“A Living Tradition”

Abstract

This article traces one method of divination, Chinese character divination, from imperial China to contemporary Taiwan. The article proves that many of the techniques used in character divination today originally stem from imperial Chinese sources. It thus shows the continued significance of this method as a living tradition in contemporary Chinese-speaking countries. The article first gives a concise overview of early sources of character divination. It then focuses on three important works from late imperial China, namely the Zichu (Realising a Character’s True Meaning), parts of the Gujin tushu jicheng (Complete Collection of Old and Contemporary Diagrams and Writings), and the Cezi midie (Secret Documents on Fathoming Chinese Characters). In these materials, the treatment of character divination gradually changed from a loose description in short narratives to rigidly defined rules and regulations. Finally, the article juxtaposes these findings with the reality of Chinese character divination in contemporary Taiwan by analysing the results of fieldwork carried out in 2014 and 2015.

本文追溯一種占卜方法——測字從古代中國到當代台灣的歷史。本文闡述當代測字的許多技術最初都源於古代的中國。因此，它表明這種方法在當代華語國家中作為活生生的傳統具有持續意義。首先，本文簡要概述測字的早期來源，然後重點介紹晚期帝國時代的三部重要著作，即《字觸》、《古今圖書集成》的部分内容和《測字秘牒》。從這些著作中可以看到測字的處理逐漸從簡短敘述中的鬆散描述變成了嚴格定義的規則和規定。最後，本文通過分析2014年和2015年進行的實地調查結果，將晚期帝國時代的測字與當代台灣測字進行了並列比較。

Keywords:

• Chinese characters
• popular religion
• divination
• cezi
• Zichu
• Gujin tushu jicheng
• Cezi midie

關鍵詞:

• 漢字
• 民間宗教
• 占卜、測字
• 《字觸》
• 《古今圖書集成》
• 《測字秘牒》

Notes

1 See Needham 1956, p. 364.

2 Bauer 1979, p. 96.

3 The status of divination has been controversial in Mainland China ever since the Kuomintang launched campaigns against superstition in the 1920s and 1930s, a trend that continued with the rise of the People’s Republic of China. See Goossaert 2006, Goossaert – Palmer 2011, and Nedostup 2009. Although practicing divination is still situated in a legal grey zone, it has gained a rapid increase in popularity since the 1980s. It has also become a focus of academic research. Most Chinese scholars, however, focus on analysing the historical dimension of divinatory methods. They tend to avoid studying contemporary practices, because they are reluctant to be stereotyped as fortune-tellers themselves. Research institutes that view divination in an academic context have been receiving greater attention in recent years. See Lackner 2011. Examples for such institutes include the Centre for Zhouyi and Ancient Chinese Philosophy (Yixue yu Zhongguo gudai zhexue yanjiu zhongxin 易學與中國古代哲學研究中心) at Shandong University (Shandong daxue 山東大學) in Jinan 濟南.

4 Elman aptly describes it as “a passionate classificatory agenda for all things, which would yield a cumulative gathering of knowledge,” when he analyses Hu Wenhuan 胡文煥 (fl. approximately 1596) and his series of works Gezhi congshu 格致叢書 (Inquiring into and Extending Knowledge). See Elman 2007, p. 148.

5 The following paragraphs constitute a brief, but not exhaustive introduction to the history of character divination. For an in-depth description of the development of character divination, see Schmiedl 2020.

6 The term chaizi appears in many sources from the Ming and Qing dynasties, but is still sometimes applied today. Cezi is a more recent term and is mainly used in contemporary Chinese. Other terms include the physiognomising of Chinese characters (xiangzi 相字), the separation of Chinese characters (biezi 別字), or the breaking of Chinese characters (pozi 破字).

7 One example for such a regularly cited text is the Zuozhuan 左傳 (Commentary of Zuo [Qiu­ming 丘明], 4th c. bce). In the Zuozhuan, certain passages include the graphic dissection of a character into its components. These components are then used to explain events and situations in the material world, thus demonstrating a close link between written characters and material truths. Three passages from the Zuozhuan are often cited in this regard: 止戈為武 (“to stop” and “spear” yields “military”); 反正為乏 (“to reverse” “correctness” yields “deficiency”); 皿蟲為蠱 (“dish” and “bugs” yields “illness”). For these passages and their contexts, see Zuozhuan, Xuangong 宣公 12.397, Xuangong 15.408, and Zhaogong 昭公 1.709.

8 Wei 緯, “weft” or “horizontal thread in a piece of fabric,” is the counterpart to jing 經, “warp” or “vertical thread in a piece of fabric.” The apocryphal texts (wei) were understood as complementing the orthodox classics of the Confucian canon (jing). The term “apocrypha” has been used in Western studies since the 1950s. See Di Giacinto 2007, p. 1. For a seminal study of the apocrypha, see Dull 1966. For a collection of fragments from the apocrypha, see Yasui Kōzan – Nakamura Shōhachi 1971–1992. For a study of apocryphal omens, see Lippiello 2001, pp. 56–65.

9 See Songshu 27.766. For an analysis of this text, see Lippiello 2000, pp. 43–44.

10 We find mention of one work in the bibliographic chapter (“Jingji zhi” 經籍志) of the Suishu 隋書 (Book of Sui), and of another in the Yanshi jiaxun 顏氏家訓 (Yan’s Family Teachings) by Yan Zhitui 顏之推 (531–591). See Suishu 34.1033. See also Yanshi jiaxun 17.39. Also, a connection to character divination is at least likely in the case of the Pozi jing. The sentence about the Pozi jing in the Yanshi jiaxun refers to other genres and titles, which make this connection quite plausible.

11 Liao Hsien-Huei considers character divination as one of the most popular divinatory methods in the Song. See Liao Hsien-Huei 2005, p. 350 n. 17.

12 Examples can be found in such texts as the Chunzhu jiwen 春渚紀聞 (Hearsay from Spring Islet) by He Wei (or He Yuan) 何薳 (1077–1145), the Beimeng suoyan 北夢瑣言 (Trivial Words from North Meng) by Sun Guangxian 孫光憲 (900–968), and the Touxia lu 投轄錄 (Record for Entertaining Guests) by Wang Mingqing 王明清 (1127 – ca. 1215). See Chunzhu jiwen, Beimeng Suoyan, and Touxia lu, Yuzhao xinzhi.

13 See Yujian: fulu 9.71.

14 Cui Wuyi, also called Cui Zunshi 崔尊師, Song dynasty, was a diviner from Chengdu, Sichuan province. See Yuan Shushan 1998, pp. 647–648.

15 See Beimeng suoyan 12.100–101.

16 A later example is the Songbai leichao 宋稗類鈔 (Categorised Jottings of Petty Matters from the Song Dynasty) by Pan Yongyin 潘永因 (dates unknown). This brush record style work in 36 chapters was probably written around the second half of the seventeenth century. See Kaderas 1998, pp. 254–257. This source also includes short narratives about practitioners of character divination. See Songbai leichao 7.666–667.

17 A short essay on character divination entitled “Cezi” appears in the Gaiyu congkao 陔餘叢考 (Miscellaneous Investigations during the Period of Leisure) by Zhao Yi 趙翼 (1727–1814). See Gaiyu congkao, pp. 733–734. Other works from the Ming and Qing are no longer extant, but their names have been preserved. See, for example, mention of two works on character divination in Siku quanshu zongmu tiyao 11.2830–2868.

18 For an example outside the field of character divination, see Elman 2005, pp. 31–34. Elman traces how Li Shizhen 李時珍 (1518–1593) approached classification and identification in the field of pharmacology in his famous Bencao gangmu 本草綱目 (Materia Medica). Li attempted to rectify the standard for classification set in Song dynasty medical texts and incorporated into his work new information about illnesses from South China.

19 See Elman 2007, pp. 137–139.

20 Elman 2005, p. 58.

21 See, for example, Zichu 1853 and Zichu 1985. All following citations are to the latter edition, hereafter ZC.

22 See Sang Lingzhi 1997.

23 The texts can be found in GJTSJC 747–748.500–521.

24 Other titles include Cezi mijue 測字秘訣 (Secret Tricks of Fathoming Chinese Characters), Chaizi michuan 拆字秘傳 (Secret Transmissions of Dissecting Chinese Characters), and Cezi mizhi 測字秘旨 (Secret Purposes of Fathoming Chinese Characters). Sometimes the variant mi 祕 is used instead of mi 秘 in these designations. Some books use both characters interchangeably: for example, one might be used on the cover of the book and the other in the text itself. Most of these works include similar material, only missing a few techniques or adding new ones. Since the exact date of print of these works is unknown, it cannot be decided conclusively which is the original work and which is the closest to it.

25 In his Zhongguo lidai buren zhuan 中國歷代卜人傳 (Historical Biographies of Chinese Diviners), Yuan Shushan 袁樹珊 (1881 – ca. 1952) includes an entry on Cheng Xing taken directly from the CZMD. It states that he is from Jiangling 江陰 in Jiangsu 江蘇 and that his style name (zi 字) is Yisan 以三. See Yuan Shushan 1998, pp. 189–190.

26 See Bauer 1979, p. 77. Bauer refers to the preface of the Cezi mizhi.

27 This comparison was inspired by Elman’s analysis of Ming dynasty categorised compendia and collectanea: “Rather than microcosms of nature, the late Ming encyclopedias created textual museums for their theater of marvels.” Elman 2007, p. 153.

28 For an overview of the classification of theoretical knowledge on divination and its practical application in imperial China, see Li – Lackner 2017, pp. 456–460.

29 There are, however, a minor number of anecdotes in the ZC that contradict this trend. Two entries in Chapter Four, for example, are styled in a way that allows for their use as templates in divination. See ZC IV–66.69 and IV–67.69–70.

30 See Genette 1997.

31 The second postface of the ZC narrates such an event with Zhou at its centre. See ZC, “Zichu ba” 字觸跋, p. 1.

32 For an in-depth description of these categories, see Schmiedl 2020, pp. 169–215.

33 For Zhou’s “Guide on How to Use the Zichu,” see ZC, “Fanli” 凡例, pp. 1–2.

34 Among others, he mentions the ten heavenly stems and twelve earthly branches tiangan dizhi 天干地支, medicine, pieces of writing, and the hexagrams gua 卦.

35 Yan Yu 嚴羽 (ca. 1195 – ca. 1245) refers to this style as “Su Li ti” 蘇李體 in his Canglang shihua 滄浪詩話 (Canglang’s Remarks on Poetry). This particular style is named after two historical figures, Su Wu 蘇武 (Former Han dynasty) and Li Ling 李陵 (died 74 bce). Both are said to have written poetry and are credited with the invention of the pentasyllabic line. See Yan – Guo 1961, p. 54. Modern scholars have questioned this attribution. On the issue of the origin of the pentasyllabic-line, see Luo Yuming 2011, pp. 127–129.

36 Yan Yu refers to this style as “Qi Liang ti” 齊梁體 in his Canglang shihua, a reference to the Qi 齊 (479–502) and Liang 梁 (502–557) dynasties. See Yan – Guo 1961, p. 48.

37 For an example of character divination in names in this chapter, see ZC I–60.16–17.

38 This example can be found in ZC I–11.3. The beginning of the anecdote refers to the original author He Daoqing 賀道慶 (Northern and Southern Dynasties). Zhou probably took the text from the Yiwen leiju 藝文類聚. See Yiwen leiju 56.1005.

39 晰以剖析為名。不外離合近是。然廋辭多出有意。晰字多從無心。廋以數字輻輳一字。晰以一字分折數字也. ZC, “Fanli,” p. 1.

40 Some anecdotes do, of course, follow Zhou’s categorisation. See, for example, ZC I–18.4 and ZC III–2.45. The second example is originally from the Shishuo xinyu 世說新語 (New Account of Tales of the World). See Shishuo xinyu jiaojian, p. 318. For an English translation of this account, see Mather 1976, pp. 292–293.

41 For an example, see ZC I–9.3.

42 For an example, see ZC III–1.45.

43 Barr describes Xu’s life after the end of the Ming dynasty as one of constant hardship. See Barr 2013.

44 See Tushan ji, pp. 2A–2B.

45 See, for example, GJTSJC 748.519.

46 Bauer 1979, p. 73.

47 See Führer 2006, p. 50.

48 See Bauer 1979, pp. 76–77.

49 凡寫兩字，止看一字，蓋字多必亂。若謀事之類，亦必移時，方可再看. GJTSJC 748.505.

50 For a rare anecdote analysing two characters, see ZC II–72.40.

51 凡人書之，家宅不寧。空字頭，豕應在亥月者也. GJTSJC 748.503.

52 如問病，見土木。及問訟，見血井。字皆凶. GJTSJC 748.505.

53 如立字，見雨下或水聲則成泣字。又如言字，見犬成獄字，問病訟皆忌之. GJTSJC 748.506.

54 For graphology in the GJTSJC, see also Bauer 1979, p. 76. The term “graphology” was first coined in November 1871 by Abbé Jean-Hippolyte Michon in the journal Journal des autographes. See Trubek 2016, chapter 9.

55 For one such example, see ZC II–9.24–25.

56 富人字多穩重，無枯淡。貴人字多淸奇，長畫肥大。貧人字多枯淡，無精神。賤人字多散亂，帶空亡. GJTSJC 748.505.

57 For an example of such later manuals, see He Yueqing 2011.

58 See Bauer 1979, pp. 77–80. Later manuals often also copy material from the other two works introduced here, the ZC and the GJTSJC. As opposed to these two works, however, the CZMD seems to be used specifically because of its systematisation.

59 Some versions lack certain techniques; others include extra techniques. Thus, for example, the technique “Zigu sange jue” 字骨散格訣 (Tricks of Character Structure Dispersing Patterns) is not always included among the techniques. See Xie Shi, Cheng Xing cezi mijue. See also Buyi chaizi michuan bairi tong. The order of techniques is also often different. Generally speaking, however, the core set of techniques stays the same in all these versions.

60 The first five chapters are called “Cezi shi’er fa” 測字十二法 (Twelve Methods of Fathoming Written Characters), “Xinyi liufa” 心易六法 (The Six Methods of the Mind Changes), “Cezi quge dafa” 測字取格大法 (The Great Method of Obtaining Patterns by Fathoming Written Characters), “Shuangju gefa” 雙句格法 (Paired Sentence Patterns Method), and “Zigu sange jue.” In some versions of the CZMD the designations of chapters differ, even if the content of the chapters stays similar. The first chapter is sometimes designated as “Cezi shi fa” 測字十法 (Ten Methods of Fathoming Written Characters), the third chapter as “Danju quge jue” 單句取格訣 (Tricks of Single Sentences Obtaining Patterns) and the fourth as “Lianju quge jue” 鍊句取格訣 (Tricks of Linked Sentences Obtaining Patterns).

61 The sixth and seventh chapter are entitled “Za zhanfu” 雜占賦 (Miscellaneous Divination Rhapsodies) and “Zhili cefa” 至理測法 (The Ultimate Principle Fathoming Method) respectively. See Yang – Song 1993.

62 The following description of techniques and examples mainly relies on two different versions of the CZMD. See the chapter “Cezi shi’er fa” in Xie Shi, Cheng Xing cezi mijue, pp. 66–72. Also see the chapter “Chaifa mijue” 拆法祕訣 (Secret Tricks of Dissection Methods) in Buyi chaizi michuan bairi tong, pp. 67–70 and pp. 72–73.

63 The techniques are ordered in different ways in different versions. Sometimes tianbi and jianbi appear first, sometimes later. See Buyi chaizi michuan bairi tong and Xie Shi, Cheng Xing cezi mijue.

64 This might be the reason why some modern popular publications only refer to these two, but ignore the other techniques. See, for example, He Yueqing 2011.

65 Pojie is, for example, not included in Buyi chaizi michuan bairi tong, possibly because of its similarity to the aforementioned techniques.

66 Contemporary popular reprints of the Meihua yishu often refer to Shao Yong as the original author of the work. See, for example, Meihua yishu.

67 For a manual of Meihua yishu, see Ming chao zhenben meihua yishu.

68 This commentary was probably added later, since it is not present in all versions of the CZMD. For this commentary, see Buyi chaizi michuan bairi tong, p. 96.

69 While the term liushu can already be found in the Zhouli 周禮 (Rites of Zhou), the first work to explain these categories with examples is the Shuowen jiezi 說文解字 (Explaining Written Patterns and Analysing Characters). For the citation in the Zhouli, see Zhouli 14.212. For more information on the six categories of creating characters as found in the Shuowen jiezi, see Winter 1998, pp. 114–116.

70 The English translations of the terms are taken from Bottéro – Harbsmeier 2008.

71 The CZMD uses the denomination xiesheng 諧聲 (indicate harmonious sound), not xingsheng 形聲 (indicate shape and sound). According to the Zhouli commentary by Zheng Xuan 鄭玄 (127–200), this designation can be traced back to Zheng Zhong 鄭眾 (d. 83 CE). See Zhouli 14.212–213.

72 The following description of techniques and examples mainly relies on two different versions of the CZMD. See the chapter “Cheng Yisan xinyi liufa” 程以三心易六法 (Cheng Yisan’s Six Methods of the Mind Changes) in Xie Shi, Cheng Xing cezi mijue, pp. 72–74. Also see the chapter “Chaifa mijue” in Buyi chaizi michuan bairi tong, pp. 70–72.

73 See the chapters “Cezi quge dafa” and “Shuangju gefa” in Xie Shi, Cheng Xing cezi mijue, pp. 74–87. Also see the chapters “Danju quge jue” and “Lianju quge jue” in Buyi chaizi michuan bairi tong, pp. 74–81.

74 似馬不能行動。求官得之少吉. Xie Shi, Cheng Xing cezi mijue, p. 87.

75 See Buyi chaizi michuan bairi tong, pp. 81–84.

76 See Li – Lackner 2017.

77 See Nedostup 2009 and Hammerstrom 2010.

78 Li – Lackner 2017, p. 473.

79 See Jones 2003.

80 See Homola 2013a, pp. 140–141.

81 As Katz shows, local religious traditions enjoy great influence in the political sphere in Taiwan. Thus, several presidential candidates appropriated omens and divination to gain supporters during the election in 2000. See Katz 2003, p. 406.

82 See Homola 2013a, p. 137. Next to these general arguments, some divinatory practices might also have survived because of other specific reasons. Bruun argues, for example, that governmental bans have never managed to eradicate ancestor worship and thus, there was always demand for specialists of fengshui 風水. See Bruun 2008, p. 48.

83 See Li Geng 2018, pp. 75–80.

84 Matthews identifies six different attitudes of Yijing diviners in reference to the relationship between science and other forms of knowledge, such as religion. See Matthews 2017, pp. 180–181.

85 For an overview of academic research on local religious traditions in Taiwan, see Katz 2003, pp. 409–411. For an overview of research on divination in Taiwan, see Homola 2013a, pp. 141–144.

86 See, for example, the works by Chiu Hei-yuan, such as Chiu Hei-yuan 1988 and 1999.

87 See Song – Yang 2003.

88 See Zhou Yingming 1997.

89 See Liu Yu 2010.

90 I have shown elsewhere that authors in Mainland China legitimate these publications on character divination by paying lip service to the state’s ban of divinatory methods, by framing character divination as a part of Chinese traditional culture, by emphasising that it deserves to be studied as part of the long history of writing in China, and by comparing it to Western forms of knowledge. See Schmiedl 2017, pp. 200–206.

91 See Huang – Zhang 2006, p. 11.

92 See, for example, the un-paginated Duzhe zhuyi shixiang 讀者注意事項 (Matters the Reader Should Pay Attention to) in Baiyun jushi 1994.

93 See, for example, the dust cover introduction of Zhao Qinghai 2012.

94 See Luo Zhengxin 1993, pp. 329–330.

95 See Homola 2013b, pp. 35–40.

96 For zhi and de in imperial Chinese texts on character divination, see, for example, ZC V–82.87 as well as Xie Shi, Cheng Xing cezi mijue, p. 83 and ZC III–31.51 as well as Xie Shi, Cheng Xing cezi mijue, p. 82 respectively.

97 Smith 1991, p. 174.

98 Of the nine diviners analysing the character, only 4 and 7 did not split the character into its upper and lower part.

99 For this technique, see Ming chao zhenben meihua yishu 1.18B–20A.

100 There might also be a connection to the popular method of selecting and changing a person’s name (qiming or quming and gaiming), in which strokes of characters are counted.

101 Bauer 1979, p. 74.

Additional information

Notes on contributors

Anne Schmiedl

Anne Schmiedl studied English Cultural Studies, Japanese Studies, and Chinese Studies in Erlangen, Germany, with research stays in Jinan, Taipei, Tokyo, Seoul, and Yamaguchi. She holds a Ph.D. in Chinese Studies from Friedrich-Alexander University Erlangen-Nürnberg (FAU) and is presently a research fellow at the Institute for Chinese Studies at FAU. Her research interests centre on literature, script theories, and the history and development of mantic arts in Chinese-speaking countries. She recently published Chinese Character Manipulation in Literature and Divination: The Zichu by Zhou Lianggong (1612–1672) (Brill 2020).