Abstract：It is well known that Li Shengduo and the others stole Dunhuang scrolls transported from Gansu to Beijing, which is a big case in the academic history of Dunhuang. Previous research is based on historical information retold by outsiders. Although many people reach the consensus, the facts and details are still unclear, and it is difficult to make a definitive case. The key to solving the case lies in proving that the possessions of Li Shengduo and the others was not legitimate. The focus of investigation and evidence collection does not lie in the stolen scrolls themselves, but on the Dunhuang fragments preserved in the National Library of China, which were torn to cover up the theft. Any fragment that shows the obvious sign of being torn and can be directly pieced together are immediate evidence of the theft.
By analysing two groups of Dunhuang scrolls in the collection of the NLC, namely BD4416+BD4438+BD4410+BD4408+BD4474+BD4503+BD4514+BD4328+BD4349+BD4521+BD4340+BD4347+BD4455+BD4513+BD4754+ BD4760+BD4436 (the Mahāpraj?āpāramitā-sūtra, Fascicle Fifty-two) and BD2376+BD2356+BD2734+BD2570+BD2395+BD2434+BD2783+BD2388+BD2396+BD2786+BD2374+BD2518+BD2703+BD2702+BD2652+BD2398+BD2402+BD2428+BD2414+BD2640+BD2772 (the Lotus Sūtra, Fascicle Two), it can be found that each group of the torn scroll was used for a very short period of time, and that the manner of breakage is very similar, the length of each fragment is roughly close to that of the others, and the characters of the Qianziwen (千字文) that encode the fragments are relatively centralized. The sign of being torn between the fragments clearly show the characteristics of human intentionality.
Similar to the two cases mentioned above, the other 34 groups, totaling 275 fragments with more than six fragments in each group, also show striking similarities in terms of the length of the fragments, their state of preservation, how they were torn up, and the distribution of the characters of the Qianziwen. These huge numbers of artificially torn fragments are typical evidence of the organized efforts of Li Shengduo and the others to cover up the theft of the Dunhuang scrolls.
Fortunately, the original inscription of Wu Xiangchen, a member of the escort committee, at the end of Dunhuang manuscript C14.701 in the collection of the National Museum of China, describes in detail the process of transporting the Dunhuang scrolls to Beijing, and provides important documentary evidence as to why Li Shengduo and the others employed the above-mentioned method of destruction. The inscription says that “the whole scroll is divided into eighteen boxes; each box of four hundred and twenty scrolls; each scroll is six meters long”， proving that the scrolls reconstructed from the fragments were originally long scrolls, and that the destruction was man-made, and each box of scrolls was numbered using four or five characters of the Qianziwen. This is precisely the reason why the word size of the above fragments of the scrolls are mostly below three to five.
Incidentally, it should be noted that some scholars have maintained, through the reconstruction of the fragments, that the torn parts between the collection of the NLC and the old collection of Li Shengduo were made by Li. This claim is difficult to believe due to a lack of analysis and argumentation.
徐浩. 李盛铎等人窃取敦煌写卷的证据[J]. 浙江大学学报(人文社会科学版), 0, (): 1-.
Xu Hao. The Evidence of Stealing Dunhuang Scrolls by Li Shengduo and the Others. JOURNAL OF ZHEJIANG UNIVERSITY, 0, (): 1-.
