From Chang’an to Nālandā:
The Life and Legacy of the Chinese Buddhist Monk
Xuanzang (602?–664)
All but one of the articles collected in this volume are selected
from over fifty papers originally presented at the first international
conference on Xuanzang 玄奘 (602?–664) and Silk Road Culture,
held in the summer of 2018 at Guiyuan Monastery 歸元寺 in Chang’an.
The Guiyuan Monastery was built during the Zhenguan reign (627–
649) of the Tang dynasty to celebrate Xuanzang’s epochal return
to Chang’an from his protracted pilgrimage to Central and South
Asia. His epic journey resulted in some of the most significant
Sanskrit-to-Chinese translations and commentaries of Buddhist
scriptures, and the records of his extraordinary exploration are no
less impressive that centuries later, still fascinates the world. This
volume of scholarship delves into aspects of Xuanzang’s life, legacy,
and impact that continues to affect us today.
JI Yun and SHI Xingding
Guiyuan Temple
00068(平)
From the Ground Up: Buddhism &
East Asian Religions
Edited by SHI Ciguang, CHEN Jinhua,
Association for the Promotion
of Xuanzang Culture
From Chang’an to Nālandā:
The Life and Legacy of
the Chinese Buddhist Monk
Xuanzang (602?–664)
Edited by SHI Ciguang, CHEN Jinhua, JI Yun and SHI Xingding
From Chang’an to Nālandā:
The Life and Legacy of
the Chinese Buddhist Monk
Xuanzang (602?–664)
Proceedings of the First International Conference
on Xuanzang and Silk Road Culture
Edited by
SHI Ciguang
CHEN Jinhua
JI Yun
SHI Xingding
COVER IMAGE: Map showing Xuanzang’s travels from Sogdians website. Courtesy
of Freer Gallery of Art and Arthur M. Sackler Gallery, Smithsonian Institution,
Washington, D.C., Map by CHIPS.
Table of Contents
From Chang’an to Nālandā:
The Life and Legacy of the Chinese Buddhist Monk
Xuanzang (602?–664)
Shi Ciguang 釋慈光
Preface . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1
1. Doctrinal Studies
1.1. Yamabe Nobuyoshi 山部能宜
A Hypothetical Reconsideration of the ‘Compilation’ of Cheng
Weishi Lun . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6
1.2. Ernest Billings (Billy) Brewster
Survivability: Vasubandhu and Saṅghabhadra on the Continuity
of the Life of a Sentient Being as Translated by Xuanzang . . . . . . . 79
1.3. Li Zijie 李子捷
The Transformation of the Theory of Zhongxing 種姓 (Skt. Gotra)
before Xuanzang’s Translations: With a Focus on the Pusa
Yingluo Benye Jing 菩薩瓔珞本業経 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 137
1.4. Dan Lusthaus
What is ‘New’ in Xuanzang’s New Translation Style? . . . . . . . . . . . 159
1.5. Richard D. McBride II
How Did Xuanzang Understand Dhāraṇī?: A View from His
Translations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 220
FROM CHANG’AN TO NĀLANDĀ:
THE LIFE AND LEGACY OF THE CHINESE BUDDHIST MONK
XUANZANG (602?–664)
2. Historical And Biographical Studies
2.1. Shigeki Moro 師茂樹
Biography as Narrative: Reconsideration of Xuanzang’s
Biographies Focusing on Japanese Old Buddhist
Manuscripts . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 252
2.2. Jeffrey Kotyk
Chinese State and Buddhist Historical Sources on Xuanzang:
Historicity and the Da Ci’en Si Sanzang Fashi Zhuan 大慈恩寺
三藏法師傳 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 270
2.3. Guo Wu 伍國
Context and Text: Historicizing Xuanzang and the Da Tang
Xiyu Ji . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 311
3. Transborder And Transcultural Perspectives
3.1. Max Deeg
How to Create a Great Monastery: Xuanzang’s Foundation
Legend of Nālandā in Its Indian Context . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 334
3.2. Arun Kumar Yadav
The Mahābodhi Temple: Centre of Indo-Chinese Cultural
Exchange . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 365
3.3. Yu Xin 余欣
Archaeological Evidence, Cultural Imagination and Image of
the Medieval World: New Perspectives on Treasures from
Qiuci. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 387
FROM CHANG’AN TO NĀLANDĀ:
THE LIFE AND LEGACY OF THE CHINESE BUDDHIST MONK
XUANZANG (602?–664)
3.4. George A. Keyworth
On Xuanzang and Manuscripts of the *Mahāprajñāpāramitāsūtra at Dunhuang and in Early Japanese Buddhism . . . . . . . . . . . . . 437
4. Appendix
4.1. Siglinde Dietz
The Xuanzang Project at the University of Göttingen . . . . . . . . . . . 498
Author Biographies . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 519
What is ‘New’ in Xuanzang’s
New Translation Style?
DAN LUSTHAUS
Harvard University
Abstract: Xuanzang is one of the rare translators of Indic material
into Chinese who was himself Chinese, and the most prolific of
any translator. His translations cover all Buddhist genres, including
āgamas, Mahāyāna sutras, devotional sutras on specific buddhas and
bodhisattvas (e.g. Maitreya, Avalokiteśvara, Amitābha, and Kṣitigarbha),
avadāna, dhāraṇīs, abhidharma, Indian commentaries, Madhyamaka, Yogācāra, hetu-vidyā, and his famous travelogue, Xiyu ji 西域記.
His translation of the Mahā-Prajñāpāramitā-sūtras (600 fascicles
in three entire Taishō volumes) is by far the largest translation in the
Chinese canon, followed by his 200-fascicle translation of the Sarvāstivādin Mahāvibhāṣa. His Heart Sūtra translation is still recited daily
throughout the East Asian Buddhist world. His translations are often
labeled the ‘New Translation’ style, indicating a break or change with
his predecessor’s efforts. This presentation will examine what is ‘new’
in his translations.
Xuanzang introduced new Chinese equivalents for Indic terms
that had already acquired standard Chinese renderings, though this
was a gradual process, as I will illustrate with a few examples (e.g., his
treatment of √kḷp terms). He is also frequently credited with being a
more literal translator, which turns out to not always be the case (e.g.
his translation of the Abhidharmakośa-bhāṣya), though often his
translations are more accurate than earlier translations of the same
texts (e.g. Vimalakīrti-nirdeśa sūtra). But these technical details are
only part of the story. What was truly new and revealing about his
From Chang’an to Nālandā: The Life and Legacy of the Chinese Buddhist Monk Xuanzang (602?–664): 000–000
159
160
philosophical orientation were his choices of texts to translate, which
included numerous texts previously translated by others that he felt
needed newer, more accurate translations, as well as new texts that
introduced new facets of Indian thinking to a Chinese audience previously unaware of these Indian developments. Taking all of this into
account, I will attempt to shed some light on Xuanzang’s thinking
and orientation beyond the usual stereotypical accounts.
Keywords: Xuanzang, translation, Yogācāra, Heart Sūtra,
Yogācārabhūmi, Asaṅga, Vasubandhu, body
Overview
uanzang 玄奘 (600–664) is credited with having introduced
a ‘new translation’ style. Usually, this is taken to mean that he
replaced previously established Chinese equivalents that had been
used for Sanskrit and Indic terms with new equivalents, many of
which became the new standard equivalents. That is certainly part
of what was new about his translations. Also, frequently Xuanzang
is acclaimed as a more ‘literal’ translator, meaning that unlike many
of his predecessors, his translations more literally reproduced Sanskrit
texts in Chinese idiom, even, at times, following Sanskrit syntax
rather than the normative Chinese syntax of the day. While Xuanzang did provide a wealth of new equivalents, and some of his translations are closer to Indian originals than were the efforts of earlier
translators, some of his translations take liberties, introducing glosses
and interpretive extrapolations; he was not the literal translator he is
often imagined to be. I shall illustrate that with some examples later.
However, while not a strict literalist, his translations were usually
‘accurate’, in the sense of conveying the meaning of the Indian texts,
even when he took liberties, made tacit implications in the Sanskrit
explicit in Chinese, rearranged the order of passages, dropped superfluous phrases, and in other ways produced a Chinese text that was
not an isomorphic representation of its Sanskrit original.
X
161
What has received less attention is the contextual agenda that
guided his choice of which texts to translate. Xuanzang was driven
to make the perilous journey to India due to a strong desire to
resolve the multitude of conflicting understandings of Buddhism
that were roiling in early seventh-century China. Buddhism in
sixth-century China had largely been a battleground of competing
Yogācāra schools—the so-called northern and southern Dilun 地論
schools, several texts and commentaries by Vasubandhu translated
in the mid-sixth century by Vimokṣaprajñā (Pimu zhixian 毘目智
仙) and Gautama Prajñāruci (Qutanboreliuzhi 瞿曇般若流支),1 and
of course the Yogācāra translations by Paramārtha (Zhendi 真諦,
499–569). During Xuanzang’s early training in China, Paramārtha’s
works, especially the Mahāyānasaṃgraha (She Dasheng lun 攝大
乘論, T no. 1592; Shelun for short), and for the more studious, the
Abhidharmakośa-bhāṣya (Apidamo jushe lun 阿毘達磨倶舍論, T no.
1558) dominated discussions. Harivarman’s Tattvasiddhi (Chengshi
lun 成實論, T no. 1646) was also still well studied at that time. Xuanzang’s Biography by his contemporary colleagues, Huili 慧立 and
Yancong 彥悰, the Da Cien si sanzang fashi zhuan 大慈恩寺三藏法
師傳 (T no. 2053; hereafter Sanzang fashi zhuan), reports that before
he went to India, Xuanzang studied and lectured on the Shelun, and
an abhidharma text, and studied well the Abhidharmakośa and Tattvasiddhi.2 Since none of the central Asian monasteries he encounAmong their translations, jointly or together: Vasubandhu’s Karmasiddhi-prakaraṇa (Yechengjiu lun 業成就論, T no. 1608), Vasubandhu’s Sanjuzujing-upadeśa 三具足經憂波提舍 (T no. 1534), Vasubandhu’s upadeśa on the
Ratnacūḍa sūtra 寶髻經四法憂波提舍 (T no. 1526), all translated in 541, Vasubandhu’s Viṃśikā-vṛtti (Weishi lun 唯識論, T no. 1588, tran. ca. 538–543), etc.
2
On Xuanzang’s early education, the Biography says:
…the Master attended the lectures given by Daoji and Baoxian on the Mahāyānasaṃgraha Śāstra and he studied abhidharma with instruction by Dharma master
Daozhen on Kātyāyana’s (or Katyāyanīputra 迦多衍尼子) Jñānaprāsthana. Not
wasting a moment of time, he studied with full effort tirelessly; and within two or
three years he thoroughly mastered the Buddhist texts of different schools…. 諸德既
1
萃, 大建法筵, 於是更聽基, 暹《攝論》,《毘曇》及震法師《迦延》, 敬惜寸陰, 勵精
162
無怠, 二三年間, 究通諸部. (T no. 2053, 50: 1.222a17–20)
[At the age of twenty, f]rom summer to winter, the Master lectured on the
Mahāyānasaṃgraha and the Jñānaprāsthana, each three times…. 法師為講《攝
論》
、
《毘曇》, 自夏及冬, 各得三遍. (T no. 2053, 50: 1.222b10–11)
Then he went to Zhaozhou and visited the reverend teacher Daoshen, from
whom he learned about the Satyasiddhi (i.e. Tattvasiddhi) Śāstra. Then he entered Chang’an and stayed at the Great Enlightenment Monastery, where he
learned about the Abhidharmakośa Śāstra from the reverend teacher Daoyue.
又到趙州, 謁深法師學《成實論》. 又入長安, 止大覺寺, 就岳法師學《俱舍論》.
(English translation from Li, Record of the Western Regions, 15–17, modified;
T no. 2053, 50: 1.222b18–20)
Note that Li, in his translation of this passage, misunderstood 毘曇 (modern
pronunciation pitan, medieval probably closer to bhidham), an abbreviation for
‘abhidharma’, as referring to Asaṅga’s Abhidharmasamuccaya, that text had not
been translated into Chinese yet, so he would have been unable to study it prior
to encountering it later in Sanskrit; it was Xuanzang himself who first translated it after returning to China. There is some speculation that a translation by
Paramārtha of the Abhidharmasamuccaya might be one of his works no longer
extant, but if Paramārtha didn’t translate the Abhidharmasamuccaya, then we
can probably rule that out. 毘曇 by itself is insufficient to identify a specific text,
it seems, unless one takes the whole phrase《毘曇》及震法師《迦延》as ‘he studied abhidharma with instruction by Dharma master Daozhen on Kātyāyana’s
(or Katyāyanīputra 迦多衍尼子) Jñānaprāsthana’. If so, then it may refer to the
Jñānaprāsthana 阿毘曇八犍度論 (T no. 1543), a core Sarvāstivādin Abhidharma
text attributed to *Kātyāyana 迦旃延, translated by Saṃghadeva (Sengqietipo 僧
伽提婆) and (Zhu) Fonian 竺佛念 in 383 CE. Xuanzang later served up another
translation of this text: Apidamo fazhi lun 阿毘達磨發智論 (T no. 1544). Also
known by the titles Apitan jing ba jiandu lun 阿毘曇經八犍度論, Jiazhanyan
apitan 迦旃延阿毘曇; this is the seventh volume of the Sarvāstivāda Abhidharma
piṭaka, often considered—along with the Mahāvibhāṣa which is a commentary
on it—the central canonical text of the Sarvāstivādins.
In general, Li Rongxi’s translation is good at the Chinese idioms and syntax,
but less reliable at identifying Sanskrit terms, names, titles. The punctuation
framing 毘曇 as a text title is from CBETA. 毘曇 is sometimes used as a shorthand title of the Abhidharmakośa, but since the Biography identifies the Kośa a
short bit later as 俱舍論, when it says Xuanzang was first introduced to it, 毘曇
would have to be something else.
163
tered on his way to India were Mahāyāna, his discussions and debates
with their scholars focused on abhidharma topics, and, according to
the Biography, in debate he was invariably victorious, indicating his
mastery even then of the intricacies of abhidharma doctrines. Even
later in life, as a full-fledged Yogācāra, he was at heart an ābhidharmika and understood Yogācāra theory through an abhidharmic lens.
Usually it is said that he went to India to obtain the complete
Yogācārabhūmi (Yuqieshi di lun 瑜伽師地論), which Paramārtha had
only partially translated (Shiqi di lun 十七地論, no longer extant), believing its comprehensive overview of Buddhist doctrine and practice
would resolve the discrepancies and disputes raging in China at that
time. That is partially true, but while in India he came to recognize
that Indians understood Buddhism quite differently from what
had developed in China. After many years in India, he had a dream
which he took as a portent that Buddhism would disappear from
India. That disturbed him, but when he spoke about it with alarm
to his Indian colleagues, their calm resignation that all things are impermanent disturbed him further. Since the full and correct range of
Buddhist teachings had not reached China yet, if it disappeared from
India then China and East Asia would always be following an incomplete and, in some ways, misunderstood Buddhism. Asking a Jain
fortune teller whether he should stay in India, where he had achieved
prestigious status, or return to China by the difficult central Asian
route, the fortune teller told him he would live a long and successful
life if he remained in India, but if he returned to China, his health
would suffer, and his life would be shortened. Alarmed that Buddhism might disappear from India, and concerned that the proper
context and understanding of Buddhism was yet to be implanted in
China, he decided to make the journey back, bringing over 600 texts
from various Buddhist and non-Buddhist schools, as well as Buddhist artworks. As the fortune teller predicted, Xuanzang did suffer
health problems at various times after his return, sometimes temporarily impacting his productivity and output, but he endeavored
to provide China with as much of the context of Indian Buddhism
as he could, becoming the most prolific translator in Chinese Buddhist history. One text alone, his translation of the Prajñāpāramitā
corpus, is 600 fascicles, fills three entire Taishō volumes, and he ac-
164
complished this during his final years, when his health continued to
decline, while simultaneously translating numerous other texts.
His prolific activity was driven by his wish to provide Chinese
Buddhists with a fuller and more accurate context for understanding
Buddhism. He didn’t write original philosophical treatises of his
own,3 so trying to determine his own thoughts on things is challenging. One type of source for discovering what he himself thought and
said would be to examine reports from his contemporaries, such as
Kuiji and Wŏnch’ŭk 圓測 (613–696), being cautious to gauge their
His travelogue, Record of Western Lands (Datang Xiyu ji 大唐西域記, T
no. 2087), was compiled by Bianji 辯機 (?–649) from meticulous notes Xuanzang had taken during his travels, indicating distances and directions so accurately that Aurel Stein and others found forgotten sites in Central Asia by following
his directions and descriptions. It contains geographic, ethnographic, anthropological and religious accounts of different localities and their people, but provides
legendary anecdotes of key figures rather than philosophical depositions. The
Cheng weishi lun 成唯識論 (T no. 1585) is mistakenly considered a compendium of ten Sanskrit commentaries on the Triṃśikā championing the opinions of
Dharmapāla, since that is how Kuiji 窺基 (632–682) claimed it was composed,
but actually, its core structure is Sthiramati’s commentary on the Triṃśikā, embellished with a wide variety of other sources, not just Triṃśikā commentaries.
This is the only translation he produced that was not a direct translation of a
text, but an edited compendium, which he produced at Kuiji’s insistence, despite
his own intention to translate each text individually. Xuanzang felt guilty about
producing this hodgepodge, which is why he determined to not omit a single
word or phrase, no matter how redundant, in his Prajñāpāramitā translation
which he started the year after finishing the Cheng weishi lun. Cf. Lusthaus, Buddhist Phenomenology, ch. 15. While Kuiji promoted the Cheng weishi lun as the
catechism for his Weishi school, the extent to which it reflects Xuanzang’s own
thinking, rather than a survey of Buddhist theories is unclear. It is Kuiji, not the
Cheng weishi lun that attributes positions to various authors, invariably claiming
which is the correct interpretation and attributing that to Dharmapāla. The only
Triṃśikā commentary that has survived in the original Sanskrit is Sthiramati’s,
and it turns out that many of the positions Kuiji attributed to Dharmapāla are
actually Sthiramati’s. Cf. Keenan, ‘A Study of the Buddhabhūmyupadeśa’, 306.
3
165
reliability. Another, sounder, way to investigate Xuanzang’s priorities
and agenda would be to survey what he selected for translation and
the order in which they appeared.
His Agenda and Motivations: Xuanzang’s Early Translations
What does his early translation activity reveal about Xuanzang’s motives and agenda? If we assume, as we probably should, that his order
of translation was not random, but cumulative, so that earlier translations were intended to provide context and requisite background to
profitably work through later works, then the order may be viewed as
a systematic syllabus of sorts. That he may have modified his subsequent efforts based on how earlier ones were received should also be
considered. Also, it was common for translators to accede to requests
by patrons and disciples, so that too would have affected his choices
over the years. We are told in his Biography that his last major translation was to be the Ratnakūta sūtra, which he began, and then, his
failing health overcoming him, he stopped after a few lines.4 In any
From Huili’s Biography of Xuanzang:
On the first day of the first month, in the spring of the first year of Linde
(664), the monk-translators and other monks of Yuhua Monastery earnestly requested that the Master translate the Maha-ratnakūta Sūtra into Chinese. Upon
seeing the sincerity of the monks, he exerted himself to translate the sutra; but
after doing just a few lines he closed the Sanskrit text and stopped the task. He
told the monks, ‘This sutra is as voluminous as the Mahā-prajñāpāramitā Sūtra.
Estimating my own strength, I shall not be able to complete this work. I am approaching my time, and it is not far-off. Now I wish to go to the Lanzhi Valley
and other places to pay my last homage to a koṭi of Buddha’s images.’ Then he
went out with his disciples, and the monks looked at one another with tearful
eyes. After worshipping the images, he returned to the monastery and engaged
exclusively in practising the Way, doing no more translation. 麟德元年春正月朔
4
一日, 翻經大德及玉華寺眾慇懃啟請翻《大寶積經》. 法師見眾情專至, 俛仰翻數
行訖, 便攝梵本停住, 告眾曰:‘此經部軸與《大般若》同, 玄奘自量氣力不復辦此,
死期已至, 勢非賒遠. 今欲往蘭芝等谷, 禮拜辭俱胝佛像.’於是與門人同出, 僧眾
166
case, his translations provide his idea of what constitutes the necessary orientation, and where he wanted to intersect with and modify
the trajectory of Chinese Buddhism. Let us examine the order of
translations during his first years back in China.
First Year
In 645, freshly returned from India via the Silk Road, the first text
he produced was a translation of the Bodhisattva-piṭaka-sūtra (Da
pusa zang jing 大菩薩藏經, T no. 310, fascs. 35–54),5 which at some
point was incorporated into the Ratnakūta sūtra compilation as the
twelfth sūtra of that collection. It provides a detailed overview of
the bodhisattva project, taking up topics such as the six pāramitās,
the Four Immeasurables, the four methods for converting people to
Buddhism, and so on—a perfect choice for an introductory text on
pursuing the bodhisattva path.
This was followed by Asaṅga’s (Wuzhu 無著) Root Verses of the
Exposition of the Ārya Teachings (Xianyang shengjiao lun song 顯揚
聖教論頌, *Prakaraṇāryavākā?, T no. 1603), a summary extract of
important themes in the Yogācārabhūmi, sometimes with verbatim
passages, sometimes with slight variations.6 To prepare Chinese readers for the full-fledged Yogācārabhūmi, he first offers these selections.
Next, he translated a sutra that clearly was of great importance
for him, the Buddhabhūmi sūtra (Fodi jing 佛地經, 1 fasc., T no.
相顧, 莫不澘然. 禮訖還寺, 專精行道, 遂絕翻譯. (Li, Record of the Western Re-
gions, 331; Sanzang fashi zhuan, T no. 2053, 50: 10.276c2–9)
5
On the Bodhisattva-piṭaka-sūtra, cf. de Jong’s review of Ulrich Pagel’s
Bodhisattvapiṭaka.
6
Since this text is unknown apart from Xuanzang’s translations, it is impossible to determine whether the discrepancies are simply due to different translation choices on Xuanzang’s part or substantial differences in the underlying Sanskrit, or even whether and how this text might have circulated India. Perhaps it
had served as a students’ primer, or perhaps even compiled by Xuanzang himself
while in India during his studies of the Yogācārabhūmi.
167
680). But this sūtra was overshadowed in subsequent tradition by
later works, possibly since this early work long preceded Kuiji’s
involvement (Kuiji, born in 632, was only twelve or thirteen years
old at the time, and would not join the saṅgha for at least another
five years, and some years after that came into Xuanzang’s inner
circle). While much later Xuanzang obviously encouraged Kuiji to
study the Yogācārabhūmi, which Kuiji wrote a commentary on, his
knowledge of the Buddhabhūmi sūtra and its composite commentaries, the Buddhabhūmyupadeśa,7 was much weaker judging by his
scant and largely uninformative remarks on them.8 The sūtra, like
Asaṅga’s verse text noted above, consists of a single fascicle, but is
necessary for following the commentaries on it contained in the Buddhabhūmyupadeśa, which Xuanzang was eager to introduce to the
Chinese audience.
His next translation was a single fascicle dhāraṇī text, Saṇmukhidhāraṇī (Liumen tuoluoni jing 六門陀羅尼經, T no. 1360).9 This was
intended in part, perhaps, to commemorate his return to China.
This was quickly followed by Asaṅga’s prose autocommentary to
the root verses that Xuanzang had recently translated, this one titled
Exposition of the Ārya Teachings (Xianyang Shengjiao lun 顯揚聖
教論, 20 fascs., T no. 1602), which he began in October of 645 and
completed in February 646. Again, these are excerpted summaries of
important themes in the Yogācārabhūmi, such as pramāṇa theory.
To sum up Xuanzang’s first year of translation, he began with an
introductory summary of the bodhisattva project; a summary of imBuddhabhūmyupadeśa (Fodi jing lun 佛地經論, T no. 1530, 7 fascs.), a
rich commentary on the Buddhabhūmi sūtra attributed to ‘Bandhuprabha (Qinguang 親光), etc.’, translated between November 12, 649 and January 2, 650.
8
For instance, he seems to not have noticed that large passages are quoted in
the Cheng weishi lun verbatim.
9
Katsumi Mimaki published a critical edition of the Sanskrit text along with
the Tibetan and Chinese and his French translation: ‘La Ṣaṇmukhī-dhāraṇī ou
“Incantation des SIX PORTES”’, and ‘La Ṣaṇmukhī-dhāraṇī ou ‘Incantation
des SIX PORTES”’. Davidson, ‘Studies in dhāraṇī literature II’, 5–61, includes
excerpts from this text.
7
168
portant themes in the Yogācārabhūmi; a key sutra important to the
Yogācāra of that day, which envisions what a Buddha land entails and
how things look through awakened cognition; and a dhāraṇī text
designed to open the six gates, i.e., one’s cognitive apprehension of
reality.
Second Year
In April of 646, Xuanzang’s translation of Sthiramati’s (Anhui 安
慧) commentary on Asaṅga’s Abhidharmasamuccaya (Dasheng
apidamo zaji lun 大乘阿毗達摩雜集論, 16 fascs., Abhidharmasamuccaya-vyākhyā, T no. 1606) appeared.10 Let’s pause to take note
of what this indicates. We are in the second year of his translation
activity, and so far nothing from Vasubandhu (Shiqin 世親) has
been tackled, but we now have three Asaṅga texts (albeit one is a
prose fleshing-out of another). Also, nothing from Dharmapāla has
been presented. Instead, Sthiramati’s commentary on an important
Asaṅga text is delivered. Also, he has presented this commentary
before having translated Asaṅga’s own root text (which he translated
in 652). Much later, Kuiji, certainly at Xuanzang’s urging, did study
this Sthiramati commentary well, and wrote his own commentary on
it.11 Notably, this text further introduces the abhidharmic approach
to Yogācāra and vice versa in great detail.
Next, Xuanzang’s famous ethnographic travelogue, the Record of
Western Lands, appeared. Not long after, Bianji, who had compiled
it, was caught up in a sex scandal involving the emperor’s daughter,
and was summarily executed. That Xuanzang had entrusted such an
important task to Bianji suggests that he was grooming him to be a
key disciple; that Bianji had connections with the court, and lived
on the grounds of the emperor’s sister rather than a monastery, is a
10
The Sanskrit for this text was rediscovered in the twentieth century, and
published by N. Tatia, Abhidharmasamuccaya-bhāṣyam.
11
Zaji lunshu ji 雜集論述記, Dasheng Apidamo zaji lun shuji 大乘阿毗達磨
雜集論述記 (X no. 796).
169
reminder of Xuanzang’s own frequent diplomatic successes with
rulers in India and Central Asia, almost all of whom took pains to
accommodate him and grant him favors. That the sex scandal did not
tinge his own reputation indicates the power of his personality and
the prestige accorded him.
Third Year
An important Vasubandhu text finally appeared the following year,
in April 647: Pañcaskandha-prakaraṇa (Dasheng wuyun lun 大乘
五蘊論, 1 fasc., T no. 1612). This text is a proto-Yogācāra work; it
still retains many elements eventually jettisoned in mature Yogācāra,
such as the concept of rūpa-prasāda (qingjing se 淸淨色), but gives
an early version of ālayavijñāna while discussing the fifth skandha,
vijñāna. Its approach, like the Abhidharmasamuccaya, was primarily
abhidharmic: defining terms with brief examples and exposition.
This was followed by working on a text that should shed important light on Xuanzang’s agenda. As mentioned previously,
Paramārtha’s translation of Asaṅga’s Mahāyānasaṃgraha had
become a dominant text among Chinese Buddhists. Surveys of
what study groups of all sects at that time were reading invariably
include the Shelun on their lists, including early Chan schools.
Xuanzang’s experience in India indicated to him that Paramārtha’s
text had engendered a variety of faulty, even pernicious ideas, that
were detrimental to a proper understanding of Buddhism and especially Asaṅga’s Yogācāra. Since he himself had lectured in China on
Paramārtha’s translation of the Shelun, he was closely familiar with
its contents, and therefore must have been all the more surprised
to discover how differently the text itself read in Sanskrit and how
Indians understood it. Rather than write a critique disputing
Paramārtha’s interpretation, he chose instead to re-translate it in a
way that would provide a superior rendering and representation of
the ideas in the original. However, instead of merely offering up his
own translation of the Shelun, or even his own translation of Vasubandhu’s bhāṣya on it, both of which Paramārtha had translated
(T no. 1593 and no. 1595, respectively), he began to address the
170
misconceptions that had developed in China by translating *Āsvabhāva’s
(Wuxing 無性) sub-commentary on Vasubandhu’s commentary,
to make sure that readers could clearly see how Indians understood
these texts. This was the *Mahāyānasaṃgrahopani-bandhana (She
Dasheng lun wuxing shi 攝大乘論無性釋, 10 fascs., T no. 1598). This
translation was finished in 649, taking more than two years to complete (April 10, 647–July 31, 649).
He had begun working on the Yogācārabhūmi (Yuqieshi di lun
瑜伽師地論, 100 fascs., T no. 1579) in July 646 and finished it in
June 648. The finished text as it has come down to us contains two
colophons that give conflicting details on dates, translation assistants,
etc., but over twenty monks assisted. The emperor drafted monks
from their home temples to work with Xuanzang on this project;
some taking dictation, some serving as copyists, copy-editors, proof
readers, etc. The Chinese version attributes authorship to Maitreya,
while the Tibetan tradition attributes authorship to Asaṅga.
While the Yogācārabhūmi project continued, he tackled the
Saṃdhinirmocana sūtra (Jie shenmi jing 解深密經, T no. 676), the
key Yogācāra sutra credited by historians with introducing signature
Yogācāra doctrines, such as vijñapti-matra 唯識, ālayavijñāna, and
trisvabhāva 三自性. This sutra had been translated several times
before.
Of the texts translated by Xuanzang up to this point, while the
Āsvabhāva’s commentary to the Shelun had not been translated
before, Paramārtha had translated the Shelun itself; and Paramārtha
had also partially translated the Yogācārabhūmi, so now it was clear
that Xuanzang was dedicating himself to revising the Chinese understanding of texts Paramārtha had made important with translations
that, in Xuanzang’s understanding, had led to misconceptions. Prior
to Xuanzang, translations of the Saṃdhinirmocana were made by
Guṇabhadra (partial), Bodhiruci, and Paramārtha.12
Xiangxu jietuodi boluomi liaoyi jing 相續解脫地波羅蜜了義經, T no. 678,
tran. btw. 435–443 by Guṇabhadra (Qiunabatuoluo 求那跋陀羅 [394–468]),
which corresponds to the tenth chapter of the Samdhinirmocana; Xuangxu
jietuo rulai suozuo shuishunchu liaoyi jing 相續解脱如來所作隨順處了義經
12
171
In September of 647, Xuanzang introduced something that had
yet to appear in China: Buddhist hetuvidyā reflecting the sharpening of Buddhist logic engineered by Dignāga. The Nyāyapraveśa
(Yinming ru zhengli lun 因明入正理論, 1 fasc., T no. 1630), which
was written by Śaṃkarasvāmin (Shangjieluozhu 商羯羅主), provided
a concise manual of Dignāga’s logic system (with a few slight variations).
To sum up so far, Xuanzang is building a foundation for bodhisattva practice, based on foundational Yogācāra texts and concerns,
from how to understand the five aggregates that comprise a person,
to the progression of practice and understanding, to tools such as
dhāraṇīs and logic. He is introducing new materials to Chinese Buddhists, but also starting to ‘correct’ previous translations—especially
those of Paramārtha, in whose texts he had been immersed before
leaving China—to bring Chinese Buddhists closer to what he had
witnessed and absorbed of Indian Buddhism.
Fourth Year
In 648, he expanded the contextual framework. The Devatā sūtra
(Tian qingwen jing 天請問經, 1 fasc., T no. 592) recounts how
Buddha responded to questions from various devas, its main theme
being to replace the three poisons of greed, hatred and delusion
by practicing Buddhist ethics and following the precepts. This
dimension of Xuanzang’s project, perhaps best appreciated by his
contemporary and sometimes associate, Daoxuan 道宣 (596–667),
(T no. 679) translated by Guṇabhadra, corresponding to the ninth chapter;
Shenmi jietuo jing 深密解脫經 (T no. 675) translated by Bodhiruci (Putiliuzhi
菩提流支 [d. 527]) in 514; Foshuo jiejie jing 佛說解節經 (T no. 677) translated
by Paramārtha in 557. Guṇabhadra’s Chinese apparently was not up to the task
of translating these texts, so the actual translators were probably Baoyun 寶雲
along with Bodhi 菩提, Fayong 法勇, and Tanwujie 曇無竭. Cf. Chu sanzang ji
ji 出三藏記集 by Sengyou 僧祐 (445–518), T no. 2145, 55: 2.12.c19–13a8. And
Radich, ‘Text T. 0679’.
172
considered one of the founding figures of the East Asian vinaya
tradition, is often neglected in favor of either his travel writings or
more scholastic works. However, ethics and precepts were important
for Xuanzang, and the early Weishi 唯識 and Japanese Hossō 法相
groups engaged in social services of various kinds, from caring for the
ill to building bridges. They are also something emphasized repeatedly in the Yogācārabhūmi and a variety of other texts that Xuanzang
translated over the years.
Next he introduced a non-Buddhist text, *Daśa-padārtha
(Shijuyi lun 十句義論, 1 fasc., T no. 2138), a Vaiśeṣika text by Candramati (Huiyue 慧月) that was eventually forgotten in India.13
Early Vaiśeṣika posited six padārthas (fundamental components of
reality), while a key Vaiśeṣika reformer, Praśastapāda, increased that
to nine padārthas, but, aside from this Candramati text preserved for
us by Xuanzang, no trace of a ten padārtha system is found in the
extant Vaiśeṣika literature. Why did Xuanzang choose this unusual
Vaiśeṣika text to translate? There are several possibilities. First, this
may have been the Vaiśeṣika manual that Indian Buddhists at that
time studied to prepare for debates with Vaiśeṣikas. Another obvious
reason is that the primary polemical targets of many Buddhist texts
during the centuries leading up to Xuanzang’s time were Vaiśeṣika
and, and while Paramārtha had translated a Sāṃkhya text,14 no one
had provided a Vaiśeṣika text in Chinese. Chinese Buddhists must
have been curious about the actual tenets of that school, being only
familiar with the narrow arguments focused against them. Also, Xuanzang valued debate as an important tool for sharpening the mind
and overcoming misconceptions; to debate, the better one knows the
opponent’s framework, the more effectively one can recognize and
exploit its weaknesses. Additionally, most Buddhist ābhidharmikas
accepted some version of atomic theory, and the Vaiśeṣika held the
most developed non-Buddhist atomic theory in India. So this text
13
This was translated into English by Ui Hakuju with the Chinese text on
facing pages, Vaiśeṣika Philosophy, 93–119. A later, improved translation can be
found in Miyamoto, Daśapadārthī, 7–25.
14
Jin qishi lun 金七十論 (Sāṃkhya-kārikā), T no. 2137.
173
was probably translated to sate the curiosity of his students as well as
to help provide some understanding of the context in which Indian
Buddhism operated.
Next, he translated Vasubandhu’s Triṃśikā (Weishi sanshi lun
唯識三十論, 1 fasc. T no. 1586), which would later become the underlying foundation for the Cheng weishi lun. The version that has
come down to us includes a variety of deviations from the Sanskrit,
as well as framing interpolations.15 Paramārtha had produced an unusual rendering of this text, the Zhuanshi lun 轉識論 (T no. 1587),
in which Vasubandhu’s text is inextricably intermingled with a commentary of uncertain authorship (some speculate it was Paramārtha’s
own commentary). That would have been the only exposure East
Asian Buddhists had to this important Vasubandhu text prior to
Xuanzang’s new translation. Compared to the Zhuanshi lun, Xuanzang’s rendering is much closer to the Sanskrit original, despite his
deviations and interpolations, and many of those are accounted for
and explained in the Cheng weishi lun, so they were neither inadvertent nor accidents nor mistakes, but deliberate interpretive overlays.
He next translated the Diamond Sutra (Vajracchedikā sūtra;
Jin’gang banruo jing 金剛般若經, 1 fasc., T no. 220), which had been
previously translated by Kumārajīva (in 401), Bodhiruci (in 509),
and Paramārtha (in 558). While Xuanzang’s translation continued
to receive attention, Kumārajīva’s version remained the traditional
favorite.
This was followed by his translation of Vasubandhu’s Introduction
to the One Hundred Dharmas (Baifa mingmen lun 百法明門論, 1
fasc., T no. 1614), a listing of the Yogācāra abhidharma system of
100 dharmas, divided into categories.16
And then finally he completed his translation of Vasubandhu’s
bhāṣya to Asaṅga’s Mahāyānasaṃgraha (She Dasheng lun Shiqin
shi 攝大乘論世親釋, 10 fascs., T no. 1597). So now he had translated *Āsvabhāva’s subcommentary on the Mahāyānasaṃgraha
Some of this is discussed in Lusthaus, Buddhist Phenomenology, Part IV.
Cf. Lusthaus, ed. ‘The One Hundred Dharmas’. http://www.acmuller.net/
yogacara/outlines/100dharmas.html.
15
16
174
and Vasubandhu’s commentary, reinforcing that he was correcting
Paramārtha’s version by pointing out how Indian Buddhists understood the text.
Fifth Year
It is not until the following year, 649, that he finally tackled Asaṅga’s root text without the commentaries, Mahāyānasaṃgraha (She
Dasheng lun ben 攝大乘論本, 3 fascs., T no. 1594). Having instructed his readers on how to read the text by providing the expositions
of Vasubandhu and his sub-commentator, Āsvabhāva, they were now
ready to enjoy Asaṅga’s text without the distortions and misconceptions introduced by Paramārtha’s popular translation. As history has
shown, however, his versions never fully eclipsed the Paramārtha versions, since, after his death there was a concerted movement by a variety of leading figures, such as Fazang 法藏 (643–712) and Wŏnhyo 元
曉 (617–686), et al., to return to the Paramārtha approach.
At this point, in 649, Xuanzang produced the following translations:
Five sutras:
Yuanqi Shengdao jing 緣起聖道經, 1 fasc. (T no. 714, Nidāna
sūtra [Sutra on the Noble Way of Conditional Co-arising]), a
sutra on pratītya-samutpāda.
Shen xiyou jing 甚希有經, 1 fasc. (T no. 689, *Adbhūta-dharmaparyāya sūtra [Sutra on the Miraculous Acts of the
Buddha]).
Wangfa zhengli jing 王法正理經, 1 fasc. (T no. 1615, Sutra
of [Maitreya’s] Correct Principles of Royal Rule), a sutra
derived from the Yogācārabhūmi, authorship therefore
attributed to Maitreya 彌勒 (rather than Śākyamuni or
Asaṅga). Actually a combination of two sutras, the first
advising on how to govern, how to correct faults, and pursue
the wholesome; the second divides people into three types
and how to help each make progress.
Zui wubi jing 最無比經, 1 fasc. (T no. 691, Supreme
175
Incomparable Sutra), extolling the benefits of faith in
the three jewels (Buddha, Dharma, and Saṅgha), which,
according to this sutra is superior to following precepts.
Rulai shijiao Shengjunwang jing 如來示教勝軍王經, 1 fasc. (T
no. 515, Rājavavādaka sūtra, Sutra in which the Tathāgata
Reveals Teachings to King Prasenajit), encouraging the king
to pursue the Dharma rather than wealth or power.
The primary audience for these texts is the emperor, whose health
was fading at this time. He died either this year or the next (depending on which source one consults), and according to several sources
turned devotedly to Buddhism at the end of his life, in part due to
Xuanzang’s influence. For the more ‘professional’ or scholarly readers he produced the following.
One Madhyamaka text:
Bhāviveka’s (Qingbian 清辯) Dasheng zhangzhen lun 大乘掌珍
論, 2 fascs. (T no. 1578, *Karatala-ratna or *Hasta-maṇi?
Jewel in the Palm).
One Sarvāstivādin Abhidharma text
Shishen zu lun 識身足論, 16 fascs. (T no. 1539, Vijñāna-kāya
pāda), the third member of the Sarvāstivādin abhidharma
canon, attributed to Devakṣema (Tiposhemo 提婆設摩).
Two texts on Bodhisattva Precepts related to the Yogācārabhūmi
Pusa jie jiemo 菩薩戒羯磨, 1 fasc. (T no. 1499, The Rituals
and Customs for Bodhisattvas), excerpted from fasc. 40 of
the Yogācārabhūmi; 羯磨 = karma, in the technical sense
of rituals, precepts and customs. It deals with ordination,
repentance, and what happens when precepts are violated.
Pusa jie ben 菩薩戒本, 1 fasc. (T no. 1501, *Bodhisattva-śīla
sūtra), excerpts from the Yogācārabhūmi on forty-two
precepts guiding a monastic’s behavior.
Since drawn from the Yogācārabhūmi, the authorship of these
two texts is attributed to Maitreya. These again highlight Xuanzang’s
176
concern with ethics, and the orderly and proper behavior expected of
sincere practitioners, especially in the context of Yogācāra practice.
One important Yogācāra commentary:
Fodi jing lun 佛地經論, 7 fascs. (T no. 1530,
Buddhabhūmyupadeśa), attributed to Bandhuprabha
(Qinguang 親光), etc. A commentary on the Buddhabhūmi
sūtra which he had translated during his first year back in
China. This composite of what appear to be at least three or
four distinct commentaries was already mentioned above.
Many key Yogācāra ideas are explained here with details not
found in other texts. Since Bandhuprabha was apparently
active at Nālandā when Xuanzang was there, this represents
the state of the art in Yogācāra thinking among Xuanzang’s
Indian contemporaries.
That brings us to the year 650—he goes on to translate more
abhidharma, Yogācāra and Prajñāpāramitā texts, and more dhāraṇī
texts—but we can stop this survey here, and turn to another issue
already raised but played out in another series of texts.
The Case of the Abhidharmakośa-bhāsya (and Mahāyānasaṃgraha)
First, to quickly summarize what has been shown so far. What is
new in Xuanzang’s approach includes not only new texts (and the
new terminology they introduced, which we haven’t discussed yet),
but a reframing of Buddhism to align Chinese Buddhism with the
theories and practices of Indian Buddhists, for whom, for instance,
logic was key, and for whom there were Mahāyānic precepts (not just
the Dharmagupta, etc. vinayas that had been adopted by East Asian
Mahāyāna Buddhists), and, of course, a more precise and accurate
presentation of Yogācāra ideas. In order to challenge the popular
translations by Paramārtha, he approached that by first translating
Indian commentaries on those texts to make clear how Indians read
those texts, and only then re-translating the root texts themselves.
As mentioned above, even prior to leaving China, Xuanzang was
177
considered an expert in the Abhidharmakośa-bhāṣya, which he had
studied in Paramārtha’s translation.
…he entered Chang’an and stayed at the Great Enlightenment Monastery, where he learned about the Abhidharmakośa Śāstra from the
reverend teacher Daoyue. He grasped the essential meanings of all
these texts by studying them only once, and could memorize whatever had passed his eyes, an ability unsurpassed even by senior scholars
of deep learning. He studied with such profundity that he could
comprehend subtle meanings and reveal what was hidden in the texts
when others failed to reach it. On more than one abstruse point he had
his own particular views.
又入長安, 止大覺寺, 就岳法師學《俱舍論》. 皆一遍而盡其旨, 經目
而記於心, 雖宿學耆年不能出也. 至於鉤深致遠, 開微發伏, 眾所不
至. 獨悟於幽奧者, 固非一義焉.17
He continued to study and learn, and when he reached Kashmir
he studied with Saṃghakīrti, who was over seventy at the time.
But as he [Saṃghakīrti] was glad to have met an intelligent person,
he exerted himself to the utmost to teach him by lecturing on the
Abhidharmakośa Śāstra before noon, the Nyāyānusāra in the afternoon, and hetuvidyā [logic] and śabdavidyā [grammar] after the first
part of the night. Thus all the scholars in the locality assembled to
attend the lectures. The Master comprehended whatever was spoken
by the teacher without missing anything. He studied the subtle
teachings with appreciation and thoroughly mastered the mysteries.
彼公是時年向七十, 氣力已衰, 慶逢神器, 乃勵力敷揚. 自午已前講
《俱舍論》, 自午已後, 講《順正理論》, 初夜後講《因明》
、
《聲明
論》. 由是境內學人, 無不悉集. 法師隨其所說, 領悟無遺, 研幽擊
節, 盡其神祕.18
Here from Li Rongxi’s English translation, see Li, Record of the Western
Regions, 17, italics added; Sanzang fashi zhuan, T no. 2053, 50: 1.222b19–23.
17
178
In Kashmir further study of the Kośa, in Sanskrit, not Chinese,
was taught by Saṃghakīrti in tandem with Saṃghabhadra’s
(Sengqiebatuoluo 僧伽跋陀羅; i.e. Xianzhong 衆賢) Nyāyānusāra,
a detailed critique of the Kośa, defending the orthodox Kashmiri
Vaibhāṣika positions from the Kośa’s misrepresentations and fallacies. Xuanzang eventually also translated the Nyāyānusāra in 80
fascicles (Apidamo shun zhengli lun 阿毘達磨順正理論, T no. 1562).
Saṃghakīrti also gave Xuanzang instruction in Buddhist logic and
Sanskrit grammar, so at this point Xuanzang was learning the Kośa
in its original Sanskrit, along with the fuller context of the Vaibhāṣika
disputes it engaged. Discrepancies between what he knew from
Paramārtha’s Chinese translation and what he was now learning
encouraged him to dig more deeply and critically into not only new
sources but the sources he thought he already knew.
Just as was the case with the Mahāyānasaṃgraha in which
Xuanzang first translated the sub-commentary, then the commentary, and finally the basic text itself, in the case of the Abhidharmakośa-bhāṣya, he first set out to translate Indian contextual material
before attempting to replace the earlier Paramārtha translation with
his new version. Saṃghabhadra had written two critiques of the
Kośa. First Xuanzang translated Saṃghabhadra’s Revealing the
Tenets of the Abhidharma piṭaka (Apidamo zang xianzong lun 阿
毘達磨藏顯宗論, T no. 1563) in 40 fascicles which he worked on
between April 30, 651 and November 26, 652. He then translated
Saṃghabhadra’s more detailed and comprehensive critique, the
aforementioned Nyāyānusāra in 80 fascicles, between February
3, 653 and August 27, 654. He was working on his translations of
the Kośa verses and the Kośa with Vasubandhu’s autocommentary
during the same period (T no. 1560 and T no. 1558, respectively;
June 3, 651–September 13, 654), but they were released only after
the Saṃghabhadra critiques. Again, this illustrates that for Xuanzang context is necessary for proper reading and understanding.
The message for modern scholars is that the Kośa should be read in
English translation quoted from Li, Record of the Western Regions, 62, with
modifications; Sanzang fashi zhuan, T no. 2053, 50: 2.231b4–8.
18
179
tandem with the Mahāvibhāṣa and Nyāyānusāra, which, since they
only survive in Chinese have been largely ignored by scholars who
work with Sanskrit and/or Tibetan materials.
Literal, Non-literal: Contextual
At the beginning, I mentioned that Xuanzang was not the strictly
literal translator he is often imagined to be, which may have surprised some of you if you have never carefully read a Xuanzang text
in tandem with its Sanskrit counterpart. Examples to illustrate this
would fill volumes, so the examples to follow are only a very brief
sampling which could be multiplied many times over. It would also
go beyond what this current paper can address to sort the ‘deviations’
from received Sanskrit counterparts into various types and further
analyze, on a case by case basis, how and why Xuanzang’s texts
don’t exactly match the Sanskrit—is it because our received Sanskrit
version(s) are later redactions that differ from what Xuanzang was
working with? Was he drawing on an exegetical tradition or principle,
written or oral, that guided his modifications? Did he misunderstand
the original text? Was he unduly influenced by prior translations and/
or translators in ways he failed to overcome? Since the translation
process often involved the main translator orally reciting and orally
translating the Indic original, with others, namely assistants, transcribing by dictation, and still others polishing, proof-reading, and
comparing what was being captured in writing against prior related
texts, and, in the case of Xuanzang, often turning the Chinese into
neat four- or eight-character phrases, which requires padding here
and abbreviating there, could deviations inadvertently have entered
in this ‘transmission’ process?
I am confident that given his prodigious output and the pace at
which he worked, he rarely proof-read the final products, perhaps at
most spot checking, or addressing concerns assistants would bring to
his attention. He was too busy translating to explain the texts to his
assistants in detail, trusting them to make sense of the texts they were
working on with minimal explanation from him.
The most famous example of this, leading to divergent com-
180
mentaries and subsequently a major political conflict, concerned
his translation of Dignāga’s Nyāyamukha, a logic manual. His ten
assistants on that translation each wrote their own commentary, but
their lack of clear understanding is evident in the fact that each had a
different interpretation from the other, often missing key points in
the logic system, so that a court Daoist, Lü Cai 呂才, to one-up the
Buddhists and prove he could understand anything no matter how
arcane or obscure, arrogantly claimed to have outsmarted all of them
and offered his own commentary. Buddhists and Daoists in the capital fought over it, outraged Buddhists submitted memorials to the
emperor to censure Lü, and rancor between Buddhists and Daoists
intensified throughout China until the emperor forced Xuanzang
to declare one way or another whether Lü’s commentary had merit.
After unsuccessfully trying to avoid passing judgement, Xuanzang
finally conceded that Lü’s interpretation was baseless, so Lü lost face.
Xuanzang never translated another logic text, to the detriment of the
East Asian Buddhist tradition. The entire eighth fascicle of Huili’s
Biography of Xuanzang documents that controversy, Huili himself
being one of the people who petitioned the emperor against Lü Cai.
To provide quick and clear examples in which Xuanzang can be
shown to be doing something other than faithfully reproducing
the Sanskrit, rather than wade into technical and complicated philological waters (such as his use of xingxiang 性相 in his Triṃśikā
translation when nothing corresponds in the Sanskrit, though he
unpacks this in the Cheng weishi lun as shorthand for zixing 自性
and xingxiang 行相, svabhāva and ākāra, i.e., what something is and
what it does, which he uses to analyze types of consciousness, though
neither term appears in the Sanskrit Triṃśikā), I will instead offer
two types of illustrations. First, there are some significant differences
between Xuanzang’s Heart Sūtra and the Sanskrit versions that have
come down to us. Second, it can be easily shown that Xuanzang deviates from the Sanskrit when lists are given, and even something as
basic as the number of items in Sanskrit and Chinese is not the same,
and further the order of items do not match all the way through.
Two lists from the Yogācārabhūmi will be used to illustrate this, both
from the Śrāvakabhūmi section.19
181
Heart Sūtra Discrepancies
Turning to the Heart Sūtra, in the Sanskrit, when Avalokiteśvara
looks down at the world, he ‘sees that the five skandhas are empty
of svabhāva’ (pañca-skandhās tāṃś ca svabhāva-śūnyān paśyati
sma), while Xuanzang’s Chinese says only that he sees that the five
skandhas are all empty (zhaojian wuyun jie kong 照見五蘊皆空).
Philosophically, failing to mention svabhāva is a significant omission
while adding ‘all’ 皆 is a trivial gloss. Additionally, the Sanskrit term
for ‘looks down’ is a pun on Avalokiteśvara’s name: Ārya Āvalokiteśvaro bodhisattvo gambhīraṃ prajñāpāramitā-cāryāṃ caramāṇo
vyavalokayati sma. The pun is lost in Chinese: 觀自在菩薩行深般
若波羅蜜多. Instead of ārya Avalokitsvara, the Chinese entitles him
‘bodhisattva’ 菩薩.
In the next section, the Sanskrit gives three paired phrases, while
Xuanzang’s Chinese gives only two (see detailed analysis in Appendix
I):
Iha Śāriputra
rūpaṃ śūnyatā, śūnyat'aiva rūpaṃ,
rūpān na pṛthak śūnyatā, śūnyatāyā na pṛthag rūpaṃ,
yad rūpaṃ sā śūnyatā, yā śūnyatā tad rūpaṃ.
Here, Śāriputra,
form is emptiness, emptiness is only form;
form is not different than emptiness, emptiness is not different
than form;
what is form is emptiness, what is emptiness that is form.
19
Deviations in lists between Sanskrit and Chinese versions of a text is not
unique to Xuanzang, but we are solely concerned with his translations in this
paper. For an example of Kumārajīva deviating from the received Sanskrit Pañcaviṃśatikasāhasrikā, cf. Orsborn, ‘Something for Nothing,’ 179.
182
舍利子!
色不異空, 空不異色.
色即是空, 空即是色.
Śāriputra,
form is no different than emptiness, emptiness is no different
than form.
Form precisely is emptiness, emptiness precisely is form.
The Chinese omits the first line of the Sanskrit.
Further down, the Chinese says: ‘No suffering, no origination [of
suffering], cessation [of suffering], nor way [to end suffering]. No
wisdom and no attainment, and nothing to be attained.’ 無苦集滅道,
無聖亦無得以無所得. The Sanskrit contains an additional phrase: ‘no
non-attainment’ (na duḥkha-samudaya-nirodha-mārgā na jñānaṃ na
prāptir n’āprāptiḥ).
Citt’āvaraṇa-nāstitvād atrasto viparyās’ātikrānto niṣṭhānirvāṇaḥ.
Because an obstructed mind does not exist, he is not frightened. He has stepped over the conceptual perversions, finally
attaining Nirvāṇa.
故心無罣礙. 無罣礙故, 無有恐怖. 遠離一切顛倒夢想, 究竟涅
槃.
Because the mind has no obstructions, having no obstructions
therefore there is no fear. Completely detached from conceptually-perverted dream thoughts, this is final Nirvana.
The Chinese adds ‘dream’ 夢.
Whether one considers such deviations significant or trivial, at
the least they indicate that Xuanzang’s text is not a strictly literal
rendering of the Sanskrit as that has come down to us, and that even
is such a short text which was one of Xuanzang’s own most treasured
recitation texts, deviations from the Sanskrit are present.
183
Lists, Terms, and Discrepancies
We now will look at two sets of lists found in the Śrāvakabhūmi of
the Yogācārabhūmi. While this can seem tedious, they are useful
for simply and unambiguously highlighting non-literal renderings,
since for the most part grammar and syntax become irrelevant and it
becomes simply a matter of matching equivalents—until they fail to
match.
The first is a list of body parts considered impure. The Sanskrit
list and Xuanzang’s translation begin in alignment, and then things
grow increasingly harder to align. The Tibetan varies from both. We
will not dwell on all the specifics and complexities, but just take note
of their more obvious non-alignments. First, here are the texts, the
Sanskrit followed by the corresponding Chinese and then Tibetan.
Śrāvakabhūmi, list of internal (= in the body) impurities
-II-3-b-(1)-i-(a) Ms.72a1L, Sh.203-1, P.95b5, D.79a5, N.83b1,
Co.84a5, Ch.428c22
tatra pratyaśubhatā katamā / āha / pratyaśubhatādhyātmam
upādāya bahirdhā copādāya veditavyā //
tatrādhyātmam upādāya / tadyathā keśā, romāṇi, nakhā,
dantā, rajaḥ, (Śbh II 60) malaṃ, tvak, māṃsaṃ, asthi, snāyu,
sirā, vṛkkā, hṛdayaṃ, plīhakaṃ, klomam, antrāṇi, antraguṇaḥ,
āmāśayaṃ, pakvāśayaṃ, yakṛt, purīṣam, aśru, svedaḥ, kheṭaḥ,
śiṅghāṇakaṃ, vasā, lasīkā, majjā, medaḥ, pittaṃ, śleṣmā, pūyaḥ,
śoṇitam, mastakaṃ, mastakaluṅgaṃ, prasrāvaḥ //20
Numbering the items for cross-referencing purposes:
tatrādhyātmam upādāya / tadyathā keśā1, romāṇi2, nakhā3,
dantā4, rajo5, malam6, tvak7, māṁsam8, asthi9, snāyu10,
sirā11, vṛkkā12, hṛdayam13, plīhakam14, kloman15, antrāṇi16, antraguṇaḥ17, āmāśayam18, pakvāśayam19, yakṛt20**,
purīṣam21, aśru22, svedaḥ23, kheṭaḥ24, śiṁghāṇakam25,
20
Matsunami, Śrāvakabhūmi, 58–60.
184
vasā26, lasīkā27, majjā28, medaḥ29, pittam30, śleṣmā31,
pūyaḥ32, śoṇitam33, mastakaṁ34, mastaka-luṁgam35, prasrāvaḥ36 /
**The Shukla edition has mūtraṃ (water from the kidneys,
urine) here instead of yakṛt.21
The corresponding Chinese passage in Xuanzang’s translation
reads:
云何依內朽穢不淨謂內身中髮毛爪齒塵垢皮肉骸骨筋脈心膽
肝肺大腸小腸生藏熟藏肚胃髀/脾腎膿血熱痰肪膏肌髓腦膜洟
唾淚汗屎尿如是等類名為依內朽穢不淨 22
This is sometimes parsed this way:
云何依內, 朽穢不淨? 謂內身中, 髮、毛、爪、齒、塵、垢、皮、肉、
骸、骨、筋、脈、心、膽、肝、肺、大腸、小腸、生藏、熟藏、肚、
胃、髀 [or 脾 ]、腎、膿、血、熱、痰、肪、膏、肌、髓、腦、膜、洟、
唾、淚、汗、屎、尿. 如是等類、名為依內朽穢不淨.
While the Sanskrit lists 36 items, this way of parsing the Chinese
yields 40 distinct items. The Tibetan text seems to contain 35 items.
བརྟེན་པ་ཡིན་པར་རིག་པར་བྱའོ། །དེ་ལ་ནང་ལ་བརྟེན་པའི་མི་གཙང་བའི་མི་སྡུག་པ་ཉིད་གང་ཞེ་ན། འདི་
ལྟ་སྟེ། སྒྲ་དང༌། སྤུ་དང༌། སེན་མོ་དང༌། སོ་དང༌། གློག་པ་དང༌། དྲི་མ་དང༌། པགས་པ་དང༌། ཤ་དང༌།
རུས་པ་དང༌། ཆུ་རྒྱུས་དང༌། རྩ་དང༌། མཁལ་མ་དང༌། སྙིང་
དང༑ མཆིན་པ་དང༌། གློ་བ་དང༌། རྒྱུ་མ་དང༌། གཉེ་མ་དང༌། ཕོ་བ་དང༌། ལོང་ཀ་དང༌། མཆེར་བ་དང༌།
ཕྱི་ས་དང༌། མཆི་མ་དང༌། རྡུལ་དང༌། མཆིལ་མ་དང༌། སྣབས་དང༌། ཞག་དང༌། ཆུ་སར་དང༌། རྐང་དང་
ཚིལ་དང་མཁྲིས་པ་དང༌། བད་ཀན་དང༌། རྣག་དང༌།
21
22
Shukla, Śrāvakabhūmi of Acarya Asanga, 203.
Yuqieshi di lun 瑜伽師地論, T no. 1579, 30: 26.428c24–28.
185
ཁྲག་དང༌། གླད་སྤྲི་དང༌། གླད་རྒྱས་དང༌། གཅིན་དང༌། དེ་ལྟ་བུ་དང་མཐུན་པ་དག་ནི་ནང་ལ་བརྟེན་པའི་
མི་གཙང་བའི་མི་སྡུག་པ་ཉིད་ཡིན་པར་རིག་པར་བྱའོ་།དེ་ལ་ཕྱི་རོལ་ལ་བརྟེན་པའི་མི་གཙང་བའི་མི་སྡིག་
པ་ཉིད་གང་ཞེ་ན། འདི་ལྟ་སྟེ། རྣམ་པར་བསྡོས་པ་དང༌། རྣམ་པར་རྣགས་
Additionally, while the Sanskrit, Chinese and Tibetan lists initially
begin with corresponding items in the same order, they soon go out
of easy alignment.
The numbering, S for Sanskrit (e.g., S1 = Sanskrit 1); C for Chinese (C1, etc.); T for Tibetan:
pratyaśubhatādhyātmam
upādāya
依內朽穢不淨,
謂內身中
nang la brten pa
Inner bases of
gang zhe na
impurity
S1. keśā
C1. fa 髮
T1. སྐྲ་ skra
fine hair, head
hair
S2. romāṇi
C2. mao 毛
T2. སྤུ་ spu
body hair
S3. nakhā
C3. zhao 爪
T3. སེན་མོ་ sen mo
nails
S4. dantā
C4. chi 齒
T4. སོ་ so
teeth
C5. chen 塵
(dust particles)
T5. གློག་པ་ glog pa,
(ulcer, sore, see
T22)
specks of dirt
S6. malam
C6. gou 垢
T6. དྲི་མ་ dri ma
stain, taint
S7. tvak
C7. pi 皮
T7. པགས་པ་ pags
pa
skin
S8. māṁsam
C8. rou 肉
T8. ཤ་ stha
flesh
S9. asthi
C9. hai 骸 (and
10. 骨 gu)
T9. རུས་པ་ rus pa
bones
S5. rajas (impurity, dirt,
dust, any small particle of
matter; the dust or pollen
of flowers)
S10. snāyu
C10. jin 筋
T10. ཆུ་རྒྱུས་ chu
rgyus
muscles, tendons,
sinews, ligaments
S11. sirā
C11. mai 脈
T11. རྩ་ rtsa
(channels, vessels)
blood vessels
186
依內朽穢不淨,
謂內身中
nang la brten pa
Inner bases of
gang zhe na
impurity
[Ch. includes
kidneys as C24
shen 腎 ]
T12. མཁལ་
མ་ mkhal ma
(kidneys)
**
C12. xin 心
T13. སྙིང་ snying
heart
--[but cf. C22,
alternate]
T14. མཆིན་པ་ mchin **
pa (liver, midriff);
but cf. T20 below
---
C13. dan 膽 (gall
bladder)
---
**
S15. kloman
C15. fei 肺
T15. གློ་བ་ glo ba
lungs
S16. antrāṇi
C16. dachang
T17. གཉེ་མ་ gnye
ma
large intestine
S17. antraguṇaḥ
C17. xiaochang
T16. རྒྱུ་མ rgyu ma
small intestine
S18. āmāśayam
(digesting nutrients)
C18. shengcang
生藏 “rawstorage”
---
āma is an
technical Indian
medical term;
āmāśayam
digestive action
associated with
the stomach and
upper torso
C19. shucang
---
pakva is a related
technical term;
pakvāśayam is
assoc. with the
large intestine
and lower torso
T19. ལོང་ཀ་ long
ka, intestines,
entrails, guts
bowels, abdomen
pratyaśubhatādhyātmam
upādāya
S12. vṛkkā (kidneys)
S13. hṛdayam
S14. plīhakam (spleen)
S19. pakvāśayam
(digested nutrients)
大腸
小腸
熟藏 ‘processed-
storage’
---
C20. du 肚
bowels/abdomen
187
pratyaśubhatādhyātmam
upādāya
---
---
依內朽穢不淨,
謂內身中
C21. wei 胃
stomach
C22. bi 髀
buttocks/thigh
[more likely: pi
脾 , spleen]
nang la brten pa
Inner bases of
gang zhe na
impurity
T18. ཕོ་བ pho
ba, (stomach,
ruminating
stomach)
stomach
---
[if 脾 , then this
would be ‘spleen’
corresponding to
S#14 plīhakam.
Stomach and
spleen 胃脾
are commonly
paired in Chinese
medicine.]
S20. yakṛt
C14. gan 肝
T20. མཆེར་བ mcher
pa (spleen)
liver
S21. purīṣam
C38. shi 屎
T21. ཕྱི་ས་ phyi sa
excrement
S22. aśru
C36. lei 淚
T22. མཆི་མ་ mchi
ma
tears
S23. svedaḥ
C37. han 汗
Tib has T23.
རྡུལ་ rdul, dirt
particle (rajas)
here, but since it
had 4. གློག་པ་ glog
pa, ulcer, sore,
for rajas above,
it is unclear how
the translator
duplicated
rajas here or
mistook sveda
for rajas. This
should probably
be amended to
རྔུལ་དང་། rngul
‘perspiration’.
sweat
---
C35. tuo 唾
T24. མཆིལ་མ་ mchil saliva
ma
S24. kheṭaḥ
C34. ti 洟
T25. སྣབས་ snabs
nasal mucuous
188
pratyaśubhatādhyātmam
upādāya
依內朽穢不淨,
謂內身中
nang la brten pa
Inner bases of
gang zhe na
impurity
How the next few items align is unclear. These are only suggestions.
S25. śiṁghāṇakam
S26. vasā (marrow;
fat, grease, lard, melted
fat, any fatty or oily
substance)
S27. lasīkā (watery
humour in the body,
lymph, serum; a tendon,
muscle)
S28. majjā (marrow,
urinary, semen
producing)
S29. medaḥ
C29. gao 膏
T26. ཞག་ zhag
(grease, oil
[liquid] fat/
butter; 2) body
oil; 3) blood clot)
greasy fat (in Ch.
medicine, 膏 can
be the fatty tissue
surrounding
organs, or stuff
that lubricates
joint capsules)
C31. sui 髓
T28. རྐང་ rkang
marrow
T27. chu sar
(lymph; several
types of disease
involving fluids
in the joints
causing arthritis
or itching, sores,
e.g. leprosy)
lymph; watery
humour
C23. shen 腎
kidneys/testes
Urogenital
system. neishen
內腎 = kidneys,
waishen 外腎
=testes
[T27a. If ཆུ་ས་ chu
sa is amended
to chu so,
‘bladder, external
and internal
urinary organs’,
then it would
correspond with
the meaning of
majjā as urinary,
semen producing,
and the Chinese
腎 with similar
meaning.
‘Marrow’ was
already expressed
by vasā / 髓 /
rkang (cf. S26;
C31; T28)
??
C28. fang 肪
T29. ཚིལ་ tshil
fat
C30. ji 肌
(muscle) / [alt: fei
肥 (fat)]
189
pratyaśubhatādhyātmam
upādāya
依內朽穢不淨,
謂內身中
nang la brten pa
Inner bases of
gang zhe na
impurity
S30. pittam
C26. re 熱 (heat/
fever)
T30. མཁྲིས་པ་
mkhris pa (bile)
‘heat’ doṣa,
‘bilious humor’
S31. śleṣmā
C27. tan 痰
(phlegm/mucous)
T31. བད་ཀན་ bad
kan (phlegm)
‘phlegmatic’ doṣa
S32. pūyaḥ
C24. nong 膿
T32. རྣག་ rnag
pus
S33. śoṇitam
C25. xue 血
T33. ཁྲག་ khrag
blood
S34. mastakaṁ
C32. nao 腦
T34. གླད་སྤྲི་ glad
spri
brain
S35. mastaka-luṁgam
C33. mo 膜
(membrane)
T35. གླད་རྒྱས་ glad
rgyas
brain membrane
S36. prasrāvaḥ
C39. niao 尿
T36. གཅིན་ gcin
urine
What happens at S18 and S19, āmāśayam and pakvāśayam, is
interesting. Xuanzang’s Chinese translation first offers neologisms
for each, shengcang 生藏 and shucang 熟藏, respectively, and then,
in addition offers two glosses on them, du 肚 bowels, abdomen and
胃 stomach. Āma is food in initial stages of digestion, i.e., ‘raw’ and
being broken down, associated with the stomach but understood
to disperse esp. in the upper body. Pakva, meaning ‘matured food’,
is food further digested, ‘mature’, ‘processed’, and nearing time of
expulsion from the body; it is associated with the large intestine and
lower torso. Āśaya means a vessel or receptacle in the body. In Chinese, the contrast between sheng 生 and re 熟 similarly signals ‘raw’
vs. ‘processed’. So those terms are apt; and cang 藏, ‘storage’ was
used in Chinese medical literature for organ systems, so that too is
apt. The Tibetan does not attempt to generate a technical neologism
for these distinctly Indian medical terms, but instead glosses them
in the same manner as Xuanzang’s gloss, but while Xuanzang’s
glosses reversed the order from the Sanskrit, the Tibetan retains the
Sanskrit order.
If, as some recensions record, C22 is pi 脾 spleen, instead of bi 髀
190
buttocks/thigh, then the Chinese has a corresponding term for S14,
plīhakam (spleen); the Tibetan counterpart at T20, mcher pa, is
closer to the place in the Sanskrit order than the Chinese, but still not
precisely aligned.
For purposes of further comparison, first a list gives the Sanskrit
terms in the order they appear in the extant Sanskrit text, with likely
Chinese and Tibetan equivalents alongside. Then another list, this
time with the Chinese order, and the Sanskrit and Tibetan bracketed alongside. Finally, the Tibetan list, with Sanskrit and Chinese
bracketed alongside. Certain terms only appear on one or two of the
lists; e.g. Chinese and Tibetan have ‘saliva’, but the Sanskrit has no
corresponding term; ‘gall bladder’ only appears in Chinese, etc.
List in order of Sanskrit terms (with attempted Chinese matching):
1.
2.
3.
4.
5.
6.
7.
8.
9.
10.
11.
12.
13.
14.
keśā; fa 髮, fine hair; T1. སྐྲ་ skra, head hair
romāṇi; mao 毛, coarse hair; T2. སྤུ་ spu, body hair
nakhā; zhao 爪, nails; T3. སེན་མོ་ sen mo
dantā; chi 齒, teeth; T4. སོ་ so
rajo; chen 塵, dust; T5. གློག་པ་ glog pa, ulcer, sore [cf. T23. རྡུལ་
rdul, dirt particle (rajas)]
malam; gou 垢, dirt; T6. དྲི་མ་ dri ma, stain, taint
tvak; pi 皮, skin; T7. པགས་པ་ pags pa
māṁsam; rou 肉, flesh; T8. ཤ་ stha
asthi; hai 骸, skeleton (and 10. 骨 gu) bones; T9. རུས་པ་ rus pa, bone
snāyu; jin 筋, muscles/tendons; T10. ཆུ་རྒྱུས་ chu rgyus, sinews,
ligaments
sirā; mai 脈, blood vessels; T11. རྩ་ rtsa, channels/vessels
vṛkkā; (can mean ‘kidneys’ or ‘heart’, so either C24. shen 腎
kidneys/testes, or compound with next term, hṛdayam; the
Tibetan treats it as ‘kidneys’ – T12. མཁལ་མ་ mkhal ma)
hṛdayam; xin 心, heart; T13. སྙིང་ snying
plīhakam (spleen); (no obvious corresponding term here in
Chinese, which has 14. dan 膽, gall bladder and 15. gan 肝,
liver; though, as mentioned above, if C22 is pi 脾, spleen,
instead of bi 髀, buttocks/thigh, then ‘spleen’ does appear in
Chinese, but in a different location on the list. ‘Liver’ is S20
[yakṛt] below, so the numbering and order between the differ-
191
15.
16.
17.
18.
19.
20.
21.
22.
23.
24.
25.
ent versions is beginning to diverge at this point. Tibetan has
T14. མཆིན་པ་ mchin pa, liver)
kloman; [C15] fei 肺, lungs; T15. གློ་བ་ glo ba
antrāṇi; [C16] dachang 大腸, large intestine; T17. གཉེ་མ་ gnye
ma, colon
antraguṇaḥ; [C17] xiaochang 小腸; T16. རྒྱུ་མ rgyu ma, intestines, bowels
(Tibetan reverses the order of large and small intestine)
āmāśayam ‘digesting nutrients’; [C18] shengcang 生藏
‘raw-storage’; T18. ཕོ་བ pho ba, stomach, ruminating stomach
pakvāśayam ‘digested nutrients’; [C20] shucang 熟藏 ‘processed-storage’; T19. ལོང་ཀ་ long ka, intestines, entrails, guts
(It might be that the Tibetan interprets āmāśayam and pakvāśayam as basic and secondary stomachs, like a ruminating
animal! But the Chinese has terms meaning ‘stomach’ and
‘bowels, entrails’ that do not correspond to anything in Sanskrit, so pho ba and long ka might be intended as parallels to
those Chinese terms instead of shaky renderings of āmāśayam
and pakvāśayam. If so, that might suggest that the Tibetan
translators had an eye on the Chinese as well as the Sanskrit,
or that a later redactor consulted the Chinese and modified
accordingly. See below.)
yakṛt; [C14] gan 肝, liver; cf. S14. plīhakam; Tibetan has
T20. མཆེར་བ་ mcher pa here meaning spleen
purīṣam; [C38] shi 屎, excrement; T21. ཕྱི་ས་ phyi sa
aśru; [C36] ti 淚, tears; T22. མཆི་མ་ mchi ma
svedaḥ; [C37] han 汗, sweat (Tibetan has T23. རྡུལ་ rdul, dirt
particle [rajas] here, but since it had T5 གློག་པ་ glog pa, ulcer,
sore, for rajas above, it is unclear how the translator duplicated rajas here or mistook sveda for rajas.)
(The Tibetan has T24. མཆིལ་མ་ mchil ma, saliva, here; the
Chinese has C35. tuo 唾, saliva later; but the Sanskrit lacks any
term for ‘saliva’)
kheṭaḥ; [C34] ti 洟, nasal mucous; T25. སྣབས་ snabs
(How the next few items align is unclear)
śiṁghāṇakam; [C30] gao 膏, greasy fat; T26. ཞག་ zhag, 1)
grease, oil [liquid] fat/ butter; 2) body oil; 3) blood clot
192
26. vasā = marrow, fat, grease, lard, melted fat, any fatty or oily
substance; [C31] sui 髓, marrow; T28. རྐང་ rkang, marrow
27. lasīkā = watery humour in the body, lymph, serum; a tendon,
muscle; [C30] ji 肌, muscle tissue (alt. fei 肥 = fat) ; T27. ཆུ་སེར་
chu s[e]r = lymph fluid
28. majjā = marrow, urinary, semen producing; [C23] shen 腎,
kidneys/testes (?)
29. medaḥ; C28. fang 肪, fat; T29. ཚིལ་ tshil, grease, fat
30. pittam = ‘heat’ doṣa, ‘bilious humous’; [C26] re 熱, heat/
31.
32.
33.
34.
35.
36.
fever; T30. མཁྲིས་པ་ mkhris pa, bile
śleṣmā = ‘phlegmatic’ doṣa; [C27] tan 痰, phlegm/mucous;
T31. བད་ཀན་ bad kan, phlegm
pūyaḥ; [C24] nong 膿, pus; T32. རྣག་ rnag
śoṇitam; [C25] xue 血, blood; T33. ཁྲག་ khrag
mastakaṁ; [C32] nao 腦, brain; T34. གླད་སྤྲི་ glad spri = brain
mastaka-luṁgam = membrane of the brain; [C33] mo 膜,
membrane; T35. གླད་རྒྱས་ glad rgyas, brain membrane
prasrāvaḥ; [C39] niao 尿, urine; T36. གཅིན་ gcin, urine
The Chinese order, with attempted Sanskrit and Tibetan equivalents:
1.
2.
3.
4.
5.
6.
7.
8.
9.
fa 髮, fine hair (keśā; T skra)
mao 毛, coarse hair (romāṇi; T spu)
zhao 爪, nails (nakhā; T sen mo)
chi 齒, teeth (dantā; T so)
chen 塵, dust (rajo = rajas = impurity, dirt, dust, any small
particle of matter; the dust or pollen of flowers; T glog pa =
ulcer, sore)
gou 垢, dirt (malam = [in med.] any bodily excretion or secretion (especially those of the dhātus q.v., described as phlegm
from chyle, bile from the blood, nose mucus and ear wax from
the flesh, perspiration from the fat, nails and hair from the
bones, rheum of the eye from the brain); T dri ma = stain, taint
pi 皮, skin (tvak = tvac = skin; T pags pa)
rou 肉, flesh (māṁsam = flesh, meat; T stha)
hai 骸, skeleton (asthi = a bone; T rus pa)
193
10. gu 骨, bones
(While the Chinese seems to list two separate ‘bone’ items,
since the Sanskrit and Tibetan both only offer one ‘bone’
term, we should probably take haigu 骸骨 as a compound for
‘bones, skeleton’. That would reduce the count by one, but
for this exercise we will follow the standard parsing and retain
the non-compounded numbering. In the chart, I have treated
haigu 骸骨 and renumbered accordingly, so the following
numbers will be one number higher than the chart.)
11. jin 筋, muscles/tendons (snāyu = any sinew or ligament in the
human and animal body, tendon, muscle, nerve, vein; [T10]
chu rgyus)
12. mai 脈, blood vessels (sirā =any tubular vessel of the body, a
nerve, vein, artery, tendon; [T11] rtsa, channels/vessels)
13. xin 心, heart ([vṛkkā?]-hṛdayam = the heart; [T12] snying)
14. dan 膽, gall bladder
(No term in Sanskrit or Tibetan corresponds. Sanskrit has
plīhakam = spleen; the Chinese alternate for C22 bi 髀 (thigh/
buttocks) is pi 脾, spleen, and [T20] is mcher pa, spleen.)
15. gan 肝, liver ([S20] yakṛt = the liver; [T13] mchin pa, liver
16. fei 肺, lungs ([S15] kloman = lungs; [T15] glo ba)
17. dachang 大腸, large intestine (antrāṇi = intestine, entrails;
[T17] gnye ma)
18. xiaochang 小腸, small intestine (antraguṇaḥ; [T16] rgyu ma)
19. shengcang 生藏, raw-storage (āmāśayam = latent nutrients
being digested [no Tib])
20. shucang 熟藏, processed-storage (pakvāśayam = digested nutrients [no Tib])
21. du 肚, bowels/abdomen [no Skt, but [T19] long ka, intestines, entrails, guts]
22. wei 胃, stomach [no Skt, but [T18] pho ba, stomach]
23. bi 髀, buttocks/thigh [if pi 脾, then this would be spleen]
24. shen 腎, kidneys/testes (majjā ? = urinary, marrow producing
semen [S28]; if intended as a translation for 12. vṛkkā [S12];
[T12 mkhal ma], then the Chinese appears in an odd location.)
25. nong 膿, pus (pūyaḥ = pus [S32]; [T31] rnag)
26. xue 血, blood (śoṇitam = blood [S33]; [T33] khrag)
194
27. re 熱, heat/fever (pittam = the bilious humour (one of the
three humours [cf. kapha and vāyu] or that secreted between
the stomach and bowels and flowing through the liver and
permeating spleen, heart, eyes, and skin; its chief quality is
heat [S30]; [T30] mkhris pa, bile)
28. tan 痰, phlegm/mucous (śleṣmā = phlegm, mucus, rheum, the
phlegmatic humour (one of the three humours of the body
[S31]; T[31] bad kan, phlegm)
(The list includes two of the three doṣas: pitta and śleṣmā, but
omits kapha. Both the Chinese and Tibetan offer interpretive
translations: pitta = Chinese ‘heat’, Tibetan ‘bile’; śleṣmā =
Chinese and Tibetan ‘phlegm’)
29. fang 肪, fat (medaḥ = fat [S29]; [T29] tshil)
30. gao 膏, greasy fat (śiṁghāṇakam [S25]; [T26] zhag)
31. ji 肌 [alt. fei 肥 = fat] muscle tissue (lasīkā = watery humour in
the body, lymph, serum; a tendon, muscle [S27]; [T27] chu sar)
32. sui 髓, marrow (vasā = marrow, fat, grease, lard, suet, melted
fat, any fatty or oily substance; brain [S26])
(Tibetan has three terms indicating ‘fat, grease’, etc. and it
is unclear which of the Sanskrit or Chinese terms indicating
something similar each is meant to indicate: T26 zhag, T27a
chu sa, T29 tshil)
33. nao 腦, brain (mastakaṁ = the head, skull [S34]; [T34] glad
spri)
34. mo 膜, membrane (mastaka-luṁgam = the membrane of the
brain [#35]; [T35] glad rgyas)
35. ti 洟, nasal mucous (kheṭaḥ = snot [S24]; [T25] snabs)
36. tuo 唾, saliva (Not in Sanskrit, but [T24] mchil ma = saliva)
37. lei 淚, tears (aśru = tears [S22]; [T22] mchi ma)
38. han 汗, sweat (svedaḥ = sweat [S23]; omitted in T)
39. shi 屎, excrement (purīṣam = feces, excrement [S21]; [T21]
phyi sa)
40. niao 尿, urine (prasrāvaḥ = urine [S36]; [T36] gcin)
如是等類, 名為依內朽穢不淨. These are what is called the basis of
internal decay and impurity.
195
Two of the three doṣas appear, but kapha is omitted.
S30. pittam = ‘heat’ doṣa, ‘bilious humous’, [C27] re 熱, heat/fever
S31. śleṣmā = ‘phlegmatic’ doṣa, [C28] tan 痰, phlegm/mucous
Tibetan order:
(omits 1. keśā, fa 髮, fine hair)
1. སྤུ་ spu, body hair (romāṇi)
2. སེན་མོ་ sen mo, nails (nakhā)
3. སོ་ so, tooth/teeth (dantā)
4. གློག་པ་ glog pa, ulcer, sore (for rajas)
5. དྲི་མ་ dri ma, stain, taint (for malam)
6. པགས་པ་ pags pa, skin (tvac/tvak)
7. ཤ་ stha, flesh (māṁsam)
8. རུས་པ་ rus pa, bone (asthi)
9. ཆུ་རྒྱུས་ chu rgyus, sinews, ligaments (snāyu)
10. རྩ་ rtsa, channels/vessels (sirā = blood vessels)
11. མཁལ་མ་ mkhal ma, kidneys (vṛkkā)
12. སྙིང་ snying, heart (hṛdayam)
([14] dan 膽, gall bladder: Neither the Sanskrit nor Tibetan
mention gall bladder; gall bladder was an important organ
system in Chinese medicine since ancient times, but unknown in Indian medicine until late medieval times.)
13. མཆིན་པ་ mchin pa, liver (yakṛt)
14. གློ་བ་ glo ba, lungs (kloman)
15. རྒྱུ་མ rgyu ma, intestines, bowels (18. xiaochang 小腸, small
intestine [antraguṇaḥ])
16. གཉེ་མ་ gnye ma, colon (17. dachang 大腸, large intestine [antrāṇi
= intestine, entrails] reversing order of small and large intestines)
17. ཕོ་བ pho ba, stomach, ruminating stomach (replaces: [19]
shengcang 生藏, life-store [āmāśayam = latent nutrients being
digested]; [20] shucang 熟藏, maturation-store [pakvāśayam =
digested nutrients])
18. ལོང་ཀ་ long ka, intestines, entrails, guts (pakvāśayam?)
(Again, it might be that the Tibetan interprets āmāśayam and
pakvāśayam as basic and secondary stomachs, like a ruminating animal. However, the Chinese lists, apart from āmāśayam
196
19.
20.
21.
22.
23.
24.
25.
26.
27.
28.
29.
30.
31.
32.
33.
and pakvāśayam, have two items not found in Sanskrit, but
that correspond to pho ba and long ka: C20. du 肚, bowels/
abdomen and C21. wei 胃, stomach. This raises the possibility
that the Sanskrit list as now extant is incomplete and that it
originally contained terms for stomach and bowels, preserved
in both Chinese and Tibetan, albeit is slightly different locations on the list. If so, then the Tibetan translation failed to
include any equivalents for āmāśayam and pakvāśayam,
which, given their somewhat unique technical application to
Indian medical theory, may simply be a matter of prudently
choosing to ignore terms too difficult to successfully translate.
Comparable lists of ‘unpleasant’ body parts are already given
in several Nikāyas (e.g. 29. Udāyīsutta of Aṅguttara Nikāya
6.3.29, Bhāradvājasutta of Saṃyutta Nikāya 35.13.127,
etc.), which tend to include udariyaṃ, ‘undigested food’ or
‘stomach contents’, but do not list āma and pakva as distinct
items, so udariya may have expanded into differing Sanskrit
expanded versions.)
མཆེར་བ་ mcher pa, spleen
ཕྱི་ས་ phyi sa, excrement (purīṣam) (S21. in Sanskrit, C38. in
Chinese)
མཆི་མ་ mchi ma, tears (aśru)
རྡུལ་ rdul, dirt particle (rajas) (This is where S23. svedaḥ, [C37]
汗 sweat appear)
མཆིལ་མ་ mchil ma, saliva
སྣབས་ snabs, nasal mucous (kheṭaḥ)
ཞག་ zhag, grease, oil (liquid) fat/ butter; 2) body oil; 3) blood
clot (śiṁghāṇakam? lasīkā? vasā?)
ཆུ་སེར་ chu ser (S27. lasīkā = watery humour in the body, lymph,
serum)
རྐང་ rkang, marrow (S26. *vasā? S28. majjā?)
ཚིལ་ tshil, grease, fat (S29. medaḥ; or śiṁghāṇakam? lasīkā? vasā?)
མཁྲིས་པ་ mkhris pa, bile (S30. pitta)
བད་ཀན་ bad kan, phlegm (S31. śleṣmā)
རྣག་ rnag, pus (S32. pūyaḥ)
ཁྲག་ khrag, blood (S33. śoṇitam)
གླད་སྤྲི་ glad spri, brain (S34. mastakaṁ)
197
34. གླད་རྒྱས་ glad rgyas, brain membrane (S35. mastaka-luṁgam)
35. གཅིན་ gcin, urine (S36. prasrāvaḥ)
Why the discrepancies? Some can be attributed to accommodating differences in medical and anatomical understandings between
the three cultures during the times of translation. While the gall
bladder played an important role in Chinese medicine as one of the
internal organ systems, Indian sources fail to mention it at all until
much later than the time when the Yogācārabhūmi was written. Why
would Xuanzang add the gall bladder when the Sanskrit never mentioned it? Perhaps to meet expectations of his Chinese audience, with
their knowledge of medicine and anatomy, so that they wouldn’t
raise doubts about limitations in Indian medical knowledge, and
by extension, of other basic components of reality. That of course
doesn’t explain the other discrepancies, and the rearrangement of
the order. Similar lists of bodily impurities appear in other Buddhist
texts, and they tend to vary from each other in some details, so it is
not inconceivable that Xuanzang’s Sanskrit text was different from
the version that came down to us. But there are too many to simply
attribute it to that. Again, to find a Sanskrit text mentioning a gall
bladder would have been anomalous at that time.
Our task here is not to solve the incommensurables between
these three lists, but merely to point out that Xuanzang’s version is
not a one-for-one strictly literal version of what he read in Sanskrit.
In addition to listing items in a different order, he included an item
important in Chinese medicine at that time but unknown in Indian
medicine: the gall bladder.
Turning to the next list, which is shorter, discrepancies again
appear, and, once again, the Chinese list is longer than the Sanskrit.
These are types of vyāyāma, ‘exertions’, or, as the context makes
clear, forms of strenuous exercise. Xuanzang renders vyāyāma with
the unusual term juewu 角武 (combative martial exercises). One presumes he intends jue (second tone) rather than jiao (‘horn’) meaning
‘dispute, contend, fight over’, rather than ‘horn’, so that the compound jue wu would mean something like physical competitions that
might have applications in combat. We will return to the ‘martial’
implication once the list has been presented.
198
The passage reads:
隨力隨能, 食噉肥膩. 增房補益, 色香味具, 精妙飲食. 過今夜分, 至
於明日, 於角武事, 當有力能. 所謂按摩, 拍毱托石, 跳躑蹴蹋, 攘臂
扼腕, 揮戈擊劍, 伏弩控弦, 投輪擲索. 依如是等, 諸角武事. 當得勇
健, 膚體充實. 長夜無病, 久時少壯. 不速衰老, 壽命長遠. 能多噉
食, 數數食已. 能正消化, 除諸疾患. 如是為於無病憍逸, 少壯憍逸.
長壽憍逸, 而食所食. 既角武已, 復作是思.23
But here we are only concerned with the actual list of exercises. 於
角武事, 當有力能. 所謂按摩、拍毱、托石、跳躑、蹴蹋、攘臂、扼腕、揮戈、
擊劍、伏弩、控弦、投輪、擲索. 依如是等, 諸角武事. The corresponding
Sanskrit passage reads: pratibalā vyāyāmakaraṇaḥ, yadutātatikriyayā
vā, nirghātena, vyāyāmaśilayā vā, ulloṭhanena vā, pṛthivīkhātena vā,
bāhuvyāyāmena vā, pādāvaṣṭambhanena vā, plavanena vā laṅghanena vā cakravyāyāmena vā / taṃ ca punar vyāyāmaṃ…
The following chart aligns them to the extent they can be aligned.
Sanskrit
Sanskrit
meaning
Wayman’s
rendering
Chinese meaning
Chinese
vyāyāmakaraṇaḥ
Combative
strenuous
exercise
the skill
of athletic
exercise24
Combative
martial activities /
exercises, by which
one becomes
strong
Juewu shi A
dang you
lineng
2
yaduta ātatīkriyayā Drawing a
vā
bow Cf. L
and M
drawing
[the bow]
3
nirghātena
1
23
24
‘destroying’
rubbing
(sic) (nirghāta [the body]
‘destroy’ >
nirgharṣaṇa
? ‘rubbing,
friction’25
角武事當
有力能
B
massage
anmo
按摩
C
Yuqieshi di lun, T no. 1579, 30: 23.409c8–17.
This column follows Wayman, Analysis of the Śrāvakabhūmi Manuscript, 156.
199
Sanskrit
Sanskrit
meaning
4
vyāyāmaśilayā vā
vyāyāmaśilayā lifting the
= exercising
contest
with a stone,
stone
Cf. E
5
ulloṭhanena vā
Rolling on the
ground?
Kicking a ball to
prevent it from
dropping
paiju
拍毱
D
6
pṛthivīkhātena vā
Excavating,
digging the Carrying stones
digging up the soil
(weight lifting)
ground
tuoshi
托石
E
7
bāhuvyāyāmena vā
Exercising
contesting
arms (or
with arm
upper body) =
H 攘臂
Leaping, jumping
tiaozhi
跳躑
F
8
pādāvaṣṭambhanena Kicking (lit.
running
vā
resolute foot)
(= I stomping)
Kicking and
stomping
cuta
蹴蹋
G
9
plavanena vā
rangbi
10 laṅghanena vā
25
Wayman’s
rendering
Chinese meaning
Chinese
[see E]
swimming
swimming
Raising hands
(to fight); roll up
sleeves to fight, to
force others, cf.
WB-DDJ 38
攘臂
H
Leaping,
jumping over
= F 蹴蹋
jumping
Stomping =
ewan
pādāvaṣṭambhana? 扼腕
I
Brandishing a
spear
huige
揮戈
J
Fencing (lit.
striking with a
sword
jijian
擊劍
K
Loading a
crossbow
funu
L
Wayman suggests nirghātena > nirghaṭṭena.
伏弩
200
Sanskrit
Sanskrit
meaning
Wayman’s
rendering
11 cakravyāyāmena vā Wheel exercise or
contesting
there
Chinese meaning
Chinese
Drawing a bow
kongxian M
控弦
Throwing a wheel, toulun
discus?
投輪
N
zhisuo
O
Casting a rope
擲索
Alex Wayman’s translation of the relevant portion:26
Not for the purpose of intoxication, not for the purpose of smartening, not for the purpose of embellishment means—a case in point—
those with enjoyment of passions, who eat food thinking: ‘Today we
are eating food that is of large quantity, has oily power as satisfying
as possible, is nourishing, nutritious, has perfect color, perfect odor,
perfect taste, is heated. When night is past, we shall be capable, be
powerful, have the skill of athletic exercise, namely, for drawing [the
bow], rubbing [the body], lifting the contest stone, digging the soil,
contesting with arm, running, swimming, jumping, or contesting
there; and, furthermore, having taken recourse to that athletic
exercise, we shall be strong, have athletic bodies, be free from illness
for serious purposes (dīkṣaṃ); and for a long time that strength will
cleave to us, and not speedily will disfigurements overcome the body
of old age; and we shall live for a very long time, and we shall be able
to eat much; and there will be proper transformation of what is
eaten, and there will be effected a reduction of faults.’ Thus one eats
for the purpose of intoxication with freedom from illness, intoxication with youth, intoxication with life.
Wayman seems to understand pādāvaṣṭambhanena as ‘running’
26
Wayman, Analysis of the Śrāvakabhūmi Manuscript, 156.
201
and, I think perhaps correctly, plavana as ‘swimming’ (one of its
meanings, others being jumping, stooping over, etc.), though Xuanzang doesn’t mention swimming (he has two different compounds
which each suggest jumping, so he seems to have read plavana and
laṅghana as two types of jumping). Wayman seems to ignore cakravyāyāmena (wheel exercise?) or simply takes it as ‘contesting’.
Now, for contrast, we offer a chart that follows the Chinese order
of items. That, as with the example of the body parts, the lists are
so misaligned that cross-referencing the Sanskrit and Chinese is not
simple, is the main point.
The chart follows Xuanzang’s order:
於角武事
當有力能
所謂
1 按摩
2 拍毱
3 托石
4 跳躑
One will become strong pratibalā
through combative27
vyāyāmakaraṇaḥ,
martial activities
anmo
massage
Nirghātena (2)
(nirgharṣaṇa?)28
One is strengthened by
strenuous exercise
‘massage’ (rubbing the
body)
paiju
Kicking a ball to
prevent it from
dropping
tuoshi
Carrying a stone
(weight lifting)
tiaozhi
leaping, jumping
vyāyāmaśilayā vā, (3)
ulloṭhanena vā, (4)
pṛthivīkhātena vā (5)
Rolling around,
digging
plavanena (8) vā
laṅghanena (9) vā
Swimming, jumping
As explained above, 角 when pronounced jue rather than the more common
jiao, means ‘dispute, contend, fight over’; sv. 角 in Kroll, Student’s Dictionary.
28
Nirghātena (‘destruction’) is clearly an error, probably for something like
nirgharṣaṇa, since the Chinese and Tibetan suggest the term means massage
or body rub. Wayman suggests that nirghātena be changed to nirghaṭṭena, I’m
guessing to derive from āghaṭṭana, friction, rubbing. lus mnye ba means to rub
the skin, i.e., massage and anmo 按摩 clearly means ‘massage’.
27
202
於角武事
當有力能
5 蹴蹋
6 攘臂
7 扼腕
8 揮戈
9 擊劍
10 伏弩
11 控弦
12 投輪
13 擲索
One will become strong pratibalā
through combative27
vyāyāmakaraṇaḥ,
martial activities
One is strengthened by
strenuous exercise
cuta
kicking and stomping
pādāvaṣṭambhanena
vā (7)
Running?
rangbi
Bare arms to fight
bāhuvyāyāmena vā (6)
Exercising arms (or
upper body)
ewan
grabbing the wrists
Grappling?
huige
Brandishing spears
jijian
Striking with a sword
(fencing)
funu
Pulling a crossbow
yadutātatikriyayā
vā, (1)
kongxian
Drawing a long bow
ditto
toulun
Throwing a wheel
cakravyāyāmena vā
(10)
Wheel exercise
zhisuo
Casting rope
Xuanzang lists twelve things, the Sanskrit only has ten. The order
in the Sanskrit is in parentheses after the Sanskrit term. But with this
list, the additional Chinese items are easier to explain. The extra two
can be attributed to Xuanzang splitting two items into two separate
items: ‘drawing a bow’ he splits into drawing a crossbow and drawing
a longbow, and the upper-body or arm work he breaks into what
may be boxing and wrestling. There are items with no parallels — the
Sanskrit terms for rolling around and digging have no Chinese counterpart, and the Chinese paiju 拍毱 has no obvious Sanskrit counterpart. Plavana has several meanings, including to jump and to swim.
203
Xuanzang seems to have perhaps taken the former meaning, while it
is possibly the latter was implied by Asaṅga.
For good measure, here is the Tibetan counterpart:
༄༅། །འདོད་པ་ལ་ལོངས་སྤྱོད་པར་བྱེད་ལ། དེ་དག་ཉིད་ལྟར་བདག་ཅག་དེང་འདི་ལྟ་སྟེ། ཟས་སྣུམ་
པ་འདོད་པའི་འདོད་ཆགས་སྐྱེད་པ། འཕེལ་བར་བྱེད་པ། བསོད་པ། ཁ་དོག་ཕུན་སུམ་ཚོགས་པ། དྲི་
ཕུན་སུམ་ཚོགས་པ། རོ་ཕུན་སུམ་ཚོགས་པ་མང་པོ་དག་ གཡོས་དྲོད་ཀྱིས་ཅི་ནུས་སུ་ཟོས་ཏེ། ཚིམ་
པར་བྱས་ན་སང་ནམ་ལངས་པར་གྱུར་པ་ན་མཐུ་དང་ལྡན་པར་འགྱུར་ཞིང་འདི་ལྟ་སྟེ། གཞུ་དགང་བའམ།
ལུས་མཉེ་བའམ། རྡོ་གདེག་པའམ། དྲིལ་བའམ། ས་བརྐོ་བའམ། སྟོབས་བརྒལ་བའམ། རྐང་འཁྱོག་
བྱ་བའམ། རྒྱུག་པའམ། རྒྱལ་ བའམ༑ མཆོང་བའམ། འཁར་ལོ་འཕང་བའི་བརྩལ་བ་བྱེད་ནུས་པར་
འགྱུར་ཏེ། བརྩལ་བ་དེ་ལ་བརྟེན་ནས་མཐུ་དང་ལྡན་པ་དང༌། ལུས་ཤིན་ཏུ་བཅགས་པ་དང༌། ཡུན་རིང་
དུ་ནད་རྣམས་མེད་པར་འགྱུར་བ་དང༌།
བདག་ཅག་གི་ལང་ཚོ་ཡུན་རིང་དུ་རྗེས་སུ་འཇུག་པར་འགྱུར་བ་
དང༑ མི་སྡུག་པར་བྱེད་པའི་རྒ་བས་ལུས་མྱུར་དུ་ཟིལ་གིས་མི་ནོན་པར་འགྱུར་བ་དང༌། ཤིན་ཏུ་ཡུན་རིང་
དུ་འཚོ་བར་འགྱུར་བ་དང༌། མང་དུ་ཟོས་ན་ཡང་ཤིན་ཏུ་སྟོབས་དང་ལྡན་པར་འགྱུར་བ་དང༌། ཟོས་སོ་ཅོག་
ཀྱང་ལེགས་པར་ཡོངས་སུ་འཇུ་བར་འགྱུར་བ་དང༌། སྐྱོན་རྣམས་ཀྱང་ སེལ་བར་བྱེད་པར་འགྱུར་རོ་ཞེས་
དེའི་དོན་དུ་ཟས་ཟ་བར་བྱེད་པ་དེ་ནི། ནད་མེད་པས་རྒྱགས་པ་དང༌། ལང་ཚོས་རྒྱགས་པ་དང༌། གསོན་
པས་རྒྱགས་པའི་དོན་དུ་ཟ་བ་ཡིན་ནོ། །དེ་དག་ཡང་འདི་སྙམ་དུ་སེམས་ཏེ། བདག་ཅག་གིས་བརྩལ་
བར་བྱས་ཟིན་པས། ཁྲུས་ཀྱི་ཆོ་ག་དག་ བྱས་ཏེ། འདི་ལྟ་སྟེ། ཆུ་གཙང་མས་ལག་པ་དག་བཀྲུ་བར་བྱ་
ཞིང༌། ལུས་དག་བཀྲུས་ནས་སྐྲ་དག་ཀྱང་བཅོས་པར་བྱ། བྱུག་པ་སྣ་ཚོགས་ཀྱིས་ལུས་དག་བྱུགས་ལ།
གོས་སྣ་ཚོགས་དང༌། མེ་ཏོག་ཕྲེང་བ་སྣ་ཚོགས་དང༌། རྒྱན་སྣ་ཚོགས་དག་གིས་ལུས་བརྒྱན་པར་ བྱའོ་
སྙམ་དུ་སེམས་ཏེ། དེ་ལ་ཁྲུས་དང༌། བཅོས་པ་དང༌། བྱུགས་པ་གང་ཡིན་པ་དེ་ནི་དེ་དག་གི་སྒེག་པ་ཡིན་
ནོ། །དེ་ལྟར་སྒེག་པར་གྱུར་པ་དག་གོས་དང༌། མེ་ཏོག་ཕྲེང་དང༌། རྒྱན་དག་ཐོགས་པ་གང་ཡིན་པ་དེ་ནི་
བརྒྱན་པ་ཞེས་བྱ་སྟེ། དེ་ལྟར་ན་སྒེག་པའི་དོན་དང་བརྒྱན་པའི་དོན་29
29
D 4036: vol. 128, folio 34a.
204
Extracting the terms we are concerned with:
gzhu dgang ba = ātatī-kriyā
lus mnye ba = nirgata. aṅga-prapīḍana
rdo gdeg pa = vyāyāma-śilā
dril ba = to roll/ twirl/ spin; to roll/ coil up; to blend
sa brko ba = excavation
stobs = strong, power
brgal ba = overcoming
rkang ’khyog bya ba (lit. ‘activities lifting the foot’) =
pādāvaṣṭambhana
rgyug pa = to run, a stick
rgyal ba = conqueror, jina; winning
mchong ba = laṅghana = jump, rush, dash, hop, leap, plunge,
pounce, skip, vault, leap up, bathe in water
I am not sure what to make of rkang ’khyog bya ba which seems
more to suggest carrying bundles of earth than ‘running’. Could brgal
pa means something like leaping over, jumping over? rgyal ba seems to
simply mean ‘winning’ (in sports, betting, war, etc.). rgyug pa would
mean to run. rdo gdeg pa clearly means to lift stones. If brgal pa and
mchong ba both mean types of jumping, then the Tibetan, like Xuanzang, sees two types. The Tibetan obviously omits the wheel toss.
The Tibetan list of exercises seems to contain ten items, like the
Sanskrit, and tracks closer to the Sanskrit than Xuanzang’s version. It
is obvious that Wayman was relying on the Tibetan for help with his
Sanskrit.
What is curious is that Xuanzang’s version is more ‘martial’ than
the other two, and explicitly declares itself such by using the term wu
武, ‘martial, war’. While the Sanskrit is not devoid of martial mentions (it does mention drawing a bow), Xuanzang frames the exercise
as ‘martial’ (武) and competitive fighting (jue 角; 角武 = martial arts),
and includes martial items with no Sanskrit counterparts: raise hands
to fight (rangbi 攘臂), grabbing the wrists (grappling?) (ewan 扼
腕), brandishing a spear (huige 揮戈), strike with a sword (jijian 擊
劍), pull a crossbow (funu 伏弩), drawing a bow (kongxuan 控弦),
throwing a wheel (?) (toulun 投輪), casting a rope (zhisuo 擲索).30 Per-
205
30
One possible explanation for Xuanzang tilting the passage toward a martial
account might be the following passage from the Mahāparinirvāṇa sūtra which
prohibits a variety of activities, including visiting military camps or engaging in
martial-like competitive or strengthening sports (I wish to thank Mark Blum for
generously sharing his soon to be published translation of this passage, along
with his annotations, though I present my own translation here):
…[Do not use fancy pillows]. Finally, don’t watch elephant competitions,
horse competitions (i.e., races), cart competition (races), weapons competitions, men competing, women competing, or bull fights, sheep fights,
competitions between water buffalos, or cock fights, or pheasant fights,
nor should you go to watch military encampments. You shouldn’t therefore listen to the musical tones (jiyue zhisheng 伎樂之聲) of clattering shells
(chuibei 吹貝), drum and horn (gujue 鼓角 [used in military activities, like
drum and bugle), qin 琴 and se 瑟 (stringed musical instruments), zheng 箏
(zither with from 5 to 16 strings; Jp: koto), flute 笛, konghou 箜篌 (Chinese
harp), or sing songs of praise (gejiao 歌叫), except when making offerings to
the Buddha.
Competitions (dou 鬪) between game-masters (shizi 師子) of games (xi
戲) like playing dice (chupu 摴蒱 = Indian dice game chaupar), entrapment
chess-like games (weqi 圍碁), and Indian chess (boluosai 波羅塞 = prāsaka,
prasena?), shooting chess (danqi 彈碁, Mark Blum explains this is ‘Described as a game of two opposing teams each originally having 6 pieces,
white and black, that are laid out facing each other and pebbles are tossed
or rolled to knock over the opponent’s pieces. The number expands to 16
in the Wei period, and 24 in the Tang. The edge of a hand-towel is used
in some manner to brush away the chips’; involving polished stones dan
彈 and a board qi 碁), Six stacks (liubo 六愽, a board game played with 15
white and black pieces pitted against each other. Six sticks were thrown to
determine each move, performing the function of dice. It came to Japan
from China prior to the Nara period.), paiju 拍毱 kicking a ball, zhishi 擲
石 hurling stones, touhu 投壺 (ancient banquet game of throwing arrows
into a pot, the winner determined by the number of arrows thrown in, and
the loser required to drink as punishment), qiandao 牽道 sport of pulling/
dragging (probably 道 here is like the dō in ken-dō, ju-dō, karate-dō, etc.),
badao xingcheng 八道行成 practicing to perfect the eight sports. One
shouldn’t watch or engage in any such games.
206
haps the Sanskrit upper body and arm exercises could be paired with
rangbi 攘臂, but the Sanskrit only mentions drawing a bow without
distinguishing between cross-bows and long bows, and there is nothing in the Sanskrit about spear/halbert or sword forms or sparring.
Exactly what the ‘wheel exercise’ or ‘wheel toss’ was—whether something like a discus, or doing something with a larger wheel or something like a medicine ball, etc.—is unclear. I am not familiar with any
specific exercise in either Chinese or Indian traditions by that name.
It may be that Xuanzang is drawing on contemporary martial exercises. If so, then the vocabulary for such things likely changed over time,
so later texts would designate these or similar exercises with different
labels. Clearly the fuller passage begins with a discussion of food,
moves to exercise, then talks about getting strong and fit, being free
from disease and living a long time as a result.
As for translation stylistics, Xuanzang seems willing to adjust his
translation for his intended audience, and even editorialize and embellish for affect. His renderings are usually ‘accurate’ in the sense of
conveying the basic meaning, but may stray from strict adherence to
the details of original Sanskrit. All translators into Chinese did so to
varying degrees as well.
Finally, one shouldn’t examine physical features such as hand and foot
or face (i.e. fortune telling by examining physical features; physiognomy and
phrenology). … [then names and prohibits types of divination]. And no
gazing at the stars in admiration (i.e. astrology 亦不仰觀虛空星宿), except
when one wishes to go to sleep. [No tales of royalty, etc.]. Finally, no
self-aggrandizement through flattery or nefarious plots against others’ lives…
其床兩頭, 不置二枕. 亦不受畜妙好丹枕, 安黃木枕. 終不觀看象, 鬪馬、鬪
車、鬪兵、鬪男、鬪女、鬪牛、鬪羊、鬪水、牛、鷄、雉、鸚鵡等鬪; 亦不故往觀看
軍陣, 不應故聽吹貝、鼓角、琴瑟、箏笛、箜篌、歌叫、伎樂之聲, 除供養佛. 摴
蒱圍碁波羅塞戲、師子象鬪、彈碁六愽、拍毱擲石、投壺牽道、八道行成. 一切
戲笑, 悉不觀作. 終不瞻相手脚面目, 不以抓鏡、芝草、楊枝、鉢盂、髑髏而作
卜筮. 亦不仰觀, 虛空星宿, 除欲解睡. 不作王家往返使命, 以此語彼, 以彼
語此. 終不諛諂, 邪命自活. (Da boniepan jing 大般涅槃經, T no. 374, 12:
11.433a7–21)
207
Important Terms
The prior examples, with the possible exception of ‘empty of svabhāva’, are more a reflection of cultural differences than issues with
implications for core Buddhist ideas. The new vocabulary introduced
by Xuanzang, on the other hand, frequently signaled philosophical
nuances that he sought to express more clearly with better semantic
equivalents. One way that the differences and overlaps between
Paramārtha and Xuanzang’s equivalents can be studied is to work
through Hirakawa’s Abhidharmakośa-bhāṣya indexes. There are hundreds of terms to examine, and it becomes clear that for both translators, one-to-one equivalences were not their desideratum, since
both rendered Sanskrit terms with varying Chinese terms, as well as
using the same Chinese term for different Sanskrit words. What does
become evident, however, when one compares Xuanzang’s rendering
of Kośa passages against Paramārtha’s previous translations, Xuanzang sometimes echoes Paramārtha even when he seems to know
a different rendering is in order, reflecting the lasting influence of
Paramārtha’s version on his own understanding of the Kośa. Scholars
have noted that at times Paramārtha’s translation of the Kośa adheres
more closely to the Sanskrit than Xuanzang’s. As his Biography
points out when discussing his early studies of the Kośa in China:
‘He studied with such profundity that he could comprehend subtle
meanings and reveal what was hidden in the texts when others failed
to reach it. On more than one abstruse point he had his own particular views’ (see fn. 17 above). But nothing as complex or in depth as
analyzing the gamut of terminology in Hirakawa’s index will be attempted here. Instead a few well known examples will have to suffice.
Earlier translators, such as Paramārtha, did not clearly distinguish
between various cognates and derivatives of the root √kḷp which
forms important terms such as vikalpa, kalpanā, kalpita, parikalpa/
parikalpita, etc., using fenbie 分別 for all of them on many occasions,
despite the different connotations and implications of each. Since
fenbie graphically evokes ‘knives’, it is often taken to mean ‘discrimination’, ‘cutting apart’ something whole. So many passages that
criticize fenbie come to be understood and translated as advocating
‘non-discrimination’, a kind of thoughtless or less-than-discerning
208
fuzziness to replace recognizing distinctions. But vikalpa often
doesn’t mean discrimination in Buddhist and especially Yogācāra
contexts, but something more akin to misguided imagining, superimposing a mistaken presupposition. Kalpanā means conceptualizing.
Parikalpa/parikalpita takes the prefix pari- which means to encircle,
surround, to emphasize that this kind of vikalpa becomes immersive,
ubiquitous. So when Paramārtha translates parikalpita-svabhāva or
parikalpita lakṣaṇa as 分別(自)性, it gives the impression that the
problem is making distinctions, which is not the case. The problem
is imposing erroneous misconceptions, false imagining, not the
ability to distinguish a door from a wall, or healthy from unhealthy
food, or wholesome from unwholesome behaviors. To bring this out,
Xuanzang replaced that translation with bianji suozhi xing 遍計所執
性 ‘pervasive presuppositions that are held’ or ‘immersed in speculative (erroneous) opinions to which one is attached’. For parikalpita
Bodhiruci had used 分別虛妄(體相), which at least had the merit of
signaling that something ‘erroneous’ (xuwang 虛妄) was involved,
but again implying to a Chinese reader that the problem was discrimination rather than false imagination.
Comparing Xuanzang’s rendering of the trisvabhāva terms brings
out some of their differences as well as their affinities:
parikalpita-svabhāva = XZ 遍計所執性 = P. 分別性
paratantra-svabhāva = XZ 依他起性 = P. 依他性
pariniṣpanna-svabhāva = XZ 圓成實性 = P. 實實性
Their rendering of paratantra is close. Paramārtha’s incidentally is
more literally accurate, since para-tantra literally means ‘dependent
on other’, which 依他 captures. Xuanzang adds 起 (依他起) to emphasize that the dependence is causal, ‘produced in dependence on
an other’. This is in line with Yogācāra’s associating paratantra with
pratītya-samutpāda. Paramārtha’s 實實性 for parinispanna-svabhāva
certainly emphasizes its superlative nature, but loses the sense of
achievement or the fulfilling of a process which the grammatical
suffix -anna signals in Sanskrit, whereas the cheng 成 in Xuanzang’s
version captures that, as well as the sense of bringing something to
perfection, becoming consummate.
209
A simpler example is replacing yin 陰, which earlier translators
used for skandha, with yun 蘊. Even when not viewed as the
counterpart to yang 陽, yin’s meanings (dark, hidden, etc.) do not
correspond very well with skandha, which means an aggregate
or heap, like a pile of straw neatly stacked. Yun on the other hand
means to gather, collect, etc., which is much closer. Yun’s meanings
include: collect; bring together; raise; contain; hide; deep; hidden;
mysterious; abstruse; (Traditional Chinese Medicine) unhealthy fermentation inside the body. Altogether a very evocative rendering for
skandha, containing both the literal sense and the more Buddhistic
connotations.31
Earlier translators had used yin 陰, ru 入, jie 界 for the basic
categories skandha, āyatana, dhātu, i.e., the five aggregates, twelve
sense-spheres, and eighteen basic factors of experience (six sense
faculties, six corresponding sense-spheres, and six-corresponding consciousnesses). Xuanzang’s equivalents for the three categories are yun
蘊, chu 處, jie 界. The earlier ru 入 (lit. ‘enter’) for āyatana probably
was meant to imply that the senses are the means by which information ‘enters’ one’s awareness. But āyatana, which in Sanskrit means
a sphere or domain, indicates both the sense faculties (indriya) and
their corresponding sense-fields (viṣaya), a notion better envisioned
by 處, a locus in which sensation occurs.
Transcriptions
Like his predecessors, Xuanzang proposed new transcriptions for the
sounds of indic words. If one reads the travel accounts of Buddhist
pilgrims over the centuries, they invariably ‘correct’ the transcriptions
of place names and personal names that their predecessors had used,
often declaring the predecessors’ renderings ‘false’, unaware that the
phonetic value of Chinese characters altered over time and from
region to region. Kūkai 空海 (774–835), in his commentary on the
Heart Sūtra, evaluates the various Chinese versions available to him,
31
Cf. Ricci Association, ‘蘊’ .
210
his main criticism of Xuanzang’s version being that its concluding
mantra is useless since it mispronounces the power-sounds of the
Sanskrit. Nonetheless, if one listens even today to Koreans chanting
the Heart Sūtra using Xuanzang’s version, one will hear something
very close to gate gate paragate parasaṃgate bodhi svāhā, whereas the
modern mandarin pronunciation of the mantra is, as Kūkai complains, far from the Sanskrit sounds it is meant to invoke.
I will only discuss one example of a transcription change since, as
far as I can tell, no one has addressed this adequately yet. The term
ālayavijñāna is a signature Yogācāra concept. Prior to Xuanzang
there was either the translation used by Guṇabhadra in his translation of the Laṅkāvatāra sūtra: zangshi 藏識 which he presumably
used to highlight its relation in that text to the tathāgatagarbha
(rulaizang 如來藏), the latter being covered over (zang) by the ālaya
which receives the coarse obscurations from the vāsanās produced by
the other consciousnesses. Or, it was transcribed, as by Bodhiruci and
Paramārtha, as 阿梨耶識 or 阿黎耶識, the latter the more frequently
occurring.32 At some point, apparently around the time of the Sui
or early Tang, the pronunciation of the second character shifted. It
had originally been lai or something close—and that pronunciation
is still preserved in Cantonese and some other dialects. It had drifted
from lai to li. So it was no longer an adequate phonetic sign. Some
were now sounding out 阿黎耶識 as ‘ā-li-ye shi’, instead of approximating ālaya. Xuanzang replaced 黎/梨 with lai 賴. Thus 阿賴耶
識 phonetically renewed rather than replaced its predecessors, so
Chinese would continue to pronounce it as ālaya rather than aliya.
However, since the phonetic value apparently shifted around the
beginning of the Tang, texts that preserved the earlier graph were
read with the shifted value and that was preserved in Korean and
32
阿梨耶, which most standard modern discussions use, actually appears
only once in Paramārtha’s version of the Shelun: 復次此識於聲聞乘由別名如來
曾顯, 如增一阿含經言, ‘於世間喜樂阿梨耶、愛阿黎耶、習阿黎耶、著阿黎耶, 為
滅阿黎耶’. (T no. 1593, 31: 1.114b26–29)
Note the variants for li 梨 here, 阿黎耶識 or 阿黎耶 are found many times in
this text.
211
Japanese. That led to some confusion over the centuries, reflected not
just in the continued used of ‘aliya’ even today by some when discussing the eighth consciousness, but it even motivated certain scholars
to attempt to derive a Sanskrit etymology for ālīya—although no
such word exists in Sanskrit.
A most pronounced form of this misconception—this parikalpita—
is demonstrated in an early essay by D. T. Suzuki, titled ‘Philosophy
of the Yogācāra’ that appeared in 1904 in Le Muséon.33 In this essay,
Suzuki posited that the original Sanskrit term was ālīya-vijñāna, and
only later became ālaya-vijñāna, apparently misled by the history
of the transcriptions and their modern Japanese (and Mandarin)
pronunciations. Interestingly, L. de la Vallée Poussin, an editor of
the journal, adds in a footnote in French that no such Sanskrit term
as ālīya is known to him. He apparently urged Suzuki to defend the
claim, so Suzuki added a ‘note additionelle’ at the end of his essay
(page 385) spelling out his theory of ālīya > ālaya. Tellingly, in his
later writings the term ālīya never appears; he had learned that ālaya
was always the Sanskrit term. It is not clear if he knew that 阿黎耶
識 had, at least during the sixth century, been pronounced ālaya shi,
and not ālīya-shi, even though one of the popular ways 黎 is still
pronounced in Japanese is rei (= Ch. lai).
33
Suzuki, ‘Philosophy of the Yogācāra’.
212
Appendix I
Heart Sūtra Discrepancies Analysis
The issue of the missing phrase gets complicated when one compares
the various Chinese editions. It is also missing in the Kumārajīva
version, and neither Kuiji nor Wŏnch’ŭk mention it in their commentaries, though Wŏnch’ŭk in particular in several places not only
compares Xuanzang’s with Kumārajīva’s, but also says that he consulted the Sanskrit original. However, the Dunhuang version (Stein
collection, S. 700, included in T no. 256), which transcribes the Sanskrit in Chinese characters, and claims this is the version of the Ci’en
master, i.e., Xuanzang (not Kuiji, as some have claimed), does include
the first phrase.34
If we compare how the different Chinese translations of the Heart
Sūtra present these two or three lines, similarities and differences
become evident. In addition to the Heart Sūtra versions by Kumārajīva and Xuanzang, there are:
Pubian zhizang bore boluomiduo xin jing 普遍智藏般若波羅蜜
多心經 (T no. 252) translated by Dharmacandra (Fayue 法月
[653–743]) in 738;
Boreboluomiduo xin jing 般若波羅蜜多心經 (T no. 253)
translated by Prajña (Bore 般若 [fl. 741−798]) and Liyan
利言 (*Candra [?], c. 707–788+), etc., ca. 790;
Boreboluomiduo xin jing 般若波羅蜜多心經 (T no. 255)
translated by Facheng 法成 in the early ninth century;
Boreboluomiduo xin jing 般若波羅蜜多心經 (T no. 254)
translated by Prajñācakra (Zhihuilun 智慧輪 [?–876]) in the
mid-ninth century;
Foshuo shengmu boreboluomiduo xin jing 佛說聖佛母般若波羅
蜜多經 (T no. 257) translated by Dānapāla (Shihu 施護 [fl.
970s]) ca. 1005.
Cf. Hurvitz, ‘Hsüan-tsang and the Heart Scripture’; and Chen, ‘On Xuanzang’s Transliterated Version of the Sanskrit Prajñāpāramitāhṛdayasūtra’.
34
213
To quickly compare the passage in question, the translated by
Dharmacandra reads:
Addressing Śāriputra: ‘Bodhisattvas and Mahāsattvas should learn as
follows: The nature of form [= *rūpatva] is emptiness, the nature of
emptiness is form. Form is not different from emptiness; emptiness
is not different from form. Form precisely is emptiness, emptiness
precisely is form. Hedonic tone, associative thinking, embodied-conditioning and consciousness are also like this.’
於斯告舍利弗,“諸菩薩摩訶薩應如是學. 色性是空, 空性是色, 色
不異空, 空不異色. 色即是空, 空即是色. 受、想、行、識亦復如是.35
So Dharmacandra includes all three lines found in the received Sanskrit, including the first line, omitted by Kumārajīva and Xuanzang.
Prajña and *Candra has:
Śāriputra, form is not different from emptiness, emptiness is not
different from form. Form precisely is emptiness, emptiness precisely
is form; also likewise for hedonic tone, associative thinking, embodied-conditioning and consciousness.
舍利子! 色不異空, 空不異色. 色即是空, 空即是色. 受、想、行、識
亦復如是.36
Both the Dharmacandra and Prajña versions reflect the longer
Heart Sūtra version, so they are not drawn from the same source as
the Kumārajīva or Xuanzang versions, and would have been translated afresh from Sanskrit. While Dharmacandra includes the line
otherwise only in the Sanskrit, Prajña’s version omits that line, like
Kumārajīva and Xuanzang.
Prajñācakra arrived in China in mid-ninth century. One of the
monasteries he worked at was the Da Xingshan Monastery, the one
35
36
T no. 252, 8: 849a27–b1.
T no. 253, 8: 849c6–8.
214
associated with the transcribed Heart Sūtra cited above. Being an
Esoteric monk, he would have been associated with the Amoghavajra lineage; Hurvitz does comment that some of the preface
has tantric overtones. Starting in 855, he transmitted, in the Da
Xingshan Monastery, teachings and new translations to Ennin 圓仁
(794–864), the famous Japanese pilgrim. His version of the passage
in question reads:
Śāriputra, form (is) empty, emptiness is seen to be form; form is not
different from emptiness, emptiness is not different from form; it is
form that is emptiness, it is emptiness that is form; the same applies
to hedonic tone, associative-thinking, embodied conditioning and
consciousness.
舍利子! 色空, 空性見色. 色不異空, 空不異色. 是色即空, 是空即
色. 受、想、行、識亦復如是.37
So Prajñācakra also includes the first Sanskrit line.
Facheng, a Dunhuang scholar and translator, including of Chinese texts into Tibetan; mid-ninth century.
Form precisely is emptiness, emptiness precisely is form; form is not
different from emptiness, emptiness is not different from form; the
same applies…
色即是空, 空即是色. 色不異空, 空不異色. 如是受、想、行、識亦復
皆空.38
So he omits the first Sanskrit line.
Dānapāla (Shihu 施護), T no. 257, tran. ca. 1005. Dānapāla was
virtually the last translator of Indian texts into Chinese.
37
38
T no. 254, 8: 850a20–22.
T no. 255, 8: 850c4–5.
215
Why is the svabhāva of the five skandhas called emptiness? It is
referred to that way [because] precisely form is emptiness, precisely
emptiness is form; form is without difference from emptiness, emptiness is without difference from form. Likewise…
何名五蘊自性空耶? 所謂即色是空, 即空是色; 色無異於空, 空無異
於色. 受、想、行、識, 亦復如是.39
He appears to reverse the phrases, omitting the last line from both
the Sanskrit as well as from the Kuiji and Xuanzang Chinese versions.
39
T no. 257, 8: 852b20–22.
216
Bibliography
Abbreviation
D
T
Derge (sDe dge bsTan ‘gyur Canon). See Bibliography,
Primary Sources, D. 4036.
Taishō shinshū daizōkyō 大正新脩大蔵経. See
Bibliography, Secondary Sources, Takakusu and
Watanabe, eds.
Primary Sources
Boreboluomiduo xin jing 般若波羅蜜多心經 [Prajñā-pāramitāhṛdaya sutra; Heart Sūtra]. 1 fasc. Trans. Xuanzang 玄奘 on July
8, 649. T no. 251, vol. 8.
Boreboluomiduo xin jing 般若波羅蜜多心經, 1 fasc. Trans. Facheng 法
成 in the early ninth century. T no. 255, vol. 8.
Boreboluomiduo xin jing 般若波羅蜜多心經. 1 fasc. Trans. Prajña
(Bore 般若 [fl. 741−798]) and Liyan 利言 (Candra? [c. 707–
788+]), etc., ca. 790. T no. 253, vol. 8.
Boreboluomiduo xin jing 般若波羅蜜多心經. 1 fasc. Trans.
Prajñācakra (Zhihuilun 智慧輪 [?–876]) in the mid-ninth
century. T no. 254, vol. 8.
Chu sanzang ji ji 出三藏記集 [Compilation of Notes on the
Translation of the Tripiṭaka. 15 fascs. Comp. Sengyou 僧祐
(445–518), T no. 2145, vol. 55.
D. 4036 = Derge edition of the Tibetan translation of the
Śrāvakabhūmi of Asaṅga’s Yogācārabhūmi: rnal ’byor spyod pa’i
sa las nyan thos kyi sa (རྣལ་འབྱོར་སྤྱོད་པའི་ས་ལས་ཉན་ཐོས་ཀྱི་ས།). Translation
into Tibetan attributed to Jinamitra and Ye-shes-sde.
Da boniepan jing 大般涅槃經 [Mahāparinirvāṇa Sūtra; The Great
Parinirvana Sutra], 40 fascs. Trans. Dharmakṣema (Tanwuchen
曇無讖 [385–433]). T no. 374, vol. 12.
Da Cien si sanzang fashi zhuan 大慈恩寺三藏法師傳 [Biography of
(Xuanzang) the Tripiṭaka Master of Dacien Temple]. 10 fascs.
Written by Huili 慧立 and Yancong 彥悰. T no. 2053, vol. 50.
Datang Xiyu ji 大唐西域記 [(Xuanzang’s) Record of Travels to
217
Western Lands].12 fascs. Comp. Bianji 辯機 (?–649). T no. 2087,
vol. 51.
Foshuo shengmu boreboluomiduo xin jing 佛說聖佛母般若波羅蜜多
經 [Perfection of Wisdom of the Sacred Buddha-Mother (Heart
Sūtra)]. 1 fasc., tran. Dānapāla (Shihu 施護 [fl. 970s]) ca. 1005. T
no. 257, vol. 8
Pubian zhizang bore boluomiduo xin jing 普遍智藏般若波羅蜜多心
經 [The Heart Sūtra, Universal Storehouse of Wisdom]. 1 fasc.
Trans. Dharmacandra (Fayue 法月 [653–743]) in 738. T no. 252,
vol. 8.
Yuqieshi di lun 瑜伽師地論 [Yogācārabhūmi śāstra; Stages of Yoga
Practice Treatise]. 100 fascs. Wrriten by Maitreya (or Asaṅga),
trans. Xuanzang between July 3, 646–June 11, 648. T no. 1579,
vol. 30.
Secondary Sources
Chen, Shu-fen. ‘On Xuanzang’s Transliterated Version of
the Sanskrit Prajñāpāramitāhṛdayasūtra (Heart Sūtra)’.
Monumenta Serica 52 (2004): 113–59.
Davidson, Ronald. ‘Studies in dhāraṇī literature II: Pragmatics of
dhāraṇīs’. Bulletin of the School of Oriental and African Studies
77, no. 1 (2014): 5–61.
Digital Dictionary of Buddhism 電子佛教辭典, edited by Charles
Muller, editor-in-chief. http://www.buddhism-dict.net/ddb/
Hurvitz, Leon. ‘Hsüan-tsang 玄奘 (602–664) and the Heart
Scripture’. In Prajñāpāramitā and Related Systems: Studies in
Honor of Edward Conze, edited by Lewis Lancaster and Edward
Conze, 103–13. Berkeley: University of California at Berkeley
Press, 1977.
de Jong, Jan Willem. ‘Review of Ulrich Pagel’s The
Bodhisattvapiṭaka: Its Doctrines, Practices and Their Position in
Mahāyāna Literature’. Indo-Iranian Journal 39, no .2 (April 1,
1996): 176–82.
Keenan, John. A Study of the Buddhabhūmyupadeśa: The Doctrinal
Development of the Notion of Wisdom in Yogācāra Thought.
Ph.D. dissertation, University of Wisconsin-Madison, 1980.
218
Kroll, Paul. A Student’s Dictionary of Classical and Medieval
Chinese. Leiden: Brill, 2017.
Li, Rongxi, trans. The Great Tang Dynasty Record of the Western
Regions. Berkeley: Numata Center for Buddhist Translation &
Research, 1996.
Lusthaus, Dan. Buddhist Phenomenology: A Philosophical
Investigation of the Ch’eng Wei-shih lun. Curzon Critical Studies
in Buddhism Series. London: Routledge, 2002.
———. ‘The One Hundred Dharmas’. http://www.acmuller.net/
yogacara/outlines/100dharmas.html
Matsunami, Yasuo, and Śrāvakabhūmi Study Group, eds.
Śrāvakabhūmi, The Second Chapter with Asamāhitā bhūmiḥ,
Śrutamayī bhūmiḥ, Cintāmayī bhūmiḥ, Revised Sanskrit Text
and Japanese Translation. Tokyo: Sankibo Press, 2007.
Mimaki, Katsumi. ‘La Ṣaṇmukhī-dhāraṇī ou “Incantation des SIX
PORTES”, texte attribué aux Sautrāntika (I): Introduction’ [The
Ṣaṇmukhī-dhāraṇī, or ‘Incantation of Six Doors’, text attricuted
to Sautrāntika (I): Introduction]. Indogaku Bukkyōgaku kenkyū
印度學佛教學研究 ( Journal of Indian and Buddhist Studies) 25,
no. 2 (1977): 972–965.
———. ‘La Ṣaṇmukhī-dhāraṇī ou “Incantation des SIX PORTES”,
texte attribué aux Sautrāntika (II): Textes et Traduction’ [The
Ṣaṇmukhī-dhāraṇī, or ‘Incantation of Six Doors’, text attricuted
to Sautrāntika (II): Texts and Translation]. Nihon Seisō gakkai
kaihō 日本西蔵学会会報 (Report of the Japanese Association for
Tibetan Studies) 23 (1977): 9–13.
Miyamoto, KeiiCh. Daśapadārthī: An Ancient Indian Literature
of Thoroughly Metaphysical Realism. Kyoto: Rinsen Book
Company, 2007.
Orsborn, Matthew. ‘Something for Nothing: Cognitive Metaphors
for Emptiness in the *Upadeśa (Dàzhìdù lùn)’. Journal of Chinese
Buddhist Studies 31 (2018): 171–222.
Pagel, Ulrich. The Bodhisattvapiṭaka: Its Doctrines, Practices and
Their Position in Mahāyāna Literature. Tring: Institute of
Buddhist Studies, 1995.
Radich, Michael. ‘Text T. 0679’. The Chinese Buddhist Canonical
Attributions. No date. https://dazangthings.nz/cbc/text/984/.
219
Ricci Association, ed. ‘Le Grand Ricci Online’. Brill. Published
online Oct. 2017 https://chinesereferenceshelf.brillonline.com/
grand-ricci.
Shukla, Karunesha, ed. Śrāvakabhūmi of Acarya Asanga. Patna: K. P.
Jayaswal Research Institute, 1973.
Suzuki, D. T. ‘Philosophy of the Yogācāra’. Le Muséon 23, no .5
(January 1, 1904): 370–86.
Takakusu Junjirō 高楠順次郎, and Watanabe Kaigyoku 渡邊海旭,
eds. Taishō shinshū daizōkyō 大正新脩大藏經 [Buddhist Canon
Compiled under the Taishō Era (1912–1926)]. 100 vols. Tokyo:
Taishō issaikyō kankōkai 大正一切經刊行會, 1924–1932.
Tatia, Nathmal, ed. Abhidharmasamuccaya-bhāṣyam. Tibetan
Sanskrit Works Series 17. Patna: K. P. Jayaswal Research
Institute, 1976.
Ui, Hakuju. The Vaiśeṣika Philosophy according to the Daśapadārthaśāstra: Chinese text with introduction, translation and notes.
Edited by F. W. Thomas. London: Royal Asiatic Society 1917.
2nd ed. Varanasi: Chowkhamba Sanskrit Series, 1962.
Wayman, Alex. Analysis of the Śrāvakabhūmi Manuscript. University
of California Publications in Classical Philology, vol. 17. Berkeley
and Los Angeles: University of California Press, 1961.
From Chang’an to Nālandā:
The Life and Legacy of the Chinese
Buddhist Monk Xuanzang (602?–664)
Proceedings of the First International Conference on Xuanzang
and Silk Road Culture
EDITORS: SHI Ciguang, CHEN Jinhua, JI Yun and SHI Xingding
BOOK DESIGN: Carol Lee
PUBLISHER: World Scholastic Publishers
560416, AMK AVE 10, 13-1001, Singapore
EMAIL: eurice.d.shih@worldscholastic.com
ISBN: 978-981-14-6185-9
FORMAT: Paperback / Softcover
DATE OF PUBLICATION: 2020-05-01
LANGUAGE: English
All rights reserved. No part of this publication may be reproduced, stored in
a retrieval system, or transmitted, in any form, or by any means (electronic,
mechanical, photocopying, recording or otherwise) without the prior written
permission of the publisher.
National Library Board, Singapore Cataloguing in Publication Data
Names: Ciguang, Shi, editor. | Chen, Jinhua, 1966- editor. | Ji, Yun, editor. |
Xingding, Shi, editor.
Title: From Chang'an to Nālandā : the life and legacy of the Chinese Buddhist
monk Xuanzang (602?-664) / edited by Shi Ciguang, Chen Jinhua, Ji Yun
and Shi Xingding.
Description: Singapore : World Scholastic Publishers, 2020.
Identifiers: OCN 1156317437 | ISBN 978-981-14-6185-9 (paperback)
Subjects: LCSH: Xuanzang, approximately 596-664. | Buddhist monks--China-Biography.
Classification: DDC 294.361--dc23