Skip to main content Accessibility help

The Bamboo Annals revisited: problems of method in using the Chronicle as a source for the chronology of Early Zhou. Part 2: The congruent mandate chronology in Yi Zhou shu

  • David W. Pankenier (a1)


The account given of the Bamboo Annals pre-Conquest chronology in Part 1 of this paper bears comparison with the internally consistent chronological sequence found in Yi Zhou shu, whose coherence has not been fully appreciated by those attempting to reconstruct the chronology. As we saw in the discussion of King Wu's reign in Part 1, Yi Zhou shu ‘Da kuang’ and ‘Wen zheng’ chapters are correct in dating the appointment of King Wu' s siblings as viceroys in the former Shang domain, which event we know took place in the Conquest year, to the 13th year of the Mandate, a finding that is fundamentally incompatible with Chou Fa-kao's, David Nivison' s, and Edward Shaughnessy' s account. The same two chapters in combination with ‘Zuo luo’ are also correct in implying that King Wu died in the 14th year. Furthermore, ‘Wen zhuan’ chapter of Yi Zhou shu correctly dates King Wen' s demise to the ‘9th year of the Mandate’ in those very words.



Hide All

45 See Part 1 of this paper, BSOAS, LV, 2, 1992, n. 39.

46 Shaughnessy, ‘The “Current” Bamboo Annals', 50, 41.

47 ibid., 41.

48 ibid., 50.

49 Nivison, ‘The dates of Western Chou', 524.

50 The 7th-year account of King Wen' s death derives from the ‘Basic Annals of Zhou’ in which Sima Qian states that King Wen died seven years after receiving the Mandate. Elsewhere, as we saw above (Part 1, n. 33), Sima Qian states that King Wen died six years after attacking the Quan Yi barbarians, which is consistent, since he implies in the’ Basic Annals of Zhou’ (Shiji, 4: 117–18) that this campaign occurred in the year following the receipt of the Mandate. However, Shang shu dazhuan dates the Quan Yi campaign to the 4th year of the Mandate (see table 4), which is corroborated by the Bamboo Annals and the Shijing, and confirmed by Yi Zhou shu which has King Wen dying in the 9th year of the Mandate, six years (inclusive) later. It appears, therefore, that Sima Qian was basically correct about the approximate timing of King Wen' s death relative to the Quan Yi campaign, but not about the timing of that campaign relative to the Conquest. Therefore, Sima Qian' s 7th-year date for King Wen' s death derives from the historian' s own presentation of the relative chronology and not from a variant’ royal calendar'.

51 Yi Zhou Shu, ‘Bao dian’ chapter has ‘3rd year’ but Xin Tang shu , 27B.604, quotes the passage as saying ‘1st year'.

52 See Part 1, n. 14.

53 All this evidence I adduced by 1983; see‘Appendix A: Chronology of Kings Wen and Wu’, in ‘Early Chinese astronomy and cosmology’, 319 ff., for annotation and explication of the texts cited. Left out of discussion here is the ‘Shi fu’ chapter of Yi Zhou shu whose dated entries display clear evidence of manipulation by the Warring States period chronologists who first attempted a reconstruction of the calendar of the Conquest year. Those dated records, the parallel ‘Wu cheng’ text quoted by Liu Xin in his ‘Canon of the Ages', and the Yinli reconstruction of the early chronology are all analysed in my ‘Reflections of the lunar aspect on Western Zhou chronology’ (T' oung Pao [1992]). The two new moon dates, together with the lunar eclipse record of 1065 B.C. from the ’ Xiao kai’ chapter of Yi Zhou shu, which identify the precise year in King Wen' s reign, the month, and the exact day, could not possibly have been retrospectively calculated since the data only became meaningful in absolute terms once King Wen' s reign and the Mandate calendar were correctly dated astronomically; see my ‘Astronomical dates in Shang and Western Zhou’. This suggests that Shaughnessy's conclusions are in need of revision in regard to the unreliability of passages from Yi Zhou shu which do not contain commentary by Zhao, Kong ; cf. Shaughnessy, ‘Authenticity of the Bamboo Annals’, 159–63.

54 See Jiujin, Chen , ’ Cong Mawangdui boshu Wuxing zhan de chutu shitan woguo gudai de suixing jinian wenti’ , Zhongguo tianwenxue shi wenji (Beijing, 1978), 165.

55 Compare the following statements about Jupiter' s annual motion, the first from the Warring States astronomer Gan De as quoted in Kaiyuan zhanjing (23:3a), and the second from the Mawangdui MS ‘Wu xing zhan ‘(Wenwu 11 [1974], 37):

(Gan de) ‘In all Jupiter takes twelve sui to complete one circuit [of the heavens]; [after] altogether 370 days, at dusk [Jupiter] enters [the void] in the west; thirty days [later] it again appears at dawn in the east.’

(MWD MS) ‘Jupiter… is visible for three [hundred sixty-five days; then at dusk it enters the void in the west and is concealed] for thirty days; [after] 395 days it again reappears in the east; [in twelve] sui it makes one circuit of the heavens…’

What is especially noteworthy, apart from the fact that the passage is obviously the product of careful, long-term observation of Jupiter's rising and setting, is the use of sui to refer to the period of Jupiter's visibility: twelve sui make up the period of the planet. Technically, sui refers, therefore, not to a conventional calendar year, but to the twelve-month period of the planet which begins and ends in successive months of the year, depending on Jupiter's ‘age’ in the cycle. The rising and setting observations of the planet for each year given in the MS confirm this usage. Hence, Jupiter sui, whether in the Zhou dynasty or in the twentieth-century A.D., are defined solely with respect to the location of the planet among the stars and are unaffected by changes in the location of the solstices and equinoxes brought about by precession.

56 See my ‘Astronomical dates ’, 10, where the locations of the planet are those computed by Stahlman, W. and Gingerich, O., Solar and planetary longitudes for years –2500 to 2000 by 10–day intervals (Madison: University of Wisconsin Press, 1963). The tables have an accuracy of plus or minus one degree. The positions of the determinative stars given there have a similar margin of error which, for the purpose in question, may be deemed inconsequential.

57 The auspicious significance of Jupiter' s stationary episodes is affirmed in Jin shu: ‘Predictions made by Jupiter as it advances or retrogrades refer to the State represented by the constellation it occupies. When Jupiter remains in a (particular) group it manifests the virtues of the State shown by that constellation, and a rich harvest can be expected. To wage war on this state would be fatal because disaster will fall upon its enemy. Happy tidings are fore told when Jupiter remains undisturbed in its path. When it advances or retrogrades and fails to observe the usual stages in its cycle, disaster will fall upon the State concerned. When this happens, launch no important projects and avoid using the army….’, tr. Yoke, Ho Peng, The astronomical chapters of the Jin Shu (Paris, 1966), 122–3 (italics mine). This probably explains why the first campaign launched by King Wu was aborted, see ‘Astronomical dates’, 15. In January of 1045, in contrast, Jupiter would have been just about to depart Quail Fire after spending eighteen months in that space astrologically associated with Zhou, hardly a propitious time to launch an attack on Shang, as Chou, Nivison and Shaughnessy would have it.

58 See Pankenier, , ‘Astronomical dates in Shang and Western Zhou’, Early China, 7, 19811982, 8–10, and esp. Pankenier, , ‘Early Chinese astronomy and cosmology: the “Mandate of Heaven “as Epiphany ’, Stanford University Ph.D. dissertation, 1983, 121 ff.

59 Nivison, , ‘The dates of Western Chou’, Havard Journal of Asiatic Studies, 43.2, 1983, 512.

60 Chou, ’Xi Zhou niandai xin kao’, 1; see also Fa-Kao, Chou, ‘Wu Wang ke Shang de niandai wenti Bulletin of the Institute of History and Philology, 56, 1, 1985; cf. Yuzhe, Zhang , ‘Halei huixing de guidao yanbian de qushi he ta de gudai lishi ’, Tianwen xuebao , 19, 1, 1978, 116–17.

61 Nivison, ‘The dates of Western Chou’, 512, for his part admits that, ‘It is likely that Jupiter's position at the time of the Conquest was thought important and was remembered. The recently discovered Li kuei inscription… though difficult to interpret, appears to say this…'.

62 It is striking that although Nivison (‘The dates of Western Chou ’, 511–12) refers repeatedly to Han conventions, Han standards, and Han astrology, he never once mentions the Mawangdui MS, ‘Wu xing zhan’. That text does not bear him out, as I point out, in regard to the application of the concept’ Quail Fire’ in actual practice.

63 Nivison, ‘1040 as the date of the Chou Conquest’, 78.

64 Nivison (‘The dates of Western Chou’, 512) explicitly accepts the criterion that Jupiter' s heliacal rising the preceding year must occur in Quail Fire for the subsequent’ civil’ year to be so designated, though he avoids mentioning whether such heliacal rising in mid-summer is the first or the second such rising in Quail Fire. If the former, then Jupiter will typically spend a full twelve months near the Bird Star in the space nominally identified as Quail Fire, if the latter, only the spring of the subsequent ‘civil’ year will be spent in Quail Fire before Jupiter, having already passed by the Bird Star many months before, advances eastward out of Quail Fire.

65 Nivison (‘The dates of Western Chou’, 511), also says of the Guoyu passage, ‘The first line… [giving the location of Jupiter in Quail Fire] is literally true of the year 1046, which was the year of the Conquest in the Shang calendar.’ By this Nivison is conceding nothing but merely means to say that his date of January 1045 for the Conquest is also the year 1046, if one speaks in terms of a supposed Shang convention according to which the solstitial month of the Zhou calendar for 1045, in which he dates the Conquest, would correspond to the last month of the Shang year 1046. In any case, the distinction between the Shang and Zhou calendars being made here is speculative, particularly in view of the proven use of the Xia convention in Yi Zhou shu to date the lunar eclipse of 1065. Nivison' s arguments in this connexion are all the more surprising considering the latitude of three months he is prepared to accept in defining the first month for the purposes of dating the bronze inscriptions throughout Western Chou. It is difficult to see the value of talking in terms of theoretically distinguishable Shang and Zhou conventions at this early date when a consistent practice is not even demonstrable for the subsequent centuries of Western Zhou; cf., Peiyu, Zhang, ‘Early China Forum’, Early China, 15, 1990, 142.

66 Nivison, ‘The dates of Western Chou’, 512.

67 This star map, dating from about 940, is in turn based on the map of stars and constellations, together with explanation and astrological commentary, made by Chen Zhuo (fl.c. 310). According to Sui shu (19.504), Chen based his work on the catalogues of stars of Shi, Shen, Gan, De, AND Wu, Xian, the fourth-century B.C. astronomers; cf. Needham, Science and civilisation in China, Vol. 2; History of scientific thought (Cambridge, 1956), 263–71. The Dunhuang MS is published in Wenwu, 3, 1966, 27–38. The explanations accompanying each region of the sky such as that translated here are also reproduced in Kaiyuan zhanjing, 64: la–11a. Lunar mansion Seven Stars, the middle mansion in Quail Fire, which as a cardinal station must contain three mansions, is referred to here as Quail Fire because it contained the Bird Star a Hydrae which represented the entire constellation.

68 Xigui, Qiu , ‘Tantan Suixian Zeng Hou Yi mu de wenzi ziliao . Wenwu, 7, 1979, 29.

69 See Needham, , Science and civilization in China, vol. 4, Physics and physical terminology. Part I: Physics (Cambridge, 1962), 140.

70 Nivison, ‘The dates of Western Chou’, 510 ff. As for the rest of Nivison's treatment of the Guoyu text, careful examination of his outline of this ‘rigorous mathematical proof’ that the text is a first-century B.C. forgery shows that his entire edifice of conjecture rests on the assumption that the ‘forger’ accepted the ‘Wu cheng’ dates for the Conquest as quoted in Han shu. This assumption is utlimately founded on the fact that the passage identifies the date of the battle as the ‘second month’ (which actually proves correct, though not in Nivison's view). Nivison's subsequent argument which proceeds from this is wholly the result of conjecture as to what the putative forger ‘must’ have had in mind, in which Nivison advances, among other things, unprecedented interpretations for key terms (e.g., chen as ‘the east moving 30–tu zodiac space 15 tu east and west of the sun, in which no star can be seen’), while ignoring parallel passages in Zuo zhuan and archaeological evidence bearing on the authenticity of the text; see ‘Early Chinese astronomy and cosmology’, 168, 181. Most extraordinary is the claim that the ‘forger’ would have been so knowledgeable in astronomy as to be capable of extrapolating astronomical conditions a millen nium earlier, which handiwork he then was able to insinuate into a well-known text whose musicological gist had already been quoted approvingly by Sima Qian (Shiji, 25.1240). And yet the very same forger was also so incompetent as to employ in his purportedly ancient record the official Han definition of the solstitial point when, as every one of his contemporaries knowledgeable in calendrical astronomy would have been aware, that location had only recently been redefined in the Tai Chu reform of 104 B.C. as a consequence of the obvious obsolescence of the nominal definition in effect throughout the Zhou. One can only wonder whom such a ‘forger’ would have thought he was fooling. As I argued in ‘Early Chinese astronomy and cosmology’, the passage can be interpreted to describe correctly celestial conditions in the fall of 1047 B.C. which would be consistent with a date of early 1046 for the Conquest, and not 1045.

71 Fang, Guben zhushu jinian jizheng, 38–9, 233.

72 Mengjia, Chen , Yinxu buci zongshu (Beijing, 1956), 212;Tso-pin, TungYinli pu (Sichuan, 1945), vol. 1, 4:3a;Pankenier, , ‘Astronomical dates’, 1718.

73 Han shu, 21B.1014.

74 Notice that this means that Di Xin's reign cannot have exceeded 40 years, while the 1045 conquest hypothesis requires him to have 41; cf. e.g., Shaughnessy ‘The “Current” Bamboo Annals’, 50.

75 Mengjia, Chen, Yinxu buci, 208–10; cf. e.g., Kane, Virginia C., ‘The chronological significance of the inscribed ancestor dedication in the periodization of Shang dynasty bronze vessels’, Artibus Asiae. 35.4, 1973, p. 341, n. 16.

76 Pankenier, ‘Astronomical dates’, 18.

77 Toyoda Hisashi demonstrates that, although the earliest literary and inscriptional sources clearly distinguish the role of King Wen as recipient of the ‘esoteric’ Mandate from the role of King Wu as the one who established the temporal authority of the Zhou over the ‘Four Quarters’, this feature of the references to the Mandate becomes rapidly attenuated. By the late ninth century, the distinction is no longer made and both Wen and Wu are spoken of as having ‘received the Mandate’. See Toyoda, ‘Shū ōchō no kunshuken no kōzō ni tsuite: “temmei no yoju”-sha o chūshin ni’ in Matsumaru, Michio (ed.), Seishū seidoki to sono kokka (Tokyo, 1980), 394, 396, 401, 405, 415.

78 An entry in the Bamboo Annals in the 12th year of Wen Ding of Shang (Fang, Guben zhushu jinian jizheng, 229), says that ‘a Phoenix alighted on Qishan’ . Shen Yue's original comment to this record identifies this as the year King Wen acceded to the throne in Zhou. In fact, this augury and the one concerning the Red Crow which is presently juxtaposed with the report of the planetary conjunction signalling the conferral of the Mandate on Zhou both refer to the same event; cf. Guoyu (1:11b), ‘When the Zhou arose, the Phoenix sang on Mt. Qi’. The report of the Phoenix augury was no doubt moved to its present location during reconstruction of the Bamboo Annals as a direct result of the common misinterpretation of a passage in Shang shu by Mencius and others; see Part 1, n. 30.

79 Jin shu, 51.1432.

80 Fang, Guben zhushu jinian jizheng, 245.

81 Here we have the main rationale for the transposition of a 17th-year record from the reign of King Cheng to the reign of King Wu during the reconstruction of the text; for another view, see Shaughnessy, , ‘On the authenticity of the Bamboo Annals’, 176.

82 Shiji, 4.149, n.4.

83 Zizhi tongjian waiji (Sibu congkan ed.), 3:13a.

84 ‘[Book Review of] Hung-hsiang, ChouShang-yin ti-wang pen-chi .’, Monumenta Serica, 19, 1960, 501–15.

85 In ‘Astronomical dates’, p. 31, n. 112, I still subscribed to this interpretation. Later, however, the additional work which led to ‘Early Chinese astronomy and cosmology’, led me to abandon this view as unconvincing for reasons which will become apparent.

86 Mengyuan, RongShitan Xi Zhou jinian’, , Zhonghua wenshi luncong , 1, 1980, 19.

87 Guangxian, Zhao , ‘Cong tianxiang shang tuiduan Wu Wang fa Zhou zhi nian ’, Lishiyanjiu , 10, 1980, 59. Zhao makes use of the same study by Zhang Yuzhe concerning the historical changes in the orbit of Comet Halley cited above as the authority for Chou Fa-kao's conclusions with regard to the Guoyu. Zhang claimed to have dated an apparition of Comet Halley to the spring of 1057, which finding is adduced by Zhao in his argument for that year as Conquest year. Unfortunately, as I have noted elsewhere, Zhang's calculations have not stood up under scrutiny by other researchers using more rigorous methods; see Yeomans, Donald K. and Kiang, Tao, ‘The long-term motion of comet Halley’, Monthly Notices of the Royal Astronomical Society, 197, 1981, 635–42. In actual fact, the return of comet Halley in question might have been observed either late in 1059, the year of the Mandate conjunction, or early in 1058, the first year of the Mandate; for full discussion, see ‘Early Chinese astronomy and cosmology’, 184 ff.

88 Nivison, ‘The dates of Western Chou’, 539.

89 Fang, Guben zhushu jinian jizheng, 260.

90 Nivison, ‘The dates of Western Chou’, 524.

91 Contrary to Edward Shaughnessy's conjecture, there is no support in the current Bamboo Annals chronology for the notion that King Wu's conquest of Shang occurred in a ‘12th year’ in the unreconstructed version of the chronicle; cf. Shaughnessy ‘The “Current “Bamboo Annals ’, 41.

92 This 1045 date for King Wu's death departs from my account in ‘Astronomical dates in Shang and Western Zhou’, 35, where the seventeen years of Wu's reign were incorrectly factored. Subsequent analysis in 1983 resolved the anomaly which was then revised for ‘Appendix A: Chronology of Kings Wen and Wu ’ in ‘Early Chinese astronomy and cosmology’. This analysis showed that the current Bamboo Annals total of seventeen years for King Wu should be factored into five years of actual rule plus the (8 + 4 =) twelve years of inflation inherent in the’ twenty-one year’ solution to the chronology devised in the third century. For discussion see ‘Transformations in King Wu' s reign’.

93 Present evidence indicates that King Cheng's reign nominally commenced in the first month of spring in 1042 B.C. after the completion of mourning for King Wu. By all accounts, King Cheng was then still under age and the Duke of Zhou assumed the role of Regent from that date. If the traditional figure of seven years for the Regency and the 1036 B.C. date for the events of the ‘Luo gao’ are both correct, as first proposed by Leopold de, Saussure, ‘La chronologie chinoise et l' avenement des Tcheou’, T' oung Pao, 23, 1932, 312, then 1035 ought to have been the first year Cheng, King ruled in his own right. However, the recently discovered He zun inscription (Wenwu, 1, 1976, 60–6; tr. Carson, Michael, ‘Some grammatical and graphical problems in the Ho tsun inscription’, Early China, 4, 1979, 41–4) suggests that 1036, ‘when the King first moved his residence to Cheng Zhou’, was King Cheng's 5th cult year, making 1040 B.C. his official first year. This is confirmed by the combination of the Bamboo Annals report of Tang Shuyu's investiture as Marquis of Jin in King Cheng's tenth year (Fang, Guben zhushu jinian jizheng, 240) and the Guoyu (10:3a) statement that in that very year Jupiter was in Great Fire (Scorpius), making the date 1031 B.C. Conceivably, the Bamboo Annals entry for King Cheng's ‘first’ year, which speaks about the Duke of Zhou's appointment to preside over the one hundred officials as well as about Cheng's cappingand ascending the throne, is a composite of two ‘first years’, the nominal first year 1042 when King Chengwas still a minor, and 1040 when he presumably reached majority and ascended the throne.

Related content

Powered by UNSILO

The Bamboo Annals revisited: problems of method in using the Chronicle as a source for the chronology of Early Zhou. Part 2: The congruent mandate chronology in Yi Zhou shu

  • David W. Pankenier (a1)


Full text views

Total number of HTML views: 0
Total number of PDF views: 0 *
Loading metrics...

Abstract views

Total abstract views: 0 *
Loading metrics...

* Views captured on Cambridge Core between <date>. This data will be updated every 24 hours.

Usage data cannot currently be displayed.