杏雨书屋藏敦煌《大般若经》写本缀合研究

A Study on Reconstructing Dunhuang Manuscripts of Mahāprajnāpāramitā-sūtra Collected in Kyōu Shooku

日本杏雨书屋《敦煌秘笈》收录了李盛铎等中国名家旧藏的敦煌真品写本精华,透露出藏经洞文物流散的信息。杏雨书屋与其他馆藏的49号《大般若经》写本可以缀合为16组,从中可以得到以下几点认识:杏雨书屋藏卷在被李盛铎等中国藏家收藏时经过了精心挑选;它们与国图藏卷关系最为密切;可缀合的杏雨书屋藏卷在藏经洞开启前可能已多残缺;伴随1910年劫余写卷解京,写卷再次遭到损毁,但杏雨书屋藏卷与国图藏卷碴口相接的撕裂应非李盛铎等中国藏家所为。

The Dunhuang Secret Documents(Tonkō Hikyū) includes the essential portion of genuine Dunhuang manuscripts collected by Haneda Tooru in the Kyōu Shooku, which were formerly collected by such famous Chinese collectors as Li Shengduo, He Yansheng, Wang Shunan, Xu Chengyao, Zhang Guangjian, and others. However, except some intact manuscripts, most of them are fragmentary, and among these fragments, quite a few are from the same scrolls that were torn apart. Via piecing fragmentary Dunhuang manuscripts from the Kyōu Shooku together with those from other institutes, we can investigate the characteristics of the manuscripts collected in the Kyōu Shooku and explore the history of how these cultural relics were scattered from the Library Cave.Based on a general survey of the Dunhuang documents that have been released by public and private institutes all over the world, according to their successive content, the match of fragmentary characters, the same formats, similar writing styles or scripts, we are able to piece forty-nine fragments of Mahāprajnāpāramitā-sūtra together as sixteen groups, namely:(1) BD6779+H531+S.9146…BD3157(This article use "H" as the abbreviation for Haneda Tooru collection in the Kyōu Shooku),(2) Дх.12468…H255,(3) H343…S. 4097,(4) BD5077…H450,(5) BD11905…BD10096+BD4854+H520,(6) BD4286+H265,(7) BD345+H668,(8) Дх. 5663+H644,(9) H243…S. 9223,(10) Дх. 10919…BD6023+H423,(11) H522…BD6635,(12) S. 7288…H195,(13) H394+BD1300,(14) BD10879+H345,(15) H248…Дх. 5831+ Дх. 5869+ Дх. 8781,(16) BD11970…BD12138+BD12149…C. 91…H88-2~8… "Дх. 1784,Дх.1830, Дх.1855"…Дх.217+BD11812…Дх.18066…Дх.6852. While these reconstructions reunite the scattered "bones and flesh" of Dunhuang manuscripts, and therefore restore the original forms of the manuscripts from the Library Cave, they also help with further researches on identifying and dating these fragmentary Dunhuang manuscripts as well as researches on the characteristics of their paper and scribal practices.To summarize, we can learn four points from the sixteen groups of fragments reconstructed above.(1) Compared with the fragments that can be reconstructed in the collections of Dunhuang manuscripts in the National Library of China(NLC), in Britain and Russia, those collected in the Kyōu Shooku include more manuscripts that contain scripture titles and colophons, and more long scrolls. Comparatively,in these manuscripts in the Kyōu Shooku collection, the writing is neat, the calligraphy is good, and there are many characters that were introduced by Empress Wu Zetian, for example, zheng() and shou(). These characters have long been regarded as a kind of symbol of manuscripts from the early Tang Dynasty by collectors. The above features reflect that the manuscripts collected in the Kyōu Shooku were carefully selected by Chinese collectors such as Li Shengduo.(2) From the amount of fragments that can be pieced together, and how well they could be matched, the collection in the Kyōu Shooku is most closely related to the collection in the National Library of China.(3) The fact that only one of the above reconstructed manuscript is complete while the other fifteen manuscripts are all incomplete to some degree indicates that most of the manuscripts may have been damaged before the Library Cave was opened, and the manuscripts collected in the Kyōu Shooku are not exceptional.(4) Among the fragments that have been pieced together, some fragments belong to the collection in Russia. It means that after Dunhuang documents were looted those left in the Library Cave were not thoroughly collected when they were packed and sent to Beijing in 1910, and the rest left in the cave were eventually collected by the Russian Oldenburg expedition team. Based on the fact that fragments from the Kyōu Shooku and those from the NLC can be pieced together in four groups, we learn that the manuscripts sent to Beijing were damaged again. However, these damages happened before Chinese collectors such as Li Shengduo had the opportunity to steal the manuscripts. Otherwise, these experts would have obtained more complete manuscripts before they were preserved rather than those fragments found scattered in China and Japan today.