Hostname: page-component-78c5997874-m6dg7 Total loading time: 0 Render date: 2024-11-10T08:07:05.355Z Has data issue: false hasContentIssue false
  • English
  • Français

The Development of Xunzi’s Theory of Xing, Reconstructed on the Basis of a Textual Analysis of Xunzi 23, “Xing E” 性惡 (Xing is Bad)*

Published online by Cambridge University Press:  26 March 2015

Dan Robins
Dan Robins*
Affiliation:
HKU SPACE Community College, Admiralty Centre, Tower 1, 18 Harcourt Road, Hong Kong
Get access Rights & Permissions [Opens in a new window]

Abstract

The section of the Xunzi called “Xing e” 性惡 (xing is bad) prominently and repeatedly claims that people's xing is bad. However, no other text in the Xunzi makes this claim, and it is widely thought that the claim does not express Xunzi's fundamental ideas about human nature. This article addresses the issue in a somewhat indirect way, beginning with a detailed examination of the text of “Xing e”: identifying a core text, removing a series of interpolations, analyzing the structure of the core text, and distinguishing between three positions that are defended there. This analysis shows that the claim that people's xing is bad is not really central to “Xing e.” More ambitiously, it supports the conclusion that Xunzi's ideas about people's xing changed over time. Though Xunzi did claim that people's xing is bad, he later abandoned the claim, and replaced it with an account of wei 偽 “artifice.”

Type
Articles
Information
Early China , Volume 26 , 2001 , pp. 99 - 158
Copyright
Copyright © Society for the Study of Early China 2002

Access options

Get access to the full version of this content by using one of the access options below. (Log in options will check for institutional or personal access. Content may require purchase if you do not have access.)

Footnotes

*

I presented a sketch of some of these arguments, with slightly different conclusions, to the 9th conference of the Warring States Working Group in October 1997. That version should appear in Warring States Papers 2 (forthcoming, 2002). I thank Bruce and Taeko Brooks for their influence on and support for my work. Also, the article has changed substantially in response to various helpful comments from Early China’s referees—I hope for the better. I also thank Donald Harper for some valuable suggestions.

References

1. Scholars who have concluded that the claim does not express Xunzi’s main idea include Wang Xianqian 王先謙, Xunzi jijie 荀子集解 (Xinbian zhuzi jicheng ed. [Beijing: Zhonghua, 1988])Google Scholar, preface; Shi, Hu 胡適, Zhongguo gudai zhexue shi 中國古代哲學史, (Hu Shi zuopin ji 胡適作品集 [Taibei: Yuanliu, 1986], vol.31), 279Google Scholar; Tang Junyi 唐君毅, Zhongguo zhexue yuan lun: yuan xing pian 中國哲學原論: 原性篇 (Tang Junyi Quanji 唐君毅全集 [Taibei: Xuesheng, 1986], vol.13), 66Google Scholar; Cua, Antonio, “The Conceptual Aspect of Hsun-Tzu’s Philosophy of Human Nature,” Philosophy East and West 27 (1977), 374CrossRefGoogle Scholar; Renhou, Cai 蔡仁厚, Rujia xin xing zhi xue lunyao 儒家心性之學論要 (Taibei: Wenjin, 1990), 74Google Scholar; and Yuquan, Tan 譚宇權, Xunzi xueshuo pinglun 荀子學說評論 (Taibei: Wenjin, 1994), 127–30Google Scholar. These discussions sometimes imply that the claim is misleading even in the context of the “Xing e” passages that include it; this point is made more explicit by Munro, Donald J., The Concept of Man in Early China (Stanford: Stanford University Press, 1969), 77–78Google Scholar; and is directly defended by Hansen, Chad, A Daoist Theory of Chinese Thought: A Philosophical Interpretation (Oxford: Oxford University Press, 1992), 336–37Google Scholar, 417n84. Munro and Hansen, along with Eno, Robert, The Confucian Creation of Heaven: Philosophy and the Defense of Ritual Mastery (Albany: State University of New York Press, 1990), 149Google Scholar, agree further that the claim is actually inconsistent with what Xunzi wrote elsewhere; Munro concludes that this part of “Xing e” may be interpolated, Hansen that the claim is interpolated within “Xing e.” Graham, A.C., Disputers of the Tao: Philosophical Argument in Ancient China (La Salle: Open Court Press, 1989), 250–51Google Scholar, further suggests that Xunzi advanced the claim as a slogan primarily to mark his position off from those of Mencius and other early Chinese philosophers. Graham is followed by Van Norden, Bryan W., “Mengzi and Xunzi: Two Views of Human Agency,” in Virtue, Nature, and Moral Agency in the Xunzi, ed. Kline, T.C. III and Ivanhoe, Philip J. (Indianapolis: Hackett Publishing Company, 2000), 127Google Scholar; and, perhaps, by Goldin, Paul Rakita, Rituals of the Way: The Philosophy of Xunzi (Chicago: Open Court, 1999), 13Google Scholar. Notable claims that Xunzi seriously and consistently thought that people’s xing is bad (or even evil) include Dubs, Homer H., “Mencius and Sün-dz on Human Nature,” Philosophy East and West 6 (1956), 216CrossRefGoogle Scholar; Schwartz, Benjamin I., The World of Thought in Ancient China (Cambridge: Harvard University Press, 1985), 291–92Google Scholar; and, at greatest length, Hutton, Eric, “Does Xunzi Have a Consistent Theory of Human Nature?” in Virtue, Nature, and Moral Agency in the Xunzi, ed. Kline, and Ivanhoe, , 220–36Google Scholar.

2. Pian 篇 is often translated as “chapter,” though it actually refers to a bamboo scroll. This translation implies that, like chapters, pian are internally unified, and contribute to a larger textual whole, but this implication is usually false of Warring States pian. I thus transcribe rather than translate the term.

3. A Concordance to the Xunzi (Institute of Chinese Studies Ancient Chinese Texts Concordance Series), ed. Lau, D.C. 劉殿爵 and Ching, Chen Fong 陳方正 (Hong Kong: Commercial Press, 1996)Google Scholar; henceforth, Xunzi. References to the Xunzi will give pian, page, and line numbers within the concordance; references to the notes of the concordance editors will give page and note numbers.

4. Knoblock, John, Xunzi: A Translation and Study of the Complete Works, vol. 3 (Stanford: Stanford University Press, 1994)Google Scholar; henceforth, Knoblock, Xunzi 3.

5. See also Knoblock, , Xunzi 3, 150–51Google Scholar.

6. Xunzi, 113n2, citing Yu Yue 俞樾.

7. 然則 is missing in some editions; see, for example, Disheng, Li李滌生, Xunzi jishi荀子集釋 (Taibei: Xuesheng, 1979), 539–49Google Scholar.

8. See also Knoblock, , Xunzi 3, 151–52Google Scholar.

9. I discuss this difficult term in part 3, “Xunzi’s Conception of Xing and Qing”; within the “Xing e” core the translation “emotions” is adequate but not ideal. I take 情性 to be a compound, and not a modifier-head structure (“emotional nature” or “essential nature”), because of other occurrences within the “Xing e” core where both terms are used as nouns in parallel phrases.

10. See also Knoblock, , Xunzi 3, 152–53Google Scholar.

11. Xunzi, 113n7, on the basis of Lu Wenchao’s 盧文弨 edition.

12. Xunzi, 113n9, citing Liang Qixiong 梁啟雄, on the basis of Yang Liang’s 楊倞 commentary. Though this insertion appears substantive, it does not affect interpretation in any significant way, since whether or not the term e 惡 “bad” is actually used there, the issue in the text is the explanation of badness.

13. This occurrence of the refrain seems to have lost the usual 其善者偽也 “when they are good it is because of artifice,” and may be dislocated. This is one of the few parts of the “Xing e” core that actually talks about artifice; it would be odd indeed if its version of the concluding formula left out the claim about artifice. Also, in its present location it interrupts an argument midway; it would fit better between sections 23.3 and 23.4. I do not incorporate these emendations into my transcription because of my conclusion below that the refrain is interpolated in all occurrences.

14. See also Knoblock, , Xunzi 3, 153Google Scholar.

15. See also Knoblock, , Xunzi 3, 153–54Google Scholar.

16. Xunzi, 114n1, on the basis of context and editions mentioned by Yang Liang.

17. Xunzi, 114n3, on the basis of parallelism and editions cited by Wang Niansun 王念孫.

18. See also Knoblock, , Xunzi 3, 154Google Scholar.

19. Xunzi, 114n5, citing Wang Xianqian 王先謙.

20. See also Knoblock, , Xunzi 3, 154–55Google Scholar.

21. Xunzi, 114n6.

22. See also Knoblock, , Xunzi 3, 155Google Scholar.

23. See also Knoblock, , Xunzi 3, 155–57Google Scholar.

24. For this reading, I follow Liang Qixiong, Xunzi jianshi 荀子簡釋 (Taibei: Daduo, 1983), 332Google ScholarPubMed. Xunzi, 115n2, reads this character as chang 嘗 “let,” following Wang Xian-qian. This also yields an acceptable sense: “Now let us try to do away with….”

25. Xianqian, Wang, Xunzi jijie, 440Google Scholar, reports this suggestion of Yu Yue’s俞樾 (see also Knoblock, , Xunzi 3, 347n24Google Scholar).

26. See also Knoblock, , Xunzi 3, 157–58Google Scholar.

27. I argue that this sentence is likely an interpolation in part 2, “The Three Core Positions.”

28. The Chinese reads: 用此觀之, 然則人之性惡明矣, 其善者偽也。 The refrain occurs in abbreviated versions four times; one of these, at the opening of “Xing e,” simply omits the concluding formula “If we look at it this way, then it is obvious that …” (I give the details of the different versions of the refrain below.)

29. For both emendations see Xunzi, 116n2.

30. Xunzi, 23/116/1–4.

31. See Xunzi, 23/117/16 and 23/117/18, for the two references.

32. Hansen, , A Daoist Theory of Chinese Thought, 336Google Scholar. Hansen fails to notice that even when the refrain is removed, “Xing e” still includes the assertion that people’s xing is bad; this is important below.

33. Xunzi conveniently divides the core text into these nine paragraphs. This division differs from my own in two major respects: my sections 23.5–23.8 correspond to just one paragraph (the fifth), and the refrain distinguishes three paragraphs in my section 23.9. Also, I assume that the occurrence of the refrain in section 23.3 is incomplete and dislocated (see n. 13 above), and that properly located it divides my section 23.3 from my 23.4. Where it is not misleading, I continue to refer to these paragraphs using my own section numbering; otherwise, I identify them as paragraphs one through nine.

34. For example: 今天下之情偽, 未可得而識也 (Now we cannot yet manage to know the world’s real and fake; Mozi 58/36/2) and 道惡乎隱而有真偽 (How does dao hide such that there are genuine and fake?; Zhuangzi 2/4/12). References are to page, pian, and line in Mozi yinde 墨子引得 (Harvard-Yenching Institute Sinological Index Series, supplement no. 21 [Beiping: Yinde bianzuanchu, 1948]); and to pian, page, and line in A Concordance to the Zhuangzi (Zhuangzi zhuzi suoyin 莊子逐字索引; Institute of Chinese Studies Ancient Chinese Texts Concordance Series), ed. D.C. Lau and Chen Fong Ching (Hong Kong: The Commercial Press, 2000).

35. This has been noticed by Fuguan, Xu 徐復觀, Zhongguo renxinglun shi: Xian Qin pian 中國人性論史: 先秦篇 (Taibei: Shangwu, 1969), 238Google Scholar.

36. Daqi, Chen 陳大齊, Mengzi xingshanshuo yu Xunzi xing’eshuo de bijiao yanjiu 孟子性善說與荀子性惡說的比較研究 (Taibei: Zhongyang wenwu gongyingshe, 1953), 32–33Google Scholar.

37. The last line of section 23.10, which opposes following one’s xing, sits more naturally with the claim that people’s xing is bad. I argue below (“The Three Core Positions”) that it is a later addition to this paragraph.

38. Cua, Antonio, “The Quasi-Empirical Aspect of Hsün-Tzu’s Philosophy of Human Nature,” Philosophy East and West 28 (1978), 3–5CrossRefGoogle Scholar.

39. It occurs outside of “Xing e” at Xunzi, 3/11/10, 4/13/2, 32/150/6.

40. The syntax of this construction is difficult to render in English; as a result, in part 1 I (mis)translated 合於 “connect with” as “as well as.”

41. Citations are to page and line number in Xunzi.

42. The parallel here is the structure: 必將待 […] 然後 […]; 然後 on its own occurs also in 23.1.

43. Yet another possibility is that we should read the denunciation of Mencius as a parenthetical aside. (In earlier drafts of this article I claimed that the denunciation was an interpolation; Donald Harper convinced me to be more flexible.)

44. There is no similar motivation for studying “Xing e” as a whole, however, since nothing suggests that it was produced as a text (rather than as a collection of similarly-themed texts).

45. On the assumption that Liu worked on texts in the order they are now recorded in the Han shu 漢書 30 bibliography, E. Bruce Brooks estimates that Liu worked on the Xunzi in 20–19 B.C. See Brooks, “The Collations of Liu Xiang,” Warring States Working Group Note 118 (privately circulated, 1997).

46. Some commentators emend sheng 生 here so that it reads xing 性 on the basis of variants in some editions of the Xunzi. See Li Disheng, Xunzi jishi, 547n7.

47. This possibility was raised by a referee for Early China. See Graham, , Divisions in Early Mohism Reflected in the Core Chapters of the Mo-tzu (Singapore: Institute of East Asian Philosophies, 1985)Google Scholar, as well as his Disputers of the Tao, 35–36.

48. This is how Robert Eno explains early Confucian discussions of tian 天 “nature,” Xunzi’s included. See his The Confucian Creation of Heaven, 154–69.

49. On this point see especially Daqi, Chen, Xunzi xueshuo 荀子學說 (Taibei: Zhongguo wenhua daxue, 1989), 43–45Google Scholar.

50. This is widely thought to differentiate his conception of xing from that of Mencius. See, for example, Daqi, Chen, Mengzi xingshanshuo yu Xunzi xing’eshuo, 6–7, 8–9Google Scholar.

51. This view is very widespread. For a recent statement, see Goldin, , Rituals of the Way, 12Google Scholar.

52. Xunzi, 22/107/22.

53. Or at least so it seems from translations (see, for example, Knoblock, , Xunzi 3, 127Google Scholar). The interpretation I favor has been given by Munro, , The Concept of Man in Early China, 66Google Scholar; and by Graham, , “The Background of the Mencian Theory of Human Nature,” in his Studies in Chinese Philosophy and Philosophical Literature (Singapore: The Institute of East Asian Philosophies, 1986), 15Google Scholar.

54. Sheng 生, used as a transitive verb, seems to mean something more like “produce,” with no implication of naturalness or spontaneity. However, I can find no text in the Xunzi where sheng 生, used as a noun or an intransitive verb, clearly has this more general sense. (The occurrence of sheng at the beginning of section 23.5 is a counterexample to this claim if we treat it as an intransitive verb rather than as a passivized transitive verb.)

55. Xunzi, 22/107/22–23.

56. But see Lüshi chunqiu, 4.3/19/10–12, for an objection to this view. Reference is to section, page, and line number in A Concordance to the Lüshichunqiu (Lüshi chunqiu zhuzi suoyin呂氏春秋逐字索引; Institute of Chinese Studies Ancient Chinese Texts Concordance Series), ed. D.C. Lau and Chen Fong Ching (Hong Kong: Commercial Press, 1994).

57. A famous example is the contrast in Mencius, 11.2, between the behavior due to water’s xing (that is, going down) and what it can be forced to do (rise into the air or flow up a hill) if subjected to external interference. Reference is to section number, and if necessary page and line number, in A Concordance to the Mengzi (Mengzi zhuzi suoyin 孟子逐字索引; Institute of Chinese Studies Ancient Chinese Texts Concordance Series), ed. D.C. Lau and Chen Fong Ching (Hong Kong: Commercial Press, 1995). Another example is a line from the Lüshi chunqiu, 1.2/2/11: “It is the xing of water to be clear; mud disturbs it, so it fails to be clear.” Zhuangzi, 8/22/13–15, offers a particularly striking example when it contrasts the xing of a duck’s legs (which is to be short) with the possibility of lengthening them. These distinctions—in terms of force, disturbance, and violence—all contrast developments that are due to xing with those that are not, and I take it that Xunzi’s references to effort and spontaneity are meant as an interpretation of the same basic distinction. (I claim below that Mencius, 8.26, agrees with Xunzi in marking this distinction in terms of effort.)

58. Xunzi, 22/111/14 and 23/113/17.

59. Xunzi, 22/108/1. I translate bing 病 as “ailment” because it seems to have a broader extension than terms such as “sickness” or “illness.” For example, in the course of the story of the silly farmer from Song, the Mencius uses it to refer to exhaustion (at Mencius, 3.2/16/4).

60. Graham, , “The Background of the Mencian Theory of Human Nature,” 9–14Google Scholar.

61. Lüshi chunqiu, 1.2/2/12–19.

62. Graham, , “The Background of the Mencian Theory of Human Nature,” 14Google Scholar.

63. One might also note that if a person’s xing is her proper course of development, then claims that attribute bad behavior to people’s xing are unintelligible. This includes not only Xunzi’s claim that people’s xing is bad, but also such statements as “It is the xing of lesser people to be moved to valor and to enjoy calamity,” and “It is people’s xing to abuse their superiors.” The statements occur at Zuo zhuan 左傳, Xiang 26.11/291/8, and Guoyu 國語, “Zhou Yu” 周語, 24/15/5, respectively. References are to reign year, paragraph, page, and line in A Concordance to the Chunqiu Zuozhuan (Chunqiu Zuozhuan zhuzi suoyin 春秋左傳逐字索引; Institute of Chinese Studies Ancient Chinese Texts Concordance Series), ed. D.C. Lau and Chen Fong Ching (Hong Kong: Commercial Press, 1995); and to section number, page, and line in A Concordance to the Guoyu (Guoyu zhuzi suoyin 國語逐字索引; Institute of Chinese Studies Ancient Chinese Texts Concordance Series), ed. D.C. Lau and Chen Fong Ching (Hong Kong: Commercial Press, 1999).

64. Graham, , “The Background of the Mencian Theory of Human Nature,” 10Google Scholar.

65. Lüshi chunqiu, 1.2/2/11.

66. Lüshi chunqiu, 15.8/89/8–9.

67. Note that when the Lüshi chunqiu refers specifically to longevity, it uses the compound expression shouchang 壽長; see Lushi chunqiu, 1.3/3/17, 2.3/9/6, 3.3/14/5, 5.4/25/3.

68. Graham recognizes the link between xing and health several times, but seems to derive it from the association with longevity. See his “The Background of the Mencian Theory of Human Nature,” 10, 13.

69. Lushi chunqiu, 3.3/14/5.

70. Zhuangzi, 12/96–98. My translation follows glosses by Li Yi 李頤, Xi Tong 奚侗, and Wang Niansun, cited by Mu, Qian 錢穆, Zhuangzi zuanjian 莊子纂箋 (Taibei: Dongda, 1985), 101Google Scholar.

71. Zuo zhuan, Zhao 25.3/387/6–8.

72. Zuo zhuan, Zhao 8.1/342/5. Graham, , “The Background of the Mencian Theory of Human Nature,” 11Google Scholar, translates xing 性 as “livelihood” here, but offers no justification for the shift in meaning, let alone for his specific translation. See also Zuo zhuan, Zhao 19.9/371/17–18, “the people enjoy their xing.”

73. Xunzi, 22/107/23. This statement immediately follows the two explications of xing from Xunzi 22 discussed on pp. 131–32 above.

74. On this, see Moran, Patrick, “Explorations of Chinese Metaphysical Concepts” (Ph.D. diss., University of Pennsylvania, 1983), 49–51Google Scholar.

75. See Graham, , “The Background of the Mencian Theory of Human Nature,” 59–65Google Scholar; and Hansen, , “Qing (Emotion) 情 in Pre-Buddhist Chinese Thought,” in Emotions in Asian Thought: A Dialogue in Comparative Philosophy, ed. Marks, Joel and Ames, Roger T. (Albany: State University of New York Press, 1995), 181–211Google Scholar. The two accounts disagree about what the single meaning of qing 情 is. Hansen takes qing to be our “reality feed-back”—the impact made by the world on our perceptual and other systems, unconditioned by social or linguistic categories. Thus, when it refers to reality, the idea is that reality is pre-social and pre-linguistic. Graham takes the root meaning of qing 情 to be close to the Aristotelian conception of essence (though pertaining to meaning rather than being), and claims that Xunzi used it to refer to emotions because he believed that emotions are the essence of human beings. Graham’s account is undermined by his inexplicable claim that qing 情 was a key term in Later Mohist philosophy of language; this is simply false (but is picked up by Hansen anyway). Also, Graham’s contention that the use of qing 情 to refer to emotions was original with Xunzi (“The Background of the Mencian Theory of Human Nature,” 64) is undermined by documents found at Guodian 郭店 in 1993. See especially the text that editors have titled Xing zi ming chu 性自命出 in Guodian Chumu zhujian 郭店楚墓竹簡, ed. bowuguan, Jingmen shi (Beijing: Wenwu, 1998)Google Scholar; for example, p. 61, strip 3 (transcribed on p. 179).

76. Graham, , “The Background of the Mencian Theory of Human Nature,” 64Google Scholar.

77. Xunzi, 18/89/11.

78. Xunzi, 22/111/14, where Xunzi states “desires are responses of the qing” (欲者, 情之應也).

79. See Graham, , “The Background of the Mencian Theory of Human Nature,” 63, 65Google Scholar; and Hansen, , “Qing (Emotions) 情 in Pre-Buddhist Chinese Thought,” 199Google Scholar.

80. See also Lushi chunqiu, 2.3/8/21–22:

天生人而使有貪有欲。 欲有情, 情有節。聖人修節以止欲, 故不過行其情也。

Nature produces people and causes them to have needs and desires. With desires, there are qing, and with qing, there is moderation. The sages cultivate moderation and thereby stop desire. Thus, they act only on their qing.

81. I shall not duplicate Hansen’s argument here; see his “Qing (Emotions) 情 in Pre-Buddhist Chinese Thought,” especially 194–202.

82. Some interpreters conclude that Xunzi also has a more religious or metaphysical side. See, for example, Machle, Edward J., Nature and Heaven in the Xunzi: A Study of the Tian Lun (Albany: State University of New York Press, 1993)Google Scholar; Eno, , The Confucian Creation of Heaven, 154–69Google Scholar; and Goldin, , Rituals of the Way, 71–74, 98–105Google Scholar. While it is true that there are a handful of passages that sit poorly with Xunzi’s more characteristic argument strategies, this is insufficient reason to think that Xunzi’s pragmatic arguments do not reflect his most fundamental views. The most we can conclude is that Xunzi was not completely consistent (this is indeed Eno’s conclusion). This should hardly surprise us, given how long Xunzi lived (on most accounts, at least eighty years).

83. Mencius, 11.2.

84. The claim that Mencius and Xunzi meant different things by xing is very common. Scholars who conclude that as a consequence Xunzi’s arguments against Mencius fail include Fuguan, Xu, Zhongguo renxinglun, 237–38Google Scholar; Graham, , Disputers of the Tao, 246, 250Google Scholar; and Siguang, Lao 勞思光, Xinbian Zhongguo zhexue shi 新編中國哲學史, vol. 1 (Taibei: Sanmin, 2001), 319Google Scholar. Of course, even if Mencius and Xunzi meant different things by xing, their disagreement could have been substantive; for suggestions, see Lau, D.C., “Theories of Human Nature in Mencius and Xunzi,” in Virtue, Nature, and Moral Agency in the Xunzi, ed. Kline, and Ivanhoe, , 206–11Google Scholar; Youlan, Feng 馮友蘭, Zhongguo zhexue shi 中國哲學史, (revised ed.; Taibei: Shangwu, 1990), vol.1, 358–59Google Scholar; and Bryan W. Van Norden, “Mengzi and Xunzi.”

85. See, for example, the role of the xin in the series of explications beginning at Xunzi, 22/107/2, the distinction between desiring and seeking at Xunzi, 22/111/4–12, and the statement at Xunzi, 21/104/10, that the xin is the ruler of the body.

86. The ideas of this paragraph are Ivanhoe’s, P.J.; see his “Human Nature and Moral Understanding in the Xunzi,” in Virtue, Nature, and Moral Agency in the Xunzi, ed. Kline, and Ivanhoe, , especially 238–40Google Scholar. David Wong develops Ivanhoe’s interpretation in his “Xunzi on Moral Motivation” in the same volume; I discuss Wong’s conclusions below.

87. I here agree with Wong, “Xunzi on Moral Motivation,” 150, and Hutton, Eric, “Does Xunzi Have a Consistent Theory of Human Nature?” 231Google Scholar.

88. Xunzi, 20/98/15–17.

89. Wong, , “Xunzi on Moral Motivation,” 148–50Google Scholar (the expression “congenial to morality” occurs on p. 150).

90. Xunzi tells this story in section 23.9, as well as at Xunzi, 9/36/1–2, 19/90/3–5, and 20/98/14–19.

91. David Nivison speculates that Xunzi might have agreed that the creation of Confucian institutions took place over generations, somewhat experimentally. See David S. Nivison’s response to Wong’s, David “Xunzi on Moral Motivation,” in Chinese Language, Thought, and Culture: Nivison and His Critics, ed. Ivanhoe, Philip J. (Chicago: Open Court, 1996), 328Google Scholar; and his The Ways of Confucianism, ed. Van Norden, Bryan W. (Chicago: Open Court, 1996), 205Google Scholar. But, as Nivison acknowledges, there is little support for this suggestion in the Xunzi: even if it is what Xunzi should have said, there is little reason to think he actually said it. (And it is not so clear that he should have said it, since it may imply that the sages’ work is not yet finished—an implication that Xunzi would not have accepted.)

92. Mencius, 11.1.

93. Note that in Mencius, 11.2, where Gaozi turns to a water analogy, he avoids the implication that led to this objection. For insightful comments on this shift, see Xi, Zhu 朱熹, Si shu zhangju ji zhu四書章句集注 (Xinbian zhuzi jicheng ed. [Beijing: Zhonghua, 1983]), 325–26Google Scholar; and Shun, Kwong-loi, Mencius and Early Chinese Thought (Stanford: Stanford University Press, 1997), 89Google Scholar.

94. Compare Xunzi, 19/95/1–2: “Without xing then there would be nothing for artifice to be added to; without artifice, then xing could not beautify itself” (無性則偽之無所加, 無偽則性不能自美).

95. Section 23.3 does not explicitly agree that people’s xing produces desires (it talks about perception instead). I nonetheless continue to take the association for granted.

96. Xunzi, 22/107/22–24.

97. Knoblock, , Xunzi 3, 127Google Scholar.

98. For this reason it is perhaps especially significant that Xunzi’s explicit recognition of a relation between xing and health (“The xing is injured—call it ailment”) comes later in this same passage (see above, p. 133).

99. Above, p. 140, I suggested that Mencius, 11.2, might support Xunzi’s interpretation; perhaps Mencius, 11.1, does as well.

100. For references see n. 84 above.

101. Graham, , “The Background of the Mencian Theory of Human Nature,” 28–32Google Scholar.

102. Mencius, 13.30. I adopt the usual assumption that the “it” in question is benevolence and duty. See also Mencius,13.21 and 14.33.

103. See, for example, Lau, D.C., Mencius (London: Penguin, 1970), 188 and 201Google Scholar (“Yao and Shun had it as their nature”); Lau, , Mencius, 185Google Scholar, renders verbal xing as “follows as his nature.”

104. Is this why Mencius, 5.1, tells us that when he talked about the goodness of people’s xing, Mencius always referred to Yao and Shun? Possibly; the only other passage in which he does both is the very similar Mencius, 14.33.

105. This is an appealing way to read the imagery in the story of the silly farmer from Song (Mencius, 2.2) and of Ox Hill (Mencius, 11.8).

106. See, for example, Knoblock, , Xunzi 3, 152Google Scholar.

107. Xunzi, 1/4/13.

108. Mencius, 8.26.

109. See Mencius, 11.7, 11.8, 11.10, and 11.17.

110. This pu 朴 is often interpreted as pu 樸 (same tone), which refers to unprocessed material, especially uncarved wood.

111. See Dao de jing, sections 15, 19, 28, 32, 37, and 57. Reference is to the traditional “Wang Bi” numbering in A Concordance to the Laozi (Laozi zhuzi suoyin 老子逐字索引; Institute of Chinese Studies Ancient Chinese Texts Concordance Series), ed. D.C. Lau and Chen Fong Ching (Hong Kong: Commercial Press, 1996). The Zhuangzi primitivist texts are concentrated in pian 8–10 in that collection.

112. See, for example, the references to Tang Junyi, Cai Renhou, and Tan Yuquan in n. 1.

113. Dubs, , “Mencius and Sün-dz on Human Nature,” 216Google Scholar; cf. Munro, , “A Villain in the Xunzi,” in Chinese Language, Thought, and Culture, ed. Ivanhoe, , 195–99Google Scholar.