1. bowuguan, Hunan sheng and kexueyuan, Zhongguo yanjiusuo, kaogu “Changsha Mawangdui er san hao Han mu fajue jianbao” , Wenwu 7 (1974), 39–48, 63Google Scholar; kexueyuan, Zhongguo yanjiusuo, kaogu and bowuguan, Hunan sheng, “Mawangdui er san hao Han mu fajue de zhuyao shouhuo” Kaogu 1 (1975), 47–57, 61Google Scholar; Shimin, Wang , “Mawangdui boshu” , in Zhongguo dabaike quanshu, Kaoguxue (Beijing and Shanghai: Zhongguo dabaike quanshu, 1986), 307–9Google Scholar. For a general discussion of Mawangdui manuscripts in English, see Riegel, Jeffrey K., “A Summary of Some Recent Wenwu and Kaogu Articles on Mawangdui Tombs Two and Three,” Early China 1 (1975), 10–15 CrossRefGoogle Scholar; Harper, Donald and Riegel, Jeffrey K., “Mawangdui Tomb Three: Documents,” Early China 2 (1976), 68–72 CrossRefGoogle Scholar; and Loewe, Michael, “Manuscripts Found Recently in China,” T'oung Pao 63 (1977), 114–25CrossRefGoogle Scholar. My account of the Mawangdui manuscripts below is based on the most recent scholarship and private communication with Mr. Chen Songchang.

2. Barnard, Noel, “The Ch'u Silk Manuscript and Other Archaeological Documents of Ancient China,” in Early Chinese Art and its Possible Influence in the Pacific Basin, eds. Barnard, Noel and Donald, Fraser (New York: Intercultural Arts Press, 1972), 77–102Google Scholar; Ling, Li , Changsha Zidanku Zhanguo Chu boshu yanjiu (Beijing: Zhonghua, 1985)Google Scholar; Zongyi, Rao and Xiantong, Zeng , Chu boshu (Hong Kong: Zhonghua shuju Xianggang fenju, 1985)Google Scholar.

3. For a useful bibliography on the studies on the Mawangdui tombs and associated artifacts, see Hiroyuki, Kondō , “Baōtai Kanbo kankei roncho mokuroku” Mi , Chūgoku shutsudo shiryō kenkyū 1 (1997), 251–199Google Scholar (sic) and the corrections and additions by the same author, “Baōtai Kanbo kankei roncho mokuroku no teisei to tsuika” , Chūgoku shutsudo shiryō kenkyū 2 (1998), 264–60 (sic,)Google Scholar.

4. The Chinese word tu has been variously translated as “diagram,” “chart,” “drawing” and so forth. Here I use “diagram” in the sense that it is “a graphic design that explains rather than represents; especially, a drawing that shows arrangement and relations (as of parts)” (Merriam-Webster Collegiate Dictionary, 10th ed., 1995). The concept of tu in China, which was the focus of the 2001 Paris conference, encompasses a wide range of symbolic, illustrative, and indexical functions. For a detailed discussion of the Han notion of tu as “diagram,” see Harper, Donald, “Communication by Design: A Study of a Mawangdui Silk Manuscript of Diagrams (Tu),” paper presented at the European and North American Exchanges in East Asian Studies Conference “From Image to Action: The Dynamics of Visual Representation in Chinese Intellectual and Religious Culture” at the Collège de France, Paris, September 3–5, 2001 Google Scholar. For a general discussion of the notion of tu in the post-Han, especially Song, period, see Reiter, Florian C., “Some Remarks on the Chinese Word T'u ‘Chart, Plan, Design”,” Oriens 32 (1992), 308–27CrossRefGoogle Scholar.

5. Han shu (Beijing: Zhonghua, 1962), 30. 1701–84Google ScholarPubMed, especially 1703,1756–62, and 1766.

6. Li Ling, Changsha Zidanku Zhanguo Chu boshu yanjiu; Rao Zongyi and Zeng Xiantong, Chu boshu; for an English summary, see Harper, Donald, “Warring States Natural Philosophy and Occult Thought,” in The Cambridge History of Ancient China: From the Origins of Civilization to 221 BC, ed. Michael, Loewe and Edward L., Shaughnessy (Cambridge University Press, 1999), 845–47Google Scholar; For an English translation of the texts on this diagram, see Ling, Li and Cook, Constance A., “Translation of the Chu Silk Manuscript,” in Defining Chu: Image and Reality in Ancient China, ed. Constance A., Cook and John S., Major (Honolulu: University of Hawai'i Press, 1999), 171–76Google Scholar.

7. Li Ling thinks the design of the inner two texts and the outer images on the diagram is intended to resemble a divination board or cosmograph (shi ). It is a model of the cosmos, which consists of a rotating round board (Heaven) on top of a square board (Earth); see Ling, Li, Zhongguo fangshu kao (Beijing: Dongfang, 2000), 89–176,178–96Google Scholar.

8. For other examples, see Tseng, Lan-ying, “Picturing Heaven: Image and Knowledge in Han China,” Ph.D. dissertation, Harvard University, 2001 Google Scholar.

9. Fujuyou, and Songchang, Chen, Mawangdui Han muwenwu (Changsha: Hunan, 1992), 151 Google Scholar. The earliest maps discovered so far are a set of seven maps drawn on wooden boards, dated to 239 b.c.e., excavated from tomb no. 1 at Fangmatan , Tianshui , Gansu. For these maps, see Shuangquan, He , “Tianshui Fang matan Qin mu churu ditu chutan” , Wenwu 2 (1989), 12–22 Google Scholar. For better color plates of these maps, see Wanru, Cao et al. eds., Zhongguo gudai dituji: Zhanguo—Yuan (Beijing: Wenwu, 1990), plates 4–16 Google Scholar.

10. Fu Juyou and Chen Songchang, Han mu wenwu, 152.

11. Fu Juyou and Chen Songchang, Han mu wenwu, 153. For the reconstructions of these maps, see Han, Mawangdui mu boshu zhengli xiaozu , “Changsha Mawangdui san hao Han mu chutu ditu de zhengli” , Wenwu 2 (1975), 35–42 Google Scholar; Qixiang, Tan , “Erqian yibai duo nian qian de yifu ditu” , Wenwu 2 (1975), 43–48 Google Scholar; Zhongmin, Han , “Guanyu Mawangdui boshu gu ditu de zhengli yu yanjiu” , in Zhongguo gudai ditu ji: Zhanguo—Yuan, 12–17 Google Scholar.

12. Han, Mawangdui mu boshu zhengli xiaozu, Mawangdui Han mu boshu , 4 (Beijing: Wenwu, 1985)Google Scholar, color plate 2, plates 83–86, text 134; Jixing, Ma , Mawangdui gu yishu kaoshi (Changsha: Hunan kexue jishu, 1992), 817–21Google Scholar; Jianmin, Li , “Mawangdui Han mu boshu ‘Yuzang maibao tu’ jianzheng” , Zhongyang yanjiuyuan Lishi yuyan yanjiusuo jikan , 65.4 (1994), 725–832 Google Scholar; Yoshio, Ikai , “Baōtai nanpō uzō. zukō” , Ryukoku shidan 103.104 (1994), 47–69 Google Scholar; Harper, Donald, Early Chinese Medical Literature: The Mawangdui Medical Manuscripts (London and New York: Kegan Paul International, 1998), 374–77Google Scholar.

13. Mawangdui Han mu boshu zhengli xiaozu, Han mu boshu, 4, text 133; Ma Jixing, Mawangdui gu yishu kaoshi, 814–17; Harper, Early Chinese Medical Literature, 372–73.

14. Ling, Li and McMahon, Keith, “The Content and Terminology of the Mawangdui Texts on the Arts of the Bedchamber,” Early China 17 (1992): 145–85CrossRefGoogle Scholar; Harper, Early Chinese Medical Literature, 359–62.

15. Mawangdui Han mu boshu zhengli xiaozu, Han mu boshu, 1, text 32. For a preliminary study of the text, see Xiang, Ling (Xueqin, Li ), “Shilun Mawangdui Han mu boshu ‘Yi Yin, Jiuzhu’” , Wenwu, 11 (1974), 21–27, 44Google Scholar; for an English translation of the “Nine Rulers” text, see Yates, Robin D. S., Five Lost Classics: Tao, Huang-Lao, and Yin-Yang in Han China (New York: Ballantine Books, 1997), 179–91Google Scholar.

16. Fu Juyou and Chen Songchang, Han mu wenwu, 148–50; Mawangdui Han mu boshu zhengli xiaozu, Han mu boshu, 4, plates 49–52, text 95. Ma Jixing, Mawangdui gu yishu kaoshi, 849–66; for a detailed discussion and translation, see Harper, Early Chinese Medical Literature, 310–27. A comparable explanatory text, called Yinshu , was discovered in tomb No. 247, dated to 186 b.c.e., at Zhangjiashan, Jiangling in Hubei, see hao, Zhangjiashan ersiqi zhujian, Han mu xiaozu, zhengli Zhangjiashan Han mu zhujian (ersiai hao mu) (Beijing: Wenwu, 2001), plates 107–18, text 285–99Google Scholar; Dalun, Gao , Zhangjiashan Han jian Yin, shu yanjiu (Chengdu: Ba Shu, 1995), 34–41 Google Scholar.

17. Fu Juyou and Chen Songchang, Han mu wenwu, 35. The editors oiHanmu wenwu followed Zhou Shirong's naming of this diagram as “Sheshen tu” (Diagram of the God of the Earth), see Shirong, Zhou , “Mawangdui Han mu de ‘shenqi tu’ bohua” , Kaogu 10 (1990), 925–28Google Scholar; Li Ling subsequently renamed it “Bibing tu” , see his “Mawangdui Han mu ‘shenqi tu’ yingshu ‘bibing tu’” , Kaogu 10 (1991), 940–42Google Scholar; on basis of the inscriptions and iconography, Chen Songchang correctly renamed it “Taiyi jiangxing tu”, see his “Mawangdui Han mu bohua “ shenqi tu' bianzheng” , Jiang Han kaogu 1993.1, 88–92; see also Zongyi, Rao, “Tushi yu cifu—Mawangdui xinchu Taiyi chuxing tu sijian” , in Hunan sheng bowuguan sishi zhounian jinian lunwen ji , ed. Hunan, shengbowuguan (Changsha: Hunan jiaoyu, 1996), 79–82 Google Scholar; and Ling, Li, “‘Sanyi’ kao” , in his Zhongguo fangshu xukao (Beijing: Dongfang, 2000), 239–52Google Scholar.

18. Fu Juyou and Chen Songchang, Han mu wenwu, 154–60; the editors named it Tianwen qixiang zazhan (Miscellaneous astronomical and meteorological divinations); the best reconstruction of this manuscript is in Zhongguo wenwu 1 (1979), 1–4, 26–29Google Scholar; cf. Loewe, Michael, “The Han View of Comets,” in his Divination, Mythology and Monarchy in Han China (University of Cambridge Press, 1994), 61–84 Google Scholar; Donald Harper has renamed it Tianwen tu, see Harper, “Communication by Design.”

19. Fu Juyou and Chen Songchang, Han mu wenwu, 162; the editors named it Guaxiang tu (Diagram of the Images of Mantic Figures); Donald Harper has renamed it Fanxin tu, see Harper, “Communication by Design.”

20. Fu Juyou and Chen Songchang, Han mu wenwu, 161.

21. Fu Juyou and Chen Songchang, Han mu wenwu, 132–43, including only Xing de B. For a detailed study of these texts, see Kalinowski, Marc, “The Xingde Texts from Mawangdui,” Early China 23–24 (1998–99), 125–202 Google Scholar. Songchang, Chen reproduced all three (A, B, C) copies of the Xing de texts in his Mawangdui boshu Xingde yanjiu lungao (Taipei: Taiwan guji, 2001)Google Scholar.

22. Fu Juyou and Chen Songchang, Han mu wenwu, 144–45. One of the diagrams has been renamed Shifa (“Methods of Using a Divination Board”); see Han, Mawangdui boshu, mu xiaozu, zhengli, “Mawangdui boshu ‘Shifa’ shiwen zhaiyao” , Wenwu 7 (2000), 85–94 Google Scholar. Some of the fragments have also been indentified as a separate diagram, named Chuxing zhan (Divination on Travel); see Lexian, Liu , Jianbo shushu wenxian tanlun (Wuhan: Hubei jiaoyu, 2003), 115–30Google Scholar. Two fragments were published in Songchang, Chen, ed., Mawangdui boshu yishu (Shanghai: Shanghai shudian, 1996), 130–37Google Scholar.

23. Not yet published. For a brief description, see Songchang, Chen, Boshu shihua (Beijing: Zhongguo dabaike quanshu, 2000), 59–61 Google Scholar.

24. Fu Juyou and Chen Songchang, Han mu wenwu, 35.

25. Yilizhushu (Shisan jing zhushu ed., 1815;rpt.Taipei:Yiwen, 1997), 28.337–34.407.I date the core of the “Sang fu” text to the Warring States period.

26. Hunan sheng bowuguan and Zhongguo kexueyuan kaogu yanjiusuo, “Changsha Mawangdui er san hao Han mu fajue jianbao,” 42; for the lacquer box, see plate 7; Wang Shimin, “Mawangdui boshu,” 307–9; for other manuscripts, especially the medical ones, see Harper, Early Chinese Medical Literature, 14–30.

27. See the discussion of other Warring States and Han canopies on pp. 73–74 below.

28. Scholars have provided different accounts of the red and black squares: Cao Xuequn counts seven red squares on the left and twelve black squares on the right, see Xuequn, Cao, “Mawangdui Han mu ‘Sangfu tu’ jianlun” , Hunan kaogu jikan 6 (1994), 227 Google Scholar; Fu Juyou and Chen Songchang count fourteen black squares and five red squares (Han mu wenwu, 36). Chen Songchang also acknowledges that there are some red pigments in the two squares in the first column on the left, but he thinks that the majority are black. Upon a close examination of the original diagram, I here follow Cao Xuequn's account of the seven red squares. Cao Xuequn also reckons the broken black lines to be the vestiges of another small canopy at the bottom of the diagram, but Chen Songchang thinks that Cao's suggestion on this point is mere speculation (personal communication, April 21,2000). These broken black lines at the bottom of the diagram are also impressed on the upper part of the diagram, but we do not know whether they are part of the original design.

29. This diagram was first published in Shirong, Zhou, “Lüetan Mawangdui chutu de boshu zhujian” , in Mawangdui yishu yanjiu zhuankan 2 (1981), 23–48 Google Scholar; and a color copy was published in Juyou, Fu and Songchang, Chen, Han mu wenwu, 36; cf. Xiaolu, Liu , Zhongguo bohua (Beijing: Zhongguo shudian, 1994), color plate 9Google Scholar.

30. Cao Xuequn, “Mawangdui Han mu ‘Sangfu tu’ jianlun,” 226–29,225. Two other references to the Mawangdui diagram are Shuguo, Chen , Qin Han lizhi yanjiu (Changsha: Hunan jiaoyu, 1993), 178–82Google Scholar and Brashier, Kenneth, “Evoking the Ancestor: The Stele Hymn of the Eastern Han Dynasty (25–220 C.E.),” Ph.D. Dissertation, University of Cambridge, 1997, 37 Google Scholar.

31. We can also use internal structural evidence to collate the manuscript. Scholars have noted the importance of developing a clear procedure and methodology for the study of bamboo slips and silk manuscripts, for example, Dexi, Zhu and Xigui, Qiu , “Qishi niandai chutu de Qin Han jiance he boshu” (first published in 1982, collected in Zhu Dexi guwenzi lunji [Beijing: Zhonghua, 1995], 137–50; especially 141–43)Google Scholar, but a systematic exposition of the methodology is needed.

32. See Boltz, William G., “ I li,” in Early Chinese Texts: A Bibliographical Guide, ed. Michael, Loewe (The Society for the Study of Early China and The Institute of East Asian Studies, University of California, Berkeley, 1993), 237 Google Scholar.

33. I use the term “system” in the sense defined in The Oxford English Collegiate Dictionary (2nd. ed., 1989), “the set of correlated principles, ideas, or statements belonging to same department of knowledge or belief.” In this article I use it in two ways: one refers to a general, abstract mourning system with a small set of basic elements and combinational principles. The other one refers to those concrete subsystems or adaptations, which are the product of the general system. Thus, I do not assume that only one system of mourning and a linear evolution existed in early China.

34. Translations of the terms follow Patricia Ebrey, Buckley, Chu Hsi's Family Rituals (Princeton, New Jersey: Princeton University Press, 1991), Glossary, 213–19Google Scholar, except for dagong and xiaogong. Ebrey translated them as “greater processed cloth” and “lesser processed cloth,” the order of which she mistakenly reversed. There are additional mourning clothes such as suicui in the Yi li. Situated hierarchically between dagong and xiaogong, suicui is a type of special mourning dress, according to the Yi li, only used by officials of local lords (zhuhou zhi chen ) when mourning the Son of Heaven. See Yi li zhushu, 32.380.

35. Zheng Xuan (127–200 c.e.) supported the theory that the three-year mourning extends to twenty-seven months, while Wang Su (195–256 c.e.) insisted on twenty-five months. For a discussion of this issue, see pp. 89–91 below.

36. Mathematically there are thirty possible combinations of the five types of mourning dress and six periods; however, not all the combinations are culturally appropriate. This is because two elements in each set combine in a hierarchical order simultaneously descending from the highest to the lowest. In order to have further distinctions, the combinations vary only in a few cases, notably the second type of mourning dress (zicui).

37. Here “nüzizi” means “daughter,” see Yanwu, Gu , “Nüzizi,” in Rizhi lu jishi (Shijiazhuang: Huashan wenyi, 1990), 269–70Google Scholar.

38. bowuguan, Gansu sheng and Zhongguo kexueyuan ed., Wuwei Han jian (Beijing: Wenwu, 1964), 133 Google Scholar, moben 20, plates 23. Three copies of the “Sang fu” texts were discovered: the two copies (A and B) are its “Transmitted Commentaries” (“Fu zhuan” ) and the other (C) includes only the main text (jing ) and the notes (ji ). Here I quote from Copy C. The version in the Yi li (Yi li zhushu 28.337–34.407) is a mixture of the jing,ji, and zhuan, which are almost identical to those in the Wuwei copies.

39. Steele, John, The I li, or Book of Etiquette and Ceremonies (London: Probsthain and Co., 1917), 2. 9–12Google Scholar, with modifications.

40. Lei, Shi , “ Yi li ‘Sang fu pian’ suo biaoxian de qinshu jiegou” , Zhongyang yanjiuyuan Minzuxue yanjiusuo jikan , 53 (1982), 1 Google Scholar.

41. Such as the discussion in the “Sangfu sizhi” chapter in the Li ji , see Li ji zhengyi (Shisan jing zhushu ed.), 49.1032. Similar discussions also appear in the “Liu de” text on the bamboo strips dated to the Middle Warring States period discovered at Guodian , Hubei, , see Guodian Chu tnu zhujian , ed. Jingzhou shi, bowuguan (Beijing: Wenwu, 1998)Google Scholar, plates 69–73, text 187–90. See also Suying, Lin , Cong Guodian jian tanjiu qi lunchang guannian (Taipei: Wanjuanlou, 2003)Google Scholar.

42. Li ji zhengyi, 34.619. Wang Guowei summarizes the system into four guiding principles: “First, treat relatives as kin; second, respect the respectable; third, honor the elderly; last, distinguish male and female.” See his “Yin-Zhou zhidu lun” , in Guantang jilin (Beijing: Zhonghua, 1959), 462 Google ScholarPubMed.

43. Yi li zhushu, 31.370. For children below the age of eight, the “Fu zhuan” regulates: “Children who have not reached the age of eight full years are spoken of as dying ‘a mourningless death.’ In such a case, one day of wailing is undertaken for every month of the child's age (yi ri yi yue ). So when a child is three months old, the father gives it a name, and when it dies it is wailed for. But if it has not yet received a name it is not wailed for.” See Steele, The I li, or Book of Etiquette and Ceremonies, 2.27–28.

44. It is conceivable that some of the kinship terms may refer to their plural forms. But since in the original text it is not explicitly expressed, and the distinction between the main line and the branches, in the case of sons, is not explicitly made, I translate them all in singular form.

45. For the Li ji passage, see Li ji zhengyi, 58.961–62; for the Xunzi passage, see Xianqian, Wang , Xunzi jijie (Beijing: Zhonghua, rpt. 1992), 372 Google Scholar; for the Gongyang zhuan passage, see Chun qiu Gongyang zhuan zhushu (Shisan jing zhushu, ed.), 9.115. On Han tomb inscriptions, the three-year mourning (i.e., twenty-five months) is often referred as “wuwu zhi yue” , or “wuwu duanren” see Yanwu, Gu, “Sannian zhi sang” , in Rizhi lu jishi, 244–45Google Scholar; see also shūsei, Kandai sekkoku, Honbunhen , , ed. Nagata, Hidemasa (Kyoto: Dohosha, 1994)Google Scholar, “Xianyu Huangbei” (dated to 165 c.e.), 130–36; “Fan Min bei” (dated to 205 c.e.), 268–72.

46. For a concise summary of the universality of kinship classification and its basic structure, see Fox, Robin, “Kinship Categories as National Categories,” in his The Search for Society: Quest for a Biosocial Science and Morality (New Brunswick and London: Rutgers University Press, 1989), 190–204 Google Scholar.

47. There is a large anthropological and historical literature on Chinese kinship; for an overview of the field, see Watson, James L., “Chinese Kinship Reconsidered: Anthropological Perspectives on Historical Research,” China Quarterly 92 (1982), 589–622 CrossRefGoogle Scholar; Freedman, Maurice, Lineage Organization in Southeastern China (London: Athlone Press, 1958)Google Scholar and Chinese Lineage and Society: Fukien and Kwangtung (London: Athlone Press 1966);Google Scholar Ebrey, Patricia B. and Watson, James. L., Kinship Organization in Late Imperial China, 1000–1940 (Berkeley, University of California Press, 1986).Google Scholar For Chinese kinship terminology of the late periods, see Hanji, Feng, The Chinese Kinship System (Cambridge: Harvard University Press for the Harvard-Yenching Institute, 1948)Google Scholar.

48. See, inter alia, Chang, K. C., “The Lineage System of the Shang and Chou and its Political Implications,” in Early Chinese Civilization: Anthropological Perspectives (Cambridge: Harvard University Press, 1976), 72–92 Google Scholar; Chun, Allen J., “Conceptions of Kinship and Kingship in Classical Chou China,” T'oung Pao 76 (1990), 16–48 CrossRefGoogle Scholar; Thatcher, Melvin P, “Marriages of the Ruling Elite in the Spring and Autumn Period,” in Marriage and Inequality in Chinese Society, ed. Rubie S., Watson and Patricia Buckley, Ebrey (Berkeley, Los Angeles, and Oxford: University of California Press, 1991), 25–57CrossRefGoogle Scholar; and Lin, Chao, Marriage, Inheritance and Lineage Organization in Shang-Chou China (Taipei: Yichih Press, 1970)Google Scholar.

49. This is a huge issue with multiple components, about which here I only mention two. For example, Jennifer Holmgren notes the importance of central Asian practices for the imperial families; see her “Imperial Marriage in the native Chinese and Non-Han States, Han to Ming,” in Marriage and Inequality in Chinese Society, 58–96.

50. Anthropological studies of the status of women in late imperial China used to assume that Chinese women stood outside the formal structure of descent groups and that strong patrilineal descent groups correlated with a weak position of women in society. See Watson, James L., “Anthropological Overview: The Development of Chinese Descent Groups,” in Kinship Organization in Late Imperial China, 1000–1940, 282 Google Scholar. But in fact, as the Mawangdui diagram shows, the descent groups in the early Han period were not strictly “patrilineal.” In their new interpretations of women's higher social status in the Song dynasty, Patricia Ebrey and others may have exaggerated the “uniquely” strong place of women's property rights in the Song dynasty. See Patricia B. Ebrey, “Shifts in Marriage Finance from the Sixth to the Thirteenth Century,” in Marriage and Inequality in Chinese Society, 97–132; for a detailed discussion of the debates over women's inheritance rights, see Kathryn Bernhardt, Women and Property in China, 960–1949 (Stanford: Stanford University Press, 1999), especially 9–46; and Beverly J. Bossier, Powerful Relations: Kinship, Status, and the State in Sung China (960–1279) (The Council on East Asian Studies, Harvard University and Harvard University Press, 1998).

51. Shi Lei, “Yi li ‘Sang fu pian’ suo biaoxian de qinshu jiegou,” 23.

52. Shi Lei, “Yi li ‘Sang fu pian’ suo biaoxian de qinshu jiegou,” 24.

53. Er ya zhushu (Shisan jing zhushu ed.), 4.61–64.

54. The kinship classification system in the Er ya is less systematic than that in the Yi li, mainly because the Er ya is an eclectic lexicon. Here I mainly follow the Yi li classification and use the Er ya in explaining kinship terms. For the Yi li classification, see Shi Lei, “Yi li ‘Sang fu pian’ suo biaoxian de qinshu jiegou”; for the Er ya classification, see Yifu, Rui , “ Er ya ‘Shi qin’ buzheng” , collected in his Zhongguo minzu jiqi wenhua lungao (Taipei: Yiwen, 1972), 2. 847–74Google Scholar.

55. Excluding from this list are ego's mother and wives and brother's wives—mother and wives fall into separate categories that are not mentioned in the relatives from the father's side; there is no mourning for brother's wives in the traditional mourning system (shusao wufu .)

56. Cao Xuequn reconstructs only three additional squares. He takes the diagram as a historical record of the family genealogy of the tomb occupant, who died at the age of about thirty and could not possibly have grandchildren or great-grandchildren at that young age. See my argument on p. 63 below.

57. The point of reference on the Mawangdui diagram should be that of an adult male not only because that is the case in the “Sang fu” text, but also because three remaining black squares on the upper section of the central axis suggest that this is a line of male descendants (for the symbolism of black on the Mawangdui diagram, see pp. 74–75 below).

58. See note 56 above.

59. Cao Xuequn, “Mawangdui Han mu ‘Sangfu tu’ jianlun,” 229.

60. Li ji zhengyi (Shisan jing zhushu ed.), 33.591; the translation follows Legge, James, Li Chi: Book of Rites (originally Li Ki, in Sacred Books of the East [Oxford: Oxford University Press, 1885]Google Scholar; rpt. and edited with introduction and study guide by Ch'u Chai and Winberg Chai [New York: University Books, 1967], 2. 42. Legge translates the phrase pangsha as “the collateral branches also were correspondingly less mourned for.” He takes pang as referring to branch lineages. The word pang here refers to the squares spread horizontally at the side of the central axis.

61. Compared to later mourning diagrams, the Mawangdui diagram omits another band of squares for the fifth grade of mourning. I think it is because they are all logical extensions of the fourth grade.

62. Chengshi, Wu , “Zhongguo gudai shehui yanjiuzhe duiyu sangfu ying renshi de jige genben guannian” , first published in 1934, collected in his Wu Chengshi wenlu (Beijing: Beijing shifan daxue, 1984), 11–29Google Scholar, especially 19–20.

63. Ding Linghua , Zhongguo sangfu zhidu shi (Shanghai: Shanghai renmin, 2000), 150–52.

64. Yi li zhusu, 30.386. Only in the case of the maternal grandparents and mother's sisters is the mourning elevated to the fourth grade (xiaogong).

65. Er ya zhushu, 4.63–64.

66. The term sheng was also used as the kinship appellation for the children of mother's brother, brother of one's wife, and the husbands of daughter and sister. See Rui Yifu, “Shi sheng zhi chengwei” and “Shi shengjiu zhi guo” , in Zhongguo minzu jiqi wenhua lungao, 2.937–48 and 991–1012. For the relationship between the kinship term sheng and bilateral cross-cousin marriage in Shang, see Chao, Marriage, Inheritance and Lineage Organization in Shang-Chou China, 29–40.

67. Er ya zhushu, 4.64.

68. Shuowen jiezi (Beijing: Zhonghua, rpt. 1963), 259 Google Scholar.

69. Hinsch, Bret, Women in Early Imperial China (Lanham, Boulder, New York, Oxford: Rowman & Lirtlefield Publishers, Inc. 2002), 10–11,161–66Google Scholar, passim. The rise of patrilinealism reflects on the elevation of the mourning for both great-great-grandparents and great-grandparents to zicui (second grade, but in different durations, see Fig. 3) in the Yi li. And it also reflects on the mourning for a wife, see Jiegang, Gu , “Fu wei qi fu sannian” , in his Shilin zashi chubian (Beijing: Zhonghua, 1963), 104–5Google Scholar.

70. Keightley, David N., “At the Beginning: The Status of Women in Neolithic and Shang China,” Nan Nü 1.1 (1999), 1–63CrossRefGoogle Scholar; Jiayin, Min ed., Yanggang yu yinrou de bianzou: Liangxing guanxi he shehui moshi (Beijing: Zhongguo shehui kexue, 1995)Google Scholar.

71. See O'Hara, Albert R., The Position of Woman in Early China According to the Lieh nü chuan, The Biographies of Eminent Chinese Women (Washington D.C.: Catholic University of America Press, 1945)Google Scholar; Raphals, Lisa, Sharing the Light: Representations of Women and Virtue in Early China (Albany: State University of New York Press, 1998)Google Scholar; Nylan, Michael, “Golden Spindles and Axes: Elite Women in the Achaemenid and Han Empires,” in Early China/Ancient Greece: Thinking through Comparisons, ed. Steven, Shankman and Stephen W., Durrant (Albany: State University of New York Press, 2002), 251–81Google Scholar. All of these works deal with representations of women; although they reflected the attitude toward women rather than their actual social position, we also have other excavated documents to support these views. For a discussion of whether women could inherit wealth, see Hinsch, Bret, “Women, Kinship, and Property as Seen in a Han Dynasty Will,” T'oung Pao 84.1(1998), 1–20CrossRefGoogle Scholar. For women as caretakers and educators, see Takao, Shimomi , Jukyō shakai to bosei: Bosei no iryoku no kanten-de miru Kan Gi Shin Chūgoku joseishi (Tokyo: Kenbun, 1994)Google Scholar. For a detailed discussion of related issues, see Brown, Miranda, “Men in Mourning: Ritual, Politics, and Human Nature in Warring States China, 453 bc-ad 220,” Ph.D. Dissertation, University of California, Berkeley, 2002 Google Scholar. For a criticism of the “social position” approach in women's studies, see Hinsch, , Women in Early Imperial China, 6–7 Google Scholar.

72. See Guangdan, Pan , “Lun Zhongguo fuquan shehui duiyu jiuquan de yizhi” (first published in X in jianshe , 3.5 [1951]Google Scholar, collected in Pan Guangdan wenji [Beijing: Beijing daxue, 2000], 10. 458–63)Google Scholar.

73. Zhemao, Cai , “Yin buci Yi Yin jiushi kao—jianlun tashi” , Zhongyang yanjiuymn Lishi yuyan yanjiusuo jikan 58.4 (1987), 755–808 Google Scholar.

74. Lai, Guolong, “Yiming suojian xingshi zhidu zhi yanjiu” , unpublished M.A. thesis, Department of Archaeology, Beijing University, 1994 Google Scholar; Kryukov, Michael V., “Hsing and Shih: On the Problem of Clan Name and Patronymic in Ancient China,” Archiv Orientâlnî 34 (1966), 535–53Google Scholar; Thatcher, “Marriages of the Ruling Elite in the Spring and Autumn Period,” 25–57; On anthropological literature on the alliance theory of kinship structures, see Claude Levi-Strauss, The Elementary Structures of Kinship (rev. ed. Boston: Beacon Press, 1969; originally published in French, 1949); and Holy, Ladislav, Anthropological Perspectives on Kinship (London and Chicago: Pluto Press, 1996), 124–42Google Scholar.

75. Shigeaki, Ochi , “Kyuzoku to sanzoku” , Kurume daigaku Hikaku bunka kenkyūjo kiyō 13 (1993), 1–22Google Scholar.

76. Runsun, Mou , “Han chu gongzhu ji waiqi zai dishi zhong zhi diwei shishi ,” in his Zhushizhai conggao (Beiiing: Zhonghua, 1987), 50–79 Google Scholar.

77. Goody, Jack, The Oriental, the Ancient and the Primitive: Systems of Marriage and the Family in the Pre-Industrial Societies of Eurasia (Cambridge and New York: Cambridge University Press, 1990), 52 CrossRefGoogle Scholar, passim. In the Confucian classics, there are also several examples of “weak” patrilineality during the Warring States and Han periods. For example, in one case in the “Tan Gong shang” chapter of the Li ji (Li ji zhengyi, 8.146), the mourning for the maternal half-brother and the paternal half-brother was not distinguished. Zheng Xuan in his commentary suggests that the mourning for paternal half-brother should be higher than that for maternal one.

78. Hinsch, “Women, Kinship, and Property as seen in a Han Dynasty Will,” 1-20.

79. Japanese scholars have done extensive researches on the changes in family structure during these periods, see articles in Tatsumi, Makino , Shina kazoku kenkyū (Tokyo: Seikatsu sha, 1940)Google Scholar; Kiyoyoshi, Utsunomiya , Kandai shakai keizai shi kenkyū (Tokyo: Kōbundō, 1967)Google Scholar; Mitsuo, Moriya , Chūgoku kodai no kazoku to kokka (Kyoto: Tōyōshi kenkyūkai, 1968)Google Scholar; and Shigeaki, Ochi, “Kan jidai no ka to kazoku” , Kurume daigaku Hikaku bunka kenkyūjo kiyō 14 (1993), 73–167 Google Scholar.

80. Tatsumi, Makino, “Kyūzoku shōkō” , in Makino Tatsumi chosaku shū (Tokyo: Ochanomizu shobō, 1985), 7.157 Google Scholar.

81. Hinsch, Women in Early Imperial China, 162.

82. Dull, Jack L., “Marriage and Divorce in Han China: A Glimpse at ‘Pre-Confucian’ Society,” in Chinese Family Law and Social Change in Historical and Comparative Perspective, ed. David C., Buxbaum (Seattle: University of Washington Press, 1978), 23–74 Google Scholar.

83. Here is not the place to discuss in detail the longstanding debate over “New Text” and “Old Text” Schools. For discussions of the controversy in English language scholarship, see Nylan, Michael, “The Chin Wen/Ku Wen Controversy in Han Times,” T'oung Pao 80 (1994), 83–144 Google Scholar; “The Ku Wen Documents in Han Times,” T'oung Pao 81 (1995), 25–50 Google Scholar; Ess, Hans van, “The Old Text/New Text Controversy: Has the 20th Century Got it Wrong?” T'oung Pao 80 (1994), 146–70 Google Scholar, and “The Apocryphal Texts of the Han Dynasty and the Old Text/New Text Controversy,” T'oung Pao 85 (1999), 29–64 Google Scholar.

84. Wuwei Han jian, 91–95, 129–32, 133–35.

85. The last phrase in the Yi li reads: “wu zhuan yong zhi dao” . Shen Wenzhuo thinks that the Wuwei text was corrupted; see Wenzhuo, Shen, “ Li Han jian yiwen shi ” , in his Zong-Zhou liyue wenming kaolun (Hangzhou: Hangzhou daxue, 1999), 285 Google Scholar.

86. Wuwei Han jian, 92,130. The same passage appears in the “Sang fu” chapter of the Yi li, see Yi li zhushu, 30.359. Translation with my modifications, see Steele, The I-li or Book of Etiquette and Ceremonial, 2.20–21. See also Raphals, Sharing the Light, 217.

87. The loci classici of the “sancong” are in the “Jiao tesheng” chapter in the Li ji (Li ji zhengyi, 26.506) and in the Yi li (Yi li zhushu, 30.359). See the discussion of “sancong” in Raphals, Sharing the Light, 217–19.

88. One might wonder that since the Mawangdui diagram is probably not intended to present all possible mourning scenarios, special cases such as the “churu” principle may have been omitted from the diagram. That is to say, the practice of the “churu” principle might have already existed. If this were the case, then the Mawangdui chart would not differ much from the Yi li and later diagrams such as the Song dynasty one (see Fig. 6 below). But the arrangement of the seven red squares on the left, especially the inclusion of the children of ego's father's sister, ego's sister, and ego's daughters, leads me to believe that the Mawangdui chart is definitely systematically different from the Song diagrams. I would argue that by these inclusions the Mawangdui diagram illustrates a distinct conception of descent group organization, a conception significantly different from those represented in late imperial diagrams.

89. During the Tang dynasty, the mourning for maternal uncles was elevated to the fourth grade (xiaogong); see the discussions in the Tang court on the issue of mourning for “outside relatives,” Gu Yanwu, “Waiqin zhi fu jie si” , Rizhi lu jishi, 262–65.

90. bowuguan, Hunan sheng, “Xin faxian de Changsha Zhanguo Chu mu bohua” , wenwu 7 (1973), 3–4 Google Scholar, Piate 1.

91. Fu Juyou and Chen Songchang, Han mu wenwu, 19 and 23.

92. Tingwan, Gong , Yu, Gong , and Jialing, Dai ed., Ba Shu Han dai huaxiang ji (Beijing: Wenwu, 1998), nos. 366–67Google Scholar.

93. See Needham, Joseph, Science and Civilisation in China (Cambridge: Cambridge University Press, 1959), 3. 210–16Google Scholar; and Cullen, Christopher, Astronomy and Mathematics in Ancient China: The Zhou, Bi Suan, Jing (Cambridge: Cambridge University Press, 1996), 50–53 CrossRefGoogle Scholar.

94. Guolong Lai, “Color Symbolism in Early Chinese Funerary Art,” a paper presented at the panel “Languages of Color in East Asian Visual Culture” at the College Art Association Ninety-first Annual Conference, New York, February 20, 2003 Google Scholar.

95. Yi li zhushu, 40.475–76.1 am grateful for Professor Donald Harper for this reference.

96. Li Ling, “Mawangdui fangzhongshu yanjiu” , in his Zhongguo fangshu kao, 407.

97. Lloyd, Geoffrey E.R. and Sivin, Nathan, The Way and the Word: Science and Medicine in Early China and Greece (New Haven and London: Yale University Press, 2002), 197 Google Scholar.

98. Li Ling, Zhongguo fangshu kao, 146–59.

99. See, inter alia, Pankenier, David W., “The Cosmo-Political Background of Heaven's Mandate,” Early China 20 (1995), 121–76CrossRefGoogle Scholar.

100. Keightley, David N., The Ancestral Landscape: Time, Space, and Community in Late Shang China (ca. 1200-1045 B.C.) (Berkeley: Institute of East Asian Studies, University of California, 2000), 37–53 Google Scholar. See also Wang, Aihe, Cosmology and Political Culture in Early China (Cambridge University Press, 2000)CrossRefGoogle Scholar, and articles in Zhongguo gudai siwei moshi yu Yin-Yang Wuxingshuo tanyuan , ed. Ai, Lan , Wang, Tao , and Fan, Yuzhou (Nanjing: Jiangsu guji, 1998)Google Scholar.

101. Zhenglang, Zhang , ““Shiyouer gong’ jiqi xiangguan wenti ,” in Jinian Gu Jiegang xueshu lunwenji (Chengdu: Bashu, 1990), 181–200 Google Scholar. The loci classici of the “tian zhi dashu” are in the Zuo zhuan (Chun qiu Zuo zhuan zhengyi [Shisan jing zhushu ed.], 58.1009) and in the Li ji (Li ji zhengyi, 26.499).

102. David S. Nivison, “The Classical Philosophical Writings,” in The Cambridge History of Ancient China: From the Origins of Civilization to 221 BC, 810–12; Harper, “Warring States Natural Philosophy and Occult Thought,” 813–84.

103. Wenjiang, Ding , “Prof. Granet's ‘La civilisation chinoise’,” The Chinese Social and Political Science Review 15.2 (1931), 265–90Google Scholar; Karlgren, Bernhard, “Legends and Cults in Ancient China,” Bulletin of the Museum of Far Eastern Antiquities 18 (1946), 199–365 Google Scholar; Graham, A. C., Disputers of the Tao (LaSalle, Ill.: Open Court, 1989), 320 Google Scholar; and articles in the recent issue of Bulletin of the Museum of Far Eastern Antiquities, 72 (2000)Google Scholar, dedicated to the topic of “correlative cosmology.” For an introduction to Granet's scholarship and bibliography, see K'un, Yang, “Marcel Granet: An Appreciation,” The Yenching Journal of Social Studies 1.2 (1939), 226–41Google Scholar and Freedman, Maurice, “Marcel Granet, 1884–1940, Sociologist,” in Marcel, Granet, The Religion of the Chinese People, translated, edited and with an introduction by Maurice Freedman (Oxford: Basil Blackwell, 1975), 1–29 Google Scholar.

104. Guolong Lai, “Rethinking Karlgren's Free Texts versus Systematizing Texts: Archaeology and the Origins of Ritual Texts in Early China,” a paper presented at the panel on “New Perspectives on Early Chinese Intellectual History” at the Society for the Study of Early China, Fifty-second Annual Meeting of the Association for Asian Studies, San Diego, March 10, 2000; and Luo, Shaodan, “Inadequacy of Karlgren's Linguistic Method as Seen in Rune Svarverud's Study of the Xinshu,” Journal of Chinese Linguistics 31.2 (2003), 270–99Google Scholar.

105. Harper, “Warring States Natural Philosophy and Occult Thought,” 813–84.

106. Porkert, Manfred, The Theoretical Foundations of Chinese Medicine: Systems of Correspondence (Cambridge: MIT Press, 1973)Google Scholar.

107. Granet, Marcel, “La vie et la mort. Croyances et doctrines de l'antiquité chinoise,” L'Annuaire del'École des Hautes Études, section des Sciences Religieuses, 1920–1921 Google Scholar, collected in his Études sociologiques sur la Chine (Paris: Presses Universitaires de France, 1990), 203–20Google Scholar.

108. Marcel Granet, “Le langage de la douleur d'après le rituel funéraire de la Chine classique,” first published in Journal de Psychologie in 1922, collected in his Études sociologiques sur la Chine, 221–42.

109. Granet, “Le langage de la douleur d'après le rituel funéraire de la Chine classique,” 237–38. This passage is translated by Donald Harper.

110. Graham, A. C., Yin-Yang and the Nature of Correlative Thinking (Institute of East Asian Philosophies Monograph Series, No. 6. Singapore: University of Singapore, 1986), 1, 8–15, 91–92Google Scholar; and his Disputers of the Tao, 313–30.

111. Dubs, Homer H. “The Custom of Mourning to the Third Year,” in The History of the Former Han Dynasty (Baltimore: Waverly Press, 1955), 3.40–42 Google Scholar; Yi, Zhao , “Liang Han sangfu wu dingzhi” , in Nian'er shi zhaji (Wang Shumin , Nian'er shi zhaji jiaozheng , [Beijing: Zhonghua, 1984], 68–69)Google Scholar; for three-year mourning in the Han dynasty, see Shuda, Yang , Han dai hunsang lisu kao (Shanghai: Shanghai wenyi, rpt. 1988), 237–68Google Scholar; Miranda Brown, “Men in Mourning: Ritual, Politics, and Human Nature in Warring States China, 453 bc–ad 220.”

112. Hu Shi ascribed to the theory of the Confucian invention of the three-year mourning, but later he changed to support Fu Sinian's theory, see Shi, Hu, “Shuo Ru” , Hu Shi wencun (Taipei: Yuandong, 1953), 4.19–20 Google Scholar; Sinian, Fu , “Zhou dongfeng yu Yin yimin” , reprinted in Hu Shi wencun, 4.82–90 Google Scholar.

113. For previous scholarship on the institution of three-year mourning, see Knapp, Keith N., “The Ru Reinterpretation of Xiao,” Early China, 20 (1995), 209–13CrossRefGoogle Scholar; Ding, Ding , “‘Sannian zhi sang’ yuanliu lunkao” , Shixue jikan 1 (2001), 7–15 Google Scholar.

114. For the two inscriptions, see jinwen, Yin-Zhou shiwen, jicheng , ed. Zhongguo, shehui kexueyuan kaogu, yanjiusuo (Hong Kong: Hong Kong Chinese University, 2001), 5. nos. 9729–30,467–68Google Scholar.

115. The graph is not clear in both inscriptions. Guo Moruo thinks it is “gui” [*k^wrju?], a loan character for “jiu” [*g(r)ju?], meaning “husband's father.” See Moruo, Guo, Liang Zhou jinwenci daxi tulu kaoshi (Beijing: Kexue), 8.212b–14aGoogle Scholar. But this decipherment and reading of the graph is, in my opinion, unreliable. The phonetic reconstructions follow Baxter, William H., A Handbook of Old Chinese Phonology (Berlin and New York: Mouton de Gruyter, 1992)CrossRefGoogle Scholar.

116. See Yirang, Sun , “Qihou hu” , in Guzhou shiyi, Guzhou yulun, (Beijing: Zhonghua, rpt. 1989), 40–41 Google Scholar.

117. Shuda, Yang, Jiweiju jinwen shuo (Beijing: Zhonghua, 1997), 35–37 Google Scholar. Zhongmian, Cen , “'Sannian zhi sangf de wenti” , Liang Zhou wenshi luncong (Shanghai: Shangwu, 1958), 300–12Google Scholar.

118. Nivison, David S., “The Key to the Chronology of the Three Dynasties: The ‘Modern Text’ Bamboo Annals,” Sino-Platonic Papers 93 (1999), 1–68 Google Scholar; “How Old Was the Traditional Chinese Three-Year Mourning?” a paper presented at American Oriental Society meeting, March 12,2000, Portland, Oregon. For the most recent formulation of the “mourning period hypothesis,” see Nivison, David S. and Shaughnessy, Edward L., “The Jin Hou Su Bells Inscription and its Implications for the Chronology of Early China,” Early China 25 (2000), 35–41 CrossRefGoogle Scholar.

119. Lun yu zhushu (Shisan jing zhushu ed.), 17.2526; the translation follows Lau, D. C., Confucius: The Analects (Penguin Books, 1979), 147 Google Scholar, with modifications; my italics.

120. The Brooks date this passage specifically to about 270 b.c.e., see E. Brooks, Bruce and Brooks, A. Taeko, The Original Analects: Sayings of Confucius and His Successors (New York: Columbia University Press, 1998), 165 Google Scholar. I regard it more generally as a Warring States text, probably of late fourth and early third century b.c.e..

121. Granet, “Le langage de la douleur d'après le rituel funéraire de la Chine classique,” 221–42; translation by Donald Harper.

122. The Li ji passage and the Xunzi passage are almost identical. Shen Wenzhuo believes that it is Xunzi who copied the Li ji text, see Shen Wenzhuo, “Lüelun lidian de shixing he Yi li shuben de zhuanzuo” , in his Zong-Zhou liyue wenming kaolun, 49–50.

123. Li ji zhengyi, 58.961–62; Wang Xianqian, Xunzi jijie, 372. Knoblock, John, Xunzi: A Translation and Study of the Complete Works(Stanford: Stanford University Press, 1988–1994)Google Scholar, 3. 70. My translation follows Knoblock's with modifications. My italics.

124. There is a disagreement among scholars as to whether nine or twelve was the “perfect” number in the ancient Chinese ritual systems. Yu Weichao and Gao Ming , “Zhou dai yongding zhidu yanjiu” (collected in Weichao, Yu, Xian Qin liang Han kaoguxue lunji, [Beijing: Wenwu, 1985], 62–114 Google Scholar) follow the Eastern Han scholar He Xiu's theory that in the ritual offering the number of bronze tripods used by the Son of Heaven was nine, the local lords seven, and so forth. But Li Xueqin, in my opinion correctly, points out that archaeological evidence would suggest that many local lords of the Spring and Autumn and Warring States periods used nine tripods in their tombs, thus the Son of Heaven must have used twelve, though we have not yet excavated an intact tomb of that rank, see Xueqin, Li, Eastern Zhou and Qin Civilizations (trans, by K. C., Chang; New Haven: Yale University Press, 1985), 460–63Google Scholar.

125. See also Li Ling, “Cong zhanbu fangfa de shuzihua kan Yin-Yang Wuxing shuo de qiyuan” , in Zhongguo fangshu xukao, 93–94.

126. Lin Qiaoyin , Sanli chenshu qiuyi (30 juan, 1803 ed.).

127. See Granet, “La vie et la mort. Croyances et doctrines de l'antiquité chinoise.”

128. Granet, La pensée chinoise, 127–248.

129. Li ji zhengyi, 32.595.

130. For the conception of “cosmic time,” see Li Ling, Zhongguo fangshu kao, 140.

131. Xu, Liu , “Chunqiu shiqi sangzang zhidu zhong de zangyue zangri” , Kaoguxue yanjiu 2 (1994), 189–200 Google Scholar.

132. See Yang Shuda, Han dai hunsang lisu kao, 237–68; although the three-year mourning was not regulated by the Han government and was not fully carried out in the early Han dynasty, the Mawangdui diagram testifies that, as a system of thought, it was already in existence by 168 b.c.e..

133. Cao Xuequn, “Mawangdui Han mu ‘Sangfu tu’ jianlun,” 227.

134. shu, Jin (Beijing: Zhonghua, ?), 20.621Google Scholar.

135. Although the main text of the Yi li does not explicitly specify how long the three-year mourning is, it assumes that twenty-five or twenty-seven months is the norm (for example, in the “Shiyu li” chapter). See the discussion on the debate between Zheng Xuan and Wang Su on the duration of the three-year mourning below. In addition to the three years, other mourning periods strictly follow the cosmological/numerological scheme.

136. See Huiyan, Zhang , “Qinqin shangsha xiasha pangsha biao” , juan 5, in Yili tu (Yangzhou, ed., 1805)Google Scholar.

137. Mingshan, Ren , Liji mulu hou'an (Jinan: Qilu, 1982), 81–82 Google Scholar; Wenzhuo, Shen, “Han jian ‘Fu zhuan’ kao” , in his Zong-Zhou liyue wenming kaolun, 173 Google Scholar.

138. Shiji (Beijing: Zhonghua, 1982), 10.433–34Google Scholar; Han shu, 4.131–32.

139. Han shu, 4.131–32; Dubs, The History of the Former Han Dynasty, 1. 266–67.

140. The Western Han scholar Liu Xiang (79–8 b.c.e.) is said to have questioned the historicity of the later perception of the Emperor Wen in the late Western Han period, but on specific doubts about the authenticity of this decree is mentioned. See Shao, Ying , “Xiaowen di” , in Wang, Liqi , Fengsu tongyi jiaozhu (Beijing: Zhonghua, 1981), 1. 93–108 Google Scholar.

141. Here Dubs follows Wang Xianqian and others, and translates “wujian” as “[let none of them] wear unhemmed [mourning] garments.” This is in fact incorrect. According to the “Wen sang” chapter in the Li ji, as part of the funerary rites, “[i]mmediately after his parent's death, [the son put off his cap, and] kept his hair with the pin in it, in the bag [of silk]; went barefoot, with the skirt of his dress tucked up under his girdle; and wailed with his hands across his breast.” See Li ji zhengyi, 56.946; and the translation, see Legge, Li Chi: Books of Rites, 2.377. Here I follow the interpretations of commentators Meng Kang and Yan Shigu , and read it as “wuxian” , “not barefoot.”

142. Han shu, 4.132; Dubs, The History of the Former Han Dynasty, 1. 268–71, with modifications.

143. The compilers of the “Treatise on Rites” in the Jin shu comment on Emperor Wen's ritual reform, saying that though Emperor Wen's system is not equal to the three-year mourning, it is close to the classic (gudian ) in spirit. See Jin shu, 20.621.

144. The practice of “taking ‘day’ in placeof ‘month’” (yiriyiyue) appears in the mourning of the death of a minor under the age of eight years, see Yi li zhushu, 31.370.

145. There are certainly different interpretations of Emperor Wen's system. The Tang commentator Yan Shigu argued that it was Emperor Wen's invention, because he understood “the three-year mourning” meant specifically to be twenty–seven months as Zheng Xuan did. For more discussions about “yi ri yi yue” in Emperor Wen's decree, see Xianqian, Wang, Han shu buzhu (Shanghai: Shangwu, 1959), 1.152–53Google Scholar.

146. See Ding Ding, “‘Sannian zhi sang’ yuanliu kaolun,” 14–15.

147. Yi li zhushu, 43.513; also see the “Jian zhuan” chapter in Li ji zhengyi, 57.956.

148. Li ji zhengyi, 42.740.

149. See Kong Yingda's commentary for “Tan Gong shang” in the Li ji (Li ji zhengyi, 6.119).

150. Wang Su's argument is unconvincing, because even if the three-year mourning lasts twenty-five months, it is still possible that in the case of a death occurring at the second half of the last month of the year, twenty-five months would be stretched to the fourth year.

151. For a brief discussion of the understandings of the three-year mourning in Chinese history, see Gu Yanwu, “Sannian zhi sang,” in Rizhi lu jishi, 239–49.

152. Shigeki, Kaizuka , “Kodai ni okeru rekishi kijutsu keitai no hensen” in Chūgoku no shigaku (Kaizuka Shigeki chosaku shū , 7 [Tokyo: Chūō k¯oron sha, 1977]), 201–30Google Scholar; and Schaberg, David, A Patterned Past: Form and Thought in Early Chinese Historiography (Cambridge, Massachusetts and London: Harvard University Asia Center and Harvard University Press, 2001), 316–24CrossRefGoogle Scholar.

153. In his review of Sterckx's, Roel The Animal and the Daemon in Early China (Albany: State University of New York Press 2002)Google Scholar, Mair, Victor H. rightly points out “the emphasis in Sinological studies of the past century and more on all aspects of the culture of early China has been overwhelmingly textual, and—with few exceptions—it is only within recent decades that scholars have begun to pay serious attention to visual materials. The challenge facing scholars of this generation and the next is to take full advantage of both textual sources and visual materials.” See Mair's review in Journal of the American Oriental Society 122.4 (2002), 841–42CrossRefGoogle Scholar.

154. See Goody, Jack, The Domestication of the Savage Mind (Cambridge University Press, 1977), 74–111 Google ScholarPubMed.

155. Harper, Donald, “The Magico-religious Conception of Tu ‘Diagram’ in Early Chinese Thought,” paper presented at the panel “Tu (Diagrams, Charts, Drawings) in Traditional Chinese Culture,” Annual Meeting of the Association for Asian Studies, Washington D.C., March 29, 1998 Google Scholar.

156. Here I assume that a text like the “Sang fu” in the Yi li already existed in the early Han dynasty; see my unpublished manuscript, “The Jing/zhuan/ji structure of the ‘Sang fu': A New Approach to the Dating of Ritual Texts.”

157. Lackner, Michael, “Argumentation par diagrammes: le ximing depuis Zhang Zai jusqu'au Yanjitu,” Extrême Orient-Extrême Occident 14 (1992), 131–68CrossRefGoogle Scholar; “Diagrams as an Architecture by Means of Word: the Yanjitu,” paper presented at the panel “Tu (Diagrams, Charts, Drawings) in Traditional Chinese Culture,” Annual Meeting of the Association for Asian Studies, Washington D.C., March 29, 1998 Google Scholar.

158. François Louis's recent interesting study of the origins of the Taiji tu (Diagram of the Supreme Ultimate) provides an excellent example of the close relationship between decorative symbols and the diagrammatic discourse on cosmology in neo-Confucian writings since the Song, and particularly in Ming and early Qing, dynasty. Although not in the exactly same configuration as the Taiji tu, cosmological diagrams (such as the cord-hook diagram, also known as the TLV design) were often used as decorative motif in early Chinese artifacts. The parallel phenomena deserve further exploration. See Louis, Francois, “The Genesis of an Icon: The Taiji Diagram's Early History,” Harvard Journal of Asiatic Studies, 63.1 (2003), 145–96CrossRefGoogle Scholar. For the early Chinese examples, see Loewe, Michael, “TLV mirrors and their significance,” in his Ways to Paradise: The Chinese Quest for Immortality (Taipei: SMC Publishing Inc. rpt. 1994), 60–85 Google Scholar; Kalinowski, “The Xingde Texts from Mawangdui,” 135–45; Li Ling, Zhongguo fangshu kao, 89–176.

159. Another example of using diagrams to analyze a text is the “Diagram of Nine Rulers” among the Mawangdui manuscripts. The Diagram is drawn on a separate piece of silk, and apparently it was not physically attached to the text of the “Nine Rulers.” It is of a later date than that of the text. The text does not observe the name taboo of the founding emperor of the Han dynasty (i.e., bang ), but the inscriptions on the Diagram observe the name taboo, and consequently changed “pobang zhi zhu” (“rulers who destroy their states”) into “poguo zhi zhu” . Therefore, the “Nine Rulers” text was written (or copied) before or during the reign of the first Han emperor (r. 206–195 b.c.e.) while the Diagram was drawn sometime between 195 b.c.e. and 168 b.c.e.

160. Duanli, Gong , Wufu tujie (Wanweibiezang, ed.; Nanjing: Jiangsu guji, rpt. 1988), 4,33Google Scholar.

161. Sui shu (Beijing: Zhonghua, 1995), 32.921Google Scholar.

162. Yizun, Zhu , Jingyi kao (Beijing: Zhonghua, rpt. 1998), 691–867 Google Scholar; He, Zhou , Lixue gailun (Taipei: Sanmin, 1998), 83–91 Google Scholar.

163. Yongwu, Huang ed., Dunhuang baozang (Taipei: Xinwenfeng, 1981), 129. 407–8Google Scholar. These three diagrams are part of the Shuyi jing (A mirror of etiquette for letter-writing and other occasions), authored by Du Youjin (mid-Tang dynasty), see Heping, Zhao , “Dunhuang xieben shuyi zhong suo kandao de bufen Tang dai shehui wenhua shenghuo” , in Zhou, Yiliang and Zhao, Heping, Tang Wudai shuyi yanjiu (Beijing: Zhongguo shehui kexue, 1995), 302–21Google Scholar. For a preliminary study of shuyi in English language scholarship, see Ebrey, Patricia, “T'ang Guides to Verbal Etiquette,” Harvard Journal of Asiatic Studies 45.2 (1985), 581–613 Google Scholar.

164. Zhou He, Lixue gailun, 86. See Yang Fu , Yi li tu (Tongzhi tang ed., published during the Qing Kangxi reign [1662–1722]), 11.39ab. I shall be examining, from a comprehensive comparative perspective, the three diagrams dating from the Han (Mawangdui), the Tang (Dunhuang) and Song dynasties (Yang Fu), in my future research.

165. Hengzhen, Shao , Yechu ji , in Yingyin Wenyuange Siku quanshu (Taipei: Taiwan Shangwu, 1983), 1215.206Google Scholar.

166. It seems that the early Han practice of inheritance was quite different from the one revealed in the story told above. Furthermore, the Han legal codes discovered at Zhangjiashan (dated to 186 b.c.e.) show that the scope of relatives and their ranking in inheritance are different from the relatives presented on the Mawangdui mourning diagram (in particular, the Zhangjiashan list includes not only daughters but also mother and wife). For the texts of the Han legal codes, see Zhangjiashan Han mu zhujian (ersiqi hao mu), plates 32–39, text 175–79,182–85. See Junming, Li ) “Zhangjiashan Han jian suojian guifan jicheng guanxi de falü” , Zhongguo lishi wenwu 2 (2002), 26–32 Google Scholar.