Fei Changfang’s Treatment of Sengyou’s Anonymous Texts

Fei Changfang’s Treatment of Sengyou’s Anonymous Texts

Michael Radich heidelbeRg UniveRsity

The catalogue entitled Lidai sanbao ji T2034, completed in 598, exerted a profound influence on the shape of the Chinese Buddhist canon. It also features a large number of new ascriptions for canonical texts. Subsequent tradition and modern scholarship have repeatedly found these ascriptions profoundly problematic. Scholarly opinion has been divided about whether the author, Fei Changfang, faithfully reported these suspect ascriptions from other sources (subsequently lost), or was himself their originator. This study analyzes three highly suspect patterns in the treatment of new ascriptions in the Lidai sanbao ji: 1) New ascriptions to the same supposed translator are arbitrarily assigned in batches to titles appearing in delimited, contiguous portions of much longer lists of anonymous texts in the earlier Chu sanzang ji ji. 2) Contradictory information is repeatedly given about the same titles in different parts of the work. 3) Information about titles affected by these two problems is ascribed to an implausibly wide range of earlier catalogues. These three patterns affect numerous separate portions of the Lidai sanbao ji. The Lidai sanbao ji, and not its putative earlier sources, is therefore the common denominator―the bottleneck where these problems collect. The most economical and plausible interpretation is that Fei himself was falsifying his information, whether deliberately or by a rather extreme scholarly negligence. These findings make it all the more imperative to critically evaluate all ascriptions appearing for the first time in the Lidai sanbao ji, including a large number of ascriptions still followed in the canon as it is used today.

intRodUction

Five extant catalogues of Chinese Buddhist texts date before the Tang dynasty: Dao’an’s 道安 (312/314–385) Zongli zhongjing mulu 綜理衆經目錄, completed in the early 380s; 1 Sengyou’s 僧祐 (445–518) Chu sanzang ji ji 出三藏記集 T2145 (CSZJJ), completed in 515; Fajing’s 法經 (d.u.) Zhongjing mulu 眾經目錄 T2146, completed in 594; Fei Changfang/Zhangfang’s 費長房 (d.u., fl. ca. 562–598 2) Lidai sanbao ji 歷代三寶紀 T2034 (LDSBJ), completed in 598; 3 and Yancong’s 彥琮 (557–610) Zhongjing mulu 眾經目錄 T2147, Author’s note: I am most grateful to Profs. Jan Nattier and Funayama Tōru for encouragement and advice as I prepared the present paper. I also thank two anonymous reviewers for JAOS, whose comments improved the paper considerably, and my research assistant, Dr. Atsushi Iseki.

1. Originally lost, but reconstructed almost in its entirety by modern scholars on the basis of citations in CSZJJ.

2. The dates of Fei’s birth and death are unknown. The last date we have for his activity is 598, the date upon which the LDSBJ was submitted to the throne. We know he was active by 562 because he reports that he witnessed in Chengdu 成都 the self-immolation of Sengya 僧崖 (488–562, on whom see James Benn, Burning for the Buddha: Self-Immolation in Chinese Buddhism [[[Honolulu]]: Univ. of Hawai’i Press, 2007], 209–12); T2034 (XLIX) 101a14–15; Ōuchi Fumio 大内文雄, Nanboku chō Zui Tō ki Bukkyōshi kenkyū 南北朝隋唐期佛教史研究 (Kyoto: Hōzōkan, 2013), 73.

3. The LDSBJ is commonly dated in the literature to 597, but as Palumbo has pointed out, the exact date upon which Fei submitted it to the throne (Kaihuang 開皇 17.12.23) corresponds to February 4, 598; see Antonello Palumbo, review of Storch, The History of Chinese Buddhist Bibliography (H-Buddhism, H-Net Reviews, April

completed in 602. Of these, the catalogues of Dao’an, Sengyou, and Fei Changfang exerted the greatest influence on the shape of the canon as we have received it, to a degree matched only by the later work of Zhisheng 智昇 (669–740) in his Kaiyuan Shijiao lu 開元釋教錄 T2154. However, it is well known (and should be known even better) that the fourth of these extant pre-Tang catalogues, Fei’s LDSBJ, is a source of numerous problematic ascriptions and dates that are still carried by texts in modern editions of the Chinese Buddhist canon. Time and again, Fei applied new ascriptions to numerous texts; and time and again, modern scholarship has found that those ascriptions are baseless and misleading. The LDSBJ was already subject to criticism by medieval Chinese bibliographers—especially by Zhisheng. In the modern era, critical studies of the received canon, and the ascriptions it contains, have devoted immense energy and space to general problems with Fei’s sources and methods and specific reassessments of individual ascriptions—for example, suggesting that some of his supposed sources may never have existed, or may have been forgeries or otherwise unreliable; or showing, where Fei’s information can be checked against other sources, that he contradicts those sources. Notwithstanding these many treacherous shoals lurking within his work, many of Fei’s new ascriptions and dates were accepted by subsequent cataloguers, and the canon as it is used today still bears the deep imprint of his influence. It therefore remains as urgent as ever to exercise the sharpest critical awareness that we can bring to bear on the ascriptions we have inherited from Fei’s work as a cataloguer.

This paper presents evidence of widespread and troubling patterns in the assignment of new ascriptions in the LDSBJ, and aims thereby to shed new light on Fei Changfang’s working method. The findings presented here, I contend, have important implications for how we should assess and use the information in the LDSBJ. Ascriptions usually dictate the dating of our texts, and canonical texts are still, very often, the richest and most basic evidence we draw upon for the study of many historical questions. Our assessment of Fei’s ascriptions therefore has potentially fundamental and far-reaching implications. I first discovered the findings presented here independently and, in an earlier draft of this paper, wrote as if they were entirely new. Subsequently, I discovered at the eleventh hour that more than eighty years ago, Sakaino Kōyō 境野黄洋 (1871–1933) had partially stolen my thunder—he had already clearly observed and sharply criticized the first of the two broad patterns in the LDSBJ studied below. 7 The credit for the original discovery of that aspect of Fei’s working style, as I analyze it below, should therefore go to Sakaino.

By rights, Sakaino’s findings should have immediately become common knowledge and sufficed to thoroughly discredit the LDSBJ as a source of ascriptions. This in turn should have forced a much more profoundly skeptical and thorough reconsideration of all received pre-Tang canonical ascriptions, many of which are fundamentally skewed by Fei Changfang’s work. Sakaino was himself so horrified by what he had found that he was moved to quite immoderate language. He says, for instance, that Fei’s “carelessness is almost beyond words”; and resulting patterns of new ascription in the LDSBJ are “so peculiar as to defy comprehension”; he complains bitterly that scholars have for centuries placed implicit trust in Fei, despite the actual enormity of his “misdeeds”; and he even declares, “Fei Changfang’s behavior is so problematic that it demands psychiatric examination.”

However, so far as I am aware, Sakaino’s discovery in fact made barely any dent in scholarly awareness. His work on this score is cited very rarely, if ever. I have seen no reference to his important work on the precise problem at issue here, even in works directly on Fei Changfang and the LDSBJ. This state of affairs surely betrays significant problems of conservatism and imperfect communications in Buddhological scholarship. In presenting my own findings here, which anatomize the problem more systematically and in greater detail than Sakaino, in a venue more visible to the present generation of scholarship, I hope that it will be possible to increase awareness of the large pattern of problems that Sakaino found so troubling.

the pRobleM (1): assignMent of new ascRiptions in aRbitRaRy batches

Nearly a century before the LDSBJ, Sengyou compiled his CSZJJ, which is our earliest extant catalogue of Chinese Buddhist translations, and justly famed as our most reliable source of information about Chinese Buddhist texts down to his time. In fascicle 4 of the CSZJJ, Sengyou supplements the list of anonymous scriptures he had already inherited from Dao’an’s earlier catalogue by presenting a list of 1,306 titles of further texts that he also regarded as anonymous—that is to say (and this is key), on the basis of the information available to him, he considered it impossible to say who had translated those texts. 9 (In the following, this list will be called “Sengyou’s list of anonymous scriptures” or just “Sengyou’s

7. Sakaino Kōyō, Shina Bukkyō seishi 支那佛教精史 (Tokyo: Sakaino Kōyō Hakushi Ikō Kankōkai, 1935). Sakaino clearly characterizes in general terms the problem of batch-wise assignment of new ascriptions to titles from contiguous portions of Sengyou’s lists, mentioning, as corpora to which it applies, ascriptions to An Shigao, Tanwulan, Faju, and Nie Daozhen (pp. 80–86). He points out that Sengyou appears to organize his lists of anonymous texts by topic, as indicated by titles (pp. 81–82); and that Fei Changfang then assigns these titles to purported translators in groups, with the effect (which Sakaino finds comical) that individual translators take on the appearance of specialists in texts on particular topics, like Brahmins or hells (p. 82). As a particularly telling instance of the damage this can do to historical understanding, Sakaino points out that it had the side effect of making scholars believe that Tanwulan was a key figure in the early transmission of proto-Tantric Buddhism to China, against all likelihood of historical veracity (pp. 82–83). Sakaino also discusses the impact of this same pattern on new ascriptions to An Shigao (pp. 80–86); Shengjian (pp. 96–98); Zhi Qian (pp. 141–45); Faju (pp. 157–59); Dharmarakṣa (pp. 177–81); Nie Daozhen (pp. 200–206); Tanwulan (pp. 211–20); and *Gītamitra [?] 祇多羅/祇多蜜 (pp. 261– 65). See also sections on the corpora of Juqu Jingsheng (pp. 867–72, esp. 867); Guṇabhadra (pp. 634–38).

8. 其の杜撰殆ど言語の外と言ふべきである, Sakaino, Shina Bukkyō seishi, 159; 實に考えられない不思議のこと, 143; しかも古來の學者が深く之を信じ、多く之によつて譯者名を定めるなど、立派な根據として採用して來たことを思ふと、長房の罪もまた大なりと言わざるを得ない, 81; 費長房の行為については、精神の鑑定を要する程の問題である, 81. 9. Sengyou’s list would reward a study in its own right for what it reveals about several types of text in circulation in his time, the probable processes by which they were formed, and the concerns of Buddhists producing and using those texts.

list.”) Of the 1,306 texts listed by Sengyou, only 193 (14.8%) can be identified with extant texts with considerable confidence. 10 Of those 193 texts, however, only 30 (15.5%) are regarded as anonymous in the present canon, which means that the remaining 163 (84.5%) have acquired attributions since the time of Sengyou.

Only one surviving catalogue was produced between the CSZJJ and the LDSBJ—Fajing’s Zhongjing mulu of 594. In Fajing’s catalogue, interlinear notes give new ascriptions to 32 of the surviving texts from Sengyou’s list. Ordinarily, this would mean that these ascriptions, to our knowledge, date to Fajing. 11 Subtracting these 32 texts for which Fajing gives ascriptions from the 193 extant texts on Sengyou’s original list, there remain 161 titles for which Fajing gives no attribution. However, 8 of Sengyou’s titles do not seem to appear in Fajing, or at least, cannot be identified confidently with a title in Fajing. 12 This still leaves 153 titles that are clearly treated as anonymous by Fajing, like Sengyou before him.

We see a striking contrast in the treatment of Sengyou’s list in the LDSBJ. As I show in Appendix 1, Fei Changfang gives new ascriptions (that is, ascriptions not found in Fajing either) for 129 of the texts in Sengyou’s list that are still extant.

When we expand our scope even further, beyond only extant texts, Fei claims to have found information about the ascriptions and dates for a huge number of previously unattributed texts, and we see a conspicuous and telling pattern: new LDSBJ ascriptions to a single translator or atelier cluster in a way that mysteriously and implausibly parallels portions of Sengyou’s list. For example, the LDSBJ ascribes a total of 54 titles to Nie Daozhen 聶道真, citing as his basis, in the general note at the end of the list of ascriptions, a/the bie lu 別錄 (in interlinear notes on specific titles in this list, Fei cites no other specific sources). Of these titles, the majority—a total of 41 texts—also appear in a short stretch of Sengyou’s list only 65 texts long (among a total of 1,306 titles), though in a different order.

To give the reader a concrete idea of what this pattern of correspondence looks like, I here list the 54 titles ascribed to Nie Daozhen in the LDSBJ, followed by a number indicating the position of the same title in Sengyou’s CSZJJ fascicle 4 (where the 41 titles in question fall between #63 and #128 in the overall sequence of 1,306 texts). 13

1. 十住經十二卷 (no correspondence) 2. 諸佛要集經二卷 (no correspondence) 3. 觀世音授記經一卷 (第二出) = CSZJJ #63 4. 寂音菩薩願經一卷 (no correspondence) 5. 大光明菩薩百四十八願經一卷 CSZJJ #75 6. 文殊師利般涅槃經一卷 CSZJJ #70 7. 師子步雷菩薩問發心經一卷 (或云問文殊師利成佛發心經) CSZJJ #82 8. 大雲密藏問大海三昧經一卷 CSZJJ #79 9. 溥首童真經一卷 CSZJJ #71 10. 寂音菩薩問五濁經一卷 CSZJJ #89 11. 無言菩薩流通法經一卷 (出大集) CSZJJ #90

10. Five titles present special problems, since they appear twice in CSZJJ, once in Sengyou’s list of anonymous texts, and once with an attribution. For the purposes of the present analysis, I have excluded those texts. For details on those texts and associated problems, see Appendix 3. 11. Whether or not we regard these ascriptions as first appearing in Fajing’s catalogue naturally depends upon our assessment of the status and date of the interlinear notes in Fajing. There are possible problems with these notes, which I hope to address in future research. See also below n. 55. For the present, however, I have treated these interlinear notes as an integral part of Fajing’s original catalogue.

12. T17, T33, T82, T91, T129, T211, T551, T1351. 13. T2034 (XLIX) 65c3–66a26; T2145 (LV) 22b16–23b1. 12. 菩薩戒要義經一卷 (出菩薩戒經) CSZJJ #104 13. 菩薩呵睡眠經一卷 CSZJJ #127 14. 菩薩呵家過經一卷 CSZJJ #126 15. 菩薩如意神通經一卷 CSZJJ #119 16. 菩薩苦行經一卷CSZJJ #116 = 菩薩諸苦行經 17. 菩薩宿命經一卷 CSZJJ #117 18. 菩薩受齋經一卷CSZJJ #122 19. 菩薩道示行經一卷 (no correspondence) 20. 菩薩求佛本業經一卷 CSZJJ #110 21. 菩薩奉施詣塔作願念經一卷 CSZJJ #118 = 菩薩奉施諸 [v.l. 詣 SYM] 塔作願念經 22. 菩薩本願行品經一卷 CSZJJ #113 23. 菩薩求五眼法經一卷 CSZJJ #125 24. 菩薩出要行無礙法門經一卷 CSZJJ #114 25. 菩薩初發心時經一卷 CSZJJ #109 = 菩薩初發心經 26. 大方廣菩薩十地經一卷 (第二出。與法護譯大同小異) CSZJJ #96 27. 菩薩戒身自在經一卷 (或云自在王菩薩問如來警戒經) CSZJJ #104 = 菩薩戒自在經 28. 菩薩三法經一卷 CSZJJ #98 29. 無言菩薩經一卷(第二出) CSZJJ #90 30. 菩薩道行六法經一卷 (no correspondence) (see CSZJJ #100 = 菩薩六法行經, but this appears below)

31. 三曼陀跋陀羅菩薩經一卷 CSZJJ #88 32. 無垢施菩薩分別應報經一卷 (即是異出離垢地經亦云應辯經) CSZJJ #86 33. 菩薩初地經一卷 [+(亦云應辯經) SYM], see CSZJJ #101 菩薩生地經 34. 儒童菩薩經一卷 (出六度集) CSZJJ #91 35. 菩薩十道地經一卷 CSZJJ #95 36. 光味菩薩造七寶梯經一卷 (出太 [v.l. 大 SYMP] 集) = CSZJJ #87 光昧菩薩造七寶梯經一卷 (抄方等大集經) 37. 菩薩緣身五十事經一卷 (與五十緣行經大同小異) CSZJJ #97 38. 菩薩戒自在經一卷 CSZJJ #103 39. 菩薩十法住經一卷 CSZJJ #94 40. 波斯匿王欲伐鴦掘魔羅經一卷 [CSZJJ #288] 41. 轉輪聖王七寶具足經一卷 [CSZJJ #275] 42. 轉輪聖王發心求淨土經一卷 [CSZJJ #276] 14 43. 文殊師利與離意女論義極似維摩經一卷 (no correspondence) 44. 文殊師利淨律經一卷 (第二出。與法護譯小異) (no correspondence) 15 45. 初發意菩薩行易行法經一卷 (出十住論) CSZJJ #128 46. 菩薩布施懺悔法一卷 (出決定毘尼) CSZJJ #123

14. These three titles (nos. 40–42 in the above list), appearing immediately after one another in the LDSBJ, appear in close proximity to one another elsewhere in Sengyou’s anonymous list: 波斯匿王欲伐鴦掘魔羅經 (#288), 轉輪聖王七寶具足經 (#275), 轉輪聖王發心求淨土經 (#276); LDSBJ T2034 (XLIX) 66a10–12, in CSZJJ T2145 (LV) 25a24, 25a11–12 respectively.

For nine of the remaining titles, I can find no record anywhere in the CSZJJ; one of these, however, is the catalogue that Fei Changfang ascribes to Nie Daozhen, 眾經錄目一卷, T2034 (XLIX) 66a21, which he draws upon elsewhere for ascriptions. 15. Listed elsewhere in the CSZJJ as an extant Dharmarakṣa translation, T2145 (LV) 7c22. Elsewhere again in the CSZJJ, a colophon also appears, noting that Nie Daozhen was amanuensis (筆受) for Dharmarakṣa’s translation, 51b8–13. The same title is ascribed to Dharmarakṣa elsewhere in the LDSBJ, T2034 (XLIX) 63a11–12, and ascribed to Nie Daozhen, with a note explicitly stating it was the second translation, 66a13.

47. 菩薩戒獨受壇文一卷 CSZJJ #108 48. 菩薩懺悔法一卷 (異本) CSZJJ #120, CSZJJ #121 49. 菩薩雜行法一卷 (no correspondence) 50. 菩薩所行四法一卷 (no correspondence) 51. 菩薩五法行經一卷 CSZJJ #99 52. 菩薩六法行經一卷 CSZJJ #100 53. 異出菩薩本起經一卷 (或無起字) CSZJJ #92 54. 眾經錄目一卷 (no correspondence)

We must recall that whatever principle Sengyou himself originally used to group these texts in such close proximity in his own list, it cannot have been ascription, since he regarded them as anonymous, and by definition, he therefore could not have taken ascription into account. But it is also impossible to believe that 41 ascriptions to Nie Daozhen, discovered by painstaking research in sources other than the CSZJJ, could have happened by pure chance to be so closely grouped in Sengyou’s list. It therefore appears that somebody has produced the bulk of this list of Nie Daozhen ascriptions by taking these 41 titles from a small stretch of Sengyou’s very long list of anonymous texts, shuffling them, and arbitrarily assigning them to Nie Daozhen, with an interspersed smattering of other materials.

Nor is this clustering of new attributions in the LDSBJ, against particular portions of Sengyou’s list, confined to this one example. We find the same pattern in the corpora ascribed to several other key translators. For Faju 法炬, for example, the LDSBJ notoriously lists 132 titles, in contrast to only 4 in the CSZJJ. In a general note, Fei says that he made up for the shortcomings of Sengyou’s ascriptions by consulting “the Jiu, bie (and other?) catalogues” 舊別諸錄. 16 Like the titles ascribed to Nie Daozhen above, the vast majority of these 132 titles ascribed to Faju (119 titles) occur close to one another in the CSZJJ, in this case in six distinct clusters:

1. nos. 7–31 from CSZJJ #522–#571, T2145 (LV) 28a8–b29; 2. nos. 32–49 from CSZJJ #745–#768, 30c12–31a8; 3. nos. 50–63 from CSZJJ #274–#293, 25a10–b2; 4. nos. 64–104 from CSZJJ #155–#199, 23b28–24a16; 17 5. nos. 105–126 from CSZJJ #615–#651, 29a16–b5; 18

6. a final short cluster, nos. 128–132 from CSZJJ #682–#685, 29c26–30a2. 19 For Tanwulan 曇無蘭, the LDSBJ equally notoriously introduces 108 new attributions, crediting him with a total of 110 translations, where previous catalogues had only listed two. 20 Here, we see the following clusters, accounting for more than half of the total list (62 titles):

1. nos. 13–22 from CSZJJ #449–#496, T2145 (LV) 27a28–c10; 2. nos. 45–56 from CSZJJ #243–#259, 24c6–23; 16. The full Faju list in the LDSBJ is found at T2034 (XLIX) 66c10–68a6. 17. Note that in this case, the length of both lists is almost the same—LDSBJ 40 texts, CSZJJ 44 texts—so that almost every CSZJJ text appears in this portion of the LDSBJ’s Faju list, though the order is very thoroughly shuffled. 18. With one slight outlier, 處中行道經 = CSZJJ #606.

19. Exceptions to this pattern: Titles no. 1–6 in the LDSBJ Faju list do not exhibit any pattern of clustering in the CSZJJ. Two texts are ascribed to Faju in the CSZJJ, viz. the 樓炭經, T2145 (LV) 9c19 (included among the first six texts in the LDSBJ), and the 福田經, 9c22; and two are listed among Dao’an’s anonymous titles in the CSZJJ, viz. the 法海經, 18b8, and the 波斯匿王喪母經, 17b2. I could find no correspondence in the CSZJJ for the 眾生未然三界經, the 諸經法菩薩名經, nor (exactly) for the 明帝釋施經 (but see 帝釋施央掘魔法服經).

20. The full Tanwulan list in the LDSBJ is found at T2034 (XLIX) 69b6–70b22. 3. nos. 57–62 from CSZJJ #607–#622, 29a8–24; 4. nos. 64–73 from CSZJJ #499–#515, 27c12–28a2; 5. nos. 80–108 from CSZJJ #797–#835, 31b11–c22. For Guṇabhadra 求那跋陀羅, the LDSBJ gives a list of 78 titles, 21 where the CSZJJ has only 13. We see the following clusters, again accounting for more than half the total list (41 titles):

1. nos. 21–34 from CSZJJ #693–#722, T2145 (LV) 30a13–b14; 2. nos. 36–51 from CSZJJ #442–#463, 27a11–b4; 3. nos. 52–61 from CSZJJ #133–#151, 23b5–23; 4. nos. 72–76 from CSZJJ #676–#690, 29c20–30a10. For Huijian 慧簡, the LDSBJ gives 25 titles, 22 where the CSZJJ has only one. We see the following clusters, again accounting for more than half the list (16 titles):

1. nos. 4–9 from CSZJJ #434–#441, T2145 (LV) 26c21–27a11; 2. nos. 14–25 from CSZJJ #201–#218, 24a17–b8. The LDSBJ listings for Dharmarakṣa 竺法護 23 are quite different from the above corpora, perhaps in part because Fei Changfang inherited a larger number of sound ascriptions. Even here, however, we find one such cluster: 1. nos. 143–163 from CSZJJ #525–#574, T2145 (LV) 28a11–c3. In other words, with two exceptions, 24 all 20 titles in this block are found within a single group of 49 titles in the CSZJJ. Finally, the LDSBJ ascribes 176 titles to An Shigao 安世高. 25 As for Dharmarakṣa, clustering is less prominent in this list, but we still find at least the following clear clusters (56 titles):

1. nos. 39–52 from CSZJJ #313–#353, T2145 (LV) 25b21–26a6. 2. nos. 63–71 from CSZJJ #660–671, 29c4–15. 3. nos. 72–75 from CSZJJ #730–#742, 30b23–c9. 4. nos. 81–102 from CSZJJ #354–#381, 26a7–b7. 5. nos. 110–117 from CSZJJ #497–#518, 27c10–28a4. 6. nos. 138–142 from CSZJJ #426–#436, 26c23–27a6. 26

This pattern surely extends beyond the examples shown above, but these examples alone suffice to show that the pattern is pervasive in the LDSBJ. Repeatedly, clusters of titles in Sengyou’s list have predominantly been assigned the same new ascription in the LDSBJ, and this pattern is seen in a large number of the LDSBJ’s ascriptions to several major translators. Indeed, the hypothesis that the CSZJJ is the LDSBJ’s source is further corroborated by the

21. T2034 (XLIX) 91a24–92b12. 22. T2034 (XLIX) 93b7–c7. 23. T2034 (XLIX) 61c26–65a7. See the study by Okabe discussed below, n. 43. 24. I cannot find an equivalent in CSZJJ fascicle 4 for the 悔過經, and the 降龍經 corresponds most closely to CSZJJ #140. 25. T2034 (XLIX) 50a21–52c14.

26. Near the end of the An Shigao list, nos. 162–173 largely correspond to texts from not one but two clusters in the CSZJJ. The resulting sequence of CSZJJ numbers runs #137, #135, #149, #728, #136, #733, #146, #731, #132, #672 (outlier), #729—in other words, most of these texts are drawn from two chunks of the CSZJJ list, #132–#149, and #728–#733, but interleaved with one another. The An Shigao list also features several possible smaller clusters, where three or four titles appear in the same relatively short segment of the CSZJJ list.

fact that frequently the same interlinear notes (or paraphrases thereof) appear in the corresponding loci in both the CSZJJ and the LDSBJ. It may seem like overkill to have shown so many examples. However, these examples were not chosen at random. As mentioned above, some of the texts on Sengyou’s list have survived to the present. 129 of the new ascriptions assigned by the LDSBJ to texts included in Sengyou’s list were accepted by later cataloguers and are still borne by the texts in question in the Taishō. Appendix 1 lists those 129 texts for the reader’s reference. As may be seen from a glance at that list, the vast majority of those surviving ascriptions (98 ascriptions) are to a small handful of translators—the same translators discussed above: Nie Daozhen (6 ascriptions), Faju (20 ascriptions), Tanwulan (13 ascriptions), Guṇabhadra (11 ascriptions), Huijian (7 ascriptions), Dharmarakṣa (14 ascriptions), and An Shigao (27 ascriptions). This means that the researcher who turns to the Taishō canon for information on the translators of these texts will encounter many of the ascriptions newly assigned by LDSBJ, as found in the lists examined above.

As Appendix 1 shows, a further 11 texts regarded as anonymous in the CSZJJ are still regarded as anonymous in the present Taishō, but carry bylines that indicate that they date to the Eastern Han. This dating, too, probably derives from the LDSBJ. As is well known, the bulk of the LDSBJ (fascicles 1–12) is organized on chronological lines, and fascicle 4 treats the Eastern Han. At the end of the fascicle we find a list of 125 anonymous texts. 27 The inclusion of these texts in this fascicle implies that they date to the Eastern Han, and this is therefore the most likely ultimate source of the dates indicated for the 11 texts in question in the Taishō.

In this list, we see a new kind of correspondence with the CSZJJ list of anonymous texts. With very few exceptions, 28 the titles in the LDSBJ’s Eastern Han anonymous list appear in exactly the same order as in the first portion of Sengyou’s list, in which Sengyou lists texts he regards as extant (CSZJJ #1–#841); but the titles in the LDSBJ’s Eastern Han list are spread throughout this portion of Sengyou’s list, rather than clustered in one place. Obviously, given that the CSZJJ list (841 titles) is much longer than the LDSBJ’s Eastern Han anonymous list (125 titles), many titles in the CSZJJ list are missing in the LDSBJ, which often “skips over” titles in the CSZJJ—sometimes as many as fifty titles at a time. However, the exact correspondence in overall order cannot be coincidental, and it is therefore certain that the compiler of the LDSBJ’s Eastern Han anonymous list was referring to the CSZJJ as he crafted his work. By itself, this fact does not necessarily mean that the LDSBJ’s Eastern Han anonymous list is suspicious. However, if we correlate titles in this list with information elsewhere in the LDSBJ, we notice yet another peculiar pattern.

the pRobleM (2): patteRns of self-contRadiction Where the CSZJJ (like Fajing’s T2146) generally k nows only a single text under a given title, the LDSBJ often features more than one notice for the same title, and these various notices give apparently contradictory information.

To begin with, fascicle 7 of the LDSBJ, which treats the Eastern Jin period, also ends with a list of supposedly anonymous titles, totaling 53 texts. By my count, 29 of those titles

27. T2034 (XLIX) 54b18–55c16. 28. Usually these exceptions amount to no more than swapping the order of two successive texts. The most striking instance of “shuffling,” which is still minor, is in LDSBJ nos. 39–46, in which the sequence of texts in the CSZJJ is #226, #219, #222, #221, #224, #225, #233, #231. also appear in the Eastern Han anonymous list in fascicle 4. In addition, the Eastern Jin list features two clusters which overlap almost perfectly with the Eastern Han list in order and content:

1. E. Jin nos. 6–14 = E. Han nos. 15–23 (perfect match); 29 2. E. Jin nos. 15–25 ≈ E. Han nos. 38–47. 30

But this obviously makes no sense. So long as we treat these lists as implying that the texts in question are to be dated to the dynasty treated by the fascicle in which they appear—which is exactly how these lists have been received by the tradition (including, in some cases, our present canon)—it would mean that the same texts are to be dated to two different periods at once. 31 (The alternative hypothesis is that Fei had suddenly discovered multiple versions of texts his predecessors only knew in one version, but that notion is unconvincing—see below.)

These contradictions, moreover, are only the tip of an iceberg. For the purposes of the present study, I have confined myself in the following to extant texts from Sengyou’s list, which is surely only a sample of a wider pattern in the LDSBJ. Even within that limited sample, however, we find the following further patterns of double listings and contradictions (as I will explain below, there is some overlap between these groups): 32

1. For 17 texts, the LDSBJ gives the name of a translator, but also includes the same text among the Eastern Han anonymous texts. 33 2. For 4 texts, the LDSBJ gives an ascription, but also includes the same text among the Eastern Jin anonymous texts. 34 3. Most egregiously, in the ruzangmu 入藏目 (“list [of texts] admitted to the canon”) in Fascicles 13 and 14, Fei lists as anonymous 64 texts among those listed in Appendix 1, for which he elsewhere gives ascriptions. 35

29. T2034 (XLIX) 74b1–6, 54c3–9. 30. T2034 (XLIX) 74b7–13, 54c22–28. No. 42 from the Eastern Han list is missing in the Eastern Jin list. Nos. 23 and 24 of the Eastern Jin list have no exact match in the Eastern Han list. See n. 101.

31. Hayashiya already noticed part of this problem; see his “Zui dai kyōroku ni kansuru kenkyū 隋代經錄に關する研究,” in Bukkyō ronsō: Tokiwa hakase kanreki kinen 佛教論叢：常盤博士還暦記念, ed. Miyamoto Shōson 宮本正尊 (Tokyo: Kōbundō shobō, 1933), 300–302. In fact, Hayashiya points out that a similar overlap is also found between a chunk of the Eastern Han list and the list for the Western Jin.

32. The following lists include three types of text not found in Appendix 1: 1) texts for which ascriptions already appear in interlinear notes in Fajing; 2) texts for which the LDSBJ ascription is at variance with that given in the Taishō; 3) texts which are still listed as anonymous and undated in the Taishō. Texts in these groups in many cases include texts for which the LDSBJ does include an ascription (or ascriptions), but where the ascription did not “stick” to become canonical.

33. In the lists in this and following footnotes, to save space, I give only the Taishō number of each text, followed by loci in the LDSBJ where the ascriptions or dates in question are given. An Shigao—T1406: 55c7, 52a18; *Lokakṣema—T350: 55b10, 53a2–3 (note that T350 is attributed to *Lokakṣema in CSZJJ, but by the title 寶積經, 6b17); Zhi Qian—T644: 55b7, 58a22; An Faqin—T816: 54b22, 65a13–20; Faju—T332: 54c2, 67b3; Bo Fazu— T330: 55a9, 66b4 (also Bo Yan, 56c14–15); Tanwulan—T22: 55a27, 69c5; T42: 55a21, 70a14; T86: 55a22, 70a15; T393: 54c20, 70a19; T1327: 55c5, 70b11; T1378b: 55c3, 70b9; Shengjian/Fajian—T696: 55b5, 83c9; Fahai— T566: 55b11, 94a14–15; T1490: 54c12, 94a16–17; Guṇabhadra—T783: 55b15, 91b25; *Guṇavarman—T1472: 54c24, 90b2 (also E. Jin). 34. Bo Fazu—T777: 74b11, 66b14; Guṇabhadra—T1670A: 74a28, 92b12; *Guṇavarman—T1472: 74b9, 90b2 (also E. Han); Tanjing—T754: 74a27, 96a25. 35. An Shigao—T151: 52a24, 117a16; T356: 52b10–11, 113a1; T499: 51b18, 117a25; T506: 52b5, 117b12; T525: 50c13, 117b19; T621: 52b15, 113a16; T724: 51c18, 117b24; T729: 51a23, 117c9; T730: 51b9, 118c22; T731:

51c16, 118c23; T733: 52a2, 117b13; T791: 51a8, 116c28; T1406: 52a18, 114b8 (also E. Han anon.); T1467a/b: 51b1, 119c9; T1470: 50a23–24, 119c3; Zhi Yao—T88: 54a5, 117c20; Kang Mengxiang—T197: 54b2, 116c9; Zhi

4. We also find among extant texts from the CSZJJ list 33 texts for which other types of double listing or contradiction appear in the LDSBJ. 36 On some occasions, further, these patterns of double listing and contradiction combine to create perfect storms of confusion. To give just a couple of examples: For the Zhangzhe Weishi suowen pusa xiuxing jing 長者威施所問菩薩修行經 T330, Fei has one listing among the Eastern Han anonymous texts; a second ascribing the text to Bo Yan 白延, citing the Shixing

Qian—T396: 58c12, 119a20; T427: 58b8, 113a5 (also Guṇabhadra, Saṅghabhara); T1356: 58b7, 113c28; T1477: 58a11, 119c8; *Kalyāṇa?—T269: 56c19, 112c27 (also Zhiyan and Baoyun); Dharmarakṣa—T391: 64a24, 118c11; T428: 64b4, 113a5; T611: 64a21, 119a6; T685: 64a27, 118c11; T2867: 64b26, 113b6; Nie Daozhen—T463: 65c7, 112c21; T483: 66a1, 114c6; T1502: 65c18, 114c8; Nie Chengyuan—T537: 65b21, 117b20 (also Shi Songgong/

Gaogong, Guṇabhadra); Faju—T113: 67c3, 116c12; T133: 67a26, 112c14; T178: 67c16, 113b13; T332: 67b3, 116c15 (also E. Han); T500: 66c22, 116c25; T501: 67c13, 117b2; T502: 67b24, 117a4; T509: 67a23, 113b18; T739: 66c20, 118c23; T794: 66c22, 117b27; Bo Fazu—T330: 66b4, 112c7 (also E. Han; also Bo Yan); T777: 66b14, 113a11 (also E. Jin); Tanwulan—T86: 70a15, 117b23; T216: 69c5, 117a21; T393: 70a19, 116c21 (also E. Han); T510: 69c12, 113b19; T1326: 70b10, 114b3; T1327: 70b11, 114b3–4 (also E. Han); T1329: 70b10, 114b3; T1378b: 70b9, 114a19 (also E. Han); T1391: 70b3–4, 114a22–23; T1393: 70a24, 70b1, 114a21, a24; F248: 70a3, 119a25; Shengjian/Fajian—T696: 83c9, 116c23–24 (also E. Han); Fahai—T1490: 94a16–17, 114b22 (also E. Han); Guṇabhadra—T427: 91b19–20, 113a5; T497: 91c26, 117a26; T537: 91c12, 117b20 (also Nie Chengyuan, Shi

Songgong/Gaogong); T540a/b: 91c17, 117c16; T747b: 91c5, 117b22; T753: 91c4, 117b8; T783: 91b25, 117b15 (also E. Han); Juqu Jingsheng—T826: 93a3, 117c19; Zhiyan and Baoyun—T269: 89c9, 112c27 (also *Kalyāṇa); Shi Songgong/Gaogong—T537: 72a13, 117b20 (also Guṇabhadra, Nie Chengyuan); Huijian—T797a/b: 93b14, 117b18; T827: 93b12, 116c16; Saṅghabhara—T427: 98b20, 113a5.

These inconsistencies between the ruzangmu and other portions of the LDSBJ fit with a pattern noticed by Hayashiya, and might be accounted for by his theories about the process by which Fei composed his ruzangmu (Hayashiya 1933: 280–82; like Sakaino’s, Hayashiya’s study is over eighty years old, but his findings still deserve to be better known). First, Hayashiya believes that Fei, like Fajing, composed his catalogue on the basis of comparison of older catalogues only, without independent direct scrutiny of the texts themselves. Second, he believes that the heart of Fei’s originally intended project was the dailu 代錄 (“catalogue by dynasties,” fascicles 4–12), but that Fajing’s catalogue, which appeared in 594 when Fei had already been at work for over half a decade (see T2034

[XLIX] 120c12–13), made it de rigeur to also include a “catalogue of texts admitted to the canon” (ruzanglu 入藏 錄). Hayashiya thinks that under the pressure this circumstance exerted, Fei added the ruzangmu in haste, and based it largely upon Fajing, even at the cost of inconsistency with other portions of his own work (see Ōuchi, Nanboku chō Zui Tō ki Bukkyōshi kenkyū, 80–81 and 105 n. 15). For instance, Hayashiya holds that the entire “Mahāyāna Vinaya” section of Fei’s ruzangmu is identical in content and order to Fajing’s section on the same material (pp. 280–81); the same is largely true of the “Mahāyāna Abhidharma” section (p. 281) (obviously this pattern of uncritically copying from whole lists conforms with much of what I have uncovered in the present paper). Hayashiya also shows that Fei copies an error made by Fajing in the treatment of Paramārtha’s Lü ershi’er mingliao lun 律二十二明了論 T1461 (see T2034 [XLIX] 119b3, 120a6; T2147 [LV] 140a15, 156a4); that the Yijiao jing lun 遺教經論 T1529 redundantly appears in two entries in Fei, because one is copied from Fajing (see T2034 [XLIX] 115a4, 120a7; T2146 [LV] 143c25); and that the “Mahāyāna sūtra” section of Fei’s ruzangmu contains texts that do not appear in his own dailu, but do appear in Fajing’s ruzanglu (p. 281). All of this evidence shows that Fei copied information uncritically from Fajing.

On the basis of such “comical” (滑稽) errors, Hayashiya concludes that Fei’s ruzangmu should not be regarded as an independent work, and he also calls it “a complete failure,” “useless for practical purposes” (pp. 282, 283, 302). At the same time (despite having noticed some overlaps in the dailu between lists of anonymous scriptures for the Eastern Han and the Jin, as mentioned above n. 31), Hayashiya’s overall evaluation of the LDSBJ is that the dailu communicates to us much valuable information from lost catalogues, and therefore, that careful use of at least the dailu is indispensable to the study of the history of canonical texts. In the present study, however, we see that inconsistencies between the ruzangmu and other parts of the LDSBJ are just one tessera in a much larger mosaic of errors; and we therefore cannot, as Hayashiya suggests, solve these problems merely by setting the ruzangmu aside. 36. I will not list detailed references for this group, but the texts I have identified that are subject to such problems are T16, T17, T20, T67, T70, T122, T145, T195, T267, T269, T365, T371, T373, T426, T492, T508, T512, T536, T537, T561, T618, T622, T688, T769, T1043, T1301, T1342, T1466, T1472, T1474, T1477, T1486, and T2027.

lu 始興錄 and Baochang 寶唱; a third ascribing it to Bo Fazu 白法祖; and a fourth entry in fascicle 13 still incongruously listing the text as anonymous. 37 Similarly, the Shi’er you jing 十二遊經 T195 is a text probably composed in some sense in China, but Fei protests too much—he ascribes one version to the obscure 彊梁婁至 (*Kalyāṇaruci?), citing the Shixing lu and Baochang; another version to *Kālodaka 迦留陀伽, citing Zhu Daozu 竺道祖, the “miscellaneous catalogue(s?) of the Jin,” and Baochang; and a third version to Guṇabhadra, citing the Gu lu 古錄. 38

Truly, it appears, Fei’s left hand knows not what his right hand does. On the most charitable reading, we might imagine that Fei had suddenly discovered multiple versions of texts for which his predecessors had known only a single translation. In this case, there might be no contradiction, because each assertion might apply to a different text with the same title—one anonymous, and another by Tanwulan, for instance. However, this hypothesis is rendered implausible by the sheer number of such new entries, in combination with the fact that scholars have repeatedly found many of these ascriptions dubious on other grounds (see immediately below). It is difficult, therefore, to avoid the conclusion that this conflicting information does indeed pertain in many cases to “the same” text, and thus, that Fei either did not notice that he was contradicting himself, or did not care. ModeRn scholaRly cRitiqUes of fei’s new ascRiptions

It comes as no surprise, then, that previous scholarship has regularly found highly dubious the new ascriptions in the LDSBJ, and indeed, that precisely the new ascriptions here under study have featured prominently among the most glaring problems. For example, Hayashiya rejects many of the ascriptions listed in Appendix 1. 39 Ui argues against many of the LDSBJ ascriptions to An Shigao, including many of the ascriptions under study here. 40 Similarly, none of the An Shigao ascriptions under study here is regarded as genuine in Zacchetti’s survey of the state of research on An Shigao; 41 and none of the same ascriptions is regarded as genuine by Nattier. 42 Okabe studied 40 new ascriptions to Dharmarakṣa appearing for the first time in the LDSBJ, and argued that none is to be trusted. 43 Bagchi pointed out long ago that Zhisheng was highly suspicious of the ascriptions to Tanwulan that he received; 44 taken together, Zhisheng’s doubts and criticisms mean that there was no work extant in his time for which he regarded the ascription to Tanwulan as impeccable. Zürcher noted the extensive

37. T2034 (XLIX) 55a9, 56c14–15, 66b4, 112c7. 38. T2034 (XLIX) 65a8–11, 70b27–c2, 91b24. Amid this profusion of heavily referenced information for three ascriptions, Fei himself only ever explicitly counts two versions of the text.

39. Hayashiya, Kyōroku kenkyū; Iyaku kyōrui kenkyū. It would be an excessively large task to list exhaustively loci in Hayashiya’s works pertinent to all the texts listed in Appendix 1 to this paper, and interested readers are encouraged to use Hayashiya’s indices, or consult relevant entries in my “CBC@” (Chinese Buddhist Canonical Attributions) database at http://dazangthings.nz/cbc/. 40. Ui, Yakukyōshi kenkyū, 437–52.

41. Stefano Zacchetti, “Defining An Shigao’s 安世高 Translation Corpus: The State of the Art in Relevant Research,” in Xiyu lishi yuyan yanjiu jikan 西域歷史語言研究集刊 / Historical and Philological Studies of China’s Western Regions no. 3, ed. Shen Weirong 沈衛榮 (Beijing: Science Press, 2010), 249–70.

42. Jan Nattier, A Guide to the Earliest Chinese Buddhist Translations: Texts from the Eastern Han 東漢 and Three Kingdoms 三國 Periods (Tokyo: The International Research Institute for Advanced Buddhology, Soka Univ., 2008). Nattier’s presentation of these attributions is negative; that is to say, if a work is not listed in her Appendix 1, “the attribution is not, at the present state of our knowledge, considered to be genuine” (p. 29). 43. Okabe Kazuo 岡部和雄, “Jiku Hōgo no yakkyō ni tsuite” 竺法護の訳経について, IBK 11.1 (1963): 148–49.

44. Prabodh Chandra Bagchi, Le canon bouddhique en Chine: Les traducteurs et les traductions (Paris: Librairie Orientaliste Paul Geuthner, 1927), 322–34. problems with “wild attributions” to Faju in “later catalogues,” and on similar grounds, questioned all ascriptions to Nie Daozhen. 45 Iwamatsu has argued that no ascriptions of extant texts to Bo Fazu are reliable, and that many of these ascriptions stem from the LDSBJ. 46 Strickmann questioned whether Huijian was ever actually the translator of any texts at all (though he is willing to entertain the possibility that Huijian might have composed some texts). 47 Mizuno Kōgen argued that a number of texts transmitted individually originally formed part of two larger collections, alternate translations of the *Ekottarikāgama and the Madhyamāgama respectively; 48 and a later quantitative study of Mizuno’s proposed alternate Madhyamāgama confirmed that this group of texts probably has a common author. 49 But this would mean that all the texts in each group were produced together, by the same translator, making a nonsense of the scattered attributions they now bear in the present canon; and in this light, it is sobering to note that 35 of Mizuno’s texts are among those studied here. 50 fei’s sUpposed soURces

When we turn to examine Fei’s claims about his sources for this rash of new and unreliable ascriptions, it only makes things look worse. In assigning these new ascriptions (again confining ourselves to Sengyou’s list as our sample), the LDSBJ variously cites something like seventeen catalogues in all (see Appendix 2 for a full list of the sources Fei claims to draw upon). This profusion of sources and concrete details might appear impressive. However, despite this apparent embarrassment of riches, there remain a large number of translators for whom Fei never cites such a specific source for a single reascription; 51 and more generally, Fei provides such specific references for only a fraction of the many new ascriptions he provides. 52 On the other hand, the range of Fei’s sources is suspiciously wide, in proportion to the use that he makes of each one—it appears that he only found a tiny amount of useful information in each. We might suspect that Fei is most concerned with impressing us with the sheer fact that he has so many sources at his disposal. Thus, among the texts here under study, Fei refers

45. Erik Zürcher, The Buddhist Conquest of China: The Spread and Adaptation of Buddhism in Early Medieval China (Leiden: Brill, 1959; 3rd ed., 2007), 70, 345 n. 254; 68. 46. Iwamatsu Asao 岩松浅夫, “Nehan gyō shōhon no hon’yakusha” 涅槃経小本の翻訳者, IBK 25.1 (1976): 245. 47. Michel Strickmann, “The Consecration Sutra: A Buddhist Book of Spells,” in Chinese Buddhist Apocrypha, ed. Robert E. Buswell, Jr. (Honolulu: Univ. of Hawai’i Press, 1990), 75–118.

48. Mizuno Kōgen 水野弘元, “Kan’yaku Chū agon to Zōichi agon to no yakushutsu ni tsuite” 漢譯『中阿含』と『増一阿含』との譯出について, Ōkurayama gakuin kiyō 大倉山学院紀要 2 (1956): 41–90 (L); idem, “Chū agon kyō kaidai” 中阿含経解題, Kokuyaku issaikyō 国訳一切経, Agon bu 阿含部 6 (rev. ed., 1969), 403–11; idem, “Kan’yaku Chū agon kyō to Zōichi agon kyō” 漢訳『中阿含経』と『増一阿含経』, Bukkyō kenkyū 仏教研究 18 (1989): 1–42 (L), Chinese translation: “Hanyi Zhong ahan jing yu Zengyi ahan jing” 漢譯《中阿含經》與《増一阿含經》, in Mizuno Kōgen, Fojiao wenxian yanjiu: Shuiye Hongyuan zhuzuo xuanji, vol. 1 佛教文獻研究：水野弘元著作選集 (一), tr. Xu Yangzhu 許洋主 (Taipei: Fagu wenhua, 2003), 509–79.

49. Jen-Jou Hung, Marcus Bingenheimer, and Simon Wiles, “Quantitative Evidence for a Hypothesis Regarding the Attribution of Early Buddhist Translations,” Literary and Linguistic Computing 25.1 (2010): 119–34. 50. Texts from Mizuno’s alternate *Ekottarikāgama originally on Sengyou’s list, but given new ascriptions in the LDSBJ (see Appendix 1): T39, T89, T122, T123, T127, T131, T133, T134, T136, T139, T140, T149, T215, T216, T508, T684; texts from Mizuno’s alternate Madhyamāgama given the same treatment: T47, T49, T50, T53, T55, T56, T58, T60, T64, T65, T66, T70, T77, T79, T82, T83, T90, T91, T92.

51. Zhi Jiangliangjie 支彊梁接 (*Kalyāṇa?), Zhu Lüyan 竺律炎, Faju (and Fali 法立), Fahai 法海, Nie Daozhen, Nie Chengyuan, Bo Fazu, Tanwujie, Zhu Fonian 竺佛念, *Guṇavarman 求那跋摩, Fayong 法勇, Huixian 慧顯, Huijian, and Saṅghabhara 僧伽婆羅.

52. This is true not just of interlinear notes on specific texts, but general notes on whole translation groups and their corpora. to only a small number of catalogues for three or more translation groups, and even then, almost always for only one text in each group. 53 The remaining sources (a dozen or so) are each cited only once or twice, for new attributions for one or two of Sengyou’s anonymous texts. 54 In other words, the sources Fei claims for his ascriptions are nearly as various and scattered as they possibly could be. 55

Against the pattern of clustering correspondences between Fei’s new ascriptions and portions of the “CSZJJ anonymous list,” this riotous parade of sources looks like nothing more than a smokescreen. I already mentioned above that it is impossible that Sengyou was grouping texts in his anonymous list by ascription, when he explicitly states that he does not know who translated them. But then it is equally impossible to imagine that it just so happened that ascriptions drawn from painstaking research in this wide range of sources fall together in clusters in Fei’s work mapping onto their sequence in Sengyou’s list.

iMplications

This brings us to the crux of the matter—the implications of this evidence for Fei’s working method, and consequently, our evaluation and use of his ascriptions.

Although scholars have long recognized that the LDSBJ contains many implausible ascriptions, opinion has been divided on the question of whether they originated with Fei himself, and some scholars have argued that he merely acted in good faith, as a conduit for inaccurate information contained in some of his numerous sources. 56 But as we have seen

53. Baochang’s catalogue (for Zhi Qian, *Kalyāṇaruci, Bo Yan, Zhi Fadu, *Kālodaka, Guṇabhadra, Tanjing); the Jiu lu (for Han anonymous texts, Zhi Qian, anonymous Wei-Wu texts, Tanwulan, Guṇabhadra); bie lu (for *Lokakṣema, anonymous W. Jin texts, Fajian/Shengjian, Tanwulan, Kumārajīva); Zhu Daozu’s catalogue (for An Faqin, Zhi Fadu, *Kālodaka, Zhiyan and Baoyun, Tanyao); CSZJJ (for Kang Senghui, anonymous E. Jin texts, Guṇabhadra, Juqu Jingsheng, Xiang gong); the Shixing catalogue (*Kalyāṇaruci, Bo Yan, Fajian/Shengjian, Shi Songgong/Gaogong, Xiang gong); the Wu catalogue (for Kang Mengxiang, Zhi Yao, anonymous W. Jin [!] texts); the “miscellaneous Jin catalogue(s)” (for An Faqin, *Kālodaka, Zhiyan and Baoyun); Fashang (for Fajian/Shengjian, *Nandi[n], Tanjing). Refer to Appendix 2 for details.

54. The Gu lu (for Han anonymous texts, anonymous Wei-Wu texts); the Cen Hao catalogue (for Yan Fotiao); Dao’an (for Dharmarakṣa, Xiang gong); Nie Daozhen (for Dharmarakṣa); Zhi Mindu (for Zhi Fadu); the “catalogue of the two Qin” (for Kumārajīva); Li Kuo (for Kumārajīva, Guṇabhadra); Daohui’s Song-Qi catalogue (for *Kālayaśas); GSZ (for *Kālayaśas); Wang Zong’s catalogue (for Tanjing); and the Zhao catalogue (for Shi Songgong/Gaogong). Once again, refer to Appendix 2 for details.

55. One remarkable feature of Fei’s claims about his sources is a striking silence—for the texts under study here, he never draws upon Fajing, his most immediate predecessor, even where interlinear notes in Fajing give new ascriptions to texts from Sengyou’s list and Fei agrees with those ascriptions (e.g., for T129, T234, T294, T350, T383, T492, T508, T512, T526, T527, T533, T536, T566, T754, T816, T1043, T1161, T1342, T1472, T1485, T1490, X15). Obviously, such agreements cannot be coincidence. In principle, there could be various reasons for this silence: 1. Fei knew Fajing’s ascriptions and drew upon them, but it did not suit his agenda to acknowledge Fajing as his source. 2. Both were drawing upon the same third source(s) in each case. (Assessment of this possibility is made more difficult by the fact that interlinear notes in Fajing almost never explicitly mention sources.) 3. Interlinear notes in Fajing are based upon the LDSBJ, and therefore postdate it, and were added later to Fajing’s catalogue. (In this case, we would also have to account for several occasions upon which ascriptions in interlinear notes in Fajing differ from the LDSBJ, e.g., for T89, T206, T310(33), T1013, T1689.) 4. Fei may have regarded Fajing as a competitor for imperial patronage, and not wished to credit him as a source of information (I am grateful to an anonymous reviewer for pointing out this possibility). This problem may not be soluble, but is an important avenue for future research. See n. 11 above.

56. For example, Tokiwa has argued that the numerous new and inaccurate ascriptions to Faju and Tanwulan were taken over by Fei from earlier catalogues, and therefore do not impugn the credibility of the LDSBJ itself; see Tokiwa, Gokan yori, 63–64. Another example may be seen in the treatment of the supposed Zhu Shixing 朱士行 catalogue of Han texts. As summarized by Tan Shibao, there have been some scholars (e.g., Feng Chengjun 馮承鈞 and Yao Mingda 姚名達) who regarded this very early and highly implausible catalogue as genuine. Among other

above, two of the most problematic patterns plaguing these lists—parallels between clusters in the LDSBJ and their counterparts in contiguous chunks of Sengyou’s list, and contradictory information about the individual titles—are widely distributed in multiple portions of the LDSBJ. This does not jibe with Fei’s claims to be drawing his information from a dozen and a half different sources, composed over a period of several centuries. For one thing, it would be another impossible coincidence if multiple sources just happened to handle the list found in the CSZJJ in the same fishy manner. For another, many of Fei’s supposed sources predate the CSZJJ, which would make such use of the CSZJJ list on their part anachronistic. The LDSBJ is the bottleneck where all these problems gather, and therefore the most economical explanation for the evidence reviewed above is that Fei himself is deliberately falsifying information. 57 Recent scholarship has seen attempts to revise our evaluations of Fei. In part, scholars have argued that Fei’s poor reputation as a cataloguer has led us to overlook his important contributions to the development of Chinese historiography, and his role in shifting notions of canonicity so that Chinese imperial imprimatur became the deciding authority for admission to the canon. 58 I have no quarrel here with such lines of argument. At the same time, one of the most important historical functions of the LDSBJ is precisely as a catalogue, that

scholars, who agreed that the catalogue had existed but considered it a forgery, there have been disagreements about its probable date and author, but most scholars (e.g., Tang Yongtong 湯用彤, Tokiwa, Hayashiya, and Lü Cheng 呂

澂) agree in speculating that it was by someone else, writing earlier than Fei himself. This means that it has been a minority position to argue, as Tan himself does, following Sakaino, that the information presented as from Zhu Shixing was forged by Fei himself. See Tan Shibao, Han Tang Fo shi tanzhen, 27, 94. 57. The strongest alternate explanation might be that Fei inherited all of these bogus ascriptions from a single lost prior catalogue. That proximate source would then constitute the bottleneck, not the LDSBJ. Such a scenario would exonerate Fei at least in the sense that in reporting these ascriptions, he showed himself only gullible, but not deliberately deceptive. The strongest candidate for such a proximate source would be Baochang’s catalogue, because it postdates most of the supposed translators and catalogues at issue here; and is supposed also have been sufficiently global in scope to have been a proximate source for such a wide range of attributions; see Tokiwa, Gokan yori, 69–72, endorsed by Palumbo, review of Storch. Note that this hypothesis would require Baochang’s catalogue to postdate the CSZJJ, since only a later catalogue could have made the problematic use of Sengyou’s list under study here. Estimates of the date of Baochang’s catalogue vary, e.g., 520 or 521 (Tanya Storch, The History of Chinese Buddhist Bibliography: Censorship and Transformation of the Tripiṭaka [Amherst, NY: Cambria Press, 2014], 51), or 518 (Henri Maspero, “Sur la date et l’authenticité du Fou fa tsang yin yüan tchouan,” Mélanges d’Indianisme offerts par ses éleves à M. Sylvian Levi [[[Wikipedia:Paris|Paris]]: E. Leroux, 1911], 129–49), or 516 (Antonello Palumbo, An Early Chinese Commentary on the Ekottarika-āgama: The Fenbie gongde lun 分別功德論 and the History of the Translation of the Zengyi ahan jing 增一阿含經 [[[Wikipedia:Taipei|Taipei]]: Dharma Drum Publishing Co., 2013], 147–48 n. 106). Moreover, this hypothesis cannot account for the fact that Fei reports that Fashang (postdating Baochang) was among his sources for some ascriptions under study here; nor the fact that he refers to the CSZJJ, against the evidence of the CSZJJ itself (the only case in which we can check him against his supposed source). Further, given that (as Hayashiya showed) the source of the LDSBJ dailu is clearly Fajing, not Baochang, this hypothesis also cannot account for the fact that the same pattern—batch treatment (and consequent botching) of entire lists—is common to both the dailu and the ruzangmu; nor, since chronological organization by dynasties was Fei’s own innovation, for the presence of similarly clumsy batch-processing, and consequent redundancy, in the lists of anonymous texts for the Eastern Han, Eastern Jin, and Western Jin periods.

Be that as it may, the ultimate point is not to arraign Fei and determine his guilt or innocence in the actual production of the information he conveys. It is far more important to assess the reliability of that information on its own terms. In that light, our present findings require the same extreme caution, whatever our speculations might be about the relative roles of Fei and his sources in producing the shambles they bequeathed to us. 58. Particularly Ōuchi, Nanbokuchō; Storch, History; idem, “Fei Changfang’s Records of the Three Treasures Throughout the Successive Dynasties (Lidai sanbao ji 歷代三寶紀) and Its Role in the Formation of the Chinese Buddhist Canon,” in Spreading the Buddha’s Word in East Asia: The Formation and Transformation of the Chinese Buddhist Canon, ed. Jiang Wu and Lucille Chia (New York: Columbia Univ. Press, 2016), 109–42; see Palumbo, review of Storch, History.

is, as a source of attributions and dates—and as the sample in Appendix 1 amply shows, our present canon has been profoundly shaped by Fei’s influence in this regard. 59 On this score, however, some scholars have gone beyond suggesting that Fei may have been a more complex figure, and attempted to rehabilitate him precisely as a bibliographer, and therefore, to rehabilitate understandings of elements of textual history for which the LDSBJ is our ultimate source. Tanya Storch, in particular, has recently authored a monograph in English on the history of Chinese Buddhist bibliography. This book could easily become a main force shaping the way students and scholars understand such questions for years to come—the very attempt at the topic is rare enough, in any language. However, by her own account, one of Storch’s central aims in that study is “to restor[e Fei] Changfang’s credibility.” 60 The findings of the present study show that were Storch to succeed in this aim, and were scholars to become more trusting, not less, of Fei Changfang’s attributions, it would be a retrograde step that could bring disastrous consequences for the study of many questions. The obscurity in which Sakaino’s findings have languished for eighty years, as I noted at the outset of this study, strongly implies that our main problem is in fact that awareness of the unreliability of the LDSBJ still needs to be increased, not reduced or “corrected.” Modern scholars concerned with texts as historical evidence, and the construction of interpretations of history on their basis, cannot afford to downplay or ignore the massive distortions that have entered our record through the LDSBJ.

It is therefore crucial that we do not conflate revisionism about Fei’s overall impact on the tradition with a rehabilitation of Fei’s reputation precisely as a cataloguer. Our attitudes to Fei’s ascriptions should be kept sharply distinct from any assessment we might make of him as a creative historiographer, apologist or propagandist for the faith, or architect of influential models of canonicity and resulting concrete instantiations of the canon. On questions of ascription and dating, the new evidence presented here urges upon us an even more rigorous hermeneutic of suspicion. More than ever, we should presume all new ascriptions in the LDSBJ to be false unless proven otherwise, and the compilation of a reliable and readily accessible list of all such ascriptions remains an urgent desideratum. 61

59. The 193 texts I have identified as still extant from Sengyou’s list constitute just over 10% of the approximately 1758 canonical texts originally presented in the tradition as translations (counting as “translations” T1–1692, T2030–2033, 2042–2049, T2865–2920). The 129 ascriptions and/or dates listed in Appendix 1 alone constitute 7.3% of such “translations.” However, Sengyou’s list is obviously only a somewhat arbitrary sample of canonical texts at large, and problematic ascriptions introduced to the tradition by the LDSBJ are certainly more numerous.

60. Storch, History, 24. This goal is reflected in the arguments and assumptions shaping much of Storch’s book, including the treatment as reliable of the bulk of Fei’s reports about numerous lost catalogues (see my Appendix 2 below). See Storch, History, 23, 26, 27–38, 96–98. In her quest to claim that Fei’s information about ascriptions is reliable, the principal obstacle that Storch confronts is the silence of Sengyou on many of the same purported sources. She handles this problem by trying to undermine Sengyou, casting him as a dogmatic and ideologically motivated propagandist, who deliberately censored the information available to him; pp. 21–24, 55–61, 65–70. She also attempts to cast doubt upon the reliability of Dao’an; pp. 31–32, 146–47 and n. 16. Storch’s positive attempts to realize this agenda are complemented by an equally striking silence, since she rarely mentions, and certainly does not seriously engage with, the nature, extent, and implications of the numerous problems already pointed out in the LDSBJ by earlier scholars (for which see once more n. 6 above). For instance, it is telling—and remarkable—that a book on the history of Chinese Buddhist bibliography can entirely omit from its bibliography one of Hayashiya’s two major works (Iyaku kyōrui kenkyū), and although it lists the other (Kyōroku kenkyū), does not, so far as I can find, cite it even once; the book evinces a similar lack of serious engagement with key works by Ui, Tokiwa, and others.

61. It always remains possible in theory, of course, that some of the new “information” conveyed to us by Fei is genuine. Some other catalogues now lost, including Baochang but also catalogues for specific dynasties, corpora, appendix 1

For approximately 129 texts from the “CSZJJ anonymous list,” 62 texts as presented in the Taishō bear ascriptions or dates first found in the LDSBJ (the count is approximate because of a few complications, which are noted below). On the basis of the evidence presented in this article, all these ascriptions should be regarded as suspect. Texts are given in Taishō order to make it easier to consult this list for individual texts. Each text is followed by the locus in the LDSBJ (T2034 [XLIX]) where the ascription or date in question is given.

七佛父母姓字經 T4, Anon., Former Wei, 60b19. 尸迦羅越六方禮經 T16, An Shigao 安世高, 52a15. 善生子經 T17, Zhi Fadu 支法度, 68a17–18. 開解梵志阿颰經 T20, Zhi Qian 支謙, 57c22. 寂志果經 T22, Tanwulan 曇無蘭, 69c5. 頂生王故事經 T39, Faju 法炬, 67a19. 鐵城泥犁經 T42, Tanwulan, 70a14. 閻羅王五天使者經 T43, Huijian 慧簡, 93b10. 63 離睡經 T47, Dharmarakṣa 竺法護, 64b21. 求欲經 T49, Faju, 67a2. 受歲經 T50, Dharmarakṣa, 64a23. 苦陰經 T53, Anon., E. Han, 55a25. 苦陰因事經 T55, Faju, 67c18. 樂想經 T56, Dharmarakṣa, 64b25. 阿耨風經 T58, Tanwulan, 69c9. 64 瞿曇彌記果經 T60, Huijian, 93b19. 瞻婆比丘經 T64, Faju, 67b16. 伏婬經 T65, Faju, 66c26. 魔嬈亂經 T66, Anon., E. Han, 55a2. 弊魔試目連經/魔嬈亂經 T67, Zhi Qian, 58b23. 數經 T70, Faju, 66c20. 尊上經 T77, Dharmarakṣa, 64b25.

or regions, were very likely extant in his time; see, e.g., Hayashiya, “Zui dai,” 236–37. Further, according to Hayashiya, we have concrete records indicating that at least the Liu Song bie lu, Fashang’s catalogue, Baochang’s catalogue, and Li Kuo’s catalogue were extant down to the time of Daoxuan’s Da Tang neidian lu 大唐內典錄 T2149; 248. This could mean that at least for those catalogues, there is a chance that Fei was conveying real information. However, even for such sources as may have existed in Fei’s time and had real independent value, the problem that confronts us, given the very great quantity of demonstrably unreliable information in the LDSBJ, is to sort the wheat (if any) from the chaff. To give just one more example of problems from the most unexpected quarters, I have elsewhere presented evidence that the LDSBJ may be misleading even for periods very close to Fei’s own time, when we might presume his information would have been most accurate and his opportunities for misrepresentation fewest; Michael Radich, “External Evidence Relating to Works Ascribed to Paramārtha, with a Focus on Traditional Chinese Catalogues,” in Shintai sanzō kenkyū ronshū 真諦三藏研究論集 [Studies of the Works and Influence of Paramārtha], ed. Funayama Tōru 船山徹 (Kyoto: Kyōto daigaku jinbun kagaku kenkyūjo /Institute for Research in Humanities, Kyoto Univ., 2012), esp. 93–94; idem, “Tibetan Evidence for the Sources of Chapters of the Synoptic Suvarṇaprabhāsottama-sūtra T664 Ascribed to Paramārtha,” Buddhist Studies Review 32.2 (2015): 263–64 and esp. n. 57. 62. Again, the exact number of texts whose dates or attributions derive from the LDSBJ depends upon the status of interlinear notes in Fajing; see nn. 11, 55.

63. Appears in the LDSBJ under the title 僧王五天使經, but with a note in SYM giving the form of the title found in the CSZJJ and T43, 疑僧字 [+或作閻字 SYM]. 64. The LDSBJ here features an interesting interlinear gloss on the title: 阿耨風經一卷 (阿耨風晉言依次) (interlinear note missing in P). This interlinear gloss, despite the fact that it has “Jin” 晉, appears only in the catalogues. 晉言 appears fairly frequently in the translations of Dharmarakṣa, T221, and texts attributed to a few other translators like An Faqin. 鸚鵡經 T79, Guṇabhadra 求那跋陀羅, 91c13. 意經 T82, Dharmarakṣa, 64a21. 應法經 T83, Dharmarakṣa, 64a22. 泥犁經 T86, Tanwulan, 70a15. 八關齋經 T89, Juqu Jingsheng 沮渠京聲, 92c23. 鞞摩肅經 T90, Guṇabhadra, 91c13. 婆羅門子命終愛念不離經 T91, An Shigao, 51b19. 十支居士八城人經 T92, An Shigao, 50c19. 相應相可經 T111, Faju, 67c15. 65 難提釋經 T113, Faju, 67c3. 波斯匿王太后崩塵土坌身經 T122, Faju, 67b2. 放牛經 = T123, Kumārajīva 鳩摩羅什, 78c5. 66 四人出現世間經 T127, Guṇabhadra, 91c7. 婆羅門避死經 T131, An Shigao, 51b24. 頻毘 [v.l. 婆 SY] 娑羅王詣佛供養經 T133, Faju, 67a26. 長者子六過出家經 T134, Huijian, 93b23. 四未曾有法經 T136, Dharmarakṣa, 64b3. 四泥犁經 T139, Tanwulan, 70a8.

阿那邠邸化七子經 T140, An Shigao, 50c18. 佛母般泥洹經 T145, Huijian, 93b22. 67 阿難同學經 T149, An Shigao, 52a12. 阿含正行經 T151, An Shigao, 52a24. 大方便佛報恩經 T156, Anon., E. Han, 54b18. 68 大意經 T177, Guṇabhadra, 91c18. 前世三轉經 T178, Faju, 67c16. 異出菩薩本起經 T188, Nie Daozhen 聶道真, 66a20. 69 十二遊經 T195, *Kālodaka 迦留陀伽, 70b27–c2. 70

65. Harrison notes that T111 is “suspiciously similar in wording” to the earlier T101(20); Paul Harrison, “Another Addition to the An Shigao Corpus? Preliminary Notes on an Early Chinese Saṃyuktāgama Translation,” in Early Buddhism and Abhidharma Thought: In Honor of Doctor Hajime Sakurabe on His Seventy-seventh Birthday, ed. Sakurabe Hajime Hakushi Kiju Kinen Ronshū Kankōkai (Kyoto: Heirakuji shoten, 2002), 15. 66. The graphically similar title 牧牛經 is ascribed in the LDSBJ to Kumārajīva, T2034 (XLIX) 78c5, and this is probably the source of the received ascription of T123. The title 牧牛經 also appears in Sengyou’s list of anonymous texts, T2145 (LV) 27b26; but no 牧牛經 is extant. In Da Tang neidian lu, 牧牛經 is listed among Kumārajīva’s translations, T2149 (LV) 253b17 (and 放牛經 is not), as in the Gu jin yijing tuji 古今譯經圖紀 T2151 (LV) 359b17–18, and Da Zhou kanding zhongjing mulu 大周刊定眾經目錄 T2153 (LV) 423c11. Then, in Kaiyuan Shijiao lu, Zhisheng identifies the two titles with one another, and lists the text as extant, among the works of Kumārajīva: 放牛經一卷 (亦云牧牛經出增一阿含第四十六卷異譯房云見別錄), T2154 (LV) 513a13. This is also the understanding of Hayashiya, Iyaku kyōrui, 366, 406 (Hayashiya rejects the ascription to Kumārajīva).

67. In the CSZJJ, the same title is elsewhere ascribed to Juqu Jingsheng, T2145 (LV) 13a12–15; and Fajing follows suit, T2146 (LV) 129b11. In addition to the ascription to Huijian, the LDSBJ also follows these earlier catalogues in featuring an ascription of this title to Juqu Jingsheng, T2034 (XLIX) 93a8–9. 68. Naitō Ryūo 内藤龍雄, “Dai hōben Butsu hō’on kyō ni tsuite” 大方便佛報恩經について, IBK 3.2 (1955): 313–15; Funayama Tōru 船山徹, Butten wa dō Kan’yaku sareta no ka: Sūtora ga kyōten ni naru toki 仏典はどう漢訳されたのか：スートラが経典になるとき (Tokyo: Iwanami shoten: 2013), 138–39; idem, “Da fangbian Fo bao’en jing bianzuan suoyinyong di Hanyi jingdian 《大方便佛報恩經》編纂所引用的漢譯經典,” tr. Wang Zhaoguo 王招國, Fojiao wenxian yanjiu 佛教文獻研究 2 (2016): 175–202. 69. Palumbo has argued that T188 probably predates T185, and is therefore early; Antonello Palumbo, “Dharmarakṣa and Kaṇṭhaka: White Horse Monasteries in Early Medieval China,” in Buddhist Asia: Papers from the First Conference of Buddhist Studies Held in Naples in May 2001, ed. Giovanni Verardi and Silvio Vita (Kyoto: Italian School of East Asian Studies, 2003), 205–7. See also Matsuda Yūko 松田裕子, “Chinese Versions of the Buddha’s Biography,” IBK 37.1 (1988): 24–33.

70. Treatment of this title is particularly complicated in the LDSBJ, which also features an ascription to the even more obscure Jianglianglouzhi 彊梁婁至 (*Kalyāṇaruci?), T2034 (XLIX) 65a8–11; and a third ascription to 興起行經 T197, Kang Mengxiang 康孟詳, 54b2. 雜譬喻經 T205, Anon., E. Han, 54b25. 猘狗經 T214, Zhi Qian, 58c7. 群牛譬經 T215, Faju, 67a6. 大魚事經 T216, Tanwulan, 69c5. 仁王般若波羅蜜經 T245, Kumārajīva, 78a23–24. 71 法華三昧經 T269, Zhiyan 智嚴, 112c27. 諸菩薩求佛本業經 T282, Nie Daozhen, 65c19. 無垢施菩薩應辯會 T310(33), Nie Daozhen, 66a2. 72 菩薩修行經 T330, Bo Fazu 白法祖, 66b4. 優填王經 T332, Faju, 67b3. 73 大乘方等要慧經 T348, An Shigao, 52b17.

寶積三昧文殊師利菩薩問法身經 T356, An Shigao, 52b10–11. 出阿彌陀佛偈 T373, Anon., E. Han, 55b24–25. 74 般泥洹後灌臘經 T391, Dharmarakṣa, 64a24. 75 迦葉赴佛般涅槃經 T393, Tanwulan, 70a19. 八吉祥神呪經 T427, Zhi Qian, 58b8. 76 八陽神呪經 T428, Dharmarakṣa, 64b4. 文殊師利般涅槃經 T463, Nie Daozhen, 65c7.

Guṇabhadra, 91b24. Terasaki has suggested that *Kālodaka himself composed the text; Terasaki Shūichi 寺崎修一, s.v. “Jūni yū kyō 十二遊經,” Bussho kaisetsu daijiten 佛書解說大辭典, ed. Ono Genmyō 小野玄妙 and Maruyama Takao 丸山孝雄 (Tokyo: Daitō shuppan, 1933–1936 [縮刷版 1999]). 71. T245 is one of the most famous examples of a scripture composed in China, but spuriously entered into the canon as a translation. The history of this ascription is somewhat complicated. An interlinear note in Fajing knows of the attempt to ascribe T245 to Kumārajīva, but overtly rejects it: “Bie lu claims that this sūtra was translated by Dharmarakṣa (!); at the head of the text, moreover, a byline states that it is a collection of buddhavacana compiled by Kumārajīva. In the opinion of the present author, the doctrinal content and style and usage of this scripture in all points appears not that of a translation by [either of] those two eminent scholars, and hence I have entered it into the catalogue of dubious [[[scriptures]]]” 別錄稱此經是竺法護譯經首又題云是羅什撰集佛語今案此經 [v.l. 語 SYM] 始末義理文詞似非二賢所譯故入疑 [+錄 SYM], T2146 (LV) 126b8–9; see Charles D. Orzech, Politics and Transcendent Wisdom: The Scripture for Humane Kings in the Creation of Chinese Buddhism (University Park, PA: Pennsylvania State Univ. Press, 1998), 75. (This is the only time in Fajing’s catalogue that bie lu is referred to in this manner.) Thus, setting aside potential problems with the interlinear notes in Fajing, the LDSBJ appears to represent not the first time the ascription of T245 to Kumārajīva 鳩摩羅什 is mentioned in the catalogues, but rather, the first time it is accepted.

72. See Kanakura Enshō 金倉円照, “Hoke kyō ni okeru Hōgo to Rajū no yakugo” 法華経における法護と羅什の訳語, in Hoke kyō no Chūgokuteki tenkai 法華経の中国的展開, ed. Sakamoto Yukio 坂本幸男 (Kyoto: Heirakuji shoten, 1972), 458. 73. The LDSBJ also lists 優填王作佛形像經 and 作像因緣經 as alternate titles for the 作佛形像, which it treats among Han anonymous texts; T2034 (XLIX) 54c2. The 優填王經, with the alternate title 作佛形像經, is treated as anonymous in fascicle 14. See also T692 (listed below). 74. Two very similar titles, 阿彌陀佛偈經 and 後出阿彌陀佛偈經, are listed among Han anonymous texts in the LDSBJ, T2034 (XLIX) 55b24–25. Both are listed again as anonymous in fascicle 13, 113b7–8. It is likely that T373 was composed in China. See Saitō Takanobu 齊藤隆信, “Kōshutsu Amida Butsu ge ni tsuite” 後出阿彌陀佛偈について, Jōdoshūgaku kenkyū 浄土宗学研究 32 (2005): 50–51; idem, “Kōshutsu Amida-Butsu ge to sono yōto” 『後出阿弥陀仏偈』とその用途, Bukkyō daigaku sōgō kenkyūjo kiyō bessatsu “Jōdokyō tenseki no kenkyū” 仏教大学総合研究所紀要別冊「浄土教典籍の研究」 1 (2006): 11–29; Nattier, Guide, 118 nn. 20, 21. 75. See n. 92.

76. Matters are unusually complicated with respect to this title in the LDSBJ. The Zhi Qian ascription is actually for the title 八吉祥經, with a note reading 見古錄亦有呪. A supposed second translation is ascribed to Guṇabhadra, 91b19–20. A very similar title, 八吉祥經, is also ascribed to *Saṅghabhara, 98b20; see T430. But the same title is also listed as anonymous in fascicle 13, 113a5. Nonetheless, this is the first time such a title is ascribed to Zhi Qian, and the present ascription of T427 therefore probably derives ultimately from this LDSBJ notice. 三曼陀跋陀羅菩薩經 T483, Nie Daozhen, 66a1. 77 六菩薩亦當誦持經 T491, Anon., E. Han, 54c19. 阿難問事佛吉凶經 T492, An Shigao, 51c22. 摩訶迦葉度貧母經 T497, Guṇabhadra, 91c26. 羅云忍辱經 T500, Faju, 66c22. 78 沙曷比丘功德經 T501, Faju, 67c13. 佛為年少比丘說正事經 T502, Faju, 67b24. 比丘避女惡名欲自殺經 T503, Faju, 67c10. 犍陀國王經 T506, An Shigao, 52b5. 阿闍世王問五逆經 T508, Faju, 67a24. 79 阿闍世王授決經 T509, Faju, 67a2. 80 採花違王上佛授決號妙花經 T510, Tanwulan, 69c12.

長者子懊惱三處經 T525, An Shigao, 50c13. 越難經 T537, Nie Chengyuan 聶承遠, 65b21. 81 樹提伽經 T540a/b, Guṇabhadra, 91c17. 摩鄧女經 T551, An Shigao, 52a6. 82 內身觀章句經 T610, Anon., E. Han, 55b4. 法觀經 T611, Dharmarakṣa, 64a21. 83 身觀經 T612, Dharmarakṣa, 64a20. 84 佛印三昧經 T621, An Shigao, 52b15. 85 自誓三昧經 T622, An Shigao, 51b5. 父母恩難報經 T684, An Shigao, 51a13. 86 盂蘭盆經 T685, Dharmarakṣa, 64a27. 未曾有經 T688, Anon., E. Han, 55a28. 作佛形像經 T692, Anon., E. Han, 54c2. 87 摩訶剎頭經 T696, Shengjian 聖堅, 83c9. 77. See Nattier, DDB, s.v. 迦羅蜜.

78. See Jan Nattier, “Now You Hear It, Now You Don’t: The Phrase ‘Thus Have I Heard’ in Early Chinese Buddhist Translations,” in Buddhism across Asia: Networks of Material, Intellectual and Cultural Exchange, ed. Tansen Sen (Singapore: Institute of Southeast Asian Studies, 2014), 45. 79. See Wu Juan, “From Perdition to Awakening: A Study of Legends of the Salvation of the Patricide Ajātaśatru in Indian Buddhism” (PhD dissertation, Cardiff Univ., 2012), 38–39, 157. 80. See Wu, “From Perdition to Awakening,” 38.

81. Again, the LDSBJ features an unusual proliferation of ascriptions for this title. In addition to the ascription to Nie Chengyuan, Fei also ascribes the variant title 曰難經 to Shi Songgong/Gaogong 釋嵩公/高公, T2034 (XLIX) 72a13. A third note ascribes the same title (giving both variant readings) to Guṇabhadra, 91c12. Finally, the same title is incongruously treated as anonymous in fascicle 14, 117b20. 82. This case is complicated by possible confusion with 舍頭諫太子二十八宿經 T1301. Elsewhere in the LDSBJ, a supposed first translation of a title corresponding to T1301 is also ascribed to An Shigao: 舍頭諫經一卷 (初出。見舊錄亦云舍頭諫太子明二十八宿經。亦云太子明星二十八宿經。亦云虎耳經), T2034 (XLIX) 51a6–7.

83. Hung has shown that T611 is largely identical with T732. Hung argues, implausibly, that T732 and T611 are by An Shigao, on the basis of “stylistic” considerations. However, the phrases he identifies are often too long to count as stylistic markers, and he pays no attention to context. It can in fact be shown that some passages also have extended parallels in An Shigao’s T602 and Kang Senghui’s 康僧會 T152(76). See Hung Hunglung 果暉法師, “Continued Study of An Shigao’s Works: The Terminology of the Mayi jing T732 and the Faguan jing T611,” IBK 55.3 (2007): 122–27(L). 84. The content of T612 is almost identical to T101(9); Nattier, Guide, 66 n. 147.

85. Nattier uses T621 as an example of ascriptions first found in the LDSBJ that are “evidently impossible”; Nattier, Guide, 15 n. 26. 86. Guang Xing argues that T684 is “authentic” (i.e., a genuine translation text), but does not support the ascription to An Shigao; Guang Xing, “A Study of the Apocryphal Sūtra: Fumu Enzhong Jing,” International Journal of Buddhist Thought & Culture 11 (2008): 105–46. 87. See n. 73.

罪業應報教化地獄經 T724, An Shigao, 51c18. 分別善惡所起經 T729, An Shigao, 51a23. 處處經 T730, An Shigao, 51b9. 88 十八泥犁經 T731, An Shigao, 51c16. 罵意經 T732, An Shigao, 51b8. 89 堅意經 T733, An Shigao, 52a2. 鬼問目連經 T734, An Shigao, 51c15. 分別經 T738, Dharmarakṣa, 64a28. 90 慢法經 T739, Faju, 66c20. 忠心經 T743, Tanwulan, 70a6. 罪福報應經 T747b, Guṇabhadra, 91c5. 91 十二品生死經 T753, Guṇabhadra, 91c4. 四輩經 T769, Dharmarakṣa, 64a24. 92 四品學法經 T771, Guṇabhadra, 91c8. 賢者五福德經 T777, Bo Fazu, 66b14. 十二頭陀經 T783, Guṇabhadra, 91b25. 出家緣經 T791, An Shigao, 51a8. 貧窮老公經 T797a/b, Huijian, 93b14. 弟子死復生經 T826, Juqu Jingsheng, 93a3. 懈怠耕者經 T827, Huijian, 93b12.

阿難陀目佉尼呵離陀經 T1013, Guṇabhadra, 92a8. 呪齒經 T1327, Tanwulan, 70b11. 93 華積陀羅尼神呪經 T1356, Zhi Qian, 58b7. 玄師颰陀所說神呪經 T1378b, Tanwulan, 70b9. 檀特羅麻油述經 T1391, Tanwulan, 70b3–70b4. 94 摩尼羅亶經 T1393, Tanwulan, 70a24, 70b1. 95 犯戒罪報輕重經 T1467, An Shigao, 51b1. 大比丘三千威儀 T1470, An Shigao, 50a23–24. 96 88. Nattier regards T730 as an “archaic translation of unknown authorship”; Nattier, Guide, 127 n. 42. 89. See n. 83.

90. A title identical to T738 features in a list of eight texts Fajing says were composed by Xiao Ziliang 蕭子良 (460–494), which he therefore includes in his catalogue of spurious texts 偽錄; T2146 (LV) 139a4, 139a7–8. However, Palumbo has suggested that it is “unlikely that it could be a wholesale concoction of the prince,” and that the “the original [text . . .] may date from the 4th c. and have been written in Central Asia”; Palumbo, Early Chinese Commentary, 201–4 and n. 48.

91. The LDSBJ entry closest to the present title of T747b, 五道輪轉罪福報應經, is treated as anonymous in the LDSBJ, T2034 (XLIX) 117b22. However, a shorter, related title, 罪福報應經一卷, is ascribed to Guṇabhadra, 91c5. This is the probable source of the received ascription of T747b. 92. This text is listed in the LDSBJ among Wei-Wu anonymous texts with various titles: 四輩經一卷 (舊錄云四事弟子經。或云四輩學經), T2034 (XLIX) 61a12. The same title is also listed as an alternate title of the 灌臘經, which is ascribed to Dharmarakṣa, 64a24. However, the 灌臘經 is still extant as T391 (and listed separately in our sources—see above). If, then, as seems likely, the present LDSBJ entry is the basis for the ascription of T769, this means that two texts have been assigned the same ascription on its basis.

93. Listed under the title 呪齒痛經. 94. In fact, we find at this point in the LDSBJ two very similar and apparently related titles, both ascribed to Tanwulan, 檀持羅麻油述神呪 and 麻油述呪經.

95. Again, we find in the LDSBJ not just one but two similar titles ascribed to Tanwulan, 摩尼羅亶神呪經 and 摩尼羅亶神咒安 [v.l. 案 SYMP] 摩經. Strickmann has shown that the eighth “book” of T1331, itself a Chinese composition or compilation, derives from T1393; Michel Strickmann, “The Consecration Sutra: A Buddhist Book of Spells,” in Chinese Buddhist Apocrypha, ed. Robert E. Buswell, Jr. (Honolulu: Univ. of Hawai’i Press, 1990), 83. 96. This case is somewhat complex. The exact title of T1470 is treated as anonymous, T2034 (XLIX) 119c3, as is a supposed second translation, 119c12. But the present ascription of T1470 to An Shigao probably dates back to the LDSBJ entry for the title 大僧威儀經四卷, 50a23–24. 沙彌尼戒經 T1474, Anon., E. Han, 54c27. 戒消災經 T1477, Zhi Qian, 58a11. 97 菩薩受齋經 T1502, Nie Daozhen, 65c18. 98 分別功德論 T1507, Anon., E. Han, 54b19. 99 阿毘曇甘露味論 T1553, Anon., Cao Wei, 60b18. 請賓頭盧法 T1689, Huijian, 93b17. 迦葉結經 T2027, An Shigao, 52b16.

appendix 2 This Appendix presents the full list of Fei’s supposed sources for the ascriptions here under study, as discussed above (all page numbers refer to T2034 [XLIX]):

– anonymous Han texts: the general note at the end of this section cites CSZJJ 僧祐律師出三藏記, Gu lu and Jiu lu 古舊二錄, and Dao’an, 55a9. – An Shigao 安世高: Jiu lu, 51a6–7; the general note at the end of this section mentions “all the catalogues” 眾錄 and “the biographies/chronicles” 諸傳, 52c4–5, c14. – *Lokakṣema 支婁迦讖: bie lu, 53a15. – Yan Fotiao 嚴佛調: the Cen Hao 岑號 catalogue, 54a20–21. 100 – Kang Mengxiang 康孟詳: the Wu catalogue 吳錄, 54b2. – Zhi Yao 支曜: the Wu catalogue, 54a8.

– Kang Senghui 康僧會: CSZJJ, 59a21. – Zhi Qian 支謙: Jiu lu, 58a11, 58a14; Baochang 寶唱, 58b7. – anonymous Wei-Wu texts: the general note at the end of this section cites only the Gu lu and Jiu lu, and then only to say that these texts are listed as anonymous in both (i.e., not necessarily as support for dating), 61b7–8. – Jianglianglouzhi 彊梁婁至 (*Kalyāṇaruci?): Shixing 始興 catalogue, Baochang, 65a8–11. – Bo Yan 白延: Shixing catalogue, Baochang, 56c14–15.

– An Faqin 安法欽: Zhu Daozu’s 竺道祖 catalogue and the “miscellaneous catalogue(s?) of the Jin era” 晉世雜錄, 65a13–20. – Dharmarakṣa 竺法護: Dao’an, 63c16–17; Nie Daozhen’s 聶道眞 catalogue, 62b8–9. – Zhi Fadu 支法度: the Zhi Mindu 支敏度 and Zhu Daozu catalogues, 68a17–18; the general note for Zhi Fadu’s texts also cites Baochang, 68a20. – anonymous Western Jin texts: the general note at the end of this section cites the Wu catalogue and bie lu 吳別二錄, 68b1. 97. The exact title as in the Taishō is listed only once in the LDSBJ, and treated as anonymous, T2034 (XLIX) 119c8. However, a 戒消伏災 [v.l. 炎 M] 經 is ascribed to Zhi Qian, 58a11, and this is the probable source of the present ascription of T1477.

98. Karashima regards T1502 as “more likely a composition made in China under the influence of the Dà āmítuó jīng 大阿彌陀經 and others”; Karashima Seishi, “On Amitābha, Amitāyu(s), Sukhāvatī and the Amitābhavyūha,” in Evo ṣudayi: Essays in Honor of Richard Salomon’s 65th Birthday, ed. Carol Altman Bromberg, Timothy J. Lenz, and Jason Neelis; Bulletin of the Asian Institute, New Series 23 (2009): 125. 99. T1507 has been very thoroughly studied by Palumbo, who argues that it was in fact composed in Chang’an between 383 and 402, but most probably between 383 and 385; Palumbo, Early Chinese Commentary; see also idem, “Models of Buddhist Kingship in Early Medieval China,” in Zhonggu shidai de liyi, zongjiao yu zhidu 中古時代的禮儀、宗教與制度, ed. Yu Xin 余欣 (Shanghai: Shanghai guji chubanshe, 2012), 313.

100. This is the only time Fei cites this catalogue, though he lists it in fascicle 15 among the catalogues he says he draws upon (including catalogues he did not see). T2149, T2154, and T2157 dutifully list this catalogue as a work that exists (or once did), but otherwise do not mention it. – Fajian/Shengjian 法堅/聖堅: the Shixing catalogue, Fashang’s 法上 catalogue, bie lu, 83c3–4; Shixing catalogue, 83c9. – Tanwulan 曇無蘭: CSZJJ; bie lu, 70a8; Jiu lu, 70a6; the general note on these translations as a group cites bie lu, 70b21. – *Kālodaka 迦留陀伽: Zhu Daozu, the “miscellaneous catalogue(s?) of the Jin” and Baochang, 70b27–c2. – Kumārajīva 鳩摩羅什: the catalogue of [[[scriptures]] translated under] the “Two Qin” 二秦錄, 78a20; Li Kuo 李廓 catalogue, 78c14; bie lu, 78a23–24. – *Nandi(n) 難提: Fashang, 72a1.

– anonymous Eastern Jin texts: the general note at the end of this section cites the CSZJJ, but only to say that Sengyou includes these titles among his newly collected anonymous scriptures, 74c16–18. 101 – Zhiyan 智嚴 and Baoyun 寶雲: Zhu Daozu (and the?) miscellaneous catalogue 竺道祖雜錄 (probably “of the Jin era,” as for An Faqin and *Kālodaka), 89c9, 89c18.

– Guṇabhadra 求那跋陀羅 (collaborating with Baoyun et al.): Li Kuo’s catalogue, 91b8; the Jiu lu, 91b24; Sengyou and Baochang, 91b19–20; the general note for these translations mentions the CSZJJ and “various catalogues” 眾錄, 92b12. – Juqu Jingsheng 沮渠京聲: CSZJJ, 93a8–9.

– Xiang gong 翔公: Shixing catalogue, Dao’an, CSZJJ, 93c25–94a1. – *Kālayaśas 畺良耶舍: Daohui’s Song–Qi catalogue 道慧宋齊錄 and GSZ, 92c10– 11. – Tanyao 曇曜: (Zhu) Daozu’s catalogue, 85a24. – Shi Tanjing 釋曇景: Baochang, Fashang and Wang Zong 王宗, 96a26. – Shi Songgong/Gaogong 釋嵩公/高公: the general note covering the three texts ascribed to him cites the Zhao catalogue 趙錄 and the Shixing catalogue, 72a13.

appendix 3

As noted in n. 10, five titles in Sengyou’s list of anonymous scriptures present particular difficulties (references to CSZJJ T2145 [LV]): – 菩薩十住行道品 [+ 經 SYP] T283, ascribed in T to Dharmarakṣa, 8b23, 22c21; – 觀世音菩薩 [+ 得大勢菩薩 SYMP] 授記經 T371, ascribed in T to Tanwujie 曇無竭, 12a28, 22b16; – 菩薩生地經 T533, ascribed in T to Zhi Qian 支謙, 7a2, 23a1; – 四天王經 T590, ascribed in T to Zhiyan 智嚴 and Baoyun 寶雲, 12c6–9, 24c17, 112c19–20; – 舍頭諫太子二十八宿經/虎耳經 T1301, ascribed in T to Dharmarakṣa, 8c17, 25b21.

In each case, the CSZJJ contains at least two entries: 1) an entry among the present list of anonymous texts in fascicle 4; 2) a separate entry (outside fascicle 4) giving an ascription for an identical title, or a title identified as pertaining to the same text. (The ascription of T590 101. Fei makes a peculiar remark at this point, which pertains to the lists of anonymous scriptures for all dynasties down to the Jin. He says, “The extant [anonymous] texts that [[[Sengyou]]] saw totaled 846 works, in 895 fascicles [exactly as at T2145 (LV) 32a1]; [I?] have already distributed them elsewhere, in the catalogues of the various dynasties [?]; I append the remainder here, deeming them to have lost their ascriptions after the Jin” 見有經本者。凡八百四十六部。合八百九十五卷。已外散入諸代世錄。所餘附此為晉下失源; T2034 (XLIX) 74c16–18. This note is difficult to understand. Does it mean that Fei himself “scattered” 散 the texts among the catalogues for the different dynasties, or that he is claiming to have found them already entered among the respective catalogues for these various dynasties, which he claims to draw upon as his sources? See n. 30.

to Zhiyan and Baoyun, moreover, is additionally corroborated in Zhiyan’s CSZJJ biography, T2145 [LV] 112c19–20.) In each case, we have only one extant text, which bears the ascription found in the alternate CSZJJ entry. For the purposes of the present analysis, I have excluded those titles, to avoid double-counting single texts.

However, with the exception of T533 (Nattier, Guide, 142–43), these ascriptions may be weak. For T283, T371, T590, and T1301, therefore, we cannot eliminate the possibility that our extant text is in fact the text listed as anonymous in CSZJJ fascicle 4, and it is the text listed with an ascription elsewhere in the CSZJJ that is lost (if indeed such a text existed). In other words, in these cases, we may be dealing with additional erroneous ascriptions dating back to the CSZJJ or before, and lent authority by the CSZJJ itself.

abbReviations CBETA Chinese Buddhist Electronic Text Association. Taishō shinshū daizōkyō 大正新脩大藏經. Edited by Takakusu Junjirō 高楠順次郎 and Watanabe Kaigyoku 渡邊海旭. Tokyo: Taishō shinshū daizōkyō kankōkai/Daizō shuppan, 1924–1932. CBReader v. 5.3, 2016. CSZJJ Chu sanzang ji ji 出三藏記集 T2145.

DDB Digital Dictionary of Buddhism. Edited by A. Charles Muller. <http://buddhismdict.net/ddb>. Editions of October–December 2016. F Fangshan shi jing 房山石經. Edited by Zhongguo Fojiao xiehui 中國佛教協會 and Zhongguo Fojiao tushu wenwuguan 中國佛教圖書文物館. Beijing: Huaxia chubanshe, 2000. GSZ Gao seng zhuan 高僧傳 T2059. IBK Indogaku Bukkyōgaku kenkyu 印度學佛教學研究. LDSBJ Lidai sanbao ji 歷代三寶紀 T2034.

M Ming [edition of the Chinese canon, indicated in T critical apparatus as 明]. P “Palace” [edition of the Chinese canon, indicated in T critical apparatus as 宮]. S Song [edition of the Chinese canon, indicated in T critical apparatus as 宋].

T Taishō shinshū daizōkyō 大正新脩大藏經, as accessed via CBETA (2016).

Taisho references follow the order: Text number (volume number, roman numerals) page|register|line number. E.g., T225 (VIII) 483b17 is text number 225, volume 8, page 483, second register, line 17.

X Shinsan dai Nippon zokuzōkyō 卍新纂大日本續藏經, as accessed via CBETA (2016). References formatted as for T.

Y Yuan [edition of the Chinese canon, indicated in T critical apparatus as 元]