The Chen-yüan lu Catalog of Doubtful and Spurious Scriptures
Zhenyuan xinding shijiao mulu (Chen-yüan hsin-ting shih-chiao mu-lu)
貞元新定釋教目録by Yuanzhao 圓照T. #2157, 55.1015-1024.
Read at the IABS, Lausanne 1999
In recent years the study of Chinese Buddhist apocrypha has greatly increased our knowledge of indigenous developments in Buddhist thought and practice, typically changing our views of the social and cultural context of Chinese Buddhism as well. Although "apocrypha" has come to be the common designation for these materials, given the overly broad range of this term on the one hand (including in its scope scriptures long accepted as canonical such as the Ta-sheng ch'i-hsin lun 大乘起信論) and on the other hand the fact that many such scriptures never pretended to be translations from Indic originals in the first place (such as the works of the San-chieh movement) we need to focus on the original categories used by the Chinese themselves--where did the judgements of orthodoxy originate and on what were they based? This in turn centers attention on the numerous scripture catalogs compiled over the years and their categories of "doubtful scriptures" (i-ching 疑經) and "spurious scriptures" (wei-ching 僞經); this is particularly important since, having been branded doubtful or spurious by a renowned cataloger, a text would no longer be copied and included in collections and so it is often solely through the entries in the catalogs that we have any information at all about these materials. Among the many catalogs that contain lists of doubtful and spurious scriptures the Chen-yüan lu is of particular interest. This catalog was compiled by the noted monk Yüan-chao at imperial order in 800 in order to update the K'ai-yüan lu (compiled 730), the catalog which became the definitive standard throughout East Asia. Unfortunately, however, the Chen-yüan lu was not itself transmitted in China after the Sung dynasty, and the edition contained in the Taishō canon (based on the Korean text) includes numerous emendations and deletions by later hands. Shortly after its composition, however, the text was obtained by Kūkai and brought to Japan. Although many of the individual fascicles were reviewed years ago by Tsukamoto Zenryū, none of the manuscripts known at that time included fascicle twenty-eight, the catalog of doubtful and spurious scriptures. Recently, however, a complete copy of the Chen-yüan lu was discovered at the Nanatsu-dera temple in Nagoya, allowing a first-hand look at the contents of this important work and a comparison with the recension contained in the Taishō canon. My paper will introduce the Chen-yüan lu, the circumstances of its discovery at the Nanatsu-dera and its significance, and survey the contents of its catalog of doubtful and spurious texts.
1) Introduction
a) This is a work-in-progress report on a planned project to translate the catalog of suspicious and spurious texts of the Zhenyuan lu and give an annotated bibliography of the contents.
b) Why catalogs? The catalogs of scripture compiled in China over a period of 700 or so years reflect a nearly unique Chinese pre-occupation with bibliography and sacred text, and it was the catalogs that gave shape to the canon itself, deciding what would be included and what would not, as well as recording various information about the texts. Indeed, the catalogs serve all of us as the first stop in our studies of any particular text. Well, if the compilation of scripture catalogs is a quintessentially Chinese activity, the study of the catalogs themselves has been of particular interest to Japanese scholars. Hayashi Tomojirō, Mochizuki Shinko, and others—in part because of their own efforts at a modern understanding and definition of the meaning of “scripture” and “canon.” My own introduction to the study of catalogs came from Okabe Kazuo, who showed me the importance of understanding the motivations and circumstances of the Chinese catalogs in order to properly assess the status and history of Buddhist texts.
(1) So it is hard to give this paper—most folks don’t care (just wanting to get information on a particular text), and the detailed listing of the various divisions of texts, comparing the number of texts listed with the actual number recorded (counting, counting . . .) and the like is tedious to do and even more tedious to listen to—so I’ve tried to lighten this up a bit and will even show some slides in case anybody wants to slip out (or snooze) when the lights go out.
c) What is apocrypha?
i) The study of so-called “apocrypha” no longer in need of justification—among others, the work of Makita Tairyō, Kamata Shigeo, the work of Antonino Forte, and the volume edited by Robert Buswell have all shown us the great value that these texts have for understanding currents of thought and practice in China as well as the political and vacillating nature of the label “suspect” or “spurious.” And there is heuristic value in occasionally reflecting on the meaning of authenticity as well—I mean, what Buddhist scripture isn’t a fake in the sense of falsely purporting to be a literal rendering of buddhavacana? In any case, and without going into all of my reasons, I prefer to avoid the word “apocrypha” altogether, or to limit is to those texts labeled as “doubtful” (ijing疑経) or spurious (weijing 偽経) in the catalogs themselves.
(1) English term not suitable (“non-canonical but still edifying, i.e., not fakes
(2) All Buddhist texts are fakes—how to distinguish?
(3) “Pseudipigrapha”, or falsely ascribed, does not cover all of the texts considered ijing or weijing. (e.g., Sanjie materials, extracts,
ii) In any case, the importance and value of the study of texts labeled spurious has been well demonstrated. With regard the catalogs of spurious scriptures, most stop w/ Kaiyuan lu. This isn’t unreasonable, considering the overwhelming normative effect of the Kaiyuan lu on later catalogs as well the collection and organization of the canons themselves—that is to say, most of the later Chinese canons (and sometimes even the structure of the scriptures themselves) were based on this catalog, and its influence is seen to this day in the Taisho canon. So too, the study of spurious scriptures and their history in the catalogs that defined them as spurious typically stops with the definitive statements in the Kaiyuan lu, for after the stamp of fraud and duplicity was imprinted by this catalog the damning of a text was hard to overcome. Such a condemnation also usually resulted in the disappearance of the text, as texts branded spurious or not included in the canon (the so-called “不入蔵録) would not be copied when a temple was building up their library or a patron wished to sponsor the merit-making activity of sutra copying. Why, then, do we find texts branded “spurious” in 730, in China—that is, in the standard-setting Kaiyuan lu—still copied as part of the official canon in Heian and Kamakura Japan? The answer lies in the Zhenyuan lu, a catalog compiled some seventy years after the Kaiyuan lu.
2) The Zhenyuan lu
a) The compiling of the Zhenyuan lu
According to the introduction, the Zhenyuan lu was compiled at the order of Emperor Detsung (779-805) by the śramana Yuanzhao of the Ximing si 西明寺沙門圓昭in Zhenyuan 16 or 800. Compiled largely to augment the Kaiyuan lu, Yuanzhao included in his catalog the texts of the Kaiyuan lu as well as texts translated or composed over the seventy-odd years that had passed since the Kaiyuan lu had been written. Having said that, it is important to note that in both structure and content the Zhenyuan lu follows the Kaiyuan lu almost exactly, down to the preface, dedication, and much (if not most) of the entries and comments. Still, the Zhenyuan lu occupies an important place in the history of Chinese Buddhism, inasmuch as it was an official catalog compiled at imperial order, and the editor, Yuanzhao, was an important figure who moved in the elite Buddhist circles of his day and had participated in several translation projects as well, giving him firsthand knowledge of new scriptures. Many changes had taken place in the seventy years since the Kaiyuan lu was compiled, including the An Lu-shan rebellion (755), the inauguration of state controlled ordination, the advent of Chinese tantrism, as well as translations by Amoghavajra, Dharmachandra, Prajna, and others. In the 70 years since the compilation of the Kaiyuan lu, the Zhenyuan lu has added works by eleven new authors/translators, a total of 269 works in 341 juan.[1]
b) Zhenyuan lu not included in the Sung, Yuan, and Ming; only Koryo, hence Taisho
As I noted, the Kaiyuan lu became the standard for the collecting of the Chinese canons, and even before the printed edition of the Sung in the late 10th century (971-83, based on the Kaiyuan lu) the Zhenyuan lu seems to have been rendered virtually superfluous, for it wasn’t included in the Sung canon nor in the later Yuan or Ming canons. It was included in the Koryo canon, however, and, as is the case with the majority of the Taisho canon, because the texts of the Koryo canon became the basis for the Taisho editions the Zhenyuan lu is also included in the Taisho. As you all know, most texts in the Taisho use the editions in the Sung, Yuan, and Ming canons as prooftexts, but in the case of the Zhenyuan lu this was not possible since it wasn’t included in these canons. For anybody who looks closely at this catalog, however, this is a problem as there are many discrepancies and irregularities that indicate the Koryo text to have been re-edited at a later date, particularly with regard to the texts newly included in the catalog since the compiling of the Kaiyuan lu. For example, the in eight chüan (translated 810 by Prajna et. al.) were both translated a number of years after the 800 compilation date of the Zhenyuan lu, yet are included in the Koryo text.[2] In addition, the numbers given in the prefaces to the individual sections often do not match the number of texts actually listed in the catalog, and it is clear from comparison to the Kaiyuan lu that the reason for this is that texts have been added and deleted after the catalog was complete and the prefaces were written. Of course, it should be added that most catalogs show evidence of continual editing past their compilation date, as for example the Lidai sanbao ji, officially listed as compiled in 597, contains a record of the first suppression of the San-chieh movement three years later, in the year 600. It is not entirely clear to me yet the exact nature of these later emendations—that is, by whom or when they were made. We do have, however, variant manuscripts that appear to reflect an earlier state of the catalog.
c) The Japanese mss of the Zhenyuan lu
As Tsukamoto Zenryū once noted, Japan is a virtual repository of T’ang culture,[3] and the many embassies, pilgrims, and merchants of Nara and early Heian brought back huge troves of Buddhist materials, many of which still remain in temple storehouses and museums, their significance not always obvious. In 805, for example, just five years after Yuanzhao compiled the Zhenyuan lu, Kūkai traveled to Ch’ang-an and stayed at the Ximing-si, Yuanzhao’s temple, and the Zhenyuan lu is listed among the texts that he brought back to Japan.[4] If, as seems evident, the Koryo edition is a later, significantly emended, recension of the Zhenyuan lu we may expect this text, perhaps, to more accurately reflect its original content.[5]
Numerous fragments of the Zhenyuan lu have been found in Japan, and many years ago both Tsukamoto Zenryū and Yabuki Keiki recognized the importance of these Japanese manuscripts, for they clearly revealed that the Koryo edition used for the Taisho canon contained numerous additions and deletions and, they surmised, reflected a much later edition of the catalog than that originally disseminated under imperial auspices in the year 800 and subsequently transmitted to Japan by Kūkai and others.
d) Importance of the Zhenyuan lu in Japan: many canons used it for the basis of their own collections.
This was an important recognition not only because it allows us to better understand the original form of the Zhenyuan lu, but also because the Zhenyuan lu—as transmitted in Japan—rather than the Kaiyuan lu seems to have been regularly used as the basis for a temple’s library or the copying or collecting of the entire canon, as was the case, for example, with the canon of Nanatsu-dera, copied at the very end of the Heian period and the repository of our elusive 28th chuan.
e) Some background and slides of Nanatsu-dera
i) The Nanatsu-dera is a temple in downtown Nagoya whose founding is attributed to Gyoki-bosatsu in the Nara period (735). Formerly a vast complex, in typical fashion it had been destroyed in fires over the years, and in post-war Japan most of its lands and buildings have been lost. Fortunately, however, its complete collection of the canon, a manuscript copy dating to the late Heian period (1176-1180, “to comply with a vow made by Yasunaga, Ason of the Ōnakatomi, Deputy Governor of Owara”), has survived. Although this canon had received attention earlier—both the canon and the “karabitsu 唐櫃,” the T’ang style chests that they were stored in, had been declared a national treasure in 1900 and a catalog of the contents published in 1968— the compilers of the catalog were not in a position to judge the content of the texts, and the scholars that were so qualified believed the Nanatsu-dera canon to be a manuscript copy of the printed Northern Song canon (completed 983) an edition that had considerable influence on all subsequent printings, including, though indirectly, the Taishō Shinshū Daizōkyō. Therefore it was not thought likely to contain any surprises. It was only about ten years ago when Professor Ochiai Toshinori realized that a number of the Nanatsu-dera texts were not based on the Song editions but rather on manuscripts of the Nara period, firing enthusiasm among scholars that the Nanatsu-dera canon might contain variants no longer transmitted after the composition of the standard-setting Kaiyuan lu and the printed Song canon. Indeed, their expectations were rewarded as they carried out a number of surveys of the texts and discovered numerous lost and hitherto unknown texts, including the only extant work attributed to Kumārajīva that is not a translation (other than his correspondence with Huiyuan), and numerous non-canonical texts—including the Sanjiefofa, Foming jing, the Piluo sanmei jing, and more. In an article published shortly after the surveys began, Professor Ochiai disclosed an amazing fifteen newly discovered texts, some of which were thought to exist no longer and some of which had never even been recorded before. Any one of these texts would have been a great find, but to encounter so many at one place and in relatively good condition has led scholars to compare the find as second only to Tun-huang in importance for the study of Chinese apocrypha, and Makita Tairyō has stated that he feels a need to completely update his epochal Gikyō no kenkyū in light of the Nanatsu-dera discoveries.
ii) Just for a visual break, let me show a few slides of Nanatsu-dera.
iii) The Zhenyuan lu as the basis for the Nanatsu-dera canon
To return to the Zhenyuan lu, I noted earlier that one aspect of its value is that apparently a number of different collections in Japan were based on it rather than the Kaiyuan lu, as is more commonly the case. The Tōji canon, for example—an important collection dating from the late Heian, is based on the Zhenyuan lu, and a late Nara period project to supply the complete canon to the kokubunji throughout the country likewise used the Zhenyuan lu as the basis for the collections. And, as I mentioned, the Nanatsu-dera canon, closely related to the Tōji canon, is also based on the Zhenyuan lu.[6] In fact the canonical catalog of the Zhenyuan lu is largely copied on the inside covers of the chests containing the texts—and because the Zhenyuan lu had moved the entire collection of thirty-five San-chieh texts from the spurious catalog and included them within the canonical section – a fact that I will return to in a moment—the initial researchers believed that perhaps at one time all of the Sanjie materials had been extant in Japan (I doubt if this was ever the case—rather, the editors of the canon were simply being scrupulous in copying the Zhenyuan lu catalog[7]). In any case, this indicates the importance of the Zhenyuan lu in studying the history of the Buddhist canon.
iv) (Parenthetically, the Nanatsu-dera discovery is only one of many exciting new manuscript caches to be receiving attention these days—the convent project of Professor Barbara Ruch, there are rumors of other Tun-huang collections in Japan, and issaikyō in other temples too promise to yield new discoveries).
3) Comparison with Kai-yuan lu and Koryo ms
An ongoing task that will only be completed as the project moves forward, but a few examples of the differences:
a) Number of texts listed and actually contained
I mentioned that the Zhenyuan lu follows the Kaiyuan lu almost exactly, and the same is true of the catalog of “doubtful and spurious scriptures.” Both the Koryo and the Nanatsu-dera mss of this catalog, contained in juan 28 of the Zhenyuan lu, claim to list 393 works in 1491 juan among the spurious texts, as compared to the Kaiyuan lu’s record of 392 works in 1055 juan. The task of comparing the entries in the two catalogs is not easy, but generally speaking and aside from minor differences in the order or titles of certain works, the two catalogs match pretty well, with several major exceptions. One difference, for example, is that the sixteen juan text of the Foming jing is listed as apocryphal in the Kaiyuan lu and condemned as filled with absurdities and confusing coarse and sacred language. In the Zhenyuan lu, however, the 16 juan Foming jing is recorded in the catalog of canonical works, with a note to the effect that it was put there by Imperial order on the 23rd day of the 10th month of the fifteenth year of the Zhenyuan era (???, 799). As an aside, the sixteen juan Foming jing is another text lost in its native land of China but that was transmitted to Japan and among those texts discovered at the Nanatsu-dera.
b) “Extra-canonical Texts”
Altogether there are 48 titles included in the Kaiyuan lu “catalog of scriptures not included in the canon” that have been discovered at Nanatsu-dera, many of which had been long lost in China, including texts of the San-chieh movement (Ochia, “Newly Found Texts,” 20-21, 48)—why were these texts copied and included as part of the canon? Let me finally turn, then, to a look at how the San-chieh materials were treated in the Zhenyuan lu.
c) San-chieh stuff
A second significant difference, and the one that initially generated my interest in the Zhenyuan lu, is the strange treatment of the San-chieh texts in the Taisho edition. First of all it is important to note that the author, Yuanzhao, was possibly a follower of the San‑chieh or at least held Xinxing, the founder, in great esteem, as witnessed by the fact that he composed at least one (and possibly two) memorial steles in Xinxing's honor. Be that as it may, the Koryo edition has an entry in the chronological catalogue (juan 10) that states: "Sramana Xinxing. Thirty-five works in forty-four juan, (Chi‑lu), as explained below in juan 28 and 30." The first unusual thing about this entry is that it says that the texts are explained in both the apocryphal (juan 28) and the canonical catalogue (juan 30). Secondly, where the texts attributed to all of the other monks whose names are included in this list are then, in order, described more fully in the prose section which follows, there is no further mention of Xinxing. Again, when one turns to the 28th juan (catalogue of apocryphal works) or the 30th juan (catalogue of the Tripitika) there is no listing of Xinxing's works or any of the San-chieh texts. As with the chronological catalogue (juan 10), however, it is apparent that the records concerning the San‑chieh materials have been removed from the 28th juan of the Koryo edition.
First, when we compare the number of texts supposed to be included in the 28th juan, 393, and the number of texts which are actually included, 357, we can see that there are thirty-six texts missing. If one compares the entries in the Kaiyuan lu and the Zhenyuan lu it is seen that these missing texts correspond to the Foming jing and the thirty-five works of the San-chieh. Now, the Koryo version includes the Foming jing in the catalogue of the canon with a note to the effect that it was put there by Imperial order in Chen‑yuan 15, 10th month, 23rd day (???? 799), but there is no further mention of the San‑chieh texts.
In the Nanatsu-dera and other Japanese recensions of the Zhenyuan lu, however, in juan 30, the catalogue of the canon (of the "Writings of the Sages") there is the following entry: "The catalogue says 'The Collected Records of the Three Levels, thirty-five works in forty-four juan. Compiled by the Sramana Xinxing of the Sui dynasty." There is also a note under the entry for the Sanjie fofa which states that none of these works had been seen in the Kaiyuan lu catalogue of canonical texts, naturally enough as this catalog includes the San-chieh materials solely in the spurious section. In any case, the Zhenyuan lu lists the complete thirty-five works, together with the number of pages of each work, the usual custom with text recorded as part of the extant canon, and an indication that Yuanzhao actually saw the texts. After the last work recorded, there is a note stating that these texts were ordered into the Tripitika in Chen‑yuan 16, 4th month, 18th day (date??—13th or 18th??-- 800).
Now, although this in itself is significant, perhaps even more interesting is that in spite of having been included in the canonical catalog, Yuanzhao also left the record of the Sanjie texts in juan 28, the catalog of spurious scriptures. As with the record in the canonical catalog, he appended a note about the imperial order to include them in the canon, as well as a record clearly testifying to the continued popularity of the Sanjie movement, noting that the movement had continued for more than 200 years and that in the capital there were fifty-five temples with Sanjie meditation halls and over 2,000 monks and nuns. It then states that the "total number or works in the catalogue of Hinayâna works and the Writings of the Sages comes to 475," as opposed to 438 in the Kaiyuan lu and 442 in the Koryo version of the Zhenyuan lu. The Koryo version has added four texts to the catalogue of the Writings of the Sages that had been newly composed or put into the canon since the writing of the Kaiyuan lu, and the Nanatsu-dera recension added two of these newly compiled works plus the thirty-five works of the San chieh‑chiao, thus the totals all agree.
Many questions are left unsolved—why did Yuanzhao include the Sanjie materials in both the spurious and the canonical catalogs? Given his reliance to the Kaiyuan lu, it seems to me that this is another example of fidelity to the catalog, underscoring once again their importance in the study of Chinese Buddhist texts. By whom and when the Sanjie materials (and other texts) were removed, however, is another question that needs more research. In any case, my point is not the history or status of the Sanjie texts but rather to highlight the value of the Zhenyuan lu for our study of Chinese and, as we can see in the case of the Japanese recensions of the Zhenyuan lu, Japanese Buddhism. Although I haven’t finished my survey, because of such significant differences among the Kaiyuan lu and Zhenyuan lu catalogs of spurious texts and the use of the Zhenyuan lu in Japan I think that we can already see that it deserves more attention.
[1] T 2157, 55.771a; these numbers are complicated, varying among editions and the actual numbers do not necessarily agree with the numbers given in the text itself; indeed, these discrepancies are among the clues that alert us to the fact that this catalog underwent a process of updating for some time (cf. Bussho kaisetsu daijiten, vol. ??, p. 49).
[2] Tsukamoto Zenryū, “Kanbon Jōganroku (Koryozobon) to Nippon no koshahon Jōganroku,” included in Zensho, 380.
[3] Tsukamoto, “Kanbon,” 373.
[4] T #2161, 55.1064a.SS
[5] A copy of the texts brought back by Kūkai was made at Tōji in the mid-1300’s and may contain a copy of the Zhenyuan lu (Tsukamoto, 377); so too with a copy of the Zhenyuan lu made at the Ninna-ji in the late 1400’s (Tsukamoto, 378).
[6] For a discussion of how catalogs were used in Heian Japan see Ochiai Toshinori, “Heianjidai ni okeru nyuzōroku to shōsoroku ni tsuite,” 475.
[7] Dôchû (d. 1281) states in his Tanyôki that he has seen the Chen yüan hsin ting shih chiao mu lu record of thirty-five works in forty-four juan, and quotes the note from the Ryûkoku MS concerning their inclusion in the catalogue. However, Dôchû also states that he has only been able to actually examine two of the thirty-five works, the San chieh chi lu in four juan, and the Fa chiai chung sheng gen chi ch'ien shen fa in one juan (cf. K'ai yüan shi chiao lu nos. 1 and 10‑11). Another record, the Kôzanji Engi, records that of the entire canon as recorded in the Chen yüan hsin ting shih chiao mu lu, the San‑chieh‑chiao materials in forty-four juan are missing.