubilai Khan and the women in his family"21 and the article "Chinese myths about the national minorities: Khubilai khan, a case study."22 Other biographies in Chinese and Japanese, such as those by Chou Liang-hsiao,23 Katsufuji Takeshi,2'' Li T'ang,2' and Otagi Matsuo26 are based exclusively on East Asian materials and scarcely consult Middle Eastern sources or secondary Westernlanguage scholarship. Khubilai's early life and his advisers and career are covered by Rossabi and in the articles by Hsiao Ch'i-ch'ing,27 Yao Ts'ung-wu,28 Hsia Kuang-nan,2' and particularly in Hok-lam Chan's articles on Khubilai's advisers Liu Ping-chung3° and Yao Shu.3' There are several detailed studies of Khubilai's capital cities, the most detailed being Nancy S. Steinhardt's dissertation on Ta-tu.32 Harada Yoshito,33 Komai Kazuchikai,34 Ishida Mikinosuke,35 and, most recently, Nancy S. Steinhardt36 have written extensively on Khubilai's summer capital at Shang-tu, and Kiselev has reported on the Soviet excavations at the ancient Mongolian capital of Khara Khorum.37 Secondary sources on Khubilai's early policies include David M. Farquhar's study of the structure of Yuan government38 and Herbert Franke's From tribal chieftain to 20 Morris Rossabi, Khubilai Khan: His life and limes (Berkeley and Los Angeles, 1988). 21 Morris Rossabi, "Khubilai khan and the women in his family," in Studia Sino-Mmgolica: Festschrift fur Herbert Franke, ed. Wolfgang Bauer (Wiesbaden, 1979), pp. 153—80. 22 Morris Rossabi, "Chinese myths about the national minorities: Khubilai khan, a case study," Central and Inner Asian Studies, 1 (1987), pp. 47—81. 23 Chou Liang-hsiao, Hu-pi-lieh (Ch'ang-ch'un, 1986). 24 Katsufuji Takeshi, Fubirai kan (Tokyo, 1966). 25 Li T'ang, Yiian Shih-tsu (Taipei, 1978). 26 Otagi Matsuo, Fubirai kan (Tokyo, 1941). 27 Hsiao Ch'i-ch'ing, "Hu-pi-lieh shih tai ch'ien ti chiu lii k'ao," Ta-lu tsa cbih, 25, nos. 1 (1962), pp. 16—23, ar>d 3 (1962), pp. 22—7. 28 Yao Ts'ung-wu, "Hu-pi-lieh han tui yii Han-hua t'ai tu ti fen hsi," Ta-lu tsa-chih, 11, no. 1 (1955), pp. 2 2 - 3 2 . 29 Hsia Kuang-nan, Yiian tai Yuri-nan shih ti ts'ung k'ao mu lu (Shanghai, 193;). 30 Hok-lam Chan, "Liu Ping-chung (1216—1274): A Buddhist-Taoist statesman at the court of Khubilai Khan," T'oungPao, 53 (1967), pp. 98—146. 31 Hok-lam Chan, "Yao Shu (1201-1278)," Papers on Far Eastern History, 22 (1980), pp. 17—50. 32 See Nancy R. S. Steinhardt, "Imperial architecture under Mongolian patronage: Khubilai's imperial city of Daidu" (Ph.D. diss., Harvard University, 1981), as well as her article "The plan of Khubilai Khan's imperial city," Artibus Asiae, 44 (1983), pp. 137—58; and her Chinese imperial city planning (Honolulu, 1990), pp. 1 5 4 - 6 0 . 3 3 Harada Yoshito, Shang-tu: The summer capital of the Yiian dynasty (Tokyo, 1941). 34 Komai Kazuchikai, "Gen nojotonarabi ni Daitonoheimenni tsuite," TiijrOTWi), 3(1940), pp. 129—39. 35 Ishida Mikinosuke, "Gen no Joto ni tsuite," Nihon daigaku zlritsu shichijutsu kinen rombun shu, 1 (i960), pp. 271-319. 36 Steinhardt, Chinese imperial city planning, pp. 150—4. 37 S. V. Kiselev, ed., Drevnemmgol'skiegoroda (Moscow, 1965). 38 David M. Farquhar, "Structure and function in the Yiian imperial government," in China under Mongol rule, ed. John D. Langlois, Jr. (Princeton, 1981), pp. 25—55. Cambridge Histories Online © Cambridge University Press, 2008 706 BIBLIOGRAPHICAL ESSAYS universal emperor and god: The legitimation of the Yuan dynasty,i9 which offers a fascinat- ing glimpse into Khubilai's political acumen. Franz Schurmann's works on the Yuan financial administration,40 give rich detail about Khubilai's economic policies. Other social and economic problems are the subject of articles by Iwao Aritaka on peasants,4' by Lo Jung-pang on the extension of the Grand Canal,42 by Chii Ch'ingyiian4* and Weng Tu-chien44 on craftsmen, and by Murakami Masatsugu,45 Otagi Matsuo,46 and, most recently, Thomas Allsen47 and Elizabeth Endicott-Wesf8 on the ortogh. The postal relay system is the subject of a book by Peter Olbricht.4^ Hsiao Ch'i-ch'ing'° and Gunther Mangold'1 are the most recent students of the Mongolian and Yuan military. The legal innovations of the Yuan are dealt with by Paul Ch'en'2 and Paul Ratchnevsky.53 Intellectual developments during Khubilai's reign are beginning to receive attention. The articles in the volume edited by Hok-lam Chan and William Theodore deBary discuss the Yuan and Neo-Confucianism.'4 The position of Buddhism is explored in articles on the discord between Buddhists and Taoists under the Yuan by Nogami Shunjo" and by Paul Ratchnevsky.'6 The conflicts between the Buddhists and the Taoists are also described by Josef Thiel." The contributions of the Tibetan 39 Herbert Franke, From tribal chief lain to universal emperor and god: The legitimation of the Yuan dynasty (Munich, 1978). 40 See Herbert F. Schurmann, Economic structure of the Yiian dynasty: Translation of chapters 93 and 94 of the Yuan shih, Harvard—Yenching Institute Studies, vol. 16 (Cambridge, Mass., 1956), and his "Mongolian tributary practices of the 13th century," Harvard Journal ofAsiatic Studies, 19(1936),pp. 304-89. 41 Iwao Aritaka, "Gendai no nomin seikatsu ni tsuite," Kuwabara Hakushi kanreki kinen toyoshi ronso (Kyoto, 1935), pp. 9 5 1 - 7 42 Lo Jung-pang, "The controversy over grain conveyance during the reign of Qubilai Qagan, 1260—94," Far Eastern Quarterly, 13 (May 1952), pp. 262—6. 43 Chii Ch'ing-yiian, "Yiian tai hsi kuan chiang hu yen chiu," Shih huo, 1 (1935), pp. 367-407; summary translation in E-tu Zen Sun and John de Francis, Chinese social history: Translations of selected studies (Washington, D . C . , 1956), pp. 2 3 4 - 6 . 44 Weng Tu-chien, "Wo-t'o tsa k'ao," Yen-ching hsiiehpao, 29 (1941), pp. 201-18. 43 Murakami Masatsugu, "Gencho ni okeru senrushi to attatsu," Toho gakuho, 13 (1942), pp. 143—96. 46 Otagi Matsuo, "Attatsu sen to sono haikei," Toyoshi tenkyu 32, no. 1 (1973), pp. 1-27; no. 2 (1973), pp. 2 3 - 6 1 . 47 Thomas T. Allsen, "Mongolian princes and their merchant partners," Asia Major, 3rd series, vol. 2, pt. 2 (1989), pp. 8 3 - 1 2 6 . 48 Elizabeth Endicott-West, "Merchant associations in Yiian China: The ortogh," Asia Major, 2 (1989), pp. 1 2 7 - 5 4 . 49 Peter Olbricht, Das Postwesen in China unter der Mongolenhemchaft im 13 und 14 Jahrhundert (Wiesbaden, 1954). 50 Hsiao Ch'i-ching, The military establishment of the Yiian dynasty (Cambridge, Mass., 1978). 51 Gunther Mangold, Das Milita'rwesm in China unter der Mongolenherrschafi (Bamberg, 1971)52 Paul Ch'en, Chinese legal tradition under the Mongols (Princeton, 1979). 53 Paul Ratchnevsky, Un code des Yiian, 4 vols. (Paris, 1937-85). 54 Hok-lam Chan and William Theodore deBary, eds., Yiian thought: Chinese thought and religion under the Mongols (New York, 1982). 55 See Nogami Shunjo, "Gendai dobutsu nikyo no kakushitsu," Otani daigaku kenkyu nempo 2 (1943), pp. 213—65, and also his "Gen no senseiin in tsuite," Haneda hakushi shoju kinen Toyoshi ronso (i960), pp. 779-9556 Paul Ratchnevsky, "Die mongolische Grosskhane und die buddhistische Kirche," Asiatica: Festschrift Friedrich Weller zum 65. Geburtstag (Leipzig, 1954). 57 Josef Thiel, "Der Streit der Buddhisten und Taoisten zur Mongolenzeit," Monumenta Sirica, 20 (1962), pp. 1 - 8 1 . Cambridge Histories Online © Cambridge University Press, 2008 BIBLIOGRAPHICAL ESSAYS 707 Buddhist monk "Phags-pa have not yet been accorded a full-scale study. Luciano Petech's "Tibetan relations with Sung China and with the Mongols"' 8 and Herbert Franke's "Tibetans in Yuan China"" and, to a lesser extent, Miyoko Nakano's study of his new script 60 are good starting points, but more research is needed on 'Phagspa's influence. The role of the Muslims has been studied by Morris Rossabi.61 Sun K'o-k'uan, in his article "Yii Chi and southern Taoism during the Yuan period," 62 and in numerous other books and articles in Chinese, has written about the new Taoist orders and the Taoist influence on the Yuan. The Christian relationship with the Mongols is described in A. C. Moule's Christians in China before the year 15^0^ and in the translations by E. A. Wallis Budge6* and by James A. Montgomery 6 ' of accounts by Nestorian monks who traveled to the Middle East and Europe. The Yuan court's cultural patronage has received some attention from scholars and has served as an antidote to the image held of the Mongols' savagery and barbarism. The new alphabet devised by 'Phags-pa was analyzed by Nicholas Poppe66 and by Nakano Miyoko. 67 Mongolian patronage of the theater is mentioned in a book by James I. Crump 68 and in Stephen H. West's article on Mongol influence on the Chinese drama. 6 ' Sherman Lee and Ho Wai-kam, Chinese art under the Mongols: The Yuan dynasty (1279-2368); 7 0 James Cahill, Hills beyond a river: Chinese painting of the Yuan dynasty 12J9—1368;7' and Margaret Medley, Yuan porcelain and stoneware,''2 all explore the Yuan influence on art, including painting, ceramics, sculpture, and textiles. Not all the foreign military campaigns initiated by Khubilai have been adequately studied. William Henthorn, in his Korea: The Mongol invasions,1* offers insights into the expansion in Korea. Sir George Sansom's A history ofJapan to 1334™ and Hori 58 Luciano Petech, "Tibetan relations with Sung China and with the Mongols," in China among equals: The Middle Kingdom and its neighbors, 10th—14th centuries, ed. Morris Rossabi (Berkeley and Los Angeles, 1983), pp. 1 7 3 - 2 0 3 . 59 Herbert Franke, "Tibetans in Yuan China," in China under Mongol rule, ed. John D. Langlois, Jr., (Princeton, 1981), pp. 2 9 6 - 3 2 8 . 60 Miyoko Nakano, A phonological study in the 'Phags-pa script and the Meng-ku tzu-yun (Canberra, 1971). 61 Morris Rossabi, "The Muslims in the early Yuan dynasty," China under Mongol rule, ed. John D. Langlois, Jr. (Princeton, 1981), pp. 2 5 7 - 9 5 . 62 Sun K'o-k'uan, "Yu Chi and southern Taoism during the Yuan period," in China under Mongol rule, ed. John D. Langlois, Jr. (Princeton, 1981), pp. 212—53. 63 A. C. Moule, Christians in China before the year 1550 (London, 1930). 64 E. A. Wallis Budge, trans., The monks of Kublai khan, emperor of China (London, 1928). 65 James A. Montgomery, trans., The history of Yaballaha III (New York, 1927). 66 Nicholas Poppe, trans., The Mongolian monuments in 'Phags-pa script, ed. John R. Krueger (Wiesbaden, '957>67 Nakano, A phonological study. 68 James I. Crump, Chinese theater in the days of Kublai khan (Tucson, 1980). 69 Stephen H. West, "Mongol influence on the development of northern drama," in China under Mongol rule, ed. John D. Langlois, Jr. (Princeton, : 9 8 i ) , pp. 434—65. 70 Sherman Lee and Ho Wai-kam, Chinese art under the Mongols: The Yuan dynasty (1279-1368) (Cleveland, 1968). 71 James Cahill, Hills beyond a river: Chinese painting of the Yuan dynasty 1279—1368 (New York, 1976). 72 Margaret Medley, Yuan porcelain and stoneware (New York, 1974). 73 William Henthorn, Korea: The Mongol invasions (Leiden, 1963). 74 George Sansom, A history of Japan to 1334 (Stanford, 1958). Cambridge Histories Online © Cambridge University Press, 2008 708 BIBLIOGRAPHICAL ESSAYS Kyotsu's dissertation, "The Mongol invasions and the Kamakura bakufu,"7' are helpful guides regarding the expedition against Japan.'6 In his "From Mongol empire to Yuan dynasty: Changing forms of imperial rule in Mongolia and central Asia,"77 John W. Dardess provides a useful survey of the campaigns against Khaidu. The expeditions in Southeast Asia, particularly those against Java, badly need additional modern research. The decline evident in the last years of Khubilai's reign also requires further study. Herbert Franke's forty-year-old study of Ahmad78 needs to be updated. His study of Sangha79 has been supplemented by a recent study by Luciano Petech.80 But more research is essential to a clearer view of the roles of these two officials as well as that of Lu Shih-jung, the last of the so-called villainous ministers of the Yuan shih. The activities of Yang Lien-chen-chia should also be reexamined, for Paul Demieville's essay on the subject, "Les tombeaux des Song meiidionaux,"81 appeared over sixty years ago. Herbert Franke's article in China under Mongol rule is the beginning of a reevaluation of Yang's role in the pillaging of the Sung tombs.82 6. MID-YOAN POLITICS The political history of the mid-Yuan period is far less well studied than that of other periods of the Mongolian Yuan dynasty. This neglect by historians can be partly attributed to the period's seeming lack of historical significance as a confused period sandwiched between the glorious and historic reign of Khubilai khaghan, the dynastic founder, and that of Toghon Temiir, its "bad last ruler." To a lesser degree, the neglect can also be attributed to the relative shortage of and lack of variety in its source materials. Unlike the periods of the early Mongolian khaghans and Khubilai, there is little 75 Hori Kyotsu, "The Mongol invasions and the Kamakura bakufu" (Ph.D. diss., Columbia University, 1967). 76 There is a huge Japanese secondary literature on the Mongol invasions. The classic study is that by Ikeuchi Hiroshi, Genko no shinkenkyu (Tokyo, 1931). More recent works include those by Yamaguchi Osamu, Mokoshurai (Tokyo, 1964; repr. 1979); HatadaTakashi, Genko (Tokyo, 1965); Abe Yukihiro, MSkoshiirai (Tokyo, 1980); AidaNiro, Moko shurai no kenkyu (Tokyo, 1971); and Kawazoe Shoji, Moko thurai kenkyu shiron (Tokyo, 1977), which presents a historical analysis of new studies. See also the chapter by Ishii Susumu in Medieval Japan, ed. Kozo Yamamura, vol. 3 of The Cambridge history of Japan (Cambridge, 1990), pp. 131—48. 77 John W. Dardess, "From Mongol empire to Yuan dynasty: Changing forms of imperial rule in Mongolia and central Asia," Monumenla Serica, 30 (1972-3), pp. 117—65. 78 Herbert Franke, "Ahmed: Ein Beitrag zur Wirtschafesgeschichte Chinas unter Qubilai," Oriens, 1 (1948), pp. 2 2 2 - 3 6 . 79 Herbert Franke, "Sen-ge: Das Leben eines uigurischen Staatsbeamten zur Zeit Chubilai's dargestellt nach Kapitel 205 der Yiian-Annalen," Sinica, 17 (1942), pp. 9 0 - 1 1 3 . 80 Luciano Petech, "Sang-ko, a Tibetan statesman in Yuan China," Acta Orientalia Academiae Scientiarum Hungaricae, 34 (1980), pp. 193-208. 81 Paul Demieville, "Les tombeaux des Song meridionaux," Bulletin de 1't.cole Franfaise d"Exlreme-Orimt, 25 (1925), pp. 458—567; repr. in his Choix d'itudcs sinologiques (1921—1970), ed. Yves Hervouet (Leiden, 1973), pp. 17—26. 82 Franke, "Tibetans in Yuan China," pp. 2 9 6 - 3 2 8 , esp. pp. 3 2 1 - 5 . Cambridge Histories Online © Cambridge University Press, 2008 BIBLIOGRAPHICAL ESSAYS 709 surviving historical material concerning the period in the Mongolian and Western languages. Of the Persian sources, Rashld al-DIn's Jami' al-Tavarikh, "Collection of chronicles," which is invaluable to the study of the history of the early Mongolian empire as a whole, covers the period only up to the reign of Temiir khaghan. This account of Temiir's reign is now available in J. A. Boyle's English translation1 as well as in Russian and Chinese translations. For the post-Temiir period, al-Qashanl's Tartkh-f Uljaytu, "History of Oljeitu,"2 and Wassaf's Tankh-i Wassaf, " Wassaf's history,"' are available only in the Persian originals and are useful mainly for the study of Yiian's relation with the western khanates. The study of the political history of the mid-Yiian, therefore, depends mainly on Chinese sources. Of the Chinese sources, the pen chi (basic annals) of various reigns and the biographies of the important figures of the period in the Yiian shih are the basic and indispensable sources for reconstructing the political history of the period. As an official history hastily put together at the beginning of the succeeding Ming dynasty, the Yiian shih is well known for its shortcomings.4 Most of historians' criticisms of the Yiian shih as a whole can be applied to the parts dealing with this period. First, there are no biographies of the principal ministers of Khaishan's and Yesiin Temiir's reigns. The accounts of the Yiian shih, therefore, must be supplemented by and corroborated with the relevant materials in the uien-chi (collected literary works) of the authors of the mid- and late Yiian, many of whom were actively involved in the government, as well as in the various later compilations of Yiian history by Shao Yiian-p'ing,' T'u Chi,6 and K'o Shao-min.7 The most useful wen-chi for this period include those by Chao Meng-fu, Ch'eng Chii-fu, Liu Min-chung, Chang Yang-hao, Huang Chin, Yii Chi, Chieh Hsi-ssu, Ma Tsu-ch'ang, Su T'ien-chiieh, Ou-yang Hsiian, Hsu Yu-jen, and Wei Su. The anthology Kuo ch'ao wen lei,8 compiled by Su T'ien-chiieh, also contains much material not available elsewhere; its coverage, however, stops at the 1310s. Another compilation by Su T'ien-chiieh, Yiian ch'ao ming ch'en shih liieh,* the biographies of important officials from the beginning of the Mongolian empire to Temiir's reign, is useful only for the early part of our period. Pichi (collections of miscellaneous notes) by mid- and late-Yiian authors including Ch'ang-ku Chen-i's Nung t'ien yii hua,'° Yang Yii's Shan chii hsin hua,11 K'ung Ch'i's 1 Rashld al-DIn, The successors of Genghis khan, trans. John A. Boyle (New York, 1971). 2 Al-QashanI, Tartkh-t Uljaytu, ed. Mahin Hambly (Tehran, 1969). 3 Wassaf, Tarikh-l Wassaf, lithograph ed. (Bombay, 1852-3), reedited by Mohammed Mahdi Isfahan! (Tehran, 1959-60). 4 Sung Lien et al. eds., Yiian shih (Peking, 1976) (hereafter cited as YS). 5 Shao Yiian-p'ing, Yiian shih leipien, Sao-yeh shan-fang ed. 6 T'u Chi, Meng tvu erh shih chi, ed. Chieh-I-i (Peking, 1934). 7 K'o Shao-min, Hsin Yiian shih (Tientsin, 1922; author's 2nd rev. ed., Peking, 1930); repr. in Erh shih wu shih, ed. Erh shih wu shih k'an hsing wu yuan hui (Shanghai, 1935; repr. Taipei, 1962-9, and other recent facsimile reprints) (hereafter cited at HYS). 8 Su Tien-chiieh, comp., Kuo ch'ao wen lei (SPTK ed.). 9 Su T'ien-chiieh, comp., Yuan ch'ao ming ch'en shih liieh, (1335 ed.; repr. Peking, 1962). 10 Ch'ang-ku Chen-i, Nung t'ien yii hua (Pao-yen t'ang pi chi ed.). 11 Yang Yu, Shan chii hsin hua (Chih pu tsu chai ts'ung shu ed.). It was translated into German by Herbert Franke, Beitrage zur Kulturgeschichte Chinas unter der fAongolenherrschaft (Wiesbaden, 1956). Cambridge Histories Online © Cambridge University Press, 2008 7IO BIBLIOGRAPHICAL ESSAYS (K'ung K'e-ch'i) Chih cheng chih chi," and T'ao Tsung-i's Nan ts'un ch'o keng lit** all are occasionally useful as supplements to the Yiian shitis, accounts. For studying institutional developments, the chih (treatises) sections of the Yiian shih are the most convenient source. The materials in the chih, however, must be supplemented by such collections of basic documents as the Yuan tien change and Tung chih t'iao ko,l> whose coverage stops at 1322 and 1315, respectively. Other useful basic materials for studying Yuan institutions include the surviving sections of the Ching shih ta tien,16 the institutional compendium compiled in 1330-1, and the several official compilations made in Toghon Temiir's reign. The latter include the Hsien t'ai t'ung chi'7 and Nan t'ai pet yao,lS both of which deal with the Censorate, and the Pi shu chien chih, which deals with the Palace Library.'9 The political history of the mid-Yuan received due attention in modern-style histories only in recent years. The best political narrative can be found in the Yiian ch'ao shih.2° This was edited by Professor Han Ju-lin, the leading authority on the Yiian history in China, but written by the staff members of the history department of Nanking University, including the prominent Yiian historians Ch'en Te-chih and Ch'iu Shu-shen. As a full-length modern historical account of the period, its publication in 1986 was an important milestone in the history of Yiian studies. An equally useful account of Yiian politics can be found in the Chung-kuo t'ung shih," which was written by another team of Chinese Yiian experts, including Ts'ai Mei-piao, Chou Liang-hsiao, and Chou Ch'ing-shu. Li Tse-fen's Yiian shih hsin chiang" provides the most lengthy, but not always a reliable, account of mid-Yuan politics. Except for the narrative histories by d'Ohsson and Henry Howorth in the last century, mid-Yiian politics did not receive the attention of Western scholars until the publication of John Dardess's Conquerors and Confucians^ seventeen years ago. Even though he was concerned primarily with the political changes in Toghon Temiir's reign, the author devoted the first two chapters of the book to the mid-Yiian as the necessary background for his discussion of late-Yiian politics. Although disagreeing with his main thesis that the restoration of Khaishan's imperial line in 1328 marked the Yiian's irrevocable secession from steppe politics and the beginning of the 12 K'ung Ch'i (K'ung K'e-ch'i), Chih cheng chih chi, invols. 3 2 1 - 4 of WuCh'ung-yao, ed., Yiiehyal'ang ts'ung shu (repr. of 1853 ed.; Taipei, 1965). 13 T'ao Tsung-i, Nan tsung ch'o keng iu (Peking, 1958). 14 Ta Yiian sbeng cheng kuo ch'ao tien cbang (Yiian lien chang) (Yiian ed., repr. Taipei, 1973) (hereafter cited as YTC). 15 A part of the t'iao-ko section of the Ta Yiian t'ung chih has survived and was reprinted in 1930 by the National Peking Library under the title T'ung chih t'iao ko (Code of comprehensive institutions). 16 For a complete list, see Su Chen-shen, Yiian cheng shu Chin shih ta tien chih yen chin (Taipei, 1984), pp. 40-50, 64-74. 17 Hsien t'ai t'ung chi in Yung-lo ta tin (Peking, i960), chiian 2608-9. 18 Nan t'ai peiyao, in Yung-lo ta tien (Peking, i960), chiian J6IO—11. 19 Pi shu chien chih (Kuang ts'ang hsiieh chiin ed.). so Han Ju-lin, comp., Yiian ch'ao shih, 2 vols. (Peking, 1986). 21 Ts'ai Mei-piao, Chou Liang-hsiao, Chou Ch'ing-shu et al., Chung-kuo t'ung shih, vol. 7 (Peking, 1983)22 Li Tse-fen's Yiian shih hsin chiang, vol. 3 (Taipei, 1978). 23 John Dardess, Conquerors and Confucians: Aspects ofpolitical change in late Yiian China (New York, 1973). Cambridge Histories Online © Cambridge University Press, 2008 BIBLIOGRAPHICAL ESSAYS 711 "thorough Confiicianization" of the Yuan political life, my chapter in the Cambridge history has greatly benefited from Dardess's incisive observations and analyses. In addition to Dardess's book, John D. Langlois, Jr.'s article on Yii Chi2'' covers the reign of Tugh Temiir through the career of this great Chinese literatus who devoted himself to enhancing Tugh Temiir's legitimacy and authority. Louis Hambis, the eminent French scholar of the Yuan period, also wrote an article on the Merkid Bayan,*5 the bureaucratic strongman active during Tugh Temiir's reign and the early part of Toghon Temiir's reign. Studies of the succession crises in the mid-Yuan and the early Mongolian Yuan period are plentiful. Yanai Wataru's classic study of the khuriltai,26 the assemblies during which the khaghans were supposedly "elected" by the Mongolian nobles, first published in 1917, is now dated. The most complete and up-to-date studies of the succession crises are two articles by Hsiao Kung-ch'in,37 which deal with the early Mongolian khaghans and the Yuan. He attributes the recurrent succession crises in the Yuan to the incomplete transformation of the Mongolian political system. Even though the "election" of a new khaghan in a khuriltai in the steppe tradition had become no more than a ritual, the concept of "election" was still manipulated by various factions to put their own candidates on the throne. Chou Liang-hsiao recently published an article that shows that in regard to ritual, the imperial successions in the Yuan still followed the Mongolian tradition.28 Two articles by Joseph Fletcher29 and another by Elizabeth Endicott-West,'0 though not dealing with the mid-Yuan, are conceptually useful in the study of the mid-Yiian succession crises. Fletcher contends that the nomads possessed no other system of succession than armed struggle, a phenomenon that he dubs "bloody tanistry." Endicott-West views the khuriltai not as an assembly of election but one of acclamation and political consultation. She uses the existence of this institution to refute the theory that Mongolian rule helped China's autocratic tradition. Apart from the works of Dardess, Langlois, and Hambis, studies of important political events and figures exist only in Chinese and Japanese. Uematsu Tadashi's article on the execution by the Yuan court in 1309 of Chu Ch'ing and Chang 24 John D. Langlois, Jr., "Yii Chi and his Mongol sovereign: The scholar as apologist," Journal ofAsian Studies, 38 (1978), pp. 99-116. 25 Louis Hambis, "Notes preliminaires a une biographic de Bayan le M&rkit," Journal Asiatique, 241 ('953). PP- 215-48. 26 Yanai Wataru, "Moko no kokkai sunawachi 'kurirutai' ni tsuite," repr. in Mokoshi kenkyu, ed. Iwai Hitosato (Tokyo, 1930), pp. 361-447. 27 Hsiao Kung-ch'in, "Lun Yuan tai huang weichi ch'engwen t'i," Yuan shih chipeifangmin tsushihyen chiu chi k'an, 7 (1983), pp. 22—39; "Lun Ta Meng-ku kuo te han wei chi ch'eng wei chi," Yuan shih chi pei fang min tsu shih yen chiu chi k'an, 5 (1981), pp. 48—59. 28 Chou Liang-hsiao, "Meng-ku hsiian han i chih yii Yuan ch'ao huang wei chi ch'eng wen t'i," Yuan shih lun ts'ung, 3 (1986), pp. 31-46. 29 Joseph Fletcher, "Turco-Mongolian monarchic tradition in the Ottoman empire," Harvard Ukrainian Studies, 3-4(1979-80), pp. 236-51, and "The Mongols: Ecological and social perspectives," Harvard Journal of Asiatic Studies, 46, (1986), pp. n—50. 30 Elizabeth Endicott-West, "Imperial governance in Yuan times," HarvardJournal of Asiatic Studies, 46 (1986), pp. 523-49. Cambridge Histories Online © Cambridge University Press, 2008 712 BIBLIOGRAPHICAL ESSAYS Hsiian,'1 two wealthy southern Chinese officials, uses theit case to illustrate the political alignments late in Temiir's reign, contending that the case was engineered by Empress Bulukhan's faction to strengthen its financial base in preparation for controlling the throne after Temiir's death. Matsuda Koichi explored the early life of Khaishan khaghan, especially his role in the steppe war against Khaidu.'2 Sun K'ok'uan, among his many pieces, wrote an article on what he calls "the Confucian rule" of Ayurbarwada's reign.» Hsiao Kung-ch'in's study of the murder of Shidebala khaghan in 1323W sees it as a result of the conservative reaction of the Mongolian and se—mu aristocracy and the bureaucracy to the sinicizing reforms carried out by the young monarch. K'uang Yii-ch'e considered the contributions of Baiju, grand councillor of the left, to Shidebala's reforms." Fujishima Tateki examined in a brief article'6 the tragic life of Khoshila, who was murdered in 1229. In sum, the existing studies of the mid-Yuan court politics are still far from being sufficient. Among other things, the "new deal" carried out by Khaishan khaghan to increase the state's revenue and the politics and policies of Yesiin Temiir's government have not been carefully investigated. The key political figures of the period — Harghasun, Li Meng, Daula-shah, and El Temiir — all merit biographical research. And the exact ideological alignments of various factions also need to be further explored. Because the Yuan state was only partly bureaucratic and partly patrimonialfeudal, the appanaged nobility, imperial relatives, and meritorious officials played an important role in imperial politics. The appanaged nobility, collectively known in Chinese as t'ou-hsia, has received much scholarly attention. In addition to the earlier studies by Wu Han, Murakami Masatsugu,37 Iwamura Shinobu,'8 Paul Ratchnevsky," Chou Liang-hsiao,4° and Hung Chin-fu41 have each written a substantial article on the t'ou-hsia as a system of political, military, judicial, and fiscal privileges and on its development in the Yuan. There is also a growing literature on the study of individual princely and noble houses. The following are particularly relevant to the scrutiny of mid-Yuan politics: Ebisawa TetsuoV* and Horie Masaaki's4' studies 31 Uematsu Tadashi, "Gendai Konan no gomin Su Sei, Cho Sen ni tsuite," Toyoibi ktnkyu, 27 (1968), pp. 46-71. 32 Matsuda Koichi, "Kaishan no seihoku Mongoria shussci," Tobogaku, 64 (1982), pp. 73—87. 33 Sun K'o-k'uan, "Chiang-nan fang hsien yii Yen-yu ju chih," in Yuan tai Han wen hua te huo lung, ed. Sun K'o-k'uan (Taipei, 1968), pp. 345—63. 34 Hsiao Kung-ch'in, "Ying-tsung hsin cheng yii Nan-p'o chih pien," Yuan shib chi peifang min tsu sbih yen chiu chi k'an, 4 (1980), pp. 36-46. 35 K'uang Yii-ch'e, "Pai-chu chi ch'i hsin cheng," Nei Meng-tu she bui k'o hsiieh, 5 (1984), pp. 59-62. 36 Fujishima Tateki, "Gen no Minso no shogai," Otani shigaiu, 12 (1970), pp. 12—28. 37 Murakami Masatsugu, "Gencho ni okeru toka no igi," Moko gakuho, 1 (1940), pp. 169-215. 38 Iwamura Shinobu, Mongoru shakai keizaishi no kenkyu (Kyoto, 1964), pp. 401-69. 39 Paul Ratchnevsky, "Zum Ausdruck t'ouhsia in der Mongolenzeit," in Collectanea Mongolica (Wiesbaden, 1966), pp. 173—91. 40 Chou Liang-hsiao, "Yiian tai t'ou hsia fen feng chih tu ch'u t'an," Yiian sbib lun u'ung, 2 (1983), pp. 53-76. 41 Hung Chin-fu, "Ts'ung t'ou hsia fen feng chih tu k'an Yiian ch'ao cheng ch'iian te hsing chih," Chung yang yen chiu yiian li shih yii yen yen chiu so chi k'an, 58(1987), pp. 483—907. 42 Ebisawa Tetsuo, "Mongoru teikoku no toho sanoke ni kansuru shomondai," Saitama daigaku kiyo, 21 (1972), pp. 3 1 - 4 6 . 43 See Horie Masaaki, "Temuge Otchigin to sono shishon," Toyo shien, 24-5 (1986), pp. 225-70; and Cambridge Histories Online © Cambridge University Press, 2008 BIBLIOGRAPHICAL ESSAYS 713 of the so-called three eastern ulus, the descendants of Chinggis khan's younger brothers; Matsuda Koichi's study of the house of the prince of An-hsi, Ananda;44 and Chou Ch'ing-shu's study on the royal house of the Onggiid tribe.45 Hsiao Ch'i-ch'ing also looked at the descendants of the four distinguished companions of Chinggis khan, demonstrating their political durability and their patrimonial-feudal and bureaucratic character.46 These past studies emphasized the t'ou-hsia as a system of hereditary privileges; more attention needs to be paid to the actual roles played by this stratum of nobility in imperial politics vis-a-vis the bureaucracy. The various Western and Central Asian ethnic groups, collectively known as the se-mu, played important but varied roles in Yuan politics. The vicissitudes of these groups' political fortunes are a topic that has attracted much attention. The following investigations are relevant to the mid-Yuan politics: Yang Chih-chiu's study of the Muslims;47 Igor de Rachewiltz's study of the Turks as a whole;48 Lu Chun-ling's and Ho Kao-chi's joint article on the Turkish Qangli, Qipchaq, and Asud tribesmen;4' T'ang K'ai-chien's study of the Tangut;50 and Sechen Jagchid's (Cha-ch'i Ssuch'in)'1 and Herbert Franke's studies on the Tibetans.'2 These studies show that the functions and fortunes of these se—mu groups in the Yuan government varied greatly. Now we should analyze the political role of the se—mu as a group in regard to their changing fortunes. Yiian's position in the Inner Asian steppes and its relations with the western khanates were important, as they affected the Yuan ruler's position as the khaghan of all Mongols but also the government's political and cultural position in China. The literature on Yiian's war and peace with Khaidu's and Du'a's camp is plentiful and still growing. W. Kotwicz, an eminent Polish scholar, published his pioneering study fifty years ago on the peace of the Mongolian world achieved in 1303." Etani 44 45 46 47 48 49 50 51 52 53 his earlier article, "Mongoru — Gencho jidai no toho san-urusu kenkyu josetsu," in Tohogaku ronshu: One Katsuioshi hakushi shoju kitten tohigahu ronshu, ed. Ono Katsutoshi hakushi shoju kinenkai (Kyoto, 1982), pp. 377-410. Matsuda Koichi, "Gencho chi no bunposei - Anseio no jirei o chushin to shite," Sbigaku zasshi, 88 ('979). PP- 1249-86. Chou Ch'ing-shu, "Wang-ku pu shih-chi," in Chung-kuo Meng-ku shih hsiieh hui ch'eng li ta hui chi nien chi-k'an (Huhehot, 1979), pp. 147—206. Hsiao Ch'i-ch'ing, "Yuan tai ssu ta Meng-ku chia tsu," in his Yiian tai shih hsin t'an (Taipei, 1983), pp. 141-230. Yang Chih-chiu, "Yiian tai Hui-hui jen te cheng chih ti wei," in his Yiian shih san lun (Peking, 1985), pp. 245-82. Igor de Rachewiltz, "Turks in China under the Mongols: A preliminary investigation of Turco-Mongol relations in the 13th and 14th centuries," in China among equals: The Middle Kingdom and its neighbors, 10th—14th centuries, ed. Morris Rossabi (Berkeley and Los Angeles, 1983), pp. 281—310. Lu Chun-ling and Ho Kao-chi, "Yiian tai te A-su Ch'in-ch'a K'ang-li jen," Wen shih, 16 (1982), pp. 117-30. T'ang K'ai-chien, "Yiian tai Hsi-hsia jen te cheng chih ti wei," Kan-su min tsu yen chiu, 1-2 (1987), pp. 10—26. Cha-ch'i Ssu-ch'in (Sechin Jagchid), Meng-ku yii Hsi-tsang li shih kuan hsi chih yen chiu (Taipei, 1978). Herbert Franke, "Tibetans in Yiian China," in China under Mongol rule, ed. John D. Langlois, Jr. (Princeton, 1981), pp. 296—328. W. Kotwicz, "Les Mongols, promoteurs de l'idee de paix universelle au d£but du XIII-c [sic] siecle," Rocznik Orientalistyczny, 16 (1953), pp. 428-34. Cambridge Histories Online © Cambridge University Press, 2008 714 BIBLIOGRAPHICAL ESSAYS Toshiyuki, a Japanese scholar, wrote an article on Khaidu's war with the Yuan.'4 Liu Ying-shen is currently the most active researcher on the Yuan's relations with the Central Asian khanates and has written several articles on the subject, comparing the Chinese and the Persian sources." The lengthy articles by Saguchi Toru, the senior Japanese Mongolist, on the 1303 peace and on later Yuan relations with the western khanates, published in 1942, still are useful in regard to the Yuan's post-1303 relations with the western khanates.'6 Thomas Allsen documented the struggle between the Yuan and the Ogodei and Chaghadai khanates for the control of Uighuristan." In addition to his study of the changing Yuan relations with the steppe, in his Conquerors and Confucians, Dardess wrote an interesting and wellresearched article on the limitations of the Yuan's efforts to control Mongolia and Central Asia from distant China, attributing the loss of Central Asia in the late 1320s to what he calls "spatial limitations."'8 The study of the political history of a period cannot entirely omit the political, judicial, economic, and cultural institutions and policies of its government. The new political institutions of the mid-Yuan, the Department of State Affairs (Shangshu sheng), which was established by Khaishan to increase the state's revenue, was the subject of a monograph by the Japanese scholar Aoyama Koryo.'9 The civil service examination system restored by Ayurbawarda was investigated by Miyazaki Ichisada,60 Yang Shu-fan,6' Yao Ta-li,62 and Ting K'un-chien.6' Hsiao Ch'i-ch'ing studied the impact of the restoration of the civil service examinations on the elite's mobility on the basis of the chin-shih of the 1333 class.64 Concerning judicial developments, Paul Ratchnevsky's Un code des Yuan6* remains a rich mine that every student of Yuan institutional history must explore. Paul H. C. Chen's monograph Chinese legal tradition under the Mongols,66 though chiefly dealing with the code of 54 Etani Toshiyuki, "Kaido no ran ni kansuru no ichi kosatsu," in Tatnura hakmhi shoju Toyosbi ronso (Kyoto, 1968), pp. 89—104. 55 Liu Ying-sheng, "Shih-chi Wo-k'uo-t'ai han kuo mo nien chi shih pu cheng," Yiian shih cbipei fang min tsu shih yen chiu chi k'an, 10 (1986), pp. 4 8 - 5 9 ; "Yiian-tai Meng-ku chu han kuo chien te yileh ho chi Wo-k'uo-t'ai han kuo te mieh wang," Hsin-chiang la hsiieh biiiehpao, 2 (198;), pp. 3 1 - 4 3 . 56 Saguchi Toru, "Jushi seiki ni okeru Gencho daiken to seihd san-oke to no rentaisei ni tsuite," Kita Ajia gakuho, 1 (1942), pp. 1 5 1 - 2 1 4 . 57 Thomas T. Allsen, "The Yiian dynasty and the Uighurs of Turfan in the 13th century," in China among equals: The Middle Kingdom and its neighbors, 10th— 14th centuries, ed. Morris Rossabi (Berkeley and Los Angeles, 1983), pp. 281—310. 58 John D. Dardess, "From Mongol empire to the Yiian dynasty: Changing forms of imperial rule in Mongolia and Central Asia," Monumtnta Serica, 30 (1972-3), pp. 117-65. 59 Aoyama Koryo, Gencho shoshosho ko (Tokyo, 1951). 60 Miyazaki Ichisada, "Gencho chika no Mokoteki kanshoku wo meguru Mo Kan kankei-kakyo fukko no igi no saikento," Toyoshi kenkyu, 23 (1965), pp. 4 2 8 - 9 1 . 61 Yang Shu-fan, "Yiian-tai k'o-chiichih-tu," Kuo li cheng chih ta hsiieh hju'eh pao, 17(1968),pp. 99—120. 62 Yao Ta-li, "Yuan tai k'o chii chih tu te hsing fei chi ch'i she hui pei ching," Yiian shih chi pei fang min tsu shih yen chiu chi k'an, 6 (1982), pp. 2 6 - 5 9 . 63 Ting K'un-chien, "Yiian tai te k'o chii chih tu," Hua hsiieh yueh k'an, 124 (1982), pp. 4 6 - 5 7 . 64 Hsiao Ch'i-ch'ing, "Yuan tai k'o chu yii ching ying liu tung, I: Yuan-t'ung yiian nien chin shih wei chung hsin," Han hsiieh yen chiu, 5 (1987), pp. 129-60. 65 Paul Ratchnevsky, Vn code des Yuan, 4 vols. (Paris, 1937-85). 66 Paul Heng-chao Ch'en, Chinese legal tradition under the Mongols: The code of 1291 as reconstructed (Princeton, 1979), pp. 108—9. Cambridge Histories Online © Cambridge University Press, 2008 BIBLIOGRAPHICAL ESSAYS 715 1291, also surveys the activities of codification in the mid-Yuan as well as its other periods. There is no single work that focuses exclusively on the mid-Yuan period's fiscal and economic policies. Rather, we must rely on the works that deal with the Yuan period as a whole. Of these, the sections on taxes, maritime transportation, paper currency, and governmental monopolies in Schurmann's Economic structure of the Yuan dynasty61 are good starting points. Tayama Shigeru wrote on the revenues and expenditure of the Yuan government.68 The monetary system receives the fullest treatment in Herbert Franke's Geld und Wirtschaft in China unter der Mongolenherrshaft, in which he contends that the Yuan never issued an excessive amount of paper currency and that the currency policy had no direct influence on the dynasty's ultimate collapse.*» Franke's thesis varies in different ways from the findings of Ch'iian Han-sheng,7° Maeda Naonori, Iwamura Shinobu,71 and P'eng Hsin-wei.7J The annual imperial grants (sui-tz'u) given to imperial relatives, which constituted a serious financial drain on the government, were examined by Shih Wei-min." The maritime transportation of grain from the lower Yangtze valley to the capital region, which was crucial to the support of the central government's economic and political stability, was considered by Wu Chi-hua. 7< There have been many studies of the cultural orientation and policies of the midYuan court. Yoshikawa Kojiro's examination of the Yuan khaghans' competence in the Chinese high culture,75 first published in 1943—5, remains the most comprehensive work on the subject. Herbert Franke's article "Could the Mongol emperors read and write Chinese?"76 parallels Yoshikawa's work but was written independently. Both articles show that most of the mid- and late-Yuan monarchs commanded varying degrees of competence in Chinese high culture. The patronage of the arts and letters by Tugh Temiir khaghan was explored by Chiang I-han and Fu Shen in their studies of a key institution created by the khaghan for that purpose, the Pavilion of the Star of Literature (K'uei chang ko).77 Fu's work also covers Toghon Temiir's reign. The translation of Chinese works into Mongolian under the auspices 67 H. F. Schurmann, Economic structum of the Yuan dynasty (Cambridge, Mass., 1956; repr. 1967), pp. 139-40. 68 Tayama Shigecu, "Gendai zaiseishi ni kansuru oboegaki - shushi no gaku o chushin co shite," in Toyo no seiji keizai (Hiroshima, 1949), pp. 191-266. 69 Herbert Franke, Geld und Wirtschaft in China unter der Mongolenherrschaft (Leipzig, 1949), pp. 57—9. 70 Ch'iian Han-sheng, "Yiian-tai te chih-pi," repr. in his Chung-kuo ching chishih lun ts'ung, vol. 1 (Hong Kong, 1972), pp. 369-416. 71 Iwamura Shinobu, Mongoru shakai keizaishi no kenkyu (Kyoto, 1968), pp. 421-32. 72 P'eng Hsin-wei, Chung-kuo huopi shih (Shanghai, 1958), pp. 409-10. 73 Shih Wei-min, "Yuan sui-tz'u k'ao shih," Yuan shih lun ts'ung, 3 (1986), pp. 144-53. 74 Wu Chi-hua, "Yiian-ch'ao yii Ming ch'u hai yun," Chung yang yen chiu yuan, li shih yil yen yen chiu so chi fan, 28 (1956), pp. 363—80; repr. in his Ming tai she hui ching chi shih lun ts'ung, 2 vols. (Taipei, 1970), vol. 1, pp. 125-54. 75 Yoshikawa Kojiro, "Gen no shotei no bungaku," in Yoshikawa Kojirg zenshu, vol. 15 (Tokyo, 1969), PP- 233-3O376 Herbert Franke, "Could the Mongol emperors read and write Chinese?" Asia Major, n.s. 3 (1952), pp. 28-41. 77 Chiang I-han, Yuan tai K'uei-fhang ko chi K'uei-chang jen wu (Taipei, 1981); and Fu Shen, Yuan tai huang shih shu hua sou ts'ang shih lu'eh (Taipei, 1981). Cambridge Histories Online © Cambridge University Press, 2008 BIBLIOGRAPHICAL ESSAYS Jl6 of various emperors was considered by Walter Fuchs,78 Herbert Franke," and Igor de Rachewiltz.80 To dispel the general misconception that the Mongols in the Yuan were generally ignorant of Chinese culture, Hsiao Ch'i-ch'ing devoted two articles to 102 Mongols, the majority of them officials in the mid- and late-Yiian periods, who mastered this culture.8' The se-mu elite's competence in Chinese was demonstrated convincingly by Ch'en Yuan in his famous book Yiian Hsi-yii jen hua hua k'ao, which was first published in the 1920s and is now available in English translation.83 But there is still no reliable work on how well the steppe culture was retained by the Yuan court and the Mongolian elite. Thus our picture of the cultural orientation of the Yiian state and the Mongolian elite remains one-sided and incomplete, as we can clearly see that the Mongols in China were progressively sinicizing, but we are uncertain about what aspects of their steppe culture they retained as the Yiian regime matured and aged as a dynasty of conquest. 7. SHUN-TI AND THE END OF YUAN RULE IN CHINA For the reign of Shun-ti, the main source is the Yiian shih, the standard dynastic history published very early in the Ming. However, it is important to note that the Yiian sbih was actually compiled in two stages. The first and major installment, covering the period from the beginning down to the reign of Ning-tsung (1332), was finished in August 1369. For the reign of Shun-ti, 1333 to 1368, a second and rather different project was required. Because no veritable record (shih lu) had been kept at the Yiian court in Shun-ti's time and because the Yiian institutional encyclopedia (the Ching shih ta lien of 1332) had not been continued, the Ming government had to conduct a special search for source materials. Accordingly, some dozen envoys were sent to different parts of the realm to collect such materials and forward them to the history office in Nanking. Sung Lien's collected works contain a fairly detailed account of how the most important of these missions, which was sent to the former Yiian capital, conducted its work. There had never been kept a veritable record for the thirty-six years of Shun-ti. The historiographers had nothing to consult, and the era was missing. Accordingly, Ts'ui Liang, minister of rites, Huang Su, the secretary, and I issued guidelines. Eleven envoys were sent to various parts of the realm to collect everything of historical relevance and send it in. What is now named Pei-p'ing was the former Yuan 78 Walter Fuchs, "Analecta zur mongolischen Ubersetzungsliteratur der Yiian-zeit," Monumtnta Serica, 11 (1946), pp. 3 3 - 4 6 . 79 Herbert Franke, "Chinese historiography under Mongol rule: The role of history in acculturation," Mongolian Studies, 1 (1974), pp. 15—26. 80 Igor de Rachewiltz, "The preclassical Mongolian version of the Hsiao-ching," ZentralasiatischeSludien, 16(1982), pp. 7 - 1 0 9 . 81 Hsiao Ch'i-ch'ing, "Yiian tai Meng-ku jen te Han hsiieh," in Kuo cbi Chung-kuopin cbiang bsiieb sbu hui i lun wen chi (Taipei, 198;), pp. 369—428; and his "Yuan tai Meng-ku jen Han hsiieh tsai t'an," in Yang Lien-sheng et al., eds., Kuo shib shib lun: T'ao Hsi-sbeng hsien sheng chiu cb'ib jung cb'ing cbu shou lun wen chi, 2 vols. (Taipei, 1988), vol. 2, pp. 373-88. 82 Ch'en Yuan, Western and Central Asians in China under the Mongols, trans, and annotated by Ch'ien Hsing-hai and L. C. Goodrich (Los Angeles, 1966), pp. 241-52. Cambridge Histories Online © Cambridge University Press, 2008 BIBLIOGRAPHICAL ESSAYS 717 capital, and Shantung was also an important place, and {it was understood that] documents of the period must survive there. In view of these considerations it was decided to send someone of official rank. Lii Chung-shan of Chang-kung [Kan-chou, Kiangsi), at the time food director [ssu shari] in the National University, was thus chosen for this mission. On 13 August 1369, Lii left [Nanking] for the north by government conveyance. On 6 September he arrived in Pei-p'ing. He gathered together all edicts, memorials, notices of appointment and dismissal, and official requests that were available as documents. Anything pertinent that was in foreign script was translated. Every kind of nondocumentary source — things relating to imperial travel, secret palace affairs, political criticism, folk songs, and all references to moral behavior, outlawry, and portents that could be found in private histories, inscriptions, and the collected writings of Confucians — was assiduously consulted. The Confucian student Wei Yii [Wei Su's elder son] and others were sent out to the prefectures of Luan [Shang-tu] and Yen-nan [southern Hopei] with strict deadlines [for submitting materials]. Lii opened his office in the former National University on 16 September and finished all the work on 30 November. Altogether eighty boxes [of materials were prepared]. Each was wrapped with Kao-li ts'ui-chih [Korean green paper?] for shipment to the Branch Secretariat, where official seals were applied, [and the whole lot was then] forwarded to Nanking. On 4 December Lii departed for Shantung. The Yellow River was then frozen, and two to three feet of snow covered the ground. He proceeded along the road by oxcart, ice accumulating in his beard and moustache with every breath he took. On 27 December he arrived in Chi-nan, where he consulted materials much along the lines of what he had already done in Pei-p'ing. In addition, he collected four hundred squeezings of inscriptions in Pei-p'ing, and one hundred more in Shantung. These were duly sealed [and forwarded], as earlier. On 9 April he finally returned to the capital [Nanking]. At length the historiographers were gathered, and I had the materials necessary to complete forty-eight [more chiian of the dynastic history}. These were presented to the throne that summer [23 July 1370].' Despite these earnest efforts at collecting information, the dynastic history's coverage of the late Yiian tends to be spotty and haphazard. The whole project was done with such speed largely for the reason that Shun-ti and his successor, Ayushiridara, were still active in Mongolia and had not relinquished their claim to be the legitimate rulers of China. The publication of the Yiian shih was intended as a clear signal to everyone within and beyond China that the Yiian dynasty should be considered as officially over and done with. Several unofficial accounts that served as sources for the later portions of the Yiian shih have survived independently. Some of these have been translated and annotated, notably Yang Yu's Shan chii hsin hua, by Herbert Franke;2 the Keng shen wai shih by Helmut Schulte-UffelagejJ and the second part of Erich Haenisch's Zum Untergang Zu/eier Reiche,* which translates a diary of Shun-ti's flight north, 1368-70. For further information, the reader should consult Herbert Franke, "Some aspects of Chinese private historiography in the thirteenth and fourteenth centuries," Historians of China andJapan, ed. W. G. Beasley and E. G. Pulleyblank.' A rich and varied body of source material for the late Yiian, as yet little exploited, 1 Sung Lien, Sung wen hsien cb'iian chi (SPPY ed.), 7, pp. 158-b. 2 Herbert Franke, Beitrdge zur Kulturgeschichte Chinas unter dtr Mongolenhcrrschaft. Das Shan-kii sin-hua del Yang Yii (Wiesbaden, 1956). 3 Helmut Schulte-Uffelage, Das Keng-sbm wai-shih. Eine Quellt zur spa'ten Mongolenzeit (Berlin, 1963). 4 Erich Haenisch, Zum Untergang zwtier Reicht (Wiesbaden, 1969). 5 Herbert Franke, "Some aspects of Chinese private historiography in the thirteenth and fourteenth centuries," in Historians of China andJapan, ed. W. G. Beasley and E. G. Pulleyblank (London, 1961), pp. 115-34. Cambridge Histories Online © Cambridge University Press, 2008 718 BIBLIOGRAPHICAL ESSAYS may be found in the collected literary works (wen chi) of contemporary people, principally but not exclusively of Chinese ancestry. Exhaustive lists of editions and of library holdings in Japan have been prepared by Yamane Yukio and Ogawa Hisashi6 and, for mainland China, by Chou Ch'ing-shu.7 The study of the late Yiian period was begun in the West only recently, and a great deal remains to be done. In China, its study goes back at least to 1936 and Wu Han's long and detailed article "Yiian ti kuo chih peng k'uei yii Ming chih chien kuo"8 (The fall of the Yiian empire and the rise of the Ming). One of the most active centers of new publication is the Research Institute for the History of the Yiian Dynasty at the University of Nanjing (Nan ching ta hsiieh li shih hsi, Yiian shih tsu), which issues the annual periodical Yiian shih chi pet fang tnin tsu shih yen chiu chi k'an9 (Studies in the history of the Yiian dynasty and of the northern nationalities). Since 1982, the Research Society on Yiian History (Yiian shih yen chiu hui) has also published an ongoing series of collected articles on Yiian history, entitled Yiian shih lun ts'ung.I0 Given that the late Yiian to some extent involves Altaic and Mongolian studies and Ming history, as well as studies of the Yiian itself, keeping abreast of new work can be difficult. Two useful newsletters with occasional coverage of work on the late Yiian are the Bulletin of Sung-Yiian Studies, formerly the Sung Studies Newsletter, and Ming Studies. 8. T H E YOAN G O V E R N M E N T AND SOCIETY The primary sources that yield the most information about the Yiian government are the Yiian shih (Official history of the Yiian, compiled 1369-70), the Yiian tien chang (Institutions of the Yiian dynasty, compiled 1320-2), and the Tung chih t'iao ko (Code of comprehensive institutions, compiled 1321).' Because the veritable records (shih-lu) of the Yiian imperial reigns are no longer extant, historians today do not have anywhere near the volume of primary source materials on the Yiian that they have for the Ming and Ch'ing periods.2 On the one hand, although the Yiian shih preserves much of the original source material, on the other hand, it tends to abridge material, at times 6 Yamane Yukio and Ogawa Hisashi, eds., Nihon genson Genjin bunsbi mokuroku (Tokyo, 1970). 7 Chou Ch'ing-shu, ed., Yiian jen wen chi pan pen mu lu (Nanking, 1983). 8 Wu Han, "Yiian ti kuo chih peng k'uei yii Ming chih chien kuo," Ch'ing hua hsiiehpao, 11 (1936), pp. 359-4239 Yiian shih chi pei fang min tsu shih yen chiu chi k'an, Research Institute of the History of the Yiian Dynasty, University of Nanjing. 10 Yiian shih lun ts'ung, Research Institute for the History of the Yiian Dynasty, University of Nanjing ed. (Peking, Chung-hua shu-chu), vol. 1, 1982; vol. 1, 1983; vol. 3, 1986. 1 For brief descriptions of the Yiian tien chang (YTC) and the T'ung chih t'iao ko (TCTK), see the handy reference dictionary entitled Chung kuo li shih ta tz'u tien: Liao, Hsia, Chin, Yuan shih (Shanghai: Shanghai tz'u-shu ch'u-pan, 1986), pp. 55—66 on YTC, and pp. 16—17 on TCTK. See also the chapter on the Yiian in Ch'en Kao-hua, Chung kuo tu lai shih shih liao hsiieh (Peking, 1983), pp. 311— 53. See also the introductory essay to the 1986 People's Republic of China edition of the TCTK: Yiian tai shih liao ts'ung k'an: T'ung chih t'iao ko (Che-chiang ku-chi ch'u-pan-she), pp. 1—5. 2 A recent reappraisal of the process by which the Yiian shih was compiled is that by Yeh Yu-ch'uan and Wang Shen-jung, "Yiian shih t'an yiian," Wen shih 27 (1986), pp. 177—94; see pp. 178—83 on the fete of the Yiian veritable records. See also Francis Woodman Cleaves, "The memorial for presenting the Yiian shih," Asia Major, 3rd series, vol. 1, pt. 1 (1988), pp. 5 9 - 6 9 . Cambridge Histories Online © Cambridge University Press, 2008 BIBLIOGRAPHICAL ESSAYS 719 merely alluding to, or summarizing, documents that are more fully preserved in the Yiian tien chang and T'ungchib t'iao ko. In this respect, the Yuan shib itself can be viewed as a secondary source, as its editors were removed by as much as three or four generations from some of the events whose descriptions they compiled. The difficult colloquial style of language of the documents in the Tung chih t'iao ko and the Yiian tien chang has deterred scholars from exploring more fully the contents of these works. Herbert Franke aptly described the peculiar language of these documents as "a sort of translationese Chinese in Mongolian word-order."3 Mongolian versions of many of these documents must have existed, as the Yiian administration was bilingual, unlike the Khitan Liao and Jurchen Chin administrations, which used Chinese as their official language. Those imperial decrees that are specifically referred to as sheng-chih in the Yiian lien chang and other Yiian sources were translated from Mongolian into colloquial Chinese.4 The reading and interpretation of sheng-chih requires some knowledge of Mongolian because the decrees incorporate Mongolian syntax. Japanese and Soviet scholars have contributed much to the deciphering of the Yiian tien chang and the T'ung chih t'iao ko. One of the most recent contributions to this field is the monograph by the Soviet linguist I. T. Zograph entitled Mongol'skokitaiskaia interferentsiia: iazyk mongol'skoi kantseliarii v kitae (Moscow, "Nauka," 1984). Zograph's work uses for analysis the Yiian period colloquial-language stone inscriptions, which were published in the People's Republic of China in 1955 by the Yuan historian Ts'ai Mei-piao.' The Japanese scholars Iwamura Shinobu and Tanaka Kenji published a two-volume annotated version of the Board of Punishments chapters {chiian 39—57) of the Yiian tien chang; appended to the first volume of their work are two useful essays, one by Tanaka Kenji and the other by Yoshikawa Kojiro, on the documentary style of language in the Yiian tien chang.6 Translations of Yiian administrative and legal regulations into French and English have enriched the field of institutional history. Paul Ratchnevsky's four-volume annotated translation of the entire monograph on punishments and laws {hsing fa chih) in the Yiian shih (chiian 102—5) a ' s 0 includes many translated portions of relevant documents from the Yiian.tien chang. The third volume of Ratchnevsky's work, the index, is invaluable for tracing Yiian institutional terminology.' Paul 3 Herbert Franke, "Chinese historiography under Mongol rule: The role of history in acculturation," Mongolian Studies 1 (1974), p. 16. 4 For an interesting discussion, see Lien-sheng Yang, "Marginalia to the Yiian tien chang," in his Excursions in Sinology (Cambridge, Mass.: Harvard University Press, 1969), pp. 126-35, e s P' PP- 126—30. 5 Ts'ai Mei-piao, Yiian tat pai huapei chi lu (Peking: K'o-hsiieh ch'u-pan-she, 1955). 6 Iwamura Shinobu and Tanaka Kenji, Gentensho ktibu dai issatsu (Kyoto: Kyoto University Press, 1964); and Iwamura Shinobu and Tanaka Kenji, Gentensho keibu dai nisatsu (Kyoto University Press, 1972). The two essays appended to the first volume are Yoshikawa Kojiro's "Gentensho ni mieta kanbun ritoku no buntai," pp. 1—4;; and Tanaka Kenji's "Gentensho ni okeru Mdbun chokuyakutei no bunsho," pp. 47-161. 7 Paul Ratchnevsky, Vn code da Yuan (Paris: Librairie Ernest Leroux, 1937); Rarchnevsky, Un code des Yuan, vol. 2 (Paris: Presses universitaires de France, 1972); Ratchnevsky and Franchise Aubin, Un code des Yuan, vol. 3, Index (Paris: Presses universitaires de France, 1977); Ratchnevsky, Un code des Yuan, vol. 4 (Paris: College de France, Institut des hautes Itudes chinoises, 198;). Cambridge Histories Online © Cambridge University Press, 2008 72O BIBLIOGRAPHICAL ESSAYS Ch'en reconstructed and translated the text of the Yiian code promulgated in 1291, the Chih yiian hsin ko.8 Some Yiian political and economic institutions are not specifically described as such in the monograph sections of the Yiian shih. The office of darughachi receives no special attention in the Po kuan chih (Monograph on official posts) in the Yiian shih, and the institution of the ortogh (Muslim merchant associations working in partnership with the Mongolian elite) receives no special section in the Shih huo chih (Monograph on financial administration). Such peculiarly Mongolian or Inner Asian institutions seem to have evaded the usual categories of information in Chinese historiographical conventions, and thus modern-day historians must draw on a wide variety of Yiian sources to piece together descriptions of these institutions. Among these other sources are the Yuan literary collections (wen chi), which provide important details on Mongolian political and fiscal policies. A useful topic index to the 170 extant Yiian and early Ming period literary collections was published in the People's Republic of China in 1979.' The ten extant Yiian period local gazetteers, reprinted in Taiwan in 1980, provide some insight into how government institutions functioned in specific locales.10 Finally, the secondary scholarship in English that has contributed most to the reconstruction of Yiian civil, military, legal, and fiscal institutions was produced by two generations of students trained by Professor Francis W. Cleaves. Among those Ph.D. dissertations directed by Professor Cleaves and later published as monographs are H. F. Schurmann's Economic structure of the Yiian dynasty: Translation of chapters 93 and 94 of the Yiian shih (1956); Ch'i-ch'ing Hsiao's The military establishment of the Yuan dynasty (1978); Paul Ch'en's Chinese legal tradition under the Mongols: The code of 1291 as reconstructed (1979); and Elizabeth Endicott-West's Mongolian rule in China: Local administration in the Yiian dynasty (Cambridge: Harvard University Council on East Asian Studies, Harvard—Yenching Institute Monograph Series no. 29, 1989). Two other as yet unpublished manuscripts that originated as Ph.D. dissertations under Professor Cleaves are Isenbike Togan's "The chapter on annual grants in the Yiian shih," 2 vols. (Ph.D. diss., Harvard University, 1973); and Yiian-chu Lam's "The first chapter of the "Treatise on selection and recommendation' for the Civil Service in the Yiian shih" (Ph.D. diss., Harvard University, 1979). Professor Cleaves's five dozen or so articles themselves treat a wide variety of topics ranging from the purely philological to the literary, cultural, and historical aspects of the Yiian dynasty. A bibliography of Cleaves's articles published between 1934 and 1985 appears in the Festschrift volume dedicated to him." 8 Paul Heng-chao Ch'en, Chinese legal tradition under the Mongols: The code of 1291 as reconstructed (Princeton, N J . : Princeton University Press, 1979). 9 See Lu Chun, comp., Yiian jtn wen chi p'irn mu fen lei so yin (Peking: Chung-hua shu-chii, 1979). 10 Sung Yiian ti fang chih tan shih ch'ichung, 12 vols. (Taipei: Kuo-t'ai wen-hua shih-yeh, 1980). 11 See Nigula Bilig. Pi wen shu. An anniversary volume in honor of Francis Woodman Cleaves. Published as the Journal of Turkish Studies, 9 (1985), pp. v—vii. Cambridge Histories Online © Cambridge University Press, 2008 BIBLIOGRAPHICAL ESSAYS 721 9. CHINESE SOCIETY UNDER MONGOL RULE The social history of the Yuan period shares many of the problems and the sources, as well as the modern scholarship that the authors of the other chapters in this volume have discussed. Yet some further comments on sources and problems of scholarship bearing specifically on problems of Yuan social history may not be out of place here. Late in the nineteenth century and early in the twentieth, contacts with Western historians introduced the Chinese to medieval European and Western Asian sources and aroused a new kind of Chinese awareness of the Mongols' Eurasian empire. Previously paid little attention in China, in that era's atmosphere of heightened nationalism, the Chinese identified with the Mongolian conquests as a valued phase of Chinese history and began to take a broader look at the steppe societies as components of the modern Chinese nation. Simultaneously, however, that same element of nationalism also led them to resent the Mongols as alien conquerors and occasionally, somewhat anachronistically, to decry the thirteenthand fourteenth-century Chinese who had collaborated with them in ruling China. In short, the earlier culture-centered attitudes toward the period began to be replaced by modern nationalistic ones, leading to quite different evaluations of that era, especially of its social history. Modern scholarship on the social history of the Yuan dynasty thus is fraught with tensions and contradictions. To complicate the scene, much recent scholarship on the period within China is not only committed to fairly rigid Marxian doctrinal views, but it must also accommodate both the claims of the period's essentially Chinese social history and those of the nomadic steppe societies to that set of rigidities, while also incorporating a large volume of new scholarly material derived from archaeological evidence and new combings of the traditional sources. Social historians are therefore constantly challenged to evaluate and make the best possible use of a large and interesting body of old and new materials. In the last half of the nineteenth and early twentieth centuries, Mongolian studies were of greatest pertinence to the Russians and the Japanese, both of whom had imperial interests in Mongolia and its bordering lands. Their scholarly researches, as eventually also the French and German, and more recently the British and American Mongolian studies, understandably were heavily philological. They also involved sinology in topics in which Chinese-language materials bore on the research into Mongolian texts. The classic example is the so-called Secret history of the Mongols, discussed in the "Note on traditional sources for Yuan history" appended to this volume. The study of Yuan period social history, like that of all fields of Yuan history, still depends heavily on sound philological and historiographical research. That is, as the focus of our scholarly endeavors broadens, the basic skills still must be The author acknowledges with deep gratitude the oral and written advice and criticism received from Professor Hsiao Ch'i-ch'ing on an earlier draft of this essay; it has been substantially revised in light of that assistance, but Professor Hsiao has not seen the final version and is not to be held responsible for any remaining shortcomings. Cambridge Histories Online © Cambridge University Press, 2008 722 BIBLIOGRAPHICAL ESSAYS those of the pioneering Mongolists and sinologists. An acceptance of that principle guides the comments that follow. Mongolian studies, of course, form an independent field of research. Many eminent Mongolists have not been students of China, yet the histories of the Mongols and the Chinese have been so intimately interlocked through much of the history of the present millennium that scholars who combine Mongolian studies with Chinese studies make an indispensable contribution. Joseph Fletcher argued forcefully and effectively for an "integrative history" that would include all of Asia, in his essay "Integrative history: Parallels and interconnections in the early modern period 15001800."' Had his untimely death not intervened, Joseph Fletcher would have applied that concept in chapters to be written for this and other volumes of The Cambridge history of China. The ideal of a more closely integrated history of China and Mongolia and other East and Inner Asian peoples is fully accepted by all the chapter authors in this volume. It is, however, an ideal difficult to achieve. A good example of that difficulty in the field of Yuan social history is provided by the work of the great Russian Mongolist B. IA. Vladimirtsov (d. 1931) whose study of nomadic feudalism was published posthumously in 1934.2 This was translated into French by Michel Carsow as La regime social des Mongols: La Fiodalisme nomade,3 into Japanese by a government agency (1941), and from Japanese into Chinese.'' That and Vladimirtsov's life of Chinggis khan' are repeatedly cited by scholars, including many of those mentioned in this chapter. There is good reason to wish to understand the nature of Mongolian society when assessing the impact of Mongol ideas about the social management of other societies that came under their rule. Yet Vladimirtsov's theory of nomadic feudalism, by which he attempted to fit Mongolian history into the same universal pattern that informs his understanding of Russian and Western history, is not without its critics. Note the review by Lawrence Krader, "Feudalism and the Tatar polity of the Middle Ages."6 Although Vladimirtsov's work has induced Chinese and Mongolian, as well as other scholars, to try to apprehend the nature of Mongolian society and to bring their understanding of it to bear on their analyses of social conditions in the Yuan period, that effort has not yet reached a high level of sophistication and usefulness. The field is moving toward "integrative history," but it has a way to go. The largest modern history of the Yuan period in Chinese does indeed include a lengthy chapter on Yuan society; that is Li Tse-fen's Yuan shih hsin chiang.1 This 1 Joseph Fletcher, "Integrative history: Parallels and interconnections in the early modern period 1500— i%oo," Journal of Turkish Studies, 9(1985), pp. 3 7 - 5 7 . 2 B. IA. Vladimirtsov, Obshchestveaayi stray Mongolov: Mongolskii kochevoy feodalizm (Leningrad, 1934). 3 Michel Carsow, trans, and ed.. La Rigime social des Mongols: La Flderalisme nomade (Paris, 1948). 4 Boris IA. Vladimirtsov, Meng-ku she hut chih tu sbib, trans. Chang Hsing-t'ang and Wu Chan-k'un (Taipei, Chung-kuo wen hua ch'u pan shih yeh wei yuan hui, 1957; rpt., 1967). Another translation directly from the Russian original is that by Liu Jung-chun, trans., Meng-ku she hui chih tu shih (Peking, 1980). 5 B. IA. Vladimirtsov, Gengis-kban, trans. Michael Carsow (Paris, 1948). 6 Lawrence Krader, "Feudalism and the Tatar polity of the Middle Ages," Comparative Studies in Society and History, I ( 1 9 5 8 - 9 ) , pp. 7 6 - 9 9 . 7 Li Tse-fen, Yuan shih hsin chiang (Taipei, 1978), vol. 5. Cambridge Histories Online © Cambridge University Press, 2008 BIBLIOGRAPHICAL ESSAYS 723 massive work includes careful investigations of many problems present in the accumulation of Ch'ing and more recent scholarship on Yuan history. It is a noteworthy addition to, and virtually the culmination of, that native tradition and at many points is a most useful work. Nonetheless, the treatment of social history, even in this very large work, is superficial. The study of Chinese social history in the Yuan period has not yet drawn as much attention as have political and institutional, literary, art, and intellectual history. In mainland China the focus has been largely on Yuan social structure, presented in simplistic class analysis and, in particular, on the popular rebellions of the late Yuan. Though often doctrinaire, this intense combing of the documentary evidence and ingenious use of the new archaeological materials have contributed significant new data. An important example of this is the compilation entitled Yuan tat nung min chart chengshih liao buipien (Compilation of historical materials relevant to the peasant wars of the Yuan dynasty).8 Part 1, compiled by Yang Ne and Ch'en Kao-hua, covers the years 1237 to 1350. Part 2, in two volumes, compiled by Yang Ne, Ch'en Kaohua, Chu Kuo-chao, and Liu Yen, covers the rebellions other than that of Chu Yiianchang, in the years 1351 to 1368. Part 3, in one volume, compiled by Yang Ne and Ch'en Kao-hua, is devoted to the rebellion of Chu Yiian-chang and the founding of the Ming dynasty, in the years from 1328 to 1367. This work will greatly facilitate a thoroughgoing reassessment of popular rebellions in the Yuan period, particularly those of the final decades of Yiian rule. A leading figure in Yiian history studies was Han Ju-lin (d. 1986), who was a student of Paul Pelliot in Paris in the 1930s and was thoroughly conversant with Western, including Soviet, scholarship. His impact on modern Chinese historical study of the Yiian can be seen in the two-volume History of the Yiian dynasty,* produced under his general supervision by a group of his associates at the Center for Yiian History Studies at the University of Nanking, which he founded and headed for many years. This work is one of the best modern dynastic histories to have been produced in mainland China. Nonetheless, its sections on social history are fragmented and thin. A collection of Han's writings, entitled Ch'iung-lu chi,la was published in 1982. The state of Yiian history studies in mainland China since 1949 is the subject of an important review essay appended to the volume Yiian shih lun chi (Essays on Yiian history) published by the Center for Yiian History Studies at the University of Nanking in 1984; it includes several important studies relevant to social history. This volume also includes a selective bibliography of articles and books published between 1949 and 1980, classified by subfields." In Taiwan, stimulated by the presence of the late Professor Yao Ts'ung-wu who 8 Yang Ne, Ch'en Kao-hua et al., comps., Yiian tai nung min chan chtng shih liao hui pien, 4 vols. (Peking, 1985). 9 Han Ju-lin, Yiian ch'ao shih (Peking, 1986). 10 Han Ju-lin, Ch'iung lu chi: Yiian shih chi hsi pei min tsu shih yen chiu (Shanghai, 1982). 11 Nan-ching ta hsiieh li-shih hsi, Yiian shih yen chiu shih, comps., Yiian shih lun chi (Peking, 1984), p. 770. Note appendices: "1949—1980 nien te Chung-kuo Yiian shih yen chiu," and "1949—1980 nien pu fen Yiian shih lun wen mu lu." Cambridge Histories Online © Cambridge University Press, 2008 724 BIBLIOGRAPHICAL ESSAYS also studied in Europe (principally in Germany) in the 1930s, a generation of Yuan specialists have been active. Some have gone abroad to gain training in Mongolian studies, a field until recently not widely offered in Chinese universities. Professor Yao's most important writings, published in Tung-pei shih lun-ts'ung12 and Yao Ts'ungwu hsien-sheng ch'iian chi,1* focus on the sinification of steppe peoples, a focus that has been emphasized in Chinese studies of the conquest dynasties, whether justified by quotations from Karl Marx or prompted by nationalistic conviction. This kind of scholarship has gained considerably in sophistication since the publication in the 1920s of Ch'en Yuan's Yuan bsi-yiijen hua hua k'ao, the 1935 revised edition of which has been reprinted in a collection of Ch'en Yuan's writings, Yuan shih yen chiu, •« Ch'ien Hsing-hai and L. Carrington Goodrich published their annotated translation of Ch'en's study in 1966, under the title Western and Central Asians in China under the Mongols: Their transformation into Chinese.I5 Most of Yao Ts'ung-wu's students in Taiwan and abroad have concentrated on political and institutional researches. A senior scholar of Yao's generation greatly influenced by him is Sun K'e-k'uan, one of the most prolific writers in recent decades on Yuan subjects. Sun's work was described and evaluated by John D. Langlois, Jr., in the volume edited by him, China under Mongol rule.l6 Sechin Jagchid also was influenced by Yao; his many writings on Yuan history combine Mongolist and sinologist skills. That ideal is well represented in the work of several contributors to our volume. These scholars' combination of Mongolian and Chinese scholarly qualifications represents one of the field's important and hopeful trends. Few subfields of Yuan social history are well developed, yet there are several in which particularly useful scholarly achievements may be noted. The potentially high value of the Chinese pi-chi (collections of scholars' notes and sketches) was pointed out by Herbert Franke, sinologue and Mongolist at the University of Munich who studied under Erich Haenisch. Franke has written on an extraordinarily broad range of periods and topics; see the Festschrift for his sixty-fifth birthday, Studia Sino-Mongolica. •' Franke presented the first and still the only published integral translation and annotation of a Yuan period pi-chi in his Beitrage zur kulturgeschichte Chinas unter der Mongolenberrschaft: Das Shan-kit sin-hua des Yang Yii.l8 He also published a general introduction to the genre, pointing out its usefulness, above all, for social history; see his "Some aspects of Chinese private historiography 12 13 14 15 Yao Ts'ung-wu, Tung-pei shih lun tt'»»g (Taipei, 1959). Yao Ts'ung-wu, Yao Ts'ung-wu hsitn shaig ch'iian cbi (Taipei, 1971—82). Ch'en Yuan, Yiian Hsi-yii jen hua hua k'ao (rev. ed. 1935); repr. in his Yiiansbihyen chiu (Taipei, 1977). Ch'en Yuan, Western and Central Asians in China under the Mongols: Their transformation into Chinese, trans. Ch'ien Hsing-hai and L. Carrington Goodrich, Monumenta Serica Monograph no. 15 (Los Angeles, 1966). 16 John D. Langlois, Jr., ed., China under Mongol rule (Princeton, 1981), pp. 2 1 2 - 5 3 . 17 See Wolfgang Bauer, ed., Studia Sino-Mongolica, Festschrift fur Herbert Franke, Miinchener Ostasiatische Studien no. 25 (Wiesbaden, 1979); and the bibliography of Frankc's writings, pp. 4 ) 1 - 7 0 . 18 Herbert Franke, Beitrage zur Kulturgescbicbte Chinas unter der Mongolenherrschaft: Das Shan-kii sin-hua des Yang Yii, Abhandlungen fur die Kunde des Morgenlandes, vol. 32, no. 2 (Wiesbaden, 19)6). Cambridge Histories Online © Cambridge University Press, 2008 BIBLIOGRAPHICAL ESSAYS 725 in the thirteenth and fourteenth centuries," in Historians of China and Japan, ed. William G. Beasley and E. G. Pulleyblank." Modern studies of the Yuan system of social classes were initiated by the preeminent Japanese historian of the Yuan period, Yanai Wataru (1875—1926), in a work known in Chinese translation as Yuan tai Meng-Han se-mu tai-yii k'ao,™ published in the mid-i93OS in a translation by Ch'en Ch'ing-ch'iian. I have not located a copy of the Japanese original edition, published first in 1916. A number of Yanai's outstanding scholarly works on Yuan period and Mongolian institutional history were published in Chinese translations in the 1930s. Yanai's work on the social classes, however, apparently was not known to Meng Ssu-ming when he wrote his extensive study Yiian tai she hui chieh chi chih tu (Social classes in China under the Yuan dynasty).21 Meng's monumental work may be considered the most important publication on Yiian social history for almost half a century, until the 1970s when several works significantly supplementing it began to appear. Also of importance from the 1930s is Chii Ch'ing-yiian's pathbreaking study of artisans in the Yiian, well known in the abridged English translation appearing in Chinese social history21 in 1956. Among the more recent works significantly supplementing the scholarship on Yiian social history are publications by Hsiao Ch'i-ch'ing, a student of Yao Ts'ungwu, Sechen Jagchid, and Francis W. Cleaves; his collected studies, called Yiian tai shih hsin t'an (New investigations into Yiian history),2* appeared in 1983. His student Huang Ch'ing-lien's monograph on the Yiian household registration system, Yiian tai hu chi chih tu yen chiu,'« was published in 1977. A number of smaller studies bearing on the social history of the period also have appeared. The time has come for a new synthesis of this field. A step toward that, albeit a preliminary one, is the book Yiian tai she hui ching chi shih kao (Draft economic and social history of the Yiian dynasty)2' by Li Kan, published in 1985. Hok-lam Chan has been the principal contributor to studies of popular beliefs and folklore in the Chin, Yiian, and Ming periods (as well as having produced many monographs and articles on other Yiian history subjects). His article "The White 19 Herbert Franke, "Some aspects of Chinese private historiography in the thirteenth and fourteenth centuries," in Historians of China and Japan, ed. William G. Beaseley and Edwin G. Pulleyblank (London, 1961), pp. 115—34. 20 Yanai wataru, "Gendai shakai san kaikyu (shikimoku ko)," Man-Sen chiri reikisbi kenkyu hokoku (December 1916); trans, into Chinese by Ch'en Chieh as Yiian tai Meng-Han se-mu tai-yii k'ao (Shanghai, n.d.; repr. Taipei, 1963). 21 Meng Ssu-ming, Yiian tai she hui chieh chi chih tu, Yenching Journal 0/Chinese Studies, Monograph series no. 16(1938; repr. Hong Kong, 1967). 22 Chii Ch'ing-yiian, "Yiian tai hsi kuan Chiang hu yen chiu: Chih jen wei yuan tai kuan chii chiang hu shih nu li te jen men," Shih huoyiieh i'an, 1 (1935), pp. 367—401. An abridged English translation appears in Chinese social history: Translations of selected studies, trans. John de Francis and E-tu Zen Sun (Washington, D.C., 1956), pp. 234-46. 23 Hsiao Ch'i-ch'ing, Yiian tai shih hsin t'an (Taipei, 1983). 24 Huang Ch'ing-lien, Yuan tai hu chi chih tu yen chiu (Taipei, 1977). 25 Li Kan, Yiian tai she hui ching chi shih kao (Wuhan, 1985). Cambridge Histories Online © Cambridge University Press, 2008 726 BIBLIOGRAPHICAL ESSAYS Lotus—Maitreya doctrine and popular uprisings in Ming and Ch'ing China"26 deals importantly with popular religion in the late Yuan as an element in the rebellions out of which the following dynasty was founded. This subject has been pursued most notably in mainland China at the Center for Yuan History Studies at the University of Nanking, where the scholar most active in this subject area has been Ch'iu Shushen. The center publishes two journals of special interest: Yuan shih chi pet fang min tsu shih yen chiu chi k'an (Quarterly for researches on Yuan history and northern border peoples history) and Yuan shih lun ts'ung (Collected essays on Yuan history). The other designated Center for Yuan History Studies in the People's Republic is located at the University of Inner Mongolia, Huhehot. The leading scholars in the field of Yuan history at the Institute of History, Chinese Academy for the Social Sciences (Peking), are Ch'en Kao-hua, who writes on social and institutional subjects, and Chou Liang-hsiao, who writes on political and pre-Yiian Mongolian history. Elizabeth Endicott-West's research on Yuan local government2' and her study of the ortogh merchant associations in pre-Yuan and Yuan times, along with that by Thomas T. Allsen, have important implications for Yuan social history.28 Morris Rossabi's new biography of Khubilai khan (1988) adds significantly to our knowledge of the reign of the greatest of the Yuan rulers, during which much of the social legislation was implemented and regularized. Rossabi's volume and his article "Khubilai khan and the women in his family"29 reveal the social history of the Yuan court and imperial household and describe the greatly different social styles of Mongolian and Chinese elites at that time. 26 Hok-lam Chan, "The White Lotus—Maitreya doctrine and popular uprisings in Ming and Ch'ing China," Sinologica, 10 (1968—9), pp. 211-33. 27 Elizabeth Endicott-West, Mongolian rule in China: Local administration in the Yuan dynasty (Cambridge, Mass., 1989). 28 Thomas T. Allsen, "Mongolian princes and their merchant partners, 1200—1260," Asia Major, 3rd series, 2 (1989), pt. 2, pp. 83—126; and Elizabeth Endicott-West, "Merchant associations in Yuan China: The ortogh," Asia Major, 3rd series, 2 (1989), pt. 2, pp. 127-54. 29 Morris Rossabi, "Khubilai khan and the women in his family," in Studia Sino-Mongolica: Festschrift fur Herbert Franke, ed. Wolfgang Bauer, Miinchener Ostasiatische Scudien no. 2 ; (Wiesbaden, 1979), pp. 153-80. Cambridge Histories Online © Cambridge University Press, 2008 BIBLIOGRAPHY Abe Takeo $ SB 11 * . "Daigen tsusei no kaisetsu ~k 7C j® ffl CO f% Ift ." In Abe Takeo, Gendaishi no kenkyii TC iX i CO 5JF 5E Tokyo: Sobunsha, 1972, pp. 253-319. Abe Takeo. "Gendai chishikijin to kakyo 7C iX &J $ A i ?4 # . " I n Abe Takeo, Gendaishi no kenkyii 7C ft !fe CO 5ff 2E . Tokyo: Sobunsha, 1972, pp. 3 - 5 3 . Abe Takeo. "Where was the capital of the West Uighurs?" In the Silver jubilee volume of the Zimbun kagaku kenkyusho. Kyoto: Ky5to daigaku jimbun kagaku kenkyujo, 1954, pp. 4 3 5 - 5 0 . Abe Yukihiro ffl 95 ffi % . Moko shurai M~^^k%. Tokyo: Kyoikusha, 1980. Aida Niro til ffl — %. Moko shurai no kenkyii M S S 5K COffl% . Tokyo: Yoshikawa kobunkan, 1971. Allsen, Thomas T. "Guard and government in the reign of the grand Qan Mbngke." HarvardJournal of Asiatic Studies, 46(1986), pp. 500—21. Allsen, Thomas T. "Mahmud Yalavach." In Yuan personalities, ed. Igor de Rachewiltz and Hok-lam Chan. Wiesbaden: Otto Harrassowitz, forthcoming. Allsen, Thomas T. "Mongol census taking in Rus', 1245—1275." Harvard Ukrainian Studies, 5 (1981), pp. 3 2 - 5 3 . Allsen, Thomas T. Mongol imperialism: The policies of the Grand Qan Mbngke in China, Russia, and the Islamic lands, 1251—1259. Berkeley and Los Angeles: University of California Press, 1987. Allsen, Thomas T. "Mongolian princes and their merchant partners, 1200-1260." Asia Major, 3rd series, 2 (1989), pt. 2, pp. 8 3 - 1 2 6 . Allsen, Thomas T. The Mongols in East Asia, twelfth—fourteenth centuries: A preliminary bibliography of books and articles in Western languages. Sung Studies Research Aids no. 1. Philadelphia: Sung Studies Newsletter, 1976. Allsen, Thomas T. "Prelude to the western campaigns: Mongol military operations in the Volga—Ural region, 1217-1237." Archivum Eurasiae Medii Aevi, 3 (1983), pp. 5-24. Allsen, Thomas T. "The Yuan dynasty and the Uighurs of Turfan in the 13th century." In China among equals: The Middle Kingdom and its neighbors, 10th-14th centuries, ed. Morris Rossabi. Berkeley and Los Angeles: University of California Press, 1983, pp. 243-80. Al-Qashani. See Kashani. Ang, Melvin Thlick-len. "Sung-Liao diplomacy in eleventh- and twelfth-century 727 Cambridge Histories Online © Cambridge University Press, 2008 728 BIBLIOGRAPHY China: A study of the social and political determinants of foreign policy." Ph.D. diss., University of Pennsylvania, 1983. Anonymous. Hsien t'ai t'ung chi 58 S ffi & . Included in Yung-lo ta tien ac 9k ~K ift , chiian 2608—9. Peking: Chung-hua shu chii, i960. Anonymous. Nan t'ai pei yao M i3 fit W . Included in Yung-lo ta tien *K $& ^C #i , chiian 2610—11. Peking: Chung-hua shu chii, i960. Anonymous. Ta Yiian hai yiin chi ^ 7C M M 15 . 2 Chiian. Ed. and reconstructed by Hu Ching $ft W.. Hsiieh-t'ang ts'ung ko 1 S S I J . Repr. in Shih liao ts'ungpien, ssu pien i. #4 ^ H H H . Taipei: Kuang-wen shu chii, 1972. Anonymous. Ta Yiian ma cheng chi ~fc ?E MI 3fc 13 . Repr. Shih liao ts'ungpien, ssu pien $L Pi H M ' E3 S . Taipei: Kuang wen shu chii, 1972. Also Kuo-hsiieh wen-k'u, 1937 ed. Anonymous. Ta Yiian ts'ang k'u chi 'X 7C M M IB . Repr. Shih liao ts'ungpien, ssu pien £ ft M M ' E3 £§ . Taipei: Kuang wen shu chii, 1972. Also Kuo-hsiieh wen-k'u, 1936 ed. Anonymous, Yiian Kao-li chi shih ft B H & V . Repr. Shih-liao ts'ungpien, ssu pien £ 14 % M E3 H . Taipei: Kuang wen shu chii, 1972. Aoyama Koryo W UJ & % . Gencho shoshosho ko 7C la (Si # ^ # . Tokyo: Meiji daigaku bungaku kenkyujo, 1951. Aoyama Koryo. "Rekidai kodai ko M ft fr M % ." Taihoku teikoku daigaku bunsei gakubushigaku ka kenkyu nempo X it » PI * 9 X ft PSB i t 9 $4 W * ¥ tt , 2 (1935). PP- 143-66. Aritaka Iwao W iS5 J8 . "Gendai no nomin seikatsu ni tsuite I * X •" 1° Kuwabara Hakushi kanreki kinen Toyoshi ronso # £ H H , ed. Kuwabara Hakushi kanreki kinen shukugakai H H If ^ l i S f . Kyoto: Kobundo, 1934, pp. 945~97'Ata Malik Juvaynl. TAe ^M/ory of the world conqueror. 2 vols. Trans. John A. Boyle. Cambridge, Mass.: Harvard University Press, 1958. 'Ata Malik Juvayni. Tartkh-i jahan gushd. 3 vols. Ed. MIrza Muhammad Qazvlni. London: Luzac, 1912—37. Aubin, Franchise. "G£ographie administrative et defense nationale en Chine: L'Exemple des dernieres annees des Chin (Kin)." In Studia Sino-Mongolica: Festschrift fur Herbert Franke, ed. Wolfgang Bauer. Miinchener Ostasiatische Studien no. 25. Wiesbaden: Franz Steiner, 1979, pp. 83—8. Aubin, Franchise. "The rebirth of Chinese rule in times of trouble: North China in the early thirteenth century." In Foundations and limits of state power in China, ed. Stuart R. Schram. London: School of Oriental and African Studies, University of London, and Hong Kong: Chinese University Press, 1987, pp. 113-46. Ayers, John. "Some characteristic wares of the Yiian dynasty." Transactions of the Oriental Ceramic Society, 29(1957), pp. 69-86. Backus, Charles. The Nan-chao kingdom and T'ang China's southwestern frontier. Cambridge: Cambridge University Press, 1981. Bacon, Elizabeth E. Obok: A study of social structure in Eurasia. Wenner—Gren Founda- Cambridge Histories Online © Cambridge University Press, 2008 BIBLIOGRAPHY 729 tion for Anthropological Research, Publications in Anthropology no. 25. New York: Wenner—Gren Foundation for Anthropological Research, 1958. Barfield, Thomas J. "The Hsiung-nu imperial confederation: Organization and foreign policy." Journal of Asian Studies, 41 (1981), pp. 45—61. Barfield, Thomas J. The perilous frontier: Nomadic empires and China. Ed. Charles Tilly. Oxford: Basil Blackwell, 1989. Barthold, Wilhelm (Vasilii V. Bartol'd). Four studies on the history of Central Asia. Trans. V. Minorsky and T. Minorsky. 3 vols. Leiden: Brill, 1956-62. Barthold, Wilhelm (Vasilii V. Bartol'd). Istoriia Turkestana. Repr. in Wilhelm Barthold, Sochineniia, vol. 2, pt. 1. Moscow: Izdatel'stvo Vostochnoi Literatury, 1963. Barthold, Wilhelm (Vasilii V. Bartol'd). Turkestan down to the Mongol invasion. Trans. T. Minorsky. 3rd ed. E. J. W. Gibb Memorial Series, n.s., no. 5. London: Luzac, 1968. Bauer, Wolfgang. Der chinesische Personenname: Die Bildungsgesetze and hauptdchlichsten Bedeutungsinhalte von Ming, Tzu und Hsiao-ming. Asiatische Forschungen. Mono- graphienreihe zur Geschichte, Kultur, und Sprache der Volker Ost-und Zentralasiens no. 4. Wiesbaden: Otto Harrassowitz, 1959. Bauer, Wolfgang, ed. Studia Sino-Mongolica: Festschrift fur Herbert Franke. Miinchener Ostasiatische Studien no. 25. Wiesbaden: Franz Steiner, 1979. Bawden, Charles R. The Mongol Chronicle Allan Tobci. Wiesbaden: Otto Harrassowitz, 1955. Bawden, C. R., and S. Jagchid. "Some notes on the horse policy of the Yuan dynasty." Central Asiatic Journal, 10 (1965), pp. 246—68. Boodberg, Peter. "Dayan, Cinggis, and Shan-yii." In Selected works of Peter A. Boodberg, comp. Alvin P. Cohen. Berkeley and Los Angeles: University of California Press, 1979, pp. 8 5 - 9 . Boyle, John A. "The burial place of the great khan Ogedei." Ada Orientalia, 32 (1970), pp. 45-5OBoyle, John A. "Dynastic and political history of the II Khans." In The Saljuq and Mongol periods. Vol. 5 of The Cambridge history of Iran, ed. John A. Boyle. Cam- bridge: Cambridge University Press, 1968, pp. 303-421. Boyle, John A. "The Journey of Het'um, king of Little Armenia, to the court of the great khan Mongke." Central Asiatic Journal, 9 (1964), pp. 175—89. Boyle, John A . , trans. The successors of Genghis Khan: Translated from the Persian of Rashtdal-Dtn. New York: Columbia University Press, 1971. Brown, William A. Wen T'ien-hsiang: A biographical study of a Sung patriot. San Francisco: Chinese Materials Center Publications, 1986. Budge, E. A. Wallis, trans. The chronography of Gregory Abu7 Faraj the son of Aaron, the Hebrew physician commonly known as Bar Hebraeus. 2 vols. Oxford: Oxford University Press, 1932. Budge, E. A. Wallis, trans. The monks of Kublai khan, emperor of China. London: Religious Tract Society, 1928. Buell, Paul D. "The role of the Sino-Mongolian frontier zone in the rise of Cinggis Cambridge Histories Online © Cambridge University Press, 2008 73° BIBLIOGRAPHY Q a n . " In Studies on Mongolia: Proceedings of the first North American conference on Mongolian studies, ed. Henry G. Schwartz. Bellingham: Center for East Asian Studies, Western Washington University, 1979, pp. 63—76. Buell, Paul D. "Sino-Khitan administration in Mongol Bukhara." Journal of Asian History, 13 (1979), pp. 121-51. Bush, Susan. "Literati culture under the Chin (1122—1234)." Oriental Art, n.s., vol. 15 (1969). PP- 103-12. Cahill, James. Hills beyond a river: Chinese painting of the Yuan dynasty. New York: John Weatherhill, 1976. Cambridge history of China. See Twitchett, Denis C., and John K. Fairbank. Centre d'etudes sinologiques de Pekin, comp. Ta Chin kuo chih t'ung chien ^. & H SS SH& (.Index du Ta Kin Kouo Tche). Peking, 1949; repr. Hong Kong: Lung-men shu-tien, 1967. Centre franco-chinois d'etudes sinologiques, comp. Ch'i-tan kuo chih t'ung chien §§ fflIH /£ ffi ^ . Chung-Fa Han hsiieh yen chiu so t'ung chien ts'ung k'an 41 & M ^ fiff 3E ffc MtifcM fJ no. 12. Peking, 1949; repr. Taipei: Ch'eng-wen ch'u-pan she, 1968. Cha-ch'i Ssu-ch'in +L ^f Wi $K (Sechin Jagchid). Meng-kupi shih hsin iping chu shih 18 l5"lft£$ft?jKit!:S . Taipei: Lien-ching ch'u-pan shih-yeh kung-ssu, 1979. Cha-ch'i Ssu-ch'in (Sechin Jagchid). Meng-ku yii Hsi-tsang li shih kuan hsi chih yen chiu WS $$'fcW.ffi.%.M%2.W<% . Taipei: Cheng-chung shu-chu, 1978. Cha-ch'i Ssu-ch'in (Sechin Jagchid). "Shuo Yuan shih chung te cha lu hua ch'ih ping chien lun Yuan ch'u te shang shu sheng IftTni + f f t f L ^ f t ^ ^ S f t l t T c S J W f^ It *M ."In vol. 1 of Cha-ch'i Ssu-ch'in, Meng-ku shih lun ts'ung W is £ It Wi . Taipei: Hsiieh-hai shu-chii, 1980, pp. 233—363. Cha-ch'i Ssu-ch'in (Sechin Jagchid). "Yuan tai Chung-tu k'ao 7C f£ •£ IP % ." Pien cheng yen chiu so nien pao 19 &ffl% #f ¥ 18 , 18(1987), pp. 31-41. Chan, Hok-lam (Ch'en Hsiieh-lin). "Chinese official historiography at the Yuan court: The composition of the Liao, Chin, and Sung histories." In China under Mongol rule, ed. John D. Langlois, Jr. Princeton, NJ.: Princeton University Press, 1981, pp. 56-106. Ch'an, Hok-lam (Ch'en Hsiieh-lin). The historiography of the Chin dynasty: Three studies. Wiesbaden: Franz Steiner, 1970. Chan, Hok-lam (Ch'en Hsiieh-lin). Legitimation in imperial China: Discussions under the Jurchen—Chin dynasty (1115-1234). Seattle: University of Washington Press, 1984. Chan, Hok-lam (Ch'en Hsiieh-lin). "Liu Ping-chung §!l S-IS (1216-74): A Buddhist-Taoist statesman at the court of Khubilai khan." T'oung Pao, 53 (1967), pp. 98-146. Chan, Hok-lam (Ch'en Hsiieh-lin). "Prolegomena to the Ju-nan i shih: A memoir of the last Chin court under the Mongol siege of 1234." Sung Studies Newsletter 10, suppl. 1 (1974). PP- 2-19. Chan, Hok-lam (Ch'en Hsiieh-lin). "Tea production and tea trade under the Jurchen—Chin dynasty." In Studia Sino-Mongolica: Festschrift fur Herbert Franke, ed. Cambridge Histories Online © Cambridge University Press, 2008 BIBLIOGRAPHY 731 Wolfgang Bauer. Miinchener Ostasiatische Studien no. 25. Wiesbaden: Franz Steiner, 1979, pp. 104-25. Chan, Hok-lam (Ch'en Hsiieh-lin). "Wang O (1190-1273)." Papers on Far Eastern History, 12 (1975), pp. 43-70. Chan, Hok-lam (Ch'en Hsiieh-lin). "The White Lotus—Maitreya doctrine and popular uprisings in Ming and Ch'ing China." Sinologka, 10 (1968—9), pp. 211—33. Chan, Hok-lam (Ch'en Hsiieh-lin). "Yao Shu (1201—1278)." Papers on Far Eastern History, 22 (1980), pp. 17-50. Chan, Hok-lam (Ch'en Hsiieh-lin), and William Theodore de Bary, eds. Yiian thought: Chinese thought and religion under the Mongols. New York: Columbia Univer- sity Press, 1982. Chan, Wing-tsit. "Chu Hsi and Yiian Neo-Confucianism." In Yiian thought: Chinese thought and religion under the Mongols, ed. Hok-lam Ch'an and William Theodore de Bary. New York: Columbia University Press, 1982, pp. 197—231. Chang Cheng-ming 5R IE BM . Ch'i-tan shih liieh % £r $L B§ . Peking: Chung-hua shuchii, 1979. Chang Chin-wu 3fi ^ H , comp. Chin wen tsui 4 X 1 . Soochow, 1895; repr. Taipei: Ch'eng-wen ch'u-pan she, 1967. Chang Ning 31 ^ . "Chi Yuan Ta-tu ch'u t'u wen wu t B T C ^ f f l t f ^ X ^ . " K'ao ku, 1972:6, pp. 25-34. Chang Po-ch'uan 3S W H . Chin tai ching chi shih liieh £ ft 3£ if £ Bg . Shenyang: Liao-ning jen-min ch'u-pan she, 1981. Chang T'ing-yii I g l et al., eds., Ming shih %M $. . 28 vols. Peking: Chung-hua shu-chii, 1974 ed. Chang Wei 3g i t . Ta Chin chi li * £ * a . TSCC ed. Chang Yang-hao 5g £ jg . Kuei t'ien lei kao SSffl& 4$ . SKCS ed. Chang Yang-hao. Mu min chung kao ft S JS ff . In Wei cheng chung kao fe £& 5& •£ . SPTKed. Chang Yuan-chi 51 7E iS . CArao shih sui pi R J i f . Taipei: Shang-wu yin-shu kuan, 1967. Ch'ang-ku-chen-i I @ I i . Nung t'ien yii hua 8kfflgfc IS . T^CC ed. Ch'ang Pi-te § & f§ , Wang Te-i I S! tg , Ch'eng Yiian-min S 7C @ , and Hou Chiin-te ^ ®: ^ , comps. Sungjen chuan chi tzu liao so yin ^ A f f i l H ^ ^ f ^ ? ! ^ vols. Taipei: Ting-wen shu-chii, 1974—1976. Chao Hung f& 8* . Meng Ta pei lu W. IS M & . In Meng-ku shih liao ssu chung f S $ ^4 ES W , ed. Wang Kuo-wei X 13 H . Peking, 1926; repr. Taipei: Cheng-chung shu-chii, 1962, 1975. Chao I jg M • Nien erh shih cha chi It ~ £ 15 IB . Ed. Tu Wei-yiin tt * S . Nien erh shih cha chi chi pu pirn i t — j& HI IB &?$£§• Taipei: Ting-wen shu-chu, 1975. Chao Kang. Ala» <j«</ /ana' /'« Chinese history: An economic analysis. Stanford, Calif.: Stanford University Press, 1986. Chao K'ang-min S M £§ , Han Wei $$fiS, et al. "Kuan yii Shan-hsi Lin-t'ung ch'u t'u te Chin tai shui yin te chi ko wen t'i H St 1*5 S ffi S Hi ± W ^ ft ft iS W « 1975:8, pp. 73-81. Cambridge Histories Online © Cambridge University Press, 2008 732 BIBLIOGRAPHY Chao Meng-fu 21 i£ M . Chao Meng-fu chi ffi i£ M %.. Ed. Jen Tao-pin ffi i t St . Hangchou: Che-chiang ku chi ch'u pan she, 1986. Chavannes, Edouard. "Inscriptions et pieces de chancellerie chinoises de l'epoque mongole." Toung Pao, 5 (1904), pp. 357-447; 6 (1905), pp. 1-42; 9 (1908), pp. 297-428. Chavannes, Edouard. "Pei Yuan lou i t $8 IS : Ricit d'un voyage dans le Nord par Tcheou Chan JH $? ." Toung Pao, 5 (1904), pp. 162-92. Chavannes, Edouard. Review of A. I. Ivanov, "Stranitsa iz istorij Si-sia (Une page de l'histoire du Si-hia; Bulletin de l'Academie imperiale des sciences de SaintPetersbourg, 1911, pp. 831-836)." T'oungPao, 12 (1911), pp. 441-6. Chavannes, Edouard. "Voyageurs Chinois chez les Khitan et les Joutchen," pt. 1. Journal Asiatique, 9th series, no. 9 (1897), pp. 377—442; pt. 2. Journal Asiatique, 9th series, no. 11 (1898), pp. 361-439. Ch'en Ch'ing-ying K S 5 I . "Yiian ch'ao tsai Hsi-tsang so feng Pai-lan wang 76 $9 ft HHfifffef6S8i ." Hsi-tsang yen chiu B Mffi % , 1983:4, pp. 29-32. Ch'en Kao-hua W ift M . "Lun Yiian tai te chiin hu fi % fi fft M P ." Yiian shih lun ts'ung 7C £ H if , 1 (1982), pp. 72-90. Ch'en Kao-hua. Yiian Ta-tu 7C ;*C 135 . Peking: Pei-ching ch'u pan she, 1982. Japanese ed.; trans. Satake Yasuhiko ft TC 4f M . Gen no Taito: Maruko Porojidai no Pekin TCCO^tP: T ; l / 3 ^ D C O i t ^ . Tokyo: ChuokSronsha, 1984. Ch'en Kao-hua, "Yiian tai ch'ien ch'i ho chung ch'i ko tsu jen min ch'i i tou cheng 7t ttwiMfQ'Pffl&teAg.&mn^."In vol.2 of Chung-kuo nung min chan cheng shih lun ts'ung + S , f t g i & ? - 5 f e f ! l ? , ed. Lu Shu-ch'ing B ffl M . Honan: Honan jen-min ch'u-pan she, 1980, pp. 286—320. Ch'en Kao-hua. "Yuan tai i fa chien lun 7Uft'<&.&ffi?I t . Wen shih X £., 11 (1980), PP- 157-73Ch'en Kao-hua. "Yiian tai yen cheng chi ch'i she hui ying hsiang 7C ft H ife 2& ^ tt # ^ W ." Li shih lun ts'ung, 1 (1964), pp. 175—217. Ch'en, Kenneth. Buddhism in China: A historical survey. Princeton, N.J.: Princeton University Press, 1964. Ch'en, Li-li. See Tung Chieh-yiian. Ch'en Ming-ta W BJ i t . Ying-hsien mu t'a M U * if . Peking: Wen-wu ch'u-pan she, 1980. Ch'en Pang-chan i$£ % S® , comp. Sung shih chi shih pen mo ^ £ ^ ^ ^ 7 ^ . 3 vols. Peking: Chung-hua shu-chii, 1977. Ch'en Pang-chan, comp. Yiian shih chi shih pen mo 7C £ ffl ~9> $ % . Kiangsi: Chiang-hsi shu-chii, 1874. Peking: Chung-hua shu-chii, 1955; repr. 1979. Ch'en, Paul Heng-chao. Chinese legal tradition under the Mongols: The code of 1291 as reconstructed. Princeton, N.J.: Princeton University Press, 1979. Ch'en Ping-ying ffl. ffi M . Hit Hsia wen tuu yen chiu H JE X %3 W % . Yin-ch'uan: Ning-hsia jen min ch'u pan she, 1985. Ch'en Shih-sung 88 1t Jfi . Meng-ku ting Shu shih kao I S f i J i . Ch'eng-tu: Ssu-ch'uan she hui k'o hsiieh ch'u pan she, 1985. Ch'en Shih-sung. "Shih lun Yiian tai chung ch'i te shao shu min tsu ch'i i 1& l i 7C Cambridge Histories Online © Cambridge University Press, 2008 BIBLIOGRAPHY 733 ."In Yuan shih lun chi 7t £ H US , ed. Nan-ching ta . hsueh li shih hsi Yuan shih yen chiu shih i$ 3JC * ^ HI £ 3* • 7t£.m9cM Peking: Jen-min ch'u-pan she, 1984, pp. 565—82. Ch'en Shu W- $t. Chin shih shih pu wu chung 3z $1 fo ffl £ S . Peking: K'o-hsiieh ch'u-pan she, i960. Ch'en Shu. Ch'i-tan cheng chih shih kao !?3 ft K ?n 5£ SI . Peking: Jen-min ch'u-pan she, 1986. Ch'en Shu. Ch'i-tan she hui ching chi shih kao K ft *fc # S ^ £ 1ffi . Peking: Shenghuo tu-shu hsin-chih san-lien shu-tien, 1963; repr. 1978. Ch'en Shu, ed. Ch'uan Liaowen I S i . Peking: Chung-hua shu-chii, 1982. Ch'en Shu. "Ha-la Ch'i-tan shuo - chien lun T'o-pa kai hsing ho Yuan tai Ch'ing tai chiu, 1956:2, pp. 67-77. Ch'en Shu. Liao tai shih hua i§ f^ $1 IS. Cheng-chou: Ho-nan jen-min ch'u-pan she, 1981. Ch'en Yuan W. M . Nan Sung ch'u Ho-pei hsin tao chiao k'ao Hi^WMifclfrilifc^ . Fu-jen ta hsueh ts'ung shu i t ^ i l l no. 8. Peking, 1941. Peking: K'ohsiieh ch'u-pan she, 1958; repr. Peking: Chung-hua shu-chii 1962. Ch'en Yuan. Western and Central Asians in China under the Mongols: Their transformation into Chinese. Trans. Ch'ien Hsing-hai and L. Carrington Goodrich. Monumenta Serica Monograph no. 15. Berkeley and Los Angeles: University of California Press, 1966. Ch'en Yuan. Yuan Hsi-yiijen Hua hua k'ao 7U B§ %& A ij§ ft % . Rev. ed., 1935; repr. in Ch'en Yuan, Yuan shih yen chiu 5t £fiTF5E • Taipei: Chiu-ssu ch'u-pan she, 1977Ch'en Yiian-ching ffi 7t ffl, , comp. Shih I'm kuang chi ^ # @f IB . Chien-an: Ch'un chuang shu yuan, 1330-3; repr. Peking: Chung-hua shu-chii, 1963. Cheng Ch'iao Iff ffi . Tung chih % i£. KSCPTS ed. Cheng Ho-sheng Iff © S . "Ch'ing ju tui yii Yuan shih hsueh chih yen chiu ft ffi Si St 7C £ ^ £ ffl % ." Shih ti hsuehpao £ Jfe ^ WL 3, no. 4 (December 1942), pp. 1-23; and 3, no. 5 (January 1925), pp. 1-22. Cheng Shao-tsung ffl %B ^ and Wang Ching-ju £ S? SP . "Pao-ting ch'u t'u Ming tai Hsi Hsia wen shih ch'uang <S5£ til ±W f ^ B K 5 C 5 d ." ATW i» -6J»^ /><7O, 1977:1, pp. I33-4ICh'eng Cho % ^ . Shih Chin lu {£ ^ ^ . Pi-lin-lang kuan ts'ung shu 8 *ft 35 II S? S ed. Ch'eng Chii-fu g ffi ^ . Hsiieh-lou chi i i i . H«-/>« Aj/e» f/&e«^ / iA« 88 Jt 5t IE Jt S e d . Also SKCS ed. Ch'eng Kuang-yii § ife IS . Sung T'ai-tsung tui Liao chan cheng k'ao ^ ^ ^ S£f S ® ^ % . Taipei: Shang-wu yin-shu kuan, 1972. Chi Yiin ffi BS et al., comps. Ssu k'u ch'uan shu tsung mu t'iyao E S K ^ t t S S B I l l l . 4 vols. Shanghai, 1931; repr. Shanghai: Shang-wu yin-shu kuan, 1934. Chia Ching-yen g l l . "T'an-ma-ch'ih chiin k'ao 8= H ffc W % . Y»<»» J^/A /«» 7C i ^ S , 2 (1983), pp. 23-42. Cambridge Histories Online © Cambridge University Press, 2008 734 BIBLIOGRAPHY Chiang Fu-ts'ung M'<&M. Sungshihhsint'an%$.$\W. chii, 1966. Taipei: Cheng-chungshu- Chiang I-han i l — i§j. Yuan tai K'uei-cbang ko chi K'uei-chang jen wu Jt Vt, i l M 18 S. ^ i|t A $J . Taipei: Lien-ching ch'u-pan shih-yeh kung-ssu, 1981. Ch'ien Ta-hsin & ~X Wf . Ch'ien yen fang chin shih wen pa wei if ffi ^ £ 5 % WL M. . Chang-sha: Lu shih chia shu, n.d. Ch'ien Ta-hsin. Ch'ien yen t'ang ch'iian shu J!t5f!5£i£i§ . 1806. 3rd rev. and expanded ed. Changsha: Lung shih chia shu, 1884. Ch'ien Ta-hsin. Ch'ien yen t'ang wen chi W W JS! 3C M • 1884; repr. Shanghai: Shang-wu yin-shu kuan, 1929. Ch'ien Ta-hsin. Nien erh shih k'ao i it — £. # S . Pref. 1782. Peking: Shang-wu yin-shu kuan, 1959 ed. Ch'ien Ta-hsin. [Pu] Yuan shih i wen chib\M\ 7t £ H X ;& . Pref. 1791. Kiangsu: Chiang-su shu-chii, 1874. Ch'ien Ta-hsin. Shih chia chaiyang hsin lu + M IS H Sf SS . Pref. 1799. Shanghai: Shang-wu yin-shu kuan, 1935; repr. 1957. Ch'ien Ta-hsin. Yuan shih shih tsupiao Jt $. Sz M S§ . Pref. 1791. Kiangsu: Chiang-su shu-chii, 1874. Ch'i-tan kuo chih R ¥r M SC . See Yeh Lung-li. Ch'i-tan kuo chih t'ung chien §£ ¥i ® ^ M Ife . 5ee Centre franco-chinois d'6tudes sinologiques. Chin Ch'i-tsung & IS Br». Nii-chen wen tz'u tien ic M 3t ? ft . Peking: Wen wu ch'u pan she, 1984. Chin Kuang-p'ing j£ ^t W- and Chin Ch'i-tsung ^ ftl ffi . Nii-chen yii yen wen tzu yen chiu ^ c K f g a ' X ^ W ^ E . Peking: Wen-wu ch'u-pan she, 1980. Chin shih & 5& . See T'o-t'o. Chin Wei-hsien i l l . Ch'i-tan U tungpei cheng ts'e % fj W * i t Be ® . Taipei: Hua-shih ch'u-pan she, 1981. Chin Yii-fu i S2 SK . Liao lingshih k'o chi lu IS K 5 ?!l M ^ . 1934; repr. as vol. 1 of LMO Chin Yuan yii wen chin ts'un lu J8iiz7Cti;&;{S#&. Taipei: T'ai-lien kuofeng ch'u-pan she, 1974. Ch'in ting Chin shih yii chieh ft S i ^ l§ ffl . Vols. 3 and 4 of C/6'/» /i»g Liao Chin Yuan san shih yii chieh tKS8!5fe7C;H3£§§ftip . Kiangsu: Chiang-su shu-chii, 1878. Ch'ing-ko-erh-t'ai S? fe W ^ {Chinggeltai], Ch'en Nai-hsiung Bg 75 tt, Hsing Fu-li M tS ffl , Liu Feng-chu 31A $ , and Yu Pao-lin ^ flf R . Ch'i-tan hsiao tzu yen chiu $1 fjr 'b ¥ 9f % . Peking: Chung-kuo she-hui k'o-hsiieh ch'u-pan she, 1985. Chinesische Gesandtenberichte. See Haenisch, Erich, and Yao Ts'ung-wu. Chiu T'ang shu S JS H . See Liu Hsu. Chiu Wu-tai shih I S t t J . J a Hsueh Chii-cheng. Ch'iu Shu-sen 6(51Sf JS and Wang T'ing 2 Si. "Yuan tai hu k'ou wen t'i ch'u i 7C f^ Z3 P [ n J U S ! ! ." Yuan shih lun ts'ung 7 C $ i l , 2 (1983), pp. 111-24. Ch5ng In-ji $P ®5 fib et al., comps. KoryS sa M M i . 3 vols. Tokyo: Kokusho kankokai, 1908—9 ed. Also 3 vols. Seoul: Asea munhaksa, 1965—71. Cambridge Histories Online © Cambridge University Press, 2008 BIBLIOGRAPHY 735 Chou Ch'ing-shu JS ft ® . "Wang-ku pu shih chi S. * SB • ffl ." In Chung-kuo Meng-ku shih hsiieh hui ch'eng li ta hui chi nien chi k'an 4 1 IIIiS:'fi'£^#ljScAl^# it. M^l . Hu-ho-hao-t'e [Huhehot]: Chung-kuo Meng-ku shih hsiieh hui, 1979, pp. 147-229. Chou Ch'ing-shu, ed. Yuan jen wen chi pan pen mulu 7Z AJCMf&'fc B $k . Nanking: Nanching. ta hsueh (Nan ching ta hsiieh hsiieh pao ts'ung k'an M M ^C # ^ $8 31 TO ) 1983. Chou Liang-hsiao. See also Zhou Liangxiao. Chou Liang-hsiao ffl M. W . Hu-pi-lieh $& jft fS.. Ch'ang-ch'un: Chi-lin chiao yii ch'u pan she, 1986. Chou Liang-hsiao. "Meng-ku hsiian han i chih yu Yiian ch'ao huang wei chi ch'eng wen t'ii? * m t? m m m % m s. &. m & m m .•• Yuan shih lun ts'ung % £ it m, 3 (1986), pp. 31-46. Chou Liang-hsiao. "Yiian tai t'ou hsia fen feng chih tu ch'u t'an Jt ft $t T £)• H $!l S W I •" Yiian shih lun ts'ung 7C £. H Wt , 2 (1983), pp. 53-76. Chou Pi-ta JS '£> ~X. Wen-chung chi ~$Cffi.%k (Chou l-kuo Wen-chung kung wen chi Ml & ffl £ & £ X * ). SKCP ed. Chu Ch'i-yiian #1 ISt M . "Ku-ssu-lo cheng ch'iian hsing ch'eng ch'u t'an Oftffli<S © fi^filcWSi ." Hsi-tsang yen chiu B sK W ?£, 1982:2, pp. 68-77. Chu Hsi-tsu Sfc # ffl. "Hsi Hsia shih chi k'ao S H ^ If # . " 5/&»o we» >«>/» /fe'rf», 3, no. 11 (1943), pp. 25-30. Chii, Ch'ing-yiian I® ^ !® . "Government artisans of the Yuan dynasty." In Chinese social history: Translations of selected studies, trans, and ed. John De Francis and E-tu Zen Sun. Washington, D.C.: American Council of Learned Societies, 1956, pp. 234-46. Chii Ch'ing-yuan. "Yiian tai hsi kuan chiang hu yen chiu: Chih jen wei Yiian tai kuan chu chiang hu shih nu li te jen men 7Ctt$'&E/ a ffl3E : MI8.&7C ." Shihhuo yiieh k'an £ « E fll , 1 (1935), pp. ft t m E P S U m 65 A f] 367-401. Ch'iian Han-sheng it ^ # . Chung-kuo ching chi lun ts'ung 41 S ^ $? l i ^ . 2 vols. Hong Kong: Hsin Ya yen chiu so, 1972. Ch'iian Han-sheng. "Sung Chin chien te tsou ssu mao i ^3sfSJ#J:s&fAj§|^ ." Chung yang yen chiu yiian, Li shih yii yen yen chiu so chi k'an, 11 (1944), pp. 425—47. Repr. in vol. 1 of Ch'uan Han-sheng, Chung-kuo ching chi shih lun ts'ung. Hong Kong: Hsin Ya yen chiu so, 1972, pp. 211—33. Ch'uan Han-sheng. "Yuan-tai te chih-pi 7C f£ (ft JK $ ." Chung yang yen chiu yiian, Li shih yii-yen yen chiu so chi k'an, 15 (1948), pp. 1-48; repr. in vol. 1 of Ch'uan Hansheng, Chung-kuo ching chi shi lun ts'ung. Hong Kong: Hsin Ya yen chiu so, 1972, pp. 369-416. Ch'uan Heng ffi % . Keng shen wai shih M ^ ft £ . Repr. as vol. 3 of Shih liao ts'ung pien $. P, W. i i , ed. Kuang-wen shu-chu pien-i so SIXitJSjKiPftT. Taipei: Kuang-wen shu-chii, 1968. Ch'iian Heng. Keng shen wai shih H $ ft A . In Pao-yen t'angpi chi 9 M S S S , ed. Ch'en Chi-ju StlUS , 1606; repr. Taipei: I-wen yin-shu kuan, 1965. Cambridge Histories Online © Cambridge University Press, 2008 736 BIBLIOGRAPHY Chuang Chung-fang 3£ •(+ ~H, comp. Chin wen ya ^ 'SC $£. 1891; photolithographic repr. Taipei: Ch'eng-wen ch'u-pan she, 1967. Chung Keng-ch'i MM.& , comp. Kan-chou fu chih #WlftM . 1779. Repr. in Chung-kuo fang chih ts'ung shu: Hua-pei ti fang 41 ill 2f i£ 18 l i ^ i^lifcJfe^J.no. 561. Taipei: Ch'eng-wen ch'u-pan she, 1976. Chuang Lien £ M . Ming Ch'ingshihshih ts'ung t'an §H ft £ V18 16. Taipei: Hsiiehsheng shu-chii, 1972. Chung-kuo she hui k'o hsiieh yuan k'ao ku yen chiu so 4" Hi %t # ?4 *P K # S W 2E fit , ed. HJ/» Chung-kuo te k'ao ku fa hsien ho yen chiu #f 41 S W # * $ S! S3 W 5E . Peking: Wen-wu ch'u-pan she, 1984. Clark, Larry V. "The theme of revenge in the Secret history of the Mongols." In vol. 2 of Aspects of Altaic civilization, ed. Larry V. Clark and Paul A. Draghi. Indiana University Uralic and Altaic Series no. 134. Bloomington: Indiana University Press, 1978, pp. 37-57Cleaves, Francis W. "The biography of Bayan of the Barin in the Yuan shih." Harvard Journal of Asiatic Studies, 19 (1956); pp. 185-303. Cleaves, Francis W. "The biography of the empress Cabi in the Yuan shih." Harvard Ukrainian Studies, 3—4 (1979-80), pp. 138-50. Cleaves, Francis W. "A Chinese source bearing on Marco Polo's departure from China and a Persian source on his arrival in Persia." HarvardJournal of Asiatic Studies, 36 (1976), pp. 181-203. Cleaves, Francis W. "Darugha and gerege." Harvard Journal of Asiatic Studies, 16 (1953). PP- 237-59. Cleaves, Francis W. "The eighteenth chapter of an early Mongolian version of the Hsiao ching." HarvardJournal of Asiatic Studies, 45 (1985), pp. 225-54. Cleaves, Francis W. "The 'Fifteen "Palace poems" ' by K'o Chiu-ssu." Harvard Journal of Asiatic Studies, 20(1957), pp. 391—479. Cleaves, Francis W. "The first chapter of an early Mongolian version of the Hsiao ching." Acta Orientalia Academiae Scientiarum Hungaricae, 36 (1982), pp. 69-88. Cleaves, Francis W. "The historicity of the Baljuna covenant." Harvard Journal of Asiatic Studies, 18 (1955), pp. 357-421. Cleaves, Francis W. "K'uei-k'uei or Nao-nao?" Harvard Journal of Asiatic Studies, 10 (1947), pp. 1-12. Cleaves, Francis W. "The lingji of Aruy of 1340." HarvardJournal of Asiatic Studies, 25 (1964-5), pp. 31-79Cleaves, Francis W. "The memorial for presenting the Yuan shih." Asia Major, 3rd series, 1 (1988), pp. 59-69. Cleaves, Francis W., trans. The secret history of the Mongols: For thefirsttime done into English out of the original tongue, and provided with an exegetical commentary. Vol. 1. Cambridge, Mass.: Harvard University Press, 1982. Cleaves, Francis W. "The Sino-Mongolian inscription of 1335 in memory of Chang Ying-jui." Harvard Journal of Asiatic Studies, 13 (1950), pp. 1-131. Cleaves, Francis W. "Uighuric mourning regulations. "Journal of Turkish Studies, 1 (1977). PP- 6 5 - 9 3 . Cambridge Histories Online © Cambridge University Press, 2008 BIBLIOGRAPHY 737 Cordier, Henri. Ser Marco Polo: Notes and addenda to Sir Henry Yule's edition, containing the results of recent research and discovery. London: Murray, and N e w York: Scribner, 1920. Crump, James I. Chinese theater in the days of Kublai khan. Tucson: University of Arizona Press, 1980. Crump, James I. "Yuan-pen, Yuan drama's rowdy ancestor." East and West, 14 (1970), pp. 473-91Dardess, John W. Confucianism and autocracy: Professional elites in the founding of the Ming dynasty. Berkeley and Los Angeles: University of California Press, 1983. Dardess, John W. "Confucianism, local reform, and centralization in late Yuan Chekiang, 1342—59." In Yuan thought: Chinese thought and religion under the Mon- gols, ed. Chan Hok-Lam and William Theodore de Bary. New York: Columbia University Press, 1982, pp. 327—74. Dardess, John W. Conquerors and Confucians: Aspects of political change in late Yuan China. New York: Columbia University Press, 1973. Dardess, John W. "From Mongol empire to Yuan dynasty: Changing forms of imperial rule in Mongolia and Central Asia." Monumenta Serica, 30 (1972—3), pp. 117-65. Davis, Richard L. "Historiography as politics and Yang Wei-chen's 'Polemic on legitimate succession'." T'oung Pao, 59 (1983), pp. 33-72. Dawson, Christopher, ed. The Mongol mission: Narratives and letters of the Franciscan missionaries in Mongolia and China in the thirteenth and fourteenth centuries. N e w York: Sheed and Ward, 1955. De Bary, William Theodore. Neo-Confucian orthodoxy and the learning of the mind-and- heart. New York: Columbia University Press, 1981. De Francis, John, and E-tu Zen Sun, trans, and eds. Chinese social history: Translations of selected studies. Washington, D.C.: American Council of Learned Societies, 1956. de Harlez, Charles J., trans. Histoire de Vempire de Kin ou empire d'Or, Aisin-GurunI-Suduri Bithe traduit du Mandchou. Louvain: Charles Peeters, 1887. Demie'ville, Paul. "Notes d'arche'ologie chinoise." Bulletin de I'Ecole Franqaise d'Extreme-Orient, 25 (1925), pp. 458-67. Repr. in Paul Demieville, Choix d'itudes sinologiques (1921—19-70), ed. Yves Hervouet. Leiden: Brill, 1973, pp. 17-26. DemieVille, Paul. "La Situation religieuse en Chine au temps de Marco Polo." In Oriente Poliano: Studi e conferenze tenute all' Is. M. E. 0. in occasione del VII centenario delta nascita di Marco Polo (1254—1954), ed. Istituto Italiano per il Medio ed Estremo Oriente. Rome: Istituto Italiano per il Medio ed Estremo Oriente, 1957, pp. 193-236. Demieville, Paul. "Les Tombeaux des Song meridionaux." Bulletin de I'Ecole Frangaise d'Extreme-Orient, 25 (1925), pp. 458-567. Repr. in Paul Demieville, Choix d'itudes sinologiques (1921—1970), ed. Yves Hervouet. Leiden: Brill, 1973, pp. 17—26. Derevianko, Evgeniia I. Mokheskie pamiatniki Srednogo Amura. Novosibirsk: Nauka, 1975- Cambridge Histories Online © Cambridge University Press, 2008 738 BIBLIOGRAPHY Dictionary of Ming biography. See Goodrich, L. Carrington. Dolezelova-Velingerova, M., and James I. Crump, trans. Liu Chih-yiian chu-kungtiao: Ballad of the hidden dragon. Oxford: Clarendon Press, 1971. Dunnell, Ruth. "Soviet scholarship on Medieval China, 1982—1987." Bulletin of Sung—Yuan Studies, 20(1988), pp. 137-42. Dunnell, Ruth. "Tanguts and the Tangut state of Ta Hsia." Ph.D. diss., Princeton University, 1983. Dunnell, Ruth. "Who are the Tanguts? Remarks on Tangut ethnogenesis and the ethnonym Tangut." Journal ofAsian History, 18 (1984), pp. 78-89. Eberhard, Wolfram. "Die Chin im chinesischen Theater." In Studia Sino-Mongolica: Festschrift fur Herbert Franke, ed. Wolfgang Bauer. Miinchener Ostasiatische Studien no. 25. Wiesbaden: Franz Steiner, 1979, pp. 345—52. Ebisawa, Tetsuo M %t W US M • "Bondservants in the Yuan." Ada Asiatica, 45 (1983), pp. 27-48. Ebisawa Tetsuo. "Mongoru teikoku no toho sanoke ni kansuru shomondai t > J ) l / f i » S J H i i i : S S t ^ i r » 1 I ." Saitama daigaku kiyo *Sf X * * ,ffi 5 , 21 (1972), pp. 31-46. Endicott-West, Elizabeth. "Hereditary privilege in the Yuan dynasty." Journal of Turkish Studies {Festschrift for Francis W. Cleaves), 9 (1985), pp. 15—20. Endicott-West, Elizabeth. "Imperial governance in Yuan times." HarvardJournal of Asiatic Studies, 46(1986), pp. 523—49. Endicott-West, Elizabeth. "Merchant associations in Yuan China: The ortogh." Asia Major, 3rd series, 2 (1989), pp. 127—54. Endicott-West, Elizabeth. Mongolian rule in China: Local administration in the Yuan dynasty. Cambridge, Mass.: Harvard University Press, 1989. Endicott-West, Elizabeth. "The Yuan." In Soviet studies of premodern China, ed. Gilbert Rozman. Ann Arbor: University of Michigan Center for Chinese Studies, 1984, pp. 97—110. Etani Toshiyuki BS § © ^ . "Kaido no ran ni kansuru ichi kosacsu jj -i (-* <J <D SL llffltZ — % ^ " In Tamura Hakushi shoju Tdyoshi ronso fflW1i±gS#P^#£ at Wl , ed. Tamura Hakushi taikan kinen jigyokai fflttff±2!'&fB;&mt#. Kyoto: Tamura Hakushi taikan kinen jigyokai, 1968, pp. 89—104. Erh shih wu shih H + S ife , ed. Erh shih wu shih k'an hsing wei yuan hui - t E J ffl fr §1 fl # . Shanghai: K'ai-meng shu-tien, 1935. Fang Kuang-ch'ang 2J U H . "Yiian shih k'ao cheng Hang p'ien 7C ^ % WZ W H •" Wenshih J A , 1988:1, pp. 229-53. Farquhar, David M. The government of China under Mongolian rule — A reference guide. Miinchener Ostasiatische Studien no. 53. Stuttgart: Franz Steiner, 1990. Farquhar, David M. "The official seals and ciphers of the Yiian period." Monumenta Serica, 25 (1966), pp. 362-93. Farquhar, David M. "Structure and function in the Yiian imperial government." In China under Mongol rule, ed. John D. Langlois, Jr. Princeton, N.J.: Princeton University Press, 1981, pp. 25—55. Feng Ch'eng-chiin £§ & $%. "Yiian tai te chi ko Nan-chia-t'ai 7C ft &3ttM ^ ^ 'a ." Cambridge Histories Online © Cambridge University Press, 2008 BIBLIOGRAPHY 739 In Feng Ch'eng-chiin, Hsi-yu Nan-hat shih ti k'ao cheng lun chu hui chi B i$. j^I $5 5£ Jfe % WL M& ^f £& tt . Kowloon: Chung-hua shu-chii Hsiang-kang fen chii, 1976, pp. 200-16. Feng Chia-sheng M %. # . Lwo shih ch'u chiao M $L W $! . Repr. in Liao shih chiao k'an chi S3 $. t£ SS IB. In 5«»£ L«<*e Chin Yuan ssu shih tzu liao ts'ung k'an 3= M 3z 7C E3 £ ff & 8 TO , ed. Chao T'ieh-han MWiM. Taipei: Ta-hua yin-shu kuan, 1971, pp. 1—260. Feng Chia-sheng. Liao shih yii Chin shih hsin chiu Wu-tai shih hu cheng chii li M £. Jfil ^ £gfWSft£5ii^#iJ . Repr. in Liao shih chiao k'an chi. In Sung Liao Chin Yuan ssu shih tzu liao ts'ung k'an, ed. Chao T'ieh-han. Taipei: Ta-hua yin-shu kuan, 1971, pp. 517-84Feng Chia-sheng. Liao shih yuan liu k'ao $& 5& M §SE % . Repr. in Liao shih chiao k'an chi. In Sung Liao Chin Yuan ssu shih tzu liao ts'ung k'an, ed. Chao T'ieh-han. Taipei: Ta-hua yin-shu kuan, 1971, pp. 585-657. Fletcher, Joseph. "Integrative history: Parallels and interconnections in the early modern period, 1500-^1800." Journal of Turkish Studies, 9 (1985), pp. 37-57. Fletcher, Joseph. "The Mongols: Ecological and social perspectives." HarvardJournal of Asiatic Studies, 46 (1986), pp. 11-50. Fletcher, Joseph. "Turco-Mongolian monarchic tradition in the Ottoman empire." Harvard Ukrainian Studies, 3—4(1979—80), pp. 236—51. Fo tsu li tat t'ung tsai % ffl M ft S t? . See Nien-ch'ang. Franke, Herbert. "Ahmed: Ein Beitrag zur Wirtschaftsgeschichte Chinas unter Qubilai." Oriens, 1 (1948), pp. 222—36. Franke, Herbert, trans. Beitrage zur Kulturgeschichte Chinas unter der Mongolenherrschaft: Das Shan-kii sin-hua des Yang Yii. Abhandlungen fur die Kunde des Morgenlandes, vol. 32, no. 2. Wiesbaden: Franz Steiner, 1956. Franke, Herbert. "Die Belagerung von Hsiang-yang: Eine Episode aus dem Krieg zwischen Sung und Chin, 1206—1207." Society and history: Essays in honor of Karl August Wittfogel, ed. G. Ulmen. Berlin: de Gruyter, 1978, pp. 351—7. Franke, Herbert. "Chia Ssu-tao (1213-1275): A 'bad last minister.' " In Confucian personalities, ed. Arthur F. Wright and Denis C. Twitchett. Stanford, Calif.: Stanford University Press, 1962, pp. 217—34. Franke, Herbert. "Chinese historiography under Mongol rule: The role of history in acculturation." Mongolian Studies, 1 (1974), pp. 15-26. Franke, Herbert. "Chinese law in a multinational society: The case of the Liao (907— 1125)." Paper presented to the History of Chinese Medieval Law Conference, Bellagio, Italy, August 1981. Franke, Herbert. "Chinese texts on the Jurchen: A translation of the Jurchen monograph in the San ch'ao pei meng hui pien." Zentralasiatische Studien, 9 (1975), pp. 119-86. Franke, Herbert. "Chinese texts on the Jurchen II: A translation of chapter one of the Chin shih." Zentralasiatische Studien, 12 (1978), pp. 413—52. Franke, Herbert. "Could the Mongol emperors read and write Chinese?" Asia Major, n.s., 3 (1952), pp. 2 8 - 4 1 . Cambridge Histories Online © Cambridge University Press, 2008 74° BIBLIOGRAPHY Franke, Herbert. Diplomatic missions of the Sung state, 960-1276. Canberra: Australian National University, 1981. Franke, Herbert. "Dschau Mong-fu: Das Leben eines chinesischen Staatsmannes, Gelehrten und Kiinstlers unter der Mongolenherrschaft." Sinica, 15 (1940), pp. 25-48. Franke, Herbert. "Etymologische Bemerkungen zu den Vokabularen der JurcenSprache." In Florilegia Manjurica in memoriam Walter Fuchs, ed. Michael Weiers and Giovanni Stary. Asiatische Forschungen no. 80. Wiesbaden: Otto Harrassowitz, 1982, pp. 7—18. Franke, Herbert. From tribal chieftain to universal emperor and god: The legitimation of the Yuan dynasty. Munich: Verlag der Bayerischen Akademie der Wissenschaften, 1978. Franke, Herbert. Geld und Wirtschaft in China unter der Mongolenherrschaft: Beitrage zur Wirtschaftsgeschichte der Yiian-Zeit. Leipzig: Otto Harrassowitz, 1949. Franke, Herbert. "Jurchen customary law and the Chinese law of the Chin dynasty." In State and law in East Asia: Festschrift Karl Biinger, ed. Dieter Eikemeier and Herbert Franke. Wiesbaden: Otto Harrassowitz, 1981, pp. 215-33. Franke, Herbert. "The legal system of the Chin dynasty." In Collected studies on Sung history dedicated to Professor James T. C. Liu in celebration of his seventieth birthday, ed. Tsuyoshi Kinugawa. Kyoto: Dohosha, 1989, pp. 387—409. Franke, Herbert. Nordchina am Vorabend der mongolischen Eroberungen: Wirtschaft und Gesellschaft unter der Chin-Dynastie. Geisteswissenschaften, Vortrage — Rheinisch— Westfalische Akademie der Wissenschaften no. G 228. Opladen: Westdeutscher Verlag, 1978. Franke, Herbert. "A note on wine." Zentralasiatische Studien, 8 (1974), pp. 241-5. Franke, Herbert. "Review of Vorob'ev, M.V., Chzhurzheni igosudarstvo Tszin'." Monumenta Serica, 32 (1978), pp. 404-8. Franke, Herbert. "The role of the state as structural element in polyethnic societies." In Foundations and limits of state power in China, ed. Stuart R. Schram. London: School of Oriental and African Studies, and Hong Kong: Chinese University Press, 1987, pp. 87—112. Franke, Herbert. "Sen-ge: Das Leben eines uigurischen Staatsbeamten zur Zeit Chubilai's dargestellt nach Kapitel 205 der Yiian-Annalen." Sinica, 17 (1942), pp. 90-113. Franke, Herbert. "Sino-Western contacts under the Mongol empire." Journal of the Royal Asiatic Society: Hong Kong Branch, 6 (1966), pp. 49—72. Franke, Herbert. "Some aspects of Chinese private historiography in the thirteenth and fourteenth centuries." In Historians of China and Japan, ed. William G. Beasley and Edwin G. Pulleyblank. Oxford: Oxford University Press, 1961, pp. 115-34. Franke, Herbert. "Some folkloristic data in the dynastic history of the Chin." In Legend, lore and religion in China: Essays in honor of Wolfram Eberhard on his seventieth birthday, ed. Sarah Allan and Alvin P. Cohen. San Francisco: Chinese Materials Center, 1979, pp. 135-53. Cambridge Histories Online © Cambridge University Press, 2008 BIBLIOGRAPHY 741 Franke, Herbert. Studien und Texte zur Kriegsgescbkbte der siidlichen Sungzeit. Asiatische Forschungen no. 102. Wiesbaden: Otto Harrassowitz, 1987. Franke, Herbert, ed. Sung biographies. 4 vols. Wiesbaden: Franz Steiner, 1976-8. Franke, Herbert. "A Sung embassy diary of 1211-1212: The Shih-Chin lu of Ch'eng Cho." Bulletin de I'Ecole Frati(aise d'Extrbne-Orient, 69 (1981), pp. 171-207. Franke, Herbert. Tangutische und chinesische Quellen zur Militdrgesetzgebung des 11. bis 13. Jahrhunderts. See Evgenii I. Kychanov. Franke, Herbert. "Tan-pa, a Tibetan lama at the court of the great khans." In vol. 1 of Orientalia Venetiana, ed. Mario Sabattini. Florence: Leo S. Olschki, 1984, pp. 157-80. Franke, Herbert. "Tibetans in Yuan China." In China under Mongol rule, ed. John D. Langlois, Jr. Princeton, NJ.: Princeton University Press, 1981, pp. 296—328. Franke, Herbert. "Treaties between Sung and Chin." In Etudes Song in memoriam Etienne Baldzs, 1st series, no. 1. Paris: Mouton and Ecole pratique des hautes etudes, 1970, pp. 55-84. Franke, Herbert. "The 'Treatise on punishments' in the Liao history." Central Asiatic Journal, 27 (1983), pp. 9-38. Franke, Herbert. "Women under the dynasties of conquest." In La Donna nella Cina imperiale e nella Cina repubblicana, ed. Lionello Lanciotti. Florence: Leo S. Olschki, 1980, pp. 23-43. Franke, Otto. Geschichte des chinesischen Reiches. 5 vols. Berlin: W. de Gruyter, vol. 1, 1930; vol. 2, 1936; vol. 3, 1937; vol. 4, 1948; vol. 5, 1952. Friedland, Paul. "A reconstruction of early Tangut history." Ph.D. diss., University of Washington, 1969. Fu Le-huan $ ^ $1. Liao shih ts'ung k'ao JS £ 8$ =#. Peking: Chung-hua shu-chii, 1984. Fu Shen W $ . Yuan tai huang shih shu hua shou ts'ang shih liieh TtiX Sk%LWM$.W, &. B§ . Taipei: Kuo-li Ku kung po wu yuan, 1981. Fuchs, Walter. "Analecta zur mongolischen Ubersetzungsliteratur der Yiian-Zeit." Monumenta Serica, 11 (1946), pp. 33—64. Fujishima Tateki M % §1flf. "Gen no Juntei to sono jidai 7G CO 11 ^ £ f CO B# ft ." Otani gakuho i\ Q ¥ $8 , 49 (March 1970), pp. 50-65. Fujishima Tateki. "Gen no Minso no shogai 7C<7)^^c7)iJiii." Otani shigaku ^C §•. ^PS , 12 (1970), pp. 12-28. Gabelentz, Hans Conon von der. Geschichte der grossen Liao aus dent Mandschu ubersetzt. St. Petersburg: Commissionaire der kaiserlischen Akademie der Wissenschaften, 1877. Geley, Jean-Philippe. "L'Ethnonyme mongol a l'6poque pre-finggisqanide (XII siecle)." Etudes Mongoles, 10 (1979), pp. 59-89. Gellner, Ernest. "Anomalies of nofixedabode." Times Literary Supplement, 13 March I98i,p. 273. Gentensho sakuin-ko 7C #!• % 'M 31 SI . See Kyoto daigaku jimbun kagaku kenkyujo Gentensho kenkyu han. Cambridge Histories Online © Cambridge University Press, 2008 742 BIBLIOGRAPHY Gernet, Jacques. Daily life in China on the eve of the Mongol invasion, 1250—1276. Trans. H. M. Wright. New York: Macmillan, 1962. Gernet, Jacques. La Vie quotidienne en Chine a la veille de I'invasion Mongole, 1250— 1276. Paris: Hachette, 1959. Gibert, Lucien. Dictionnaire historique et giographique de la Mandchourie. Hong Kong: Imprimerie de la Societe des Missions-Etrangers, 1934. Golden, Peter B. "Imperial ideology and the sources of political unity amongst the pre-Cinggisid nomads of western Eurasia." Archivum Eurasiae Medii Aevi, 2 (1982), pp. 37-76. Goodrich, L. Carrington, and Chaoying Fang, eds. Dictionary of Ming biography. 2 vols. New York: Columbia University Press, 1976. Groeneveldt, W. P. "The expedition of the Mongols against Java in 1293 A.D." China Review, 4 (January—February 1876), pp. 246—54. Grube, Wilhelm. DieSprache undSchrift derjulen. Leipzig: Otto Harrassowitz, 1896; repr. Peking: Licoph Service, 1939. Haeger, John W. "Marco Polo in China? Problems with internal evidence." Bulletin of Sung and Yuan Studies, 14 (1978), pp. 2 2 - 3 0 . Haenisch, Erich. Zum Untergang zweier Reiche: Berichte von Augenzeugen am den Jahren 1232—33 und 1268—70. Wiesbaden: Otto Harrassowitz, 1969. Haenisch, Erich, and Yao Ts'ung-wu. Trans and ed. Peter Olbricht and Elisabeth Pinks. Meng-ta pei-lu und Hei-ta shih-lueh: Chinesische Gesandtenberichte u'ber die friihen Mongolen 1221 und 1237, nach Vorarbeiten von Erich Haenisch und Yao Ts'ung-wu ubersetzt und kommentiert von Peter Olbricht und Elisabeth Pinks: Eingeleitet von Werner Banck. Wiesbaden: Otto Harrassowitz, 1980. Halperin, Charles J. Russia and the Golden Horde: The Mongol impact on medieval Russian history. Bloomington: Indiana University Press, 1985. Hambis, Louis, trans, he chapitre cviii du Yuan che: Les Fiefs attribues aux membres de la fami He imperiale et aux ministres de la cour mongole d'apres I'histoire chinoise officielle de la dynastie mongole. Monographies du T'oung Pao, vol. 3. Leiden: Brill, 1954. Hambis, Louis, trans. Le chapitre cvii du Yuan che: Les Genealogies imperiales mongoles dans I'histoire chinoise officielk de la dynastie mongole. Avec des notes supplementaires par Paul Pelliot. T'oung Pao suppl. no. 38. Leiden: Brill, 1945. Hambis, Louis. Gengis khan. Paris: Presses Universitaires de France, 1973. Hambis, Louis. "L'Histoire des Mongols avant Genghis-khan d'apres les sources chinoises et mongoles, et la documentation conservee par RaSId-al-Dln." Central Asiatic Journal, 14(1970), pp. 125-33. Hambis, Louis. "Notes preliminaires a une biographie de Bayan le Markit." Journal Asiatique, 241 (1953), pp. 215—48. Hambis, Louis. "Notes sur I'histoire de Coree a l'6poque mongole." T'oung Pao, 45 (X957). PP- I 5 1 - 2 1 8 Hamilton, James R. Les Ouighours a I'epoque des cinq dynasties d'apres les documents chinois. Paris: Presses Universitaires de France, 1955. Han Ju-lin $?ffi# . Ch'iung lu chi: Yuan shih chi hsi pei min tsu shih yen chiu =f ft Cambridge Histories Online © Cambridge University Press, 2008 BIBLIOGRAPHY 743 . Shanghai: Shang-hai jen-min ch'u-pan she, 1982. Han Ju-lin, comp. Yiian ch'ao shih 7C $1 £ . 2 vols. Peking: Jen-min ch'u-pan she, 1986. Han Woo-keun (Han U-gun). The history of Korea. Trans. Lee Kyung-shik. Honolulu: East-West Center Press, 1971. Han Yin-ch'eng @ l l . "Lin, Fu chou chien chih yii Che shih yiian liu S!ffi'N\ M S H #f .R M M ." Ning-hsia she hui k'o hsiieh $E g t t # W 3E, 1981:1, pp. 63-7. Hana, Corinna. Bericht iiber die Verteidigung der Stadt Te-an wdhrend der Periode K'aihsi, 1205 bis 1208. Wiesbaden: Franz Steiner, 1970. Haneda Toru 33ffl~$ . "Gench5 ekiden zakko 7C §8 P W % # . Toyo bunko so kan; Eiraku daiten, keisei daiten tan-sekima fukusei bon no fuhen M Pfc 3C M M TO • 7*m±& ' «1tt*fttt#raa«*COWII . Tokyo: Toyo bunko, 1930. Rpt. in Haneda Hakushi shigaku rombun shu, jokan, rekishihenWi E8 W ± £ US ft 3C £i • - L # : /S i ^ (Kyoto, Toyoshi kenkyu sokan 3, 1-2, pp. 32-114). Hanedu Toru. "Moko ekiden k5 H: "4f P IS %." Tokyo, Tcfyo kyokai chosabu gakujutsu hokokuMW&^M&U^ffi^^itWoL 1, 1909.Repr. in Haneda Hakushi shigaku rombun shu, jokan, rekishihen (Kyoto: Toyoshi kenkyu sokan, 3, 1—2, pp. 1—31). Harada, Yoshito. Shang-tu: The summer capital of the Yuan dynasty. Tokyo: Toa kokugakukai, 1941. Hartwell, Robert. "A cycle of economic change in imperial China: Coal and iron in northeast China, 750-1350." Journal of the Economic and Social History of the Orient, 10(1959), PP- I O 2 ~59Hatada Takashi MfflSi. Genko: Moko teikoku no naibu jijo 7U5& : S^WIJScoftSB S t f . Tokyo: Chuokoronsha, 1965. Hayden, George A. Crime and punishment in medieval Chinese drama: Three Judge Pao plays. Harvard East Asian Monographs no. 82. Cambridge, Mass.: Harvard University Press, 1978. Henthorn, William E. Korea: The Mongol invasions. Leiden: Brill, 1963. Hino Kaisaburo B S i HfiP. "Teian koku J E $ i . " l n vol. 6 of Ajia rekishi jiten T ^ / S S I f t . Tokyo: Heibonsha, i960, p. 388. Ho, Ping-ti. "An estimate of the total population of Sung-Chin China." In ttudes Song in memoriam Etienne Baldzs, 1st series, no. 1, ed. Francoise Aubin. Paris: Mouton and Ecole pratique des hautes 6tudes, 1970, pp. 33-53. Ho Ping-ti. Studies on the population of China, 1368—1953. Cambridge, Mass.: Harvard University Press, 1959. Holmgren, Jennifer. "Marriage, kinship and succession under the Ch'i-tan rulers of the Liao dynasty (907-1125)." Young Pao, 72 (1986), pp. 44-91. Holmgren, Jennifer. "Observations on marriage and inheritance practices in early Mongol and Yiian society with particular reference to the levitate." Journal of Asian History, 20(1986), pp. 127—92. Holmgren, Jennifer. "Yeh-lii, Yao-lien and Ta-ho: Views of the hereditary prerogative in early Khitan leadership." Papers on Far Eastern History, 34 (1986), pp. 37-81. Cambridge Histories Online © Cambridge University Press, 2008 744 BIBLIOGRAPHY Holt, P. M., Ann K. S. Lamb ton, and Bernard Lewis, eds. The Cambridge history of Islam. 2 vols. Cambridge: Cambridge University Press, 1970. Hoog, Constance, trans. Prince Jin-gim's textbook of Tibetan Buddhism. Leiden: Brill, 1983. Hori Kyotsu. The Mongol invasions and the Kamakura bakufu. Ph.D. diss., Columbia University, 1967. Horie Masaaki M tt SS BM . "Mongoru - Gencho jidai no toho san-urusu kenkyu josetsu e y z T j W - T G l B ^ f t t W l R ^ H ^ ) M i ? E H ." In Tohdgaku ronshu: Ono Katsutoshi hakushi shoju kinen MJJ9^M '• /> it , ed. Ono Katsutoshi hakushi shoju kinenkai 'bW B ¥ If ± 31 M IE Kyoto: Ryukoku daigaku Toyo shigaku kenkyukai, 1982, pp. 377-410. Horie Masaaki. "Temuge Otchigin to sono shison 7 k, V "% %V ^ ¥ > £ •? 0) •? M ." Toyo shien M W- £l ?E , 24—25 (1986), pp. 225-70. Hou Jen-chih IS C £ and Chin T'ao & 81. Pei-ching shih hua it M £. ag. Shanghai: Jen-min ch'u-pan she, 1980. Howorth, Sir Henry Hoyle. "The northern frontagers of China. Pt. V: The Khitai or Khitans." Journal of the Royal Asiatic Society, n.s., 13(1881), pp. 121-82. Hsia, Chih-tsing. The classic Chinese novel: A critical introduction. New York: Columbia University Press, 1968. Hsia Kuang-nan H % l$f. Yuan tai Yun-nan shih ti ts'ung k'ao mulu Jt ft i t ^ $. M W. % @ $k . Shanghai: Chung-hua shu-chii, 1935. Hsia Wen-yen I X f , comp. Tu hui poo chien II ft ff £ KHCPTS ed. Hsiang-mai # Si . Pien wei lu ¥$fk%. . Taisho Tripitaka, vol. 52, pp. 751-81. Hsiao Ch'i-ch'ing WifllBt. "Hu-pi-lieh shih tai ch'ien ti chiu lii k'ao k. & f}\ B# ft r W&V IS] ^ ." Ta-lu tsa chih 'X ^ % ^ v , 25, no. 1 (July 1962), pp. 16-22; 25, no. 2 (July 1962), pp. 25-8; 25, no. 3 (August 1962), pp. 22-7. Hsiao, Ch'i-ch'ing. The military establishment of the Yuan dynasty. Cambridge, Mass.: Harvard University Press, 1978. Hsiao, Ch'i-ch'ing. "Yen Shih, 1182—1240." Papers on Far Eastern History, 33 (1986), pp. 1 1 3 - 2 8 . Hsiao Ch'i-ch'ing. "Yuan tai k'o chu yii ching ying liu tung - i Yiian-t'ung yuan nien chin shih wei chung hsin 7tft®m$l^¥kUm J ^ t ^ 'P 't- ." Han hsiieh yen chiu SI ^ ffi % , 5 (1987), pp. 129-60. Hsiao Ch'i-ch'ing. "Yuan tai Meng-ku jen Han hsiieh tsai t'an 7C ft ~& A W . " I n Kuo shih shih lun: T'ao Hsi-sheng hsien sheng chiu ch'ihjung ch'ing chu shou lun wen chi m$.mm •• m%mft±A&$k&$imm'*m , ed. Yang Lien-sheng t§ M IS . 2 vols. Taipei: Shih-huo ch'u-pan she, 1988, vol. 2, pp. 373-88. Hsiao Ch'i-ch'ing. "Yuan tai Meng-ku jen te Han hsiieh 7 C f t S ' S f A 6 ^ i l ^ . " I n Kuo chi Chung-kuo pien chiang hsiieh shu hui i lun wen chi' M %* 4 1 M ift 8 1 9 ffi # Wk tit X H (Proceedings of the international conference on Chinese border area studies), ed. Lin En-hsien # S i I . Taipei: Kuo-li cheng-chih ta-hsiieh, 1985, pp. 369-428. Hsiao Ch'i-ch'ing. Yiian tai shih hsin t'an Jt ft !£ Iff $fc . Taipei: Hsin wen-feng ch'upan kung-ssu, 1983. Hsiao Ch'i-ch'ing. "Yiian tai ssu ta Meng-ku chia tsu 7C ft E3 X. W 'fi %• 8£ ." In Cambridge Histories Online © Cambridge University Press, 2008 BIBLIOGRAPHY 745 Hsiao Ch'i-ch'ing, Yiian tai shih hsin t'an 7€ ft, £ Iff W- . Taipei: Hsin wen-feng ch'u-pan kung-ssu, 1983, pp. 141-230. Hsiao Ch'i-ch'ing. "Yiian tai te ju hu: Ju shih ti wei yen chin shih shang te i chang 7C ft #1 §? J3 : m± ak®.mm$. ±ff3~M ." Tung fang wen hua MHXit, 16 (1978), pp. 151-78. Repr. in Hsiao Ch'i-ch'ing, Yiian tai shih hsin t'an 7E "ft Si $f W . Taipei: Hsin wen-feng ch'u-pan kung-ssu, 1983, pp. 1—58. Hsiao Ch'i-ch'ing. "Yiian-t'ung yiian nien chin shih lu chiao chu 7C %£ ft ^f M i f&t $ ft ." Shih huo yiieh k'an & W. R TU , 13 (1983), pp. 72-90, 147-62. Hsiao Hsiin IF fa, comp. KM kung i lu Sfe % Ji % . In Pei-p'ing k'ao i t ¥ *$. Peking: Pei-chung ch'u-pan she, 1963. Hsiao Kung-ch'in IS #) HI. "Lun Ta Meng-ku kuo te han wei chi ch'eng wei chi H . ." Yiian shih chi pei fang min tsu shih yen chiu chi k'an fil , 5 (1981), pp. 48-59. Hsiao Kung-ch'in. "Lun Yiian tai huang wei chi ch'eng wen t'i Mi 7C •ft fi &. 3£ & PH M ." Yuan shih chi pei fang min tsu shih yen chiu chi k'an 7 C i R MW , 7(1983). PP- 22-39. Hsiao Kung-ch'in. "Ying-tsung hsin cheng yii Nan-p'o chih pien 5 Z. %k •" Yiian shih chi pei fang min tsu shih yen chiu chi k'an 7C£ % % fll , 4 (1980), pp. 36-46. Hsin T'ang shu $f S 1f . See Ou-yang Hsiu and Sung Ch'i. Hsin Wu-tai shih $f £ ft £ . See Ou-yang Hsiu. Hsu Fan ff PL . Yiian tai li chih yen chiu Jt \X ^ $0fiff3E . Peking: Lao tung jen shih ch'u pan she, 1987. Hsu Meng-hsin '&-1? 3£ , comp. San ch'ao pei meng huipien H 19 ifc r§. # H . Ed. of Yiian Tsu-an It ffl ^ , 1878. Rpt. in 4 vols. Taipei: Wen hai ch'u pan she, 1966. Hsu Sung ffi fe , comp. Sung hui yao chi kao ^ # W ft $?. Photogtaphic repr. of ms. Peking, 1936; repr. Peking, 1957; repr. Taipei: Hsin wen-feng ch'u-pan she, 1975Hsu tzu chih t'ung Men ch'ang pien M 'M & H § ^ M . See Li T'ao. Hsu Yu-jen p f f I . Chih cheng chi II IE %k . Liao-ch'eng, Shantung, 1911; repr. in vol. 7 of Yiian jen wen chi chen pen ts'ung k'an Jt A ~$C % f£ ^ H flj , comp. Wang Te-i I ^ S . Taipei: Hsin wen-feng ch'u-pan kung-ssu, 1985, pp. 1-367. Hsiieh Chii-cheng ^ S IE et al., eds. CA/« Wu-tai shih f 5 R $ . Peking: Chunghua shu-chii, 1976 ed. Hu Chih-yu 49 fft Ji.. Tzu-shan ta ch'iian chi % Oj A ^ M . SKCS ed. Hu Chu S8 I*J. Ch'un-pai chai lei kao S e » i i . raCC ed. Hu Ts'ui-chung S9 ff 4". Y«w« J W AJ» pien S S S S ' . SKCS ed. Huang Chin ^ ?f . Chin-hua Huang hsien sheng wen chi & & M 9c $L X %.. SPTKed. Huang Chin. Huang Wen-hsien chi Sf ~X SK ft . TSCC ed. Huang Ch'ing-lien H ?t M . Yuan tai hu chi chih tu yen chiu 7C ft F> It U & W % . Taipei: Kuo-li T'ai-wan ta-hsiieh wen-hsiieh yiian, 1977. Huang Ch'ing-yiin H JS It . "Kuan yu Pei Sung yii Hsi Hsia ho yiieh chung yin chiian ch'a te shu Hang wen t'i ISSI)Kifc*IISHft«J*ffliS5KW Chung hsiieh li shih chiao hsiieh + § I S S i , 1957:9, pp. 19—20. Cambridge Histories Online © Cambridge University Press, 2008 746 BIBLIOGRAPHY Huang, Ray. "Ming fiscal administration." In The Ming dynasty, 1368—1644, pt. 2, vol. 8 of The Cambridge history of China, ed. Frederick W. Mote and Denis C. Twitchett (forthcoming). Huang, Ray. Taxation and governmentalfinancein sixteenth-century Ming China. Cam- bridge: Cambridge University Press, 1974. Huang Shih-chien ^ B$ £ . "Chen-chin yii Yuan ch'u cheng chih M ± H 7C W Be J&." Y«<*» JA//& /»» «'«»^ ?t £ it I t , 3 (1986), pp. 193-204. Huang Shih-chien. "Mu-hua-li kuo wang hui hsia chu chiin k'ao ^ § S i I ( I T S i t . . " Y»«A JJWA lun ts'ung Jt £ H SI , 1 (1982), pp. 57-71. Huang Shih-chien. "Ta Yuan t'ung chih k'ao pien 'X 7C 5§ M % $#." Chung-kuo she hui k'o hsiieh 41 @ tt # £f ^ , 2 (1987), pp. 157-71. Huang Shijian (Huang Shih-chien). "The Persian language in China during the Yuan dynasty." Papers on Far Eastern History, 34 (1986), pp. 83-95. Huber, Edouard, "Etudes Indochinoises: V. — La Fin de la dynastie de Pagan." Bulletin de I'Ecole Frangaise d'Extreme-Orient, 9 (1909), pp. 633—80. Hucker, Charles O. The censorial system of Ming China. Stanford, Calif.: Stanford University Press, 1966. Hucker, Charles O. A dictionary of official titles in imperial China. Stanford, Calif.: Stanford University Press, 1985. Hucker, Charles O. "The Yuan contribution to censorial history." Chungyangyen chiu yuan li shih yii yen yen chiu so chi k'an * 5*iW S f i l t S a flf^SJft ftflJ . Bulletin of the Institute of History and Philology, Academia Sinica, extra vol. 4 (i960), pp. 219-27. Hulsewe, A. "Review of Christian Schwarz-Schilling, Der Friede von Shan-Yuan (1005 n. Chr.): Ein Beitrag zur Geschichte der chinesischen Diplomatic." T'oung Pao, 47 (1959). PP- 445-74Hummel, Arthur W., ed. Eminent Chinese of the Ch'ing period. 2 vols. Washington, D. C : U.S. Government Printing Office, 1943—44. Hung Chin-fu ift sfe "S . "Ts'ung t'ou-hsia fen feng chih tu k'an Yuan ch'ao cheng ch'uan te hsing chih" S f e S T ^ ^ S U S ^ T C ^ J & S W t t M . Chung yang yen chiu yuan li shih yii yen yen chiu so chi k'an, 58 (1987), pp. 483—907. Hung Chin-fu. Yuan tai chien ch'a chih tu yen chiu 7nf^la^$!lSSf^E . 2 vols. Taipei: Kuo-li T'ai-wan ta-hsiieh li-shih-hsueh yen-chiu-so, 1972. Hung Chiin 8S $1. Yuan shih i wen chengpu Jt £ W X Mffli.Yuan-ho, Kiangsu: ed. of Lu Jun-hsiang W. M ^ , 1897. Hung Hao i& 6£. Sung mo chi wen tS ^ £2 K!. Repr. in vol. 3 ofLiao-hai ts'ung shu & MM9 , ed. Chin Yii-fu 3z§iit5t , Ta-lien: Liao-hai shu-she, 1931-4; repr. Taipei: I-wen yin-shu kuan, 1971 [?]. Also TSCC ed. Idema, Wilt L., and Stephen H. West. Chinese theater 1100—1450: A source book. Wiesbaden: Franz Steiner, 1982. Ikeuchi Hiroshi ?fe ft 'S.. Genko no shinkenkyu 76 jg <T> $f 9f % . 2 vols. Tokyo: Toyo bunko, 1931. Ikeuchi Hiroshi. Man-Sen shi kenkyu. Vol. 3: Chusei S$ B £ if % : "f.ift . Tokyo: Yoshikawa kobunken, 1963. Cambridge Histories Online © Cambridge University Press, 2008 BIBLIOGRAPHY 747 Imai Hidenori ^ # ^ J3 . "Kincho ni okeru jikan meigaku no hatsubai # §5 IC Jft 4>^M^m<r>mW ." ToAo shukyo % H m ft , 45 (197 5), pp. 48-70. Imai Hidenori. "Kindai Joshin no shinko — Saiten wo chushin to shite sfe ft ix. It CO •fs ffll — S ^ fc t ' t i U i ." In Toyogaku ronshu: Mori Mikisaburo Hakushi shoju kinen SC#*Sil£:$EH$HgB1S±SiliS:& ,ed. Mori Mikisaburo Hakushi shoju kinen jigyokai^f H ^ H f i | 5 f f ± ^ # f B ^ : ^ l | # . Kyoto: Hoyu shoten, 1979, pp. 773-90. Inosaki Takaoki #fl£fE£ PJ. "Gendai no take no sembaiken to sono shiko suru igi 7C «<7)t5-CO»M«i f C O f f l f r t ^ * * ." Toyoshi kenkyii * # J & B t S B , i 6 (September 1957), pp. 29-47. Inosaki Takaoki. "Gendai shasei no seijiteki kosatsu 7C ft tt Pi CO iS ?n W # flf ." TffywA/ £e»4:>« * # £ 5f 5£ , 15 (July 1956), pp. 1-25. Ishida Mikinosuke E EB f£ ^ 8fj . "Gendai no kogeika Neparu no ozoku Aniko no den ni tsuite jtft <F> I l f l f\ — ; ! / £ ! H / B 1 * <7) # (C i t l» X •" Moko gakuho W. ~& # ^ , 2 (1941), pp. 244-60. Ishida Mikinosuke. "Gen no J5to ni tsuite 7C CO J l i P tC St l^ T ."In vol. 1 of N//6o« daigaku soritsu shichijunen kinen rombunshu 0 ^^C^ilJlZl-b + ^ S ^ i i A^ , ed. Nihon daigaku B ^ ^: #i . Tokyo: Nihon daigaku, i960, pp. 271-319. Ishida Mikinosuke. "Jurfica." In lkeuchi hakushi kanreki kinen Toyoshi ronso ?fe ?3 tt ± S H S ! E ; § : ^ # £ i i l j i : . Tokyo: Sayubo kankokai, 1940, pp. 39—57. Repr. in Ishida Mikinosuke, Toa bunkashi soko M 35 X -fb $L H # . Tokyo: Toyo bunko, 1973. PP- 71-86. Ishii Susumu. "The decline of the Kamakura bakufu." In Medieval Japan, ed. Kozo Yamamura, vol. 3 of T/6e Cambridge history of Japan. Cambridge: Cambridge University Press, 1990, pp. 128—75. Isono, Fujiko. "A few reflections on the anda relationship." In vol. 2 of Aspects of Altaic civilization, ed. Larry V. Clark and Paul A. Draghi. Indiana University Uralic and Altaic Series no. 134. Bloomington: Indiana University, 1978, pp. 81-7. ISzamc, N. "L'Etat feodal mongol et les conditions de sa formation." Etudes Mongoles, 5 (1974), pp. 127-30. Iwai, Hirosato. "The source and meaning of Ta'chen, the dynastic title of P'u-hsien Wan-nu." Memoirs ofthe Research Department oftheToyo bunko, 9(1937), pp. in—61. Iwamura Shinobu & ti iS-. Mongoru shakai keizaishi no kenkyii -t > 3 ' ;H± # $2 $? $. CO 57F % . Kyoto: Kyoto daigaku jimbun kagaku kenkyujo, 1968. Iwamura Shinobu and Tanaka Kenji B3 41 BS — , eds. (Koteibon) Gentensho: Keibu (t£ isL^f) 7 C ^ ^ ; JPJcf|5.2 vols. Kyoto: Kyoto daigaku jimbun kagaku kenkyujo, 1964, 1972. Iwasaki Tsutomu $a 4S "ti. "Seiryofu Banrashi seiken shimatsu ko H®)ffif S J i S ft #& ^ % ." Tohogaku MJi^,47 (1974), pp. 25—41. Iwasaki Tsutomu. "Seiryofu seiken no metsubo to Soka zoku no hatten B M ffi 2& W. (Dffl,t: t mWM<n i i ."In Suzuki Shun sensei koki kinen Toyoshi ronso t£fc<£. # £ M f S ,ed. Suzuki Shun sensei koki kinen Toyoshi ronso Cambridge Histories Online © Cambridge University Press, 2008 748 BIBLIOGRAPHY henshuhnkai &*«$fc££#f2¥m#£§tSiai*g«# .Tokyo: Yamakawa shuppansha, 1975, pp. 73—88. Iwasaki Tsutomu. "S5ka jo Kokushira seiken no seikaku to kito ^ W ^ I f l W l S i & S <n&&t±m ."Chu6daigakuAjiashikenkyu*¥$:-XmTyT$LW!$i ,2 (1978), pp. 1-28. Jagchid, Sechin. See also Ch'a-Ch'i Ssu-ch'in and C. R. Bawden. Jagchid, Sechin. "Kitan struggles against Jiirchen oppression: Nomadism versus sinicization." Zentralasiatische Studien, 16 (1982), pp. 165—85. Jagchid, Sechin. "The Kitans and their cities." Central AsiaticJournal, 25 (1981), pp. 70-88. Jagchid, Sechin. "Patterns of trade and conflict between China and the nomads of Mongolia." Zentralasiatische Studien, 11 (1977), pp. 177—204. Jagchid, Sechin, and Paul Hyer. Mongolia's society and culture. Boulder, Colo.: Westview, 1979. Jagchid, Sechin, and Van Jay Simons. Peace, war, and trade along the Great Wall: Nomadic-Chinese interaction through two millennia. Bloomington: Indiana University Press, 1989. Jan, Yun-hua. "Chinese Buddhism in Ta-tu: The new situation and new problems." In Yiian thought: Chinese thought and religion under the Mongols, ed. Hok-lam Chan and William Theodore de Bary. New York: Columbia University Press, 1982, pp. 375-4I7Johnson, Douglas L. The nature of nomadism: A comparative study ofpastoral migrations in southwestern Asia and northern Africa. University of Chicago, Department of Geography. Research paper no. 118. Chicago: University of Chicago Press, 1969. Johnson, Linda Cooke. "The wedding ceremony for an imperial Liao princess: Wall paintings from a Liao dynasty tomb injilin." Artibus Asiae, 44(1983), pp. 107-36. Johnson, Wallace. The Tang code. Vol. 1: General principles. Princeton, NJ.: Princeton University Press, 1979. Juvaynl. See 'Ata Malik Juvaynl. Juzjani. See Minhaj al-Din Juzjani. Kahle, Paul. "Chinese porcelain in the lands of Islam." Transactions of the Oriental Ceramic Society, 18 (1940— 1), pp. 27—46. Kanda Kiichiro Wffl5 — IP . "Gen no Bunso no furyo ni tsuite % <F> X §? <T) J \ZWL^T ." In Haneda Hakushi shoju kinen Toyoshi ronso ^ E 0 W ± S » i i g , ed. Haneda Hakushi kanreki kinenkai 3 e f f i 1 t ±J © M ! E & # . Kyoto: Toyoshi kenkyukai, 1950, pp. 453-68. Kane, Daniel A. "The Sino-Jurchen vocabulary of the Bureau of Interpreters. Ph.D. Diss., Australian National University, 1975. Kane, Daniel A. The Sino-Jurchen vocabulary of the Bureau of Interpreters. Bloomington: Indiana University Research Institute for Inner Asian Studies, 1989. Kao Wen-te S X S . Meng-ku nu li chih yen chiu W £ U & *J W % . Koke Khota: Nei Meng-ku jen-min ch'u-pan she, 1980. Kao Wen-te. "Yiian T'ai-ting te shou nien cheng wu 7C 3? S if? H ^ SE H ." In vol. 1 of Min tsu shih lun ts'ung £5 M i I t H , ed. Chung-kuo she hui k'o hsiieh yiian Cambridge Histories Online © Cambridge University Press, 2008 BIBLIOGRAPHY 749 min tsu yen chiu so min tsu li shih yen chiu shih QM S l i $ B f ? S i . Chung-kuo she hui ko, Peking: Chung-hua shu-chii, 1987, p. 38. Kao Yu-kung. "Source materials on the Fang La rebellion." HarvardJournal of Asiatic Studies, 26(1966), pp. 211—40. Kara, D. [Kara, Gyorgy]. Knigi mongol'skikh kochevnikov. Moscow: Nauka, 1972. Kara, D., E. I. Kychanov, and V. S. Starokov, "Pervaia nakhodka chzhurchzhen'skikh rukopisnykh tekstov na bumage." Pis'mennye Pamiatniki Vostoka, 1969, pp. 223-38. Karmay, Heather. Early Sino-Tibetan art. Warminster: Aris and Phillips, 1975. Kates, George N. "A new date for the origins of the Forbidden City." Harvard Journal of Asiatic Studies, 7(1943), pp. 180-202. Kato Shigeshi M B 38 . Shina keizaishi kosho 3fc ffl Mffi56 % i@ . 2 vols. Toyo bunko ronso f i ^ X t f i l f no. 34. Tokyo: Toyo bunko, 1952-3. Katsufuji Takeshi § f S . Fubirai kan M- >& f\ W . Tokyo: Jimbutsu shuraisha, 1966. Kawazoe Shoji. "Japan and East Asia." In Medieval Japan, ed. K020 Yamamura, vol. 3 of The Cambridge history of Japan. Cambridge: Cambridge University Press, 1990, pp. 396-446. Kawazoe, Shoji Jl| W BB — .Moko shiirai kenkyii shiron^L ~£j ${ 3)5 W ^E S& Ml- Tokyo: Yuzankaku, 1977. Kennedy, E. S. "The exact sciences in Iran under the Saljuqs and Mongols." In The Saljuq and Mongol periods, ed. John A. Boyle, vol. 5 of The Cambridge history of Iran. Cambridge: Cambridge University Press, 1968, pp. 659-79. Khazanov, Anatoli M. Nomads and the outside world. Trans. Julia Crookenden. Cambridge: Cambridge University Press, 1984. Kirakos, Gandzaketsi. Istoriia Armenii. Trans. L. A. Khanlarian. Moscow: Nauka, 1976. Kiselev, Sergei V., ed. Drevnemongol'skie goroda. Moscow: Nauka, 1965. Kiyose, Gisaburo N. A study of the Jurchen language and script: Reconstruction and decipherment. Kyoto: Horitsu bunkasha, 1977. K'o Shao-min $J &B iS . Hsin Yuan shih #f 7t £ . Tientsin: T'ui keng t'ang Jl $t S , 1922; 2nd rev. ed. Peking: 1930. Repr. in Erh shih wu shih I f l S , ed. Erh shih wu shih k'an hsing wei yuan hui — + H £ : P J f T § i J l # . Shanghai, 1935; repr. Taipei: K'ai-ming shu-tien, 1962—9. Komai Kazuchikai 15 # ft] 3? . "Gen no Joto narabi ni Daito no heimen ni tsuite 7t <T)± S5 M If il * SB <T> ¥ ® i; I t l> T ." Tda ronso M 55 31 £ , 3 (1940), pp. 129-39. Kotwicz, Wladyslaw, "Les Mongols, promoteurs de l'idee de paix universelle au debut du XH-e [sic] siecle." In La Pologne au Vl-e congres international des sciences historiques. Warsaw: 1933, pp. 199—204. Repr. in Rocznik Orientalistyczny, 16 (1950, actually published 1953), pp. 428-34. Koyama Fujio <ds iXl % ± ^ . "Pasupa moji aru Shina furu toji A I E X ? i •§ 5 " Gasetsu fttft, 1 (1937), pp. 23—31. Cambridge Histories Online © Cambridge University Press, 2008 75° BIBLIOGRAPHY Kracke, Edward A. Civil service in early Sung China, 960-1067. Harvard-Yenching Institute Monograph Series no. 13. Cambridge, Mass.: Harvard University Press, Krader, Lawrence. "Feudalism and the Tatar polity of the Middle Ages." Comparative Studies in Society and History, 1 (1958-9), pp. 76-99. Ku Yen-wu W i& 5& , ed. Huang Ju-ch'eng M ikfiS. Jih chih lu chi shih: wai ch'i cbung B ftl H ^ P : ft-kM . Shanghai: Shang-hai jen-min ch'u-pan she, 1985. K'uang Yu-ch'e M. ¥B Wt. "Pai-chu chi ch'i hsin cheng W tt J& K Sffl&."Nei Meng-ku she hui k'o hsueh ft SI £ tt t Pi 9 , 1984:5, pp. 59-62. Kubo, Noritada. "Prolegomena on the study of the controversies between Buddhists and Taoists in the Yuan period." Memoirs of the Research Department of the Toyo bunko, 25 (1967), pp. 39—61. K'ung Ch'i (K'ung K'e-ch'i) ?L ?¥ . Chih cheng chih chi S IE S IB . In Yiiehya fang ts'ung shu # fl !§£ iK ft , ed. Wu Ch'ung-yao ffi # $1. Repr. Taipei: I-wen yinshu kuan, 1965; vols. 321-4. Kuwabara Jitsuzo. "On P'u Shou-keng iff #f S : A man of the western regions who was the superintendent of the Trading Ships' Office in Ch'iian-chou towards the end of the Sung dynasty, together with a general sketch of trade of the Arabs in China during the T'ang and Sung eras." Memoirs of the Research Department of the Toyo bunko, 2 (1928), pp. 1-79; 7 (1935), pp. 1-104. Kychanov, Evgenii I. "From the history of the Tangut translations of the Buddhist canon." In Tibetan and Buddhist studies commemorating the 200th anniversary of the birth of Alexander Csoma de Korb's, ed. Louis Ligeti. Budapest: Akademiai Kiad6, 1984, pp. 377-87Kychanov, Evgenii I. Izmennyii zanovo utverzhdennyi kodeks devisa tsarstvovaniia nebesnoe protsvetanie (1149-1169). Vol. 1, Moscow: Nauka, 1988; Vol. 2, Moscow: Nauka, 1987; Vol. 3, Moscow: Nauka, 1989; Vol. 4, Moscow: Nauka, 1989. Kychanov, Evgenii I. "Les Guerres entre les Sung du nord et le Hsi-Hsia." In Etudes Song in memoriam Etienne Baldzs, ed. Francoise Aubin, 2nd series, no. 2. Paris: Mouton, 1971, pp. 106—18. Kychanov, Evgenii I. "Mongol-Tangutskie voiny i gibel' gosudarstva Si Sia." In Tataro—Mongoly v Azii i Evrope, ed. S. L. Tikhvinskii. 2nd ed. Moscow: Nauka, 1977. PP- 46-61. Kychanov, Evgenii I. "Mongoly v Vl-pervoi polovine XII v." In Dal'nii Vostok i sosednie territorii v srednie veka, ed. V. E. Larichev. Novosibirsk: Nauka, 1980, pp. 136-48. Kychanov, Evgenii I. "Monuments of Tangut legislation (12th-13th centuries)." In Etudes Tibetaines. Actes du XXIXe Congres international des Orientalistes, July 1973. Paris: L'Asiatheque, 1976, pp. 29—42. Kychanov, Evgenii I. "O nekotorykh naimenovaniiakh gorodov i mestnostei byvshei territorii Tangutskogo gosvdarstva." In vol. 1 of Pis'mennye pamiatniki i problemy istorii i kul'tury naradov vostoka: XL Godichnaia nauchnaia sessia LO IV. AN. SSSR {Tezisy). Moscow: Nauka, 1975, pp. 47—51. Cambridge Histories Online © Cambridge University Press, 2008 BIBLIOGRAPHY 751 Kychanov, Evgenii I. Ocherk istorii tangutskogo gosudarstva. Moscow: Nauka, 1968. Kychanov, Evgenii I. "Svod voennykh zakonov Tangutskogo gosudarstva 'Iashmovoe Zertsala upravleniia let tsarstvovaniia Chzhen' — Kuan' (1101 —1113)." In Pis'mennyepaniiatniki vostoka, 1969. Moscow: Nauka, 1972, pp. 2 2 9 - 4 3 . Kycanov {Kychanov], Evgenii I., and Herbert Franke. Tangutische und chinesische Quellen zur Militdrgesetzgebung des 11. bis 13. Jahrhunderts. Munich: Bayerischen Akademie den Wissenschaften, Philosophische—Historische Klasse Abhandlungen: Neue Folge, vol. 104, 1990. Kychanov, Evgenii I. Vnov' sobrannye dragotsennyeparnye izrecheniia. Moscow: Nauka, 1974. ^ Kyoto daigaku jimbun kagaku kenkyujo Gentensho kenkyu han B. $$ ^k. ^ A X ?4 ^W^E0f7CftSW5E , comp. Gentensho sakuin-ko zoku han 7C ft % % 31 $5 . Kyoto: Kyoto daigaku jimbun kagaku kenkyujo, 1957. Repr. as Gentensho sakuin-ko. Taipei: Wen-hai ch'u-pan she, 1973. Laing, Ellen Johnston. "Patterns and problems in later Chinese tomb decoration." Journal of Oriental Studies, 16(1978), pp. 3—20. Lam, Ruby. "The role of shu-yuan in Yuan China." Unpublished paper. Lam, Yuan-chu. "On the Yuan examination system: The role of northern Cheng-Chu pioneering scholars." Journal ofTurkish Studies, Festschrift for Francis W. Cleaves, 9 (1985), pp. 15-20. Lamb, H. H. Climate: Present, past and future. 2 vols. London: Methuen, 1977. Langlois, John D . , Jr., ed. China under Mongol rule. Princeton, N J . : Princeton University Press, 1981. Langlois, John D . , Jr. "Political thought in Chin-hua under Mongol rule." In China under Mongol rule, ed. John D. Langlois, Jr. Princeton, N.J.: Princeton University Press, 1981, pp. 1 3 7 - 8 5 . Langlois, John D . , Jr. "Yii Chi and his Mongol sovereign: The scholar as apologist." Journal of Asian Studies, 38 (1978), pp. 99-116. Lao, Yan-shuan. "The Chung-t'ang shih-chi of Wang Yiin: An annotated translation with an introduction." Ph.D. diss., Harvard University, 1962. Lao, Yan-shuan. "Southern Chinese scholars and educational institutions in early Yuan: Some preliminary remarks." In China under Mongol rule, ed. John D. Langlois, Jr. Princeton, N J . : Princeton University Press, 1981, pp. 107-33. Ledyard, Gari. "The Mongol campaigns in Korea and the dating o( the Secret history of the Mongols." Central Asiatic Journal, 9(1964), pp. 1-22. Lee, Ki-baik. A new history of Korea. Trans. Edward W. Wagner. Cambridge, Mass.: Harvard University Press, 1984. Lee, Sherman E., and Wai-kam Ho. Chinese art under the Mongols: The Yuan dynasty (1279-1368). Cleveland: Press of Case Western Reserve University, 1968. Len'kov, Vitalii D. Metallurgiia i metalloobrabotka u Chzhurchzhenei v XII veka (po materialam issledovanii Shaiginskogo gorodishcha). Novosibirsk: Nauka, 1974. Leslie, Donald D. The survival of the Chinese Jews: The Jewish community of K'ai-feng. Leiden: Brill, 1972. Lewis, Bernard. "Egypt and Syria." In The central Islamic lands, ed. P. M. Holt, Ann Cambridge Histories Online © Cambridge University Press, 2008 752 BIBLIOGRAPHY K. S. Lamb ton, and Bernard Lewis, vol. iA of The Cambridge history of Islam. Cambridge: Cambridge University Press, 1970, pp. 175-230. Li Fan-wen. See also Pu P'ing. Li Fan-wen ^ ?§~X."Hsi Hsia i min tiao ch'a chi ffi S ft S M 38 13 ." Ning-hsia she hut k'o hsiieh 9 M tt # * 4 * , 1981:1, pp. 38-62. Li Fan-wen. W// HM» j w cA/« lun chi 2§ 3fiff% W& IB . Yin-ch'uan: Ning-hsia jen min ch'u pan she, 1983. Li Fan-wen. "T'ung-yin" yen chiu |H] "& 9f %. Yin-ch'uan: Ning-hsia jen min ch'u pan she, 1986. Li Han ^ JS and Shen Hsiieh-ming it 4P W . "Liieh lun Hsi tsu tsai Liao tai te fa chan mm %m&M ft &!&&." Sung Liao Chin shih lun ts'ung * $ 8 £ £ & * , 1 (1985), pp. 277-94. Li Hsi-hou ^ $§ J?. "Shih lun Liao tai Yii-t'ien Han shih chia tsu te li shih ti wei IS H$gtt£ffl#ft£;8!Wag£ififfi ." Sung Liao Chinshih lun ts'ung * & & £ t i l l , 1 (1985), pp. 251-66. Li Hsin-ch'uan $ '0 ff. Chien-yen i lai ch'ao yeh tsa chi i t ifc £A 3f5 91ffiF$g IB. Repr. in vols. 21—2 of Sung shih tzu liao ts'ui pien ^•^S.'S.^WM, 1st series, ed. Chao T'ieh-han t& $8 * . Taipei: Wen-hai ch'u-pan she, 1967. Also KHCPTS ed. Li Kan 3* # . Yuan tai she hui ching chi shih kao 7 U f ^ % t # ^ P £ I S . Wuhan: Hupei jen-min ch'u-pan she, 1985. Li Pai-yao $ 6 M et al., eds. P« CA1/ shu % ^ # . Peking: Chung-hua shu-chu, 1972 ed. Li Ssu-ch'un $ ® #fi . y»«« 1/&/6 &»<>/& 7C i # . Shanghai: Chung-hua shu-chii, 1926. Li T'ang 2£ fll . V«rf» Shih-tsu 7C 1Sffl. Hong Kong: Hsiang-kang Hung-yeh shuchii, 1978; Taipei: Ho-Lo t'u-shu ch'u-pan she, 1978. Li T'ao ^ M. et al., comps. H;« /z» chih t'ung chien ch'ang pien M 'S fe IS ^ J* SI . Peking: Chung-hua shu-chii, 1979-ongoing ed. Also Chekiang, 1881; repr. Taipei: Shih-chieh shu-chii, 1961. Li Tse-fen ^ iJ 3F . Ch'eng-chi-ssu han hsin chuan BE & ca rT ff W . Taipei: Chunghua shu-chu, 1970. Li Tse-fen. Y«<*» iA/'/6 Aj»« fA/awg 7C £ §f i l . 5 vols. Taipei: Chung-hua shu-chii, 1978. Li Wei ^ i£f. Hsi Hsia shih yen chiu 2§ S i 5f ^£ . Yin-ch'uan: Ning-hsia jen min ch'u pan she, 1989. Li Wen-hsin ^ X is. Liao-ning shengpo wu kuan ts'ang Liao tz'u hsiian chi H ^ 4i W 1^J fil Iffi M ^ S ^ . Peking: Wen-wu ch'u-pan she, 1962. Li Yuan $ $8 . Ch'ing t'ang lu W #t & . 12th century. In oiwaw 35 oiShuofu aft S5, comp. T'ao Tsung-i M ^ $$ . Fascim. repro. Taipei: Shang-wu yin-shu kuan, 1972. ^ ^ Lwo CA/n Yuan chuan chi san shih chung tsung hoyinte SSsfeTCfUfB — + S ^ ' a ' ? I W 1 (Combined indices to thirty collections of Liao, Chin, and Yuan biographies). Harvard—Yenching Institute Sinological Index Series, vol. 35. Peking: 1940. Liao Lung-sheng B S $£. "Pei Sung tui T'u-fan te cheng ts'e Cambridge Histories Online © Cambridge University Press, 2008 BIBLIOGRAPHY 753 Kuo li T'ai-wan Shih-fan la hsiieh li shih hsiiehpao (1976), pp. 141-77Liao Lung-sheng. "Sung Hsia kuan hsi chung te ch'ing pai yen wen t'i ^ JC BS % 41 fit eiF»1I ." Shih huo yiieh k'an & « ft TO, 5 (1976), pp. 462-9. Liao shih Mi.. See T'o-t'o. Lie, Hiu. Die Mandschu-Sprachkunde in Korea. Indiana University Publications, Uralic and Altaic Series, vol. 114. Bloomington: Indiana University Press, 1972. Ligeti, Louis. "Les inscriptions djurtchen de Tyr. La formule om mani padme hum." Ada Orientalia Academiae Scientarum Hungaricae, 12 (1961), pp. 5-26. Ligeti, Louis. "Les Noms mongols de Wen-tsong des Yuan." T'oung Pao, 27 (1930), pp. 57-61. Ligeti, Louis. "Note pr&iminaire sur le dechiffrement des petits caracteres Joutchen." Ada Orientalia Academiae Scientarum Hungaricae, 3 (1953), pp. 221-8. Ligeti, Louis. "Le Tabghatch, un dtalecte de la langue Sien-pi." In Mongolian studies, ed. Louis Ligeti. Amsterdam: B. R. Griiner, 1970, pp. 265—308. Lindner, Rudi Paul. "What was a nomadic tribe?" Comparative Studies in Society and History, 24 (1982), pp. 689-711. Liu Ch'i 81 fl$, comp. Kuei ch'ien chih If ?H i& . Repr. Peking: Chung-hua shu-chii, 1983. Liu Feng-chu 89 A SS and Yu Pao-lin I S # . "Nii-chen wen tzu Ta Chin te sheng t'o sung chiao k'an chi &M~$t^l:&'&B?&i&tfl.&} si . "In Min tsu yu wen lun chi S M B§ ef l i &, ed. Min tsu yu wen pien chi pu g M i§ X S ft 95 . Peking: Chung-kuo she-hui k'o-hsiieh ch'u-pan she, 1981, pp. 292-344. Liu Hsii §1J Hfej et al., eds. Chiu Tang shu B )9 18 . Peking: Chung-hua shu-chii, 1975 ed. Liu Min-chung SI W. 4 1 . CA»»g «« chi 4" 31 ^ . SKCS ed.; also repr. in vol. 1206 of Ying yin Wen-yuan ko Ssu k'u ch'iian j / 6 « p £ n X i i ^ E 3 ^ ^ : * . Taipei: Shang-wu yin-shu kuan, 1986, pp. 1—200. Liu Ming-shu £!! IS f& . "Yuan tai chih hu k'ou ch'ing ts'e 7C ft Z P P # ffi ." Chung-kuo wen hua yen chiu hui k'an ^ H X f t S f ^ E A f J , 7 (1947), pp. 101-6. Liu Ying-sheng 841 SB S#. "Shih chi Wo-k'uo-t'ai han kuo mo nien chi shih pu cheng £.MT%ffl-aiPrM3iit&'m&W. . " Y«*» JA/A C/6/ peifang min tsu shih yen chiu chi k'an 7C £ & i t ?? K & £ PF 2E JH flj , 10 (1986), pp. 48-59. Liu Ying-sheng. "Yuan ch'ao yii Ch'a-ho-t'ai han kuo te kuan hsi 7C 19 J&l IS •& a ff a W B3 l i ." y«a» JA//& /«« /J'«»£ 7C i i t ^ , 3 (1986), pp. 56-81. Liu Ying-sheng. "Yuan tai Meng-ku chu han kuo chien te yiieh ho chi Wo-k'uo-t'ai han kuo te mieh wang T C f t l l ^ s i f r f f l ^ l f t i f t f f i R & i l i & i f H t f J i g c t : ." Hsin-chiang ta hsiieh hsiieh pao Df II 'X IP ^ $3 , 2 (1985), pp. 31-43. L;a y« j/M c/&; 89 M * S . T5CC ed. Lo Chi-tsu P ISffl.,comp. Liao Han ch'en shih hsipiao M iH & tS ^ IS. Repr. as no. 35 in vol. 4 of Liao shih hui pien ffl £ Ik JH , ed. Yang Chia-lo ^§ f^ IS . Taipei: Ting-wen shu-chii, 1973. Lo Chi-tsu. Liao shih chiao k'an chi S £ 131 W) IB . 1st ed., 1938; 2nd rev. ed., Shanghai, 1958; repr. in Sung Liao Chin Yuan ssu shih tzu liao ts'ung k'an ^ JS ife 7C Cambridge Histories Online © Cambridge University Press, 2008 754 BIBLIOGRAPHY E3 £ "§ Pi W. fl] , ed. Chao T'ieh-han t& i t ^ . Taipei: Ta-hua yin-shu kuan, 1971, pp. 261-515. Lo Ch'iu-ch'ing M 3=fc & . "Sung Hsia chan cheng chung te fan pu yii pao chai ^ M.m&tt>t&wmmmm :• ch'ung-cbihsuehpao msmm,6< 1966-7), PP . 223-43. Lo Hsien-yu l i f t . "Yiian ch'ao chu ti Han hua shu i J S U t t ^ S i f b M i l t . " Min tsu yen chiu S S B t ^ S , 1987:5, pp. 67—74. Lo Jung-pang. "The controversy over grain conveyance during the reign of Qubilai Qaqan, 1260—94." Far Eastern Quarterly, 13 (1954), pp. 263—85. Lo Jung-pang. "Maritime commerce and its relation to the Sung navy." Journal ofthe Economic and Social History of the Orient, 12 (1969), pp. 57-101. Lou Yao ® m . Pet hsingjih lu % ft 0 ft . TSCC ed. Lu Chun-ling It t$£ ^t and Ho Kao-chi {5J jgj g|. "Yuan tai te A-su, Ch'in-ch'a, K'ang-li jen 76 ft #J P? 8 • & ^ • fli M A ." U^« shih X * , 16 (1982), pp. 117-30. Lu Ming-hui S. BMft. "San shih nien lai Chung-kuo Meng-ku shih yen chiu kai k'uang H + ^ J U ^ S S S i S f ^ E W S i ."In Meng-ku shih yen chiu lun wen chi W~& £.ffi$csfc3i:M ,ed. Lu Ming-hui et al. Peking: Chung-kuo she-hui k'ohsueh ch'u-pan she, 1984, pp. 240—5. Lynn, Richard J. Kuan Yiin-shih. Boston: Twayne, 1980. Ma Tsu-ch'ang H ffi ft . Ma Shih-t'ien wen chi M, E ffl %. M . Ming (1368-1644) ed.; repr. in vol. 6 of Yiian jen wen chi chen pen ts'ung k'an T c A X S l ^ ^ i B f J , ed. Wang Te-i 3ifc!§l£ . Taipei: Hsin-wen feng ch'u-pan kung-ssu, 1985, pp. 509—681. Ma Tuan-lin JS *S B§ . Wen hsien t'ung k'ao & JK S # . In SM /'«»^ + JS . Shanghai: Shang-wu yin-shu kuan, 1936. Maeda Masana WifflIE =S . &«<«' »o rekishi-chirigakuteki kenkyu MBW W i ^ E . Tokyo: Yoshikawa kobunkan, 1964. Maeda Naonori Id B9 IE ft . "Gencho jidai ni okeru shihei no kachi hendo % t l S t t t - S & W F C O f f l ^ t ^ i l ."In Maeda Naonori, GwrAo shi no kenkyu <T> ff % . Tokyo: Tokyo daigaku shuppankai, 1973, pp. 107-43. Maejima Shinji wl Hll 1a # . "Senshu no Perushiyajin to Ho Juko Sffl f i H ^ ." Shigaku £ ^ , 25 (1952), pp. 256-321. Makino Shuji ft H5 i^ — . Gendai kotokan no taikeiteki kenkyu 7 t f t ^ ) # ' S W i i ^ W fiff 5£ . Tokyo: Taimedo, 1979. Makino, Shuji. "Transformation of the shih-jen in the late Chin and early Yiian." Ada Asiatica, 45 (1983), pp. 1-26. Mangold, Gunther. Das Militdrwesen in China unter der Mongolenherrschaft. Bamberg: Aku Fotodruck, 1971. Martin, Henry Desmond. The rise of Chingis khan and his conquest of north China. Baltimore: Johns Hopkins University Press, 1950; repr. New York: Octagon, 1971. Marugame Kinsaku % H j& fF. "Korai to So to no tsuko mondai iU M h ^ t <D 81 ." Chosen gakuhdmU ^ S , 17(1960), pp. 1-50; 18(1961), pp. 58-82. Cambridge Histories Online © Cambridge University Press, 2008 BIBLIOGRAPHY 755 Matsui Shuichi fe # 3f — . "Roryu hanchin ko I t M IF £ % ." M/g«ife« z<axr/W * ^ U S , 68(1959), pp. 1397-1432. Matsuda Koichi fS EH # — . "Gencho chi no bumposei — Anseio no jirei wo chushin to shite K,m%i<F>ft&lffl -fl&3.<n&m£'P>b£ I T " Shigaku zasshi j £ m&U , 88 (1979), pp. 1249-86. Matsuda Koichi. "Kaishan no seihoku Mongoria shutsujin il ^, •> "t' > GO H i t t y 3* U T * m ." Tohogaku M ft m , 64 (1982), pp. 73-87. Matsuura Shigeru 1SffllIS,. "Kindai Joshin shizoku no kosei ni tsuite sk f^ iZ H R S W i i C O U t •" Toyoshi kenkyu M ?¥ £ iff ?E , 36, no. 4 (March 1978), pp. 1-38. Medley, Margaret. Yiian porcelain and stoneware. New York: Pitman, 1974. Medvedev, Vitalii E. Kul'tura Amurskikh Chzhurchzhenei konets X—Xl vek (po materialam gruntovykh mogil'nikov). Novosibirsk: Nauka, 1977. Meng Ssu-ming W S 99 . Yiian tai she hut chieh chi chih tu 7C f£ l i # PW IS $!j §1 . Yenching Journal of Chinese Studies, Monograph series no. 16. Peking, 1938; repr. Hong Kong: Lung-men shu-tien, 1967. Meng Ta pei liieh W M ffi B§ . See Chao Hung. Mikami Tsugio H ± # ^ . Kindai Joshin shakai no kenkyu jfcft&Xtti^COfiff^u. Vol. 1 of Mikami Tsugio, Kinshi kenkyu ife i£ 5f <5E . Tokyo: Chuokoron bijitsu shuppan, 1972. Mikami Tsugio. Kindai seijiseido no kenkyu & ft jgf fe $ij S CO 57f ^E . Vol. 2 of Mika mi Tsugio, Kinshi kenkyu ^ i 9f 5E • Tokyo: Chuokoron bijitsu shuppan, 1972. Mikami Tsugio. Kindai seiji, shakai no kenkyu & f t i & ? p t t W < 7 ) 9 f 2 5 • Vol. 3 of Mikami Tsugio, Kinshi kenkyu ife $. 9F 5E . Tokyo: Chuokoron bijitsu shuppan, 1973Ming shih BM £ . See Chang T'ing-yii. Minhaj al-DIn Juzjanl. Tabaqat-i nastri. Ed. W. Nassau Lees. Calcutta: Calcutta College Press, 1964. Minhaj al-DIn Juzjanl. Tabaqat-i nastri. 2 vols. Trans. H. G. Raverty. New Delhi: Oriental Books Reprint Corporation, 1970. Mino, Yutaka. Ceramics in the Liao dynasty: North and south of the Great Wall. New York: China Institute in America, 1973. Miyazaki Ichisada "Bflfft? 5£ . "Gencho chika no MSkoteki kanshoku wo meguru Mo Kan kankei - kakyo fukko no igi no saikento 7 C $ t e T c 7 ) K * W & $ £ £ a 6 <' %mmm% - f 4 $ I S ( 0 S i W S « t t .-Toyoshi kenkyu Mft&m9c , 2 3 (1965). PP- 428-91. Mole, Gabriella. The T'u-yii-hun from the Northern Wei to the time of the Five Dynasties. Rome: Istituto Italiano per il Medio ed Estremo Oriente, 1970. Moriyasu Takao & % # ^ . "Uiguru to Tonko •b 4 7 )V t Sc 41 . " I n Tonko no rekishi S I W I S , vol. 2 oiKoza Tonko i t & Wi. !§ , ed. Enoki Kazuo W. — W.. Tokyo: Dai to shuppansha, 1980, pp. 297-338. Moses, Larry. "A theoretical approach to the process of Inner Asian confederation." Etudes Mongoles, 5 (1974), pp. 113—22. Mostaert, Antoine, and Francis W. Cleaves. Let Lettres de 1289 et 1305 des ilhan Cambridge Histories Online © Cambridge University Press, 2008 BIBLIOGRAPHY Arghun et Oljeitii a Phillipe It Bel. Cambridge, Mass.: Harvard University Press, 1962. Mote, Frederick W. "Confucian eremitism in the Yuan period." In The Confucian persuasion, ed. Arthur F. Wright. Stanford, Calif: Stanford University Press, 1960, pp. 202—40. Mote, Frederick W. "The growth of Chinese despotism: A critique of Wittfogel's theory of oriental despotism as applied to China." Oriens Extremus, 8 (1961), pp. 1-41. Mote, Frederick W. "Yuan and Ming." In Food in Chinese culture: Anthropological and historical perspectives, ed. K. C. Chang. New Haven, Conn.: Yale University Press, 1977. PP- I 9 5 - 2 5 7 Mote, Frederick W., and Denis C. Twitchett, eds. The Ming dynasty, 1368-1644, pt. i , vol. 7 of The Cambridge history of China. Cambridge: Cambridge University Press, 1988. Mote, Frederick W., and Denis C. Twitchett, eds. The Ming dynasty, 1368—1644, pt. 2, vol. 8 of The Cambridge history of China (forthcoming). Moule, Arthur C. Christians in China before the year 1550. London: Society for Promoting Christian Knowledge, 1930. Moule, Arthur C. Quinsai, with other notes on Marco Polo. Cambridge: Cambridge University Press, 1957. Moule, Arthur C. The rulers of China. London: Routledge & Kegan Paul, 1957. Moule, Arthur C , and Paul Pelliot. Marco Polo: The description of the world. 2 vols. London: Routledge and Sons, 1938. Mullie, Joseph L. "Les Anciennes Villes de l'empire des grands Leao au royaume mongol de Barin." T'oung Pao, 21 (1922), pp. 105-231. Mullie, Jozef [Joseph] L. De Mongoolse prins Nayan. Mededelingen van de Koninklyke Vlaamse Academie voor Wetenschappen, Letteren en Schone Kunsten van Belgie. Klasse der Letteren, vol. 26, no. 3. Brussels: Paleis der Academien, 1964. Mullie, Joseph [Jozef] L. "Une planche a assignats de 1214." T'oungPao, 33(1937), pp. 1 5 0 - 7 . Mullie, Joseph L. "Les Sepultures de K'ing des Leao." T'oung Pao, 30 (1933), pp. 1-25. Munkuev, Nikolai TS. Kitaiskii istochnik 0 pervykh mongol'skikh khanakh. Moscow: Nauka, 1965. Munkuev, Nikolai TS. "Zametki o drevnikh mongolakh." In Tataro-Mongoly v Azii i Europe, ed. S. L. Tikhvinskii. 2nd ed. Moscow: Nauka, 1977, pp. 377-408. Murakami Masatsugu fcf -h IE — . "Gencho ni okeru senfushi to attatsu 7C iji? IC St -It -5 M Jfr e] £ f$ J&, ." Toho gakuho (Tokyo) M 3fr # WL , 13 (1942), pp. 143-96. Murakami Masatsugu. "Gencho ni okeru toka no igi 76 $ ] | C fit t t ^ & T CO S ^ •" Moko gakuho ^ ' f i ' ^ S , 1 (1940), pp. 169-215. Murakami Masatsugu. "Mongoru cho chika no hoyusei no kigen ^ 7 3 )]/ IB $3 T <T) H g, $!] cr> ® j® ." Toyo gakuho M # ¥ ^ , 44 (1961), pp. 305-39. My dans, Shelley, and Carl Mydans. "A shrine city, golden and white: The seldomvisited Pagan in Burma." Smithsonian Magazine, October 1974, pp. 72—80. Cambridge Histories Online © Cambridge University Press, 2008 BIBLIOGRAPHY 757 Nacagdorz, S. "L'Organisation sociale et son deVeloppement chez les peuples nomades d'Asie Centrale." ttudes Mongoles, 5 (1974), pp. 135-44! Nakano, Miyoko. A phonological study in the 'Phags-pa script and the Meng-ku tzu yiin. Canberra: Australian National University Press, 1971. Nan-ching ta hsiieh li shih hsi Yuan shih yen chiu shih S a ^ ^ U ^ x S B f % H , comp. Yuan shih lun chi yt $. H $k . Peking: Jen-min ch'u-pan she, 1984. Needham, Joseph. Clerks and craftsmen in China and the West. Cambridge: Cambridge University Press, 1970. Needham, Joseph. "Medicine and Chinese culture." In Joseph Needham, Clerks and craftsmen in China and the West. Cambridge: Cambridge University Press, 1970, pp. 263-93. Needham, Joseph, et al. Heavenly clockwork: The great astronomical clocks of medieval China. 2nd ed. Cambridge: Cambridge University Press, 1986. Needham, Joseph, et al. Civil engineering and nautics, pt. 3 oi Physics and physical technology, vol. 4 of Science and civilization in China. Cambridge: Cambridge Univer- sity Press, 1971. Nevskii, Nikolai A. Tangutskaiafilologiia.2 vols. Moscow: Izdat. vostochnoi literatury, i960. Nieh Ch'ung-ch'i M ^ lK. "Sung Liao chiao p'ing k'ao ^5 j@ 55 Jfl # . " Yen-ching hsiieh pao M JK ^ ^ , 27 (1940), pp. 1-51. Repr. in vol. 2 of Nieh Ch'ung-chi, Sungshih ts'ung k'ao ^ $. 31 # . Peking: Chung-hua shu-chii, 1980, pp. 283-387. Nien-ch'ang & # . Fo tsu li tai t'ung tsai WfflM ft JS R . In vol. 49 of Taishoshinshu daizokyo ;*C IE ff 1^ * l& H ,ed. Takakusu Junjiro % ^ IH ?K g|5 and Watanabe Kaikyoku jS JB M 1&. Tokyo: Taisho issaikyo kankokai, 1929-34, pp. 477-735. Niida Noboru C # B3 S . Chiigoku hosei shi kenkyu: Keiho ^ S r S f t J i W ^ : ! ] ^ . Tokyo: Tokyo daigaku shuppankai, 1959. Nishida Tatsuo S EH M M . Seikago no kenkyu: Seikago no sai kosei to Seika moji no kaidoku mnmomft •• ^w.m<r>nm^.fSMX^<nm^ .2vois. Kyoto: Zayuho kankokai, 1964—6. Nogami Shunjo I? -h ft W>. "Gendai dobutsu niky5 no kakushitsu 7E ft j® % — #fe (D B f l . " Otani daigaku kenkyu nempo ;*: & * P ffi 5E ^ $B, 2 (1943), pp. 213-65. Nogami Shunjo. "Gen no senseiin ni tsuite jt <7) S iS K (I WL l> T •" 1° Haneda hakushi shoju kinen Toyoshi ronso 3 8 E B 1 f ± ® l l t 3 ^ * # i ^ ^ . Kyoto: Toyoshi kenkyukai M W £. 9f % # , 1950, pp. 779—95. Nogami Shunjo. Ryo Kin no Bukkyo JS ^ CD ^ tfc . Kyoto: Heirakuji shoten, 1953. Noguchi Shuichi W • M ~~. "Gendai kohanki no ogo juyo ni tsuite 7C ft t£ ^ S9 CO i l f c J g J I S K o ^ T ." Shigaku ^ ^ , 56(1986), pp. 53-83. Noguchi Shuichi. "Gendai Mushu cho no ogo juyo ni tsuite: Genshi Shoohyo ni kansuru ichi kosatsu T c f t ^ ^ W c O l ^ g ^ (C I t l» T T TEA J ^ 1 ^ C! BS i" •§ — ^ ^ . " In Ajia shominzoku ni okeru shakai to bunka: Okamoto Yoshiji sensei taikan kinen ronshu T V T&RM C ^ ft § tt # t X ft : K * ^ [ - ^ 4 Ji W S ^ «i $k , ed. Okamoto Yoshiji sensei taikan kinen ronshu kankokai H * ® - 5fe ^ jg t ffi ^ ^ m f!) tT # . Tokyo: Kokusho kankokai, 1984, pp. 271-305. Cambridge Histories Online © Cambridge University Press, 2008 758 BIBLIOGRAPHY Novey, Janet. "Yii Ching, a Northern Sung statesman, and his treatise on the Ch'itan bureaucracy." Ph.D. diss., Indiana University, 1983. Nung sang chi yao H Mi W I? . SPPY ed. Ohsson, Constantin M. d'. Histoire des mongols depuis Tchinguiz-khan jusqu'a Timour Bey ou Tamerlan. 4 vols. The Hague: Les Freres Van Cleef, 1834. Okazaki Seiro P3 «S iff SP . "Seika no Ri Genko to tokuhatsu rei S S CO $ 5c * £ J5M<$ ." Tohogaku M ~fi P , 19 (1959), pp. 77-86. Okazaki Seiro. Tanguto kodaishi kenkyu 5>y' *f' — h 1& f^ 5& • Kyoto: Kyoto daigaku Toyoshi kenkyukai, 1972. Okladnikov, Aleksei, P., and Anatolii P. Derevianko. Dalekoe proshloe Primor'ia i Priamur'ia. Vladivostok: 1973. Okladnikov, Aleksei P., and V. E. Medvedev. "Chzhurchzheni Priamur'iapodannym arkheologii." Problemy Dal'nego Vostoka, 1974:4, pp. 118—28. Olbricht, Peter. Das Postwesen in China unter der Mongolenherrschaft im 13. und 14. Jahrhundert. Wiesbaden: Otto Harrassowitz, 1954. Olschki, Leonardo. Guillaunu Boucher: A French artist at the court of the khans. Baltimore: Johns Hopkins University Press, 1946. Olschki, Leonardo. Marco Polo's Asia. Berkeley and Los Angeles: University of California Press, i960. Onogawa Hidemi ' h i ? JH ft 3S . Kinshi goi shusei 4 t g i l l S . 3 vols. Kyoto: Kyoto diagaku jimbun kagaku kenkyiijo, 1960-2. Osada Natsuki ft EH X ISt. "Joshin moji to genson shiryo i X X ? t Mfefm ." Rekishi kyoiku, 18, pt. 7 (1970), pp. 25—31. Oshima Ritsuko. "The chiang-hu in the Yiian." Ada Asiatica, 45 (1983), pp. 69—95. Otagi Matsuo f S S S . "Attatsu sen to sono haikei &|Si$8 £ £ <T)1M ft •" Toyoshi kenkyu 32, no. 1 (1973), pp. 1-27; 32, no. 2 (1973), pp. 23—61. Otagi Matsuo. Fubirai kan %$. & f[\ ff . Tokyo: Fuzambo, 1941. Otagi Matsuo. Kittan kodai ski no kenkyu IgftS'ft^COfiJf^E . Toyoshi kenkyu sokan M # A W % St T!J no. 6. Kyoto: Toyoshi kenkyukai, 1959. Otagi Matsuo. "Ri Dan no hanran to sono seijiteki igi: Moko cho chika ni okeru Kanchi no hokensei to sono shtikensei e no tenkai ^ t t C O ® l L i •? < Toyoshi kenkyu S ^ J B f t , 6 (August—September 1941), pp. 253-78. Ou-yang Hsiu © H i^ , comp. Hsin Wu-tai shih Sf H ft, £. (original title: Wu-tai shih chi E ft $. IB ). Peking: Chung-hua shu-chii, 1974 ed. Ou-yang Hsiu, and Sung Ch'i ^ ^ , comps. Hsin Tang shu M ff It . Peking: Chung-hua shu-chii, 1975 ed. Ou-yang Hsiian i i S . Kuei chai chi £. W * . SPTK ed. Pai Kang fi M. "Kuan yii Hu-pi-lieh fu hui Han fa te li shih k'ao ch'a M Hfc M >& %\ I t l S W i i t i ." Chung-kuo shih yen chiu 41 M & W % , 4 (1981), pp. 93-107. Pai Pin S % , ed. Hsi Hsia shih lun wen chi S I $ i 3 t l . Yin-ch'uan: Ning-hsia jen-min ch'u pan she, 1984. Pai Pin, and Shih Chin-po 5 ^ JS. "Ta Yiian Su-chou lu Ye-k'o ta lu hua ch'ih shih hsi Cambridge Histories Online © Cambridge University Press, 2008 BIBLIOGRAPHY 759 chihpei * (l919)> PP- 6 8 - 8 0 . Pan, Jixing. "On the origin of rockets." T'oung Pao, 73 (1987), pp. 2 - 1 5 . Parker, Edward H. A thousand years of the Tartars. London and Shanghai: Kelly & Walsh, Ltd., 1895; 2nd ed. London: Kegan Paul, Trench, Tribner; and New York: Knopf, 1924; repr. New York: Dorset, 1987. Pelliot, Paul. Les Mongols et la papaute. Paris: Librairie Auguste Picard, 1923; repr. Peking: Licoph Service, 1939. Pelliot, Paul. Notes critiques d'histoire Kalmouke, vol. 1. Paris: Librairie d'Amerique et d'Orient, i960. Pelliot, Paul. Notes on Marco Polo. 3 vols. Paris: Imprimerie Nationale, Librairie Adrien-Maisonneuve, 1959, 1963, 1973. Pelliot, Paul. Review of E. Haenisch, "Die letzten Feldziige Cinggis Han's und sein Tod. Nach der ostasiatischen Ueberlieferung" (Asia Major, 9 {1933], pp. 5 0 3 51). T'oung Pao, 31 (1934), pp. 157-67Pelliot, Paul, and Louis Hambis, trans. Histoire des campagnes de Gengis Khan, Chengwou Ts'in-TchengLou. Leiden: Brill, 1951. P'eng Hsin-wei S2 ft ^ . Chung-kuo huo pi shih 41 D 1? 9? £ . Shanghai: Shang-hai jen-min ch'u-pan she, 1958. P'eng Pai-ch'uan tl> If HI . Tai-p'ing chih chi t'ung lei ± ¥ fa & $E HI . In vol. 10 of Shih yuan ts'ung shu M M IS t , ed. Chang Chiin-heng 51 $3 W . 1917. Repr. Taipei: Ch'eng-wen ch'u-pan she, 1966. P'eng Ta-ya % *. $£, comp., and Hsu T'ing & H , annot. Hei Ta shih liieh "§ . In Meng-ku shih liao ssu chung f S J S H l , ed. Wang Kuo-wei £ Peking, 1926; repr. Taipei: Cheng-chung shu-chii, 1962, 1975. Petech, Luciano. "Sang-ko, a Tibetan statesman in Yuan China." Acta Orientalia Academiae Scientiarum Hungaricae, 34 (1980), pp. 193-208. Petech, Luciano. "Tibetan relations with Sung China and with the Mongols." In China among equals: The Middle Kingdom and its neighbors, loth—i/ith centuries, ed. Morris Rossabi. Berkeley and Los Angeles: University of California Press, 1983, pp. 173-203. Peterson, Charles A. "First Sung reactions to the Mongol invasions of the north, 1211-1217." In Crisis and prosperity in Sung China, ed. John W. Haeger. Tucson: University of Arizona Press, 1975. Peterson, Charles A. "Old illusions and new realities: Sung foreign policy, 12171234." In China among equals: The Middle Kingdom and its neighbors, ioth-i4th centuries, ed. Morris Rossabi. Berkeley and Los Angeles: University of California Press, 1983, pp. 2 0 4 - 3 9 . Petrushevskii, Ilia P. "Pokhod mongol'skikh voisk v sredniuiu Aziiu v 1219-1244 gg. i ego posledstviia." In Tataro-Mongoly y Azii i Evrope, ed. S. L. Tikhvinskii. 2nd ed. Moscow: Nauka, 1977, pp. 107—39. Pinks, Elizabeth. Die Uiguren von Kan-chou in der friihen Sung-Zeit. Wiesbaden: Otto Harrassowitz, 1968. (Po-na peri) Erh shih wu shih ( H £ft $ ) — + £ £ . See Erh shih wu shih. Cambridge Histories Online © Cambridge University Press, 2008 760 BIBLIOGRAPHY Pope, John A. Fourteenth-century blue and white: A group of Chinese porcelains in the Topkapu Sarayi Muzesi, Istanbul. Washington, D.C.: Freer Galley of Art, 1952. Poppe, Nicholas. "Jurchen and Mongolian." In Studies on Mongolia: Proceedings of the first North American conference on Mongolian studies, ed. Henry G. Schwartz. Bellingham: Center for East Asian Studies, Western Washington University, 1979, pp. 30-7. Poppe, Nicholas, trans. The Mongolian monuments in 'Phags-pa script. Ed. John R. Krueger. Wiesbaden: Otto Harrassowitz, 1957. Poucha, Pavel. "Uber den Inhalt und die Rekonstruktion des ersten mongolischen Gesetzbuches." In Mongolian studies, ed. Louis Ligeti. Amsterdam: B. R. Griiner, 1970, pp. 377-415. Pu P'ing h ¥ (Li Fan-wen). "Hsi Hsia huang ti ch'eng hao k'ao I I I $ i i # . " Ning-hsia she hui k'o hsiieh * S tt # P, P , 1981:1, pp. 70-82. Pulleyblank, Edwin G. "A Sogdian colony in Inner Mongolia." T'oung Pao, 41 (1952), pp. 317-56. QashanI, Abu al-Qasim 'Abd Allah Ibn 'Ali. Tarikh-i (Jljaytu. Ed. Mahin Hambly. Tehran: Bungahi Tarjameh va Nashr-i Kitab, 1969. Rachewiltz, Igor de. "The Hsi-yu lu B j ^ & by Yeh-lu Ch'u-tsai W&&M ." Monumenta Serica, 21 (1962), pp. 1—128. Rachewiltz, Igor de. "More about the preclassical Mongolian version of the Hsiaoching." Zentralasiatische Studien, 19 (1986), pp. 27—37. Rachewiltz, Igor de. "Muqali, Bol, Tas and An-t'ung." Papers on Far Eastern History, 15 (1977). PP- 45-62. Rachewiltz, Igor de. "Personnel and personalities in north China in the early Mongol period."Journal ofthe Economic andSocialHistory ofthe Orient, 9(1966), pp. 88-144. Rachewiltz, Igor de. "The preclassical Mongolian version of the Hsiao-ching." Zentralasiatische Studien, 16(1982), pp. 7—109. Rachewiltz, Igor de. "Qan, qa'an and the seal of Giiyiig." In Documenta Barbarorum: Festchrift fur Walter Heissig zum 70. Geburststag, ed. K. Sagaster and M. Weiers. Wiesbaden: Otto Harrassowitz, 1983. Rachewiltz, Igor de, trans. "The Secret History of the Mongols." Papers on Far Eastern History, 4(1971), pp. 115-63; 5 (1972), pp. 149-75; JO (1974), pp. 5582; 13 (1976), pp. 41-75; 16 (1977). PP- 27-65; 18 (1978), pp. 43-80; 21 (1980), pp. 17-57; 23 (1981), pp. 111-46; 26 (1982), pp. 39-84; 30 (1984). pp. 81-160; 31 (1985), pp. 21-93. Rachewiltz, Igor de. "Some remarks on the ideological foundations of Chinggis khan's empire." Papers on Far Eastern History, 7 (1973), pp. 21-36. Rachewiltz, Igor de. "Some remarks on the language problem in Yuan China." Journal of the Oriental Society of Australia, 5 (1967), pp. 65-80. Rachewiltz, Igor de. "Some remarks on Toregene's edict of 1240." Papers on Far Eastern History, 23 (1981), pp. 38-63. Rachewiltz, Igor de. "Turks in China under the Mongols: A preliminary investigation of Turco-Mongol relations in the 13th and 14th centuries." In China among equals: The Middle Kingdom and its neighbors, 10th—14th centuries, ed. Morris Cambridge Histories Online © Cambridge University Press, 2008 BIBLIOGRAPHY 761 Rossabi. Berkeley and Los Angeles: University of California Press, 1983, pp. 281-310. Rachewiltz, Igor de. "Yeh-lii Ch'u-ts'ai (1189-1243): Buddhist idealist and Confucian statesman." In Confucian personalities, ed. Arthur F. Wright and Denis C. Twitchett. Stanford, Calif.: Stanford University Press, 1962, pp. 189-216. Rachewiltz, Igor de, et al. Index to biographical material in Chin and Yuan literary works. 1st series (with Miyoko Nakano). Canberra: Australian National University Press, 1970. 2nd series (with May Wang). Canberra: Australian National University Press, 1972. 3rd series (with May Wang). Canberra: Australian National University Press, 1979. Rachewiltz, Igor de, Hok-lam Chan, Hsiao Ch'i-Ch'ing, and Peter W. Geier, eds. In the service of the khan: eminent personalities of the early Mongol-Yiian period (12001300). Wiesbaden: Otto Harrassowitz, 1992. Rail, Jutta. Die vier grossen Medizinschulen der Mongolenzeit: Stand und Entwicklung der chinesischen Medizin in der Chin- und Yu'an-Zeit, Wiesbaden: Franz Steiner, 1970. Rashld &l-Dln.Jami' al-Tavdrikh, vol. 1, pt. 1. Ed. A. A. Alizade. Moscow: Nauka, 1968. Rashld a\-Dln. Jdmi' al-Tavdrikh. 2 vols. Ed. B. Karlml. Teheran: Iqbal, 1959. Rashid al-Dln. Sbornik letopisei, vol. 1, pt. 2, Trans. O. I. Smirnova. Leningrad: Nauka, 1952. Rashld al-Dln. Shu 'ab-ipanjgdnah. Manuscript, Topkapi Sarayi Museum, Catalogue no. 2932. Rashld al-Dln. The successors of Genghis khan. Trans. John A. Boyle. New York: Columbia University Press, 1971. Ratchnevsky, Paul. "Les Che-wei £taient-ils des Mongols?" In vol. 1 of Milangesdesinologie offerts a Monsieur Paul Demih/ille. Bibliotheque de l'lnstitut des Hautes Etudes Chinoises, vol. 20. Paris: Presses Universitaires de France, 1966, pp. 2 2 5 - 5 1 . Ratchnevsky, Paul. Cinggis-khan: Sein Leben und Wirken. Wiesbaden: Franz Steiner, 1983. Ratchnevsky, Paul. "Die mongolischen Grosskhane und die buddhistische Kirche." In Asiatica: Festschrift Friedrich Welter zum 65. Geburtstag, ed. Johannes Schubert. Leipzig: Otto Harrassowitz, 1954, pp. 4 8 9 - 5 0 4 . Ratchnevsky, Paul. "Uber den mongolischen Kult am Hofe der Grosskhane in China." In Mongolian studies, ed. Louis Ligeti. Amsterdam: B. R. Griiner, 1970, pp. 417-43Ratchnevsky, Paul. Un code des Yuan. 4 vols. Vol. 1: Paris: Librairie Ernest Leroux, 1937; vol. 2: Paris: Presses Universitaires de France, 1972; vol. 3: Paris: Presses Universitaires de France, 1977; vol. 4: Paris: College de France, Institut des Hautes Etudes Chinoises, 1985. Ratchnevsky, Paul. "Zum Ausdruck 't'ouhsia' in der Mongolenzeit." Collectanea Mongolica: Festschrift fur Professor Dr. Rintchen zum 60. Geburtstag. Wiesbaden: Otto Harrassowitz, 1966, pp. 173-91. Rerikh, Iurii N. (George N. Roerich). "Tangutskii titul dzha-gambu Kereitskogo." Kratkie soobshcheniia instituta naradov Azii, 44 (1961), pp. 41—4. Cambridge Histories Online © Cambridge University Press, 2008 762 BIBLIOGRAPHY Rockhill, William. The journey of William of Rubruck to the Eastern parts of the world. London: Hakluyt Society, 1900. Roerich, George. The blue annals. 2nd ed. Delhi: Motilal Banarsidass, 1976. Rogers, Michael C. "The late Chin debates on dynastic legitimacy." Sung Studies Newsletter, 13 (1977), pp. 57-66. Rogers, Michael C. "The myth of the battle of the Fei River (A.D. 383)." T'oung Pao, 54 (1968), pp. 50-72. Rogers, Michael C. "National consciousness in medieval Korea: The impact of Liao and Chin on Koryo." In China among equals: The Middle Kingdom and its neighbors, ioth—i4th centuries, ed. Morris Rossabi. Berkeley and Los Angeles: University of California Press, 1983, pp. 151—72. Rogers, Michael C. "The regularization of Koryo-Chin relations (1116-1131)." Central Asiatic Journal, 6(1961), pp. 51-84. Rogers, Michael C. "Studies in Korean history, II: Koryo's military dictatorship and its relations with Chin." T'oung Pao, 47 (1959), pp. 42-62. Rorex, Albright. "Some Liao tomb murals and images of nomads in Chinese paintings of the Wen-chi story." Artibus Asiae, 45 (1984), pp. 174-98. Rossabi, Morris, ed. China among equals: The Middle Kingdom and its neighbors, 10th— 14th centuries. Berkeley and Los Angeles: University of California Press, 1983. Rossabi, Morris. "Chinese myths about the national minorities: Khubilai Khan, a case study." Central and Inner Asian Studies, 1 (1987), pp. 47-81. Rossabi, Morris. "Khubilai Khan and the women in his family." In Studia SinoMongolica: Festschrift fur Herbert Franke, ed. Wolfgang Bauer. Munchener Ostasiatische Studien no. 25. Wiesbaden: Franz Steiner, 1979, pp. 153-80. Rossabi, Morris. Khubilai Khan: His life and times. Berkeley and Los Angeles: University of California Press, 1988. Rossabi, Morris. "The Muslims in the early Yuan dynasty." In China under Mongol rule, ed. John D. Langlois, Jr. Princeton, N.J.: Princeton University Press, 1981, PP- 257-95Rossabi, Morris. Voyager from Xanadu: Rabban Sauma and thefirstjourney from China to the west. New York: Kodansha, 1992. Roux, Jean-Paul. "Le Chaman gengiskhanide." Anthropos, 54 (1959), pp. 401-32. Rozman, Gilbert, ed. Soviet studies ofpre-modern China: Assessments of recent scholarship. Ann Arbor: Center for Chinese Studies, University of Michigan, 1984. Rozman, Gilbert. Urban networks in Ch'ing China and Tokugawa Japan. Princeton, N.J.: Princeton University Press, 1973. Saccheti, Maurizia Dinacci. "Sull'adozione del nome dinastico Yuan." Annali, Istituto Orientate di Napoli, 31 (1971), pp. 553-8. Saeki Tomi & f6 8 and Chikusa Masaaki * '& W * . So no shin bunka % <T>ffi£ it. Vol. 6 of Toyo no reikishi S W <7) M ^ , ed. Saeki Tomi. Kyoto: Jimbutsu oraisha, 1967. Sagaster, Klaus. Die weisse Geschichte. Wiesbaden: Otto Harrossowitz, 1976. Saguchi Toru {£ • M . "Juyon seiki ni okeru Gencho daikan to seiho san-oke to norentaisei ni tsuite"+ ES 1tt 22 U fit It Cambridge Histories Online © Cambridge University Press, 2008 BIBLIOGRAPHY 763 t P T ." Kita Ajia gakuho ifc 55 SB 55 ^ $8 , 1 (1942), pp. 151-214. Sainson, Camille. Nan-tchao ye che, histoireparticuliere de Nan-tcbao. Paris: Imprimerie Nationale, Ernest Leroux, editeur, 1904. Sang Hsiu-yiin IS 3f S . "Chin shih Wan-yen shih hun chih chih shih shih ^iijLfc, H .R #§ $!j £ IS P ." Chung yang yen chiu yuan It shih yii yen yen chiu so chi k'an*P ik WSEKEimffflSEBrHlfl (Bulletin of the Institute of History and Philology, Acadmia Sinica), 39(1969), pp. 255-88. Sansom, George B. A history ofJapan to 1334. Stanford, Calif.: Stanford University Press, 1958. Sat5 Hisashi fe M ft . Kodai Chibetto shi kenkyii S f t ^ ^ ' V S S S T ^ . 2 vols. Kyoto: Kyoto daigaku Toyoshi kenkyukai, 1958—9. Sayf ibn Muhammad. Ta'rtkh-i namah-i Harat. Ed. Muhammad Zubayr al-Siddlql. Calcutta: Baptist Missionary Press, 1944. Schulte-Uffelage, Helmut, trans, and ed. Dans Keng-shen wai-shih: Eine Quelle zur spa'ten Mongolenzeit. Ostasiatische Forschungen, Sondetreihe Monographien no. 2. Berlin: Akademie-Verlag, 1963. Schurmann, Herbert F. Economic structure of the Yuan dynasty: Translation of chapters 93 and94 of the Yuan shih. Harvard-Yenching Institute Studies, vol. 16. Cambridge, Mass.: Harvard University Press, 1956; repr. 1967. Schurmann, Herbert F. "Mongolian tributary practices of the thirteenth century." HarvardJournal of Asiatic Studies, 19(1956), pp. 304-89. Schurmann, Herbert F. "Problems of political organization during the Yuan dynasty." In vol. 5 of Trudy XXV Mezhdunarodnogo kongressa vostokovedov. Moscow: Izdatel'stvo Vostochnov Literatury, 1963, pp. 26—30. Schwartz-Schilling, Christian. Der Friede von Shan-yiian (1005 n. Chr.): Ein Beitrag zur Geschichte der chineschen Diplomatie. Asiatische Forschungen no. 1. Wiesbaden: Otto Harrassowitz, 1959. Seifeddini, M. A. "Monety s nadpis'iu 'ulug mangul ulus-bek'." Numizmatika i epigrafika, 9(1971), pp. 115-21. Sekino Tadashi US W M and Takejima Takuichi Yl H # — . RyoKinjidai no kenchiku -2 vols of plates. Tokyo: T5ho to sono butsuzo ! I £ B 3 F f t y j 8 ! ? h g / $ & bunka gakuin Tokyo kenkyujo, 1934. Serruys, Henry. The Mongols in China during the Hung-wuperiod. Melanges Chinois et Bouddhiques no. 11. Bruges: L'Institut Beige des Hautes Etudes Chinoises, 1959. Serruys, Henry. "Remains of Mongol customs in China during the early Ming." Monumenta Serica, 16(1957), pp. 137—90. Serruys, Paul. "Notes marginales sur le folklore des Mongols Ordos." Han-Hiue SI ^P : Bulletin du Centre a"Etudes Sinologiques de Pe'kin, 3 (1948), pp. 15-210. Shao Hsun-cheng SBfllIE . "Yuan tai te wen hsiieh yu she hui 7t ft W % ? H ft # . " T'u shu yiieh k'an ffl 1? ^ TO , 3 (1943). Repr. in Yuan shih lun ts'ung 7C S. at IS , 1 (1982), pp. 221-4. Shao Yiian-p'ing HP IS ¥ , comp. (Hsu hung Men lu) Yuan shih leipien ( Wi %Lffi& ) 7t £ H H . Jen-ho Shao shih C fO 3P ft , 1699; repr. in Sung Liao Chin Yuan pieh Cambridge Histories Online © Cambridge University Press, 2008 764 BIBLIOGRAPHY shih ^ M 3s 7Z S!l £ , comp. Hsi Shih-ch'en, ^ 1S E . Ch'ang-shu: Hsi shih Saoyeh shan-fang, 1797. Shao Yiian-p'ing, comp. Yuan shih leipien ft $. $} £§ . 1699. Repr. in vols. 49—58 of Shih liao ts'ungpien hsiipien S f i S i S H , ed. Kuang-wen shu-chii S X f i . Taipei: Kuang-wen shu-chii, 1968. Shen Kua i tfiS. Meng hsi {ch'ijpi t'an chiao cheng W- 8£ % WLftS!, ed. Hu Tao-ching SE H fi¥ . 2 vols. Shanghai: Shanghai ch'u pan kung ssu, 1956; repr. Taipei: Shihchieh shu-chii, 1965. Sheng-wu ch'in cheng lu chiao chu S S i t i E i l f t f t . I n Meng-ku shih liao ssu chung W S £ #4 E ffl , ed. Wang Kuo-wei 3E H I t . Peking, 1926; repr. Taipei: Chungcheng shu-chii, 1962, 1975. Shiba, Yoshinobu. Commerce and society in Sung China. Trans. MarkElvin. Ann Arbor: Center for Chinese Studies, University of Michigan, 1970. Shiba Yoshinobu $f & ft im . Sodai shogyoshi kenkyu ^ ft IS % £ 5f 5E . Tokyo: Kazama shobo, 1968. Shiba, Yoshinobu. "Sung foreign trade: Its scope and organization." In China among equals: The Middle Kingdom and its neighbors, ioth-i4th centuries, ed. Morris Rossabi. Berkeley and Los Angeles: University of California Press, 1983, pp. 89-115. Shih Chin-po . & & Wi . Hsi Hsia Fo chiao shih liieh ffi S <$ tfe £. B§ . Yin-ch'uan: Ning-hsia jen min ch'u pan she, 1988. Shih Chin-po. " 'Hsi Hsia i ching t'u' chieh S S W & M M ." Wen hsien.X «R , 1 (1979). PP- 215-29. Shih Chin-po. Hsi Hsia wen hua H S 3t •fb . Ch'ang-ch'un: Chi-lin chiao yii ch'u pan she, 1986. Shih Chin-po and Pai Pin & JS . "Ming tai Hsi Hsia wen ching chiian ho shih ch'uang ch'u t'an 18 ft S 1 1 . ® # W 5 tt «) « ." /CWo ^0 A J M ^ 1977:1, pp. 143-64Shih Chin-po, Pai Pin, and Huang Chen-hua Sf M W . "Wen-hai" yen chiu~% Peking: Chung-kuo she hui ko hsiieh ch'u pan she, 1983. Shih Chin-po, Pai Pin, and Wu Feng-yun % ($ £ . Hsi Hsia wen wu B Peking: Wen wu ch'u pan she, 1988. Shih, Chung-wen. The golden age of Chinese drama: Yuan Tsa-chii. Princeton, N.J.: Princeton University Press, 1976. Shih Kuo-ch'i Jfi D V . Chin shih hsiang chiao & £ e¥ ft . TSCC ed. Shih Wei-min £ % Ji . "Yuan sui tz'u k'ao shih 7C ^ i§ =# If ." y««» i i i /«»ts'ung jt i l i fs , 3 (1986), pp. 144-53. Shimada Masao A ffl IE 615 . Ryo no shakai to bunka M <T)tt# t X -fb . Tokyo: Kobundo, 1956. Shimada Masao. i?>or/6o kansei no kenkyu %t W& U O W 5E. Toyo hoshi ronshu S ^¥ & £ li ^ no. i. Tokyo: Sobunsha, 1978. Shimada Masao. Ryocho shi no kenkyu 18 $ft 5& (7) W ?E . Tokyo: Sobunsha, 1979. Shimada Masao. Ryodai shakai shi kenkyu . I f t t t t i B f ^ E . Kyoto: Sanwa shobo, 1952. Cambridge Histories Online © Cambridge University Press, 2008 BIBLIOGRAPHY 765 Shimada Masao. Ryosei no kenkyu SI Si C7) ffl -J5 . Tokyo: Nakazawa insatsu kabushiki kaisha, 1954; repr. 1973. Shimada Masao. So-shu jo ffl ttl $£ . Tokyo: Bunkodo shoten, 1955. Shiratori, Kurakichi, ed.; Yanai, Wataru; Inaba, Iwakichi; and Matsui, Hitoshi. Beitrage zur historischen Geographie der Manschurei. (Translation of their Manshu rekishi chin"). 2 vols. Tokyo: Verlag der Siidmandschurischen Eisenbahn A.G.; vol. 1, 1914; vol. 2, 1912. Shiratori Kurakichi fillo, Yanai Wataru 15 ?3 2 , Matsui Hitoshi fe # ^ , and Inaba Iwakichi f§ M S ^ . Manshu rekishi chiri 8§ #N M £ ife S . 2 vols. Tokyo: Minami Manshu Tetsudo kabushiki kaisha, 1913; repr. Tokyo: Maruzen, 1940. Shkoliar, Sergei A. Kitaiskaia doognestrel'naia artilleria. Moscow: Nauka, 1980. Shu Fen §? $t . Liao shih kao Jf $L $S . Wu-han: Hu-pei jen-min ch'u-pan she, 1984. Sinor, Denis. "The Inner Asian warriors." Journal of the American Oriental Society, 101 (1981), pp. 133-44. Sinor, Denis. "The legendary origin of the Turks." In Folklorica: Festschrift for Felix J. Oinas, ed. Egle Victoria Zygas and Peter Voorhies. Bloomington, Ind.: Research Institute for Inner Asian Studies, 1982, pp. 223—57. Sinor, Denis. "On Mongol strategy." In Proceedings of the fourth East Asian Altaistic conference, ed. Ch'en Chieh-hsien. T'ai-nan: National Ch'eng-kung University, 1975, pp. 238-49. Siren, Osvald. "Chinese sculpture of the Sung, Liao and Chin dynasties." Bulletin ofthe Museum of Far Eastern Antiquities (Ostasiatiska Samlingarna), 14 (1942), pp. 4 5 - 6 4 . Skelton, R. A., trans. The Vinland map and the Tartar relation. New Haven, Conn.: Yale University Press, 1965. Smith, John M. " 'Ain Jalut: Mamluk success or Mongol failure?" HarvardJournal of Asiatic Studies, 44 (1984), pp. 307—45. Smith, John M. "Mongol and nomadic taxation." Harvard Journal ofAsiatic Studies, 30(1970), pp. 4 6 - 8 6 . Smith, John M. "Mongol manpower and Persian population. "Journal of the Economic and Social History of the Orient, 18(1975), pp. 270—99. Sogabe Shizuo ^ ft 815 H W. . Nissokin kahei koryushi B * & : K f l l £ 8 f t £ . Tokyo: Hobunkan, 1949. Ssu-ma Kuang 5] % 3fc . Su-shui chi wen &%• IB Ml . TSCC ed. Ssu-ma Kuang et al., comps. Tzu chih t'ung chien Sf ?p jffi ££ . Peking: Chung-hua shu-chii, 1956 ed. Stein, M. Aurel. Innermost Asia, vol. 1. Oxford: Clarendon Press, 1928. Stein, Rolf A. "Leao-Tche." T'oungPao, 35 (1939), pp. 1 — 154. Stein, Rolf A. "Mi-nag et Si-hia: Geographie historique et 16gendes ancestrales." Bulletin de 1't.cole Franqaise d'Extreme-Orient, 44(1947—50), pp. 223-65. Steinhardt, Nancy S. Chinese imperial city planning. Honolulu: University of Hawaii Press. 1990. Steinhardt, Nancy S. Chinese traditional architecture. New York: China Institute in America, 1984. Cambridge Histories Online © Cambridge University Press, 2008 766 BIBLIOGRAPHY Steinhardt, Nancy S. "Currency Issues in Yiian China." Bulletin of Sung—Yiian Studies, 16(1980), pp. 59—81. Steinhardt, Nancy S. "Imperial architecture under Mongolian patronage: Khubilai's imperial city of Daidu." Ph.D. diss., Harvard University, 1981. Steinhardt, Nancy S. "The plan of Khubilai Khan's imperial city." Artibus Asiae, 44 (1983), pp. 137-58. Su Ch'e M Wi, Luan ch'eng chi 3k$iM. 3 vols. Peking: Chung-hua shu-chii, 1984 ed. Su Chen-shen M M ^ . Yuan cheng shu Ching shih ta tien chih yen chiu 7C i& H M t£ ^ $i 12.ffl% . Taipei: Chung-kuo wen-hua ta-hsiieh ch'u-pan pu, 1984. Su Shih M W . Tung-p'o chih lin M * i£ # . TSCC ed. Su T'ien-chueh M 3z S , comp. Kuo ch'ao wen lei MfflXM . SPTK ed. Su T'ien-chueh. Tzu-hsi wen kao 8£ SI 3t $5 . Repr. in Yiian tai chen pen wen chi hui k'an T c f ^ i ^ ^ ^ J U f t f ! ] . Taipei: Kuo-li chung-yang t'u-shu kuan, 1970. Su T'ien-chueh. Yiian ch'ao ming ch'en shih liieh 7t H £ E V B§ . Chien-an, 1335; repr. Peking: Chung-hua shu-chii, 1962. Su T'ung-ping I? HI ^ . See Chuang Lien. Sui andT'ang China, 389—906. See Twitchett, Denis C , ed. Sui shu' W • . See Wei Cheng. Sun Chii-yiian 3& 3S III . "Ch'ing-t'ang lu chi kao # ^ £& $S $5 ." Hsi-tsang yen chiu H^cW^E , 1982:2, pp. 144—55. Sun K'o-k'uan J& !& "M.. "Chiang-nan fang hsien yii Yen-yu ju chih S. Si SS H J8I S $6 HI ?p ." In Yiian tai Han wen hua te huo tung 7cf^MX'fk6*)S§S), ed. Sun K'ok'uan. Taipei: Chung-hua shu-chii, 1968, pp. 345—63. Sun K'o-k'uan. Meng-ku Han chiin chi Han wen hua yen chiu KS'SlJSS.SIlSC'fbW ?E . Taipei: Wen-hsing shu-tien, 1958. Sun K'o-k'uan. "Yii Chi and southern Taoism during the Yiian period." In China under Mongol rule, ed. John D. Langlois, Jr. Princeton, N.J.: Princeton University Press, 1981, pp. 212—53. Sun K'o-k'uan. "Yuan ch'u Li T'an shih pien te fen hsi it W & SI V ^ W # #f ." Ta-lu tsa chih ~X W. SI IS , 13, no. 8 (1956), pp. 7-15. Sung biographies. See Franke, Herbert, ed. Sung jen chuan chi tzu liao so yin ^ A <$ 13 IS ^4 % 31 .See Ch'ang Pi-te. Sung Chiung ^ §£ . Yen-shih chi M 5 * . SKCS ed. Sung Lien * B . Sung Wen-hsien kung ch'iian chi ^X^&±M . SPPY ed. 1936; repr. Taipei: Chung-hua shu-chii, 1965. Sung Lien et al., eds. Yuan shih 7C.£. 15 vols. Peking: Chung-hua shu-chii, 1976 ed. Sung shih 3= £ .See T'o-t'o. Ta Chin kuo chih ~X fz H ;£ . See Yii-wen Mou-chao. Ta Chin tiaofa lu X £ ^ $ $ . T5CC ed. 7a Y«a» j&«d» chih tsa chi X % IT $*J ^S 15 . In vol. 9 oiHsiieh shu ts'ungpien ¥ $f 3S H , ed. Chi-Fo-t'o J6WK. Taipei: I-wen yin-shu kuan, 1971. Ta Yiian sheng cheng kuo ch'ao tien chang ^C7CliBfSS?ft^ . Facsimile repr. of rev. and expanded 1303 ed., Taipei: Kuo-li ku-kung po-wu yiian, 1972. Tai Hsi-chang IS Wa % • Hsi Hsia chi BS S 12 . 1924; repr. in vol. 4 of Chung hua wen Cambridge Histories Online © Cambridge University Press, 2008 BIBLIOGRAPHY 767 shih ts'ung shu 41 H ~$t £ M • , ed. Wang Yiin-wu £ if E . Taipei: Hua-wen shu-chii, 1968. Takejima Takuichi ftS#~. Ryo Kin jidai no kenchiku to sono butsuzo JS 3z Bf ft ff) M S£ t ^ W ft ^ . Tokyo: Ryubun shokyoku, 1944. Takigawa Masajiro H III St ^ fiP and Shimada Masao Jl EH IE @$i . Ryoritsu no kenkyu HI # 2l $f 5E . Tokyo: Osaka yogo shoten, 1943. Tamura Jitsuzo H ft * IS . Chugoku seifuku ocho no kenkyu 't'JSffiBIiEIBWif^E . 2 vols. Kyoto: Toyoshi kenkyukai, 1964-71. Tamura JitsuzS. Keiryo no hekiga %t ^ <75 S fi . Kyoto: Dohosha, 1977. Tamura Jitsuzo. "The legend of the origin of the Mongols and problems concerning their migration." Ada Asiatica, 24 (1973), pp. 1-19. Tamura Jitsuzo and Kobayashi Yukio 'b ft ff H . Keiryo 85 H! . Kyoto: Kyoto daigaku bungakubu, 1953. T'ang Ch'eng-yeh Ml M M . Sui Wen-ti cheng chih shih kung chib yen chiu Pi X ^ ®f fe ^ SJ /£ 9T 5E . Taipei: Shang-wu yin-shu kuan, 1967. T'ang K'ai-chien j© Ml iH . "Yuan tai Hsi Hsia jen te cheng chih ti wei 7C iX S X A &5 if£ ?o ift'S ." Kan-su min tsuyen chiu Vi M S M W ^E , nos. 1-2 (1987), pp. 10-26. T'ang lit shu i S W W. 18 . Peking: Chung-hua shu-chii, 1983 ed. Tao, Jing-shen. "Barbarians or northerners: Northern Sung images of the Khitan." In China among equals: The Middle Kingdom and its neighbors, ioth—i4th centuries, ed. Morris Rossabi. Berkeley and Los Angeles: University of California Press, 1983, pp. 66-86. Tao, Jing-shen. "The influence of Jurchen rule on Chinese political institutions." Journal of Asian Studies, 30(1970), pp. 121—30. Tao, Jing-shen. The Jurchen in twelfth-century China: A study of sinicization. Seattle: University of Washington Press, 1977. Tao, Jing-shen. "Political recruitment in the Chin dynasty." Journal of the American Oriental Society, 94(1974), pp. 24-35. Tao, Jing-shen. Two sons of heaven: Studies in Sung-Liao relations. Tucson: University of Arizona Press, 1988. Tao, Jing-shen. "Yii Ching and Sung policies toward Liao and Hsia, 1042-44." Journal of Asian History, 6(1972), pp. 114—22. T'ao Chin-sheng $9 W £ (Jing-shen Tao). Chin Hai-ling ti te fa Sung yii Ts'ai-shih chan i te k'ao shih 4 i i $ W t t * i * E i l S W % K . Taipei: Kuo li Taiwan ta hsiieh wen hsiieh yiian, 1963. T'ao Chin-sheng. Sung Liao kuan hsi shih yen chiu 5)? jg BS % 5& 5f ^E . Taipei: Lienching ch'u-pan shih-yeh kung-ssu, 1984. T'ao Hsi-sheng W 3$ 3? . "Yiian tai Mi-le Pai-lien chiao hui te pao tung Jt ft 31 Wl 6 M %L # &} m Wl ." Shih huo yiieh k'an * « H Vi , 1 (1935), pp. 152-5. T'ao Tsung-i $sl m Wt . Nan ts'un ch'o keng lu M ft 111 Sf ^ . Peking: Chung-hua shuchii, 1959 ed. T'ao Tsung-i. Shu shih hui yao • £. # S . Pref. 1376. Hung-wu [1368-98] ed.; 2nd ed. Wu-chin, 1929; repr. Shanghai: Shang-hai shu-chii, 1984. Tartaro-Mongoly v Azii i Europe. See S. L. Tikhvinskii. Cambridge Histories Online © Cambridge University Press, 2008 768 BIBLIOGRAPHY Taskin, V. S., trans. E Lun-i (Yeh Lung-li), litoriia gosudarstva Kidanei (Tsidan' go chzhi): Perevod s kitalskogo, wedenie, kommentarU i prilozheniia. In Pamiatniki pis'mennosti vostoka, 81, no. 35. Moscow: Nauka, 1979. Tayama Shigeru EB li-1 3? . "Gendai zaiseishi ni kansuru oboegaki — shushi no gaku wo chushin to shite 7 t t t B*iS5& \Z ffl t Z> * » - 4£ £ <F> $ £ <£ >b t L T."In Toyo no seiji keizai M # CO & j£ IS §t . Hiroshima: Mokukoku shoten, 1949, pp. 191—266. Taylor, Keith W. The birth of Vietnam. Berkeley and Los Angeles: University of California Press, 1983. Terentyev-Katansky, A. P. "The appearance, clothes and utensils of the Tanguts." In The countries and peoples of the East, ed. D. A. Olderogge. Moscow: Nauka, 1974, pp. 215-24. Thiel, Joseph. "Der Streit der Buddhisten und Taoisten zur Mongolenzeit." Monumenta Serica, 20(1961), pp. 1—81. Thiele, Dagmar. Der Abschluss eines Vertrages: Dip/omatie zwischm Sung und Chin Dynastie, 1117—1123. Miinchener Ostasiatische Studien no. 6. Wiesbaden: Franz Steiner, 1971. T'ien Ju-kang. "Mongol rulers and Chinese pirates." History Today, 33 (1983), pp. 33-8Tietze, Klaus. "The Liao—Sung border conflict of 1074—76." In Studia SinoMongolica: Festschrift fur Herbert Franke, ed. Wolfgang Bauer. Wiesbaden: Franz Steiner, 1979, pp. 127—51. Tikhvinskii, S. L., ed. Tartaro-Mongoly v Azii i Europe, Moscow: Nauka, 1977. Ting K'un-chien T M @ . "Yuan tai te k'o chu chih tu 7t ft #J $i # ffl ft ." Hua hsiiehyueh k'an M ^ B ¥0 , 124 (1982), pp. 46-57; 125 (May 1982), pp. 28-51. Togan, Isenbike. "The chapter on annual grants in the Yuan shih." 2 vols. Ph.D. diss., Harvard University, 1973. T'o-t'offiiflftet al., eds. Chin shih ^.i. . 8 vols. Peking: Chung-hua shu-chii, 1975 ed. Also 2 vols. Taipei: Kuo-fang yen-chiu yuan, 1970 ed. T'o-t'o et al., eds. Liao shih 3?5& . 5 vols. Peking: Chung-hua shu-chii, 1974 ed. T'o-t'o et al., eds. Sung shih % $.. 40 vols. Peking: Chung-hua shu-chii, 1977 ed. Torii Ryuzo B g SI W.. Kokogakujoyori mitaru Ryo no bunka: Zufu % i f i i • ) ! + £ ' 5 S < 7 ) ^ ' ( f c i : BBH4 vols. Tokyo: Tohobunka gakuin, Tokyo kenkyusho, 1936. Torii Ryuzo. Ryo no bunka wo saguru HE <T) % it £ $i -5 . Tokyo: Shokasha, 1937. Torii Ryuz5. "Ryodai no hekiga ni tsuite 3§ f^ <T) § | fi [I Wt I* T ." Kokka M ¥ , 490, pp. 272-80; 491, pp. 283-9; 492, pp. 313-17; 493. PP- 343-50 (1931). Toyama Gunji ft- lil W fe . Kinchoshi kenkyu ^ 19 ^ W % . Toyoshi kenkyu sdkan M f f j f f l ^ l f l no. 13. Kyoto: TSyoshi kenkyukai, 1964. Ts'ai Mei-piao H j l M , Chou Liang-hsiao ffl M. K , and Chou Ch'ing-shu SO ^ 8J . Chung-kuo t'ung shih. •£ IS j§ Sfc , vol. 7. Peking: Jen min ch'u pan she, 1983. Ts'ai Mei-piao M Hti&, ed. Yuan taipai hua pei chi lu 7C ft S IS W M ^ . Peking: K'o-hsiieh ch'u-pan-she, 1955. Tseng Lien © W. . Yuan shu 7C 8 . Shao-yang: Tseng shih Ts'eng-i t'ang, 1911. Cambridge Histories Online © Cambridge University Press, 2008 BIBLIOGRAPHY 769 Ts'ui Wen-yin I X S . Chin shih jen ming so yin sfe $. A £ % 31 . Peking: Chunghua shu-chii, 1980. Tsung Tien m ft . K'o Chiu-ssu shih liao fl A S S S . Shanghai: Shang-hai jen-min ch'u-pan she, 1963. Tsunoda, Ryusaku, and L. C. Goodrich. Japan in the Chinese dynastic histories. South Pasadena, Calif,: P. D. and lone Perkins, 1951. Tu Lien-che. "Hung Chun." In vol. 1 of Eminent Chinese of the Ch'ing period, ed. Arthur O. Hummel. Washington, D.C.: U.S. Government Printing Office, 1943. PP- 360-1. Tu Wei-yun tt It & . Ch'ing tai shih bsiieh yu shih chia ft ft £ <P H £ %. . Taipei: Tung-ta t'u-shu yu-hsien kung-ssu, 1984. Tu Yu t t ft . Tung tien Si ft . In Shih t'ung + U , comp. Wang Yiin-wu I f S . Shanghai, 1935-7; repr. Taipei: Hsin-hsing shu-chii, 1965. Tu Yii-t'ing ttl?. Yuan tai Lo-lo-ssu shih liao chi k'ao 7GftIIS8f£il e HB=# . Chengtu: Ssu-ch'uan min-tzu ch'u-pan she, 1979. T'u Chi M ^r. Meng-wu-erh shih chi * 7C IS £ 13 . Wu-chin, Kiangsu: Chieh I-i, 1934; repr. Peking: Ku-chi ch'u-pan she, 1958; Peking: Chung-hua shu, 1962. Also Taipei: Shih-chieh shu-chii, 1962; repr. 1968. Tucci, Guiseppe. Tibetan painted scrolls. 2 vols. Rome: La Libreria dello stato, 1949. Tung Chieh-yiian. Master Tung's western chamber romance (Tung Hsi-hsiang chu-kung- tiao): A Chinese chantefable. Trans. Li-li Ch'en. Cambridge: Cambridge University Press, 1976. T'ung chih t'iao ko 81 VI IS f& . Peking: Kuo-li Pei-p'ing t'u-shu-kuan, 1930. Twitchett, Denis C. "Merchant, trade and government in late T'ang." Asia Major, 14 (1968), pp. 63-95. Twitchett, Denis C , and John K. Fairbank, eds. Sui and T'ang China, 589—906, pt. 1. Vol. 3 of The Cambridge history of China. Cambridge: Cambridge University Press, 1979. Tzu chih t'ung chien i?ftaIS US . See Ssu-ma Kuang. Uematsu, Tadashi tSfejE . "The control of Chiang-nan in early Yuan." Ada Asiatica, 45 (1983), pp. 49-68. Uematsu Tadashi. "Gendai Konan no gdmin Shu Sei Cho Sen ni tsuite: Sono chusatsu to zaisan kambotsu wo megutte 7 c f ^ f l C ^ ( 7 ) S E S ^ f S ? i ^ ( ^ ' 5 ' ' i T: * mm®. t$t <&•%&.$: It) <r)X ." Toyoshi kenkyu t ^ * i ? E ,27 (1968), pp. 46-71. Uematsu Tadashi. Gentensho nendai sakuin 71; ft 3§£ 3 s ft ^ 31 . Tokyo: Dohosha, 1980. Uematsu, Tadashi. "Institutions of the Yuan Dynasty and Yuan society." Gest Library Journal, 5 (1992), pp. 57-69. Uematsu Tadashi. "Ishu Shigen shingaku narabini kaisetsu^fc ftfM7C^ft&J M t / t C ffi 18 ." Toyoshi kenkyu M W £. Sf % , 30 (1972), pp. 1-29. Umehara Kaoru #5 W.ffland Kinugawa Tsuyoshi ft. ill 3£ , comps. Ryo Kin Genjin denki sakuin S I ^ 7 c A < S n H ^ 3 l . Kyoto: Kyoto daigaku jimbun kagaku kenkyujo, 1972. Cambridge Histories Online © Cambridge University Press, 2008 77° BIBLIOGRAPHY Vasary, Isrvan. "The origin of the institution of basqaqs." Ada Orientalia Academiae Scientarum Hungaricae, 32 (1978), pp. 201—6. Viktorova, L. L. Mongoly: Proiskhozhdenie naroda i isloki kul'tury. Moscow: Nauka, 1980. Visdelou, Claude de. Histoire abregee de la Tartarie. In vol. 4 of Bibliotheque orientate: Ou dictionnaire universel, contenant tout ce qui fait connoitre des peuples de I'Orient, ed. Barthelemy d'Herbelot de Modainville. Maestricht: J. E. Dufour & P. Roux, 1780. Vladimirtsov, Boris IAkovlevich. Gengis-khan. Trans. Michel Carsow. Introduction historique par Rene Grousset. Paris: Librairie d'Amerique et d'Orient AdrienMaisonneuve, 1948Vladmirtsov, Boris IA. The life of Genghis khan. Trans. Prince D. S. Mirsky. Boston: Houghton Mifflin, 1930. Vladimirtsov, Boris IA. Obshchestvennyj stroj Mongolov: Mongol'skii kochevoj feodalizm. Leningrad: Nauka, 1934. Vladimirtsov, Boris IA. La Regime social des Mongols: Le Feodalisme nomade. Trans. Michel Carsow. Paris: Librairie d'Amerique et d'Orient Adrien-Maisonneuve, 1948. Vladimirtsov, Boris IA. Meng-ku she hui chih tu shih W & ft # ffrj S. £ . Trans. Chang Hsing-t'ang H ft JS . Taipei: Chung-kuo wen hua ch'u pan shih yeh wei yuan hui, 1957; repr. Taipei: Shih Chung hua ch'u pan shih yen wei in hui, 1967. Vladimirtsov, Boris IA. Meng ku she hui chih tu shih. Trans. Liu Jung-chun SI H $£ . Peking: Chung-hua she hui k'o hsiieh ch'u pan she, 1980. Vladimirtsov, Boris I A. Moko shakai seidoshi M ~&fct# 8HJ K £ . Trans. Gaimusho chosabu. Tokyo: Gaimusho, 1934. Voegelin, Eric "The Mongol orders of submission to the European powers, 1245— 1255." Byzantion, 15 (1940-1), pp. 378—413. Vorob'ev, Mikhail V. Chzhurchzheni i gosudarstvo Tszin' (X V.—1234 g.): Istoricheskii Ocherk. Moscow: Nauka, 1975. Vorob'ev, Mikhail V. "O estestvennykh naukakh v chzhurchzhen'skom gosudarstve Tszin'." In Istoriia, kul'turayazyki narodov Vostoka. Ed. YU. A. Petrosian. Moscow: Nauka, 1970, pp. 145-9. Wada Sei Si ffl » . "Teian koku ni tsuite & $ M \Z gfc ^ T ." Toyo gakuho Mftm $6 , 6(1915); repr. with revisions in Wada Sei, Toa shi kenkyu: Manshu hen M 35 36 ffl % '• 85 M U. Toyo bunko ronso $ # i f i f no. 37. Tokyo: Toyo bunko, 1955, pp. 161-89. Waldron, Arthur N. The Great Wall: From history to myth. Cambridge: Cambridge University Press, 1990. Waley, Arthur. The travels of an alchemist: The journey of the Taoist Ch'ang-ch'un from China to the Hindukush at the summons of Chingiz khan. Recorded by his disciple Li Chih-ch'ang. London: Routledge and Sons, 1931. Wan Ssu-t'ung M Sff IH , ed. Keng shen chiin i shih §£ • 8 ift 9 . Repr. in vol. 4 of Shih liao ts'ungpien JS. i$ 3ii§ . Taipei: Kuang-wen shu-chii, 1968. Wan Ssu-t'ung. Liao ta ch'en nienpiao M ^ El ^F $i . Repr. as no. 33 in vol. 4 of Liao Cambridge Histories Online © Cambridge University Press, 2008 BIBLIOGRAPHY 771 shih huipien SI i 9k M , ed. Yang Chia-lo $§ ^ IS . Taipei: Ting-wen shu-chii, Wang Ch'eng 3:fi§. Tung tu shih liieh M IP V B§ . Repr. in vols. 11—14 of Sung shih tzu liao ts'uipien *fe $. jtf $4 W M , 1st series, ed. Chao T'ieh-han $& W. %f . Taipei: Wen-hai ch'u-pan she, 1967. Wang Ching-ju 3E M ® . Hsi Hsia yen chiu B H S 3E . 3 vols. Vol. 1: Peking: Kuo li chung yang yen chiu yiian li shih yii yen yen chiu so, 1932, special suppl. no. 8. Vol. 2: Peking: Kuo li chung yang yen chiu yiian li shih yii yen yen chiu so, 1933, special suppl. no. 11. Vol. 3: Peking: Kuo li chung yang yen chiu yiian li shih yii yen yen chiu yiian, 1933, special suppl. no. 13. Wang Chung 31 J£ . "Lun Hsi Hsia te hsing ch'i l i S§ X (ft H £S ." Li shih yen chiu m$.ffi% , 1962:5,pp. 20-32. Wang E~5.™%.Ju-nani shih S B I f . TSCC ed. Wang Feng 3£ & . W» <*'/ 06/ {§ Si ft . TSCC ed. Wang Gung-wu. "The rhetoric of a lesser empire: Early Sung relations with its neighbors." In China among equals: The Middle Kingdom and its neighbors, 10th—14th centuries, ed. Morris Rossabi. Berkeley and Los Angeles: University of California Press, 1983, pp. 47-65Wang Hui-tsu ffiSffi. Yiian shih pen cheng 7t &. $ i l . 2 vols. Peking: Chung-hua shu-chii, 1984 ed. Wang Kuo-wei £ 1 $ . "Chin chieh hao k'ao ife J? *& =# ." Yen-ching hsiieh pao, 1 (1927), pp. 1-14. Wang Kuo-wei. "Ta-ta k'ao te M # ." In Wang Kuo-wei, Kuan t'ang chi tin # . Repr. Peking: Chung-hua shu-chii, 1959, pp. 634—86. Wang Min-hsin i g, ft. "Ch'i-tan te 'ch'ai ts'e i' ho 'tsai sheng i' C S 4 I J " Kukungt'ushu chi k'an & "g U t M VI, 3, no. 3(1973), pp. 31-52. Wang Min-hsin. "Shan-yuan ti meng te chien t'ao iSSHISMWI^Bt." SA/A huoyiieh k'an ^ K ^ TO , n.s., 5 (1975), pp. 97-107. Wang P'u i $? TW»g A»/' ><?o S # S ' . 3 vols. Shanghai: Shang-wu yin-shu kuan, 1935; repr. Peking: Chung-hua shu-chii, 1955. Wang P'u et al.: Wu-tai huiyao £ fi # H . 3 vols. Shanghai: Shang-hai ku-chi ch'upan she, 1978 ed. Also KHCPTS ed. Wang Shih-tien 3E ± 15 , ed. P/ j£« n&/m nW& S ' t S S . i'/CCi' ed. Wang Te-i 3: H SS , comp. Yiianjen wen chi chenpen ts'ung k'an 7 £ A X f t ^ ^ ^ f l J . 8 vols. Taipei: Hsin wen-feng ch'u-pan kung-ssu, 1985. Wang Ting 3.B. Fen chiao lu $iM$&. Pref. 1089. In Pao yen t'ang pi chi WM^.% S , ed. Ch'en Chi-ju IS6 IS ffi . 1606; repr. Shanghai: Wen-ming shu-chii, 1922. Wang Yao 3E ft . "Hsi Hsia Hei-shui ch'iao pei k'ao pu B S IS * ffi W # fit." C/&«ng ><*»#OT/nttaAJ«>A >«<»» hsiieh pao t f ^ S ^ ^ ^ P ^ , 1978:1, pp. 51-63. Wang Yiin 31W . Ch'iu-chien hsien sheng ta ch'iian wen chi ft ffl % £. ^C £ £ M • SPTKed. Wassaf ('Abd Allah Wassaf al-Hadra). Tarikh-i Wassaf. Lithograph ed. Bombay: 1852-3, re-ed. Mohammad Mahdi Isfahan!. Tehran: 1959—60. Watson, William. T'ang and Liao ceramics. New York: Rizzoli, 1984. Cambridge Histories Online © Cambridge University Press, 2008 772 BIBLIOGRAPHY Wei Cheng & S4 , Ling-hu Te-fen ^ ! E S $ , et al., eds. 5*/ shu it • . Peking: Chung-hua shu-chii, 1973. Wei Shou IS J& . W« JA« iS It . Peking: Chung-hua shu-chii, 1974. Wei Su M M . Wei T'ai-p'u chi 1&±&%k . Wu-hsing: Liu shih Chia yeh t'ang Si R Wi OH ?=£. , 1913. Repr. in vol. 7 of Yiianjen wen chi cbenpen ts'ung k'an S A X $ t 3 s 18 fll , comp. Wang Te-i 5 tlJ Wt. . Taipei: Hsin wen-feng ch'u-pan kung-ssu, 1985, pp. 369-485. Wen Yu-ch'eng M. 3s.fi!c. "Yuan An-hsi wang yii tsung chiao 76 3c Bi 3i ft ^? tit ." K'ao ku yii wen wu % "£ H 3t ^ , 1984:4, pp. 95—7. Wei Yuan, M M . Yuan shih hsin pien Jt $. Sf M . Kiangsu: Shen wei t'ang, 1905. Weng Tu-chien ^ ® f i t . "Wo-to tsa k'ao %t Bft 38 # ." Yenching hsiieh pao, 29 (1941), pp. 201—18. West, Stephen H. "Jurchen elements in the northern drama Hu-t'ou-p'ai." T'oung Pao, 63 (1977). PP- 273-95. West, Stephen H. "Mongol influence on the development of northern drama." In China under Mongol rule, ed. John D. Langlois, Jr. Princeton, NJ.: Princeton University Press, 1981, pp. 434—65. West, Stephen H . Vaudeville and narrative: Aspects of Chin theater. Miinchener Ostasiatische Studien no. 20. Wiesbaden: Franz Steiner, 1977. Wittfogel, Karl A., and Feng Chia-sheng. History of Chinese society, Liao (907-1125). Transactions of the American Philosophical Society, n.s., vol. 36. Philadelphia: American Philosophical Society, 1949. Wright, Arthur F., ed. The Confucian persuasion. Stanford, Calif.: Stanford University Press, i960. Wright, Arthur F., and Denis C. Twitchett, eds. Confucian personalities. Stanford, Calif.: Stanford University Press, 1962. Wu Chi-hua ^ M % . "Yuan ch'ao yii Ming ch'u hai yiin" 7C Wl H 93 %} U & . Chungyang yen chiu yuan, Li shih yii yen yen chiu so chi k'an, 28(1956), pp. 363—80. Repr. in Wu Chi-hua, Ming tai she hui ching chi shih lun ts'ung W f ^ t t ^ S s ^ i l j g . 2 vols. Taipei: Taiwan hsiieh sheng shu chii, 1970, vol. 1, pp. 125—54. Wu Chi-yu. "Sur la version tangoute d'un commentaire du Louen-yu conservee a Leningrad." T'oung Pao, 55 (1969), pp. 298-315. Wu Han ^ B£. "Yuan ti kuo chih peng k'uei yii Ming chih chien kuo, 7E i$r EH £. $i ffi Mm Z M m ' ." Ch'ing-hua hsueh-pao rf * ^ S , 11 (1936), pp. 359-423. Wu Hsiao-ling ^ BS # et al., comps. Ta Chin kuo chih t'ung chien 'X jfe M Mffi$& (Index du Ta Kin Kouo-tche). Peking: Centre franco—chinois d'£tudes sinologiques, 1949. Wu Kuang-ch'eng ^ « ^ . Hsi Hsia shu shih B I f » . Pref., 1826; repr. in vols. 88—91 of Shih liao ts'ung pien hsii pien A Pt WL 31 H H , ed. Kuang-wen shu-chii pien iso J l i X i t ^ i i i i i f t f . Taipei: Kuang-wen shu-chii, 1968. Wu, K. T. (Wu Kuang-ch'ing). "Chinese printing under four alien dynasties." HarvardJournal of Asiatic Studies, 13 (1950), pp. 447—523. Cambridge Histories Online © Cambridge University Press, 2008 BIBLIOGRAPHY 773 Wu-tai shih chi E. ft £ §H . See Ou-yang Hsiu. Wu T'ien-ch'ih £k3^W. Hsi Hsia shih kao B I $ S . Chengtu: Ssu-ch'uan jen-min ch'u-pan she, 1980. 2nd rev. ed. Chengtu: Ssu-ch'uan jen-min ch'u-pan she, 1983. Wylie, Turrell V. "The first Mongol conquest of Tibet reinterpreted." Harvard Journal ofAsiatic Studies, 37 (1977), pp. 103—33. Yamaguchi Osamu ill D & , Moko shurai W ~& SI ^5'. Tokyo: Togensha, 1964, repr. 1979. Yamamura Kozo, ed. Medieval Japan. Vol. 3 of The Cambridge history of Japan. Cambridge: Cambridge University Press, 1990. Yanai Wataru 15 P9 5 . "Gendai shakai san kaikyu (shikimoku ko) 7C ft tt # EL Pit m. ( fe B %")," Man-Sew «&/WrafetfA//fettfeya £ab£« $SS£%3M5&W3E?8£ . 1916. Trans, into Chinese as Yuan tai Meng Han Se-mu taiyii k'ao 7C ft SK 31 fe S ^ Ml # by Ch'en Chieh W ft and Ch'en Ch'ing-chuan ^ Jt ft . Shanghai, n.d.; repr. Taipei: Shang-wu yin-shu-kuan, 1963. Yanai Wataru. "Moko no kokkai sunawachi 'kurirutai' ni tsuite W S <Ti H # IP *> F ^ U ;l/ ^ A J tC Sfc l> T •" M/^a** z^j/&/ A * H IS , 28 (1917), no. 4, pp. 47; no. 5, pp. 457-84; no. 7, pp. 688-722. Repr. in Yanai Wataru, Mokoshi kenkyti W. S £ if ?E , ed. Iwai Hirosato S # ^C S . et al. Tokyo: T5ko shoin, 1930, pp. 361-447. Yanai Wataru. Mokoshi kenkyu I S S f ?S . Tokyo: Toko shoin, 1930. Yang Chia-lo ^ ^ IS , ed. Liao shih huipien 8! $. $k M . 10 vols. Taipei: Ting-wen shu-chii, 1973. Yang Chia-lo, ed. Liao shih hui pien pu J8 5& ft HI ffl . Taipei: Ting-wen shu-chii, 1974Yang Chih-chiu ^ iS $. . Vaa» jA/A ja» /an TC $L E. fifa . Peking: Jen-min ch'u-pan she, 1985. Yang Chih-chiu. "Yuan tai Hui-hui jen te cheng chih ti wei TcftEIUlAfKi&^pftS &. ." In Yang Chih-Chiu, Yuan shih san lun Jt £ H Mr. Peking: Jen-min ch'u-pan she, 1985, pp. 245-82. Yang Lien *t 31. ATaa« Yun-shihp'ing chuan M H 5 ff IS . Urumchi [Ti-hua]: Hsinchiang jen-min ch'u-pan she, 1983. Yang, Lien-sheng. Excursions in sinology. Cambridge, Mass.: Harvard University Press, 1969. Yang, Lien-sheng. "Marginalia to the Yuan tien-chang." Harvard Journal of Asiatic Studies, 19 (1956), pp. 42-51; repr. in Lien-sheng Yang, Excursions in sinology. Harvard—Yenching Institute Studies no. 24. Cambridge, Mass.: Harvard University Press, 1969, pp. 126-35. Yang, Lien-sheng. Money and credit in China: A short history. Harvard—Yenching Institute Monograph Series, vol. 12. Cambridge, Mass., 1952; repr. Cambridge, Mass.: Harvard University Press, 1971. Yang, Lien-sheng. "The organization of Chinese official historiography: Principles Cambridge Histories Online © Cambridge University Press, 2008 774 BIBLIOGRAPHY and methods of the standard histories from the T'ang through the Ming dynasty." In Historians of China and Japan, ed. William G. Beasley and Edwin G. Pulleyblank. Oxford: Oxford University Press, 1961, pp. 44-59. Yang Ne %k afi . "Yuan tai nung ts'un she hui yen chiu 7C ft §k W ?± # W % ."Li shih yen chiu M. £ 5f % , 1965:4, pp. 117—34. Yang Ne and Ch'en Kao-hua W iS5 W*, et al., comps. Yuan tai nung min chart cheng shih liao huipien 7i>i%!&k&.ffl,^'£.')&'£ M. 4 vols. Peking: Chung-hua shu-chii, 1985. Yang Shen # W . Nan-chao yeh shih fH IB I? &. . Repr. Taipei: Hua-wen shu-chii, 1969. Yang Shu-fan S§ SS $F . "Yuan tai k'o chu chih tu 7C ft & V> U & ." Kuo li cheng chih ta hsiieh hsiieh pao H It ife #} ;*: IP # $g , 17 (1968), pp. 99-120. Yang Wei-chen It $£ W . Tung-wei-tzu wen chi M £t ? i * . SiT*: ed. Yang Yu *§ SS. Shan chu hsin hua ill @ Sf IS .TOCCed. Yang Yii-mei ^§ W M. "Yiian tai Chiang-nan t'ien fu shui chih k'ao 7G ft 21 M E9 SK $£ $!l % ." Chung-kuo li shih hsiieh hui shih hsiieh chi k 'an * Mffi* * # * P * TO, 21 (1989), pp. 143-70. Yao Sui m m . Mu an chi ®L Jt *•. 5PT/C ed. Yao Ta-li ft ^C ^ . "Yiian tai k'o chii chih tu te hsing fei chi ch'i she hui pei ching 7C ft*4^$"JSWf?/8&*&##£ ." Yiian shih chi pei fang min tsu shih yen chiu chik'an T c i R i t T s S ^ £ © f 5 S * T i J , 6 (1982), pp. 26-59. Yao, Tao-chung. "Ch'iu Ch'u-chi and Chinggis khan." Harvard Journal of Asiatic Studies, 46(1986), pp. 201-19. Yao Ts'ung-wu Wi. #£ ^ . "A-pao-chi yii Hou T'ang shih ch'en Yao K'un hui chien t'anhuachilu M ft Wt m '&m ffi E * 1* # M MIS * « ." Wen shih che hsueh pao 3C A @ P f f i , 5 (1953), pp. 91—112. Repr. with revisions in vol. 1 of Yao Ts'ung-wu, Tung-pei shih lun ts'ung M. ifc ^ H fl . Taipei: Cheng-chung shu-chii, 1959, pp. 217-47. Yao Ts'ung-wu. "Ch'i-tan chiin wei chi ch'eng wen t'i te fen hsi S f t S f i i S f l f l S W ^ - f f ."Wen shih che hsiieh pao X S ^ ^ I S , 2(1951), pp. 8 1 - u i . Repr. in vol. 1 of Yao Ts'ung-wu, Tung-pei shih lun ts'ung M i t £ H ^ . Taipei: Chengchung shu-chii, 1959, pp. 248—82. Yao Ts'ung-wu. "Hu-pi-lieh han tui yii Han hua t'ai tu te fen hsi iS & f}\ ff §i 1$ 81 it W. & #J ^ Vi ." Ta-lu tsa chih ^ M H U. , 11, no. 1 (July 1955), pp. 22-32. Yao Ts'ung-wu. "Hu-pi-lieh p'ing Sung i hou te nan jen wen t'i f@> ')& >?!J tfjfflA fSi M ." In vol. 7 of Yao Ts'ung-wu hsien sheng ch'iian chi $,'& , ed. Ch'en Chieh-hsien ^ffl^fe and Cha-ch'i Ssu-ch'in +L ^ Sf ft . Taipei: Cheng-chung shu-chii, 1982, pp. 1—86. Yao Ts'ung-wu. "Liao Chin Yiian shih chiang i (chia): Liao ch'ao shih HM siz 7C £ P « ( ¥ ) : i f j , " comp. T'ao Chin-sheng PS # £ . Repr. as vol. 2 of Y*o Ts'ung-wu hsien sheng ch'iian chi Wi 'it % 5fc £ jfe ^ , ed. Ch'en Chieh-hsien Bi if 5fc and Cha-ch'i Ssu-ch'in IL *f Sf $C . Taipei: Cheng-chung shu-chii, 1972. Yao Ts'ung-wu. "Liao Tao-tsung Hsiian-i huang hou shih hsiang tz'u yiian yii te wen hua te fen hsi 31 H m K & S B + § PI SS i( W X fb M » *f ." Wen shih che hsiieh pao X ^ © ^ S , 8 (1958), pp. 97-134. Cambridge Histories Online © Cambridge University Press, 2008 BIBLIOGRAPHY 775 Yao Ts'ung-wu. "Shuo A-pao-chi shih tai te Han ch'eng Kuo hsiieh chi k'an M I? #5 fll , 5 (1935), pp. 53—78. Repr. in vol. 1 of Yao Ts'ung wu, Tung-pei shih lun ts'ung W. i t S& !§ IS. . Taipei: Cheng-chung shu-chii, 1959, pp. 193-216. Yao Ts'ung-wu. "Shuo Ch'i-tan te na po wen hua !S£ Mfl-W ^ & 3t it ." In vol. 2 of Yao Ts'ung-wu, Tung-pei shih lun ts'ung ^ i t 5& t§ I t . Taipei: Cheng-chung shuchii, 1959, pp. 1-30. Yao Ts'ung-wu. Tung-pei shih lun ts'ung M Jtj £ IS St . 2 vols. Taipei: Cheng-chung shu-chii, 1959. Yao Ts'ung-wu. Yao Ts'ung-wu hsien sheng ch'iian chi #fc &. U % ifc £ ^ , ed. Ch'en Chieh-hsien B£ ft ft and Cha-ch'i Ssu-ch'in ft 3f Iff UK. (Sechin Jagchid). 7 vols. Taipei: Cheng-chung shu-chii, 1971—82. Yao Ts'ung-wu. "Yuan Shih-tsu Hu-pi-lieh han: T'a te chia shih, t'a te shih tai yii t'a tsai wei ch'i chien chung yao ts'o shih 7C1ft ffl & <B M if : ftfe W ^ 1ft • fi W fflfiSJSJfi ." In vol. 6 of K«o Ts'ung-wu hsien sheng ch'iian chi , ed. Ch'en Chieh-hsien W. ft ft and Cha-ch'i Ssu-ch'in +L fF #f tfc . Taipei: Cheng-chung shu-chii, 1972, pp. 399—416. Yeh Ch'ien-chao MfBWl. Chin lu chih yen chiu £ ft £ if 3E . Taipei: Shang-wu yinshu kuan, 1972. Yeh Hsin-min M $f S . "Wo-ch'ih-chin chia tsu yii Meng Yiian han t'ing te kuan hsi & * if ^ R H S 7CflFi i 6*/ BS {^ ." N« Meng-ku ta hsiieh hsiieh pao (Che she pan) ft g £ * ¥ #5 $8 ( S t t i S ) , 1988:2, pp. 14-26. Yeh Lung-li M IS ffl , comp. Ch'i-tan kuo chih 18fl-S ^ . Shanghai: Shang-hai kuchi ch'u-pan she, 1985. Yeh Lung-li. 5«e also Taskin, V. S. Yeh Tzu-ch'i M •? ^ . 7i'ao mu tzu ^ ^ ? . Peking: Chung-hua shu-chii, 1959 ed.; repr. 1984. Yen Chien-pi US! $8 3B . "Nan Sung liu ling i shih cheng ming chi chu ts'uan kung fa hui nien taik'ao l ^ / ^ & t V I E ^ g f t f l t g S I & ^ f t ^ ." Yen-ching hsiieh pao MB.^%1 , 30(1946), pp. 27—50. Yip, Hon-ming. "The class system of Yiian society: A critique of Meng Siming's Yiiandai shehui jieji zhidu." Journal of Asian Culture, 4(1980), pp. 82—106. Yoshikawa Kojiro ^ Jl| ^ >fcfiP. "Gen no shotei no bungaku S W l f W t i ." In vol. 15 of Yoshikawa Kojiro zenshu cf ill # ^K 8P ^ H . Tokyo: Chikuma shobo, 1969, pp. 2 3 2 - 3 0 3 . Yoshikawa Kojiro. Gen zatsugeki kenkyu 7t H 0'J W ?£ . Tokyo: Iwanami shoten, 1948; 2nd rev. ed. 1958. Trans, by Cheng Ch'ing-mao M ft M as Yiian tsa chiiyen chiu 7C $£ $9 9F 3E . Taipei: I-wen yin-shu kuan, i960. 7C ft 3z ^ IS . " I n vol. 14 of Yoshikawa Kdjiro. "Genkyo kinsen ki Yoshikawa Kojiro zenshu 1§ Jl| # >^ 615 ^ # . Tokyo: Chikuma shobo, 1968, pp. 453-68. Yii Chi MM. Tao-yuan hsiieh ku lu i t S & * U.. KHTS ed. Also SPTK ed. YiiChi. 7«0-;w0»/e/£00it@&$[ Yuan Fu-chou lu ju hsueh 7C^fflgSffiiP, 1345; repr. in vols. 5 and 6 of Yiian jen wen chi chenpen ts'ung k'an Cambridge Histories Online © Cambridge University Press, 2008 776 BIBLIOGRAPHY comp. Wang Te-i I S S , Taipei: Hsin wen-feng ch'u-pan kung-ssu, 1985; vol. 5. PP- 251-644; vol. 6, pp. 1-471Yii Ch'ueh £ ffl . Ch'ing-yanghsien sheng wen chi W Wift £ £ H . SPTKed. Yii Ta-chiin $fX& . "Lun T'u Chi te 'Meng-wu-erh shih chi' t§ M ^ tfj M 7t IS A 12 ." Yiian shih lun ts'ung it £. It WL , 3 (1986), pp. 219-30. Yii-wen Mou-chao =? X fS HS . Ta Chin kuo chih * & S ;£ . KHT5 ed. Yii-wen Mou-chao. 7a Chin kuo chih chiao-cheng ~X ife M i£ tt SE . Ed. Ts'ui Wen-yin ft X £P . Peking: Chung-hua shu chii, 1986. Yiian ch'ao cheng Mien lu 7C f! % M & . TSCC ed. Yiian Chi (Kuo-fan) t S (M M). y«a» JMI >«? cA/« /«» 0&1 7C £ W 5E i i M . Taipei: Shang-wu yin-shu kuan, 1974. Yuan Chiieh ft m . Ch'ingjung chu shih chi i t # g ± * . 5PTK ed. Yuan Hao-wen' % JF P3 , comp. G&»»£ f/6o» <:/&/ 4"fflM . SPTK ed. y««» jen wen chi chen pen ts'ung k'an % A % M & # Wt TO'.. 5Ve Wang Te-i. y«<»» J/&/A 7C$.. See Sung Lien. Yiian Ta-tu k'ao ku tui 7C ^C H5 # ^ W . "Chi Yiian Ta-tu fa hsien te Pa-ssu-pa tzu wen wu g e j C ^ S P a S W A S E ^ X ^ I ." *T'<»o <*« # * , 1972:4, pp. 54-7. Yiian Ta-tu k'ao ku tui. "Pei-ching Hou Ying-fang Yiian tai chii chu i chih it % IB 3 5 ^ 7 C f ^ S t t 3 f l i t ." K'ao ku % ft , 1972:6, pp. 2-15. Yuan Ta-tu k'ao ku tui. "Yuan Ta-tu te k'an ch'a ho fa chiieh 7tJK%frtflffl)M%Q$l M ." K'ao ku, 1972:1, pp. 19-28. Yiian tai nung min chan cheng shih liao huipien 7tftBkB,W.¥-!£l%lilLM.See Yang Ne. Yiian tien chang TC^M . See Ta Yiian sheng cheng kuo ch'ao tien chang. Yule, Henry. The book of Ser Marco Polo. 3rd ed. rev. by Henry Cordier. 2 vols. London: Murray, 1903. Yung-lo ta tien * t& ~%.ft. Fragments. Peking: Chung hua shu chii, i960; repr. Taipei: Shih-chieh shu chii, 1962. Zhou Liangxiao (Chou Liang-hsiao). "On Khubilai khan." Social Sciences in China. 2 (1981), pp. 177-94- Cambridge Histories Online © Cambridge University Press, 2008