1. An initial version of this paper was delivered at the Annual Meeting of the American Historical Association in New York, 28 December 1985. It was presented in revised form at the International Conference on Shang Culture at Anyang, 13 September 1987. The final version was submitted on 8 April 1988. In arriving at the hypotheses offered here, an effort has been made to take into account as fully as possible the present state of our knowledge about the matters at hand, although it is obvious that these views will require further consideration and refinement as pertinent new information becomes available in the future.

2. According to David Nivison's calculations, dynastic Shang began in 1575 B.C. (“The Dates of Western Chou,” HJAS 43.2 [12 1983]:562)Google Scholar; but see also his revised chronology for Western Zhou (“1040 as the Date of the Chou Conquest.” Early China 8 [1982–1983]: 76–78)CrossRefGoogle Scholar. David Pankenier, on the other hand, proposes the date 1554 B.C. (“Astronomical Dates in Shang and Western Chou,“ Early China 7 [1981–1982]:21Google Scholar).

3. The Bamboo Annals, itself a part of the long history of archaeological discovery in China, is said to have been unearthed in A.D. 279 (A.D. 281, according to Nivison, , HJAS 43.2 [12 1983]:496)Google Scholar, together with other texts, after lying buried since the beginning of the third century B.C. Presently there are two versions: the Jinben, considered to be of post-Song date, and the Guben, the recension reconstituted by modern scholars. In both versions, the text records in brief entries, major events, along with seemingly minor ones, from the time of the legendary emperors, Huang Di, Yao, and Shun, down to the end of the fourth century B.C.

If it is in the nature of a miracle that the Annals survived into modern times in any form, it would be even more astonishing if it were found to contain any semblance of historical fact relating to Shang. Yet precisely this sort of verification has been forthcoming from the oracle bone inscriptions recovered at Anyang which have confirmed the king list of the final twelve Shang rulers ds provided by the Annals (cf. Keightley, David, “The Bamboo Annals dnd Shang-Chou Chronology,” HJAS 38.2 [12 1978]:423–424Google Scholar, also see Shaughnessy, Edward, “On the Authenticity of the Bamboo Annals,” HJAS 46.1 [06 1986]: 149–180)Google Scholar. The integrity of the Annals concerning the period before 1300 B.C. obviously remains uncertain; but there is nevertheless reason to believe that it ought to be taken more seriously than other textual references to this early period. At least one scholar today, David Pankenier, feels convinced that an astronomical phenomenon recorded in the Annals as having taken place during the reign of the last Xia ruler can now be securely identified with the planetary conjunction of mid-December 1576 B.C. (“Astronomical Dates,” pp. 18–21; Pankenier, , “Mozi and the Dates of Xia, Shang, and Zhou: A Research Note,” Early China 9–10 [1983–1985]:175, 179–180Google Scholar).

The Bamboo Annals will be quoted in the edition of Shiming, Fang and Xiuling, Wang, Guben Zhushu jinian jizheng (Shanghai, 1981)Google Scholar, which includes, besides the authors' reconstruction of the Guben text, the recension of the Jinben text by Wang Guowei, Jinben Zhushu jinian shuzheng. Page numbers will be given in the form Guben/Jinben when both texts are referred to; otherwise the version cited will be specified. Translations from the text follow those of Legge, James, The Chinese Classics, vol. 3Google Scholar; The Shoo King (Hong Kong, 1960), pp. 108–141Google Scholar.

4. Fang, pp. 38/233.

5. The Guben records only six capitals; the Jinben, seven (Fang, pp. 21–37/216–233).

6. Fang, pp. 29/244.

7. In answer to David Keightley's suggestion that Anyang may not have been the site of the late Shang capital at all, but rather a “cult and mortuary center” (“The Late Shang State,“ in Keightley, David N. ed., The Origins of Chinese Civilization [Berkeley, 1983], p. 533CrossRefGoogle Scholar, Keightley, , Sources of Shang History [Berkeley, 1978], p. xiv, n. 6Google Scholar), it can be argued that, even if this were the case, the rigorous schedule of ancestral sacrifice requiring the participation of members of the royal clan would mean that the capital itself could at least not have been any great distance away. The presence of the oracle bones at Anyang and the king's role in the divination procedure point to the same conclusion.

8. Fang. pp. 21/216.

9. Fang, pp. 26/221.

10. Fang, pp. 27/221.

11. Jinben: Fang, p. 222; Geng does not occur in Guben.

12. Fang, pp. 27/222; Guben omits the rulers other than Zu Ding.

13. Fang, pp. 28/223.

14. Fang, pp. 29/224. For the transfer of the Shang capitals prior to Pan Geng, compare also Tiehan, Zhao, “Shuo Yin Shang Bo ji Cheng Tang yihou zhi wu qian,“ in Gushi kaoshu (Taibei, 1965), pp. 211–238Google Scholar, and Heng, Zou, Xia Shang Zhou kaogu lunwen ji (Beijing, 1980), pp. 203–210Google Scholar.

15. Kaogu 1978.1:2–4Google Scholar.

16. Kaogu 1984.1:37–40, pls. 3–4Google Scholar; Kaogu 1986.4:318–323, pls. 7–8Google Scholar.

17. Kaogu 1974.4:234–248, pls. 2–5Google Scholar.

18. Kaogu 1983.3:206–216Google Scholar.

19. Kaogu 1976.4:260, fig. 3, pl. 5:4–6Google Scholar. The first frontlet comes from d burial (M4) assigned to the end of Period II; Kaogu 1984.1:37Google Scholar; 38, fig. 3:1, 40, pl. 4:1 (color). The second example is from the Period IV burial M11; Kaogu 1986.4:321, fig. 6, top, 323Google Scholar; pl. 7:1 (color). The presence of these two frontlets at Erlitou suggests that the chariot may already have been in use by this period.

20. Kaogu 1975.5:305, fig. 4:3, pl. 9:2Google Scholar; Kaogu 1976.4:260, fig. 4, pl. 5:3Google Scholar; Kaogu 1978.4:270, fig. 1:2, pl. 12:1Google Scholar; Fong, Wen, ed., The Great Bronze Age of China (New York, 1980), no. 1Google Scholar; Kaogu 1983.3:203, fig. 9:4, 6, pl. 1:1 (color)Google Scholar; Kaogu 1986.4:320, figs. 4, 5, pl. 7:2, 5 (color), pl. 8:6, 7Google Scholar.

21. Evidently the only vessel from Erlitou to exhibit an elementary form of surface decor is the well-known jue included in the Great Bronze Age of China exhibition (Kaogu 1978.4:270, fig. 1:2, pl. 12:1Google Scholar; Wen Fong, ed., The Great Bronze Age of China, no. 1).

22. Kaogu 1978.4:270, fig. 1:3, pl. 12:3Google Scholar; Wen Fong, ed., The Great Bronze Age of China, no. 3.

23. Kaogu 1978.4:270, fig. 2:1Google Scholar; Wen Fong, ed., The Great Bronze Age of China, no. 2; Kaogu 1983.3:204, fig. 10:5, 6, pl. 1:4 (color)Google Scholar; Zhongguo meishu quanshu: Gongyi meishu bian, 9: Yuqi (Beijing, 1986), p. 23, fig. 45Google Scholar.

24. Kaogu 1976.4:229–233, 262, fig. 6, pl. 6:1, 2 (color)Google Scholar; Kaogu 1986.4:323, fig. 9, pl. 7:3, 4 (color), pl. 8:1–3Google Scholar.

25. Wenwu 1961.4–5:73–74Google Scholar; Kaogu xuebao 1985.1:58–59Google Scholar.

26. Wenwu 1961.4–5:77Google Scholar; Chang, K. C., The Archaeology of Ancient China, 3d ed. (New Haven, 1977), pp. 232–233Google Scholar. A revised estimate is provided by David Keightley, who calculates that the construction of the wall required only twelve and a half years (“Religion and the Rise of Urbanism,” JAOS 93.4 [1973]:530–531Google Scholar; “The Cradle of the East: Supplementary Comments,” Early China 3 [1977]:58)Google Scholar.

27. Wenwu 1977.1:22–25Google Scholar; An Jinhuai, Chin-huai, “The Shang City at Cheng-chou and Related Problems,” in Chang, K. C., ed., Studies of Shang Archaeology (New Haven, 1986), pp. 26–27Google Scholar.

28. Wenwu 1983.4:7–8Google Scholar.

29. For the definition of Styles I through V, see Loehr, Max, “The Bronze Styles of the Anyang Period,“ Archives of the Chinese Art Society of America 7 (1953):42–53Google Scholar; Ritual Vessels of Bronze Age China (New York, 1968), pp. 13–14Google Scholar. At Zhengzhou, Style I is represented, for example, by the ding illustrated in Henan chutu Shang Zhou qingtong qi, vol. 1 (Beijing, 1981), no. 12Google Scholar. An early version of Style IV is seen so far only once, on a swing-handle you (Wenwu 1983.3:54, figs. 14, 15, pl. 1 [color], left)Google Scholar.

The style of bronze decor characteristic of the period of the da kong--the archaeologically unattested period of the missing capitals between Zhengzhou and Anyang--is the subject of a separate article now in progress. For preliminary work on this topic, see Huber, , “Some Anyang Royal Bronzes: Remarks on Shang Bronze Decor,“ in Kuwayama, George, ed., The Great Bronze Age of China: A Symposium (Los Angeles, 1983), pp. 19–24Google Scholar. The topic is also under investigation by Thorp, Robert, “The Growth of Early Shang Civilization: New Data from Ritual Vessels,” HJAS 45.1 (06 1985):44–51Google Scholar.

30. Wenwu 1975.6:64–68, pl. 1 (color)Google Scholar; Wen Fong, ed., The Great Bronze Age of China, no. 11.

31. Wenwu 1975.6:64Google Scholar.

32. Kaogu 1983.1:52–53Google Scholar; 56, figs. 18, 19, 58, fig. 25; pl. 4:5. The four fang ding differ slightly in their degree of stylistic development, with the largest example from H1 (H1:2, Kaogu 1983.1:56, fig. 18, pl. 4:5Google Scholar) recognizable as the latest. The advanced stage of this vessel is indicated, not so much by its surface decoration, as by certain features of its shape, which include its less extensive rim and more narrow handles. Although the vessels technically belong to Styles I and II, all would appear to date from the last third of the Zhengzhou period. Their lateness among the finds from this site argues neither for nor against the proposed identification of Zhengzhou as either Bo or Ao.

33. Only five other fang ding of comparable dimensions and weight have ever been found in China. All were unearthed at Anyang and date from the first half of the Late Shang period. In each case it Is certain that the vessel was cast on behalf of a high ranking member of the royal clan. Three vessels are inscribed with the name of a royal consort. These include the two Hou Mu Xin fang ding excavated from Tomb 5 at Xiaotun and the Hou Mu Mou fang ding. The other two vessels--the ox and the deer fang ding--were unearthed from the large cruciform royal tomb 1004 at Houjiazhuang. The heights and weights of these five vessels are as follows:

34. Among the more recent Western publications devoted to the Chinese Bronze Age, the questions whether Erlitou is a manifestat ion of Xia or Shang culture and whether Zhengzhou should more 1ikely be identified with Bo or Ao have been largely brushed aside, in favor of a fairly noncommittal view about these issues. K. C. Chang, on the other hand, has concerned himself deeply with these matters and approaches them with a degree of flexibility. Although initially hesitant to disavow the identification of Zhengzhou as Ao (The Archaeology of Ancient China, 3d ed., pp. 270–271)Google Scholar, he has become increasingly persuaded that at least the final phase at Erlitou may be Xia: “A view to identify Erh-li-t'ou I and II with Hsia and Erh-li-t'ou III and IV with Hsia's terminal capital site would be consistent with the history of Erh-li-t'ou itself. In short recent work at the Erh-1i-t'ou cul ture sites tends to strengthen the view that regards the culture not as Early Shang but as Hsia” (Shang Civilization [New Haven, 1980], pp. 344–345Google Scholar; compare also pp. 270–271; Chang, , “The Origin of Shang and the Problem of Hsia Culture in Chinese Archaeology,” in Kuwayama, George ed., The Great Bronze Age of China: A Symposium, pp. 12, 14Google Scholar, and The Archaeology of Ancient China, 4th ed. [New Haven, 1986], pp. 316, 336)Google Scholar. It should be noted however that Alexander Soper as early as 1966 already voiced his suspicion that Zhengzhou might not be Ao commentina that “one miaht be tempted to aive Erh-li-t'ou an even more dramatic role as a possible Hsia royal residence” (“Early Middle and Late Shang: A Note,” Artibus Asiae 28 [1966]: 23)Google Scholar.

In the long run, perhaps the most valuable contribution to a real solution of the problem resides in David Pankenier's work on astronomical dates, which is also of great significance regarding the historicity of Xia. The results of his study have inclined him toward identifying Zhengzhou with Bo on the bas i s of the agreement of the date he proposed for the beginning of Shang--1554 B.C.--with the radiocarbon dates ascertained for Zhengzhou of ZK 177: 1572 ± 140 B.C. and ZK 178:1598 ± 110 B.C. (“Astronomical Dates,” p. 24).

In China, there have been in the past nearly as many different opinions as there were scholars involved in the debate. An excel lent summary of the various points of view has been compiled by Weizhang, Yin, “You guan Xia wenhua tansuo de jige wenti,” Wenwu 1984.2:55–62Google Scholar. His own view, based upon an exceedingly detailed analysis of the archaeological evidence, of a significant break between Erlitou II and III and his belief that Periods III and IV belong within Shang, is most completely explained in “Erlitou wenhua zai tantao,” Kaogu 1984.4:352–356Google Scholar.

The scholar most strongly urging the Identification of Zhengzhou, not with Ao, but with Bo, is Heng, Zou (Lunken ji, pp. 192–203)Google Scholar. Although his conclusions are largely persuasive, not all his argumentation is equally convincing. The textual evidence he compiles for associating Zhengzhou with 80 (“Zhengzhou Shang cheng ji Tang du Bo shuo,” Wenwu 1978.2:69–71Google Scholar) has been disputed in an article by Jia, Shi (“Zheng Bo shuo shangque,” Kaogu 1980.3:255–258, 287)Google Scholar. His use of Eastern Zhou ceramics found at Zhengzhou bearing an impressed character supposedly reading “Bo” (Wenwu 1978.2:69–71Google Scholar) has also been received with skepticism (see Songlin, Zhang, “Zhengzhou Shang cheng quyu nei chutu de Dong Zhou taowen” Wenwu 1985.3:76–79Google Scholar). For reference to more recent views on these matters, see notes 52 and 53.

35. Kaogu 1977.4:229Google Scholar; Zhongguo kaogu xuehui diyici nianhui lunwen ji: 1979 (Beijing, 1980), p. 47Google Scholar.

36. For a discussion of the Xia cultural sphere in traditional historical geography (Shanxi, southwest and western Henan) see, Haloun, Gustav, Seit wann kannten die Chinesen die Tocharer oder Indogermanen überhaupt? (Leipzig, 1926), pp. 22–31Google Scholar; Tiehan, Zhao, “Xiadai zhu di suo ju kao,” in Gushi kaoshu, pp. 62–73Google Scholar, Heng, Zou, Lunwen ji, pp. 220–251Google Scholar.

37. Kaogu 1959.11:592–600Google Scholar, Chang, K. C., Shang Civilization, pp. 342–343Google Scholar.

38. Wenwu 1983.3:8–20Google Scholar. Excavations of the wall at Wangehenggang are not yet complete, but according to the report, the west wall is estimated to be roughly 92 meters long, and the length of the south wall is more than 82 meters.

39. Kaogu 1980.4:373Google Scholar.

40. Kaogu 1978.1:17Google Scholar; Kaogu 1977.4:229Google Scholar.

41. Fang, pp. 19/215.

42. Furthermore, the time span that would remain for the development of the bronze styles of the Zhengzhou period would be still further reduced in years if allowance is made for the probable “gap” of some decades existing between the most advanced bronzes from Zhengzhou and the earliest examples from the time of the Anyang capital. See note 29.

43. Repeated references in the passages concerning Xia in the Bamboo Annals to the Luo and the Yi rivers also favor the proposition that the Xia royal residences were in the Luoyang-Yanshi area. The text tells that when the third ruler Tai Kang was forced to abandon Zhenxin, he “went hunting beyond the Luo” (Jinben: Fang, p. 203). During the reign of Jie it is said that the “Yi and Luo became dry” (Jinben: Fang, pp. 212-219). Further mention of the Luo also occurs in the entries for the pre--Xia rulers. The Yellow Emperor “sacrificed to the river Luo,” as did Yao after him (Jinben: Fang, pp. 190, 194). Mention of the Luo and Yi are absent, on the other hand, from the sections dealing with Shang.

44. Pankenier, , “Astronomical Dates,” p. 21Google Scholar.

45. Kaogu 1984.6:488–504, 509Google Scholar; Kaogu 1984.10:872–879Google Scholar; Kaogu 1985.3:322–335Google Scholar; Zhongyuan wenwu 1985.3:85–94Google Scholar, Jianghan kaogu 1987.1:41–47Google Scholar.

46. Kaogu 1984.6:489Google Scholar. The east wall is 1,710 meters in length; the west wall, 1, 640 meters. The south wall, which lies under Tazhuang, has been disturbed by the changing course of the Lo River.

47. Kaogu 1985.4:322, 324, 326Google Scholar. On its inner side, the wall enclosing this courtyard was constructed as a contiguous series of apartmentst each approximately five meters square.

48. Ibid., p. 335.

49. Kaogu 1988.2:128–129, 140Google Scholar. D5 lies above the remains of a walled enclosure, approximately forty meters square, which had fallen into disuse by the middle Nanguanwai period (ibid., pp. 130, 140). According to An Jinhuai. the middle Ndnguanwai period is coeval with or slightly earlier than the Lower Erligang period (Kaogu xuebao 1973.1:91)Google Scholar.

50. Zhongyuan wenwu 1985.3:87Google Scholar.

51. Kaogu 1984.10:874Google Scholar, Zhongyuan wenwu 1985.3:85–87Google Scholar.

52. E.g., Shilin, Huang, Zhongyuan wenwu 1985.3:85–91Google Scholar; Shilin, Huang, “Zaitan Yanshi cheng ji qi xiangguan wenti,” paper presented at the International Conference on Shang Civilization, Anyang, 10–16 09 1987, pp. 1–25Google Scholar; Yousheng, Fang, Jianghan kaogu 1987.1:41–47Google Scholar.

53. Jinhuai, An, “Shilun Zhengzhou Shang cheng he Yanshi Shang cheng de zaowan guanxi,” paper presented at the International Conference on Shang Civilization, Anyang, 10–16 09 1987, pp. 12–14Google Scholar. An Jinhuai was convinced of the identification of Zhengzhou and Ao as early as 1961 (see Wenwu 1961.4–5:73–80Google Scholar).

54. In this case the city wall at Yanshi presumably would have been built by soldiers of the Shang army and by conscripts taken from the Xia.

55. Zhao Tiehan has raised the question whether the term bo originally signified a du yi (capital city) or merely a place where a she (altar) was established. It is his impression that early in its history the name represented by the character bo must have lost its unique reference to a specific locality and that a number of traditions linking it to a variety of places arose. Later commentaries came to gloss its occurrences with distinguishing attributes, such as Bei Bo, Nan Bo, Xi Bo, and others (Gushi kaoshu, pp. 159–210, 222–233). Compare Zhongmian, Cen, Huanghe bianqian shi (Beijing, 1957), pp. 95–102Google Scholar; Heng, Zou, Lunwenji, pp. 184–192Google Scholar.

56. Yousheng, Fang, Jianghan kaogu 1987.1:44–46Google Scholar, Huang Shilin, on the other hand, regards the matter of the Donggong as mere conjecture (Zhongyuan wenwu 1985.3:90–91)Google Scholar.

57. From the account in the Annals of the Shang campaign against the Xia, it can be inferred that, using Bo as a point of departure, the Shang conducted expeditions against those who stood in the way of their planned conquest, namely the Luo, the Jing Wen, the Kunwu, and the Gu (Jinben: Fang, pp. 213–214). Given the number of campaigns involved, Bo must have been at a considerable remove from the Xia cities.

58. Kaogu 1984.6:509Google Scholar.

59. Kaogu 1984.10:874, figs. 4–5Google Scholar, Henan chutu Shang Zhou qingtong qi 1.99Google Scholar.

60. For example, certain sherds from the fifth stratum of Trench 1, classified by the authors of the excavation report as “earlier than the Upper Erligang period” (Kaogu 1984.6:493)Google Scholar, are described by An Jinhuai as belonging to the period of Erlitou III to IV (“Shilun Zhengzhou Shang he Yanshi Shang cheng de zaowan guanxi,” p. 6 and passim). Also see Zhongyuan wenwu 1985.3:92–94, 101Google Scholar.

61. Tai, Liet al., Kuadizhi jijiao (Beijing, 1984), ch. 3, p. 177Google Scholar. Also see, Daoyuan, Li in Guowei, Wang, Shuijingzhu jiao (Shanghai, 1984), ch. 7, p. 243Google Scholar. The so-calld Ao granaries, established in the vicinity of Zhengzhou in Qin times (Mi, Huangfu, Dinang shiji, in Shuijingzhu jiao, ch. 4, p. 243)Google Scholar, do not corroborate a direct link between Zhengzhou and the Shang city of Ao.

62. Jinben: Fang, p. 221; also see Mi, Huangfu, Diwang shiji in Shuijingzhu jiao, ch. 4, p. 243Google Scholar. It should be kept in mind that if the Huanghe made its bend to the northeast earlier in its course than it presently does--as apparently was the case in Chunqiu times, and possibly before--then Zhengzhou would have been at a considerably greater distance from the Huanghe than it is today and would have fit the description “on the He” even less than it does now. See, Maspero, Henri, “Contribution à l'étude de la société chinoise à la fin des Chang et au debut des Tcheou,” Bulletin de l'école française d'extrème orient 46 (1954):338Google Scholar; Cen Zhongmian, Huanghe bianqian shi, ch. 4, kexueyuan, Zhongguo, Zhongguo ziran dili: Dimao (Beijing, 1980), map p. 100Google Scholar.

63. Jinben: Fang, p. 204. Both versions agree in stating that during his reign Xiang resided at Shang (6/204).

64. Zhong Kang occurs in the Jinben only (Fang, p. 204). Likewise the Guben only refers to Xiang's residence at Zhenxin (Guben: Fang, p. 7).

65. Jinben: Fang, p. 204.

66. Jinben: Fang. p. 206.

67. Jinben: Fang, p. 207.

68. Jinben: Fang, p. 208.

69. Jinben: Fang, p. 211.

70. Jinben: Fang, p. 213.

71. Jinben: Fang, p. 214. The battle at Mingtao must have been one of the earliest major battles in China to be fought with Bronze Age weaponry. See note 57 for references to the immediately preceding campaigns. The Lord of Shang apparently returned from the area of his campaign to Bo, where he began his reign as the first Shang king.

72. Kaogu xuebao 1979.3:301–375, pls. 1–14Google Scholar.

73. Kaogu xuebao 1979.3:311Google Scholar.

74. Kaogu xuebao 1979.3:307, fig. 7Google Scholar.

75. Kaogu xuebao 1979.3:326, fig. 20Google Scholar; 345, fig. 18; 357, fig. 45.

76. Kaogu xuebao 1979.3:367, fig. 52: 1–3, 9Google Scholar.

77. Kaogu xuebao 1958.3: pl. 3:12, 1, 4Google Scholar.

78. Bohong, Wang, et al., Fengxi fajue baogao (Beijing, 1962), pp. 43–69, pls. 25–26Google Scholar.

79. Kaogu 1964.4:174–178, pls. 2–4Google Scholar; Kaogu 1977.4:229Google Scholar.

80. Kaogu xuebao 1982.4:427–476, pls. 1–10Google Scholar; Zhongguo kaogu xuehui diyici nianhui lunwen ji: 1979, p. 47Google Scholar.

81. Miaodigou yu Sanliqiao (Beijing, 1959), pls. 62–63, 66Google Scholar; Kaogu 1977.4:229Google Scholar.

82. Kaogu 1983.1:30–42, pls. 4–7Google Scholar.

83. Kaogu xuebao 1983.1:55–91, pls. 9–18Google Scholar, Zhonguo kaogu xuehui diyici nianhui lunwen ji: 1979, p. 47Google Scholar.

84. However, in the area of Yuxian, southeast of Zhengzhou, the pottery from several sites exhibits a clear and indisputable connection with that in Shandong--a connection persisting over a considerable time. At the Yuxian site of Gushuihe a general similarity to the pottery of the Oawenkou culture may be observed; but two stemmed goblets, one with a flat-bottomed cup (Kaogu 1978.1:29, fig. 8:14)Google Scholar and one with a round body (Kaogu 1978.1:29, fig. 8:22)Google Scholar. are virtually identical to two examples of the Dawenkou type discovered at Gangshan, Zouxian, Shandong (Kaogu 1963.7: pl. 5:7 and 5:5, respectively)Google Scholar. The pottery from the Gushuihe site corresponds closely to many types from the third period dt Dahecun, as evidenced by a comparison of vessels from the two sites (Kaogu 1978.1:28, fig. 7:3–6Google Scholar and Kaogu xuebao 1979.3:326, fig. 10:1–5, 8, 9)Google Scholar. The Gangshan pottery is roughly similar to that from the late period at the Shandong site of Wanyin in Yanzhou, which has been dated (ZK 317) to 3275 ± 130 B.C. (Kaogu xuebao 1978.4:417)Google Scholar. Again, the ding from both sites are closely alike (see (Kaogu 1963.7: pl. 6:10, 12Google Scholar; and Kaogu 1979.1:11, fig. 8:32, 38)Google Scholar. It should also be observed that the mortar, one of the characteristic vessels of the late Henan Longshan and Erlitou periods, appears for the first time at Gdngshan (Kaogu 1963.7:356, fig. 6)Google Scholar.

Although no exact parallels can be offered, the highly stylized, tall wheel-made drinking vessels of polished black ware found with typical late Henan Longshan pottery in the second and third levels at Wadian in Yuxian (Wenwu 1983.3:47, figs. 30–31; pl. 3:1–2)Google Scholar are surely influenced by the achievements of the classical Longshan potters of Shandong, whose work is familiar to us from finds at Donghaiyu in Rizhao (Kaogu 1976.6: pl. 6:1–4Google Scholar) and Yaoguanzhuang in Weifang (Kaogu 1963.7: pl. 2:1, 2, 4Google Scholar).

Thus, although it now seems that the influence of Shandong on the pottery at Dahecun subsided almost entirely by the period of the late Henan Longshan (Dahecun Period VI), communication with the east evidently lasted into the “late period at Yuxian, less than one hundred kilometers to the southeast. In view of the Yuxian finds, it seems almost surprising that evidence for a similar continuing communication with the east has not been documented for the Zhengzhou area as well.

85. The most striking of the pottery forms at Shangjie--a round-bodied ding ornamented with cord impressions and circumferential bands of relief appliqué and supported by wedge-shaped legs in radial position (Kaogu 1966.1:5, fig. 3:3Google Scholar)--finds a near parallel among the wares from the third sector at Erlitou, which have been ascribed to Period II (Kaogu 1984.7:586, fig. 4:3Google Scholar). More advanced versions of this vessel type, belonging to the period of Erlitou III, on which the outer edges of the wedge-shaped legs reach nearly to the top of the body of the vesse1, are recorded from the Meishan site in Linruxian (Kaogu xuebao 1983.4:467, figs. 1–2Google Scholar). The flat-bottomed pen from Shangjie (Kaogu 1966.1:3, fig. 14Google Scholar) is identical to an example from Period II at Erlitou (Kaogu 1984.7:586, fig. 15Google Scholar) and to another of the same period from Oongtnagou at Luoyang (Kaogu 1966.1:20 fig. 5:10Google Scholar) The shape of the li vessel from Shangjie with its comparatively tall and pointed legs (Kaogu 1966.1:5, fig. 3:10Google Scholar) appears less advanced than the more sturdy specimens from the third period at Erlitou (Kaogu 1984.7:587, fig. 6:9Google Scholar).

86. K. C. Chang, who predicted that “an important Early Shang culture will someday be found here [at Shangqiu],” calls attention to the obstacles presented to the archaeologists choosing to excavate in this area of the Huanghe flood plain, where the changing course of the river has created an accumulation of silt, in some places as much as eight meters deep (“The Origin of Shang and the Problem of Hsia in Chinese Archaeology,“ in Kuwayama, George, ed., The Great Bronze Age of China: A Symposium, p. 14Google Scholar; Archaeology of Ancient China, 4th ed., pp. 336–339Google Scholar; also see Tuan, Yi-fu, China [Chicago, 1969] pp. 18–19Google Scholar; Zhongmian, Cen, Huanghe bianqian shi, pp. 91–94Google Scholar; Kexueyuan, Zhongguo, Zhongguo ziran dili: Dimao, pp. 95–106Google Scholar.

87. Kaogu 1978.1:37, figs. 1, 2Google Scholar; 38, fig. 3, pls. 10, 11; Zhongguo kaogu xuehui diyici nianhui lunwen ji: 1979, p. 47.

88. Zhongguo kaoguxue zhong tan shisi niandai shouji 1965–1981 (Beijing, 1983), pp. 81–82Google Scholar.

89. Wenwu 1983.3:21–36, pl. 2Google Scholar. The incompletely excavated wall is judged to have been approximately square with each side measuring 185 meters long. Pottery from the lowest level (Period I) is thought to show an affinity to that of the later phase at Dawenkou, in Ningyang, Shandong. It should be mentioned that the late Henan Longshan wares are also represented in the Anyang area at the Baiying site near Tangyin (see Kaogu 1980.3:196, fig. 4, 199, fig. 7Google Scholar).

90. Huber, Louisa F., “The Traditions of Chinese Neolithic Pottery,” BMFEA 53 (1981):59–66Google Scholar; Huber, , “A Commentary on the Recent Finds of Neolithic Painted Pottery from Ta-ti-wan, Kansu,” Early China 9–10 (1983–1985):9Google Scholar.

91. Huber, , “Traditions,” pp. 78–84Google Scholar.

92. Ibid., pp. 102–125.

93. Tours of inspection are recorded for Yao (Jinben: Fang, p. 193); Shun (Jinben: Fang, p. 196); Yu (Jinben: Fang, p. 201); Qi (Jinben: Fang, p. 202); and Mang (Jinben: Fang, p. 208), but only for the first Shang ruler Cheng Tang (Jinben: Fang, p. 217). According to the Bamboo Annals, Yu, the first of the Xia rulers, resided at a place named Yangcheng (Guben: Fang, p. 1) or Ji (Jinben: Fang, p. 200). Of the following rulers, Tai Kang's residence is given as Zhenxin (Guben: Fang, pp. 4–5), where Zhong Kang is also said to have lived (Jinben: Fang, pp. 203–204), Xiangls residence is given as Zhenguan (Guben: Fang, p. 7), or as Shang and then Zhenguan (Jinben: Fang, p. 204). Shao Kang, late in his reign, moved to Yuan (Jinben: Fang, p. 207). Under the ruler Zhu, the residence was changed to Laoqiu (Fang, pp. 8/207). The residences of the succeeding rulers are not specified until the time of Yin Jia and Kong Jia who lived at Xihe (Fang, pp. 13/210). The final ruler Jie, as previously mentioned, is said to have resided at Zhenxin (Fang, pp. 15/212).

94. Kaogu 1983.3:203, fig. 9:9Google Scholar. A related motif decorates the corners of a jade handle discovered earlier at Erlitou (Kaogu 1976.4: pl. 6:2Google Scholar, right [color]; 229–233, 259–263).

95. Kaogu 1978.1:3Google Scholar; Wenwu 1984.2:61Google Scholar; Kaogu 1984.4:355Google Scholar.

96. Examples of small ceramic pouring vessels have been recovered from the following Henan Longshan sites: Sanliqiao III, Shanxian (Miaodigou yu Sanliqiao, pl. 88:4a–b); Chigouzhai, Guangwuxian (Andersson, J. G., BMFEA 15 [1943]: pl. 35:1Google Scholar); Buzhdozhai, Mienchixian (ibid., pl. 35:2); Xiaoban'gou, Mengjia (Kaogu 1974.4:251, fig. 11:3Google Scholar); Meishan I, Linruxian (Kaogu xuebao 1983.4:438, fig. 11:27Google Scholar); Meishan, II (Kaogu xuebao 1983.4:451, fig. 21:22Google Scholar); Baiying, Tangyin, northern Henan (Kaogu 1980.3:199, fig. 7:5. 6Google Scholar); Erlitou (early period) (Kaogu 1965.5:219, fig. 7:5Google Scholar).

97. See Huber, Louisa F., “The Relationship of the Painted Pottery and Long-shan Cultures,” in Keightley, David N., ed., The Origins of Chinese Civilization, pp. 207–209Google Scholar. Regarding the possible influence of western metallurgy on the beginnings of the Chinese Bronze Age, it is important to note that the first find of a cast copper knife--definitely of a non-Chinese type--along with a cast copper socketed axe was made at a Qijia site in Minxian, Gansu, (Kaogu 1985.11:978, fig. 2:10, 11Google Scholar).

98. Kaogu 1976.4: pl. 5:4–6Google Scholar; pl. 6: lower.

99. Kaogu 1965.5: pl. 2:2; pl. 1:11Google Scholar; Kaogu 1978.1:21, fig. 6:2Google Scholar (Erlitou Period II at Luoyang Dongmagou); Huber, “Traditions,” pl. 99:243, 244.

100. Kaogu xuebao 1980.3: pl. 9:1Google Scholar; 364, fig. 31:8; Huber, “Traditions,“ pl. 93:230.

101. Kaogu xuebao 1980.3: pl. 10:4Google Scholar; Huber, “Traditions,” pl. 95:235.

102. Kaogu 1983.3:204, fig. 10:7Google Scholar; a second gui likely to have been an eastern import has also been found at Erlitou; see Kaogu 1975.5: pl. 9:1Google Scholar.

103. Salmony, Alfred, Carved Jade of Ancient China (Berkeley, 1938), pl. 33:1–2Google Scholar; Important Chinese Works of Art: The Collection of Mr. and Mrs. Richard C. Bull (Sotheby Parke Bernet auction catalog for the sale of 6 12 1983), no. 174Google Scholar.

104. Kaogu 1972.4:57Google Scholar, fig. 2

105. Kaogu 1983.3: pl. 1:2 (color)Google Scholar.; Huber, “Traditions,” pl. 96:236.

106. Loehr, Max, Ancient Chinese Jades from the Grenville L. Winthrop Collection (Cambridge, Mass., 1975), no. 329Google Scholar; Salmony, Carved Jades of Ancient China» pl. 25:2. For a modern copy of this jade type, see Hansford, S. Howard, Jade Essence of Hills and Streams : The Von Oertzen Col lection of Chinese and Indian Jades (New York, 1969), no. A 54Google Scholar.

107. Kaogu 1978.4:270, fig. 2:1Google Scholar; Wen Fong, ed., The Great Bronze Age of China, no. 2; Kaogu 1983.3: pl. 1:4 (color)Google Scholar.

108. Compare, for instance, the jades recently unearthed from the burials at Taosi, Xiangfenxian, Shanxi (Kaogu 1983.1:38, fig. 10Google Scholar).

109. Kaogu xuebao 1982.3:310, fig. 25Google Scholar; 314, fig. 29:2–3. The painted decoration on some of the Xiajiagang jades (e.g., Kaogu xuebao 1982.3:309, fig. 13:3–4Google Scholar) is very similar in character to the designs incised on the chisel recovered from the late Shandong Longshan site of Liangchengzhen (Kaogu 1972.4:57, fig. 2Google Scholar; Huber, “Traditions,” pl. 96:236), suggesting a continuity in the evolution of the eastcoast jade tradition.

It should be noted that although the zhang is represented among the objects recovered from one of the large pits (K1) at the Guanghan site of Sanxingdui in Sichuan, there is no reason to think that this type of jade object was the invention of Sichuan artists (Wenwu 1987.10:8, fig. 12:3, pl. 3:2Google Scholar). The precise date of the Guanghan jades is not yet established, and, while some material remains from this site indicate contact with the Zhong Yuan as early as the late Xia culture of the Erlitou period, the bronzes found there suggest that the Guanghan elite remained under the influence of the Shang culture for a century or more (ibid., p. 14).

110. Fang, pp. 8/207.

111. Fang, pp. 10/208.