Manuscripts of the Mahāparinirvāṇa-mahāsūtra from Dunhuang: Preliminary Arrangement According to its Scroll Division

Abstract

The paper considers one of the methods of manuscript classification applied to the Chinese translation of Mahāparinirvāṇa-mahāsūtra from Dunhuang. Given the fact that the beginnings and endings of some scrolls of its different versions do not correspond, researchers identify several types of scroll division (fen juan 分卷). This paper attempts to reconstruct one of these types based on Daboniepanjing chao 大般涅槃經鈔 (“Digests of the Mahāparinirvāṇa-mahāsūtra”) manuscripts from Dunhuang.

Full Text

Semyon Ryzhenkov Manuscripts of the Mahāparinirvāṇa-mahāsūtra from Dunhuang: preliminary arrangement according to its scroll division Abstract: The paper considers one of the methods of manuscript classification applied to the Chinese translation of Mahāparinirvāṇa-mahāsūtra from Dunhuang. Given the fact that the beginnings and endings of some scrolls of its different versions do not correspond, researchers identify several types of scroll division (fen juan 分卷). This paper attempts to reconstruct one of these types based on Daboniepanjing chao 大般涅槃經鈔 (“Digests of the Mahāparinirvāṇa-mahāsūtra”) manuscripts from Dunhuang. Key words: Chinese Buddhism, Dunhuang, manuscripts, Mahāparinirvāṇa-sūtra, Mahāparinirvāṇa-mahāsūtra, scroll division, digests of sūtras, 北敦 6363, 北敦 3386, 北敦 2838 The Mahāparinirvāṇa-mahāsūtra (Daboniepanjing 大般涅槃經, hereinafter - MPNMS) is believed to have been written around the 2nd or 3rd c. AD. The full Sanskrit version of the MPNMS has not remained intact. The MPNMS was an important scripture among the Buddha-nature corpus of texts since it was the first of this kind to reach China, and it played a significant role in the dissemination of the Buddha-nature doctrine. There are two full versions of the sūtra, known as Northern (beiben 北本) and Southern (nanben 南本), both of which are found in Dunhuang cave library. The Northern version1 is a translation of Dharmakùema (Tanwuchen 曇無讖, 385-433) made between AD 421 and 430.2 It consisted of 40 volumes (juan 卷) and was completed in two stages: first, a text of 10 volumes was translated, which corresponded to approximately six volumes of an earlier translation by Buddhabhadra in terms of volume and content; second, the translation of the remaining 30 volumes was completed. The text of the © Semyon Yurievich Ryzhenkov, Institute of Oriental Manuscripts, Russian Academy of Sciences 1 T. 374. 2 CHEN JINHUA 2004, 215-263. MPNMS is heterogeneous. The researchers believe that its second part was written later. The first 10 volumes are sometimes called “the core portion” (qianfen 前分). The Southern version[26] was made, based on the “Northern” one, in AD 436 by Huiyan 慧嚴 (363-443), Huiguan 慧觀 (375?-445?) and others. The text was split into chapters in the same way as in the Buddhabhadra (Fotuobatuoluo 佛陀跋陀羅, 359-429) and Faxian 法顯 (337-422) six-volume translation with some minor stylistic changes. The translation consisted of 36 volumes, mainly due to the greater amount of text in each scroll rather than abridgements. Preliminary figures indicate that the total number of MPNMS manuscript fragments from Dunhuang is over 3,000 items. The archive of the National Library of China possesses the largest number of fragments and full scrolls of the sūtra (over 700 items). The vast majority of Dunhuang copies of the sūtra contain the text of its Northern version. However, sometimes, with a small fragment, we cannot establish with certainty which of the two versions it belongs to. It is also impossible to work out even the approximate number of copies solely on the basis of these data, since the manuscripts are represented both by full scrolls and fragments of different size, some very small indeed. We can get more accurate information by putting the fragments together, but many of them do not fit together precisely, so sometimes we can only make assumptions that they belong to one and the same copy on the basis of the handwriting, paper etc. We face the same problems when attempting to correlate the various volumes of the sūtra. Apart from differences in paper, handwriting, sheet size, etc., the sūtra copies are also distinguished by differing scroll divisions (fen juan 分卷). In other words, while the overall number of volumes is the same (40), the beginning and end of some scrolls do not match those in other copies. The present paper attempts to classify the surviving copies of the Northern version by scroll division type. Obviously, we can only classify those manuscripts that have either the beginning or the end, or full scrolls. For that purpose we need to reconstruct the possible types of scroll division, a task which is made possible thanks to a number of Dunhuang manuscripts. Among them is a series of documents containing a list of MPNMS volumes with indication of their beginnings (tou 頭) and ends (wei 尾). We believe that these documents served as a check list for the monks who copied the sūtra to help them in the standardization of the text.[27] These are the following manuscripts: P.3150, P.5047 (held in the National Library of France), S.1361 (held in the British Library), 北 6612v (held in the National Library of China) and Ф-271 (held in the IOM, RAS). Their contents were deciphered and published by Fang Guangchang,[28] so I am not going to include that process the present paper. The data provided by the manuscripts show four possible types of scroll division. Jing Shengxuan made up a classification table, in which the sūtra manuscripts were sorted by these types of division.[29] His research has shown that a considerable portion of the manuscripts do not accord with any of “check lists” in the five aforementioned manuscripts. We should also note that none of these types of division represented by the Dunhuang lists of MPNMS accord with the Taishō Tripiñaka version. Do such manuscripts represent a new type of division, or they are just variations of the existing ones? To clarify this issue, I decided to analyze a number of manuscripts labelled and catalogue as Daboniepanjing chao 大般涅槃經鈔 or Daboniepanjing yiyao 大般涅槃經義要 (“Digests of the Mahāparinirvāõa-mahāsūtra”). The published catalogues of Dunhuang collections contain over twenty manuscripts that have been given these labels by modern catalogue compilers. Most of them date from approximately 7th-8th cc. AD. They consist of MPNMS fragments arranged in an order that differs from the canonical version. Amongst these documents three typologically different kinds of texts are found - a) wasted pages (marked with dui 兌 “deleted”) conglutina- ted together; b) random writings; c) well-organized sūtra extracts (yiyao 義要). Making digests of sūtras was quite common in medieval China. Nevertheless, bibliographers tended to regard such texts negatively, and digests were placed in the category of apocryphal texts and dubious sūtras. Sengyou 僧祐 (445-518) expressed concerns that two such texts, which he dated as being from the reign of Emperor Wu of Southern Qi (483-493), while not fake and promoting the teaching, might at some point in the future be mistaken for the original.[30] For our purposes we consider the following manuscripts: 北敦 6363 (北 6604), 北敦 3386 (北 6610) and 北敦 2838 (北 6607) from the collection of the National Library of China. All three take the form of a digest made up of quotations from the “core portion” of the “Northern” version of the sūtra (1-10 vols.) that were carefully copied in the order of the canonical version, divided by titles with the volume numbers and have been dated to around 7th-8th cc. The colophon of some lost MPNMS manuscript dated the equivalent of AD 721 reads: (開元九年臘月十三日馬奉錄於此經中略取要義) On the 13th day of the 12th month of 9th year of Kaiyuan reign, Ma Fenglu slightly extracted the essentials from this sūtra.[31] Of course, we cannot therefore conclude that three manuscripts in the Beijing collection are the “essentials” made by a certain Ma Fenglu in 721, nor indeed can we judge the authenticity of that colophon. Moreover, in the various digests the quotes from the MPNMS are not always identical, but generally include the same fragments with few differences. However, the date of this colophon is in line with the estimated dating of these manuscripts, which might also prove that the making of such digests of the sūtra was practiced in the 7th-8th cc. The following table presents a comparison of the technical characteristics of these three manuscripts: 北敦 6363 北敦 3386 北敦 2838 condition beginning mutilated beginning mutilated beginningand end both mutilated content Preface (mtd); MPNMS quotations (vols. 1-10) MPNMS quotations (vols. 4-10) MPNMS quotations (vols. 3-6) titles each vol., except the first (卷第二, 卷第三 etc.) each vol. (卷第五, 卷第六 etc.) vols. 4, 5. The title of vol. 6 is omitted (卷第四, 卷第五) dating 7th-8th cc., Tang. dynasty 7th-8th cc., Tang. dynasty 7th-8th cc., Tang. dynasty. script kaishu kaishu kaishu length 3.4+1245 cm 1061 cm 5.5+260 cm width 26 cm 25.5 cm 28.1 cm length of a single sheet 45.5-46.2 cm 40 cm 36.5-37 cm 北敦 6363 北敦 3386 北敦 2838 top margin no data ~2.8 cm ~2.8 cm bottom margin no data ~3.4 cm varies width of frame no data ~1.85 cm varies characters per line no data 17 17 lines per sheet ~28 23 22-24 lines (total) no data 605 167 These characteristics can help us to reconstruct the presumed type of scroll division in the core sūtra part of the original text that served as a source. For example, in the 北敦 6363 manuscript the last quotation from the first volume ends with 却坐一面[32], while the text after the second volume title 卷第二 begins with 爾時娑羅10 which might not accord with the division common for all extant versions of this volume. In all known versions the second volume starts with 爾時會中11. The vast majority of copies of the first and the second sūtra volumes share this common division. The only exception is the S.3707 (MPNMS vol. 1), which ends with 皆亦如是12. The second volume of this set ought therefore to begin with 爾時會中[33], so the S.3707 list should belong to the same divisional type as the master copy that served as a source for 北敦 6363. The data obtained are best presented as a table. I have used alphabetical labels to identify the types of division given in the aforementioned “check lists”: а. 北 6612v, also S.1361 and Ф-271, b. 北 6612v, c. P.3150, d. P.5047. The type of division reconstructed from the “Digests of the Mahāparinirvāõa-mahāsūtra” is shown here as (e). The cells with shelfmarks contain the ending of the last quotation of the volume and the beginning of the next one. The (e) type is highlighted in grey and in cases where it accords with other types the corresponding cells are also highlighted in grey. vol. number type of division beginning line (according to Taishō, vol. 12) type of division concluding title (according to Taishō, vol. 12) 1 Taishō 365c06 Taishō 371c08 a 365c06 a 371c08 b b c c 14 d d e 365c06 e 371b11 北敦 6363 365c07 北敦 6363 366a16~371b12 2 Taishō 371c08 Taishō 379a06 a 371c08 a 379a06 b b c c d d e 371b12? e 379a06 北敦 6363 366a16~371b12 北敦 6363 377b22~379c[34] 3 Taishō 379a13 Taishō 385b06 a 379a13 a, e 384c25 b b 385b13 c c d d e? 379a13? 3.5? ? 3.6? ? 北敦 6363 377b22~379c14 北敦 6363 384c25 北敦 2838 ? 北敦 2838 384c25~c27 vol. number type of division beginning line (according to Taishō, vol. 12) type of division concluding title (according to Taishō, vol. 12) 4 Taishō 385b13 Taishō 390b08 a, e 384c27 a, e1 390b13 b 385b13 b c c d ? d ? ? e2? 391b05 ? ? ? 391b29 北敦 6363 384c25 北敦 363 395b29~c17 北敦 3386 389b9~395b27 北敦 2838 384c25~c27 北敦 2838 391a10~391b6 5 Taishō 390b15 Taishō 396c11 a, e1 390b15 a 396c10 b b c c d d e2? 391b06(?) e 398a12 ? 391c03(?) 北敦 6363 395b29~c17 北敦 6363 398a12~398b12 北敦 3386 389b9~395b27 北敦 3386 397b27~398b12 北敦 2838 391a10~391b6 北敦 2838 ? 6 Taishō 396c18 Taishō 402c11 a 396c18 a 402с10 b b c c 404a29 d d vol. number type of division beginning line (according to Taishō, vol. 12) type of division concluding title (according to Taishō, vol. 12) e 398a13 e 404a29 (?)Дх-3369 396c06 (?)Дх-3369 ? 北敦 6363 398a12~398b13 北敦 6363 403a14~406b03 北敦 3386 397b27~398b13 北敦 3386 403a14~406b03 7 Taishō 402c18 Taishō 408c23 a 402c18 a 411a06 b b 408c22 c 404b01 c d d e 404b01 e 411a06 ? ? ? 411b16(?) 北敦 6363 403a14~406b03 北敦 6363 410b29~411b25 北敦 3386 403a14~406b03 北敦 3386 410b29~411b25 8 Taishō 409a15 Taishō 409a15 a 411a07 a 417b13 b 409a19 b 416a10 c c 417b13 d d e 411a07 e? 417c01 P.2342 411b17 ? ? 北敦 6363 410b29~411b25 北敦 6363 417a29~417c01 北敦 3386 410b29~411b25 北敦 3386 417a29~417c01 9 Taishō 416a18 Taishō 422b27 a 417b14 a 422b27 b 416a18 b vol. number type of division beginning line (according to Taishō, vol. 12) type of division concluding title (according to Taishō, vol. 12) c 417b14 c d d e 417c01 e 422b27 北敦 6363 417a29~417c01 北敦 6363 422b12~423a4 北敦 3386 417a29~417c01 北敦 3386 422b12~423a4 10 Taishō 422c06 Taishō 428b13 a a b b c c d d e 422c06 e ? ? 432a06 北敦 6363 422b12~423a4 北敦 6363 428b12~ 北敦 3386 422b12~423a4 北敦 3386 428b12~ The data obtained make it possible to classify the sheets according to their types of scroll division. The following table has been borrowed from the work of Jing Shengxuan and updated with the newly data. Where possible, the shelfmark is accompanied by an approximate dating. The shelfmarks are given in Chinese characters (for Chinese collections). The manuscripts from the National Library of China have two sorts: old (北) and new (北敦), apart from newly catalogued items for which only the new type is used. type of division beginning or/and ending line (according Taishō, vol. 12) shelfmarks of manuscripts from Dunhuang juan 1 (e) 如是~如是 (365c06~371b11) S.3707 (~500) type of division beginning or/and ending line (according Taishō, vol. 12) shelfmarks of manuscripts from Dunhuang (a) (b) (c) (d) ?~灑地 (?~371c08) S.1317, 甘圖26, S.3707, 北 6287 (北敦 6298), S.1550, 津藝 200, S.3153, S.6943, 北 6285 (北敦 845) + 北 6289 (北敦 544) + 北 6290 (北敦 686) juan 2 (e) 爾時~是法 (371b12(?)~379a05) 北敦 14507 (5-6th cс.) (a) (b) (c) 爾時~是法 (371c14~379a05) 北敦 14954 (7-8th cс.) 爾時~? (371c14~?) 北敦 14954 (7-8th cс.) ?~是法 (?~379a05) 北 6293 (北敦 2322) (6th c.), S.829 (7th c.), 北 6295 (北敦 1997) (5-6th cс.), S.4500 (7-8th cс.), S.6098, 北敦 14507 (5-6th cс.). juan 3 (a) (e) ?~色象 (?~384c25) S.2835 (6th c.), S.2876 (early 6th c.), 北 6299 (北敦 4355) (6th c.), 北敦 14946 (6th c.), 北敦 15323 (8-9th cс.), 北敦 15151 (6th c.) (b) (c) ?~所覆 (?~385b06) 北敦 15323 (8-9th cс.), 北 6302 (北敦 2370) + 北 6300 (北敦 7654) + 北 6300 (北敦 7654) + 北 6303 (北敦 7516) + 北 6304 (北敦 2726) + 北 6307 (北敦 7462) (5-6th cс.), S.4720, 北 6298 (北敦 1215) (6-8th cс.), S.172 (7th c.), Ф-184 (8-9th cс.), S.6742 (7th c.), 北敦 13842 (8-9th cс.), 北敦 14459 (7-8th cс.) juan 4 (e2) ?~墮落(?~391b05) 北 6308 (北敦 6588) (5-6th cс.), 津藝 022 (522), 北 6309 (北敦 7949)(5-6th cс.) (b) (c) (e1) 佛復~那含 (385b13~390b07) 北敦 13843 (9-10th cс.) type of division beginning or/and ending line (according Taishō, vol. 12) shelfmarks of manuscripts from Dunhuang (a) (b) (c) ?~那含 (?~390b07) 北 6306 (北敦 1946) (7-8th cс.), 北 6305 (北敦 5261) (8th c.), 北 6311 (北敦 2676) + S. 433 (7-8th cс.), S.2115, 北 6538 (7-8th cс.) (?) ?~涅槃 (?~391b29) S.3518 (588) juan 5 (e) ?~惡道 (?~398a12) 北 6316 (北敦 1131) (7-8th cс.), 北 6317 (北敦 3405) (5-6th cс.), 北 6319 (北敦 5733) (5-6th cс.), 北 6318 (北敦 1038) (5- 6th cс.), 中央圖書館 81 (e1) 爾時~惡道 (390b15~398a12) 北敦 13874 (8-9th cс.), 北敦 13875 (7- 8th cс.) (a) (b) (c) 爾時~正法 (390b15~396c10) S.1966 (7-9th cс.), 北 6539 (北敦 663) (9-10th cс.), 北 6321 (北敦 2760) (7- 9th cс.), S.5384 (7-9th cс.), 北敦 14949 (708) juan 6 (e) (c) 爾時~菩薩 (398a13~404a29) S.2393 (6th c.), 北 6323 (北敦 1470) (7- 8th cс.), S.2864 (7th c.), 中央圖書館 73 (a) (b) 善男~經典 (396c14~402c10) 北 6324 (北敦 3173) (8th c.), 北 6325 (北敦 3975) (9-10th cс.), 北敦 13844 (8- 9th cс.) juan 7 (e) 復次~思議 (404b01~411a06) 北敦 13845 (7-8th cс.), 中書店14 (7th c.) (a)(e) ?~說已 (?~411a06) S.67 (6th c.), 北 6327 (北敦 3430 (5-6th cс.), 北 6334 (北敦 1209) (7-8th cс.), 津藝 328, 北敦 14484 (7-8th cс.) (a) 爾時~說已 (402c18~411a06) 北 6326 (北敦 1358) (6th c.), 北 6326 (北敦 1358) (6th c.) type of division beginning or/and ending line (according Taishō, vol. 12) shelfmarks of manuscripts from Dunhuang (b) (c) ?~思議 (?~408c22) S.6573 juan 8 (?) (e)? 善男~涅槃 (411a07~417c01) S.883, 北 6542 (北敦 89), Ф-82, 北 6330 (北敦 1983), P.2205, S.4876, 散 68, 北敦 14464, 北敦 14550, 中央圖書館 82, Ф-82 (7-9th cс.), 北敦 1983 (7-8th cс.), S.4876, 北敦 14464 (7-8th cс.), 北敦 14550 (7- 8th cс.), S.883 (7th c.), 北 6542 (北敦 89) (7-9th cс.) (a) 善男~滅沒 (411a07~417b13) 北敦 13846 (8-9th cс.) (a)(c) ?~滅沒(?~417b13) S.6942 (7th c.) (b) ?~等鳥 (?~416a10) S.130, Ф-74 (7-9th cс.), 北 6333 (北敦 3653) (7-8th cс.) juan 9 (e) 迦葉~不久 (417c01~422b27) S.93 (7th c.), 北 6543 (北敦 3714) (6- 7th cс.), 上博 4 (7-8th cс.), 上博 4 (7-8th cс.), 上博 61 (7-8th cс.), 北 6335 (北敦 2136), 北敦 13847 (8th c.) (b) 復次~不久 (416a18~422b27) S.4788, S.6510 (6th c.) juan 10 the scroll division is the same for all manuscripts Different versions of MPNMS manuscripts were circulating in Dunhuang during the entire period spanned by the Dunhuang manuscripts collection. For its core portion there are versions with at least four different scroll divisions ((a), (b), (c), and (e) versions). The (e) type copies were circulating from the 5th c. till 10th c. or later, but the majority date from the 5th-6th cc., with some (considerably fewer less) from the 7th-8th cc. By contrast, manuscripts belonging to the (b) type were copied mostly in later centuries (8th- 10th cc.). The “Essence of MPNMS” manuscripts were copied from the (e) version. The scheme of its division is presented below. 1. from 如是 to 如是 (365c06-371b11) 2. from 爾時 to 是法 (371b12-379a06?) 3. from 佛復 to 色像 (379a13?-384c25) 4. from 爾時 to 那含 (384c27-390b07) (e1) / from 爾時 to 墮落 (384c25-391b05) (e2)[35] 5. from 爾時 to 惡道 (390b15-398a12) (e1) / from 迦葉 to 惡道 (391b06-398a12) (e2) 6. from 爾時 to 菩薩 (398a13-404a29) 7. from 複次 to 說已 (404b01-411a06) 8. from 善男 to 涅槃 (411a07-417c01) / from 善男 to 滅沒 (411a07- 417b13?) 9. from 善男 to 不久 (417c01-422b27) / from 複次 to 不久 (417b14?- 422b27) 10. from 爾時 to 病人 (422c06-428b13) From this preliminary classification, we can only know that all types of division were probably in use in all periods when the Dunhuang cave library was accumulating its stocks (5th to 10th cc.). That means that these different types were not standardized from the very beginning of sūtra circulation until the library was sealed up in the first part of 11th c. The modest attempts to standardize the copying process that were made by unknown scribes did not change the overall situation. Moreover, the division into volumes of xylographic editions of Chinese Tripiñakas produced in the following centuries is not uniform either and might be compared with manuscript from Dunhuang. From time to time the copyists tended to deal quite freely with texts, randomly splitting them in order, for example, to save paper. Further investigation will allow us to produce a more detailed reconstruction of the history of the MPNMS text. Abbreviations MPNMS: Mahāparinirvāõa-mahāsūtra T.: Taishō Buddhist Canon mtd: mutilated References CHEN Jinhua 2004: “The Indian Buddhist Missionary Dharmakùema (385-433) - A new dating of his arrival in Guzang and of his translations”. T’oung Pao, XC, 215-263. FANG Guangchang 方廣錩 1997: Dunhuang fojiao jinglu jijiao 敦煌佛教經錄輯校 [Consolidation and comparison of Buddhist catalogues from Dunhuang]. Vol. 1-2. Nanjing: Jiangsu guji chubanshe. IKEDA On 池田温 1990: Chūoku kodai shahon shikigo shūoku 中国古代写本識語集録 [Collected colophons of ancient Chinese manuscripts]. Tokyo: Tōkyō daigaku tōyō bunka kenkyūjo. Dunhuangxue da cidian 1998: 敦煌學大詞典 [Big dictionary of Dunhuang studies]. Ed. by Ji Xianglin 季羨林. Shanghai: Cishu chubanshe. JING Shengxuan 景盛軒 2009: Daboniepanjing yiwen yanjiu《大般涅槃經》異文研究 [Study of textual discrepancies of the Mahāparinirvāõasūtra]. Chendu: Sichuan chuban jituan bashu shushe. KAMATA Shigeo 鎌田茂雄 2010: Zhongguo fojiao tongshi 中國佛教通史 [Comprehensive history of Chinese Buddhism]. Tr. to Chin. by Guan Shiqian 關世謙. Taipei: Foguang chubanshe. KUO Liying 2000: “Sur les apocryphes bouddhiques chinois”. Bulletin de l’Ecole français d’Extrême-Orient 87, 677-705. Taishō: 大正新脩大蔵經 Taishō shinshu daizōkyō [Taishō Revised Tripiṭaka]. Ed. by