Published online by Cambridge University Press: 26 March 2015
From the outside, it would almost seem as if the authors and the editor of New Sources of Early Chinese History (hereafter, Sources) had completed their job before it was due. After all, this volume, which is no. 3 in the Early China Special Monograph Series,1 is the first Western book exclusively devoted to introducing a broad range of Chinese epigraphical materials: inscriptions in bones, turtle plastrons, and bronze as well as manuscripts written on stone tablets, wooden and bamboo strips, and silk.
A review of Edward L. Shaughnessy, ed., New Sources of Early Chinese History: An Introduction to the Reading of Inscriptions and Manuscripts (Berkeley: Society for the Study of Early China and Institute of East Asian Studies, University of California, Berkeley, 1997), provides the framework for this article. I am grateful first and foremost to Donald Harper, who initially invited me to write an article about early Chinese manuscripts and who with his encouragement saw me through times when I thought I could never finish it. He has also proved unfailing in providing needed information and materials. The decision to write this article in the form of a review has been my own, however, and neither Prof. Harper nor anyone else except me is responsible for the opinions and criticism expressed herein. In addition, I would like to thank Ulrich Lau, Peng Hao , Chen , Lin Suqing , and Tomiya Itaru for their sharing of valuable information. The Academia Sinica in Taibei has been an excellent environment for compiling the materials in this article. Final additions were made to this review in May 2000.
1. Published by the Society for the Study of Early China and the Institute of East Asian Studies, University of California, Berkeley.
2. Hummel, Arthur W., ed., Eminent Chinese of the Ch’ing Period (Washington, D.C.: U.S. Government Printing Office, 1943–44), 827 Google Scholar; Sources, 6.
3. Loewe, Michael, ed., Early Chinese Texts: A Bibliographical Guide (Berkeley: Society for the Study of Early China and Institute of East Asian Studies, University of California, Berkeley, 1993)Google Scholar.
4. Texts mentions manuscripts and inscriptions from early as well as medieval times for the following entries (where additional manuscripts exist that Texts does not mention, I add in parentheses the names of the sites that produced these manuscripts and the numerical code in the Database of Early Chinese Manuscripts introduced below, pp. 306–36): Chuci , 53; Chunqiu and commentaries, 71–73; Guanzi 248–49 (see also Yinqueshan, Database VI.01); Hanshu , 131; Heguanzi , 138–39; Huangdi neijing , 199–200; Laozi , 281-86 (see also Guodian, Database III.04); Liezi , 301; Lunyu , 319 (see also Mawangdui, Database V.17; Tokushima, Database IX.03); Mozi , 339 (see also Changtaiguan, Database III.02); Shangshu , 383; Shiji , 407; Shijing , 420; Shuowen jiezi , 435; Sunzi bingfa , 448–49 and 453–54; Xiaojing , 148–49; Yanzi chunqiu , 484; Yijing , 223 (see also Wangjiatai, Database IV.06); Yili , 241–42; Zhanguoce 31, 8–9; Zhuangzi , 61–62 (see also Zhangjiashan, Database V.09). Not mentioned in Texts are early Chinese manuscripts identical with or similar to Da Dai Liji , Guoyu , Kongzi jiayu , Liji , Shuoyuan , Xinshu , and others.
5. It is also noteworthy that, having appeared in the same series, the two volumes obviously owe their existence and outlook to a large extent to more or less the same individuals. This is not to disregard or belittle the considerable contribution of those authors who contributed to only one of the volumes. Judging from the tables of contents, four scholars in particular (E. Shaughnessy, M. Loewe, A.F.R Hulsewé, and W.G. Boltz) contributed almost 50% (Sources) or 40% (Texts) of the text in each volume. Three of them are also credited with editorial work in one or both of the volumes (Texts, ix–x).
6. A number of traditionally received texts that are possibly or even certainly of Han or Pre-Han origin have not been discussed in Texts, but these are generally of fragmentary nature and/or of disputed origin. For a listing and short discussions of these, see Wanli, Qu 13 , Xian Qin wenshi ziliao kaobian in Qu Wanli quanji , vol. 4 (Taibei: Lianjing, 1983), 451–503.Google Scholar However, Qu's identification of certain texts as forgeries of later times need not inevitably be followed, as recent archaeological findings have shown. In a number of instances. Texts also fails to mention important secondary studies. A detailed discussion of these has to be reserved for an exclusive review of Texts.
7. This is probably not a coincidence since the bibliography is entitled “Suggestions for Further Reading.” But some references are given in both the essays and the bibliography, a practice that one wishes had been followed for all cited references.
8. This argument was provided by Donald Harper in a personal communication.
9. See Zhongguo kaoguxue nianjian 1992 1992, 326; and Zhongguo wenwubao , January 5, 1992, 1–2.
10. This may be the reason Loewe avoids both “inscriptions” and “manuscripts” throughout his essay and rather uses the neutral “documents.”
11. For inscriptions on mirrors, see the research by Suqing, Lin , “Liang- Han jingming chutan” , Zhongyang yanjiuyuan Lishi yuyan yanjiusuo jikan 63.2 (1993), 325–70 Google Scholar; and “Liang-Han jingming huibian” Guwenzixue lunwenji (Taibei: Guoli bianyiguan, forthcoming). For inscribed weights and measures of volume made of metal or clay, see Long, Qiu et al., Zhongguo gudai duliangheng tuji (Beijing: Wenwu, 1981)Google Scholar. On official Han seals, the most convenient compilation is Weizu, Sun , Liang-Han guanyin huikao (Hong Kong and Shanghai: Daye and Shanghai shuhua, 1993)Google Scholar.
12. For the convicts’ epitaphs, see Wenwu 1972.7 and 1982.3; and Kaogu 1972.4. For the Warring States clay inscription, see Guwenzi yanjiu 14 (1986), 177–95 Google Scholar; Zhongguo kaoguxue yanjiu lunji (Xi’an: San Qin, 1987), 336–42 Google Scholar; Koshi shunjū 6 (1989), 75–79 Google Scholar; and Zhongguo wenwubao, September 1, 1991, 3. For writings on pottery in general, see the rubbings of inscriptions in Ming, Gao , Gutaowen huibian (Beijing: Zhonghua, 1989)Google Scholar; and Yong, Wang and Miao, Li , eds., Zhongguo gudai zhuanioen (Beijing: Zhishi, 1990)Google Scholar. The former includes marks on pottery as early as Shang and Western Zhou; the latter deals with specimens from Late Warring States Qin to the Nanbeichao period with a general historical introduction on the subject. Dictionaries on pottery graph forms have been written by Zhongyi, Yuan , Qindai taowen (Xi’an: San Qin, 1987)Google Scholar; and Ming, Gao and Yinghui, Ge , Gutaowenzi zheng (Beijing: Zhonghua, 1990)Google Scholar.
13. The best reference for early stone inscriptions for which the stones or at least rubbings are still extant is Hidemasa, Nagata , ed., Kandai sekkoku shūsei (Kyoto: Dōhō, 1994)Google Scholar. To this should be added the exceptionally rich Fei Zhi bei inscription excavated in 1991. See Wenwu 1992.9,37–42 Google Scholar; and Dalu Zazhi 94.2 (1997), 1–13 Google Scholar.
14. As, for example, in his “Textual Criticism and the Tui, Ma Wang Lao tzu,” Harvard Journal of Asiatic Studies 44:1 (1984), 185–224 Google Scholar.
15. Lao, D.C., Chinese Classics: Tao Te Ching (Hong Kong: Chinese University Press, 1982)Google Scholar; Henricks, R., Lao-tzu: Te-tao ching-. A New Translation Based on the Recently Discovered Ma-wang-tui Texts (New York: Ballantine, 1989)Google Scholar; Mair, V., Tao Te Ching: The Classic Book of Integrity and the Way (New York: Bantam, 1990)Google Scholar; Möller, H.-G., Laotse Tao Te King: Die Seidentexte von Mawangdui (Frankfurt a.M.: Fischer Taschenbuch Verlag, 1995)Google Scholar.
16. Susan Weld, “Covenant in Jin’s Walled Cities: The Discoveries at Houma and Wenxian” (Ph.D. diss., Harvard University, 1990). See pp. 23–28,37,46–49,57–63,74–78, 323–40, 351–54, 389, 394–403, 406–19 for examples of the only slightly edited text of her essay in Sources. This is not indicated in Sources.
17. According to the “Finding List of Inscribed Bronzed Vessels Cited” in Sources of Western Zhou History, 313–18.
18. Compare, for example, the translations of the bamboo-strip text Shiwen (Ten questions) in Sources, 251, and Harper, Donald, “The Sexual Arts of Ancient China as Described in a Manuscript of the Second Century b.c.,” Harvard Journal of Asiatic Studies 47:2 (1987), 550 CrossRefGoogle Scholar and 552. For other translations see his “The Wu Shih Erh Ping Fang: Translation and Prolegomena” (Ph.D. diss. University of California, Berkeley, 1982) and “The Conception of Illness in Early Chinese Medicine as Documented in Newly Discovered 3rd and 2nd Century b.c. Manuscripts,” Sudhoffs Archiv 74 (1990), 210–35 Google Scholar. A complete, albeit unannotated, translation of the Spellbinding text from the Shuihudi manuscripts has been published by Harper in Religions of China in Practice, ed. Lopez, Donald (Princeton: Princeton University Press, 1996), 241–50 Google Scholar. See also, Harper’s complete translation of the Mawangdui medical manuscripts in his Early Chinese Medical Literature: The Mawangdui Medical Manuscripts (London: Kegan Paul International, 1998)Google Scholar.
19. Preparations for the joint Sino-Japanese excavations at Niya began as early as 1988. For results which include some new Kharosthi documents, see Zhong-Rigongtong Niya yiji xueshu diaocha baogaoshu (Nitchū kyōdō Niya iseki gakujutsu chōsa hōkokusho ), vol. 1 (Urumqi and Kyoto: 1996)Google Scholar. I would like to thank Prof. Nagasawa Kazutoshi for making this reference available to me. For the Sino-Japanese excavations that started in 1998 at the Chang'an site, see Zhongguo shibao June 6,1998,14.
20. A proceedings volume of this conference is being edited by Sarah Allan and Crispin Williams.
21. Occasionally, even the layout of an epigraphical source can be tricky. See, for example, Loewe's, Michael reading of Etsingol strip no. 502.22/502.9A (or TD2 no.5 in his Records of Han Administration, vol. 2, 256–59 Google Scholar). Loewe rigidly assumes a layout that is indeed often seen in multi-strip documents, which he describes in Sources, 173–74, as follows: “The two sets of binding cords separated the text into three sections or registers. Within each section the columns proceed from right to left. After completing the section at the head of the strip, the reader would pass on to the middle and then to the lowest section.” However, the correct reading sequence for the strip in question (as transcribed in Gan, Lao , Juyan Hanjian: shiwen zhi bu [Taibei: Academia Sinica, 1960]Google Scholar, and other subsequent transcriptions) does not conform to this rule. The first column of text on this strip clearly reads straight down the entire length of the strip; only then, oddly, the text in the remaining two columns has to be read one section after the other as described by Loewe.
22. Cf. Postgate, Nicholas, Wang, Tao, and Wilkinson, Toby, “The Evidence for Early Writing: Utilitarian or Ceremonial?,” Antiquity 264 (1995), 474 Google Scholar.
23. See Tsuen-hsuin, Tsien, Written on Bamboo and Silk (Chicago: University of Chicago Press, 1962), 90 Google Scholar and 160. Interpretations of early graph forms always carry an element of uncertainty, of course. For example, there is no way to positively prove that ce did indeed consist of wooden or bamboo strips. Rather, some think of it as plastron pieces bound together by strings.
24. Tsien, Written on Bamboo and Silk, 160.
25. Zhongshu, Wang, Han Civilization (New Haven: Yale University Press, 1982), 175 Google Scholar. Xia Nai once even stated that 20,000 to 30,000 Han tombs were discovered in the decade between 1950 and 1961 alone; see Nai, Xia, Xin Zhongguo de kaogu shouliuo (Beijing: Wenwu, 1961), 74 Google Scholar.
26. Muzhou, Pu , Muzangyu shengsi (Taibei: Lianjing, 1993), 4 Google Scholar.
27. Cf. Maspero, Henri, Les documents chinois de la troisième expédition de Sir Aurel Stein en Asie Centrale (London: Trustees of the British Museum, 1953), 21 Google Scholar. Once (Sources, 233, n.42), Harper states that before the hemerological texts from the Qin tomb of Shuihudi were found, “the oldest almanacs were Dunhuang manuscripts.” This is true if the author indeed refers to Han dynasty wooden strips found in Dunhuang, not if he refers to the famous paper rolls found in Dunhuang cave monasteries. Especially in the formative stage of the academic field of Dunhuang Studies efforts were made to include both the wooden and the paper documents under this heading (cf. Dunhuangxue jikan 6 [1984]), and even today occasional lip-service is paid to this ideal. But in actual practice, research on both types of materials has been conducted very much separately in terms of individual scholars as well as in terms of publications. Thus, “Dunhuang manuscripts” today usually evokes the image of the paper rolls rather than the wooden strips.
28. In the Journal of the Royal Asiatic Society, April 1965, 13–26.
29. Ōba’s article is reprinted in his Kankan kenkyū (Kyoto: Dōhō, 1992)Google Scholar.
30. This had already been pointed out by Hulsewé in 1987, besides some other orthographical and logical irregularities in the document; “Han China,” 269, n.31, et passim.
31. This account relies on an article by Kageyama Terukuni in the August 1996 issue of Gekkan Chūgoku tosho (no. 89), a periodical published by Uchiyama shoten in Tokyo. On p. 6, Kageyama relates the contents of a report in the February 11,1996, issue of Shodō bijutsu shinbun (no. 539). I am grateful to Prof. Tomiya Itaru for providing me with a copy of Kageyama’s article. Forgeries from what appears to be the same source were also sold in Taiwan; see the forthcoming report of Shijie, Luo , “Paoshui zhujian tuoshui chuli ji qi zhen-wei panduan” , Jiandu xuebao 17 (1999)Google Scholar.
32. See Sommarström, Bo, Archaeological Researches in the Edsen-gol Region, Inner Mongolia (Stockholm: Sino-Swedish Expedition, 1956), 1 Google Scholar; and Hidemasa, Nagata , Kyoen Kankan no kenkyū (Kyoto: D¯h¯, 1989), 45, n.28Google Scholar.
33. In his first discussion of this document (in “Han China”) Hulsewé cited the ji, but declared himself unable to give a satisfying explanation of its meaning. Now, he has unfortunately opted for deleting it altogether and to do so without further notice. As can be easily seen, the graph ji is not indifferent to the entire argumentation of the author. If Xiongnu is not the subject of the sentence, but a transitive object to ji (“those who attack the Xiongnu [and make them surrender?]”), the issue would not be one of enfeoffing foreigners (a practice that certainly existed), but of rewarding those who gained merit in battle. This need not inevitably be true, though. Ji could also be a kind of epitheton ornans, a loan for a word that still has to be identified, the final graph of a sentence that starts on another strip, or, as Hulsewé obviously thought, just an error. In any case, a reference to its existence should have been made.
34. While in recent years the number of newly discovered early Chinese manuscripts keeps rising steadily, the number of Chinese scholars who are turning their variously inspired attention to these sources increases at an even faster pace, so much so that by now it is becoming increasingly difficult for any single person to keep abreast of all new monograph studies, let alone articles, that appear on different aspects of these sources. The last nine months alone have occasioned at least three international conferences dedicated to three different caches of manuscript finds (for the proceedings see IV.5A below). The following sections have benefited from a talk on new developments in early manuscript studies held by Chen Wenhao at the Academia Sinica, Taipei, on December 19, 1998. I would also like to thank my teacher Hsing Itien and my friend Luo Shijie for providing some materials cited herein.
35. To speak of a computerized full-text database of inscriptions or manuscripts is somewhat misleading, since it is only the transcriptions of these texts that can be used in a keyword search. Images of the documents themselves can be scanned, but not analyzed by the computer. Conducting a keyword search with such a database of transcriptions cannot provide absolute guarantee that one will find all extant evidence, because transcriptions of early Chinese epigraphical documents to a varying degree contain an element of interpretation and uncertainty. For a description of the database at the Institute of History and Philology, Academia Sinica (accessible via the homepage of the Institute), see xiaozu, Jiandu zhengli , “Shiyuso cang Hanjian zhengli gongzuo jianbao” in Juyan Hanjian bubian , ed. xiaozu, Jiandu zhengli (Taibei: Zhongyang yanjiuyuan Lishi yuyan yanjiusuo, 1998), 1–14 Google Scholar. For examples of how to use computerized databases for the research on epigraphical sources see Zenggui, Liu , “Diannao zai Hanjian yanjiu zhong de yingyong” , Xinshixue 2.4 (1991), 1–36 Google Scholar. A database of digitized images of and some basic information on manuscripts found by Aurel Stein, which also serves as means of preservation, is presently established by the British Library as part of the International Dunhuang Project and can be accessed from the website of the project (http://idp.bl.uk). For details of an available computerized database with several kinds of transcriptions and digitized images of the bamboo strips from Baoshan, see Atsutshi, Iwamoto “Konpyūta riyō ni yoru Hōzan Sokan, Sokei moji no kenkyū” , in Chūgoku shutsudo shiryō kenkyū 3 (1999), 82–100 Google Scholar. Another computerized database including transcriptions as well as some images of the finds from Mawangdui, Wuwei, Shuihudi, Ymqueshan, Etsingol (Juyan), and several smaller caches can be purchased on CD-Rom with the CHANT Center, Institute of Chinese Studies, Chinese University of Hong Kong.
36. This database has been translated and enlarged from a Chinese version, which I originally compiled for a study group of wooden strips at the Academia Sinica, Taibei, on behalf of Professor Hsing Itien (Xing Yitian) using earlier compilations as well as archaeological reports. The Chinese version also benefited greatly from several pages of detailed criticism by my friend Wang Ding in a letter from Berlin. For the English version I have once again checked all entries with the references that are cited under the “Reproductions/Transcriptions” section of each entry as far as these were available (which has mostly been the case).
37. There are a few wooden documents even as late as the Qing dynasty, but these are written in Manchu. For a short introduction to these and other wooden documents in Japanese, Korean, and Kharosthi, see Sinica 2 (1991), 5 Google Scholar. A detailed review of Japanese wooden strips is provided with every issue of the periodical Mokkan kenkyū
38. See Toshifumi, Nishikawa , “Yinwan Kanbo kantoku san, shigō mokutoku ni tsuite” , Ōryō shigaku 24 (1998), 79, n. 15Google Scholar.
39. This tomb together with tomb no. 336 has been reported on in Wenwu 1992.9, where the tombs were mistakenly identified as 127 and 136 respectively. I would like to thank Peng Hao for sharing this information.
40. Cf. n. 39 above.