The Key to the Chronology of the Three Dynasties: The "Modern Text" Bamboo Annals

by David Shepherd Nivison


Summary of the Argument

  1. Shaughnessy's discovery that a slip in the Annals chronicle for Cheng Wang of Zhou has been moved to the end of the chronicle for Wu Wang shows that the Bamboo Annals chronology is at least as early as the fourth century BCE, and is earlier than any other known. So in trying to reconstruct exact dates one must begin with this.
  2. The Annals date for the conjunction of 1059 BCE is 1071, back 12; its dates for the reign of Wen Wang are 1113-1062; thus the correct dates should be 1101-1050. Other chronologies give Wen 50 years only. Inscriptions show that a Zhou king nonally had a post-mourning "accession" in his third year; so Wen's dates must be 1101/1099-1050. This is confirmed by the lunar eclipse on day bingzi (13) in the first month of Wen's 35th year, = 1065, as mentioned in Yi Zhou shu 23 "Xiao Kai."
  3. Mourning-completion indirectly indicates a royal calendar with first year 1056 (1058 being "mandate year" after the conjunction). If Mu Wang's first year was 100 years after the beginning of Zhou, it could be 956. This is confirmed by bronze inscription dates (taking lunar phase terms to mark approximate lunar quarters).
  4. The dates for kings two through four, assuming Annals reign lengths, therefore should be these: Cheng Wang, 1037/35-1006 (2 + 30); Kang Wang, 1005/03-978 (2 + 26); Zhao Wang, 977/75-957 (2 + 19), confirmed by the Xiao Yu ding inscription (979). Wu Wang died three years after the conquest, which must be 1040, on Qing Ming Day, confirmed (e.g.) by the final line of the "Da Ming" ode in the Shi jing.
  5. The Zhou Gong Regency was then misdated as the seven years before Cheng's 30 years, making the conquest 1045 (still reflected in other dates in the present Annals). The Yin Li conquest date is 1070, back 25 from 1045; if the Yin Li date for the first year of Shang, 1579, is also back 25, then the correct date is Pankenier's date 1554.
  6. Pankenier's conjunction date 1953 for Shun 14, and Pang's solar eclipse date 1876 for Zhong Kang 5, plus two-year intervals between reigns, yield complete Xia Dynasty dates to 1561-1555 for next-to-last king Fa; so the last king Jie is an invention. This chronology seems verified by the fIrst day of Kong Jia, 17 Feb 1577, = jiazi (01). Thus gan names of kings were apparently determined by first days of their reigns.
  7. This gan hypothesis is confirmed by its successful application to all the Shang kings, who all have gan names. Confirmations of resulting exact dates for all of Shang include these: (1) Two traditional errors are explained, (a) the impossibly long 75 years for Tai Wu; and (b ) the misdating of Yong Ji before Tai Wu rather than after him. Also (2) Shang bone inscriptions show that Wu Yi did die (as in the Annals) during a hunt "in the He-Wei area" in his 35th year, the correct date being 1109 BCE.
  8. Analysis of middle and late Western Zhou bronze inscriptions together with Annals reign lengths yields exact dates for Zhou kings from Mu Wang through You Wang.

Summary, Appendices

  1. Dukes of Lu: Xian Gong's reign should be 23 years rather than the 32 years in the Shiji; and Li Gong's succession date was the "first year" of Gong Wang, but was Gong Wang's accession date (915) rather than his succession date (917); also, the reign of the first duke Ho Qin was incorrectly lengthened 3 years in the Annals as an indirect consequence of the moving of "Shaughnessy's slip."
  2. The Late Shang ritual cycle: 70 or more inscriptions for the eastern campaign of Di Xin must be dated to 1077-1076, the campaign beginning on 29 September 1077, with the annual zai sacrifice to Shang Jia. From this starting point for analysis, the first days of sacrifice years from 1120 through 1041 are calculated, confirming dates for the Wu Yi, Wenwu Ding and Di Yi reigns.
  3. Pre-Zhou chronicle: Successive revisions from ca. 427 to 300 are explained, showing that dates in the Annals back to Huang Di are systematically related to Xia and post- Xia dates. (Therefore the original text must have begun with Huang Di. )
  4. Conquest dates other than the correct date 1040 explained and refuted: notably, the "13th year" theory of Liu Xin and many later scholars; and the incorrect but widely accepted date 1027, thought to be required by a' quotation from the Annals by Pei Yin.
  5. Chronology, Huang Di through Western Zhou: tabular summaries.
  6. Dated Western Zhou bronze inscriptions and lunar quarters theory: Absolute dates are computed for 56 bronzes that have full internal dates, using the four quarters (yue xiang) theory. The system is demonstrated by a tabular presentation of a day-by-day analysis of the Zhou conquest campaign in the first half of 1040 BCE.
  7. The slip text of the Bamboo Annals: Legge and other later scholars have argued that the Annals text and dates were reworked after the discovery of the book ca. 280 CE, so that the Annals dates are worthless for recovering ancient chronology. These objections are analyzed and refuted.
  8. The Wei revision and the "modern text": Errors and lacunae in the Wei (4th century) part of the chronicle are here examined, showing that the "modern text" probably derives from a copy of the work that was being done by the Jin court scholars before it was finished; and that the creators of the original text ca. 318-299 altered the dating of the reign of Huicheng Wang (as king) for political reasons, making it 335-319 rather than 334-319 BCE.