1. Shaughnessy, Edward L., “On the Authenticity of the Bamboo Annals,” Harvard Journal of Asiatic Studies 46, no. 1 (06 1986): 149–180CrossRefGoogle Scholar. As always, I am grateful to Professor David S. Nivison, whose general influence is apparent on every page of this study, and to Professor David N. Keightley whose editorial pen has helped to make a complex argument more intelligible.

2. Shaughnessy, , “On the Authenticity of the Bamboo Annals,” p. 179Google Scholar, n. 59.

3. Legge, James, The Chinese Classics, volo 3, The Shoo King or the Book of Historical Documents (1883; reprint ed., Hong Kong: Hong Kong University Press, 1960), p. 588Google Scholar; Lunheng (Sibu belyao ed.), 1.12b.

4. For both of these chronologiess see Shaughnessy, , “On the Authenticity of the Bamboo Annals,” p. 177Google Scholar.

5. Legge, , The Chinese Classics, vol. 3, Prolegomena, p. 149Google Scholar.

6. Xiangyong, Fan, Guben zhushu jinian jijiao dingbu (Shanghai: Renmin chubanshe, 1962), p. 26Google Scholar, citing Jinshu (Beijing: Zhonghua shuju, 1974), 51.1432Google ScholarPubMed.

7. It is generally accepted without question that the “Ancient” Bamboo Annals preserves an earl 1er textual tradition than does the “Current” Bamboo Annals. In the case of this emendation, however, there is some evidence to suggest that the “Current” version is just as early as the “Ancient” version. The “Fengbao” chapter of the “Jizhong” tradition of the Yi Zhou shu , shown in Shaughnessy, “On the Authenticity of the Bamboo Annals,” pp. 159-165, to have been composed spuriously in the late third century A.D., begins with the following passage:

It was the twenty-third ritual cycle, the first day of the month was gengzi (day 37); the archer-lords of the nine regions all came to Zhou. The king was at Feng. Early in the morning, taking his place in the Lesser Court, the king announced to Duke Dan of Zhou, saying: “Wuhu! The many lords have all come to celebrate the arduous military service (against) Shang. How am I to maintain them? How am I to act?”

Although in the present arrangement of the Yi Zhou shu, this text is placed among chapters understood to belong to King Wen's reign, I would suggest that the “king” was originally intended to be King Wu. If the references to celebrating military achievements over the Shang and the king's solicitation of the Duke of Zhou's advice did not suffice to demonstrate this, then the placement of this conversation in Feng, which according to the Bamboo Annals, the Shiji, and the Shangshu dazhuan was not constructed until one or two years before King Wen died, leaves little doubt about it. It is equally clear that the king could not be King Cheng since, according to the Bamboo Annals, the Duke of Zhou died in the twenty-first year of Cheng's reign. That this audience with the many lords in celebration of the victory over Shang is dated to the “twenty-third year” requires that the date be based on a hybrid Wen/Wu regnal calendar that includes nine “mandate” years for King Wen and eleven years of pre-conquest personal reign for King Wu. Thus, this twenty-third year would be equivalent to the fourteenth year of King Wu's personal reign. Except for the “Current” Bamboo Annals, there is no other extant source that explicitly states that King Wu personally reigned for eleven years before the conquest. (One passage in the Lüshi chunqiu does seem to suggest that King Wu had already reigned twelve years before the conquest, but the text is not unambiguous; Lüshi chunqiu, 14.8a.) Thus, it would seem that the composer of the “Feng bao” text could only have based this twenty-third-year date for King Wu's post-conquest period on the arrangement of the Bamboo Annals represented in the “Current” version of that text, and, consequently, that this arrangement dates from no later than the late third century A.D.

8. Shiji (Beijing: Zhonghua shuju, 1959), 4.120Google ScholarPubMed. Accounts in other chapters of the Shiji contain somewhat contradictory dates, but in no case do they require a fundamental revision in the narrative provided by the “Zhou benji”; see, for example, “Qi Tai gong shijia” 32.1479, and “Lu Zhougong shijia” , 33.1515.

9. Shiji, 61.2123.

10. For the fourth century B.C. date of the “Yinli” chronology, see Zuobin, DongYinli pu (Nanqi, Sichuan: Guoli Zhongyang Yanjiuyuan lishi yuyan yanjiusuo zhuankan, 1945)Google Scholar, Shang 4.2a-4b; Pankenier, David W., “Astronomical Dates in Shang and Western Zhou,” Early China 7 (1981–1983: n. 4Google Scholar.

11. The most complete information on the “Yinli” chronology, especially as it is quoted by Xuan, Zheng, is found in Maoshi zhengyi (beiyao, Sibu ed.):16, 1.1a-2bGoogle Scholar. Conflicting analyses of the “Yinlio” chronology and particularly of its date for the Zhou conquest are suggested by Dong Zoubin, who regards 1070 B.C. as the “Yinli” date of the conquest (Yinli pu, Shang 4.2a), Chen Mengjia who gives the date as 1076 (Yinxu buci zongshu [Beijing: Kexue chubanshe, 1956], pp. 211–12)Google Scholar, and Tang Lan , who argues for 1075 (“Zhongguo gudai lishi shang de niandai wenti” , Xin jianshe 1955.3:48–52)Google Scholar. In all cases, however, it is assumed that King Wu's regnal calendar continued the calendar initiated by King Wen upon his receipt of the mandate.

12. Hanshu (Beijing: Zhonghua shuju, 1962), 21B.1015–16Google Scholar.

13. Legge, , The Chinese Classics, 3:320ffGoogle Scholar.

14. Shaughnessy, , “On the Authenticity of the Bamboo Annals,” pp. 163–164, nn. 39, 40Google Scholar.

15. Shaughnessy, , “On the Authenticity of the Bamboo Annals,” pp. 164–165Google Scholar.

16. The entry for the twelfth year of King Wu's annals also contains a ganzhi year-designation (xinmao, #28), but this year, which was that of the Zhou conquest, was obviously considered in some sense to be the first year of the dynasty. That these ganzhi year-designations were indeed inserted into the text by its third-century A.D. editors is demonstrated in “On the Authenticity of the Bamboo Annals,” p. 171, fig. 3; when the annals of King Cheng's reign are written in forty-graph strips, for the strip transposed from his reign to that of King Wu to be written intact on a single strip, it is necessary to delete the two ganzhi graphs in the annals of the first year.

17. For a statistical analysis of the Bamboo Annals, see Shaughnessy, , “On the Authenticity of the Bamboo Annals,” pp. 154, n. 13Google Scholar. Of 214 discrete entries in the Western Zhou portion of the text, only seven could be considered as portents of this type.

18. Shangshu dazhuan (congkan, Sibu ed.), 2.16aGoogle Scholar; Shiji, 4.118.

19. In the Bamboo Annals entry for Di Xin's thirty-fourth year, the target of the Zhou attack is written as Qi whereas in the forty-fourth year entry it is written as Li . It has long been agreed that these two graphs both refer to the same state; for a thorough demonstration of this, see Jiegang, Gu and Qiyu, Liu, “Shangshu ‘Xibo kan Li’ jiaoshi yilun” Zhongguo lishi wenxian yanjiu jikan , 1 (1980): 46, n. 3, and 51–53Google Scholar.

20. Shangshu dazhuan, 2.17a.

21. Mi, Huangfu, Diwang shiji (jicheng, Congshu ed,), p.28Google Scholar.

22. Yi Zhou shu (jicheng, Congshu ed.), 3–58Google Scholar. For a discussion of the textual history of the Yi Zhou shu demonstrating the Warring States date of the “Kong Zhao” tradition, see Shaughnessy, , “On the Authenticity of the Bamboo Annals,” pp. 159–62Google Scholar.

23. Nivison, David S., “The Dates of Western Chou,” HJAS 43, no. 2 (12 1983):481–580Google Scholar; Pankenier, David W., “Astronomical Dates in Shang and Western Zhou,” Early China 7 (1981–1982):2–37CrossRefGoogle Scholar.

24. Mi, Huangfu, Diwang shiji, 28Google Scholar.

25. Pankenier, , “Astronomical Dates in Shang and Western Zhou,” p. 5Google Scholar.

26. Ibid.

27. Nivison, , “The Dates of Western Chou”; see especially, pp. 499–524Google Scholar. Nivison has subsequently publicly published one research note (“1040 as the Date of the Chou Conquest,” Early China 8 [1982–1983]:76–78Google Scholar) and privately published at least two further papers (“Is 1040 the Date of the Chou Conquest,” Duduan 01 [6 01 1984]Google Scholar; “1040 or 1045,” Duduan 05 [25 07 1984])Google Scholar, in which he has suggested that the date of 1045 B.C. given for the Zhou conquest in “The Dates of Western Chou” is perhaps wrong and should be changed to 1040 B.C. The primary reason for Ni vi son's change is a reinterpretation of the immediate pre-conquest chronology given in the Shiji. Not only do I find this reinterpretation unconvincing, but the date 1040 does not, it seems to me, satisfy the several primary calendrical data for the date of the conquest. In this paper, when I attribute the conquest date of 1045 B.C. to Ni vi son, it should be understood that I refer exclusively to the argument given in his study, “The Dates of Western Chou.”

28. Jiegang, Gu, “Yi Zhou shu ‘Shifu’ pian jiaozhu xieding yu pinglun” Wenshi 2 (1963), pp. 1–42Google Scholar; Shaughnessy, Edward L., “‘New’ Evidence on the Zhou Conquest,” Early China 6 (1980–1981):57–79CrossRefGoogle Scholar.

29. This date is corrected on the basis of its quotation in the “Lüli zhf” chapter of the Hanshu (21B.1015-16).

30. See Shaughnessy, , “‘New’ Evidence on the Zhou Conquest,” p. 70, n. 3Google Scholar; for an expanded version of the calendar presented here, see pp. 68-69 of that work.

31. Zuobin, Dong, Zhongguo nianli zongpu (Hong Kong: Hong Kong University Press, 1960), 1:124Google Scholar. Although the regnal chronology given by Dong is extremely questionable, the almanac is still useful and accurate in its correlation of lunarmonths with ganzhi day-designations.

32. Legge, , The Chinese Classics, 3:420–21, 452Google Scholar.

33. Zuobin, Dong, Zhongguo nianli zongpu, 1:125Google Scholar. It should be noted that Dong regards the month beginning with jiachen (#41) to be the second month of the Zhou year, but since neither the first month of the year nor the inter-calation schedule had been rigorously systematized by the beginning of the Western Zhou, a discrepancy of one month is not unusual.

34. This corpus has been collected and analyzed by Zuobin, Dong, Yinli pu, Xia 9.48a–63bGoogle Scholar; see too Mengjia, Chen, Yinxu buci zongshu, pp. 301–09Google Scholar; Nivison, , “The Dates of Western Chou,” p. 501Google Scholar.

35. References to published collections of oracle-bone inscriptions are as given in Keightley, David N., Sources of Shang History: The Oracle-Bone Inscriptions of Bronze Age China (Berkeley: University of California Press, 1978), pp. 229–31Google Scholar, with the qualification that they have been uniformly translated into pinyin romanization. In this and subsequent studies, I propose to use notations of the form “I.D.” to indicate the periodization of inscriptions, with Roman numerals I through V corresponding to the regnal periods of Dong Zuobin1 s five-period periodization (i.e., I: Wu Ding; II: Zu Geng, Zu Jia; III: Lin Xin, Geng Ding; IV: Wu YI, Wen Ding; V: Di Yi, Di Xin; see Zuobin, Dong, “Jiaguwen duandai yanjiu li” , Zhongyang yanjiusuo jikan waibian 1, no. 1 [1933]:323–324)Google Scholar, and capital let ters designating the diviner-group to which the inscription belongs (i.e., D: Dui ; Bin ; L: Li ; c: Chu ; He: He ; WM: Wuming [No Diviner-group]; H: Huang ; and also D-B for the Dui-Bin Transitional-group and D-L for the Dui-Li Transitional-group. This type of dual notation is preferable to either of the single notations now in use, since a single period often included more than one diviner group and a single diviner group often extended over two or more periods.

36. Hand copies and transcriptions of the original inscriptions, as well as full references for the oracle-bone collections cited, can be found in Zuobin, Dong, Yinli pu, Xia 9.48a–50aGoogle Scholar.

37. While the excess of only two, or perhaps just one, days between these two dates invites speculation that an extra day or days may have been assigned inadvertantly to one month, consideration of numbers 5 through 9 confirms that there must have been an intercalation. Otherwise, even if number 1 marked the first day of the ninth month and the first month of the second year began on day 32, the earliest possible day for it, this would still require a span of 121 days for four months. Considering that a lunar cycle Is 29.53 days, four lunations should properly last 118 days. It would seem quite unthinkable that over a course of four months â mistake of three days such as this could go unnoticed.

In addition to the Information available in this corpus of inscriptions involving the Renfang campaign, Dong Zuobin has demonstrated that all “thirteenth month” inscriptions date to either Period I.B or II.B/C (Yinli pu, Xia 5.2b–4a)Google Scholar, and, more important, has illustrated a late II.C scapula-bone (Yicun 399) with the following three inscriptions (Xia 5.11b-12a):

Crack on guiwei (day 20), Xiong divining: “In the coming ten-day week there will be no harm.” Sixth month.

Crack on guichou (day 50), Da divining: “In the coming ten-day week there will be no harm.” Sixth month.

Crack on guihai (day 60), Da divining: “In the coming ten-day week there will be no harm.” Sixth month.

This bone requires at least a thirty-one-day interval between “sixth month” notations, also suggesting the usage of a mid-year inter-calary month. The difference between this mid-year inter-cal ation and the earlier end-of-year inter-calation suggests that in addition to the “New School” innovations in the Shang ritual calendar introduced during Zu Jia's reign, there was also at this time an important innovation in this aspect of calendrical science.

38. Nivison, , “The Dates of Western Chou,” p. 501Google Scholar. This correlation is particularly impressive given that Dong's almanac indicates that an inter-calary month would have been required after the ninth month of 1077 B.C., precisely when the inscriptions require an inter-calary month.

39. The Shiji chronology of the final seven years of King Wen's reign is given in the following table.

Years given above in brackets are inferred from the Shiji account's use of “ming nian” (next year) to separate the various events. The implications of this lack of greater specificity vis-à-vis the chronology given in the Bamboo Annals can be seen most clearly in the case of the attacks on Li and Yu . The Shiji actually separates the attacks on Li, Yu and Chong , dating them to sequential years. That the Bamboo Annals chronology, In which these attacks are said to have occurred in the same year, is here preferable can be seen from two points. First, Sima Qian's use of the indefinite “ming nian” to date these events instead of regnal year-notations suggests that there was a series of events that he knew occurred in the time between King Wen's reception of the “mandate” and his death and that he was simply spacing these events out over those years. That Sima assigned the conquest of Chong and the transfer of the capital to Feng to the same (sixth) year probably Indicates only that he had run out of years to continue listing the events in successive years. It is noteworthy that the sequence of these events in both the Bamboo Annals and in the Shiji is otherwise identical. Second, and more important, both Li and Yu were located in the same vicinity, just north of the Yellow River and south of the southern tip of the Taihang Mountains. From a practical military viewpoint, an attack on one of these states would certainly have required an attack, in the same year, on the other. Whether the attack on Chong belongs in the same year (as per Bamboo Annals) or in the next (as per Shiji) is open to question. But since Chong was probably located directly across the Yellow River from Li and Yu, there is some reason to accept the Bamboo Annals here too. For the geography of this campaign, see Jiegang, Gu and Qiyu, Liu, “Shangshu ‘Xibo kan Li’ jiaoshi yilun,” pp. 51–59Google Scholar.

40. Shiji, 4.116.

41. I am preparing a study of the chronology of the entire Western Zhou dynasty and will publish it in due course.