Articles by alphabetic order
 Ā Ī Ñ Ś Ū Ö Ō
1 2 3 4 5 6 7 8 9 0

Defining the Chinese Buddhist Canon: Its Origin, Periodization, and Future

From Tibetan Buddhist Encyclopedia
Jump to navigation Jump to search
Ue no ghg.jpg

Journal of Chinese Buddhist Studies (2015, 28: 1–34)

New Taipei: Chung-Hwa Institute of Buddhist Studies

ᷕ厗ἃ⬠⬠⟙䫔Ḵ⋩ℓ㛇ġ 枩 1–34炷㮹⚳ᶨ䘦暞⚃⸜炸炻㕘⊿烉ᷕ厗ἃ⬠䞼䨞㇨

Defining the Chinese Buddhist Canon: Its Origin, Periodization, and Future

Fang Guangchang 㕡⺋拑

Professor, Department of Philosophy, Shanghai Normal University Translated by Xin Zi and Jiang Wu


This article is written by an eminent Chinese author of the Chinese canon and provides the author’s insights on how to understand the history of the canon in relation to Chinese Buddhism. Fang discusses the issues of historical periodization of the canon in history and various possibilities for the future development of the canon. This paper examines the history of the Chinese Buddhist canon by reflecting upon various essential issues. Fang first clarifies the origin of various name of the canon and proposes a working definition for the study of the canon. He also provides a periodization scheme of the canon and divides the history of the canon into the period of manuscript editions, the period of printed editions, the period of printed editions in modern times, and the period of digital editions. The author provides a detailed analysis of the characteristics of the canon during different periods and also makes suggestions for future studies.


Chinese Buddhist canon, Dunhuang manuscript, periodization, printing, digitization

斄㕤㻊㔯⣏啷䴻䘬⸦ᾳ⓷柴 㕡⺋拑 ᶲ㴟ⷓ䭬⣏⬠㱽㓧⬠昊⒚⬠䲣㔁㌰ 㐀天

㛔㔯䓙ᷕ⚳⣏啷䴻䞼䨞䘬叿⎵⬠侭㑘⮓炻娛徘Ḯᷕ⚳ἃ㔁⎚ᷕ⣏啷䴻 䘬㺼嬲ˤ㕡⺋拑㔁㌰忂忶⮵㻊㔯⣏啷䴻䞼䨞ᷕ䘬ᶨṃ㟡㛔⓷柴䘬⍵⿅炻㍸ ὃḮᶨᾳ斄㕤㻊㔯⣏啷䴻㬟⎚䘬ℐ㘗侫⮇ˤἄ侭椾⃰慸㶭Ḯ⣏啷䴻䘬⎵䧙 ␴Ἦ㸸炻᷎㍸↢Ḯᶨᾳ斄㕤⣏啷䴻䘬⭂佑ẍ⇑忚ᶨ㬍䞼䨞ˤ㛔㔯怬⮯⣏啷 䴻䘬㬟⎚忚埴Ḯ⤪ᶳ↮㛇烉⮓㛔㗪㛇ˣ⇣㛔㗪㛇ˣ役ẋ⇣㛔㗪㛇炻ẍ⍲㔠 䡤⊾㗪㛇ˤἄ侭⮵㭷ᾳ㗪㛇⣏啷䴻䘬䈡溆忚埴Ḯ℟橼↮㜸炻᷎㊯↢ḮṲ⼴ 䘬䞼䨞㕡⎹ˤ 斄挝娆烉 㻊㔯⣏啷䴻ˣ㔎䃴⮓㛔ˣ↮㛇ˣ⌘⇟⎚ˣ㔠⫿⊾ There have been different views on the specific time when Buddhism was first transmitted into China. One relatively credible version says in the first year of the Yuanshou reign (2 B.C.) of Emperor Aidi ⑨ⷅ in the Han dynasty, Yi Cun Ẳ⬀, an envoy from the central Asian state of Scythia, dictated the Buddha Snjtra (Futu Jing 㴖Ⰸ䴻) to Jing Lu 㘗䚏, a student of the Imperial College. While Yi Cun followed the Indian tradition of transmitting the snjtra orally, Jing Lu wrote it down according to the Chinese custom. Therefore, this text is the first translated Buddhist scripture in writing, suggesting that Buddhist scriptures and the Buddhist religion were transmitted into China simultaneously (Fang 1998b, 24–7).

Central Asian monks came for missionary work, and Chinese monks also went to Central Asia and India for Buddhist scriptures. As a result, Buddhist scriptures were translated into Chinese one after another. No longer subordinated to the mainstream teachings such as Confucianism and Taoism, Buddhism gradually became an independent and organic component of Chinese thought and culture.

In response, Chinese Buddhist scriptures have developed into a massive collection of the canon, which has been called Dazangjing ⣏啷䴻 (Jpn. Daizokyǁ, Kr. Daejanggyeong) or literally the “Great Storage of Scriptures.” The content of the Chinese Buddhist canon is related to many academic fields such as philosophy, history, sociology, language, literature, astronomy, geography, medicine, and so on. As a result of cross-cultural communication, the Chinese Buddhist canon influenced the entire Chinese cultural sphere profoundly. It also contains abundant information for studying Chinese and East Asian culture.

In this paper I will offer a panoramic view of the history of the Chinese Buddhist canon by reflecting upon various essential issues. I will first clarify the etymological origin of the term Dazangjing and propose a working definition for the study of the canon based on my identification of three essential elements in canon formation: selection criteria, structural system, and external markers.

This chapter also provides a periodization scheme of the canon based on my working definition and divides the history of the canon into the period of manuscript editions, the period of printed editions, the period of printed editions in modern times, and the period of digital editions. In addition, I will provide a detailed analysis of the characteristics of the canon during different periods and make suggestions for future studies.

Etymology of the Term Dazangjing

The Chinese Buddhist canon was often referred to as Dazangjing, a Buddhist term invented by Chinese people and without a counterpart in Sanskrit. The creation of the term Dazangjing resulted from the synthesis of Chinese and Indian cultures and the development of Chinese Buddhism. In my view, the evolution of its meaning is related to three factors: Chinese views on translating Buddhist scriptures, the popular devotion to the Three Refuges (sanbao ᶱ⮞) during the Southern and Northern dynasties, and the massive production of Chinese Buddhist scriptures.

Etymologically, dà ⣏ (great) is a modifier that signifies the scope of the canon, suggesting that its content reaches the limit of time and space. Zàng 啷 (storage) is a paraphrase of the Sanskrit word piΛaka, which means “cases” or “baskets” for storage. Because paper was not introduced in ancient India, scriptures were carved or written on palm leaves, which were made into the so-called “palm-leaf Buddhist scriptures.” Indian monks usually put these palm-leaf scriptures in cases or baskets, namely, piΛaka. Therefore, piΛaka has gradually become a measuring unit and an alternative name for Buddhist scriptures. Scriptures of different categories were stored in different piΛakas. For example, scriptures under the categories of sǍtra, vinaya, and abhidharma were stored in three separate “baskets,” which is where the name TripiΛaka comes from.

-ưng 䴻 (scripture) is translated from the Sanskrit word snjtra, whose original meaning is “running through.” Buddhists in ancient India believed that if flower petals were bound with strings, they would not be blown away by the wind. Similarly, collecting the words of Buddha’s teaching would preserve them forever so that they could be handed down to later generations. Therefore they were called snjtras. The Chinese character jƯng 䴻 originally referred to vertical lines in fabric, with an extended meaning of “constancy.” Thereupon, the word Mưng embodies the Chinese traditional thought that truth can last forever, as Heaven does.

As KumƗrajƯva’s 沑㐑伭Ṩ (334–413) famous student Sengzhao 慳₏倯 (384–414?) says in his Commentary to the VimalakƯrti Sutra (Zhu Weimo jing 㲐䵕㐑䴻), “-ưng means being constant. Although things have been changing from ancient times to the present, the cardinal truth does not change at all. Neither nonbelievers nor Buddha’s disciples could make any changes. That’s why Mưng is considered constant and eternal” (T 38: 327c). It seems that Mưng, the Chinese translation of snjtra, reflects Chinese Buddhistsboundless devotion to and faith in Buddha’s teaching, though this translation does not correspond exactly to the original meaning of the word.

In ancient India, the word snjtra only refers to one of the three “baskets,” comprising snjtra, vinaya, and abhidharma. However, in Chinese, the meaning of Mưng has been expanded gradually. There are three usages of this word. First, it is equivalent to snjtra in Indian Buddhism, referring to all the translated Buddhist scriptures transmitted from India. Second, since the beginning of the transmission of Buddhism into China, Chinese people always called all the texts Mưng, including vinaya and abhidharma. Third, it has been used in phrases such as Dazangjing, which includes Chinese Buddhist works written and edited by Chinese people.

The Chinese Buddhist canon has been a research subject since the twentieth century. Scholars attempted to find out when the word Dazangjing first appeared and usually assumed that this phrase was created during the Sui dynasty (Daizǀkai 1990, 22). According to the record written by Guanding 㿴 枪 (561–632), Separate Biography of Tiantai Master Zhizhe of the Sui Dynasty (Sui Tiantai Zhizhe dashi biezhuan 昳⣑冢㘢侭⣏ⷓ⇍⁛), Zhiyi

柿 (538–597) “copied fifteen sets of Dazangjing” (T 50: 197c) throughout his life. This suggests that the word Dazangjing appeared during the Sui dynasty. I used to hold this opinion as well. However, at the suggestion of Japanese scholar Fujieda Akira 喌㝅㗫, I carefully examined the nuanced meaning of the few lines at the end of this biography and found that this paragraph is actually not Guanding’s original writing, but supplementary remarks added by a Master Xian 戹㱽ⷓ (d.u.). Furthermore, all the information on this Master Xian needs further study. So we cannot take the appearance of Dazangjing in Guanding’s biography of Zhiyi as the evidence proving that the term first appeared in the Sui dynasty.

To clarify the origin of the first use of the term Dazangjing, I have examined a number of sources. My findings suggest that the term Dazangjing ⣏啷䴻 must have been invented before the Buddhist persecution around 845, or during the Zhenyuan 屆⃫ reign of Emperor Dezong ⽟⬿ (785–805) in the Tang dynasty at the latest.

In the first place, when collating Dunhuang manuscripts, I found the occurrence of this term in two obscure manuscripts: first, Catalog of the Indian MahƗyƗna and HƯnayƗna Snjtra, Vinaya, and Abhidharma in the Great Tang (Xitian daxiaocheng jinglülun bingjian zai Da Tang guonei dushu mulu

大⣑⣏⮷Ḁ䴻⼳婾᷎夳⛐⣏Ⓒ⚳ℏ悥㔠䚖抬), which appeared in both the Pelliot and the Stein collections as P 2987 and S 3565, respectively; second, Number of Scriptures in the Great Storage of the Great Tang (Da Tang Dazangjing shu ⣏Ⓒ⣏啷䴻㔠), which was preserved only in the Pelliot collection as P 3846. Based on the scribal style, the former two Dunhuang manuscripts must have been written during the period when the Allegiance Army (Guiyi Jun 㬠佑幵) controlled Dunhuang (851–1036).

However, after analyzing the contents, I believe that these two documents date back no earlier than Emperor Xuanzong’s Ⓒ䌬⬿ period (712–756) and no later than the Buddhist persecution around 845. As for the manuscript numbered P 3846, it must have appeared after the Buddhist persecution around 845. Therefore, I concluded that the word Dazangjing appeared in the period between Emperor Xuanzong’s time and the Buddhist persecution during the Huichang reign (Fang 2002a, 1).

Second, when I was searching the Chinese Buddhist canon in electronic format, I found a sentence saying, “those hundreds and thousands of copies of liturgy were again abbreviated from the Bodhisattva Dazangjing” (℞䘦⋫枴 㛔炻⽑㗗厑啑⣏啷䴻ᷕ㫉䔍 T 39: 808a), in the first fascicle of Commentary on Essential Secret Teachings of the Great Yoga of the Adamantine Crown (Jingangding jing dayujia mimi xindi famen yijue 慹∃枪䴻⣏䐄ụ䦀⭮⽫⛘ 㱽攨佑始). Although this text is not contained in any edition of the Chinese Buddhist canon in Chinese history, Taishǁ Canon collects it based on a Japanese reprint. The Japanese version originated from KǍkai’s 䨢㴟 (774–

835) collection during his visit to China in the Tang dynasty from 804 to 806.

Finally, I found a record in the book Record of Inheritance of Two Sets of

Great Teaching (Liangbu dafa xiangcheng shizifufaji ℑ悐⣏㱽䚠㈧ⷓ屯Ẁ㱽

姀), written by Haiyun 㴟暚 (d.u.) of the Tang, saying that “according to the Sanskrit version, this text was translated into six scrolls. In addition, one comprehensive scroll was compiled to teach the procedure of practice and chanting. In total, there were seven scrolls, which were made into one whole set to be put into Dazangjing” (T 51: 785c). Haiyun’s work was written in the eighth year of the Taihe ⣒␴ reign (834) of Emperor Wenzong Ⓒ㔯⬿ (809–840) in the Tang dynasty, before the Buddhist persecution during the Huichang reign.

Both documents mentioned before, P 2987 and S 3565 in the Dunhuang manuscripts, contain the phrase Xitian Dazangjing 大⣑⣏啷䴻 (the Indian Buddhist canon), which is a massive collection of 84,500 scrolls. From this we know that the Chinese who created the term Dazangjing did not limit it to meaning the Chinese Buddhist canon, but actually used it as a common term for all Buddhist literature. Nonetheless, in ancient times, Buddhist communities in other traditions continued to transmit their own scriptures and use their own traditional terms.

For example, HƯnayƗna Buddhist literature is called Tripi৬aka; Tibetan Buddhist literature is named Ganggyur 䓀䎈䇦 or Tanggur ᷡ䎈䇦. The phrase Hanwen Dazangjing 㻊㔯⣏啷䴻 (the Chinese Buddhist canon) was first used by Japanese scholars at the beginning of the twentieth century. Chinese Buddhists, focusing on the integration of Buddhist literature written in different languages, also used a series of new terms, such as the PƗli Buddhist canon (Bali Dazangjing ⶜⇑⣏啷䴻), the Southern Buddhist canon (Nanchuan Dazangjing ⋿⁛⣏啷䴻), the Tibetan Buddhist canon (Zangwen Dazangjing 啷㔯⣏啷䴻), the Mongolian Buddhist canon (Mengwen Dazangjing 呁㔯⣏啷䴻), the Manchu Buddhist canon (Manwen

Dazangjing 㺧㔯⣏啷䴻), the Tangut Buddhist canon (Xixia Dazangjing 大

⢷⣏啷䴻), and so on. For the purpose of comparison, the term Dazangjing in Chinese Buddhism naturally evolved into Hanwen Dazangjing 㻊㔯⣏啷䴻 (Chinese Buddhist canon). Therefore, Hanwen Dazangjing and Dazangjing are only different names from different historical and linguistic backgrounds that actually have the same referent: that is, the Chinese Buddhist canon.

The Definition of the Chinese Buddhist Canon

In China, people at first called the Buddhist canon zhongjing 䛦䴻 (Myriad Scriptures), yiqiejingġ ᶨ↯䴻 (All Scriptures), and zangjingġ 啷䴻 (Storage of Scriptures). The term Dazangjing only appeared in the Tang dynasty. If we examine these terms carefully, we can see that their emergence and changes reflected the Chinese conception of Buddhist scriptures. However, when they first appeared, these names were not clearly defined but simply followed longtime conventions of usage. In modern times, along with the development of scholarly research on Buddhism, the Buddhist canon has increasingly attracted people’s attention, and scholars have attempted to define the meaning of the Chinese Buddhist canon.

What is “the Chinese Buddhist canon”? Twenty years ago, my definition was “the whole collection of Chinese Buddhist literature” (Fang 2002a, 3). When I review that definition now, I find it not very accurate. The connotation of the so-called “whole collection of Chinese Buddhist literature” should be the Buddhist literature written in Chinese, and the extension should be all Chinese Buddhist literature. However, the fact is that all the literature collected in the Chinese Buddhist canon is written in Chinese, but not all of it is Buddhist literature.

For example, there are works such as SƘφkhya kƘrikƘ (Jin Qishi Lun 慹ᶫ⋩婾) and VaiğHΙikadağapadƘrthağƘstra (Shengzong Shiju Yilun ⊅⬿⋩⎍佑婾), which belong to the SƘՉkhya school 㔠婾㳦 or VaiğHԕika school ⊅婾㳦 in India. In addition, not all Chinese Buddhist literature was collected into the Chinese Buddhist canon. A large number of Buddhist texts exist outside it. Moreover, the formulation “a whole collection of the Chinese Buddhist literature” cannot demonstrate the fact that the Chinese Buddhist canon is an organic unity with fixed content, internal logical structure, and external appearance, such as case number.

When ancient scholars were compiling the Chinese Buddhist canon, they had certain selection criteria, structural designs, and identification methods. Based on this fact, twenty-six years ago I proposed the theory of “three essential elements” of the Chinese Buddhist canon: selection criteria, structural system, and external identification markers. My definition did not highlight these elements, so now I improve it and express it as follows:

Including essentially the translated Buddhist scriptures of past ages as the core of its content, the Chinese Buddhist canon is the collection of the Chinese Buddhist classics and related literature organized according to certain structures and with some external identification markers.

Here, I add the “three essential elements” to the above definition as modifiers. First, I use the wording “including essentially the translated Buddhist scriptures of past ages as the core of its collection” to indicate the selection criteria because the Chinese Buddhist canon has incorporated all the translated Buddhist scriptures as its core. I also put in a quantitative limit, expressed as “including essentially,” to distinguish the Chinese Buddhist canon from the collections of abridged scriptures such as Essential Texts from the Canon (Zangyao 啷天). In addition, I use the phrase “related literature” to show that the Chinese Buddhist canon includes some non-Buddhist literature. Historically, the Chinese Buddhist canon collected scriptures of the Indian

6Ƙৄkhya and Vaiğeৢika schools. Taishǁ Canon also has a catalog of “NonBuddhist Religions” (Waijiaobu ⢾㔁悐), including Taoist, Manichean, and Nestorian scriptures.

My definition does not stress the structure of the Three Baskets (tripiΛaka)—snjtra, vinaya, and abhidharma—which form a specific structure. In the history of Indian Buddhism, this structure never became a universal way of classifying Buddhist scriptures. It was the same in the history of Chinese Buddhism, and this way of classification was given up long ago. For example, Taishǁ Canon totally abandoned it. To look forward to the future, there is no possibility of it surviving. Therefore, I did not stress the traditional structure of TripiԮaka as the core of the canon.

One of the functions of a definition is to explain the essential aspects of the research subject by its connotations and extensions. A definition should be able to describe every single stage in the development of the subject. A formulation such as “take the TripiԮaka as the core” is only suitable for a certain historical period, and therefore cannot define the entire history of the canon. Unlike this narrow definition, my formulation—“including essentially the translated Buddhist scriptures of past ages as the core of the canon”— already covers the contents of the traditional TripiԮaka. Periodization Criteria

For thousands of years, the content, structure, and appearance of the Chinese Buddhist canon has been changing. In order to study the canon, dividing its history into historical periods is necessary. To determine these periods, we need a set of feasible criteria.

Most scholars take dynastic change as their criterion for determining the periods of Chinese Buddhism. I do not agree with that. Instead, I prefer to determine the periods of Chinese Buddhism by the inherent logic of its development (Fang 1998a). And I do not think that we can apply the periodization standard of Chinese Buddhism to the history of the Chinese Buddhist canon, because the Chinese canon exists independently and has its own history.

To explore the inherent logic of the development of the Chinese Buddhist canon, we have to examine the various factors that stimulated its transformation. The following five factors have affected the evolution of the Chinese Buddhist canon.

Chinese Buddhism

As a collection of books that has recorded and reflected the history of Chinese Buddhism, the Chinese Buddhist canon is conditioned by the development of Chinese Buddhism from beginning to end. Therefore, the canon evolves with the changes in Chinese Buddhism. (See below for details.)

The Factors Irrelevant to Buddhism

The Chinese Buddhist canon was also affected by Chinese feudal dynasties. China had been a highly centralized autocratic empire since Emperor Qin Shihuang 䦎⥳䘯 united the country in 221 B.C. The state power had the supreme position, which could not be counterbalanced by other forces. In history, the Chinese ruling class supported or suppressed Buddhism based on their own interests. Dominating the relationship, the state power imposed its own will upon Buddhism. After the 1911 Revolution, the power of Chinese feudal dynasties over Buddhism no longer existed. However, political and intellectual factors outside Buddhism continued to affect the compilation of the Chinese Buddhist canon to some extent.

The Compilers

Different editions of the canon were compiled by different people. All the differences of time, place, compilers’ guidelines, principles, scholarship, and method determine the differences among various editions. It is also necessary to consider the gaps between the compilers’ subjective expectations and objective realities, and the interactions arising from them.

Physical Form and Printing

After Buddhism was introduced to China, the physical form of Chinese books changed from bamboo and wooden slips to silk, and then to paper. The way of producing books shifted from handwritten to block printing and to typographic printing. A series of printing technologies, such as photocopying and laser typesetting, has emerged since the end of the nineteenth century. Along with the development of digital technology, a revolution of form and book production has taken place in recent years. All this has brought about major changes in book format and the appearance of the Chinese Buddhist canon.

Bookbinding and Layout

Since Buddhism was transmitted into China, there have been various ways of binding in the history of the Chinese book, such as butterfly binding, stitched binding, whirlwind binding, concertina binding, wrapped-back binding, modern paperback and hardcover, and even the e-book. Bookbinding and layout have thus become unavoidable topics in research on the Chinese canon.

Due to these five factors, the Chinese Buddhist canon has shown different physical appearances as time has gone by. In my opinion, these five factors affected the Chinese Buddhist canon in five different ways. If we plan to study the Chinese Buddhist canon from one perspective, then we should consider one of these factors as the criterion for periodization. To some extent, most of these could serve as the criterion. However, in practice, the five factors are not compatible and thus cannot be integrated. Since the twentieth century, scholars have mainly focused on researching block-printed editions. Later, the study of handwritten manuscript editions appeared on scholars’ horizon. Because the Chinese Buddhist canon is a kind of book collection, taking the historical development of the Chinese book as the periodization criterion for the canon is also a convenient choice.

Considering all these factors, I divide the development of the Chinese Buddhist canon into four periods: hand-writing, block-printing, modern printing, and digital. In the following, I will outline the evolution of the canon in each stage and its characteristics.

The Hand-writing Period

The hand-writing period is the beginning and foundation of the Chinese Buddhist canon, which can be divided into six stages.

The Preparation Stage: First to Fifth Century

This stage corresponds roughly to the period from the transmission of Buddhism into China to Daoan’s 忻 ⬱ (312–385) time, during which Buddhism was first considered equal to “the Daoist Learning of Emperor Huangdi and Master Laozi” (Huang-Lao zhi xue 湫侩ᷳ⬠) and then was subordinated under Neo-Taoism. Despite the emergence of some excellent Buddhist scholars, Chinese Buddhism had not become independent due to lack of clear self-consciousness. Although Daoan once questioned the practice of “matching the meaning” (geyi 㟤佑), which could be considered a vague sense of self-consciousness, he himself could not completely get rid of the influence of “geyi” because of the historical conditions.

The Chinese translation of Buddhist scriptures was also in a chaotic situation during this stage. The quantity of translated scriptures was considerable. Some translators intentionally translated sectarian scriptures and some monks even traveled to Central Asia and India to seek new scriptures. The overall situation was that the translators would translate whatever scriptures they came across, whether complete or not.

From Daoan’s catalog, Comprehensive Catalog of Scriptures (Zongli zhongjing mulu 䵄䎮䛦䴻䚖抬), we can see that Chinese Buddhists did not realize or perceive the necessity to distinguish and collate Buddhist scriptures under the categories of MahƘ\Ƙna and HưnayƘna. The transmission of scriptures also varied geographically. There was no unified or standardized edition of the Chinese Buddhist canon all over the country. This accorded with the level of development of Chinese Buddhism and the political situation of the sixteen separate states in the Eastern Jin dynasty (316–420).

Daoan was the first person to raise the issue of apocryphal scriptures in the history of Chinese Buddhism. In addition, Daoan’s catalog, Zongli zhongjing mulu, following the elaborate Chinese bibliographical tradition, attempted to record detailed information about translators and the time and place of translation of every scripture in chronological sequence.

This shows that the formation of the Chinese Buddhist canon has been closely connected to traditional Chinese culture and thought since the early time. Daoan would not have thought in terms of “the three elements of the Chinese Buddhist canon” that I have used in the definition. But in fact, the issue of apocryphal scriptures in his work did involve the first element— criterion of selection. Therefore, the chaotic state of Chinese Buddhist scriptures was also the preparation stage of the Chinese Buddhist canon.

The Formation Stage: Fifth to Sixth Century

This stage corresponds to the period from KumƘrajưva coming to China to Fei Changfang’s 屣攟㇧ (dates unknown) compilation of the Record of the Three Jewels through the Ages (Lidai sanbao ji 㬟ẋᶱ⮞姀) in the Sui dynasty. During his stay in China, KumƘrajưva translated NƘJƘrjuna’s 漵㧡 (d.u.) MƗdhyamika theory systematically and introduced a new world to Chinese monks, allowing them to study authentic Indian Buddhism. From then on, Chinese Buddhism obtained a clear self-consciousness and developed independently, and conflicts among Confucianism, Daoism, and Buddhism began to take place.

Based on a deepened understanding of Buddhism, KumƘrajưva’s disciple Huiguan ㄏ奨 (d.u.) developed the theory of “Classification of Teaching into Five Periods” (wushi panjiao Ḽ㗪⇌㔁). After him, different classification theories emerged. Their purpose was to organize different concepts from Indian Buddhism that had been transmitted into China into an organic system. The emergence of the “Classification of Teaching” was a significant event in the history of Chinese Buddhism because it helped to spread Buddhism in China and stimulated the creation of indigenous Buddhist schools in the Southern and Northern dynasties and the Sui and Tang dynasties. Such classifications involve the second element of the Chinese Buddhist canon that I have defined—the structure. Therefore, “Classification of Teaching” meant much to the formation of the Chinese Buddhist canon.

It was the Separate Catalog of All Scriptures (Zhongjing bielu 䛦䴻⇍抬), by an anonymous author, that introduced the “Classification of Teaching” into the organization of Buddhist scriptures. The author of this catalog absorbed Huiguan’s “Classification of Teaching into Five Periods” and developed many categories such as “Catalog of MahƘ\Ƙna Scriptures” (Dacheng jinglu ⣏Ḁ䴻

抬), “Catalog of HưnayƘna Scriptures” (Xiaocheng jinglu ⮷Ḁ䴻抬), “Catalog of Universal Teachings of Three Vehicles” (Sancheng tongjiaolu ᶱḀ忂㔁 抬), “Catalog of the Great Vehicle of Three Vehicles” (Sancheng zhong Dacheng lu ᶱḀᷕ⣏Ḁ抬), and “Catalog of Undecided Scriptures of the Great and Lesser Vehicle” (Daxiaocheng bupanlu ⣏⮷Ḁᶵ⇌抬). This was a tentative but helpful attempt to determine a structural system for arranging Buddhist scriptures. Around that time, Sengyou’s ₏䣸 (445–518) catalog, Compilation of Notes on the Translation of the TripiΛaka (Chusanzang jiji ↢ ᶱ啷姀普), appeared and made great contributions to the preservation of documents by separating original scriptures from apocryphal ones. But in the classification of Buddhist scriptures, Sengyou’s catalog retrogressed to the level of Daoan’s catalog (Zongli zhongjing mulu), not adopting the new classification scheme developed in the anonymous Separate Catalog of All Scriptures.

Later on, many catalogers, such as Li Kuo 㛶⹻ in the Northern Wei dynasty (386–534), Baochang ⮞ⓙ in the Liang dynasty (502–557), and Fashang 㱽ᶲ in the Qi dynasty (550–577), compiled their own catalogs. This shows that Chinese monks tried to grasp the essential characteristics of the Chinese Buddhist canon in order to collate and distinguish Buddhist scriptures and to design an organizational structure. In these catalogs, we can see clearly that the development of Chinese Buddhism, the organization of the Indian Buddhist canon, and the Chinese cataloging tradition influenced the canon profoundly.

According to the documents handed down from ancient times, Wei Shou’s 櫷㓞 (507–572) “TripiԮaka Prayer of the Northern Qi Dynasty” (Beiqi sanbu yiqiejing yuanwen ⊿滲ᶱ悐ᶨ↯䴻ョ㔯) and Wang Bao’s 䌳墺 (513–576) “TripiԮaka Prayer of the Zhou Dynasty” (Zhou zangjing yuanwen ␐啷䴻ョ㔯) prove that in the Southern and Northern dynasties (420–589), the governments of the Northern Qi (550–577) and Northern Zhou (557–581) had compiled the Chinese Buddhist canon. In the Southern dynasties, Baochang’s ⮞ⓙ Catalog of All Scriptures (Zhongjing mulu 䛦䴻䚖抬) was written under the command of Emperor Wudi 㬎ⷅ (464–549) in the Liang dynasty (502–557), with the purpose of compiling the Chinese Buddhist canon.

According to Dunhuang manuscripts, governments began compiling the Chinese Buddhist canon even earlier. There is a batch of Buddhist scriptures copied by Dunhuang official scribes from 511 to 514 in the Northern Wei dynasty. The official scribes employed in Dunhuang copied scriptures for years. What they were copying could not be one single volume, and it must be part of the Chinese Buddhist canon. The remaining volumes mentioned above include both MahƗyƗna and HƯnayƗna Snjtras and MahƗyƗna and HƯnayƗna Abhidharma. Most of these are ordinary Buddhist scriptures rather than those that one could gain merit by copying, such as the Lotus SǍtra and the Diamond SǍtra. This could also prove that the scriptures copied by official scribes in Dunhuang are components of the Chinese Buddhist canon.

After 1,500 years, these fifteen scrolls survived, and there were two copies of fascicle 14 of Xuanzang’s 䌬⤀ (602–664) Discourse on the Establishment of Consciousness-only (Cheng Weishilu ㆸ ⓗ 嬀 婾 ). This shows that these remaining scrolls belong to at least two sets of the canon. If we consider the year recorded in fascicle 8 of Discourse on the Establishment of Consciousness-only, they could have belonged to three different sets of the canon. The discovery of the transcriptions in Dunhuang also proves that Buddhist beliefs flourished there during the Northern Wei dynasty. At least, copying scriptures had become an official undertaking west of the Yellow River, showing that Buddhism was a significant social force at that time.

Some documents in Dunhuang manuscripts also demonstrate that as early as the later fifth century, private sponsorship of the canon was a trend in north China. Evidence can be found in fascicle 6 of SaΥyuktƘbhidharmahΩdaya (Za apitan xinlun 暄旧㭿㙯⽫婾, S. 00996): a colophon attached to the end of that volume records that a person named Feng Jinguo 楖㗱⚳ created ten sets of the canon (Yiqiejing ᶨ↯䴻, All scriptures), each including 1,464 scrolls. These figures indicate the size of the Chinese Buddhist canon at this time. It is a pity that only one of the 1,464 scrolls in total made by Feng Jinguo has been discovered. During the formation stage of the handwritten manuscript canon, popular devotion to the Three Treasures or Refuges, which includes the Dharma, became another driving force for the creation of the Chinese canon. Traditionally, the Three Refuges—Buddha, Dharma, and SaՉgha—are essential components of Buddhism. Therefore, all three are worshipped by Buddhists.

As the embodiment of “Dharma Treasure,” Buddhist scriptures are worshipped as well. Here, the dividing line between philosophical Buddhism and faith-based Buddhism had been very clear. The ordinary people’s major Buddhist activities were confession and merit accumulation. Specifically, making private copies of Buddhist scriptures and reciting and upholding the scriptures constituted their daily Buddhist practice. Descriptions of copying, reciting, and worshipping Buddhist scriptures, and the merit that those activities could create, can be found in many scriptures. Those descriptions helped to promote the activities of faith-based Buddhism.

The apocryphon, Snjtra of AvalokiteĞvara [Promoted by] Lord Gao (Gaowang guanshiyin jing 檀䌳奨ᶾ枛䴻) is one of the products and proofs of those activities. Stronger physical evidence of the worship of Buddhist scriptures is the twenty-fascicle version and sixteen-fascicle version of BuddhabhƘΙita-buddhanƘma SǍtra (Fo ming jing ἃ⎵䴻) found in Dunhuang, which show how the initial twelvevolume version translated by Bodhiruci 厑㍸㳩㓗 (d.u.) during the years of Zhengguang’s 㬋 ⃱ reign in the Northern Wei dynasty (520–524) was developed into the thirty-fascicle version in TripiΛaka Koreana. As devotion to the Three Jewels (sanbao ᶱ⮞ ) spread, more and more people valued Buddhist scriptures and took their veneration as an important Buddhist practice, thus promoting the creation of the Chinese Buddhist canon (Fang 1990, 470–89).

China’s profound cultural heritage was another important factor in the formation of the Chinese Buddhist canon, as the Chinese have a strong consciousness of being a great civilization. China has self-awareness and a sense of superiority, which Chinese people consciously spread and perpetuated by all kinds of methods.

This high self-consciousness of civilization has been manifested in collecting, maintaining, and preserving the books from the past in order to sort out, analyze, and integrate different thoughts, by which later generations can “cultivate themselves, regulate family, govern the country, and pacify the world.” Following Confucius, scholars engaged in compiling books, generation by generation. After the great unification achieved during the Qin and Han dynasties, all Chinese emperors took it as an essential activity to collect, compile, and catalog books. This cultural tradition was profound and magnificent, and formed certain social conditions ready to absorb and digest foreign cultures such as Buddhism.

After Indian Buddhism was introduced into China, it went through confrontation and domestication within Chinese traditional culture. Indian Buddhism transformed Chinese traditional culture greatly and also changed itself, gradually developing into Chinese Buddhism, which was tightly connected with Chinese traditional culture and became one of the three major Chinese traditions, together with Confucianism and Daoism.

It was against this background that the unified Chinese Buddhist canon—corresponding to the unified political empire—took form. In contrast to India, which was never truly unified, every unified Chinese dynasty would compile a standard history for the previous dynasty and as well as its own edition of the Chinese Buddhist canon.

Fei Changfang’s catalog, Records of the Three Jewels through the Ages (Lidai sanbao ji 㬟ẋᶱ⮞姀), compiled in 597, marks the end of the formative period of the Chinese Buddhist canon. Judging from its title, one can easily tell that this work is a direct product of the “Cult of Three Jewels.” Scholars of later generations criticized this work for not following the stylistic rules and layout of scriptural catalogs, and Taishǁ Canon even put his work into the category “History and Biography” rather than “Catalogs.” This happened because those scholars did not fully understand the social and historical background against which Fei’s work was compiled.

Fei Changfang invented the classificatory rubric of Register of Canonical Texts (Ruzang lu ℍ 啷 抬 ) and corrected the previous convention of listing titles under categories such as “Derivative Scriptures” (Biesheng ⇍䓇, meaning excerpts from a complete scripture), “Doubtful Scriptures” (Yihuo 䔹 べ ), and “Apocrypha” (Weiwang ‥ ⤬ ). Fei’s innovation shows that the Chinese Buddhist canon had evolved from the stage of spontaneous dissemination to that of theoretical sophistication. Therefore, the canon had taken shape in both practice and theory. The stylistic rules and layout of Fei’s catalog (Lidai sanbao ji) were adopted by later influential scriptural catalogs such as Daoxuan’s 忻 ⭋ (596–667) Neidian lu ℏ ℠ 抬 , Da Zhou lu ⣏ ␐ 抬 , Zhisheng’s 㘢㖯 (d.u.) Kaiyuan lu 攳⃫抬, and Zhenyuan lu 屆⃫抬, and they became the most basic catalogs.

The Stage of Structural Systematization: Sixth to Ninth Century

This stage corresponds to the time period from the completion of Fei Changfang’s catalog to the Buddhist persecution in 845, during which scholarmonks who were in charge of the compilation of scriptural catalogs tried to work out the structure of the Chinese Buddhist canon from different angles. Zhisheng’s Catalog of Buddhist Works Compiled during the Kaiyuan Period (Kaiyuan shijiao lu 攳⃫慳㔁抬), which integrated the accomplishments of earlier scholars, became a model for later generations. Zhisheng’s contribution to the structural system of the canon and the catalog of Buddhist scriptures represented the highest level of Chinese Buddhist bibliographical study in ancient times.

As the Chinese Buddhist canon was developed, especially with the appearance of “combined cases” (hezhi ⎰ⷁ), the issues of “external markers” of the canon were put on the agenda. In the formation period, the “title-label method” (Jingming biaozhi fa 䴻⎵㧁娴㱽) employed one character from the title of the text in the canon to label each individual scroll or case. The “fixed shelf storage method” (dingge chucun fa ⭂㟤⃚⬀㱽) was also invented. In the Dunhuang region ruled by Tibetans, there appeared the “verse-based case number method” (jisong zhihao fa ῰枴塇嘇㱽), which employed characters in popular liturgical verses to mark each case.

At that time, scholar-monks studied further translated scriptures and Buddhist thought, writing a large number of works in response. Different Chinese Buddhist schools took form, and scholars from these schools wrote many works in order to elaborate their own doctrines. In addition, a variety of Chinese Buddhist writings, such as annals, liturgical texts, catalogs, translations, scripture extracts, and other faith-based works, appeared in great numbers.

Some of these Chinese works were collected into the canon, but most were excluded by monk-compilers because the Chinese Buddhist canon mainly collected the translated scriptures. If it is said that in the first two stages, the development of the canon coincided with that of Chinese Buddhism, after this stage of structural systematization, the orthodox canon tended to be fixed and could not really reflect the progress of Chinese Buddhism. To supplement the main canon and make up for this deficiency, there appeared “separate canons” (Biezang ⇍啷). The Vinaya school compiled their own VinayapiΛaka 㭿⯤啷, while the collection of doctrinal works of the Tiantai school was also popular. In addition, the Zen school also collected works on Chan, naming their collection “Chan Canon” (Chanzang 䥒啷). These phenomena deserve our attention.

The Stage of National Unification: Ninth to Tenth Century

This stage corresponds to the period from the Buddhist persecution in 845 to the printing of Kaibao Canon 攳⮞啷 in 983. Before the persecution, the development of the canon was relatively stable. Both the main canon and the separate canons mentioned above were expanding in scale. However, the 845 persecution was a heavy blow to Chinese Buddhism. Almost all the scriptures and images were destroyed in most regions of the country. Afterward, Buddhism recovered gradually.

Temples all over the country used Zhisheng’s catalog as the standard to rebuild the canon for themselves or for local regions, which in fact promoted the unification of different editions of the canon and created opportunities and the social environment for the appearance of printed canons. However, such a unification based on Zhisheng’s catalog also resulted in a dilemma for including translated scriptures, such as the Tantric scriptures translated by Bukong ᶵ䨢 (705–774, Amoghavajra), because these were translated after Zhisheng’s catalog was written.

Another important factor that facilitated the unification of the canon was the appearance and spread of the court editions. These were compiled to accumulate merits for the imperial families. Relying on abundant human and material resources, the court editions were usually carefully copied, wellcollated, produced with fine paper, and well-made. Because they often contained the new scriptures translated in official translation bureaus, they were ranked highest in quality among all editions of the canon.

The imperial family also bestowed the canon to different regions. Therefore, compilers from different regions built up new canons or supplemented local canons according to the court edition. As a result, the court edition in fact helped to unify and regulate the Chinese Buddhist canon in different regions.

The intervention by the imperial court should be considered another important factor in the formation and unification of the canon. Before the Kaiyuan period (713–741), issues about how the canon was compiled or what kind of canon was going to be created were only relevant to local Buddhist groups and were never considered by the court.

Even though there was a court edition, the court was only responsible for funding it and left monks to compile it. The imperial family had little effect on the structure, the content, or the methods of compilation of the court editions. After the Kaiyuan period, Emperor Xuanzong 䌬⬿ (685–762) in the Tang dynasty prevented some scriptures from being included in the canon. This practice was adopted and strengthened by emperors from later generations. The scriptures translated by monks had to be approved by the court or they would not be collected into the canon.

Since the Qin and Han dynasties, China had been basically under unified political power, and emperors had the supreme authority. The imperial intervention in canon formation was, actually, a kind of guided control of political power over divine power. To perpetuate their own long-term stability, Chinese feudal rulers would never have approved the rise of any independent religious power; instead, they had to integrate religious power into their power structure and put it under their control. The imperial intervention in the contents of the canon was a sign that the political power controlled the development of Buddhism and made it part of the imperial bureaucratic system.

As for external markers of the canon, during this stage, the “Labeling Method by the Thousand Characters Classic” (Qianziwen zhihao fa ⋫⫿㔯塇嘇㱽) was adopted. This character-based call number system quickly replaced the storage-based method, verse-based method, and other methods. It spread further after the national unification of the canon and was adopted by the editions after Kaibao Canon. From another perspective, the appearance of the “Labeling Method by the Thousand Characters Classic” also helped to promote the progress of the national unification of the canon.

After the 845 persecution, different editions of the canon were gradually unified under the framework of Register of Canonical Texts in Zhisheng’s catalog, while editions from different places still showed differences. This situation was caused by regional variations of local Buddhism in different places, the distinction between doctrinal and faith-based devotional traditions, various needs for having a canon, the unstable nature of handwritten manuscript editions, and so on. Due to these reasons, various editions of Zhisheng’s catalog appeared. The situation was different from that before the 845 persecution during the Huichang 㚫㖴 reign, since the various editions were based on a standard checklist, namely Zhisheng’s catalog. Because of this variety, three canonical systems, representing the “Central Plain” region, the northern region, and the southern region, appeared in the block-printing period.

The Stage of the Coexistence of Handwritten and Block-Printed Copies: Tenth to Early Twelfth Century==

This stage corresponds to the period from the printing of Kaibao Canon to the end of the Northern Song dynasty (960–1127) , during which, even though block-printed copies of the canon first appeared, handwritten copies were still popular. As the block-printed copies became more numerous, the number of handwritten copies began to decline. Therefore, during this period, blockprinted and handwritten copies coexisted.

In China, this lasted from the printing of Kaibao Canon to the end of the Northern Song dynasty; in Japan, this period was much longer, lasting until the Edo period (1615–1868). We have found several handwritten editions from this period, such as the Jinsushan edition 慹䱇Ⱉ啷䴻, the Faxisi edition 㱽╄⮢啷䴻, and the Daheningguo edition ⣏␴䓗⚳啷䴻. In addition, there are many copies written in gold or silver ink. Extant handwritten manuscripts handed down from the Northern Song dynasty are now considered first-rate cultural relics for their fine paper and elegant handwriting. Thus it can be seen that the form and function of this type of canon tended to be more and more faith-based.

==The Stage of Pure Merit Accumulation: Early Twelfth to Early Twentieth Century

This stage roughly corresponds to the period from the Southern Song dynasty (1127–1279) to the Qing dynasty (1644–1911), during which block-printed editions of the canon completely replaced handwritten ones in China and became the major means of production and circulation. However, the handwritten editions did not disappear completely. Although their philosophical function declined, their faith-based function was highlighted. The handwritten editions in this period were mainly written in gold or silver ink. According to the research materials we have now, people did not stop making gold or silver-lettered canons until the Qing dynasty.

Nowadays, although no one tries to make any complete gold or silver-lettered copies of the canon, there are still some people making handwritten, gold- or silverlettered, and even blood-written copies of scriptures, in order to gain merit. All the above can be seen as evidence of purely merit-oriented devotion. In recent years, people have been making copies of scriptures for the sake of calligraphic demonstration, which is a sign of the popularity of Buddhist culture. We can expect that the art-oriented and merit-oriented copy-making will be long-standing traditions.

As discussed above, in the first four stages of the history of handwritten editions, there were only handwritten copies, while in the last two stages, editions of other forms coexisted with the handwritten ones. Our division of such stages can only provide an overview; the actual situation is much more complicated. In some cases, when a stage had already ended, the following one did not start immediately. Or sometimes two stages overlapped. The disparity among regions in China added much more complexity. Thus, this kind of periodization is only for convenience.

The very fact that handwritten canons were copied by particular groups of people gives rise to some basic features, that is, the uniqueness of each individual copy. This means that each copy of the canon or scripture copied by hand is the only one extant in the world. This contrasts sharply with blockprinted canons, for copies printed from the same set of blocks are totally the same. Thus, there must be differences in the copies of the same scripture that were hand-copied by different people, and copies of the same scripture copied by the same people at different times. This formal uncertainty—or, in other words, scribal/textual fluidity—is another feature of the handwritten canon. The combination of uniqueness and fluidity determines its basic characteristics, summarized in the following:

1. Differences in number of lines and number of characters in each line (xingkuan 埴 㫦 ), design of boundary lines (jielan 䓴 㪬 ), and calligraphic style;

2. Differences in scribal style such as redaction of the content and the use of different kinds of Chinese characters;

3. Differences in textual content of different editions due to addition, deletion, accidental omission, and scribal errors;

4. Differences in the division of fascicles and chapters due to the use of different master copies;

5. Differences in the scope of inclusion and structure of the canon due to regional, temporal, and personal variations in their creation.

Because of the uniqueness and fluidity of the handwritten canon, we have to seek commonalities while preserving minor differences by temporarily ignoring details and searching for a system of textual transmission based on the original master copies. Here we must establish the concept of “text lineage” (chuanben ⁛㛔, literarily “transmitted texts”). For every handwritten copy, whether it is a copy of scripture or a whole set of the canon, if it is not the original, it must have a master copy from which it was made. Both original copies and later copies form a system of text lineage. The handwritten copies that belong to the same text lineage are considered the same edition of the canon.

Then, the question is how to distinguish different systems of text lineage. Here I rely on the “three essential elements” of the canon proposed in 1988, namely selection criteria, structural system, and external marker, to solve the issue. These three elements focus on content, structure, and physical markers of the canon respectively. The internal characteristics of the canon can be presented in the aspects of content and structure, while the external ones are shown by physical markers. We can use these three elements to evaluate and examine any handwritten edition of the canon. If all the three elements are changed, the edition of the canon is changed and will be considered a new edition. If only external markers are changed, the canon is still considered unchanged. But if the content or structure is changed, the edition of the canon with the new internal elements will be considered a new one. Therefore, the decisive factors distinguishing the editions of the handwritten canon are their internal characteristics.

To study an edition of the handwritten canon, the most important procedure is to analyze its catalog, because the content and structure of that edition are presented in the catalog. No matter how different the handwritten editions’ external markers are, they are considered to belong to the same system of text lineage if they are based on the same catalog. Because of the uniqueness and fluidity of handwritten editions, even when there are subtle differences in structure or content between two sets of a canon, we still consider that they belong to the same system of text lineage.

For example, compared with Great Kaiyuan Canon 攳⃫⣏啷, the several sets of the canon on which Kehong’s Phonetic Glossaries of Buddhist Snjtras (Kehong yinyi ⎗ 㳒枛佑) is based, included editions of scriptures that were not in Zhisheng’s catalog (Kaiyuan lu). However, we still consider the canon Kehong worked on as a variation of Kaiyuan Canon. For another example, although a canon compiled at Longxing Monastery 漵冰⮢ in the Dunhuang region under Tibetan rule added several new scriptures not contained in Daoxuan’s catalog (Da Tang neidian lu ⣏Ⓒℏ℠抬), we still consider it to belong to the system of Daoxuan’s catalog.

Because of the role catalogs played in the study of manuscript canons, we must study various Buddhist catalogs in depth and among them identify the catalogs for the canon. As for the text lineages of manuscript scriptures or canons and their transformation, there are very few people who have done serious research, and it should become a focus in the future.

In order to determine if a copy of manuscript scripture belongs to a canon or not, we need to examine the physical copy by first checking if there are external markers such as the case number (zhihao 塇嘇). If there is a case number, this copy must belong to a canon. Then we can look at the colophon and see if it mentions whether the creation of the copy was for a canon or not. Finally, we can establish links within a group of copies through comparison and thus establish their identity.

The Block-Printing Period

More research is needed to determine when the earliest printed materials first appeared in Chinese history. The earliest block-printed canon was Kaibao Canon, printed in the early Song dynasty; the latest was Piling Canon 㭿昝啷, printed in the late Qing dynasty and early Republican China. During the 1,000 years in between, more than twenty editions of the block-printed canon were produced. However, despite the number and size of the court editions and private editions, the development of the Chinese Buddhist canon was declining along with Chinese Buddhism. New editions such as the First Supplement to Jiaxing Canon (Jiaxing Xuzang ▱冰临啷) and the Second Supplement to Jiaxing Canon (Jiaxing you xuzang ▱冰⍰临啷) collected more Chinese Buddhist works than ever before. During the block-printing stage, the overall structure of the canon did not change very much. In external form, the Chinese canon evolved from the scroll style (juanzhouzhuang ⌟庠

墅) to the accordion-folding style (jingzhezhuang 䴻㉀墅), and then to the stitched-booklet style (xianzhuang 䶂墅). Compared with the handwritten canon, the most important feature of the block-printed canon is its uniformity. The copies made from the same set of carved wood blocks have the same format and layout. Therefore, the engraved blocks on which the copies are based become the most persuasive basis for distinguishing different editions of the block-printed canon.

Once the blocks are engraved, the internal characteristics such as content and structure are fixed. Because of the enormous scope of a whole set of the canon, a considerable number of blocks were used in the printing process. For example, Kaibao Canon used 130,000 blocks. In order to manage them systematically and to identify each individual block, engravers created block numbers and carved them to create “external markers” of a canon on the blocks as well. Thus, the blocks, which reflect all “three elements of the canon,” become the basis for us to distinguish different editions of the block-printed canon, just as catalogs are the basis for the study of the handwritten canon.

We can therefore establish a principle that if the blocks are different, even though two sets of the canon were compiled under the same catalog, they are different editions. This is very important for clarifying the confusion of several editions of the block-printed canon. For example, there is a question about how many Liao dynasty (916-1125) texts recently found in a wooden pagoda in Ying County ㅱ䷋, Shanxi Province, belong to the canon. Although scholars have been debating it for a long time, there is still no consensus.

They did not realize that a common criterion of evaluation is needed. If we pay more attention to the blocks and compare them carefully, we can discover that only copy No. 7 belongs to the canon. Also, the Liao edition has a biggercharacter version and a small-character version. In addition, printed copies from Fangshan Stone Canon (Fangshan shijing ㇧Ⱉ䞛䴻) were produced during the Liao and Jin dynasties but were based on Liao Canon. Questions remain about how the three were related. After we consider blocks as the only criterion to distinguish different editions of the canon, we can declare confidently that the bigger-character edition, the small-character edition, and the Fangshan Stone Canon edition existed simultaneously but belong to different editions of the canon (Fang 2004). In a similar vein, although Zhaocheng Canon (Zhaochengzang 嵁❶啷) and the First Korean Canon (Chuke Gaoli zang ⇅⇣檀渿啷) were reprints of Kaibao Canon, these were all independent editions because they have their own blocks.

Analyzing the blocks for an edition not only covers details such as the size of the blocks, number of characters in a line and number of lines, boundary design, and block numbers, but also refers to the overall condition of a whole set of blocks. Many editions of the canon had been repaired and supplemented. For example, the blocks of Qisha Canon (Qisha zang 䢏䞪啷) were engraved in the Song dynasty, supplemented in the Yuan dynasty (1271–1368), and repaired in the Ming dynasty (1368–1644). Puning Canon (Puning 㘖⮏啷) was supplemented with scriptures of the esoteric tradition. Yongle Southern Canon (Yongle nanzang 㯠㦪⋿啷) was supplemented in the Wanli reign (1572–1620) in the Ming dynasty.

In addition, damage cannot be avoided, and blocks used for a long period need repairs. For example, Chongning Canon (Chongning zang ⲯ⮏啷) was repaired several times. When the Cultural Relics Publishing House (Wenwu chubanshe 㔯䈑↢䇰䣦) reprinted Qing Canon or Dragon Canon (Qing Long Zang 㶭漵啷) in recent years, original blocks were rearranged, supplemented, and repaired on a large scale. If the main body of a set of blocks is not changed but only repaired and partly supplemented, we consider it a new version of the same edition. That is to say, one edition of the canon may have different copies made from the same blocks and different versions caused by supplements and repairs. Thorough studies of different versions of one edition of the canon will be an important task in canon research. For example, it is significant to clarify the relationship among the original version, the supplemented version, and the repaired version of Zhaocheng Canon carved during the Jin dynasty. If the alleged Hongfa Canon (Hongfa zang ⻀㱽啷) is indeed a supplemented version of Zhaocheng Canon, it will not be considered an independent edition.

The Modern Printing Period

The modern printed editions can be classified into two types based on the method of production: typographic printed and photographic printed.

Typographic Printed Editions

There are two kinds of typographic printed canons—metal-type printed editions and laser-typeset printed editions. The metal-type editions are printed with movable metal type and conform to modern book standards. The editions of this type include the Kǁkyǁ edition (.ǁkyǁ]ǁ ⻀㔁啷), the Zǁkuzǁkyǁ edition (Dai Nihon Zǁkuzǁkyǁ ⣏ 㖍 㛔 临 啷 䴻 ), and the Taishǁ edition (Taishǁ]ǁ ⣏ 㬋 啷 ) in Japan. In China, there were the Pinjia edition (Pinjiazang 柣 ụ 啷 ) and the Puhui edition (Puhuizang 㘖 ㄏ 啷 ). Laser typesetting uses the technology of laser plate-making. In China, canons printed in this way include the Wenshu edition (Wenshu dazangjing 㔯㬲⣏啷 䴻), which was aborted halfway through, and the Foguang edition ἃ⃱⣏啷 䴻, which is in progress.

The modern printed editions have the advantages of sharp fonts, practical bookbinding style and layout, and a large amount of information. What deserves special mention here is that the appearance of the modern printed canon is connected to the rise of modern academic research on Buddhism. The Taishǁ edition, an example of the newly compiled canons, has high academic value not only in its collation and punctuation but also in its unique design of a classification system and its scientific and practical index. With such advantages, the modern printed editions replaced the block-printed ones as soon as they appeared. Although the level of technology used in metal-type printing and in lasertypeset printing is different, both require manual input and typesetting by computer. Because it is hard to avoid typos even with careful collation, there is no essential distinction between these two types of modern printing in terms of accuracy.

Photographic or Facsimile Editions

There are also two types of photographically printed editions. The first type does not make any changes to the original copy. The modern photographically printed versions of First Southern Canon (Chuke nanzang ⇅⇣⋿啷), Yongle Northern Canon (Yongle beizang 㯠㦪⊿啷), Qing Dragon Canon (Qing Long Zang 㶭漵啷), and Pinjia Canon are all of this type. The second type resets and reedits the original copies. Editions of the second type include the modern Taiwan edition and the Beijing edition of Chinese TripiΛaka (Zhonghua Dazangjing ᷕ厗⣏啷䴻). Now the ancient block-printed canons have become cultural relics and the photographic printed editions can present photocopies of the original canon. Although the photocopies are not the same as the original blocks, they are convenient for researchers studying ancient editions of the canon. The photographically printed copies usually employ modern bookbinding and layout, making reading more convenient. Therefore, the photographically printed copies are very popular. Determining the Independent Status of Modern Printed


The catalog on which a particular edition is based is the main criterion used to determine whether a modern printed edition is an independent one or not. If the catalog shows independent content and structure, we will consider the edition a new one. In addition, we have to focus on the historical transmission of that edition, namely, the original version and collated version.

According to these two criteria, all the typographically printed editions are new editions. Because the cost of typesetting is high and the process provides more room for rearranging the content, modern compilers take the opportunity to compile a new edition of the canon. In terms of text lineage, for example, the Taishǁ edition is very different from the Pinjia edition, mainly because they used different master copies for printing and different copies for collation.

The status of photographically printed editions is more complicated to determine. As mentioned above, there are two types of photographically printed editions. The first one does not make any changes to the original copies, while the other resets and reedits their content and structure. The first type follows the original catalog and keeps the original form, so we consider the photographically printed editions of this type the same edition as the original copy. The Qisha edition, which was photographically printed in the 1930s, is a special edition. Because the original copy was missing a number of volumes, some scriptures from other editions were added to this reprint, as explained by the compilers in the instructions and catalog attached to the end of this edition.

The photographically printed version of the Qisha edition did not cause confusion in the recognition of editions and did not change the arrangement of the original, reflecting the true face of the ancient blockprinted Qisha edition. So we still consider the photographically printed version the same edition as the original Qisha version. The supplemented part in the photographic printed version resembles the supplemented engravings in ancient block-printed editions.

The Beijing edition of Chinese TripiΛaka can be taken as an example of the other kind of modern printed canon that rearranges the order of the original content. Although the Beijing edition of Chinese TripiΛaka is based on Zhaocheng Canon, the whole collection of Chinese TripiΛaka contains about 10,000 fascicles of texts. The total number of texts in Zhaocheng Canon, however, accounts for half of the content in the Beijing edition of Chinese TripiΛaka. The rest includes texts taken from other editions of the canon and even texts input through computer.

Compared with Zhaocheng Canon, the structure and order of content of the Beijing edition were changed enormously, and it made many supplements and repairs to Zhaocheng Canon. What’s more, its catalog is different. Therefore, it is not a simple photographic version of Zhaocheng Canon but a new edition. (Some people still believe, inaccurately, that the Beijing edition of Chinese TripiΛaka is a photographic version of Zhaocheng Canon.) We have to admit the independent status of this edition, which inherited the contents of Zhaocheng Canon and eight other editions. So, when we evaluate the Beijing Chinese TripiΛaka, we cannot ignore its legacy from Zhaocheng Canon. In sum, in the study of modern printed editions, we have to take both their catalogs and their text lineages into consideration.

The Digital Period

With the rapid development of technology, the digital age of books has come, and the Chinese Buddhist canon entered the digital world in the 1980s. The digitization of the canon has greatly progressed in the past twenty years. The process can be divided into two stages.

Initial Stage

The major feature of the initial stage is “media transformation,” in which the paper medium of the Chinese Buddhist canon was transformed into digital data. Correspondingly, similar to the division of typographic and photographic printings, the digital canon in the initial stage also has two types: one is manual input, and the other OCR (Optical Character Recognition) input.

In the mid-1990s, TripiΛaka Koreana completed the digital transformation first. This success laid a solid foundation for future works. After the completion of the digital version of Taishǁ Canon, the digital transformation spread to Chinese Buddhist academic communities throughout the world. The major advantage of the digital canon is its full text retrieval system through which instantaneous retrieval, storage, and spread of information brings great convenience for researchers.

OCR scanning of a whole collection of the canon was first finished in the late 1990s. Now the scanned editions of the canon include Taishǁ Canon, Yongle Northern Canon, and Qing Dragon Canon. Although the scanned canons cannot realize the function of full-text retrieval, the information contained in paper copies that would take several bookshelves to store can be condensed on a small hard disk. Compared with photographically printed editions, the scanned ones indeed have more advantages. Completion of the scanning paved the way for the advanced stage of the digital period. Although the electronic texts created in this initial stage have many advantages, the deficiencies of the typographic and photographic printed versions still exist in them because they are basically simple transformations of media. In addition, there are two more problems in the electronic canon:

(1) Textual variations in different electronic versions

Early in the initial stage of the digital period, there was great enthusiasm for textual input of scriptures and the entire canon, as many people were working in this area. However, the quality of text input by different people or groups is different. Therefore, there might be several electronic versions of different quality for Taishǁ Canon. However, in the past seventeen years, after resource integration and competition, the version of Taishǁ Canon created by the Chinese Buddhist Electronic Text Association (CBETA) in Taiwan has been recognized by the public, while other versions have disappeared gradually.

The early electronic texts input by CBETA contain many mistakes, but the texts were gradually collated and corrected. Thus, early texts are different from later texts. Since the electronic texts created by CBETA were released gradually, different versions of the same electronic texts have been spread around the world. Due to lack of information about when the texts were input, ordinary readers may have difficulty distinguishing different versions.

(2) Public credibility of electronic texts

It can be either an advantage or a disadvantage that electronic texts are easy to revise. By revising, electronic texts may be perfected. However, revision can cause instability of these texts, negatively affecting their public credibility. The electronic version of Taishǁ Canon created by CBETA has been widely accepted among Buddhist researchers, despite its mistakes and typos. Because the electronic version has corrected many typographical errors in the original, we consider its quality better than or as good as the original Taishǁ Canon.

But when careful scholars search CBETA’s electronic version, they still have to check against the original Taishǁ Canon when they need to indicate the sources of quotations. Therefore, we have to determine how to establish public credibility of electronic texts. This is of course not a problem Buddhist scholars have to deal with alone. Rather, it is linked to public opinion about electronic texts in the whole society. In my opinion, one way to gain public trust in electronic texts of the Chinese Buddhist canon is to number the published texts, build databases for different versions of electronic texts, and keep records of revisions.

All electronic versions that appeared in the initial stage are simple transformations of certain original editions of the canon. Since they all depend on the original editions, they do not have independent status and do not belong to new editions of the canon. Catalog is still the criterion for distinguishing different editions.

The Advanced Stage

The major feature of the advanced stage of the digital canon is the hypertext link, namely, showing the hyperlinks of different data resources on the same screen in order to meet different readers’ demands. There are two levels of hyperlinks in the canon: the ordinary hypertext at the low level and the interactive hypertext at the high level.

One representative of the so-called ordinary hypertext is the Chinese Buddhist TripiΛaka Electronic Text Collection 暣⫸ἃ℠普ㆸ created by CBETA in April 2004. Its basic content includes the first fifty-five volumes and the eighty-fifth volume of Taishǁ Canon, and ten volumes under the category of historical biography in the Newly Compiled Japanese Supplemented Canon (Shinsan Dai-Nihon zokuzǁkyǁ 㕘个⣏㖍㛔临啷䴻).

The desktop structure of this collection is designed according to the principle of “combining texts and their commentaries together” (yishu lishu ẍ䔷晠㚠) first proposed by Yang Wenhui 㣲㔯㚫, thus changing the content and structure of the canon and providing new functions that the traditional canon does not have.

The so-called interactive hypertext is a new form of the canon based on active reader participation on the Internet (Fang 2002b). The future development of the digital canon remains to be seen. In my opinion, in the next few years or decades, a newly collated digital canon with the scanned images on the top and electronic texts below will appear; so will a hypertext canon with an interface that can accommodate multiple editions and allow switching back and forth among them. The new digital canon will not only be a simple database but also provide researchers with different kinds of research tools.

As new forms of the canon appear, the criteria for evaluating different editions in the digital era will change as well. They will be neither internal content and structure, nor external markers; the “three elements” of the Chinese Buddhist canon will be abandoned in the advanced stage of the digital period. The amount of information contained in the digital canon and the types, numbers, and functions of the research tools that the digital version can provide will be the criteria to evaluate different versions of the canon. This is a leap forward in quality.

In the digital era, the traditional paper-based canon will never disappear, and will develop toward the format of luxury bookbinding and layout to enhance its faith-based functions. The ideal canon should have three functional forms—the philosophy-oriented type, the faith-based type, and the research database type (Fang 1997). It is very difficult for the previous versions of the canons to achieve all three. The combination of digital and paper-based editions will realize this ideal.


Chen Shiqiang 昛⢓⻟, ed. 2000. Zhongguo fojiao baikequanshu: Jingdian juan ᷕ⚳ἃ㔁䘦䥹ℐ㚠烉䴻℠⌟. Shanghai: Shanghui guji chubanshe.
Chikusa Masaaki 䪢㱁晭䪈. 2000. Sǁ Gen ban Daizǁkyǁno keifu ⬳⃫䇰⣏ 啷䴻̯䲣嬄. In Chikusa Masaaki, 6ǁ Gen Bukkyǁ bunkashi kenkyǍ ⬳⃫ ἃ㔁㔯⊾⎚䞼䨞, 271–362.Tokyo: KyǍko Shoin.

Ciyi ヰ⿉, ed. [1979] 1993. Foguang dacidian ἃ⃱⣏录℠. Reprint, Beijing: Zhongguo shumu wenxian chubanshe.
Daizǁkai, ed. [1964] 1990. Daizǁkyǁ: seiritsu to hensen ⣏哝䳴烉ㆸ䩳̩⢱
怟. Reprint, Kyǁto: Hyakkaen.
Ding Fubao ᶩ䤷ᾅ. [1922] 1984. Foxue dacidian ἃ⬠⣏录℠. Reprint, Beijing: Wenwu chubanshe.
Fang Guangchang 㕡⺋拑. 1990. Guanyu Dunhuang yishu “Foshuo Foming jing” 斄㕤㔎䃴怢㚠˪ἃ婒ἃ⎵䴻˫. In Dunhuang Tulufan yanjiu lunwen ji 㔎 䃴 ⎸ 欗 䔒 ⬠ 䞼 䨞 婾 㔯 . Beijing: Hanyu dacidian chubanshe, 470–89. Reprint in Vol. 2 of Fang Guangchang, Dunhuangxue Fojiaoxue luncong 㔎䃴ἃ⬠婾⎊, 125–53. Hongkong: Zhongguo Fojiao wenhua chubanyouxiangongsi, 1998.

———. 1997. Lun dazangjing de sanzhong gongneng xingtai 婾⣏啷䴻䘬ᶱ 䧖≇傥⼊ン. Zongjiao zhexue ⬿㔁⒚⬠ 3, no. 2: 140–50.

———. 1998a. Fojiao zhi ἃ㔁⽿. Shanghai: Shanghai renmin chubanshe.

———. 1998b. Futujing kao ˪㴖Ⰸ䴻˫侫. Fa Yin 㱽枛, no. 6: 24–7.

———. 2002a. Ba zhi shi shiji Fojiao dazangjing shi ℓ—⋩ᶾ䲨ἃ㔁⣏啷 䴻⎚. Taibei: Foguangshan wenjiaojijinhui.

———. 2002b. Zixun shidai de Fojiao muluxue 屯妲㗪ẋ䘬ἃ㔁䚖抬⬠. Fojiao tushuguan guanxun ἃ㔁⚾㚠棐棐妲 29: 15–21.

———. 2004. Liao Dazi Zang de dingming yu cunben ˪怤⣏⫿啷˫䘬⭂⎵ 冯⬀㛔. Zhongguo xueshu ᷕ⚳⬠埻 2: 162–89.

———. 2006. Zhongguo xieben dazangjing ᷕ⚳⮓㛔⣏啷䴻䞼䨞. Shanghai:
Shanghai guji chubanshe.

———. 2007. Pilingzang chutan 㭿 昝 啷 ⇅ ㍊ . Paper delivered at the International Conference on the Chinese Buddhist Canon, 16–20 September, in Shanghai.
Ikeda On 㰈䓘㹓. 1990. ChǍgoku kodai shahon shikigo shǍroku ᷕ⚳⎌ẋ⮓ 㛔嬀婆普抬. Tǁkyǁ: ǀkura Shuppan.
Lan Jifu 啵⎱⭴, ed. 1994. Zhonghua fojiao baikequanshu ᷕ厗ἃ㔁䘦䥹ℐ
㚠. Tainan: Zhonghua fojiao baike wenxian jijinhui.
Li Fuhua 㛶⭴厗 and He Mei ỽ㠭. 2003. Hanwen fojiao dazangjing yanjiu 㻊㔯ἃ㔁⣏啷䴻䞼䨞. Beijing: Zongjiao wenhua chubanshe.
Mochizuki Shinkǁ 㛃 㚰 ᾉ Ṑ and Tsukamoto ZenryǍ ⠂ 㛔 ┬ 昮 . 1984. Mochizuki Bukkyǁ daijiten 㛃㚰ἃ㔁⣏录℠, expanded ed. Tǁkyǁ: Sekai Seiten Kankǁ Kyǁkai.
Ren Jiyu ả 两 グ , ed.1998. Zongjiao dacidian ⬿ 㔁 ⣏ 录 ℠ . Shanghai:
Shanghai cishu chubanshe.

———. 2002. Fojiao dacidian ἃ㔁⣏录℠. Nanjing: Jiangsu guji chubanshe.
Storch, Tanya. 2014. The History of Chinese Buddhist Bibliography: Censorship and Transformation of the Tripitaka. Amherst: Cambria.

———. 2016. Fei Chang Fang’s Records of the Three Treasures throughout the Successive Dynasties and its Role in the Formation of the Chinese Buddhist Canon. In Spreading Buddha’s Word in East Asia: Formation and Transformation of the Chinese Buddhist Canon, edited by Jiang Wu and Lucille Chia. The Sheng Yen Series in Chinese Buddhist Studies. New York: Columbia University Press.

Tsien Tsuen-hsuin. 1985. Paper and Printing. In Vol. 5, Part I of Science and Civilisation in China, edited by Joseph Needham.Cambridge: Cambridge University Press.

Tu Aming 㜄㬋㖶. 2016. The Creation of the CBETA Electronic Tripitaka Collection in Taiwan. In Spreading Buddha’s Word in East Asia: Formation and Transformation of the Chinese Buddhist Canon, edited by Jiang Wu and Lucille Chia. The Sheng Yen Series in Chinese Buddhist Studies. New York: Columbia University Press.

Zhongguo dabaike quanshu chubanshe ᷕ⚳⣏䘦䥹ℐ㚠↢䇰䣦, ed. 1988. Zhongguo dabaike quanshu: Zongjiao ᷕ⚳⣏䘦䥹ℐ㚠烉⬿㔁. Beijing:
Zhongguo dabaike quanshu chubanshe.