Skip to main content

“When Red Pigeons Gathered on Tang's House”: A Warring States Period Tale of Shamanic Possession and Building Construction set at the turn of the Xia and Shang Dynasties

  • SARAH ALLAN (a1)

“When Red Pigeons Gathered on Tang's House” (Chi jiu zhi ji Tang zhi wu 赤之集湯之屋) is a Warring States period bamboo manuscript written in the script of the Chu state. It concerns figures that are well known in historical legend: Tang , the founder of the Shang dynasty; his wife; his minister Yi Yin 伊尹, here called by the title xiaochen 小臣 [minor servitor]; and the last king of the Xia dynasty, here called simply the Xia Lord (xia hou 夏后). These figures have their familiar identities, but the tale recorded in the manuscript is unique and has no apparent political or philosophical import. The protagonist, Xiaochen, is Tang's cook, but he does not play the role of founding minister raised up by a future king. Moreover, he is associated with a nexus of motifs associated with shamans, including spirit possession. He acquires clairvoyance after eating a soup of magic red birds (jiu 鳩, [pigeons] or hu 鵠 [cranes]) intended for Tang. After fleeing from an angry Tang, he is possessed by a spirit-medium raven. He then cures the illness of the Xia Lord by having him move his house and kill the yellow snakes and white rabbits under his bed. One rabbit escapes and the story concludes that this is why parapets are placed on houses, suggesting that the context of the story was the construction of a building. Thus, it may have been similar to a historiola, narrated in a ritual to sanctify houses after the placement of the parapet, thus preventing illness among the inhabitants.

Hide All

1 Xueqin, Li 李學勤, (ed.), Qinghua Daxue cang Zhanguo zhujian, vol. 3 清華大學藏戰國竹簡 (Shanghai, 2012), part 1, pp. 22–23, 105–117; part 2, pp. 166–170. Hereafter abbreviated as Qinghua zhujian.

2 Qinghua zhujian, vol. 1, part 1, pp. 2–5, pp. 35–46; part 2, pp. 127–134.

3 Xueqin, Li 李學勤, “Lun Qinghua jian Bao xun de jige wenti” 論清華簡《保訓》的幾個問題, Wenwu (2009), no. 6, p. 76.

4 Jia Lianshang 賈連翔 Qinghua Daxue chutuwenxian yanjiu yu baohu zhongxin 清華大學出土文獻研究與保護中心, “Shi jiezhu shuzi Qinghua Daxue cang Zhanguo zhujian jianbei huahen xianxiang” 試借助數字建模方法分析清華大學藏戰國竹簡簡背劃痕現象, in Chutu wenxian yu Zhongguo gudai wenming guoji xueshu yantaohui 出土文獻與中國古代文明國際學術研討會, Huiyi lunwenji 會議論文集, pp. 356–368. For a study of the lines cut on the reverse sides of the Han Dynasty bamboo manuscripts in the collection of Peking University, see Wei, Han 韓巍, “Xi Han zhushu Laozi jianbei huahen de chubu fenxi” 西漢竹書老子簡背劃痕的初步分析, in Yanjiusuo, Beijing Daxue Chutu Wenxian 北京大學出土文獻研究所, Beijing Daxue cang Xi Han zhushu 北京大學藏西漢竹書 (Shanghai, 2012), pp. 227335.

5 Qinghua Daxue cang Zhanguo zhujian, vol. 1, part 1, pp. 2–5, pp. 35–46, part 2, pp. 127–134.

6 For evidence for the lines on the back that Yin zhi and Yin gao were continuous, see Sun Peiyang 孫沛陽, “Jiance bei huaxian chutan” 簡冊背劃綫初探, Chutu wenxian yu guwenzi yanjiu 出土文獻與古文字研究, no. 4 (December 2011), pp. 449–462. For the three manuscripts, see Xiao Yunxiao 肖蕓曉, “Shilun Qinghua zhushu Yi Yin sanpian de guanlian” 試論清華竹書伊尹三篇的關聯 See also Note 4 above.

7 Ibid. Xiao Yunxiao, “Shilun Qinghua zhushu Yi Yin sanpian de guanlian”.

8 Allan, Sarah, “On Shu 書 [Documents] and the origin of the Shang shu 尚書 [Ancient documents] in light of recently discovered bamboo slip manuscripts”, Bulletin of the School of Oriental and African Studies, LXXV, 3 (October 2012), pp. 547557.

9 Qinghua Daxue cang Zhanguo zhujian, vol. 1, part 2, p. 127.

10 See Allan, Sarah, Buried Ideas: Legends of Abdication and Ideal Government in Early Chinese Bamboo-Slip Manuscripts (Albany, NY, forthcoming), Chapter 2.

11 This graph is made up of 鳥+咎. Jiu 咎 as a word means ‘inauspicious omen’, but I take it here as the phonetic and the graph as a loan for jiu 鳩. The Tsinghua editors suggest hu 鵠. For discussion, see the main text below and Note 49.

12 The Tsinghua editors take zhi 之 as er 而. Following Huang Jie 黃傑, “Chu du Qinghua jian (san) Chi jiu (cong niao) zhi ji Tang zhi wu biji”初讀清華簡 (三)《赤咎(從鳥)之集湯之屋》筆記,” (see[3/15/2013]), I read as zhi 之 and punctuate afterwards.

13 Graph missing at the end of the slip, probably the place name given in Slip 5, graph 3. See Note 15 below.

14 The name Ren Huang is not found in other texts. Wang Ning 王寧, “Du Qinghua jian san Chi hu zhi ji Tang zhi wu sanzha” 讀清華簡三《赤鵠之集湯之屋》散劄, (see reads the graph 巟 as 媓, a title for a royal wife, and takes Ren as her given name.

15 Qinghua zhujian reads this graph as 廷, meaning “palace”. Yang Mengsheng 楊蒙生, “Du Qinghua zhujian Chijiu zhi ji Tang zhi wu biji” 讀清華竹簡《赤鵠之集湯之屋》 筆記, in Chutu wenxian yu Zhongguo gudai wenming guoji xueshu yantaohui, Huiyi lunwenji, pp. 373–374, reads the graph as xing 省, “inspect”, and takes “Xiaochen kui”小臣饋 as its object. However, Huang Jie 黃傑, “Chu du Qinghua jian [san] Chi jiu [cong niao] zhi ji Tang zhi wu biji 初讀清華簡(三)《赤咎(從鳥)之集湯之屋》筆記”[3/15/2013], suggests that it is a place name and the missing graph at the end of Slip 1 should be the same character; this seems the simplest and most direct solution.

16 Following Qinghua zhujian. The reading is uncertain; the original graph is zhou 洀.

17 This graph is partially obscured, but has 示 on left. It should be associated with Xiaochen's sudden need to lay down by the road, so I translate it as “put a spell on”. Yang Mengsheng “Du Qinghua zhujian Chijiu zhi ji Tang zhi wu biji”, argues that the right side may be deciphered as bei 孛. However, this is an unknown character and there is no obvious loan, so it does not identify the intended word.

18 The Chu graph is made up of {疒+未}. Qinghua zhujian reads it as mei 眛, “blind”, but Xiaochen could “see but not speak”, so the reading mei 寐, “sleepy”, which shares the same phonetic, suggested by Huang Jie, “Chu du Qinghua jian [san] Chi jiu (cong niao) zhi ji Tang zhi wu biji,” seems more likely.

19 The word read 撫 is written as {示+蕪}. This reading is uncertain, but the context suggests a rite to ease his pain [chu 楚], as Qinghua zhujian suggests.

20 This graph is probably the name of a heart illness. See Feng Shengjun 馮勝君, “Du Qinghua san Chi hu zhi ji Tang zhi wu zhaji” 讀清華三《赤鵠之集湯之屋》劄記, in Chutu wenxian yu Zhongguo gudai wenming guoji xueshu yantaohui, Huiyi lunwenji, pp. 251–252, who has a similar understanding but takes this graph as a joined character.

21 Slip 9, graph 2, and Slip 13, graph 3 are transcribed as ce 刺 in Qinghua zhujian, which makes sense, though the evidence is weak.

22 疴 is written as . The phonetic is 可. The semantic element, ne 疒, indicates illness. The semantic element, kun, insects  , also appears in many graphs that represent words for illness. Thus, the word should be some kind of disease or skin condition, presumably caused by the animals beneath the bed. Possibly relevant is that people who live with rabbits can contract “rabbit fever” from them which is manifested as skin lesions.

23 See Note 19.

24 The original graph is written as: . I transcribe it as 屯, following Yang Mengsheng, “Du Qinghua zhujian Chijiu zhi ji Tang zhi wu biji”, pp. 375–376. Qinghua zhujian suggests the original graph should be transcribed as lu 鹿 and conjectures that it means 存, “preserve”.

25 The original graph is 埤; following Qinghua zhujian, 170, Note 30. This is discussed further below.

26 Qinghua zhujian reads this graph, which also occurs in Rui Liang fu bi 芮良夫毖, Slip 6, as ding 丁. See discussion in the main text below.

27 See Frankfurter, David, “Narrating power: the theory and practice of the magical historiola in ritual spells”, in Ancient Magic and Ritual Power, Meyer, Marvin and Mirecki, Paul, (eds) (Leiden, 1995), pp. 457476. See also Raz, Gil, The Emergence of Daoism: The Creation of a Tradition (London, 2012), pp. 119120. I am grateful to Gil Raz for bringing this concept to my attention.

28 Sorensen, Jorgen Podemann, “The argument in ancient Egyptian magical formulae”, Acta Orientalia XLV (1984), p. 8, as cited by Frankfurter, p. 466.

29 Chang, K. C., “Some dualistic phenomena in Shang society”, in Early Chinese Civilization: Anthropological Perspectives, Harvard-Yenching Institute Monograph Series, XXIII, (Cambridge, MA, 1976), p. 100.

30 I adopted the term “founding minister” from Granet, Marcel's Danses et légendes de la Chine anciennes (Paris, 1959). See Allan, Sarah, “The identities of Taigong Wang in Zhou and Han literature”, Monumenta Serica XXX (1973), pp. 5799. Because a parallel role is played by Shun and Yu 禹 in the pre-dynastic era and by some of the ministers of the hegemons, I have extended it to apply the meaning to those who act in this role, even when the ruler was not a founding king. See also, Allan, Sarah, The Heir and the Sage (San Francisco, 1981).

31 References to Yi Yin's kitchen work include: Hanfeizi jishi (Taipei, 1974), p. 829 (juan 15.2); p. 222 (juan 4.12); Lüshi chunqiu jiaoshi, p.1225 (juan 18.6) p. 1514 (juan 22.2); Mozi jiaozhu 墨子校註 (Beijing, 1993), p. 97 (juan 2); Shi ji, p. 2153 (juan 63); and Zhuangzi jishi 莊子集釋, p. 814 (juan 8, pian 23). For carrying the ding, see Zhanguo ce zhengjie 戰國策正解 (Taipei, 1976), p. 51 (juan 6 xia); Huainanzi Hongliu jijie 淮南子鴻烈集解 (Beijing, 1989), p. 633 (juan 13), p. 845 (juan 19); and Shi ji, p. 3182 (juan 124). For a more complete list, see Sterckx, Roel, Food Sacrifice and Sagehood in Early China (Cambridge, 2011), p. 67, Note 63.

32 Mozi jiaozhu, p. 77 (juan 2 zhong).

33 Lüshi chunqiu jiaoshi, p. 1093 (juan 17. 5).

34 Shi ji, p. 94 (juan 3).

35 Allan, The Heir and the Sage, pp. 48–49.

36 Sterckx, Food Sacrifice and Sagehood in Early China, see esp. pp. 65–82.

37 See Note 14 above.

38 Guoen 遊國恩, You (ed.), Tian wen zuanyi 天問纂義 (Beijing, 1982), pp. 350351.

39 Lüshi chunqiu jiaoshi, p. 739 (juan 14.2).

40 Allan, The Shape of the Turtle, pp. 43–54.

41 Pearce, David and Lewis-Williams, David, Inside the Neolithic Mind (London, 2005).

42 Allan, Sarah, “Erlitou and the formation of Chinese civilization: toward a new paradigm”, Journal of Asian Studies LVI, 2 (2007), pp. 461496.

43 See Allan, Sarah, “The Taotie Motif on early Chinese ritual bronzes”, in Chinese Zoomorphic Imagination, edited by Jerome Silbergeld and Eugene Wang (Honolulu, in press).

44 Allan, Sarah, “He flies like a bird, he dives like a dragon; who is that man in the tiger mouth? Shamanic images in Shang and early Western Zhou art”, Orientations, XLI, 3 (April, 2010), pp. 4551. See also Childs-Johnson, Elizabeth, “Jade as Confucian ideal, immortal cloak, and medium for the metamorphic fetal pose”, Enduring Art of Jade Age China, vol. 2 (New York, 2002), pp. 1524.

45 “Shangdai de shenhua yu wushu” 商代的神話與巫術, Yanjing Xuebao XX (1936), pp. 485–576.

46 Sukhu, Gopal, The Shaman and the Heresiarch: A New Interpretation of the Li Sao (Albany, NY, 2012), see especially, pp. 75–85.

47 Bowuguan 湖北省荊州博物館, Hubeisheng Jingzhou, Jingzhou Tianxingguan er hao Chu mu 荊州天星觀二號墓 (Beijing, 2003), p. 184, Fig. 154 (M2: 60).

48 Hou Naifeng 侯乃峰, “Chi hu zhi ji Tang zhi wu de ‘chi hu’ huo dang shi ‘chi jiu’” 《赤鵠之集湯之屋》的‘赤鵠’或當是‘赤鳩’,[3/15/2013]. See also Chen Pengyu 陳鵬宇, “Qinghua jian Chi hu zhi ji Tang zhi wu shenhua yuansu shuzheng” 清華簡《赤咎(从鳥)之集湯之屋》神話元素贖證, in Chutu wenxian yu Zhongguo gudai wenming guoji xueshu yantaohui, Huiyi lunwenji, pp. 349–350 who cites later evidence for a tradition of jiu as birds of omen. I would also like to thank Chen Pengyu for bringing the lacquer figure from Tianxingguan illustrated below to my attention.

49 The archaic reconstructions herein are those of Zhengzhang Shangfang 鄭張尚芳 as found on the website They are also published in Shangfang, Zhengzhang, Shanggu yinxi 上古音系 (Shanghai, 2003).

50 Hou Naifeng, “Chi hu zhi ji Tang zhi wu de ‘chi hu’ huo dang shi ‘chi jiu’.

51 See Sterckx, Food, Sacrifice, and Sagehood, p. 74, and note 83. Sterckx points out that baskets labelled “boiled cranes” and crane meat were found in Mawangdui Tombs 1 and 3.

52 Sterckx, Roel, The Animal and the Daemon in Early China (Albany, NY, 2002), p. 174.

53 Lüshi chunqiu jiaoshi, 677.

54 Shi ji 史記, pp. 2785–2786 (juan 105).

55 Particularly interesting is a passage in the Lüshi chunqiu jiaoshi, p. 843 (juan 15.3) that states that Tang personally shot Yi Yin (with a bow and arrow), so the Xia ruler would believe that he had fled to him. Possibly the author was trying to reconcile conflicting accounts about Yi Yin travelling to Xia as a spy and fleeing from Tang.

56 Shi ji, p. 1358 (juan 28).

57 Xingzu, Hong 洪興祖, ed. Chu ci bu zhu 楚辭補註 (Beijing, 1983), pp. 8889. The commentary states that one edition has 兔 for 菟. The Taiping Yulan 太平御覽 (Tainan, Taiwan, 1968), vol. 1, juan 4, 10 (p. 155) also cites this line with the character tu 兔 [rabbit]. Ke, Yuan 袁珂, Zhongguo shenhua tonglun 中國神話通論 (Chengdu, 1991), p. 235, follows Wen Yiduo 聞一多 in taking gu tu 顧菟 as referring to the toad in the moon rather than the rabbit and the toad as a transformation of Chang E 嫦娥. However, the Mawangdui banners include both rabbits and toads and I suspect that several myths about the moon may have been conjoined at the end of the Warring States period or in the early Han.

58 Hunansheng Bowuguan 湖南省博物館 and Zhongguo shehui kexueyuan kaogu yanjiusuo 中國社會科學院考古研究所, Changsha Mawangdui yi hao Han mu 長沙馬王堆一號漢墓 (Beijing, 1973), vol. 2, pl. 77; Jiejun, He 何介鈞, ed., Changsha Mawangdui er, san hao Han mu 長沙馬王堆二、三號漢墓 (Beijing, Wenwu chubanshe, 2004), p. 104 and colour plate 20.

59 Taiping Yulan, juan 4, 11 (p. 156). also cites a line attributed Fu Xuan 傅玄 (Jin dynasty) that is not found in in the current “Tian wen” which refers to the “white rabbit” as pounding medicine, presumably that of immortality stolen by Chang E: “What is there in the moon, that the white rabbit pounds medicine?” (月中何有, 白兔擣藥). Han dynasty stone reliefs also show the rabbit pounding a mortar, but it seems likely that mythical traditions have been combined and it is not clear when this occurred.

Recommend this journal

Email your librarian or administrator to recommend adding this journal to your organisation's collection.

Journal of the Royal Asiatic Society
  • ISSN: 1356-1863
  • EISSN: 1474-0591
  • URL: /core/journals/journal-of-the-royal-asiatic-society
Please enter your name
Please enter a valid email address
Who would you like to send this to? *



Altmetric attention score

Full text views

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

Abstract views

Total abstract views: 0 *
Loading metrics...

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

Usage data cannot currently be displayed