This article is a short version of the second part of a book that the author is preparing for publication. The book discusses the relationship between early Chinese art and ancestral worship, the most important religious form of early China. The first part of the book examines the nature and development of liqi --the art of ritual vessels--from prehistoric times to the Eastern Zhou, and the third part focuses on Han funerary art. The present article was first written for the 1987 Symposium on Han Art at the University of Michigan and the 1987 meeting of the American Oriental Society (Western Branch) at Berkeley. I want to thank Professors Jeffrey Riegel, Ken DeWoskin, David Keightley, and William Boltz for their support. This article was submitted in final form on 20 February 1988.

1. Keightley, David N., “The Religious Commitment: Shang Theology and the Genesis of Chinese Political Culture,” History of Religions 17.3–4 (1978):217CrossRefGoogle Scholar.

2. Cf. Li ji, in Yuan, Ruan, Shisanjing zhushu (Beijing: Zhonghua, 1980), pp. 1605–1607Google Scholar. Different dates have been attributed to Li ji. Most likely, this book was compiled during the early Han. See Legge, J., Li Chi: Book of Rites (New York: University Books, 1967), vol. 1, p. xlviGoogle Scholar. Because my discussion in this article concerns some general principles of early Chinese religion and art, Li ji and other later documents are used as secondary sources complementing archaeological evidence.

3. Li ji, pp. 1495, 1508, 1589; see too, Chang, K. C., Art, Myth and Ritual (Cambridge, Hass.: Harvard University Press, 1983), pp. 37–41Google Scholar.

4. Institute of Archaeology, Yinxu Fu Hao mu (Beijing: Wenwu, 1980), pp. 221–228Google Scholar.

5. Zhou li records a general and perhaps idealized plan of the capital city. Zhou li, in Yuan, Ruan, Shisanjing zhushu, pp. 927–28Google Scholar. The archaeological excavations at Anyang, Zhengzhou, and Panlongcheng have all demonstrated that large ritual structures were built inside a town, but burial districts were located outside. See, Chang, K. C., Shang Civilization (New Haven, Conn.: Yale University Press, 1980), pp. 90–95, 110–124, 272–283, 297–306Google Scholar.

6. Li ji, p. 1589.

7. Li ji, p. 1258, translation based on Legge, , Li Chi, pp. 103–104Google Scholar. Legge translated the character jia as “clan,” but the corresponding term for “clan” in Li ji (p. 1508) is zu . Here the character jia means “ineage” or “extended family.”

8. Li ji, p. 1508; see too, Chang, , Art, Myth, and Ritual, pp. 15–16Google Scholar.

9. “[A noble man] maintains his ancestral temple and represents sacrifices reverently at proper seasons. This brings order to his “lineage and clan.” Li ji, p. 1611.

10. Li ji, pp. 1439, 1441, 1595.

11. Waley, A., trans., The Book of Songs (New York: Grove Press, 1978), pp. 239–280Google Scholar; see too, Chang, , Art, Myth, and Ritual, pp. 10–15Google Scholar.

12. Chang, , Shang Civilization, pp. 272–283, 297–306Google Scholar.

13. Thorp, R., “Origins of Chinese Architectural Style: The Earliest Plans and Building Types,” Archives of Asian Art 36 (1983): 22–26Google Scholar.

14. Zhen, Dai, Kaogongji tu (Shanghai: Shangwu, 1955), pp. 104–107, 113Google Scholar. Zou Heng first noted similarities in architectural plan between the Erlitou building and the Three Dynasties ancestral temple reconstructed by Dai Zhen. He also contended that the Erlitou building was a temple. Heng, Zou, Shang-Zhou kaogu (Beijing: Wenwu, 1979), p. 27Google Scholar.

15. Arnheim, R., New Essays on the Psychology of Art (Berkeley: University of California Press, 1986), p. 83Google Scholar.

16. Thorp, , “Origins of Chinese Architectural Style,” pp. 26–31Google Scholar.

17. Li ji, p. 1416; translation based on Legge, , Li Chi, vol. 1, pp. 370–371Google Scholar; emphasis mine.

18. This idea is discussed in a book the author is preparing for publication. Zoomorphic images on Shang and Western Zhou ritual paraphernalia were not the ultimate objects of worship. They are not self-contained; rather, they appear as the surface elements of ritual objects. In fact, Chinese art before the Eastern Zhou was essentially noniconic; spirits and deities remained shapeless, and thus ancestral deities, the most important subjects of religious worship in ancient China, were represented merely by plain wooden tablets. More likely, these zoomorphic images functioned to “animate” inanimate materials--a piece of metal or stone; similar beliefs existed in many traditions in the ancient world, along with a universal belief in the magic power of eyes or masks. See, Gombrich, E. H., The Sense of Order (Ithaca, N.Y.: Cornell University Press, 1979), pp. 257–262Google Scholar. This suggestion agrees with K. C. Chang's assumption that ritual bronzes, as well as bronze decoration, were a means of religious communication between man and divinities (Chang, , Art Myth and Ritual pp. 56–80)Google Scholar TMs suggestion may also explain an essential characteristic of Shang and Western Zhou zoomorphs: their protean shapes and incessant permutations These zoomorphs are not formulated icons but assume endless variations; they combine features from different animal species but are never turned into naturalistic representations. These varying imaaes seem to attest to a painstaking effort to create metaphors for an intermediate state between the supernatural emd reality--something that one could depict but could not portray.

19. Institute of Archaeology, Yinxu Fu Hao mu, pp. 4–6Google Scholar. Remains of above-ground mortuary houses from the Later Shang dynasty have been also found in Dasigong village. Dezhi, Ma and others, “1953 nianqiu Anyang Dasikongcun fajue baogao” 1953 , Kaogu xuebao , 1953.9:25–40Google Scholar.

20. Chang, , Shang Civilization, pp. 119–124Google Scholar.

21. Zhou li, pp. 757–759.

22. Li li, pp. 1595–1596.

23. It is important to distinguish “funerary rituals” (sangli from “grave sacrifices” (muji ) in ancient Chinese ritual practices. The former was carried out immediately following a death, but the latter refers to rituals practiced routinely in graveyards. The three ritual canons describe the “funerary ritual” at great length but omit all mention of the “grave sacrifice.“

24. The origin of “grave sacrifice” has been the focus of a debate among Chinese scholars for some two thousand years. Scholars of the Eastern Han held similar opinions on this issue. A typical statement was provided by Wang Chong : “Ancient people held ancestral sacrifices in temples; nowadays people have the custom of sacrificing in graveyards” (Pansui, Liu, Lunheng jijie [Beijing: Guji, 1957], p. 469.Google Scholar) Similar statements also appear in the writings of two distinguished Eastern Han historians, Ying Shao and Cai Yong . Ying Shao reported: “In ancient times the ritual of ‘grave sacrifice’ did not exist …. Emperors [of the present] hold ceremonies in the first month of the year at the Yuan ling Mausoleum” (quoted in Fan Ye , Hou Han shu [Beijing: Zhonghua, 1965], vol. 2, p. 99)Google Scholar. The same idea was expressed even more categorically in a statement by Cai Yong quoted in note 58 below.

Confusion about the origin of grave sacrifices first appeared during the Jin dynasty. Sima Biao's statement that “the mausoleum sacrifice had been practiced during the Western Han” (Hou Han shu, p. 3103) contradicted the belief of Eastern Han historians that the ceremony had not existed unti1 it was created by Emperor Ming. Going even further, Sima Biao argued that the “grave sacrifice” had actually been initiated during the Qin dynasty: “In ancient times there did not exist the ritual of ‘grave sacrifice.“

The Han mausoleums all contained a qin hall because the Han followed the regulation of the Qin” (Hou Han shu, pp. 3199–3200).

This issue became a favorite topic in the Qing dynasty. Some scholars, such as Xu Qianxue and Gu Yanwu , disagreed with Sima Biao and returned to the view of Han historians (Qianxue, Xiu, Duli tongkao 94[1696]: 1bGoogle Scholar; Rucheng, Huang, Rizhi lu jishi [Hubei: Chongwen, 1872], 15:1–3)Google Scholar. Other scholars, represented by Yan Ruoju and Sun Yirang , derived their evidence from historical writings and the Confucian classics to prove that the ritual of “grave sacrifice” had not only existed during the Qin and Western Han but had actually been practiced by disciples of Confucius and even by King Wu of the Zhou dynasty (Ruoju, Yan, Duli congchao , in Huang-Qing jingjie , 20:14a–bGoogle Scholar; Yirang, Sun, Zhou li zhengyi [Shanghai: Zhonghua, 1934])Google Scholar.

This debate has been renewed during recent years. In 1982, Yang Hongxun declared that the ritual practice of grave sacrifice can be traced back as far as the Shang dynasty. He provided archaeological evidence including the sacrificial ground discovered at the Shang royal cemetery, the remains of Shang-Zhou above-ground mortuary houses, and the “design of the royal cemetery of the Zhongshdn kingdom” (“Guanyu Qindai yiqlan mushang jianzhu de wenti“ , Kaogu 1982.4:403Google Scholar). Opposed to him was the noted historian, Yang Kuan , who followed Gu Yanwu and Xu Qianxue and concluded that the pre-Han funerary structures found in archaeological excavations “could only be qin-halls--the ‘retiring room’ for the soul of the deceased, rather than ‘offering halls’ in which people held ritual sacrifices to their ancestors” (“Xian-Qin mushang jianzhu he lingqin zhidu” Wenwu 1982.1:31Google Scholar; “Xian-Qin mushang jianzhu wenti de zaitantao” , Kaogu 1983.7:636–640Google Scholar). A third scholar, Wang Shimin returned to Sima Biao's opinion, saying that the official “grave sacrifice” was initiated by the First Qin Emperor (Shimin, Wang, “Zhongguo Chunqiu Zhanguo shidai de zhongmu” , Kaogu 1981.5:465)Google Scholar.

In my opinion, this debate has been caused by different understandings and uses of the term muji, or “grave sacrifice.” Yan Ruoju, Sun Yirang, and Yang Hongxun all understood the word in its general, broad sense as any kind of ritual held in a graveyard; Sima Biao and Wang Shimin understood it in a narrower sense as the imperial sacrifices held in mausoleums; the Han dynasty scholars, together with Gu Yanwu, Xiu Qianxue, and Yang Kuan, used the term in a specific sense to mean the most important official ancestral sacrifice held in the royal mausoleums in the first month of the year--that is, the shangling ceremony initiated by Emperor Ming of the Eastern Han in A.D. 58.

These scholars, in short, have focused on two different historical problems: one is the origin of the ritual practice of grave sacrifice; the other is the formulation of the official “mausoleum sacrifice.“ Both kinds of ritual are important for understanding the development of ancestral worship in ancient China.

25. Li ji, p. 1275.

26. Li ji, p. 1292.

27. Two reconstruction plans of the Zhongshan mausoleum have been provided. Hongxun, Yang, “Zhanguo Zhongshan wangling ji Zhaoyutu yanjiu” , Kaogu xuebao 1980.1:119–137Google Scholar; Xinian, Fu, “Zhanguo Zhongshanwang Cuo mu chutu Zhaoyutu jiqi lingyuan guizhi de yanjiu” , Kaogu xuebao 1980.1:97–119Google Scholar. In addition to the plan of the Zhongshan Mausoleum, an important Warring States mausoleum was excavated in Guweicun , Hebei; it has been identified as a royal burial of the state of Wei . The plan of the mausoleum was similar to that of the Zhongshan Mausoleum: the square funerary park, encircled by double walls, was centered on three individual sacrificial structures built in a row on a terrace. See Hongxun, Yang, “Zhanguo Zhongshan wangling,” pp. 131–12Google Scholar.

28. Zhou li, p. 786. This passage records that the “zhongren” or officer in charge of tombs, was “to regulate the measurement of the tumuli and the number of trees [planted in the graveyards] according to the rank of the deceased.“

29. Shizhe, Zhu, Shangjun shu jiegu dingben (Beijing: Guji, 1956) 5:74Google Scholar.

30. Weiyu, Xu, Lüshi chunqiu jishi (Beijing: Zhongguo Shudian, 1985) 10:8bGoogle Scholar.

31. General excavation reports and studies on the Lishan Mausoleum include: Shaanxi Provincial Administrative Committee of Cultural Relics, “Qinshihuang ling diaocha baogao” , Kaogu 1962.8:407–411Google Scholar; Thorp, R., “An Archaeological Reconstruction of the Lishan Necropolis,” in Kuwayama, G., ed., The Great Bronze Age of China: A Symposium (Los Angeles: Los Angeles County Museum of Art, 1983), pp. 72–83Google Scholar, Kuan, Yang, Zhongguo gudai lingqin zhidushi yanjiu (Shanghai: Guji, 1985), pp. 183–201Google Scholar.

32. Qian, Sima, Shi ji (Beijing: Zhonghua, 1959), 6:265Google Scholar.

33. See above note 18.

34. Wei, Han, “Fengxiang Qingong lingyuan zuantan yu shijue jianbao” , Wenwu 1983.7:30–37Google Scholar, Shaanxi Provincial Archaeological Institute, “Qin dongling diyihao lingyuan tanchaji” , Kaogu yu wenwu 1987.4:19–28Google Scholar.

35. Xuehui, Lishan, , “Qin dongling tancha chuyi” , Kaogu yu wenwu 1987.4:86–88Google Scholar. My discussion does not follow the Lishan Xuehui's suggestion that the Lishan Mausoleum was one of the Lintong Qin royal tombs.

36. Remains of buildings have been found between the two walls of the funerary park. Pottery vessels unearthed there are stamped with inscriptions such as Lishanyuan (Lishan Mausoleum) and Lishan siguan (the sacrificial officer of the Lishan Mausoleum), and thus identify these buildings as departments of ritual affairs. See Kangmin, Zhao, “Qinshihuang ling yuanming Lishan” , Kaogu yu wenwu 1980.3:34–38Google Scholar.

37. This architectural complex consisted of four individual buildings arranged in an east-west row. The largest was about 20 meters long and 3.4 meters wide, with an elaborate limestone doorway. Its walls were constructed of plastered stone, and the walkways were paved with stones. Other remains, including post holes, large bricks, and decorated tiles, suggest that these buildings were wooden-framed structures with tiled roofs. Kangming, Zhao, “Qinshihuang lingbei 2, 3, 4 hao jianzhu yizhi” , Wenmu 1979.12:13–16Google Scholar.

38. Yong, Cai, Du duan , in Rong, Cheng, Han-Wei congshu 14:20Google Scholar, see, too, Kuan, Yang, Zhongguo gudai lingqin, p. 638Google Scholar.

39. Xuehui, Lishan, “Qin dongling tancha,” p. 89Google Scholar.

40. Quoted in Xuehui, Lishan, “Qin dongling tancha,” p. 89Google Scholar.

41. Important archaeological reports regarding excavations in the outer district of the Lishan Mausoleum include: Archaeological Team of the Qin Terra-cotta Army. “Lintongxian Qinyongkeng shijue diyihao jianbao” , Wenwu 1975.11:1–18Google Scholar; “Qinshihuang ling dongce dierhao b1ngma yongkeng zuantan shijue jianbao” , Wenwu 1978.5:1–19Google Scholar; “Qinshihuang ling dongce disanhao bingma yongkeng qingli jianbao” , Wenwu 1979.12:1–12Google Scholar; “Lintong Shangjiaocun Qininu qingli jianbao” , Kaogu yu wenwu 1980.2:42–50Google Scholar; “Qinshihuang ling dongce majiukeng zuantan qingli jianbao” , Kaogu yu wenwu 1980.4:31–41Google Scholar, “Qinshihuang ling xice Zhaojiabeihucun Qin xingtu mu” , Wenwu 1982.3:1–11Google Scholar; “Qinshihuang lingyuan peizangkeng zuantan qingli jianbao” , Kaogu yu wenwu 1982.1:25–29Google Scholar.

42. The predynastic Qin temple system is not clear. According to Shi ji vol. 6, p. 266Google Scholar, the most important royal temples were located in the two capitals, Yong and Xianyang. Other records, however, suggest that from King Zhao's reign a temple was built near a deceased ruler's mausoleum; see Kuan, Yang, Zhongguo gudai lingqin, p. 24Google Scholar.

43. Qian, Sima, Shi ji, 6:241Google Scholar.

44. Qian, Sima, Shi ji, 6:266Google Scholar.

45. Li ji, p. 1292; translation based on Legge, , Li Chi 1:155–156Google Scholar.

46. Zhongshu, Wang, Han Civilization, trans. Chang, K. C. (New Haven: Yale University Press, 1982), pp. 175–179Google Scholar.

47. Hung, Wu, The Wu Liang Shrine: The Ideology of Early Chinese Pictorial Art (Stanford University Press, forthcoming)Google Scholar.

48. Fuyi, Luo, “Xiang Tajun shicitang tizi jieshi” , Gugong bowuyuan yuankan 1960.2:178–180Google Scholar; emphasis mine.

49. Quoted in Kongyang, Zhu, Lidai lingqin beikao (Shanghai: Shenbao, 1937), 13:5aGoogle Scholar.

50. Li ji (p. 1589): “The mass of ordinary officers and the common people had no ancestral temple. Their dead were left in their ghostly state”; trans. Legge, , Li Chi, 2:206Google Scholar.

51. According to Sanfu huangtu and the biography of Wei Xuancheng in Han shu, a Western Han mausoleum contained four major architectural units: the qin or “retiring hali,“ the biandian . or “side hall,“ the yeting or place for concubines, and the imperial offices. Gu, Ban, Han shu, (Beijing: Zhonghua, 1962), 73:3115–3116Google Scholar.

52. According to textual evidence, free-standing sculptures of the Qin. such as the famous Twelve Golden Men, were erected above ground only in the imperial palace. See Qidn, Sima, Shi ji 6:239Google Scholar. From the epoch of Emperor Wu of the Western Han, large stone figures for funerary purposes began to be placed on the ground in front of mausoleums. It has been reported that remains of such statues were found in front of Emperor Wu's tomb, flanking the Spirit Path; see Mincong, Xie, Zhongguo lidai diling kaolue (Taibei: Zhongzheng, 1979), p. 58Google Scholar. The authenticity of this report may be supported by the existence of stone statues in front of Huo Qubing's tomb. Huo was a famous general under Emperor Wu, and his tomb was one of the “satellite burials” of the emperor's mausoleum.

From the beginning of the Han dynasty, a “satellite burial” system was established. In this system, an emperor's tomb was the focal point of the mausoleum complext surrounded by the tombs of his empress, concubines, relatives, and meritorious ministers and generals. The scale of such a burial is astonishing. According to archaeological surveys, at least 175 satellite burials surrounded Emperor GaozuTs tomb, 34 surrounded Emperor Jing's, and 59 surrounded Emperor Xuan's. See Qingzhu, Liu and Yufang, Li, “Xi-Han zhul ing diaocha yu yanjiu” , Wenwu cankao ziliao 1982.6:1–15Google Scholar; Yang Kuan, Zhongguo gudai lingqin, Appendix I.

53. Qian, Sima, Shi ji, 99:2725–2726Google Scholar.

54. Pizhong, Wanget al., “Han Jingdi Yang ling diaocha jianbao” , Kaogu yu wenwu 1980.1:36–37Google Scholar; Hongtao, Li and Pizhong, Wang, “Han Yuandi Welling diaochaji” , Kaoku yu wenwu 1980.1:40–41Google Scholar. It is also reportea tnat a large architectural foundation, 51 meters long and 29.6 meters wide, was found southeast of Emperor Xuan's mausoleum. The reporter has identified this building as a qin-hall, but, because the foundation was located outside the funerary park, it was more probably the remains of a miao. See the Archaeological Team of the Du Mausoleum, “1982–1983 nian Xi-Han Duling de kaogu gongzuo shouhuo” 1982–1983 , Kaogu 1984.10:887–894Google Scholar.

55. Gu, Ban, Han shu, 43:2130Google Scholar.

56. For example, we read in texts that King Wuling of the Zhao passed on the throne to his younger son in the Zhao ancestral temple; that Ying Zheng , the future emperor of Qin, stayed overnight in the Qin ancestral temple at Yong before becoming the heir; and that all the emperors of the Western Han designated their crown princes in the Great Ancestral temple dedicated to the founder of the dynasty (Qian, Sima, Shi ji, 43:1812Google Scholar; Zongxiang, Zhang, Jiaozheng sanfu huangtu [Shanghai: Gudian Wenxue, 1958], p. 43)Google Scholar. Again, according to Han shu 73:3115–3116Google Scholar, only daily meals were offered to a deceased emperor in his qin, while more important monthly sacrifices were held in his miao.

57. Ye, Fan, Hou Han shu, 2:99Google Scholar.

58. For example, Cai Yong stated in A.D. 172 after attending the “mausoleum sacrifice” that year: “I heard that the ancients did not sacrifice to their ancestors in graveyards, and I had doubted the need for the ritual of ‘mausoleum sacrifice’ practiced by the present court. Only after I witnessed the dignified manner of the ceremony today, did I begin to realize its original motivation. Now I understand the complete filial piety and sincerity of Emperor Filial Ming [who initiated the ritual]. It would be improper to replace this ritual with the old custom.” Ye, Fan, Hou Han shu, p. 3101Google Scholar.

59. Ibid., pp. 27–28:3194–3195.

60. Ibid., p. 32.

61. Ibid., pp. 1–87.

62. Rucheng, Huang, Rizhi lu jishi 15:2–3Google Scholar.

63. Ye, Fan, Hou Han shu, pp. 123–124Google Scholar.

64. Yi, Zhao, Haiyu congkao, Dushu taji congkan (Taibei, 1960), 1, no.3:32abGoogle Scholar.

65. Powers, Martin, “Pictorial Art and Its Public in Early Imperial China,” Art History 7.2 (1984): 135–163Google Scholar.

66. During the Eastern Han, the funerary shrine of a family was open to the public. This is most clearly indicated by the inscription engraved on the An Guo shrine dated to A.D. 158: “You, observers, please offer your pity and sympathy. Then may your longevity be as gold and stone and may your descendants extend your line for ten thousand years. You, boys who herd horses and tend sheep and cows, are all from good famlies. If you enter this hall, please do not destroy the hall or make any trouble: this will cause disaster to your descendants. We are stating clearly to people of virtue and kind-heartedness within the four seas: please scrutinize these words and do not ignore them.“ See Falin, Li, Shandong Han huaxiangshi yan jiu (Jinan: Qilu Shushe, 1982), p. 101Google Scholar.

67. See the articles by Powers, M., DeWoskin, K. J., and Hung, Wu in Stories from China's Past (San Francisco: Chinese Cultural Center, 1987):54–82Google Scholar.

68. Shou, Chen, Sanguo zhi (Beijing: Zhonghua, 1959), p. 81Google Scholar; Xuanling, Fang, Jin shu (Beijing: Zhonghua, 1974) 20:634Google Scholar.

69. To my knowledge, at least eight such tombs have been discovered--four in Shandong, three in Henan, and one in Sichuan. See Institute of the Wu Family Shrines, Jiaxiang County, “Shandong Jiaxiang Songshan faxian Han huaxiangshi“ , Wenwu 1979.9:1–6Google Scholar, Committee of Cultural Relics, Jining District, “Shandong Jiaxiang Songshan 1980 nian chutu de Han huaxiangshi” , Wenwu 1982.5:60–70Google Scholar; Institute of Cultural Relics, Jiaxiang County, “Jiaxiang Wulaowa faxian yipi huaxiangshi” , Wenwu 1982.5:71–78Google Scholar, Archaeological Team of the Henan Provincial Cultural Bureau, “Henan Nanyang Dongguan Jin-mu” Kaogu 1963.1:25–27Google Scholar, Rulin, Wang, “Henan Nanyang Xiguan yizuo gumu zhongde Han huaxiangshi” , Kaogu 1984.8:424–426Google Scholar; Museum, Nanyang, “Nanyang faxixsan Dong-Han Xu Aqu muzhi huaxiangshi” , Wenwu 1974.8:73–75Google Scholar; Yanxiang, Xie “Sichuan Pixien Xipu chutu de Dong-Han canbei” , Wenwu 1974.4.67–71Google Scholar.