1. The Shanghai Museum bamboo texts, as these strips are now known, are being published serially: Shanghai bowuguan cang Zhanguo Chu zhushu , ed. Ma Chengyuan (Shanghai: Shanghai guji); Volume 1 was published in 2001, Volume 2 in 2002, Volume 3 in 2003 (it was actually released only on 18 April 2004), Volume 4 in 2004, and Volume 5 in 2006; subsequent volumes are expected at the rate of about one per year. For an account of the purchase of these strips and related issues, see “Ma Chengyuan xiansheng tan Shangbo jian” , in Shang boguan cang Zhanguo Chu zhu shu yanjiu , ed. Mingchun, Liao and Yuanqing, Zhu (Shanghai: Shanghai shudian, 2002), 1–8Google Scholar. Ma notes that the first purchase, made through the intercession of Cheung Kwong-yue , came at the beginning of 1994 (pp. 1-2). A second purchase of 497 strips, apparently deriving from the same tomb, was not made until the end of 1994 (P. 4).

2. For these strips, see Baoshan Chu jian , ed. Hubei sheng Jing Sha tielu kaogudui (Beijing: Wenwu, 1991)Google ScholarPubMed, and Guodian Chu mu zhujian , ed. Jingmen shi bowuguan (Beijing: Wenwu, 1998)Google Scholar. Like the tomb that produced the Shanghai Museum bamboo strips, the Guodian tomb was also robbed—twice—first in August of 1993, and then again in October of the same year. It was the second looting that prompted local archaeologists to excavate the tomb, in the course of which they recovered the bamboo strips. For an account of the excavation, see Hubei sheng Jingmen shi bowuguan, “Jingmen Guodian yihao Chu mu” , Wenwu 1997.7: 35–48Google Scholar.

3. The Shanghai Museum invited the Shanghai Institute of Atomic Nuclei (Shanghai Yuanzihe yanjiusuo ) of the Chinese Academy of Sciences to perform a C-14 dating of the strips. The result was 2257 ± 65 BP or 306 ± 65 B.C.E. (assuming that the Institute used the international standard of 1950 as Present); see “Ma Chengyuan xiansheng tan Shangbo jian,” p. 3. This certainly confirms the antiquity of the strips, and is also generally consistent with the date of the calligraphy on them.

4. In addition to the Zhou Yi, the volume also contains three other texts: Zhong Gong (28 strips, most of them fragmentary), Heng Xian (13 strips, all intact), and Pengzu (8 strips, 4 of which are intact).

5. Photographs of the portion of this manuscript that contains the hexagram and line statements, which is to say the same portion of the text contained in the Shanghai Museum bamboo-strip manuscript, are available in Mawangdui Han mu wenwu , ed. Juyou, Fu and Songchang, Chen (Changsha: Hunan, 1992), 106–17Google Scholar.

6. Many of the initial studies of this manuscript have been posted on one of two websites: <http://www.jianbo.org> and <http://www.confucius2000.com>. Chen Renren , “Shanghai bowuguan cang Zhanguo Chu zhushu Zhou Yi yanjiu zongshu” , Zhou Yi yanjiu 2005.2:16-27, contains a listing of thirty-five articles either posted to one of these two web-sites or else published in the journal Zhou Yi yanjiu. To these should be added Liu Dajun , Jin bo zhushu Zhou Yi zongkao (Shanghai: Shanghai guji, 2005).

7. Maozuo, Pu has just published a two-volume study of the Shanghai Museum Chu Zhou Yi manuscript and its context: Chu zhushu Zhou Yi yanjiu: jian shu Xian Qin Liang Han chutu yu chuanshi Yi xue wenxian ziliao (Shanghai: Shanghai guji, 2006)Google Scholar. The book is divided into two parts, the first presenting the Shanghai Museum manuscript and the second comprising a conspectus of materials for the development of the Yi jing tradition through the Eastern Han dynasty. The first part reproduces several portions of the Shanghai bowuguan cang Zhanguo Chu zhushu Volume 3 presentation of the manuscript: the color photographs of the individual strips; the strip-by-strip discussions of the text; and the comparison of the Shanghai Museum manuscript, the Mawangdui manuscript, and the received text. It also includes a discussion of the black and red symbols found on the manuscript, somewhat differently organized from that of Shanghai bowuguan cang Zhanguo Chu zhushu Volume 3 but with no apparent difference in interpretation. In addition to these sections, this first part of the book adds a lengthy general introduction to the manuscript (“Chu zhushu Zhou Yi gaikuang yu yanjiu” and three new appendices: a bare transcription of the manuscript (“Chu zhushu Zhou Yi yuanwen” ), a glossary of phrases in the manuscript (“Chu zhushu Zhou Yi cimu jieshi” ), and a concordance of characters in the manuscript (“Chu zhushu Zhou Yi zhuzi suoyin” ). The second part of the book is divided into two chapters. The first of these chapters surveys archaeological materials related to the Yi jing. These include hexagram symbols found in oracle-bone and bronze inscriptions and on other implements, and hexagram symbols and related texts found in bamboo-strip texts, including those from Baoshan , Xincai , Tianxingguan , Guodian , and Fuyang ; and all Yi jing related texts from Mawangdui, the Xiping Stone Classics, as well as the Guicang texts from Wangjiatai . The second chapter presents quotations and uses of the Yi in transmitted literature, together with a complete text of the received version of the Yi jing. Despite the redundancies with Shanghai bowuguan cang Zhanguo Chu zhushu Volume 3, the book serves as a convenient sourcebook for the early development of the Yi jing tradition in China.

8. For this strip, see Rao Zongyi , “Zai kaituo zhong de xunguxue: Cong Chu jian Yi Jing tandao xinbian Jingdian shiwen de jianyi” , paper presented to the first Guoji xunguxue yantaohui , Zhongshan daxue , Gaoxiong, Taiwan, April, 1997; cited by Zeng Xiantong , Zhou Yi Kui gua guaci ji liusan yaoci xinquan” , in Xiantong, Zeng, Guwenzi yu chutu wenxian congkao (Guangzhou: Zhongshan daxue, 2005), p. 193Google Scholar. See also, Songchang, Chen, Xianggang Zhongwen daxue Wenwuguan cang jiandu (Hong Kong: Xianggang Zhongwen daxue, 2001), p. 12Google Scholar.

9. The significance of these symbols will be discussed in detail below, pp. 12-16.

10. Because references to Zhou Yi line statements in the Zuo zhuan do not make any mention of these line tags, employing instead a fairly cumbersome system of identifying lines by comparing two different hexagrams, some have suggested that these line tags were a later development; see, for example, Jingchi, Li, Zhou Yi tanyuan (Beijing: Zhonghua, 1978), pp. 65–66Google Scholar. This manuscript serves as valuable evidence for the use of these line tags already in the late fourth century B.C.E.

11. For a brief description of the sequence of the Mawangdui Yi jing manuscript and how it differs from that of the received text, see Shaughnessy, Edward L., “A First Reading of the Mawangdui Yi Jing Manuscript,” Early China 19 (1994), 52–53Google Scholar.

12. Numerous other characters throughout the manuscript are similarly attested, including (but by no means limited to) for de for de for duo for fu for heng for hou for jue for lin for miao for qi for ru for shen or for sui for su for xiao for ya for yi for yin for yuan for zai for zheng for zhong or for zhu etc. Other graphic forms have attested precedents in oracle-bone or bronze inscriptions: for bin for dun for gu while still more have parallels in other contemporary Chu manuscripts: for sha for sheng for shi , or for xiang . There are also many instances of classifier difference (that is, the use of a different signific from that found in the character of the received text): for dian for fu for fu for gong for gu for hu for hou for kou for lü for qin , or for sheng , etc.

13. There are, of course, numerous “phonetic loans,” as for instance for bu for chu for fen for fu for fu for fu for gou for guo for han for hui for jun for pan for pan for peng for wei for yin for you for you for zhu , etc. Phonetic loans are especially prevalent in the cases of binomial words, as, for example, for zhiti for tihao for ciqie . However, on the whole, it would seem that there are somewhat fewer phonetic loans than in other contemporary texts with received counterparts, and even fewer than in the case of the Mawangdui Yi jing manuscript.

14. There are also numerous variora within the received tradition; for compendia and studies of these, see Li Fusun , Yi jing yiwen shi (Huang Qing jingjie xubian ed., 1888); Qinting, Xu, Zhou Yiyiwen kao (Taibei: Wuzhou, n.d.)Google Scholar; and Xinchu, Wu, Zhou Yi yiwen jiaozheng (Guangzhou: Guangdong renmin, 2001)Google Scholar.

15. The Mawangdui manuscript, probably copied in the 170s B.C.E., that is after the death of Liu Bang and before the accession of Liu Qi, writes this phrase as , and the Fuyang Zhou Yi manuscript, buried in 165 B.C.E., writes , more or less identical with the Shanghai Museum manuscript; for the Fuyang manuscript, see Ziqiang, Han, “Fuyang Han jian Zhou Yi shiwen” , Daojia wenhua yanjiu 18 (2000), 20Google Scholar. Whether this means that the copying of the Fuyang manuscript predated Liu Bang's reign or death, or that it simply failed to observe the taboo on his name is unclear.

16. Strip 46 contains only the single character ji “auspicious,” and the hollow three-sided red square with an inset solid black square symbol. Although several other hexagrams also end with the character ji, and thus a strip with just this one character might also belong to any one of them, since the symbol here matches that on strip 44 after the hexagram name, the editor is almost surely correct in placing the strip here.

17. Heng, Gao, Zhou Yi gujing jittzhu (Shanghai: Kaiming, 1947), 164Google Scholar. Gao was followed in this reading by Richard Kunst, who also proposed this emendation in his translation of the text (“an old pitfall contains no game”); see Kunst, Richard A., “The Original ‘Yijing’: A Text, Phonetic Transcription, Translation, and Indexes, with Sample Glosses” (Unpublished Ph.D. dissertation: University of California, Berkeley, 1985), 334–35Google Scholar

18. It is unclear to me whether the following phrase, zhui bi lou “it is tattered old clothes” in S and weng bi lou “the jar is broken and leaking” in R, should be read together with the opening phrase, and if so whether its different readings in the manuscript and the received text are significant or not.

19. I offer this merely as a possibility, but one that does not seem to me to be very likely. Pu Maozuo reads as a phonetic loan for fu , which the Fang yan says was a dialect word for “to catch”; see Shanghai bowuguan cang Zhanguo Chu zhushu (san), p. 198. In any event, the manuscript here is ambiguous at best, and certainly does not provide obvious support for the traditional interpretation of the image as a “well.”

20. Of course, some of the ambiguity of a text such as the Zhou Yi has arisen in the course of the long exegetical tradition as readers have become ever further removed from the original compositional context. As the Shanghai Museum bamboo-strip manuscript becomes better studied, it will surely provide important evidence to reconsider other cases of phonetic or especially graphic variants. For now, I will offer just one example that seems fairly conclusive to me. This is found in the “image” statement of the Jiu wu line of Guai hexagram (hexagram #43 in the received text).

S/33:

R:

Although Zhu Xi (1130-1200) had explained the xian lu of the received text as the “the purslane (a type of edible green) of today” (jin machi xian ), an early gloss by Meng Xi (c. 90-40 B.C.E.) of the Han dynasty indicating that it was the name of an animal (shou ming ) has led such modern scholars as Gao Heng and Richard Kunst to emend xian to huan , a type of mountain goat, and to regard lu as the original form of lu “to hop,” thus reading the phrase, in Richard Kunst's translation, as “a mountain goat hopping lickety-split”; see Kunst, “The Original ‘Yijing’,” pp. 324-25. In the manuscript, both of the graphs in question are clearly written with grass significs (in the second character, the liu component is a shortened form of the lu of the received text), indicating that at least the copyist of the manuscript understood both characters to represent a type of plant, consistent with Zhu Xi's later gloss.

21. Shaughnessy, Edward L., “The Fuyang Zhou Yi and the Making of a Divination Manual,” Asia Major 3rd ser. 14.1 (2001), 15Google Scholar.

22. LiShangxin, , “Chu zhushu Zhou Yi zhong de teshu fuhao yu guaxu wenti” , Zhou Yi yanjiu 2004.3,24Google Scholar, suggests that there is a seventh symbol, which he describes as a large solid red square with a smaller solid black square set inside it. The only example of this comes at the end of Yi hexagram, strip 24. Pu Maozuo treats it as a three-sided hollow red square with an inset smaller solid black square (what I refer to below as “B”). Since the red coloring here is faded, and since this would be the only example of this putative seventh symbol, it seems best to follow the editor's description in the following discussion.

23. However, he uses only three brief sentences to explain away the third hexagram with mis-matched symbols, Yi, which intervenes between these in the received sequence of hexagrams. Since this hexagram has a B symbol at the head and a C symbol at the tail, Pu says that it must follow after a B group and come before a C group.

24. Shanghai bowuguan cang Zhanguo Chu zhu shu (san), p. 259.

25. Shanghai bowuguan cang Zhanguo Chu zhu shu (san), p. 260.

26. To my knowledge, only three other attempts to explain these symbols have been published to date. The first is Jiang Guanghui , “Shangbo cang Chu zhushu Zhou Yi zhong de teshu fuhao de yiyi” , at <http://www.jianbo.org/admin3/list.asp?id=ii99> 16 May 2004. Jiang suggests that the symbols have no intrinsic relationship with the text of the Zhou Yi, but rather were added to the manuscript by diviners as a sort of aide mémoire to help locate hexagrams. He goes on to suggest, by way of a “theoretical model,” that the sequence of hexagrams in the Shanghai Museum manuscript must have been more or less similar to that of the received Zhou Yi. Unfortunately, the theory behind this model is not at all clear in Jiang's brief presentation. The second discussion is more detailed: Li Shangxin's “Chu zhushu Zhou Yi zhong de teshu fuhao yu guaxu wenti,” mentioned above at n. 22. Li begins by pointing out flaws in Pu Maozuo's presentation, including several discrepancies between descriptions given in his study of the text (shiwen ) and in the appendix devoted to the symbols. He also notes that, while Pu argues for a hexagram sequence different from that of the received text, the sequence he suggests is thoroughly influenced by the received sequence. Li observes simply that there is too little evidence to support Pu's reconstruction. In its place, he proposes that the manuscript's sequence is essentially the same as that of the received text, and further suggests a model for the sequence of the symbols. According to Li, hexagrams paired by virtue of their hexagram pictures (usually by inversion of the hexagram picture, as for example in the case of Shi #7 and Bi #8, or in the eight cases in which such inversion results in the same hexagram picture by conversion of all six lines to their opposite form, as in the case of Qian #1 and Kun #2) always share the same symbol. On the other hand, Li goes on to suggest that different symbols must be given to hexagrams created by converting hexagram pictures that can be inverted (for example, Shi #7 can be considered to convert into Tongren #13), or by exchanging top and bottom trigrams of hexagram pairs that cannot be inverted (for example, Qian #1 and Kun #2 exchange trigrams to produce Tai #11 and Pi #12). Unfortunately, of these examples suggested by Li, neither Qian or Kun nor Tai or Pi is present in the manuscript, and so it seems impossible to test his explanation. The third discussion, Chen Renren , “Shangbo Yi teshu fuhao de yiyi yu biaoshi yuanze” , at <http://www.jianbo.org/admin3/2005/chenrenren001.htm>, 4 July 2005, is an exhaustive critique of the previous attempts by Pu Maozuo, Jiang Guanghui and Li Shangxin to detect some principle behind the placement and nature of these symbols, concluding not only that all three of these attempts are seriously flawed but—especially given the fragmentary nature of the Shanghai Museum manuscript—that no consistent explanation is likely to be persuasive.

27. The following presentation is essentially identical with Hanyi, Xia, “Shilun Shangbo Zhou Yi gua xu” , Jianbo 1 (2006), 97–105Google Scholar.

28. The top of strip 3, about 23 cm long, is missing. Presumably it would have carried the last two characters of the Liu si line, the eight characters of the Jiu wu line, and the first sixteen characters of the Shang liu line.

29. A similar analysis might hold for strips 48 and 50, the first of which is broken at 13.4 cm from the top and the second at 12.6 cm from the top—the beginnings of Gen #52, and Jian #53, hexagrams—the intervening strip 49 is intact and contains the final portion of Gen.

30. Du Yu, “Chunqiu Zuo zhuan jijie Hou xu” , in Chunqiu Zuo zhuan zhengyi (Sibu beiyao ed.), “Jiaokan ji Hou xu” 1a.

31. For this title and description of the text, as also for the other texts discovered at Ji jun and mentioned below, see Fang Xuanling , Jin shu 51 (“Shu Xi zhuan” ) (Beijing: Zhonghua, 1974), 1433. This text seems to correspond to a description in the earlier Jin shu of Wang Yin (284-354; written between C.E. 320 and 340), which states that “among the ancient texts there were hexagrams of the Changes, similar to the Lianshan and Gui cang ” (gu shu you Yi gua, si Lianshan Gui cang ); quoted by Xun, Ouyang, Yiwen leiju (Shanghai: Shanghai guji, 1999), 40.732Google Scholar. With the discovery, in 1993 at Wangjiatai in Jiangling , Hubei, of a third-century B.C.E. manuscript of the Gui cang, several scholars have recently surmised, reasonably, it seems to me, that this Yi yao yinyang gua discovered in Ji jun was probably none other than the Gui cang; for the first suggestion of this connection, see Wang Mingqin , “Gui cang yu Xia Qi de chuanshuo: Jianlun tai yu jitan de guanxi ji Diaotai de diwang” , Huaxue 3 (1998), 212-26; see, too, Ning, Wang, “Qin mu Yi zhan yu Gui cang zhi guanxi” , Kaogu yu wenwu 2000.1: 49–50, 55Google Scholar; Yuanqing, Zhu, “Wangjiatai Gui cang yu Mu tianzi zhuan” , Zhou Yi yanjiu 2002.6: 9–13Google Scholar; Junhua, Ren and Ganxiong, Liang, “Gui cang Kun Qian yuanliu kao” , Zhou Yi yanjiu 2002.6:14–23Google Scholar.

32. The Lu shi of Luo Ping and Luo Bi (completed in 1170) quotes a passage that is extremely similar to the received text of the Shuo gua commentary of the Yi jing but as coming from the Gui cang; Lu shi (Sibu beiyao ed.), 1.15b (“Fa hui” ). It seems likely that this derived from the Gua xia Yi jing discovered at Ji jun, and was probably appended to the end of the Yi yao yinyang gua/Gui cang.

33. For published photographs of this manuscript, see above, n. 5. For concise listings of the Mawangdui library, see Yuqian, Pian and Shu'an, Duan, Ben shiji yilai chutu jianbo gaishu (Taipei: Wanjuanlou tushu, 1999), pp. 32–42Google Scholar, and Enno Giele, “Database of Early Chinese Manuscripts,” at http://www.lib.uchicago.edu/earlychina/res/databases/decm/mss.html#064.

34. For translations of these texts, see Shaughnessy, Edward L., I Ching: The Classic of Changes, The First English Translation of the Newly Discovered Second-Century B.C. Mawangdui Texts (New York: Ballantine Books, 1996)Google Scholar.

35. For introductions to this discovery, see Pingsheng, Hu, “Fuyang Han jian Zhou Yi gaishu” , Jianbo yanjiu 3 (1998), 255–66Google Scholar; Ziqiang, Han, “Fuyang Han jian Zhou Yi yanjiu” , Daojia wenhua yanjiu 18 (2000), 63–132Google Scholar; Edward L. Shaughnessy, “The Fuyang Zhou Yi,” 7-18. A complete transcription is given in Ziqiang, Han, “Fuyang Han jian Zhou Yi shiwen” , Daojia wenhua yanjiu 18 (2000), 15–62Google Scholar, and in Ziqiang, Han, Fuyang Han jian Zhou Yi yanjiu (Shanghai: Shanghai guji, 2004)Google Scholar.

36. See, above, n. 15, for evidence that, unlike the Mawangdui manuscript, the Fuyang manuscript did not observe a taboo on the name of Liu Bang (i.e., Han Gaozu; r. 202-195 B.C.E.).

37. For an introduction to these texts, see Tsien, Tsuen-hsuin, Written on Bamboo and Silk: The Beginnings of Chinese Books & Manuscripts, 2nd ed. (Chicago: University of Chicago Press, 2004), pp. 78–85Google Scholar. For the fullest study of the Yi jing inscription, see Wanli, Qu, Han shi jing Zhou Yi canzi jicheng (Nangang: Academia Sinica, 1961)Google Scholar.