XIN IN THE XUNZI

Hiu-Chuk Winnie Sung April, 2010

Abstract

This thesis examines the theories of action, decision-making and human in the Xunzi, an ancient Chinese text from the (475 BCE-221BCE). It focuses in particular on Xunzi’s concept (heart/mind). The discussion begins with a challenge to a widely-accepted but often unsupported assumption in the literature that human desires (yu) are sometimes set in tension against (the intentions of) the heart/mind (xin) As a corollary of this view, it is also often suggested that yu is associated with a lack of morality, while xin is the moral capacity. This thesis proposes an original understanding of yu, that is, that yu is causally inert and hence does not stand in opposition to xin, as if yu and xin belonged to two separate human capabilities or areas of function. The discussion then proceeds to investigate the nature of xin. It is proposed that xin has a natural preference for certain objects and may, in that light, initiate action to pursue or even modify the pursuit (qiu) of these objects. On this view, the basis upon which xin deliberates (lu) is not necessarily a moral one. For Xunzi, making the right decision, and doing the right thing, is the result of xin being properly trained to know Dao. On this account, both desirable and undesirable behaviours are attributed to xin. The latter occurs because xin does not fully or appropriately know Dao. An underlying assumption is that once xin knows Dao, it simultaneously understands the importance of abiding by it and will set out to do so. An important implication of this way of understanding Xunzi’s conception of xin is that it also bears on the way we understand Xunzi’s theory of human nature (xing).

Acknowledgements

First and foremost, I would like to express my heartfelt gratitude to Karyn Lai, my principal supervisor and teacher, for all the support, guidance and helpful advice she has generously given me throughout my candidature. I thank Karyn for all the assurance, patience and encouragement she gave me, which allowed me to explore different possibilities. And I thank her for all the knowledge and insights she has unreservedly shared with me. I would like to thank the University of New South Wales for the external research grant it provided, which permitted me to conduct eight months of research in Hong Kong. I thank the Research Centre for and Culture, Chinese University of Hong Kong, for accepting me as a visiting graduate student and for allowing me to take part in the vibrant academic activities of their philosophy department. During this time, I had the opportunity to meet and learn from prominent scholars whose passion in Chinese philosophy has inspired me on both academic and personal levels. I would like to thank Professor Xiaogan Liu for kindly allowing me to audit two of his classes, in which I learned a lot about methodological issues from Professor Liu. I am deeply grateful for all the learning opportunities he gave me. I have benefited from discussions with Professor Roger Ames, Professor Kim-Chong Chong, Professor David Wong, Professor Chung-ying Cheng, and Professor Gu-ying Chen. I thank them for the useful suggestions they gave me in understanding Xunzi’s thought. I am grateful to Siu- fu Tang for sharing with me a bibliography of his dissertation and for allowing me to cite his unpublished work. I also benefited from discussing Xunzi’s thought with Doil Kim, Wai-wai Chiu, Kai-chiu Ng, Guanghua Lin and Leo Chan. My earlier papers related to this thesis were presented at the 2008 World Philosophy Congress, the 2008 Australasian Society for Asian and Comparative Philosophy Conference, and at the postgraduate philosophy seminar at UNSW. I thank the participants for their comments, especially Peter Wong, Jeremy Seligman, Koji Tanaka, Michaelis Michael, and Simon Lumsden. I am grateful to Professor Vincent , from whom I received my first course of instruction in Chinese philosophy at the University of Toronto in 2004. Without

Professor Shen’s help, I would not have been in a position to start my graduate studies in Chinese philosophy. I want to express my deep gratitude to Professor Jeffrey Riegel. I thank Professor Riegel for charitably taking his time to read the Xunzi text with me on a regular basis, for helping me to understand Xunzi, and for alerting me to other relevant aspects of which I was not aware. Professor Riegel’s in-depth and critical comments have shaped my way of reading and understanding the Xunzi. I am grateful to him for sharing with me his insights into early Chinese thought and for allowing me to consult and cite some of his unpublished works. I must acknowledge the very large debt I owe to Professor Kwong-loi Shun. The guidance I received from Professor Shun during my stay in Hong Kong is of enormous value to me and extends well beyond my present study. I learned from Professor Shun the ways to conduct textual analysis and the importance of methodology. His advice has helped to define my research direction and influenced the way I think through issues in Xunzi’s thought and in early Chinese thought generally. I am grateful for the time he generously spent on reading my papers and for the numerous helpful suggestions he gave me. Finally, let me thank all my teachers who played different but equally meaningful roles at each stage of my life. I am very fortunate to have received so much guidance along the way. All the mistakes in this thesis are my own. I thank my friends and family for their support and care. I especially thank my husband Jack for his love and support. Most of all, I thank my parents for their love and encouragement of learning.

Contents

Conventions and Citations i

Background 1

1 A Challenge to Antagonisitc Interpretations 1.1 Overview of Xin and Yu 10 1.2 Secondary Interpretations 18 1.3 Antagonistic Interpretations 29 1.4 Passage 22.5a 32

2 Xin and Yu 2.1 Making Choices 50 2.2 Relation between Xin and Yu 62 2.3 Further Issues 73

3 Orientation of Xin 3.1 Natural Inclinations of Xin 82 3.2 Zhic 89 3.3 Dao 95 3.4 Xin and Dao 113

4 Basis for Xin’s Judgement 4.1 The Issue 119 4.2 Five Interpretations 122 4.3 Two Bases 126 4.4 Proposed Interpretation 132 4.5 A Remaining Difficulty 141

5 Xin and Xinga 5.1 Xunzi on Xinga 144 5.2 Xunzi and 151 5.2.1 Mencius on Xin 156 5.2.2 Differences Between Xunzi’s and Mencius’s Xin 161 5.3 Four Complications 166

Conclusion 173

Appendix: Situating Xunzi’s Xin in Context 176

Character List 203

Bibliography 205

Conventions

The main purpose of the present study is to investigate Xunzi’s thought through the concept of xin ಴.1 Since my focus is on the concept of xin as it is presented in the text, I have set aside the issues concerning the authorship and chronology of Xunzi’s works and will continue to use the personal “Xunzi” in my discussion to refer to the ideas expressed in the text. The nature of this study does not call for a defence of or challenge to Xunzi’s thought, but rather a faithful representation of Xunzi’s view. I will mainly rely on conducting textual analyses of Xunzi’s thought by examining the connotations of terms, the meaning of particular passages, and the relation among passages. Since Xunzi has made it explicit that his project is to rectify many of the terminologies at the time, which he regards as problematic and as having caused people confusion, it is reasonable to assume that Xunzi is especially cautious with his definitions of concepts and that Xunzi could be using the terms differently from what is generally accepted at his time. This makes cross-referencing with other texts more difficult. Hence, I will be more careful when it comes to using materials beyond the Xunzi. Rather than using other texts to explain Xunzi’s view or substantiate my interpretation, I will rely on relevant information from related texts to strengthen my interpretation. I will avoid using Western philosophical notions to discuss Xunzi’s thought for two main reasons. First, this is an attempt to avoid imposing alien frameworks and concepts on Xunzi’s thought. Second, I would also like to avoid generalising Western philosophical concepts. For example, it could be tempting in my analysis of the relation between xin and yu  to say that Xunzi’s conception of xin is different from the Western notion of dualistic body-mind framework. But if I were to make such a claim, I would risk defeating my own purpose in suggesting that Xunzi’s concept of xin is in some way akin to the concept of mind. At the same time, I would also be generalising the entire tradition of the philosophy of the mind under a dualistic body-mind framework, when there is in fact much debate and controversy about the relation between body and mind in

1 It is noted by many scholars that the text Xunzi is not the work of Xunzi alone. See Robert Eno’s detailed discussion of the nature of the Xunzi text (1990, pp.134-8).

i

the Western philosophical tradition as well. This is not to say that a meaningful comparative study between Xunzi’s concept of xin and the Western conception of the mind is not possible. However, this can only be appropriately done if both sides are thoroughly examined. Given the scope of this study, I can only focus my attention on examining Xunzi’s concept of xin and spelling out Xunzi’s conception of xin on its own terms. Although the Xunzi text is relatively coherent and consistent compared with other texts in the same period, it is still fragmented. For this reason, I take on the task of extracting from textual evidence the hidden assumptions and underlying claims that Xunzi would have adopted in order for him to remain consistent. In doing so, I will apply the principle of charity and grant a lot of Xunzi’s claims to be true in the first place. Based on these assumptions and underlying claims that I have extracted, I will at various point propose my own interpretations of Xunzi and adopt these proposed interpretations for my understanding of other issues. The advantage of this method is that it can extract insights from Xunzi’s thought and bring to light the philosophical issues that have concerned Xunzi but are not fully articulated. The disadvantage of this approach is that some of the questions I pose might not even concern Xunzi or that some of the implications I draw from these assumptions are not intended by Xunzi himself. In view of this potential danger, I will be cautious in my moves not to alter too much the original spirit of Xunzi’s thought. In order to avoid framing the readers’ understanding of a particular concept or passage in the Xunzi, I have tried to provide a literal translation of the Chinese passages that keeps the meaning close to the original. A downside of this practice is that my rendering might make the sentences seem grammatically strange in English. I will also provide detailed analyses of key terms for my analysis, such as “yu ,” “ ,” and

“dao ,” and leave them untranslated afterwards. In cases where the translation does not bear on different interpretations, I will adopt Knoblock’s translation, although modifications will be made where appropriate. Even for passages where I have not adopted Knoblock’s translations, I still provide references to Knoblock’s translation in cases where I have consulted Knoblock’s translations. For convenience, I have adopted

ii

common English translations for other keys terms in the Xunzi that, in my view, do not significantly affect my arguments. For example, I have translated “junzi ” as

“superior person,” “ ” as “Heaven,” “shen ” as “body,” and “ ” as “rites/ritual practices.” Readers with a particular interest in these concepts might find my translations of these important terms unsatisfactory. In my translations, I have also tried to use gender-neutral terms such as “people” to discuss Xunzi’s views. It is highly possible, however, that Xunzi only refers to “men” in some instances. Given the scope of this project, it is unavoidable that I have to occasionally adopt common translations for convenience. It should be noted that these translations, if not explained in detail, are only adopted tentatively. In my references to secondary resources, I have converted all Chinese terms in Wade-Giles to .

Citations

Xunzi: All references to the Chinese text of the Xunzi are to the chapter and line numbers

in Xunzi yinde  ಚ. All references to the English translations of the Xunzi are to

chapter and section, volume, and page numbers in John Knoblock’s translation.

Guanzi . All references are to chapter, page, and line numbers in Zhuzi jicheng 

ᬫ.

Hanfeizi . All references are to chapter, page, and line numbers in Zhuzi jicheng

ᬫ.

Hanshi waizhuan द. All references are to the volume and page numbers in the

Sibu chongkan ᣩ੯ৃ edition.

iii

Huainanzi . All references are to the volume, page, and line numbers in Zhuzi

jicheng ᬫ.

Mengzi . All references are to Yang Bojun trans., Mengzi jinzhu jinyi ࡯࡯ ,

with book numbers 1A to 7B substituted for book numbers 1 to 14.

Guoyu !. All references are to the volume, section, and page numbers in Xu Yuangao

comp., Guoyu jijie !ᬫ".

Laozi ᒪ. All references are to chapter numbers in Liu Xiaogan, gujin ᒪੱ࡯.

Liji %. All references are to the volume, page, and line numbers in the Sibu chongkan

ᣩ੯ৃ edition.

Lushi chunqiu &'(). All references are to book and chapter numbers in John

Knoblock and Jeffrey Riegel trans., The Annals of Lu Buwei.

Lun Yu *!. All references are to book and passage numbers in Yang Bo-jun trans.,

Lunyu yizhu *! . Also consulted Lau trans., The .

Mozi +. All references are to chapter, page, and line numbers in Zhuzi jicheng ᬫ

.

iv

Shijing ᎞. All references are to part, book, and ode numbers in James Legge trans.,

The Book of Poetry.

Zhuangzi ᗳ. All references are to chapter and page numbers in Guo -,

Zhuangzi jishi ᗳᬫ.. Also consulted Burton Watson trans., Chuang Tzu.

Yantielun /0*. All references are to the volume, page, and line numbers in the Sibu

chongkan ᣩ੯ৃ edition.

Yijing 1᎞. All references are to the commentaries and the Ta Chuan (TC) sections in

Richard Wilhelm trans., The I Ching.

Yinwenzi 2໨. All references are to the chapter name, page, and line numbers in Zhuzi

jicheng ᬫ.

Zhanguoce 4 5. References are to the volume number in the Sibu chongkan ᣩ੯

ৃ edition.

Zuozhuan ఃद. All references are to the volume and page numbers in Yang Bojun,

Chunqiu zuozhuanzhu ()ఃद .

v Xunzi

Xunzi (ca. 310 BCE- ca. 215BCE), also known by the personal name

7 or the epithets Xun Qing 8, is generally considered one of the three great

Confucian thinkers of the Chinese classical period, along with (ca. 551BCE- ca. 479BCE) and Mencius (ca. 385BCE- 312BCE).1 Xunzi is the author of the text bearing his name, Xunzi. Unlike the Lunyu and the Mencius, which are written in dialogue form, the Xunzi is written in treatise form and reads coherently. Xunzi’s skilful employment of rhetorical techniques and reasoned arguments adds vigour to his writing. Xunzi’s discussions demonstrate his awareness of complex issues during his time. Xunzi would seriously consider and examine the views of other major thinkers of his time before he goes on to rebut or criticise them. This may be why David Nivison describes Xunzi as “the first philosopher in China who could be described as ‘academic’ in the modern sense.”2 In the text, Xunzi argues against the idea of (ca. 480BCE- ca. 390BCE), Zhuangzi (ca. 420BCE- ca. 350BCE), Song Xing (ca. 360BCE- ca. 290BCE), and (ca. 350 BCE). Although Xunzi proclaims himself a follower of Confucius and defends key elements of the Confucian tradition in his work, such as the importance of Confucian rituals and the authority of sage-kings, Xunzi harshly criticises Mencius’s teachings. Xunzi’s disagreement with Mencius adds another layer of complexity to the already intricate system of his thought, rendering the issue of situating him in the Confucian tradition all the more challenging. Little is known for certain about the biography of Xunzi. The few surviving sources that contain information about Xunzi are ’s Shiji ੹% (Records of the

1 Xunzi’s dates are uncertain. Qian Mu assigns a birth date as early as before 339BCE (2001, p.387) while John Knoblock dates Xunzi’s birth between 315-305BCE (1982-83, p.34). This discrepancy in dating is mainly due to inconsistent records about the age Xunzi visited (see note 8). Scholars generally agree that Xunzi lived past 238BCE. It is recorded that Mencius studied with the disciple of Confucius’s grandson, Zisi. Unlike the Zisi-Mencius lineage, Xunzi does not have a readily traceable lineage. Wang Zhong in his Xunqingzi Tonglun links Xunzi to the school of Zixia (Wang Xianqian, pp.21-4). Liang Qichao places Xunzi in the lineage of Zhonggong (1970: 45). Scholars consider Zhonggong as Yong whose style name was Zhonggong. Yang Liang further identifies Zigong as Ran Yong. This will mean Xunzi is a disciple of Zigong. However, both of these views are not widely accepted by scholars today. 2 Nivison, 1999, p.791. Background

Historian),3 Liu Xiang’s Sunqing xinshu xulu :8;ྒྷ=ᧁ (Preface to New Version of the Book of Sun Qing), Ying Shao’s Fengsu Tongyi ?ࣚAᑶ, the Zhanguoze 4 5

(Strategies of the Warring States),4 the Hanfeizi ,5 the Hanshi waizhuan 

द,6 the Yantielun /0* (The Debate on Iron and Salt), and the Xunzi itself.7 What we can gather from these sources is that Xunzi was born in the state of ឌ during

China’s Warring States period (479-221 BCE), which began with the disintegration of the and ended with the unification of the empire under the (221-206BCE).

Xunzi is said to have pursued his studies at the in the state of Qi D,8 where Xunzi was exposed to rigorous intellectual exchanges between thinkers from different persuasions.9 Nonetheless, it is not clear if Xunzi had ever directly debated with any of the major figures at the Jixia Academy, for there are no records of any actual debate between Xunzi and other contemporary Jixia scholars. Masayuki Sato notes that Xunzi was probably too young to have been allowed to exchange his views with other prominent Jixia scholars; therefore, he could only summarise and respond to his predecessors retrospectively in writing.10 One thing for certain is that Xunzi was actively involved in the Jixia Academy. It is also recorded that Xunzi was once the oldest master at the Jixia Academy. This is evidenced by his promotion to the honorary rank of

3 Shiji 74. ¡ ¢ 4 Zhanguoce “Qi juan ” 5 The “Nansan” chapter of the Hanfeizi mentions Xunzi’s visit to King Huai of (38/286/6). 6 Hanshi waizhuan 3/30-1.

7 ¤

Yantielun “Lun Ru £ ” 2/19/7. 8 There are two views regarding the age Xunzi went to Qi. The Shiji and Liu Xiang’s preface record that Xunzi travelled to Qi at the age of “fiftieth” whereas the Fengsutongyi says that Xunzi went to Qi when he was fifteen. The “fiftieth” reading is favoured by most of the Song editions of the text (with the exception

of Chao Gong-wu ¥ ¦ § who reads “fifteenth”). Hu Shi, Ye Shao-jun, and Burton Watson also adopt this ©

view. The Siku ¨ editors agree with the fiftieth reading. Lu Wen-chao, Qian Mu, Wang Shu-min, and Sato argue for the “fifteenth” reading. See Wang Xianqian’s introduction and Sato, 2003, pp.46-53 for a detailed discussion of the issue. 9 There are also scholarly doubts concerning the history of the Jixia Academy and whether Jixia existed as an institutionalised academic organisation. Nathan Sivin, for example, argued that a Jixia Academy never existed. See Sivin, 1995. For a detailed survey of scholars who visited the Jixia Academy, see Jin, 1930. See also Knoblcok vol.1, pp.11-16 for scholars who were active during Xunzi’s days. 10 Sato, 2003, p.52.

2 Background libationer (jijiu EF) at the official ancestral sacrifices.11 Because of his active involvement in the Jixia Academy, there are grounds to believe that Xunzi was well acquainted with the intellectual currents of thought of the time.12 There are records that

Xunzi travelled to the states of Zhao ឌ, Qin G and H and he probably died shortly after the unification of China by Qin.13 Two of Xunzi’s most famous students include the leading figure of the Legalist tradition, Hanfei , and the prime minister of Qin, Lisi

IJ. Other students, albeit less famous, were instrumental in editing and transmitting some of the most influential Confucian classics, including the Guliang Commentary to the Springs and Autumns, the Zuozhuan to the Springs and Autumns, and the Record of Rites of Dai the Elder.

Studies of Xunzi’s Philosophy

Xunzi’s thought continued to be exalted for a while after his death. In the Han dynasty, the imperial librarian Liu Xiang ৖L (ca. 77-6BCE) collected the corpus of Xunzi’s own

11 The libationer is a person who is in charge of the wine sacrifice ritual ceremony. It is not clear as to whether the role of a libationer implies the rank of the head of Jixia Academy. The general agreement is that Xunzi became a senior master at the Academy. Knoblock further suggests that Xunzi was once the head of the restored Jixia Academy (Knoblock vol 1: p.34). 12 Although Shen Buhai, Yang Zhu, Hui Shi and Zhuangzi were not members of the Jixia Academy, there is general evidence that their positions were well known in the Academy. 13 The Yantielun suggests that Xunzi went to Chu in the time of tyrant Min (2.14a), when Jixia scholars

found their advice to no avail to King Min and scattered abroad (“Lunru” chapter). Xunzi later on returned

to Qi but was slandered and left for Chu. The prime minister of Chu, Chun Shenjun , invited Xunzi 

to Chu and appointed him as the Magistrate of Lanling  (Shiji 74). The Xunzi records dialogue

between Xunzi and rulers and statements of Qin, Qi, and Zhao (K 8.2, 16.4, 15.1a, 16.6). Liu Xiang also 

wrote in preface of Sunqingshu that Xunzi met King Zhao of Qin   and discussed military affairs



   with Sun Bing   in front of King Xiao Cheng of Zhao . Hu Shih however argued that Xunzi could not have met Sun Bing because that would make him at least 134 years old when Chun Shenjun died (see also Wang Xianqian’s preface p.32). The “Ruxiao” and “Qiangguo” chapters of the Xunzi also record

Xunzi’s visit to Qin. The Hanfeizi suggests that Xunzi had seen King Hui of Yan  (“Nansan” chapter) around 316BCE. The Yantielun, on the other hand, suggests that Xunzi was active around 220BCE (4.7b). By Liang Qichao’s calculation, Xunzi would have lived up to 161 years old. Besides, the account in the Hanfeizi implies that Xunzi is in the same generation as Mencius. However, Liu Xiang’s preface says Xunzi lived more than a hundred years after Mencius. Besides, there is no record of Xunzi’s involvement during the time of King Xuan (see Knoblock vol 1: 32).

3 Background writings and interpolations by his followers and produced a collection of 32 books called the Sunqing xinshu :8;ྒྷ (The New Version of the Book of Sun Qing). But it was not until the Tang dynasty, in 818 CE, that the first commentary on the Xunzi appeared. Yang

Liang MN14 collated the text, reorganised the 32 books into the sequential order we now have, provided thorough commentaries on Xunzi’s works, and named the text Xunzi. This delay in the appearance of a detailed commentary might have contributed to the decline of Xunzi’s influence. Han Yu ൅ (618-907), an important literary writer of the Tang dynasty, suggests that Mencius is the last transmitter of the Confucian way and that Xunzi’s understanding of the Confucian way is not thorough.15 In the Song dynasty, Xunzi’s status suffered a further, more rapid decline. The prominent neo-Confucians of the Song and Ming Dynasties revered Mencius. Since Xunzi unabashedly attacks

Mencius, studies on Xunzi were marginalised. The Cheng brothers, Cheng Hao PQ

(1032-1085) and Cheng PR (1033-1107), criticised Xunzi’s theory of human nature as being “biased and impure” with regard to the Confucian tradition.16 Sᅠ

(1130-1200), another major Song philosopher, followed the Cheng brothers and held a low opinion of Xunzi. When Zhu Xi compiled the Four Books and excluded the Xunzi, it became obvious that Xunzi’s teaching was marginalised. It is not until Qing dynasty that we saw a revival in the studies of Xunzi. Lu Wen- chao U໨V (1717-1796), Xie Yong WX (1719-1795), and Wang Nian-sun Y೔:

(1744-1832) offered important revised editions incorporating philological exegeses and annotations that are vital to the study of the Xunzi. Two important scholars, Qian Da-xin

[\] (1728-1804) and Hao Yi-xing ^_` (1757-1825), defended Xunzi’s teaching as “pure” and argue that the teachings of Mencius and Xunzi share the same objective. Yu Yue aឃ (1821-1907) also included the Xunzi in his evaluations of

14 Yang Liang’s date was unknown but the preface of his annotated work is dated 818CE. 15 

Han Yu, “Yuan Dao  ”. 16

Zhu Xi, Jin lu ji zhu   .

4 Background various early thinkers. One of the most significant works in this period is the work of

Wang Xian-qian Ycd (1842-1917). In his Xunzi Jieji ᬫ" (A Collection of

Explanations of Xunzi, 1891), Wang Xian-qian collected almost all editions and commentaries of the Xunzi up to his time, made corrections, and provided his own commentaries. Today, Xunzi Jieji serves as one of the essential references for study of

Xunzi. Other important scholars who discussed Xunzi’s theories include Wang Zhong e f (1744-1794), Liu Tai-gong ৖g෼ (1751-1825), Zhang Tai-yan ijk (1868-1935) and Tan Si-tong lସn (1865-1898). Around the same time in Edo Japan, Ogyu Sorai opqࣹ (1666-1728) produced what is considered to be the first Japanese commentary on Xunzi called Doku Junshi (Reading Xunzi, 1763). Although Ogyu Sorai’s understanding of Xunzi is harshly criticised by Tsukada Taiho ॿt\u (1745-1832),

Doku Junshi remains an influential work in Japanese scholarship. Another work of great importance is the Junshi zochu (Additional Annotations of Xunzi, 1820) by Kubo Ai vw x (1759-1832). Kubo provided detailed commentary that took into account old editions and early Japanese studies of Xunzi. Igai Hikohiro yz{| (1761-1845) also made important additional notes. This period is mostly about textual and philological studies rather than philosophical discussions. From the early twentieth century to the present day, there has been a consistent increase in the studies of Xunzi. Zhong Tai }~, Yu Xing-wu ፾, Liang Qi-xiong

‚ƒᬩ, Liu Shi-pei ৖ †, and Yuan Ting-zhuo ‡౲‰ are some of the most important scholars whose comprehensive studies on Xunzi are indispensable to modern translators and interpreters. Li Di-sheng IŠp later compiled a modern Chinese commentary that consults nearly all the abovementioned works. More recently, Chinese scholars such as Mou Zhong-shan ‹Œ࠲, Chen Da-qi ᫶\D, Tang Jun-yi (Tang Chun-

5 Background i) , Lao Si-guang (Lao Ssu-kuang) ‘’“, Zheng-tong ”•A, and Xu Fu- guan –ದ˜ have discussed the thoughts of Xunzi from a more Western philosophical perspective. While we saw an increase in Chinese and Japanese works on Xunzi, English- language studies of Xunzi still cannot be considered prolific. Compared with studies on Confucius and Mencius, there are significantly fewer systematic studies and analyses of the Xunzi in English scholarship.17 James Legge first offered a translation of the “Xing e” chapter of the Xunzi in his Chinese Classics in 1893. However, Legge has not properly included the Xunzi as one of the Chinese classics. Instead, he attached his translation of “Xing e” as an appendix to his translation of the Mencius. This move might have created the impression among Western scholars that the Xunzi is not an important piece of classical Chinese work. In the three decades that follow Legge’s translation, Xunzi’s work had been generally neglected. In 1924, J.J. L. Duyvendak translated and provided scholarly notes of the “Zheng ming” chapter of the Xunzi. In 1927, Homer Dubs published the first English-language book that is devoted entirely to Xunzi.18 A year later, Dubs translated nineteen chapters of the Xunzi.19 Dubs’s works have certainly ignited research interest in Xunzi in the English-speaking world. Since then, a number of scholars have discussed the philosophy of Xunzi.20 However, during the period of 1928 to 1998, there were only two English-language books on Xunzi published.21 Up to the present stage, there are only three substantial English translations of the Xunzi.22 All of

17 For a list of English language studies on Xunzi published in the twentieth century, see Wang Ling-kang, 2003, 26-37. 18 Dubs, Hsuntze: Moulder of Ancient Confucianism. 19 Dubs, The Works of Hsuntze: Translation from the Chinese, with Notes. 20 There are a considerable amount of journal articles and book chapters on Xunzi, including Creel 1931, D.C. Lau 1953, Munro 1969, Schwartz 1985, Yearly 1980, Graham 1989, Nivison 1991, Ivanhoe 1991, Van Norden 1992. 21 Cua 1985 and Machle 1993. 22 The three substantial translations of the Xunzi are: Burton Watson’s Hsun-tzu: Basic Writings (1963), Homer Dubs’s The Works of Hsuntze (1966), and John Knoblock’s Xunzi: A Translation and Study of the Complete Works (1988, 1990, 1994). Among them, only Knoblock has provided a complete translation of the entire text. There are some translations of important sections or chapters. For example, J. J. L Duyvendak and Y. P. Mei have translated the “Zhengming” chapter, Edward Machle the “Tianlun” chapter, Dan Robins the “Xing e” chapter, and Eric Hutton has translated several key chapters. This stands in contrast to the many translations of the Lunyu and the Mencius in English.

6 Background these indicate that, for a long time, Xunzi’s thought has received little systemic and comprehensive treatment in the literature.

Recent Research on Xunzi’s Philosophy

In the past decade, there has been a notable increase in publications on Xunzi.23 These recent studies on Xunzi are carefully conducted and they offer valuable insights into Xunzi’s thought. Nevertheless, the scope of studies in Xunzi is still limited. Many of the earlier scholars were more concerned to use Xunzi’s philosophy, especially his theory of human nature, to extend understanding of Mencius’s philosophy or that of the Confucian tradition. There was little interest in Xunzi as a legitimate thinker himself and hence little work was done to examine his ideas more comprehensively. Xunzi’s thought was usually discussed in relation to or in comparison with that of Mencius. Recent scholars are gradually shifting and broadening the focus of studies in Xunzi on human nature to include Xunzi’s views on social and political order and language. As a corollary, concepts such as li , yi ᑶ, and tong lei ᎈš have begun to receive more scholarly attention. Compared with the commonly-discussed concepts mentioned above, the concept of xin receives notably less attention. The focus of earlier studies in Xunzi is his theory of human nature. When scholars debated whether Xunzi is an orthodox Confucian, their focus is usually on Xunzi’s theory of human nature. As noted previously, Han Yu of the Tang dynasty is probably the first to suggest that Mencius is the last transmitter of the Confucian way. The neo-Confucians unequivocally revered Mencius and scorned Xunzi and it had become official since then that Xunzi was unorthodox. In the late Qing dynasty,

Tan Si-tong lସn (1865-1898), Xia Zeng-you ›œ (1865-1924), and Liang Qi-chao

‚ž᝾ (1873-1927) started an “anti-Xun movement” because they regarded Confucius’s

23 Paul Goldin (1999), Masayuki Sato (2003), Janghee Lee (2005), and Kurtis Hagen (2007) have published monographs on Xunzi. There were two published collections of essays on Xunzi: P.J. Ivanhoe and T. C. Kline edited Virtue, Nature, and Moral Agency in the Xunzi (2000) and A. S. Cua, Human Nature, Ritual, and History (2005). While Aaron Stalnaker’s Overcoming Our Evil (2006) is a comparative study of Xunzi and Augustine, a substantial part of it contains detailed examination of Xunzi’s thought.

7 Background teachings as tainted by Xunzi.24 It is interesting to note that in spite of the low opinion neo-Confucians have of Xunzi, Tan, Xia, and Liang actually believed that Xunzi’s philosophy had dominated the entire intellectual history of China up to their time. Liang, for example, maintains that

…the classical masters of the Han Dynasty, whether of the Modern or Ancient Text Schools, had all evolved from [Xunzi]; therefore, over the past two thousand years, even if the schools had often changed, all of them had changed uniformly within the framework of the [Xunzi] school. The school of Mencius was discontinued and hence that of Confucius also declined.25

Around the same period in the West, Dubs calls Xunzi “the moulder of ancient Confucianism” and puts forward the controversial argument that Xunzi’s teachings continue orthodox Confucian tradition.26 According to Dubs, Xunzi’s teaching had been regarded as orthodox for more than a millennium before the Song Dynasty and even Zhu Xi himself was following the teaching of Xunzi. Unfortunately, Dubs has provided little evidence to substantiate these claims.27 Herrlee Creel, on the other hand, sees Xunzi as “the destroyer of the principles of Confucius.”28 For Creel, Mencius’s theory of human nature was in agreement with the philosophy of Confucius on all levels. Since Xunzi’s philosophy is in opposition to that of Mencius, Creel argues, Xunzi has “perverted” the teaching of Confucius.29 Creel further disagrees with Dubs in saying that Xunzi’s influence only started to decline in Zhu Xi’s time. Creel points out before the Song dynasty, there were notably more commentaries on the Mencius than on the Xunzi, with a stark contrast of nine to one. Besides, when both Mencius and Xunzi were admitted to the temple of Confucius in 1084, Mencius’s image was placed just below the immediate whereas Xunzi’s image was placed among the twenty-two

24 In a poem Xia wrote for Liang, Xia describes Xunzi as a “demon” that blots out the sun (Liang, 1959, p.100). 25 Liang, 1959, p.99. Wang Zhong of the Qing dynasty notes in his Xunqingzi Tonglun that important Han scholars were influenced by Xunzi. See Wang Xian-qian’s Xunzi Jijie 2007: 22. 26 See Dubs 1927 and 1930. 27 Dubs, 1930, p. 235. 28 Creel, 1929, p.75. 29 Creel, 1953, p.134.

8 Background worthies. While Mencius was given the rank of gong (duke), Xunzi was given the two- level lower rank of bo (earl).30 All of these suggest that Xunzi was not held in high esteem before the Song dynasty and Mencius was the key figure that eclipsed Xunzi’s influence. This scholarly avoidance of Xunzi’s theory of human nature will inevitably have an impact on Xunzi’s theory of xin because the concept of xin is intimately linked to the concept of xinga and is a core concept in pre-Qin theories of human nature. Confucian scholars favour Mencius’s concept of xin and its implication that moral cultivation is about nurturing and expanding xin because its emphasis is on the internal rather than external. They find it difficult to accept Xunzi ‘s concept of xin because it relies on external means to correct one’s problematic nature. This seems to go against the tradition that emphasises the cultivation of certain internal ethical attributes in forming a moral character, rather than relying on external means to enforce morality.31 What I propose to do in the present study is to offer a new angle. The present study will focus on analysing Xunzi’s concept of xin in the Xunzi and take the concept of xin as the perspective from which I read Xunzi. I will analyse Xunzi’s concept of xin and and highlights a range of issues such as Xunzi’s view on agency, decision making, and human nature.

30 Creel, 1931. 31 For example, Liang Qi-chao in Chinese Political Thought criticised Xunzi’s philosophy for being too “mechanistic,” which does not accord with the ethical standard of Confucianism but more with the legalist tradition. See 1968, Chapter 4.

9 CHAPTER 1

A Challenge to Antagonistic Interpretations

1.1 Overview of xin and yu

The goal of this chapter is to examine the relation between the concepts xin ಴ and yu

 in the Xunzi. In order to prepare the ground for my own interpretation, the present chapter focuses on unearthing and discarding some predominant assumptions about xin and yu that could do injustice to Xunzi’s view. This chapter is divided into four sections. The first section is an overview of Xunzi’s descriptions of xin and yu. The second section discusses secondary interpretations, followed by the third section that brings to light a characteristic, but often overlooked, understanding of the subject matter in the literature, which I call the antagonistic interpretations. On the antagonistic interpretations, both xin and yu have causal roles in action. In the fourth section, I will primarily rely on Passage 22.5a of the Xunzi as textual evidence to challenge the assumptions behind antagonistic interpretations. In what follows, I will provide a general overview of Xunzi’s descriptions of xin and yu. It is clear throughout the Xunzi that xin is in a leadership position in relation to the body. This can be observed in the way Xunzi differentiates xin from the body and calls it the ruler:

಴ᒮ  হࡽ੪ࡽ

Xin is the ruler (jun ) of the body and the master (zhu ) of the spiritual

intelligence. It issues orders but does not receive orders.1

Here, when Xunzi says that xin issues orders, he is most likely saying that xin issues

1 Xunzi 21/44-5; K 21.6a/3/105. 1. A Challenge to Antagonistic Interpretations orders to the bodily senses. Elsewhere, it is also said that xin is the ruler (jun ) that gives orders to the five senses.2 The impression that xin is in a supreme position is intensified by his subsequent claim that xin acts on its own and is not subject to manipulation.3 The reason that xin is in a ruling position is that the senses do not zhic (know, understand) but xin does:

಴ྞಬಬᏐᒵᏐಬಶಎ 

ዅ!॥#$ࡋ !&಴ಬ'ࡥ)&*&

Xin has zheng zhic. Because of zheng zhic, and by means of the ear, it is possible to know sound; by means of the eye, it is possible to know shape. But it must first wait until the senses have registered their respective categories, then it is possible. The five senses register but does not zhic. If xin zheng zhic but cannot offer an explanation, then people will call this “not zhic.”4

c c Zhi  is usually translated as “knowledge.” Knoblock translates zhi as “awareness,” probably because there is one passage that says grass and wood do not have zhic but the animals do.5 However, Xunzi seems to mean something more than just awareness when he uses the term zhi in relation to xin. It is worth noting that in the last sentence of the above passage, Xunzi posits a difference between zheng zhic and zhic. Zheng means “evidence” or “verification.” In this regard, zheng zhic should be about collecting evidence or data supplied by the senses. The “zhic” in “zheng zhic” can be understood as “awareness,” meaning that xin is aware of the data of the senses. But the “zhic” in “zhic sound” and “zhic shape” should denote a higher level of zhic that requires xin’s offering of an explanation (shuo '). In another passage, it is said that the way horses respond to neighs and oxes respond to lows is not zhic because they are naturally constituted this way.6 These altogether suggest that zhic refers to something more than just reflexive response or awareness. The requirement of offering an explanation implies that xin should be able to make sense of the data

2 Xunzi 17/12; K 17.3a/3/16. 3 Xunzi 21/45-6; K 21.6a/3/105. 4 Xunzi 22/19-21; K 22.2e/3/13. 5 Xunzi 9/69; K 9.16a/2/103-4. 6 Xunzi 3/25; K 3.8/1/177.

11 1. A Challenge to Antagonistic Interpretations supplied by the senses and form a judgment about them. In addition, Xunzi makes a connection between humans having inborn zhic and that xin never stops storing or moving.7 In this light, we can understand zhic as a distinguishing capacity of xin and that this capacity is always enabled. It is clear that xin has considerable influence over how senses react to their preferred objects. Since senses do not have zhi, it is by virtue of xin’s capacity to zhic that the sensory objects may be known. According to Xunzi, it is by xin that sounds and shapes are known. The senses do not know and since their individual capacities are not interchangeable, they need a higher-order mode of operation that processes the data they gather and forms judgements about them.8 Or else, they will only be registering lines and shapes without knowing whether the very thing in front of them is indeed an object of preference. Suppose it is human nature to prefer fine meat. Even if a person is offered the finest meat it still awaits xin to register the evidence provided by the senses and judge that it is a piece of meat. And then, actions are issued accordingly. As such, if xin is not operating properly, discernment of sensory objects will be affected. This explains why Xunzi thinks that a person’s perception of sensory objects will be distorted if her xin is unstable.9 Elsewhere, it is said that the senses will not be aware of their preferred objects when xin is filled with fear.10 This further suggests that the state of xin determines the extent to which the senses function properly. Although xin is said to have control over the senses, it should not be taken to mean that there is a mind-body dichotomy, that xin is the “mind” and the senses belong to the realm of the “body.” There are also descriptions about how xin is similar to the bodily senses in that they each have their own preferred objects, for example, the eyes yu the fullest of colours and the ears yu the fullest of sounds.11 Just like how the eyes by nature have beautiful colours as its preferred object, xin is said to have lic

7 Xunzi 21/35; K 21.5d/3/104. 8 Xunzi 17/11-2; K 17.3a/3/16. 9 Xunzi 21/67-70; K 21.8/3/108-9. 10 Xunzi 22/85; K 22.6e/3/137-8. This sharply contrasts with passage that says the senses yu the fullest

of their objects (Xunzi 11/46-7; K11.4/2/156) because yong  means less than ordinary, which is clearly a state that is less than fullest. 11 Xunzi 11/46-7; K 11.4/2/156.

12 1. A Challenge to Antagonistic Interpretations

12 + (benefits, profit) as its preferred object. We can infer from this that xin is also predisposed to like certain things. Another point of similarity between xin and the senses is that they are capable of differentiating among their preferred objects:

,ᕽ቗ࡾ0ᒵ1࿾ႚ456ᒵࡾ07ᗇᯄ࿦;ᤇ6઎ࡾੰ0?@

ABCᕓEF6@ࡾG0HIটᅘ၀MNOࡾ,0P'QRSTUV

Wപࡾ಴0YY

(1) Forms, colors, and designs are differentiated by the eye. Pitch and timbre, bass and treble, modal keys and rhythm, and odd noises are differentiated by the ear. Sweet and bitter, salty and bland, pungent and sour, and distinctive tastes are differentiated by the mouth. Fragrances and stenches, perfumes and rotten odors, putrid and rancid smells, foul and sour odors, and distinctive strange smells are differentiated by the nose. Pain and itching, cold and heat, smoothness and roughness, and lightness and heaviness are differentiated by the body. (2) (Shuo, gu,) 13 pleasure, anger, sorrow, joy, like, dislikes, and yu are differentiated by xin.14

Just like the eye has the capacity to process visual data, xin is the capacity that discriminates between a range of feelings and yu.15 Note that the things that are said to be differentiated by xin in (2) include yu and the rest of the list is almost the same as Xunzi’s definition of qinga:

12 See section 3.1 for discussion on lic being the preferred object of xin. In the “Wang ba” chapter, we are also told that xin yu relaxation and ease (Xunzi 11/46-7; K 11.4/2/156). Cf. Xunzi 23/26; K 23.2a/3/154. 13 Commentators differ in their interpretations of the first two characters, shuo and gu. I will not investigate the meaning of these two terms here. My focus is on the succeeding terms that parallel the definition of qing. 14 Xunzi 22/17-9; trans., K 22.2d/3/129. 15 Note that the objects differentiated by the senses are external objects whereas the objects differentiated by xin are inner sensations. There is a disagreement about whether Xunzi means that the sensory objects are external. Hu Shih translates “size, shape, color, and texture differ with different eyes.” J.J.L. Duyvendak disagrees and argues, “Not the objective reality of things is doubted, but the text says that the senses enable us to perceive the really existing differences between things” (1924, p.231). I will return to this passage in 3.3.

13 1. A Challenge to Antagonistic Interpretations

Z[പSTUV*ചY

The feelings of like and dislike, of delight and anger, and of sorrow and joy that are given from birth is called qinga.16

a It should be noted that Xunzi uses the term “qing ച” in two senses, one referring to what is truly the case of something17 and one referring to inborn feelings. It is in the latter sense that Xunzi provides his own definition of “qinga.” The term “qinga” that appears in the above quote is usually translated as “emotion.”18 However, Xunzi’s usage of qinga could be quite different from the usage of “emotion” in ordinary English, for it is not obvious that qinga refers to mental states that can give rise to action or to a ange of phenomena that can be viewed independently of a subject. I will therefore use the term “feeling” in a loose sense throughout my discussion in order to avoid unnecessary presuppositions.19 From the above two quotes, we can infer that qinga for Xunzi refers to a range of natural feelings that cannot be modified. This idea is explicitly expressed in his definition of qinga. It can also be inferred from the parallel Xunzi draws between (1) and (2), in which xin is discussed along with bodily senses and qinga with sensory objects. If the parallel between (1) and (2) is read in a strict sense, then qinga is more like a sensory object or stimulus that is perceived by xin. Xin can detect and differentiate qinga but cannot change these spontaneously aroused feelings. Qinga is just “there”. This understanding of Xunzi’s concept of qinga is also consistent with Xunzi’s view on qinga expressed in Passage 22.5a, a passage we will consider in detail in section 1.4. This reading also fits well with Xunzi’s idea that human beings shares one and the same qinga.20 If qinga can be modified by xin, then the same sensation (e.g. eating fine meat) could be pleasure to one person but pain to another, depending on the modifiable states of xin. In this sense, it is more difficult to see why Xunzi thinks that everyone has the same qinga.

a Xunzi defines yu  in relation to qing :

16 Xunzi 22/3; K 22.1b/3/127. 17 See section 3.3 for discussion of qinga used in this sense. 18 See, for example, Dubs, 1996b, p.281; Watson, p. 144; K 22.1b/3/127. 19 I thank Professor Kwong-loi Shun for alerting me to this point. 20 For example, Xunzi 3/36; K 3.10/1/179. Xunzi 11/76; K 11.7a/2/159.

14 1. A Challenge to Antagonistic Interpretations

ᒮചංY

a 21 Y Yu is the response of qing .

This definition of yu may be understood in different ways. It is possible that yu is a necessary conseqence of qinga uor itself that isy a response and can be independent of qinga. I will leave this issue open-ended for now and discuss it in section 2.3. It suffices our purpose for now to see that yu arises in response to qinga. Yu is commonly translated as “desires” in the literature.22 It should be noted that Xunzi uses the term “yu” in two different senses. One sense of yu is what we may call a broad notion of yu, which refers to the psychological state directed towards a favourable object. The objects of yu, in this broad sense, can virtually encompass everything imaginable. In the Xunzi, this broad sense of yu is mostly used as a verb in the text and is not used as a concept. Some instances where yu is used in this broad sense are:

^_`a॥b

The scholar behaves in this way because he [yuverb (would like, want)] to cultivate his own person.23 and,YY

Y

cdeໟ࡯༢ख़jklm

If you [yuverb (would like, want)] to observe a millennium, you must look at today. If you [yuverb (would like, want)] to know ten thousands or a million, you must scrutinize one or two.24

In the above occurrences, the term “yu” simply refers to a kind of favourable attitude. Xunzi’s official definition of yu that we saw above is concerned with a narrower sense

21 Xunzi 22/63; K 22.5b/3/136. 22 For example, see translations of “yu” by Dubs, J. J. L. Duyvendak, Y. P. Mei, Watson, and Knoblock. 23 Xunzi 2/26; K 2.7/1/155. 24 Xunzi 5/31, K 5.4/1/206.

15 1. A Challenge to Antagonistic Interpretations of yu, which confines yu to inborn and spontaneous responses. It is this narrow sense of yu that is discussed by Xunzi as a concept. I take Xunzi to be speaking in this narrow sense of yu when he talks about the problems associated with blindly pursuing the objects of yu and when he rebuts his contemporaries’ views on dealing with yu. Hence, my subsequent analysis of Xunzi’s concept of yu will be concerned with this narrow sense of yu. Although Xunzi has not elaborated on how each of the feelings identified in the definition of qinga will affect yu, we can infer a positive correlation between yu and the feeling of liking or fondness (hao [). There are many instances in the Xunzi where the object of yu is identified with the object of hao. There is a passage that says common people hao benefits and another parallel passage that says the common people’s yu for benefits cannot be eliminated:

[+n)&oࡾpჽr

None of those commoners who hao profit will want to consider Qi home.25

ᬵtu&vwn+

Although Yao and Shun are not able to eliminate the yu for profit in people...26

Another passage says that both the gentleman and the petty man hao honour whereas another passage says that both sage Yu and tyrant Jie yuverb honour:

[xപᡠ[+പz{|}ࡥ~

In hao honour and detesting disgrace, in hao profit and detesting harm, the gentleman and the petty man are the same.27

xപᡠ{€~

Yu honour and detest disgrace. In this both [sage] Yu and [tyrant] Jie were the

25 Xunzi 16/30; K 16.4/2/242. 26 Xunzi 27/65; K 27.63/3/222. 27 Xunzi 4/33; K 4.8/1/190.

16 1. A Challenge to Antagonistic Interpretations

same.28

And as earlier mentioned, it is said that senses are predisposed to like certain objects. For example, the eyes hao colours and the ears hao sound. Elsewhere, the senses are described as yu those very preferred objects:

[ᕽ[ੰ[઎಴[+‚{ጳ„ ࡥചZᒮY

The eyes hao colours, ears hao sounds, the mouth hao tastes, xin hao profit…all these are products of humans’ qinga xinga (inborn feelings given by birth/what is true of them).29

^ࡥചᎫᕽᎫᒵੰᎫ઎GᎫ@಴Ꭻ‡

It is the qinga of humans that their eyes yu the most intense of colours, their ears yu the richest of sounds, their mouths yu the most intense of flavours, their noses yu the richest of aromas, their xin yu effortlessness/pleasure.30

Recall that xin also has its preferred object, that is, lic. Xunzi would probably not object to us saying that xin yuverb lic.31 Xunzi thinks that all human beings have certain common objects of yu. It is frequently discussed in the text that humans have the common yu for certain things such as beautiful clothes, delicious food, wealth, power etc.32 It is also said that all human beings will have the same objects of yu when they are in similar circumstances:

দࡥྞl~‰Š‹ŒŽ[+പz{ࡥ„ྞ

{ಎᒮ{€~

28 Xunzi 12/66; K 12.8/ 2/185. 29 Xunzi 87/23/25-6; K 23.2a/3/154. It is not clear whether the character “qing” here means “what is genuine of something” or “feelings that are given from birth,” as defined in Chapter 22. 30 Xunzi 39/11/46-7; K11.4/2/156. 31 See section 3.1 for further discussion. 32 Xunzi 40/11/74-6; K 11.7b/2/160 and Xunzi 11/4/60-1; K 4.11/1/193).

17 1. A Challenge to Antagonistic Interpretations

All human beings have one thing in common: when hungry, they yu food; when cold, they yu warmth; when exhausted, they yu rest; they like profit and dislike harm. Such are what humans are born to have, which do not wait to be developed, which is what Yu and Jie have in common.33

While Xunzi believes that there are some objects of yu that are commonly shared by everyone, sages and tyrants alike, Xunzi also seems to think that the objects of yu are subject to change. It is said that through carefully weighing and calculating options, one can decide on what should be the objects of yu.34 Elsewhere, it is said that learning can even make one not yu to see and talk about things that are contrary to the teaching of sages.35 In Xunzi’s discussion that human nature is bad, he maintains that following the yu of the senses will lead to disorder and lewdness. Goodness, in Xunzi’s view,

36 arises from wei  (artifice, modification), where wei is defined in terms of the activities of xin.37 This raises the intriguing question as to how exactly xin relates to yu so that goodness is possible despite human nature being bad. Before we examine this relation in detail, let us begin with some secondary discussions on this subject.

1.2 Secondary Interpretations

There are only a few extended discussions of the relation between the concepts xin and yu in the Xunzi. Although there are a number of mentions of how xin influences yu in the literature, they are usually embedded in discussions on other subject matter rather than being treated extensively as a topic in its own right. The drawback of such an approach is that when accounts of the relation between xin and yu are confined within the framework of a larger project, the explanation of these concepts tends to be rushed and hence often hides unsupported theses about both concepts, xin and yu, and their relation. In this regard, I will highlight two difficulties associated with the ususal way scholars have apprehended the subject matter.

33 Xunzi 10/4/42-4, K 4.9/1/191. 34 Xunzi 86/22/72-8; K 22.6b-c/3/137. 35 Xunzi 3/1/47-9; K 1.14/1/142. 36 Xunzi 87/23/1; K 23.1a/3/150. 37 Xunzi 83/22/4; K 22.1b/3/127.

18 1. A Challenge to Antagonistic Interpretations

One difficulty is that interpreters often conflate two questions in studies of Xunzi: the question of how xin and yu are related and the question of what the reason for xin’s evaluation is. Very often, discussions of the relation between the concepts xin and yu are confined within the framework of moral evaluation. For example, the argument could run along the following line: xin evaluates on the basis of how best to satisfy the overall set of yu. So, on ocassion xin may override immediate desires so as to better satisfy the overall set of desires. Hence, xin makes prudential choices, not moral choices. Another example would be an argument that says xin evaluates in accordance with moral principles. In cases where the objects of yu are not morally permissible, xin will suppress yu.38 The ground for attributing a moral quality to xin is the understanding that xin brings yu in line with what is morally permissible. In these two examples, it is unclear whether an understanding of how xin evaluates is used to inform an understanding of how xin relates to yu, or the other way around. It seems that both arguments are dealing with two questions at the same time and are consequently answering the questions of how xin and yu are related and that of how xin evaluates in reference to each other. The tendency to conflate these two questions is problematic because it reveals a confusion about the distinction between two questions that are of a different nature, one concerns the nature of the relation between two concepts while the other concerns the nature of xin’s evaluation. Dealing with these two questions simultaneously raises the risk of hastily dismissing one or both of the questions involved, or of hiding possible tensions in Xunzi’s thought, or of generating self-validating readings of Xunzi. This is not to argue that the issue of how xin evaluates does not shed light on the issue of how xin relates to yu, or vice versa. My contention is that these two questions are important questions that call for extensive investigation and should therefore be treated separately on their own terms rather than using one question to frame the answer to the other question. Another difficulty with existing interpretations about the relation between the concepts xin and yu is that interpreters are often vague in their discussions about the relation between xin and yu. One possible reason is that both concepts are laden with assumptions. This problem pertains to both Chinese language studies and English language studies. In studies conducted in Chinese, there is a tendency to use the same Chinese term to explain an important term in the classical texts that requires extensive

38 These two examples are oversimplified versions of two lines of interpretation proposed in the literature. I will elaborate on this point in my discussion of specific interpretations.

19 1. A Challenge to Antagonistic Interpretations analysis. Such an approach usually overlooks and leaves unexamined the subtleties embedded in Xunzi’s usages of these terms in his intellectual. The term jie , for example, can be understood in at least two different senses. One means to control in a violent sense while the other means moderation in a milder sense. Each reading of “jie” will yield a significantly different interpretation of Xunzi from the other. However, in most Chinese language studies, the same term jie is used to explain Xunzi without much analysis or elaboration.39 But to an English reader, it still begs the question as to which of the two senses of jie the interpreter has in mind. In this regard, English studies have the advantage of being more sensitive to the different connotations embedded in specific Chinese terms. Nonetheless, when the terms xin and yu are conveniently translated as “mind” and “desire” in English, certain assumptions from Western philosophical traditions about the mind and desire that inherit will be carried over into our understanding of Xunzi’s concepts of xin and yu. Moreover, it is inviting to subsume discussions of the relation between xin and yu under the familiar Western philosophical topic of “mind and desire.” As a corollary, the guiding questions behind scholarly interpretations could well be questions that have long interested discussions of mind and desire in Western philosophical traditions; however, these questions about mind and desire are not necessarily the questions that concern Xunzi himself. In light of the two difficulties discussed above, my goal in the following is to only focus on examining Xunzi’s conception of the relation between xin and yu. The question that concerns the basis of xin’s evaluation will be reserved for Chapter 4. In the following, I will first lay bare certain assumptions about the relation between xin and yu from a sample of scholarly works that I consider as representative of the different lines of interpretations on the subject matter. I will then challenge these assumptions by using a passage in Chapter 22 of the Xunzi as an entry point. The purpose of this exercise is to prepare the ground, with minimal assumptions, for my own interpretation of the subject matter. Here, I point out two related tenets that underlie common interpretations of the relation between xin and yu.

39 See for example Li 1979, Liang 1983, and Wei 1992.

20 1. A Challenge to Antagonistic Interpretations

(a) The attribution of causal roles in action to both xin and yu, that is, understanding xin and yu as both capable of causing action by themselves. (b) Connected to the view that both yu and xin have causal roles in action is the view that both xin and yu have inherent moral characteristics. Yu is often characterised as wayward and bad or at least something that brings about consequences that are disorderly, while xin, when properly cultivated, is the overriding moral capacity of humanity. Where there is interaction between xin and yu, xin is usually thought to overcome or redirect yu.

In the following discussion, I will demonstrate the predominance of these assumptions in the literature by assessing a number of interpretations that explicitly or implicitly suggest the abovementioned tenets.40 Homer H. Dubs, considered the first scholar to comprehensively introduce Xunzi’s thought to an English speaking audience,41 claims that there is no sharp distinction between reason and emotion in Xunzi as in ’s division of the soul. He states:

In contrast to this theory we may place that of Plato that plants have appetites; animals have will in addition to the appetite; and men have, besides these two functions, the rational principle…In [Xunzi] there is no such thing as a purely cognitive relation; such a world as that of much of Western thought is entirely foreign to him.42

On the other hand, he argues that the mind is separated from desires because Xunzi divides the human “soul” into the mind and the desires. His understanding of the relation between the xin and yu is that:

For Xunzi the dichotomy of the soul is into the “mind” or “will” and the “desires” or “emotions”.43

40 In my discussion of secondary interpretations, I will retain the English translations scholars adopted so as to represent their views with minimal distortion. 41 His book Hsuntze: The Moulder of Ancient Confucianism (1927), and his translation of 19 books (pian) of Xunzi The Works of Hsuntze (1928) paves the way for later research on Xunzi’s thought. 42 Dubs 1966a: 52. 43 Dubs 1966a: 171.

21 1. A Challenge to Antagonistic Interpretations

And that:

Xunzi keeps the reason and will together as the mind, and separates off only the appetites or emotions.44

In both these statements, Dubs sustains a clear distinction between mind (xin), on the one hand, and desires or emotions (yu), on the other. His dichotomy nominates reason and willing as capacities of the mind, carried out in isolation of the “appetites or emotions.” If the mind consists of reason and will, then Dubs’s interpretation still upholds a division between reason and emotion in Xunzi, which can eventually be translated into a dichotomy between the mind and desires. Dubs’s difficulty in breaking away from a dichotomous framework is partly due to his attribution of causal role in action to yu. Dubs defines “desire” in terms of “an impulse to seek for certain objects” and “a seeking.” According to Dubs:

[T]here are two elements in desire as a fundamental human capacity; there is desire in the sense of an impulse to seek for certain objects, and there is desire in the sense of a seeking or a purpose. 45

Dubs is suggesting here that yu alone can carry out the act of seeking its objects because the whole process from having an impulse to seeking the objects could have nothing to do with xin. This can explain why xin and yu are conceived antagonistically by Dubs. Another related reason for the ambiguity in Dubs’s account of xin and yu is that he has made a distinction between “principle” and “inner principle.” According to Dubs:

The mind can be controlled by principle, and then even if the desires are many, the mind can control them. But if the mind is not controlled by principle, though the desires be few, yet there would result disorderly conduct.46

44 Dubs 1966a: 173. 45 Dubs, 1927, p.51. 46 Dubs, 1966a, p.173.

22 1. A Challenge to Antagonistic Interpretations

A paragraph later, Dubs says:

Other philosophies which do not follow the Confucian Way but give some inner principle will likewise fail. For judgments in the matter of desire are unusually difficult.47

Unfortunately, Dubs has not given sufficient explanation of what exactly the following of “principle” and “inner principle” means. Based on his discussion, it seems that “principle” is derived from standards established by the sage kings, which Dubs calls “the Confucian Way,” whereas an “inner principle” is derived from yu:

Indeed for the ordinary man, his mind is certain to go astray; it is only the superior man or Sage who can conform to the three conditions of emptiness, concentration, and unperturbedness, and can judge by an inner standard; others must follow the standard that the Sages worked out, namely the Confucian Way, in order to avoid prejudice and error. 48

Dubs further says that there will be “disorderly conduct” when xin follows an inner principle and that there will be “orderly conduct” when xin follows the principle.49 Although Dubs has not explicitly said why there will be disorderly conduct if xin is not following the principle, given his claim that orderly conduct results from xin controlling yu, he probably regards disorderly conduct as resulting from xin not controlling yu. In other words, when xin is guided by an inner principle (that is its own judgment), it gives free rein to yu; when it is guided by principle (that is the Confucian Way), xin controls yu.50 If this is the case, then the role of xin is simply to allow or disallow yu from producing certain effects. In the case of disorderly conduct, xin cannot be held as the direct cause of disorder because yu is the agent that pursues the objects. Xin simply fails in its job to overcome yu but it itself is not the active agent producing the disorder. Thus, yu is still understood by Dubs as possessing causal role in action and is separate from xin. Yu is capable of producing bad effects

47 Dubs, 1966a, p.174. 48 Dubs, 1966a, p.173. 49 Ibid. 50 On Dubs’s understanding, only the sages’ inner principle is in accordance with external principle.

23 1. A Challenge to Antagonistic Interpretations by itself whereas xin, when following principle, produces good effects. In this regard, xin and yu are clearly antagonistic, for the two can have very different and even conflicting orientations. Lee Yearley’s analysis of Xunzi’s xin is one of the first extended studies in English that focuses on the concept of xin.51 He argues that there are two conflicting concepts of xin for Xunzi: xin as “director” and xin as “spectator”. When Yearley describes xin as a director, he says that:

[Xunzi’s] model of mental direction, however, emphasizes conflict and mastery. The mind directs the recalcitrant and selfish spontaneous movements of the self, forcing them to follow the mind’s direction rather than their own natural inclinations.52

The antagonistic terms such as “conflict” and “mastery” suggest that there is an opposition between xin and yu as two separate agents.53 For Yearley, short-term desires are “narrowly selfish desires” whereas long-term desires are “desires that take a wide view of the self’s needs.”54 The mind as a director overcomes short-term desires in order to better satisfy long-term desires. Yearley gives the example of someone who spontaneously desires food but gives the food to an elderly person because he desires to be fed when he is old. The role of the mind is to take a comprehensive long-term view of the needs and interests of the self and overrule short-term, “spontaneous” desires from taking actions that will prevent satisfaction of desire in the long run. Yearley’s interpretation has taken yu as capable of causing action in and of itself, for it assumes that yu could pursue its objects when xin does not interfere. The other aspect of xin is as a spectator. According to Yearley:

Nevertheless, the key to detachment remains the mind’s capacity to be a spectator. [Xunzi’s] perfected mind operates on the desires in a fashion that resembles certain experiences that most people have had at some time…On

51 Lee Yearly’s article “Hsun Tzu on the Mind: His Attempted Synthesis of Confucianism and ” was published in 1980. 52 Yearley, 1980, p.468. 53 It is also unclear if Yearley has assumed that that there is a separation between “mind” and the “self” where the latter actually refers to desires. 54 Yearley, 1980, p.466.

24 1. A Challenge to Antagonistic Interpretations

certain occasions, our minds detachedly view objects to which we normally are attached.55

The spectator xin can objectively watch the activities of yu without being in any way involved with them. Simply put, this aspect of xin is unaffected by yu. Here, Yearley reaffirms the position that xin and yu are antagonistic to each other. Yu can pursue its own objects independently of spectator xin while spectator xin is also detached from all activities of yu. On this level, not only are xin and yu separate, they do not even interact with each other. Yearley’s understanding of xin and yu is analogous to the conflict a mother may have with her son. The son may want to do something but his mother stops him because she knows he is not sufficiently mature to tell what he really wants, or what he should want is not what he now wants. Perhaps when the son grows up he will realise that what he really wants is not in conflict with his mother’s advice at all. But ultimately, the son is the one who is carrying out the actions. Although his mother might overrule some of the actions that he wants to take, it still means that it is the son’s actions that are halted, not the mother’s action. Similarly, when there is a conflict between xin and yu on Yearley’s account, xin is simply stopping the movement of yu. Hence, xin and yu are understood as two independent agents and that it is possible to have a collision between two separate roles in causing action. In David Nivison’s analysis of Xunzi’s philosophy, the relation between xin and yu is placed in terms of the long-term satisfaction of desires. On this point, Nivison’s analysis is somewhat similar to Yearley’s understanding. According to Nivison, yu moves blindly towards its objects. Xin, on the other hand, can prudentially assess yu and make judgments about how yu can be best satisfied. In this sense, the reason xin stops yu is not that yu is inherently bad but that momentary satisfaction of yu will prevent long-term satisfaction of yu. Nivison writes:

In any case, it is clear that Xunzi thinks that my judgments about what is the best course – best on some ground – can modify the action that desire alone would determine…[It] is possible to regulate one’s seeking, even if not to get rid of desires themselves; and it is possible to approach satisfaction, even if

55 Yearley, 1980, p.474.

25 1. A Challenge to Antagonistic Interpretations

not to attain complete satisfaction. And so it is possible to make an overall choice of courses of action that optimizes satisfaction and minimizes disorder, danger, and anxiety…When in fact people choose ill, they are simply confused: in a sense they have failed to add and subtract, and take one immediate satisfaction in place of many long-term satisfactions.56

On Nivison’s interpretation, xin and yu are again conceived antagonistically in a number of ways. Translating yu as “desire,” Nivison asserts that yu has an independent, even primary, causal role in decision-making. He states that, “our desires taken by themselves tend to lead us into conflict” and xin can modify the “[actions] that desire alone would determine.”57 This shows that Nivison has understood yu in terms of its causal role in action, that is, its capacity to bring about actions with disorderly consequences. It is obvious that, for Nivison, the activities reside in different domains of decision making: whereas yu functions “automatically,” xin is capable of making intelligent and properly considered decisions. In a discussion of the senses, Nivison states specifically that desires (yu) are not necessarily subject to deliberation:

the senses’ seeking of their objects is automatic and will go forward unless the faculty of intelligent decision in us brings it to a stop or slows or redirects it.58

Not only does such an account of the relation between xin and yu attribute agency to both xin and yu, it also polarises the different domains of xin and yu. Whereas yu is a “given,” that is, installed in human nature, xin is a regulative capacity that makes judgments that bear on the activity of yu and prevent yu’s automatic pursuits. On this view, it seems that yu can by itself take a course of action, even though its movement is automatic and blind. The role of xin is to check the movement of yu where necessary. Antonio Cua seems to think that xin limits and transforms the movement of yu. This can be observed in Cua’s discussion of the transformation of “natural desires”

56 Nivison, 1996c, pp.208-209. 57 Nivison, 1996c, p.208, italics mine. 58 Nivison, 1996c, p.86. Although Nivison refers to the “senses” rather than desires in this passage, this passage is embedded within a larger discussion about optimum satisfaction of human desire wherein he draws a connection between senses and desires.

26 1. A Challenge to Antagonistic Interpretations into “reflective desires.” Although Cua has not explicitly said that xin transforms yu, we can infer that Cua thinks that xin regulates yu by changing the way yu behaves. In showing how xin relates to yu, Cua has quoted a passage in the Xunzi, which can be literally translated as the following:

੪‘l’ ੪‘಴“஭ᬼ#੪‘

The one yu, which is received from Heaven, is zhi ’ (regulated, worked

upon), which is received by the many […] of xin, until it becomes difficult to categorise what is received from Heaven. 59

Cua takes the above line to mean:

…the moral transformation of emotions and feelings is essentially a change, not in the sense of direction alone, but in an alternation of the form of expression of natural emotions and means of satisfaction of material desires.60

Cua has not elaborated on what he means by “the form of expression” of yu. What we do know is that before xin comes to transform yu, yu already has a form of expression of its own. Cua describes this crude form of expression as “a mere biological drive.”61 Cua gives the example of one’s natural desire for food, which has no regard for the proper ways of satisfaction that accord with ritual practices. This suggests that the “expression” of yu, for Cua, should have included the way yu seeks its objects. Without the intervention of xin, yu would just proceed to pursue objects in a wayward manner. Cua further writes:

These desires are reflective desires, because they have passed the test of moral approval. To pursue these second-order desires in a moral way requires an exercise of a second-order volition, that is, a second-order desire to constitute our will as a moral agent rather than an amoral agent.62

59 Xunzi 22/58-9; See Section 1.4 for detailed analysis of this line. 60 Cua, 2005c, pp.49-50 n28. 61 Ibid. 62 Cua 2005c, p.50.

27 1. A Challenge to Antagonistic Interpretations

In Cua’s view, reflective desires are transformed through the exercise of second-order volition. To be more explicit, Cua is saying that the exercise of second-order volition is taken in the form of xin issuing approvals to natural desires because natural desires themselves are blind responses. The very fact that there is the requirement for “moral approval” from xin shows that yu can potentially go against xin. The task of xin, then, is to re-steer the course of yu to a track that is morally permissible. Cua’s depiction of the relation between xin and yu strongly suggests that yu can by itself seek its own objects. The role of xin is to modify the behaviour of yu. In this sense, I consider Cua as also adopting the assumption that yu in and of itself has a causal role in action. In a similar vein, T. C. Kline talks about how xin can direct and transform yu into a more complex motive. Kline differs with Cua in that Kline does not regard this newly transformed motive as yu anymore. According to Kline, since this new motive does not emerge from human nature, labelling it as yu will contradict Xunzi’s definition of yu that says yu is what comes from Heaven.63 According to Kline:

This new motive can incorporate a wide range of cognitive descriptions and evaluations as well as rest on sensitive perception and understanding of the nature of external factors.64

Although Kline is eager to distinguish the newly transformed motive from yu, he also thinks that yu alone can “lead to forms of expression that encourage chaos and conflict”65 and for that reason, yu should be countered and redirected. The following passage exemplifies Kline’s conception of the relation between xin and yu:

…Xunzi believes we have two primary sources of motive force in the self. One source is the disposition of our nature that, in responding to our environment, produces desires. The second source springs from the capacity of our xin to both understand and judge possible courses of action, to approve and disapprove. Because we are usually motivated by approvals and disapprovals and not desires, Xunzi recognizes the fact that individuals may find themselves approving courses of action that they do not necessarily want

63 Kline, 2000, p.161. 64 Ibid. 65 Ibid.

28 1. A Challenge to Antagonistic Interpretations

to approve. This conflict between desire and approval reveals a second reason that desires are problematic. They potentially create conflict within the self.66

It is quite clear from the above that Kline attributes a causal role in action to both xin and yu. Although xin usually gets to cause actions, it is also possible that yu can cause actions alone. In this light, Kline understands the relation between xin and yu as one about how the two causal forces in action compete against each other with one force (yu) being redirected by another force (xin) into a proper course. The above overview is by no means an exhaustive review of scholarly discussions of the relation between xin and yu in the Xunzi. Nonetheless, it represents some of the major discussions about the relation between xin and yu in the literature. I hope it is clear by now that, among the various interpretations of the relation between xin and yu that have been proposed, there underlies a commonly accepted tenet that both xin and yu can cause action and are therefore capable of causing actions in their own right. Under this assumption, whether it controls yu or redirects yu, the relation between xin and yu is essentially characterised in terms of the interaction between two agents that can both cause action. Let us now delve deeper into the implications of this assumption.

1.3 Antagonistic Interpretations (AI)

In this section, my task is to make explicit the implications of attributing causal roles to both xin and yu. In all of the analyses above, the implications of the views – whether xin curtails the activity of yu, or whether it re-steers yu toward different objects or goals – is that xin and yu are antagonistic. However, it is also the case that in these views, there is little explicit discussion of the connection between xin and yu. Hence, the analysis in this section focuses on the potential implications of (a) attributing causal role in action to xin and yu and (b) rendering their relationship an antagonistic one. It does not focus specifically on the views outlined by each of the scholars above. Instead, I will label those interpretations that attribute causal agency to both xin and yu as “antagonistic interpretations (AI).”

66 Kline, 2006, p.239. Italics mine.

29 1. A Challenge to Antagonistic Interpretations

I categorise antagonistic interpretations of the relation between xin and yu into two kinds. The first can be called the “path-blocking” interpretation. It holds that xin monitors and controls yu by determining whether it is permissible to pursue the objects of yu. If xin, for whatever reason, deems an object of yu as impermissible to pursue, it can exercise control on yu, thereby thwarting yu from pursuing their objects. Dubs, Yearley, and Nivison follow this line of interpretation. The second type of AI can be called “path-re-steering” interpretations. It holds that xin can redirect yu to some other new objects such that yu pursues the objects that xin approves. The interpretations put forward by Cua and Kline lean towards the path-re-steering model. We can invoke a water analogy to explain the path-blocking and path-re- steering interpretations. On this model, yu is analogous to the water flowing in a drain and the objects of yu are analogous to the destination. Xin is the valve positioned in between that can affect the flow of water. On the path-blocking model, the valve can either block or allow the water to flow but cannot change its path. In other words, xin can merely stop yu from pursuing its objects but cannot change its objects; the path- re-steering model, on the other hand, has a valve that can either block or re-steer yu to some new paths. After yu has been reshaped by xin, xin no longer has to stop yu pursuing objects. The main difference between the path-blocking and the path-re- steering interpretation concerns the changeability of objects of yu. The path-blocking interpretation holds that the objects of yu are predetermined and cannot be changed whereas the path-re-steering interpretation holds that the objects of yu can be determined by xin. Common across these two interpretations is the understanding that xin and yu are two antagonistic causal forces in action. It is obvious that the path- blocking interpretation, which understands xin as blocking yu, has already assumed that xin and yu are two opposing forces in action. Although the path-re-steering interpretation holds that yu can eventually be brought into agreement with xin, it still stipulates an initial conflict between xin and yu as two causal forces in action or else there would not arise the need to re-steer yu. To make my point of disagreement with the AI models clearer, we can consider the following three assumptions: (1) there are predetermined objects of yu, (2) yu can itself cause into actions to pursue its own objects, and (3) (1) will automatically lead to (2). The path-blocking model would accept all three assumptions. For instance, if yu has food as its object, it will automatically pursue food unless it encounters certain hindrance. The path-re-steering model would reject

30 1. A Challenge to Antagonistic Interpretations

(3) by arguing that the movement of yu towards its predetermined objects is not necessarily automatic because yu could be re-steered to some other directions. Nonetheless, both models would accept (2) and it is in this sense that I group both of them as “antagonistic interpretations (AI).” (1) is the obvious assumption that interpreters who advocate AI adopt. They understand the predetermined objects as potentially problematic in certain circumstances and therefore the pursuit of these objects should be checked. (2) is a more hidden claim underlying AI, which understands yu as capable of causing action by itself. Since Xunzi also discusses situations in which the objects of yu are not being pursued and since AI proponents hold that yu can pursue its objects by itself, their explanations for cases where the objects of yu are not obtained have to be something along the line that says yu is being controlled by xin. If we think in terms of the hydraulic model, we see that the AI interpreters have taken the water flow, i.e. yu, to be what is reaching the output. This means that AI conceptual models assume at the outset that yu can pursue its objects by itself. The role of xin is to control yu so as to stop yu from pursuing objects in an automatic way that is likely to cause disorder. In instances where the objects of yu are not obtained, according to the AI interpreters, it is because xin has thwarted yu’s pursuit of them. I agree that (1) is a reasonable assumption because Xunzi does say explicitly that humans are naturally drawn toward certain objects of yu67 and it is problematic when the pursuit of these objects is left unchecked.68 The acceptance of (1), however, does not entail (2) is also true. My following task is to call into question assumption (2) by arguing that there is insufficient textual evidence to substantiate (2). If I can successfully discard (2), it follows that (3) is also discarded. I primarily rely on passage 22.5a of Xunzi because it has been widely cited by AI interpreters to support their theses. As my analysis will show, there is no ground for us to think that yu in and of itself can cause action.

67 Xunzi 11/46-7; K 11.4/2/156. 68 Xunzi 19/1.

31 1. A Challenge to Antagonistic Interpretations

1.4 Passage 22.5a

Passage 22.5a explicitly discusses the relation between xin and yu.69 This passage also appears frequently in literature supporting AI. Since it talks about how the pursuit of the objects of yu will follow xin’s approval, it is usually taken by AI interpreters to mean that xin can override yu. When Xunzi writes this passage, he is responding to

a two main contemporary proposals about yu. One proposal is that order (zhi –) depends on the elimination of yu. This view is probably first advocated by Song Xing.70 Another proposal is that order depends on the reduction of yu. Mencius is likely to be the target of criticism here because Mencius regards a reduction of yu as the key to nurturing xin.71 Since “order (zhia)” is included in Xunzi’s definition of goodness,72 I will assume in my subsequent discussion that maintaining order exemplifies goodness for Xunzi and hence is regarded as an ideal state by Xunzi. Xunzi rebuts these two proposals about yu by distinguishing the difference between what is within and what is beyond human control and arguing that order does not depend on eliminating or reducing yu. The following is a literal translation of the passage:

[1] দ—–ಎwᒮࡾ˜™ ྞᒮদ—–ಎšᒮࡾ

™ “ᒮYY

[2] ྞ0#„›œ–࠿“š0#ചໟœ–࠿YY

[3]Y&ಎಚŸᒮ &ಎಚ੪‘Ÿᒮ ੪‘಴

YY

[4]Y੪‘l’ ੪‘಴“஭ᬼ#੪‘YY

[5]Yࡥ„¡¢ࡥപ›¡¢ࡥྞ „£›ᒮœ&„›

&ࡾ„ࡾ›YY

69 Xunzi 22/56-62; K 22.5a/3/135. 70 Xunzi 18/114-7; K 18.10/3/48. 71 Mencius 7B35. 72 Xunzi 23/37.

32 1. A Challenge to Antagonistic Interpretations

[6] Q¤਄&੣಴§಴¨቗ᬵ“©प –&੣

਄¤಴«಴¬቗ᬵš©§ ࠿Y

[7]YQ–࠿­ ಴ࡒ ച&Ÿ॥­Ÿ॥ࡒᬵ¯°

ಚ¬¢YY

[1] (i) All those who maintain “order depends on the elimination of yu” are those who lack the means to dao yu and are therefore troubled by the fact that yu exists at all. (ii) All those who maintain “order depends on the reduction of yu” are those who lack the means to jie yu and are therefore troubled by the fact that yu are many. [2] (i) Having yu and not having yu, different category, life and death, not those of order and disorder. (ii) The numerousness and fewness of yu, different category, the count of qing (natural likes and dislikes), not those of order and disorder. [3] (i) Yu does not depend on whether its obtainment is ke (possible/permissible), (ii) but the one who pursues (qiu zhe) follows what is ke. (iii) That yu does not depend on whether the obtainment of their objects is possible/permissible is received from Heaven. That the one who pursues follows ke is received from xin. [4] The one/singular/common yu, which is received from Heaven, is worked upon/regulated (zhig) by that which is received by the many of xin/made into many by xin, until it becomes difficult to categorise what is received from Heaven. [5] What human beings yu most is life, what humans hate most is death. Even so, there are some who follow life and end up with death. It is not that they do not yu life and they yu death. [In this case], it is not possible to live but possible to die. [6] The reason yu is excessively intense but action does not reach the point is that xin stops it. If what xin makes possible/permits is in line with lia, then although yu are many, how could it harm order? The reason yu does not reach a point as intense as action is that xin causes it. If what xin makes

33 1. A Challenge to Antagonistic Interpretations

possible/permits loses lia, then although yu are few, how could it stop at disorder? [7] Therefore, order and disorder reside in what is ke by xin, not in what is yu of qing (natural likes and dislikes). If one does not seek where it resides but instead seeks it where it does not reside, even if you claim that you obtain it, you indeed lose it.

Since [1] is the conclusion of Xunzi’s argument, we can put it aside for now and look at how the succeeding lines are used as premises to support [1].

Line [2]

(i) ྞ 0# „› œ–࠿

(ii) “š 0# ചໟ œ–࠿

(i) Having yu and not having yu, different category, life and death, not those of order and disorder. (ii) The numerousness and fewness of yu, different category, the count of qing (natural likes and dislikes), not those of order and disorder.

[2] can be read in two possible ways. One reading is to understand [2i] as saying that having yu and not having yu are two matters in two different categories.

Yang Liang ±² offers this reading in his commentary. He thinks that the difference between the presence and absence of yu is like the difference between life and death. Having yu and not having yu are two different categories of things.73 Watson,74 Knoblock, Liang Qi-xiong,75 Zhang Shi-tong,76 and A. C. Graham,77 among others, adopt this reading. A second way to read [2i] is to understand it as saying that the whole issue of having and not having yu is in the category of life and death, not in the

73 Wang Xian-qian, p.427. 74 Watson, p.154. 75 Liang Qi-xiong, p.321. 76 Zhang Shi-tong, p.253. 77 Graham 1989, p.252

34 1. A Challenge to Antagonistic Interpretations category of order and disorder.78 This reading is suggested by Y. P. Mei.79 Despite the obvious popularity of the first way of reading, I think there are both textual and philosophical considerations against the first reading and favour the second reading. In [2ii], Xunzi is trying to say that the quantity of yu belongs to the different category of qinga, not that of order and disorder. The structure of the first half for both [2i] and [2ii] are clearly the same. If we follow the first reading, then [2ii] should mean the numerousness of yu belongs to one category and the fewness of yu belongs to another category. If “life and death” are the names for the two categories of having yu and not having yu in [2i], then we would also expect to see two names for the categories of numerous yu and few yu in [2ii]. However, the first reading will not fit with the second half of [2ii]. “The number of qing ചໟ” is describing the category as a whole rather than as two different categories.80 It is obvious that [2ii] means the issue concerning the numerousness and fewness of yu is in the category of qinga, not that of order and disorder. And since [2i] and [2ii] are written in response to those who maintain that yu should be eliminated and reduced, as mentioned in [1], these two sentences are supposed to express a similar meaning. Then, [2i] should mean the issue concerning having or not having few yu is altogether in the “life and death” category not the “order and disorder” category. The former is beyond human control but the latter is within human control. Hence, the second reading yields a more consistent interpretation of [2]. Clearing up the meaning of [2] is important because it sets the stage for the rest of Xunzi’s argument. The difference between these two readings of [2] is subtle but it affects how we interpret Xunzi. On the first reading, Xunzi would be making a trivial point. All he says in [2i] is just that having yu is in the category of life. Since it is common knowledge that yu pertains to living things, Xunzi is not offering anything new to his opponents. His opponents would have no trouble accepting the position that yu is in the category of life and maintain at the same time that yu should be eliminated or reduced. Hence, the first reading leaves us with difficulty because it is

78 !

Wang Nian-sun thinks that the words “life and death” are corrupt and emends the text to “the %#& device of xing (nature) " # $” so that it structurally parallels “the count of qing ” in [2ii]. I hesitate to emend the text as Wang did. Even if we read [2] as it is given, the second reading is still more plausible because it renders a more consistent interpretation of [2ii]. 79 Mei, p.247. 80 I owe this point to Jeffrey Riegel.

35 1. A Challenge to Antagonistic Interpretations not clear why exactly Xunzi thinks having yu does not pose a problem to order. If I am correct in following the second reading, then Xunzi is making a clear and definite point: the existence and amount of yu do not affect order because these are matters in other categories that are beyond human control. Life and death and feelings are phenomena that take place irrespective of human behaviour or action, or those of any other being. Analogously, the presence of yu is a natural phenomenon independent of human creation or artifice. Likewise, for Xunzi, yu is merely “there” irrespective of human creation. This is a direct rebuttal of the propositions that good order depends on eliminating yu or that order depends on reducing yu.

Line [3]

(i)Y&ಎಚY(ii) Ÿᒮ  (iii) &ಎಚ੪‘Ÿᒮ 

੪‘಴Y

(i) Yu does not depend on whether its obtainment is ke (possible/permissible), (ii) but the one who pursues (qiu zhe) follows what is ke. (iii) That yu does not depend on whether the obtainment of their objects is possible/permissible is received from Heaven. That the one who pursues follows ke is received from xin.

81 In [3], we are told that qiu Ÿ is affected by what is made possible/permitted (ke ) by xin. Since qiu in this context is discussed along with the obtainment of objects ಚ, we can infer that qiu here refers to the pursuit of objects of yu. This impression is

81 ( ' A survey of the usage of ke in the text shows that ke ' and ke yi seem to be used interchangeably to mean “possible” or “permissible” and it is unclear whether Xunzi draws a distinction between the two. Ke yi is sometimes used to mean “possible” as in “a petty person ke yi become a gentleman (junzi),” “a man on the street ke yi become a sage,” “a person ke yi travel on foot across the world” etc. In these examples, ke yi refers to what is logically or practically possible. Ke does not only refer to possibility, however. There are also instances wherein ke yi is used to mean “permissible,” such as “it is not ke yi for a ruler to be incautious” and “it is not ke yi to blame Heaven (tian).” These two senses of ke as “possible” and “permissible” are related since they both prevent certain actions from being carried out. It will suffice our present purpose to simply understand ke as a condition that makes the pursuit of objects of yu possible. See section 3.3 for further discussion of the way Xunzi understands ke in relation to xin’s choice.

36 1. A Challenge to Antagonistic Interpretations confirmed by passage 19.1a, which says that qiu arises because the objects of yu are not obtainable:

ࡥ„ྞ&ಚ&vŸ

Humans are born with yu. Yu but not obtain [], then it cannot be the case that there is no pursuit [qiu].82

Xunzi is saying that xin affects the pursuit of the objects of yu. The crucial question concerns the subject of qiu.83 If yu carries out qiu, then [3] is suggesting that the movement of yu is subject to the control of xin. If yu cannot carry out the act of qiu, then [3] is not concerned with the movement of yu. These two understandings of the agent of qiu will yield very different interpretations of the relation between xin and yu. On the former reading, if xin permits, then the act of pursuing will be carried out by yu; if xin does not permits, the movement of yu is blocked. AI interpreters would adopt this reading of [3]. The latter reading, on the other hand, only tells us that xin determines the act of pursuing itself. We are not able to infer from this information alone whether it means xin allows yu to carry out the act of pursuing. Most of the translations and commentaries of [3] have not clearly identified the agent of qiu. Li Di-sheng ³´„ interprets [3] as saying that yu depends on ke to form the behaviour of qiu, suggesting that yu pursues its objects. Yang Liang’s commentary, for example, vaguely suggests that yu automatically pursues (qiu) but its pursuit can follow the permission of xin. Liang Qixiong vaguely refers the agent of qiu to human beings. Ogyu Sorai’s µ„ಋ· commentary is one of the few that implies xin is the agent of qiu. According to Ogyu Sorai, [3i] and [3ii] mean that:

[i] For the yu of humans, xin does not ask whether its object can be obtained or not. [Line 2ii] discusses the pursuit of humans. Therefore, (…) necessarily pursues what can be obtained and does not pursue what cannot be obtained.84

82 Xunzi 19/1; K 19.1a/3/55. 83 I am indebted to Professor Kwong-loi Shun for alerting me to this point. 84 Ogyu Sorai, pp.276-7.

37 1. A Challenge to Antagonistic Interpretations

In my translation of the above quote, it becomes obvious that there is no subject in the second sentence. As far as we can tell, Ogyu Sorai’s interpretation relates xin to de instead of relating yu to de. It is likely that, for Ogyu Sorai, the subject of pursuit in [ii] refers back to the xin he discussed in [i]. The assumption that yu is the agent of qiu is problematic for three reasons. First, there is no outright textual evidence that shows yu and the agent who qiu (qiu zhe) are the same thing. In passage 22.5a, we can only infer that qiu refers to the pursuit of the object of yu. Even though passage 19.1a discusses qiu in close relation with yu, all we can infer is that there is a connection between yu and qiu. There is simply insufficient grounds for us to make the further claim that yu performs the act of qiu. Second, it is emphasized in [3i] and [3iii] that yu does not depend on what is possible/permissible (ke) but qiu depends on ke. The fact that Xunzi has differentiated yu and qiu suggests that there is a significant distinction between yu and qiu. It would be absurd for Xunzi to then make the contradictory claim in [3ii] that yu does not depend on whether the obtainment of its objects is ke but their actions to pursue their objects depend on ke. The major consideration against taking yu to be the agent of qiu comes from line [3ii]. It is clearly said that qiu follows (cong ) what is permissible. There are two possible interpretations of this sentence, depending on how we understand the subject of qiu. If yu is taken to be the agent of qiu, then [3ii] is saying that yu follows what is ke when pursing their objects. The AI interpreters would adopt this understanding. Another possibility, as suggested by Ogyū Sorai, is to take xin as the agent of qiu. In order to adjudicate between these views, we can turn to the term

“cong ” in this sentence. “Cong” is used verbally here to mean “to follow” or “to comply with.” The question relevant to our investigation then is to ask what kind of compliance the term “cong” denotes. Is the agent of qiu coerced to comply or does the agent comply voluntarily? Let us examine how the meanings of cong bear on these two interpretations. If yu is the subject of qiu, cong should denote a kind of coerced compliance. Since yu is a kind of natural and spontaneous response, there cannot be any considerations about whether its object can be pursued or not ([3i]) if yu is to pursue its objects. In the absence of intervention, yu will simply proceed and pursue its object. For yu to cong xin’s approval, it must be the case that yu is under coercion to

38 1. A Challenge to Antagonistic Interpretations follow xin. However, an examination of the usages of “cong” shows that the term is used to refer to a kind of deliberate following rather than coerced compliance. In the texts of the same period, there is no obvious suggestion that the term connotes coercion. In the Lunyu 2.4, Confucius reported that he was able to cong what his xin desires without overstepping the line at the age of seventy, implying that he was still not able to cong his xin without transgressing them even when he was aware of the boundaries when he was at a younger age.85 This suggests that cong is more of an act that arises from deliberation rather than mere compliance with external force. In the Xunzi, one passage discusses the difference between ministers who cong mandate and those who go against mandate.86 Another example is that the son of Heaven should not cong the yu of people.87 The occurrences of “cong” in these instances are used in the sense of deliberate or voluntary following. Passage 22.6a elaborates on what it means to cong what is ke and can therefore be read as a supplement to [3ii]:

দࡥ)& ॥w॥&˜)ᗆ& ˜ᒮྞ¹

ྞࡥº“പ»š¼ჽ^º&፪ᬻº¿»À૆

As a general rule, all people cong what they regard as ke and abandon what they regard as not ke. There is no instance of someone who understands that there is nothing to compare with dao and yet not cong dao. Consider the case of the person who likes travelling south, yu south, cannot be more; hate north, cannot be less. Surely he would not depart from south and travel north just because he could not cover all the southern ground!88

The first line of this passage is strikingly similar to [3ii]. It says that people follow what they ke. The subsequent example of someone who likes going south will travel south illustrates this point. The decision to travel south is a voluntary one made on the basis of personal preference. This suggests that cong should be a kind of deliberate following that involves thinking. In light of passage 22.6a, we can be more certain that “cong” in [3ii] cannot be a compulsory compliance but one that is deliberate and

85 Lunyu 2.4. 86 Xunzi 13/10; 13.2/2/199. 87 Xunzi 4/72; K 4.12/2/195. 88 Xunzi 22/67-9; trans., K 22.6a/3/136, modified.

39 1. A Challenge to Antagonistic Interpretations voluntary. And since it is said that the agent of qiu is one who cong, and since the agent who cong is capable of deliberation, it is unlikely that Xunzi would regard yu itself as the agent of qiu.

Line [4]

੪‘l’ ੪‘಴“஭ᬼ#੪‘

The one/singular/common yu, which is received from Heaven, is worked upon/regulated (zhig) by that which is received by the many of xin/made into many by xin, until it becomes difficult to categorise what is received from Heaven.

Xunzi is roughly saying here that the “the one yu,” which was originally received

g from tian, becomes difficult to categorise after it has gone through zhi ’. The ambiguity mainly surrounds how we understand the phrase “the many of xin (xin zhi џ).” But since this line might be corrupt,89 it is unclear what exactlyן duo “many” is. I follow the traditional reading that “the many of xin” refers to the many calculations and measurements of xin.90 In this sense, Xunzi would be saying that natural yu endowed by nature is being “zhig” by the many calculations and measurements of xin. What is left to explain is the term “zhig,” which can be loosely translated as “to control.” Unfortunately, there has not been enough scholarly attention given to provide a fuller explanation of what kind of regulation zhig denotes. There are different ways in which xin can be said to “control” yu that yield different interpretations of [4]. Xunzi could be using “zhig” to mean (i) xin limits the movement of yu, just like a prison guard oversees and checks the behaviour of prisoners. “Zhig” in [4] could also mean (ii) xin manages yu to make it more functional. This sense of zhig is like how a tailor manages a piece of cloth. The tailor works on the cloth in a way that takes into consideration the kind of material she has

89

There are two main editions of this line. In the Yuan Dynasty editions, text ji ) usually replaced duo

* *

. Liang Qixiong suspects that the original text should have both duo and ji ) as xin zhi duo ji (p.322). See Knoblock vol. 3, p.135 n 97. 90 *

Yang Liang emends text “duo” to “ji” ). Commentators like Zhang Shi-tong and Li Di-sheng generally followed this interpretation of Yang Liang.

40 1. A Challenge to Antagonistic Interpretations and the specific design she wants. Let us now determine which sense of regulation Xunzi is likely to be using. If Xunzi uses “zhig” in the first sense of regulation, then his emphasis is on xin’s control over the movement of yu. Guo Song-dao ÂÃÄ interprets [4] to mean that the one yu given at birth is quickly scattered around at the order of xin and gradually loses its genuine, old form so that it cannot be sought in the category to which it originally belonged. Guo seems to be making an implicit suggestion that for xin to zhig yu is for xin to give commands to yu. According to Gao, yu departs from its original position when it starts following xin. This means that its original position is one that xin does not approve.91 Similar accounts can be found in Waston’s translation of [4], which understands “zhig” in terms of “to direct action”:

Thus a single [yu] which has sprung from the inborn nature may be directed and controlled in many ways by [xin], until it becomes difficult to identify it with the original [yu].92

By saying that yu is “directed” by xin, Watson’s translation suggests that yu is initially going to act in a way that is unapproved by xin. In the second sense of regulation, “zhig” refers to how xin manages or works out yu in a certain way. This is different from the first interpretation because it does not presuppose that yu has a predetermined orientation that is different from xin and hence does not attribute a causal role to yu in action. Li Di-sheng’s commentary on [4] hints at this interpretation. According to Li, there is one common yu for everyone. But since xin has many considerations and calculations, it has to decide what can be the objects of yu and, among these objects, what are the weightier and more urgent ones. Hence, there is a divergence between action and yu. Although Li’s explanation is not entirely clear, his interpretation is significantly different from AI in that it does not presuppose yu takes bad actions and needs to be controlled by xin. Li also calls our attention to the difference between yu and objects of yu and the difference between yu and action. He seems to be suggesting that for Xunzi, objects of yu and action are not in the same category as yu. Then, for xin to zhig yu is for xin to work out

91 Wang Xianqian, pp.427-8. 92 Watson, p.155.

41 1. A Challenge to Antagonistic Interpretations the ways that it can satisfy yu by determining what objects and how much of them can be pursued. In order to adjudicate between these two interpretations, we can examine the connotation of “zhig.” If “zhig” has the connotation of controlling or limiting certain actions, then it supports the first interpretation; if “zhig” only emphasises working out things in a certain way, then the second interpretation is more plausible. One textual observation about the usages of “zhig” is that the object of zhig is usually subject to control, as in the case of “ruler zhig minsters,”93 or “ruler zhig the people.”94 This might create the impression that to zhig something is to be in a position of controlling action, as in rulers controlling the actions of ministers. Nonetheless, the observation that zhig is a relation from superior to inferior is alone insufficient to show that zhig is concerned with regulating action as such. It is also plausible to explain these occurrences of zhig in the sense that the ruler allocates duties to his ministers.95 The second reading is more plausible because “zhig” is sometimes used as a verb when discussed along with Å (legal punishments) and du ి (standard measurements),96 as in the sages are making or working out protocols and standards. Furthermore, when zhig is used as a verb, the object of zhig does not always have anything to do with being checked. For instance, there are numerous references to “zhig names” and “zhig rituals and music” in the Xunzi. 97 When zhig is used in relation to names, rituals (lib ) and music, the term is used verbally to mean improvising or making. In other early texts, zhig is also used in a non-violent sense to denote the activity of working out something. In the Mozi, for example, we find passages that speak of employing good tailors when a particular piece of clothing cannot be zhig.98 There are references to zhig land and zhi properties in the Mencius.99 These usages suggest that zhig refers to the process of allocating or working out an undifferentiated piece so as to make it functional.

93 Xunzi ; K 10.1/2/120. 94 Xunzi 10/31; K 10.5/2/125. 95 I am indebted to Jeffrey Riegel for helping me think through this point. 96 E.g. Xunzi 8/97, Xunzi 82/21/83, Xunzi 87/23/28. 97 E.g. Xunzi 22/7, 27/130. 98 Mozi 9/22/10. 99 Mencius 1A:7 and 7A:22.

42 1. A Challenge to Antagonistic Interpretations

The second reading of zhi is also consistent with Xunzi’s view that humans

g 100 should zhi the (’Ç). Since Xunzi cannot emphasise enough the idea that something received from Heaven is beyond the control of human beings, it is absurd for him to claim that humans can control the action of tian. On the second reading, “zhig the mandate of Heaven” means that humans can work out the mandate of Heaven to make it functional for humans. This reading is in line with the general theme of the Tian Lun chapter. As the above textual observations suggest, zhig should be more suitably understood as working out something instead of controlling the action of something. Understanding zhig as the act of working out things instead of controlling things can better explain the “one yu” in [4]. The very notion of working out implies that the object before zhi is undifferentiated or unworked on. In this sense, “one yu” simply means that yu is natural and has not yet been worked on. If the second reading more closely captures Xunzi’s position, then [4] is saying that undifferentiated yu is worked out by the many calculations of xin. There is no insinuation that because yu is heading towards its own direction, its action needs to be checked. The emphasis of “zhig” is on the process of working out certain things. Hence [4] cannot be a basis for AI interpreters to argue that xin controls the movement of yu.

Line [1]

ÈÉÊYদ—–ಎwᒮࡾ˜™ ྞᒮYÈÉÉÊYদ—–ಎšᒮ

ࡾ™ “ᒮY

(i) All those who maintain “order depends on the elimination (qu) of yu” are those who lack the means to dao yu and are therefore troubled by the fact that yu exists at all. (ii) All those who maintain “order depends on the reduction (gua) of yu” are those who lack the means to jie yu and are therefore troubled by the fact that yu are many.

100 Xunzi 17/44; K 17.10/3/21.

43 1. A Challenge to Antagonistic Interpretations

Here, Xunzi is contesting the propositions that order is contingent upon the elimination (qu w) or reduction (gua š) of yu. Let us focus on the notion of jie yu first.“Jie ” can be loosely translated as “to regulate.” Since sholars who advocate AI endorse the view that xin regulates the movement of yu, their reading of jie is along the line that implies the causal force of yu is checked. Masayuki Sato, for example, understands Xunzi’s conception of jie in relation to how xin should control yu. He writes:

Since Xun Zi emphasized the role of volitional mind/heart in behaviour, it was not difficult for him to incorporate the concept of jie into his argument that human desire should be controlled (by volitional mind/heart)…In this context as far as the concept of jie is concerned, human desire should be controlled in an oppressive way, particularly if the will (of a person’s mind/heart) has decided to victimize the physical body in order to live up to the morality of that person.101

Implied in Sato’s account is the assumption that yu by itself will lead to actions that are not in line with morality. By controlling yu, xin can cause actions that are in line with morality. Sato’s interpretation encapsulates the core of an antagonistic interpretation of jie yu. It is especially true for path-blocking interpreters because they believe that xin suppresses yu. The path-re-steering interpreters would object to Sato’s claim that xin controls yu in an “oppressive” way, arguing that xin can transforms yu in a positive way. Nonetheless, path-re-steering interpreters would still endorse the basic claim that xin controls the movement of yu. By understanding the movement of yu as being obstructed or redirected by xin, it is inevitable that a rather aggressive notion has to be incorporated into the meaning of jie. Since it is presumed that there is a conflict between xin and yu, for xin to jie yu would mean for xin to forcefully counteract yu. On this reading, the term “jie” takes on a violent connotation. An examination of the usages of jie in early texts shows that jie has the

b connotation of proportioning or segmenting. In the Mencius, the content of li Ë is

101 Sato, 2004, pp.377-8

44 1. A Challenge to Antagonistic Interpretations defined in terms of the jie and adornment of and yi.102 Since Mencius’s project is one that encourages people to extend ren and yi, jie in this context is more likely to be about adjusting the proper balance between ren and yi. In the , there is one passage that talks about suitably jie eating and thinking. There is no suggestion that the act of eating or thinking should be aggressively controlled. Rather, it is only saying that one should not overeat or overthink.103 In another passage, there is reference to jie yu. It is said that the reason this balance is lost is due to human feelings, such as joy and anger, losing their measure. When one jie the five yu, joy and anger will not lose their measure. And when joy and anger lose their measure, one should make a plan Ì, presumably about how to manage joy and anger within a suitable degree.104 The rest of the passage talks about the disadvantages of losing the balance of things and the benefits of restoring the balance and alignment of life. We can infer from this that jie yu is about dealing with the feelings in a way that it does not disturb a balance. Another Guanzi passage that corresponds to this idea of managing feeling is in “Xin Shu II,” which conveys the point that music can jie anger while rituals can jie music.105 It is more reasonable to understand the author as saying that rituals can coordinate the performance of music than saying that music should be violently stopped or reduced by rituals. Then, for music to jie anger should also be about how music coordinates or manages anger. There is no clear example in the Guanzi that shows jie concerns violent reduction of an action. In the Lushi chunqiu, a connection is made between shi  (fit, to make suitable) yu and jie one’s nature. The point is that the sages do not dwell in what is excessive in nurturing their nature.106 We can then infer that, in the case of Lushi chunqiu, well-performed jie will result in harmony and balance. Our understanding of jie can be further illuminated by a passage in the Zuozhuan, in which Physician He tells the Duke of Jin, who is suffering from excessive contact with women at that time, to jie his sexual activities.107 The physician then uses an analogy of musical performance called five jie 8 to explain

102 Mencius 4A:27. 103 Guanzi 49/272/1-2. 104 Guanzi 49/271/5. 105 Guanzi 37/223/15. 106 Lushi chunqiu 1/3.2. 107 I am indebted to Jeffrey Riegel for alerting me to this passage and for allowing me to consult his paper “The Zuozhuan Lecture of Physician He on Passion and Disease.”

45 1. A Challenge to Antagonistic Interpretations proper sexual practices to the ruler.108 According to Physician He, the speed and sequence that the five musical notes occur should be adjusted in relation to one another. Once the notes are in tune, the sounds are harmoniously blended together. Music should then stop at this point because anything beyond the optimal sound would lead to extraneous moves and excessive sounds. Since the five notes can only be stopped when the harmonious sound is played, it is important for the person who performs jie to know when to stop. The question of when to stop is explained by Physician He’s further comments about the relation between five jie and the six energies 2 of Heaven. The five jie is defined by Physical He as the sequence of the six energies, which are manifested in the five colours, the five tastes, the five notes, and the four seasons. This means that five jie must embody certain cosmological pattern. Jie, then, is a form of order that is done by setting the suitable proportions for things in a way that corresponds to a greater cosmological pattern. In light of these observations about the meaning of jie in early texts, let us

a return to the Xunzi. In Xunzi 16.8, it is said that yi ᑶ (propriety) is jie the person within and jie the thousand things without.109 Yang Liang takes the term “jie” in this passage to mean “limit” or “restrain,” whereas Wang Xianqian takes it to mean “fitting” or “suitable.” Moreover, the next sentence says that a person who jie will

“produce peace Δ for the ruler above and create a “fine-tuned balance 4” for those below.110 This understanding of jie coheres with the meaning of jie we discussed earlier. Hence, Wang Xianqian’s interpretation is more plausible because the emphasis of jie is not on the act of stopping or reducing something but on setting the suitable proportions for things so that each can have optimal functioning in the overall balance of things. Since the person who performs jie must be able to cope with other things and align with nature, this explains why Xunzi seems to see a correlation among the abilities to be at ease (Î) with performing jie on rituals, to readily respond to changing circumstances, to accomplish things when time demands, and to

108 ,-.

Zuozhuan Zhaogong + /1221-3. I am indebted to Jeffrey Riegel for alerting me to this passage in the Zuozhuan. 109 Xunzi 17/78; K 16.8/2/248. 110 Xunzi 16/78-9.

46 1. A Challenge to Antagonistic Interpretations harmonise the common people.111 In fact, Sato himself has conducted a detailed examination of the meaning of jie and concluded, “Jie denotes a balanced condition, an appropriate mixture of nature’s ingredients.”112 But as I mentioned earlier, Sato’s understanding of jie in passage 22.5a of the Xunzi reads an aggressive notion into the term that is not found in his own observation about the meaning of the term. Sato also examines the notion of jie in the Lushi chunqiu, for he regards the Lushi chunqiu and the Xunzi as sharing their main ideas.113 As we saw, the emphasis of jie is more on balanced proportion rather than reduction. An example can help illustrate the difference. If a person says she is going to reduce her sugar intake, then she means she will consume less sugar than she used to. In contrast, if a person says she is going to jie her sugar intake, she only means that she will consume a suitable portion that is in proportion to other nutrients (supposing that there is certain balanced diet standard). She is not necessarily going to consume less sugar than before. It is even possible that she will consume more if the standard for balanced diet requires. If “jie” in [1ii] is not concerned with violently holding back the movement of yu, we have one more reason to challenege the assumptions beind AI. Xunzi is more likely to be saying in [1ii] that xin can set the right proportions of yu such that no excessive yu will occur to cause imbalances and disrupt the order of things. Let us now return to [1i]. Xunzi maintains that order depends on dao yu and jie yu. Dao is used verbally here to mean channelling or guiding. Again, this notion of channelling is open to interpretation. It could mean that (1) xin can block yu and thereby determine when yu is allowed to go through (path-blocking); or that (2) yu is initially travelling in a wrong direction and needs to be directed by xin (path-re- steering); or (3), in the straightforward sense, yu needs to be guided. On the first two interpretations, xin guides the movement of yu. In other words, there are two causal roles involved in causing action. The third interpretation, on the other hand, does not assume outright that there are competing causal forces in action invovled. As our analyses of [3], [4], and [1i] have shown, we have no basis at this point to attribute

111 Xunzi 8/60; K 8.7/2/76. Another example of the relation between jie and an can be found in Xunzi 75/19/103; K 19.9c/3/70. 112 Sato, 2000, p.21. 113 Although Sato wants to show that jie in the Lushi chunqiu means moderation of desire, it is not entirely clear what he means by “moderation.” Ibid.

47 1. A Challenge to Antagonistic Interpretations separate causal roles in action to both xin and yu. I will therefore opt for the third interpretation.

Line [6]

Q¤਄&੣಴§಴¨቗ᬵ“©प –&੣਄

¤಴«಴¬቗ᬵš©§ ࠿

The reason yu are excessively intense but action does not reach the point is that xin stops it. If what xin makes possible/permits is in line with lia, then although yu are many, how could it harm order? The reason yu does not reach a point as intense as action is that xin causes it. If what xin makes possible/permits loses lia, then although yu are few, how could it stop disorder?

It is further expressed in [6] that the strength of yu is irrelevant to action. AI interpreters would understand [6] as saying xin stops the movement of yu; however, for both the phrases “xin causes it” and “xin stops it,” the object pronoun “it (zhi )” should refer back to action, rather than yu. Xunzi’s point seems to be that xin can cause action instead of xin controlling yu. This reading is supported by [5], which says that what human yu most is life, and what they hate most is death and yet there are some who follow life and end up with death. It is not that they do not yu death, but that the action of pursuing of life is not made possible by xin ([5]). Since the act of pursuing the objects of yu can only be made possible by xin, this explains why Xunzi believes that order and disorder arise as a result of whether the object of yu is pursued. This understanding can help us make better sense of Xunzi’s claim that the source of order or disorder lies in xin ([7]). In Chapter 3, I will show that the action that xin is putting a stop to is the action that xin would have first undertook when it operates on an instinctual level. Passage 22.5a is an important clue to our understanding of Xunzi. Although my analysis of this passage alone cannot supply a full picture of Xunzi’s understanding of the relation between the concepts xin and yu, it has enabled me to challenge the tenuous assumptions of AI about the relation between the concepts xin

48 1. A Challenge to Antagonistic Interpretations and yu and thus clear the ground to better approximate Xunzi’s view. So far, our analysis has been framed in negative terms. In the following chapter, I will draw on relevant passages to reconstruct Xunzi’s picture of the relation between the concepts xin and yu.114

114 I am deeply grateful to Professor Jeffrey Riegel, whose comments on passage 22.5a and suggestions have helped me think through numerous important issues in this passage.

49 CHAPTER 2

Xin and Yu

2.1 Making Choices

In this chapter, I will argue that, for Xunzi, only xin can decide on the course of action to choose. For this reason, the issue of whether yu will conflict with xin does not even arise. This implies that yu itself cannot cause action but does not preclude the view that yu could be a factor in xin’s decision. I will then discuss the implications of the proposed interpretation on our understanding of Xunzi’s concept of xin. In the Xunzi, it is said that the pursuit of the objects of yu can be limited by lu

Ï:

ᬵ&wŸ&ಚÏᒮŸ

Although yu cannot be eliminated, when what is pursued is not obtainable, lu zhe would like to moderate pursuit (jie qiu).1

Xunzi defines lu in terms of the activities of xin. In view of this connection, the concept of lu should be a good place to start our discussion of the relation between xin and yu. A literal translation of Xunzi’s definition of lu is like the following:

[1] Z[പSTUV*ചച಴ჽຬ*Ï

The feelings of like and dislike, of delight and anger, and of sorrow and joy that one has birth are called qing; qing being so and xin wei zhi (for the sake of qing/to this end/because of it) chooses. This is called lu.2

a Although lu is defined in terms of how xin relates to qing ച , if we recall that yu is defined as “a response of qinga,”3 we come to see a linkage between xin and yu

1 Xunzi 22/66; K 22.5b/3/136. 2 Xunzi 22/3; K 22.1b/3/127. See discussion in section 2.2 for analysis of this quote. 3 Xunzi 22/63; K 22.5b/3/136. 2. Xin and Yu embedded in the concept lu. This leads us to ask what has been implied about the relation between xin and yu in this succinct definition. “Lu” is often rendered into English as “thinking,” “deliberation,” and “reflection.” In the literature, Xunzi’s conception of lu is often overlooked. Scholars usually take Xunzi’s definition of lu as evidence for the cognitive capacity of xin. Most of the time, lu is assumed to be about the way xin chooses feelings. My following analysis of [1] intends to expose the problems with such a reading of Xunzi. The major English translations of [1] are:

Dubs: the mind selects from among the emotions by which it is moved – this is called reflection.4

Watson: When the emotions are aroused and the mind makes a choice from among them, this is called thought.5

Knoblock: The emotions being so paired, the mind’s choosing between them is called “thinking.”6

While the translators’ renderings of the term “lu” vary, they are more or less in

7 agreement that ze ຬ is an act of xin choosing the feelings. Corollary to the view that some feelings must have been regarded by xin as more appropriate than others, scholars tend to make the inference that xin is trying to approve or disapprove yu. This implies that yu is ultimately what causes action. Michael Puett translates [1] quite differently:

The qing being thus, the mind makes them choose.8

Puett cites [1] to show that Xunzi “clearly prioritizes mind over qing.”9 Puett might have regarded this as an evident point and therefore has not gone into detail to explain

4 Dubs, 1966b, p.281. 5 Watson, p.144. 6 Knoblock 22.1b/3/127. 7 Duyvendak and Y. P. Mei also understand “ze” in [1] as choosing among feelings (Duyvendak, 1924, p.225; Mei, p.54). 8 Puett, 2004, p.53.

51 2. Xin and Yu his observation. According to Puett’s reading, it is qing rather than the mind that carries out the choosing. This suggests even more strongly that yu can by itself cause action. For convenience, I will call the reading that takes qinga to be the object of xin’s choice a “common reading” of [1]. My reservation with the common reading of [1] is that it heavily resembles the dichotomous framework between reason and emotion in a way that does not do justice to Xunzi’s concept of lu. Typically when there is a tension between reason and emotion, it is emotion that has to be brought under the control of reason. Since Xunzi’s concept of lu is in some way congenial to such an understanding, it is tempting for us to subsume Xunzi’s understanding of lu under our contemporary understanding of deliberation. Janghee Lee, for instance, follows this line of reading and understands lu as xin choosing the proper emotional state. Lee briefly explains Xunzi’s lu by way of an example:

When we see someone slip and fall on icy ground, we tend to laugh without thinking of his or her pain. Once we put ourselves in his or her shoes, we stop laughing and worry about his or her condition. This process of thinking, which discerns one’s emotional state and chooses the proper one, is the crucial part of critical reasoning that constitutes the autonomous function of xin.10

This example reveals Lee’s assumption that emotional states alone cause one to act. Suppose the feeling of humour will cause us to laugh at the person who falls on the ground, whereas the feeling of sympathy will cause us to help that person. On Lee’s account, the role of xin would be to distinguish feeling of humour from feeling of sympathy and, in this scenario, choose the latter. Once the feeling of sympathy is chosen by xin, the feeling can move us to help the person in need. Siu-fu Tang also understands lu in [1] as xin’s choosing among the identified feelings. Based on this reading, Tang argues that since xin makes judgements by approving or disapproving, lu is the act of approving or disapproving certain kinds of yu.11 In cases where an inappropriate yu is approved, Tang thinks that it is because xin

9 Ibid.

10 Lee, p.41.

8 ?@ A DE D G HI

/ 0 1 2 3 6 7 ( 3>

; < = BC

11 F

4 5 % 9 : % 9 J(Tang, 2008a, p.14).

52 2. Xin and Yu is “dominated” or “misled” by yu.12 The language Tang uses to describe yu strongly suggests that yu in and of itself constitutes another causal force in action, which competes against xin. It is just that yu loses most of the time. In view of my earlier discussion about there being insufficient textual evidence for us to ascribe to yu a causal role in action, I hesitate to take the common translation of [1] as granted. My proposed reading is to take “wei zhi” in [1] to mean “for this” or by extension, “in light of this.” The line is rendered:

Qinga is so and because of this, xin chooses. This is called “deliberation.”

On the proposed reading, [1] is saying that xin makes a choice because of the feelings. This differs significantly from the common reading, which understands lu as xin’s choosing among feelings. On the latter reading, feelings are the object of choice. The former reading, on the contrary, does not assume that the feelings are objects of choice. It only understands Xunzi as saying that the action of choosing (ze) takes place because of qing. The question as to what the object of choice is for xin is left open. Some examples will help illustrate what it means to make a choice because of X but the choice itself does not necessarily involve X. Suppose Jane has just been offered a position in an offshore branch of her organisation. Effectively, this is a promotion for which she has been waiting for a long time. However, moving offshore would entail moving offshore without her familiy as well as being away from her friends. In view of these considerations, Jane turns down the offer. An outsider can describe Jane’s situation as, “Given her family and friends, Jane makes a choice.” On the common reading of [1], this will read Jane is choosing between her family and friends, which is clearly not the case. On the proposed reason, the description about Jane means that her family and friends are the reasons for Jane having to make a choice about her job. In speaking of her own situation, Jane may say, “I chose my family and friends.” This may make it sound as if the objects of Jane’s choices were (a) her family and friends; and (b) her work. However, the objects of her choice in this case are between her two work choices: to stay in her current position or to take up the new offshore position. Her reasons for decision include, among others, her desire to be with her family and friends. In a

12 Tang, 2008b, pp.17, 21.

53 2. Xin and Yu similar vein, my reading of [1] takes qing to be one of the reasons for xin’s choice, not the object of xin’s choice. By identifying qing as a reason for xin’s choice, the proposed translation does not warrant the conclusion that xin is serving the interests of qing or yu. Eric Hutton translates [1] as the following:

When there is a certain disposition and the heart makes a choice on its behalf, this is called “deliberation.”13

There is a subtle yet significant difference between Hutton’s translation and the common reading of [1] in that Hutton does not suppose that the feelings (qing) are an object of choice. It only says that xin makes choices. Hutton’s translation suggests another way of understanding [1]. However, it presupposes that xin acts in the interest of qinga when it chooses. In order to avoid conflating the question of how xin and yu are related and that of what is the basis of xin’s evaluation, I doubt whether it is necessary to assume that “wei ჽ” means “for the sake of” or “on behalf of.” The proposed reading is not committed to such a position. We could well imagine that Jane has accepted her offshore position even though she knows full well that taking the new job will not serve the interest of her family and friends. Again, a number of considerations could have been factored in Jane’s decision and, in the case of Jane deciding to relocate, her family and friends do not become the reason for her choice. The rationale behind her decision does not concern us here. The point is that while Jane’s family and friends can be a reason for her choice, the choice itself does not necessarily serve the interests of her husband and children. Similarly, on the proposed reading, all we are told in [1] is that feelings are reasons that xin chooses, but this cannot be taken to mean that xin is acting on behalf of feelings or yu. And we cannot infer from [1] alone what xin’s motivation is when it chooses, for there is a difference between the reason (X) for xin’s choice and the reason xin has X as its object of choice. My analysis thus far has only shown that we cannot make any hasty inference about the underlying rationale of xin’s choice from [1]. It does not preclude the possibility that xin indeed tries to satisfy yu for Xunzi. I will reserve this subject matter for discussion in Chapter 4.

13 Hutton, 2005, p.292.

54 2. Xin and Yu

I defend the proposed reading on three main grounds. First, it keeps the meaning of [1] close to the original text with minimal alteration or unnecessary assumptions. In addition, given Xunzi’s repeated stress on the point that feelings themselves are neither right nor wrong, it is unlikely that he thinks feelings themselves need to be approved or disapproved. It leaves open questions regarding the object and basis of choice. Second, the proposed reading avoids the kind of tension that is present in the common reading of [1]. It does not understand ze as about choosing particular feelings over other feelings. Granted that Xunzi does mean by lu that xin is choosing among feelings, as the common reading has understood, it is difficult to see how it works on a practical level. The very notion of choice (ze) implies that certain feelings in the set of qing cannot be obtained at the same time because some feelings are mutually exclusive. For example, it is difficult to imagine how the feeling of happiness (S) and the feeling of anger (T) can be chosen at the same time. This point is especially reinforced by Knoblock’s translation, which suggests that feelings come in pairs: “The feelings being so paired, the mind’s choosing between them is called ‘thinking.’14 On Knoblock’s reading, it would be impossible for xin to choose both the feeling of happiness and the feeling of anger if these two feelings come as a pair. However, the idea that for xin to choose qinga is for xin to choose either happiness or anger runs against our intuition. There are numerous situations in life where a decision involves a complex set of feelings, rather than one particular feeling. Suppose I am on a diet but have a craving for sweets. I know that eating a chocolate bar will make me happy but at the same time also make me angry with myself because I broke my diet plan. If I proceed to eat the chocolate bar, it is difficult to see which feeling xin has chosen. One might say that there must be a stronger feeling that outweighs the weaker feeling, that is, xin chooses the stronger happy feeling as opposed to the weaker angry feeling.15 But if that is the case, it is actually inaccurate to say that xin chooses “happiness.” By choosing “happiness,” xin has not eliminated “anger” or “sadness.” Suppose xin chooses the happy feeling, the very decision has already included the angry feeling. When I proceed to eat the chocolate bar, there will

14 K 22.1b/3/127. 15 This touches upon the question as to whether all feelings are ultimately commensurable for Xunzi. This does not concern us here because my aim is to simply show that ze is not about xin’s choosing of a particular feeling.

55 2. Xin and Yu still be an angry feeling entailed by my choice. On the common reading of [1], there would be a tension in Xunzi’s thought. The proposed reading avoids this problem. It understands Xunzi as saying that given the different feelings involved in eating the chocolate bar, xin makes a choice about eating it or not. My third argument against the common reading draws on a related passage we mentioned before:

Ñ॥ಶÒ$Ï॥പᒮ

Ñ॥+ಶÒ$Ï॥zᒮ

ÓÔÕÖ$×॥പ੩Ù

See something that can be yu, must comprehensively lu what can be detestable of it. See something that is beneficial, must comprehensively lu repeatedly what can be harmful with it. Weigh the two together and then carefully calculate it and decide whether to choose or give up in regards to yu and dislike.16

The above quote is often quoted in association with [1] to explain that feelings are the object of xin’s choice. A close examination of the passage, however, shows that its ideas are more consistent with my proposed reading than with the common reading. At first glance, the above quote seems to suggest that lu is concerned with making a choice among feelings, for it talks about what should be taken or given up on the basis of being likeable or detestable. However, if we read the text carefully as a whole, we can make two observations that challenge the common reading of [1]. First, what is to be chosen is the object of yu, not yu itself. There is a marked difference between what is desirable and the desire itself. In this case, the choice concerns the former rather than the latter. Second, according to the quote above, lu refers to the (person’s) grasp of the range of possibilities and implications resulting from the (person’s) pursuit of the specific object and deliberating whether the object in question will elicit other unpleasant feelings. This implies that certain feelings are not aroused at the time of deliberation. Hence, it is unlikely that lu involves choosing among feelings that are already aroused. The common reading of [1], however, understands the arousal of

16 Xunzi 3/46-7; K 3.13/1/180.

56 2. Xin and Yu feeling as occurring prior to deliberation as, for example, suggested in Watson’s translation.17 What is being conveyed is the idea that feelings are the reason xin has to choose. It is said that when an object that elicits positive feelings comes into sight, one has to think about the negative feelings that the same object might also elicit. It is by weighing and calculating these two possibilities together that a choice can be made. This shows that, for Xunzi, the agent does not necessarily have certain feelings that he would have had if the agent were to lu. This is consistent with my proposed reading, which understands Xunzi as saying that feelings are reasons for xin having to choose.18 The proposed interpretation also understands xin as capable of taking feelings, whether positive or negative, into consideration when it chooses a course of action. However, it contests the idea that lu is about xin rising above feelings and deciding whether certain feelings should be denied or allowed. On the proposed interpretation, lu is about xin having to make a choice because of feelings. What this seems to imply is that there is a range of possibilities for xin to choose from and that each possibility is associated with feelings. In the simple chocolate-craving scenario we considered, for example, xin can choose from a range of possibilities: I can eat or not eat; if I eat, I can eat a little or eat a lot. If I eat, I will be very happy and very angry at the same time; if I eat a little, I might be slightly happy but still very angry. In this sense, feelings constitute the background of possibilities for xin’s consideration.19 The proposed reading is compatible with the view that feelings are what differentiate one possibility from another. What it does not assume is the view that xin is choosing among different feelings so that the chosen feeling itself can give rise to action. The idea that lu involves xin entertaining the various possibilities can be observed in many classical texts. For example, it is said that when one is in a state of fullness, one should lu the possibility of insufficiency and that one should lu the possibility of danger when one is in a state of security.20 In the Zuozhuan, it is said that one should lu about the possible ways one’s fortune can be spent. 21 In the Guoyu,

17 “When the emotions are aroused and the mind makes a choice from among them is called thought” (Waton, p.144). 18 Since our present concern is the relation between xin and yu rather than the basis of xin’s choice, I will not discuss here whether it is suggested in the above quote that xin would choose a course of action that yields the greatest degree of happiness. I will examine this issue in Chapter 4. 19 I am grateful to Professor Jeffrey Riegel for helping me think through this point.

20 Xunzi 7/24.

,KL . 21 M

Zuozhuan Zhaogong + /1519.

57 2. Xin and Yu the end of lu results in no more changing one’s course of actions.22 Furthermore, the

b b term “lu” is often paired with “si Ú,” where the term “si ,” in turn, is associated with

23 rong Û, meaning deep and therefore a large capacity to hold things. Interestingly, being deep is also linked with choosing in the Xunzi. For instance, it is said that xin is deep (rong) and there are no limits on what it can choose24 and that xin does not make careful choices when its lu is not deep.25 This further suggests that lu is about taking into account a wide range of possibilities that are brought to light by sib. In light of the observation that lu has the connotation of entertaining possibilities in classical usages as well as in Xunzi’s own definition of lu, we can extract two important points about Xunzi’s understanding of lu. First, lu is a process of xin trying to entertain different possiblibities of choices. This requires xin to comprehend a bigger picture, an implication that is encapsulated in the expression “back and forth (qian hou Ò$),” meaning a comprehensive view of things. Second, the peculiar set of feelings that is entailed by each possible choice can be what differentiates one possibility from another for xin’s considerations. If this choice is not about the range of different feelings xin can choose from, what is this choice about then? If we recall the quote about weighing options, we find grounds to suppose that lu refers to xin choosing between different courses of action.

Qࡥ਄&ࡾ&ᕸÔࣨÞ&ßOà áࡥࡾჽNNà âࡥ

ࡾჽOãࡥࡾഠ NOÔ&ßåæ ࡥࡾჽççæ പࡥ

ࡾჽåãࡗࡥࡾഠ å炘ᒮੱ࡯ßÔ‚ᬻ˜êᕧຬ{ì

ࡾíîl©ಚ॥ïðటîlòóôჽ& ॥ໟY

Thus, when humans act, they cannot not weigh on some instrument. If the beam of the balance is not properly aligned, a heavier object may rise up so

PQ

22 O N Guoyu Wuyu N 19/6/547. It is interesting to note that commentator Wu Zeng-qi understands not having lu in terms of the xin being settled. 23 My understanding of sib is helped by Arthur Waley’s discussion of the origin of the word sib. According to Waley, “si-ing was an observation-post in the market, from which the overseer could observe which stall-holders were cheating” (Waley, p.44). I am indebted to Professor Kwong-loi Shun for alerting me to this passage. 24 Xunzi 21/46-47. 25 Xunzi 4/32.

58 2. Xin and Yu

that people will think it light and a lighter object may sink down so people will think it heavy. This is why people become deluded about light and heavy. If the counterweight is not right, then misfortune may be entrusted to yu and deemed as fortune and good fortune may be entrusted to dislike and deemed as misfortune. This is the reason people become deluded about fortune and misfortune. Dao, from antiquity to the present, has been the properly aligned balance…To abandon Dao and chooses by one’s own standards from within is like exchanging two for one. How could there be gain? Anyone who would actually exchange the yu that has accumulated for a hundred years for the dislike of a single moment does not understand the count.26

The first half of the passage talks about the importance of weighing in accordance with Dao before moving (dong ਄). Towards the end of the passage, it is said that if one chooses (ze ຬ) in a way that disregards Dao, then one will not be weighing accurately. I assume that the “ze” in this passage is used in the same sense as the “ze” in [1] and therefore understand Xunzi as saying that the choice xin makes will give rise to a course of action.27 Further support for the view that lu is about xin choosing a course of action comes from the observation that when Xunzi discusses xin, his focus is often on how xin causes action. In describing xin as the ruler of the body, Xunzi says that:

಴ᒮ  হࡽ੪ࡽᕧõᕧ«ᕧöᕧ੩

ᕧ¿ᕧ§

Xin is ruler (jun ) of the body and the master (zhu ) of the spiritual

intelligence. It issues orders but does not receive orders. On its own it prohibits or causes, renounces or selects, acts or stops.28

26 Xunzi 22 /71-8; K 22.6b-c/3/137. 27 Chad Hansen raises the idea that lu is about xin choosing a course of action but has not gone into detail to explain why he takes action to be the outcome of lu (Hansen: 320). Hansen has not drawn any implications on his reading of lu on the relation between xin and yu because the subject matter does not concerns Hansen in his discussion of lu. 28 Xunzi 21/44-5; K 21.6a/3/105.

59 2. Xin and Yu

Here, Xunzi is probably talking about how xin orders the body to move in a certain way. Moreover, the six verbs that are used to describe xin are all concerned with the act of causing or stopping action.29 Consider also the passage:

ࡥ„¡¢ࡥപ›¡¢ࡥྞ „£›ᒮœ&„›&

ࡾ„ࡾ›Q¤਄&੣಴§‚&੣਄¤಴«

Y

What human beings yu most is life, what humans hate most is death. Even so, there are some who follow life and end up with death. It is not that they do not yu life and they yu death. [In this case], it is not possible to live but possible to die. The reason yu is excessively intense but action does not reach the point is that xin stops it…The reason yu does not reach a point as intense as action is that xin causes it.30

According to the above passage, xin causes (shi «) someone to stop a course of action that pursues life or cause someone to take a course of action that pursues death. It is evident that when Xunzi talks about the importance of xin, the emphasis is predominantly on xin’s ability to cause action. In light of this observation and our earlier discussion about the difficulties involved in understanding lu as xin choosing a particular feeling, it is more plausible for us to understand lu as xin choosing a course of action. Elsewhere, it is said that if a person’s lu is not clear, her action will become problematic:

29

Out of these six terms, it is probably least obvious that “jin R (prohibit/forbid)” is concerned with action taking. An examination of the usages of suggests that jin is largely concerned with actions. In the Xunzi, for example, there are references to jin picking up objects that have been left behind (Xunzi 27/127; K 27.109/3/235), jin aggressive behaviour (Xunzi 27/40; K 9.8/2/99), and jin attending to what is unimportant (Xunzi 12/111; K 12.11/2/191). Other texts of the same period also use jin in the sense of forbidding certain actions such as jin killing people (e.g. Lushi chuqiu 1/5.5; Mozi 28/137/4), jin farming (Guanzi 73/361/8; Mozi 25/108/16), jin cutting wood (Guanzi 41/143/3; Huainanzi 5/70/4) etc. The impression that jin is mainly concerned with actions is further confirmed by the frequent association between jin and zhib S (stop, halt). There seems to be a general assumption that if X is jin,

then X will be zhib, which is reflected by the frequent use of jin zhib.29 As a combination, jin and zhib it

U U V is often contrasted with ling T and xing and also xing and shi (e.g. Mencius 1B:16; Guanzi 1/2/4, 12). 30 Xunzi 22/59-61; K 22.5a/3/135.

60 2. Xin and Yu

দc÷ྞዔ¨಴&×ù÷&࿾úÏ&࿾û×üYদc÷ྞዔ¨಴

&×ù÷&࿾úÏ&࿾û×üংং¿ᒮÑþࡾჽÑ

ࡾჽ$ࡥংং॥ ᤒᒮឃð  ࡾჽ   হ

ࡾჽ}᪵࠿॥ Y

As a general rule, when examining things about which there are doubts, if xin is not inwardly settled, then external things will not be clear. If my lu is not clear, then it is not possible to settle what is so of a thing and what is not so of it. Someone walking along a road in the dark may see a fallen stone and think it a tiger crouching in ambush, or he may see an upright tree and think it a standing man. The darkness has beclouded the clarity of his vision. Someone who is drunk may jump across a ditch a hundred paces wide, thinking it a drain half a pace wide, or may stoop down to go out to the city gate, thinking it a small doorway. The wine has disordered his spirit.31

The above passage indicates that lu can affect action. Distorted judgements are the result of unclear lu and will lead to odd actions that would not have been carried out when lu is clear. The connection between lu and action can also be observed in

b passages that discuss lu along with xing (¿ action). In two passages, morally worthy action is correlated with clear lu.32 Xunzi also contrasts the petty person’s lu and xing with that of the superior person. It is said that when the petty person lu, it is difficult for him to understand; when the petty person acts, it is difficult for him to be secure. On the contrary, it is easy for the superior person to understand when he lu and it is easy for the superior person to be secure when he acts.33 Thus, it is evident how xin lu affects what action is to be carried out. Sorting out this difference is conceptually important to our understanding of Xunzi. On the common reading, lu is an act of xin deciding which feeling is the appropriate one to manifest. On the proposed view, lu is an act of xin deciding what course of action one should take. In a way, we can still say that when xin chooses a

31 Xunzi 21/67-70; trans., K 21.8/3/108-9, modified. 32 E.g. Xunzi 4/24, 17/25, 18/105. 33 Xunzi 4/35-7.

61 2. Xin and Yu course of action, it is also choosing a set of feelings that go with it. However, this is very different from saying that xin is approving or disapproving particular feelings as such. It does not presuppose that feelings are aroused as they are already into motion prior to lu. Lu refers to xin choosing a course of action and the nature of this decision allows xin to take into account the feelings that are attached to actions. In the following, I will apply my understanding of lu to discuss the relation between xin and yu.

2.2 Relation between Xin and Yu

34 Xunzi defines yu as a response ying ං of qing. In order to better understand what

Xunzi means by “response,” we need to be mindful of not imposing our contemporary understanding of response as an agent’s action or behaviour that results from stimuli. Textual observation of the usages of “ying” suggests a causal connection between X and the ying of X. For example, there are references to the ying to a question35 and the ying of common people to a noble person.36 In addition, “ying” is frequently used in conjunction with “bian ” (change, alteration) to mean “to cope with an altered situation.”37 It is fairly clear that ying is induced by something else. It is also interesting to note that the Shuowen explains “ying” as “dang ዅ,” which roughly means “fitting” or “matching.” This suggests that there is certain expectation of what the ying of something should be like. The ying of X cannot be anything but has to be in a way that matches with X. This also allows us to better understand a passage wherein Xunzi describes the harmony between the superior and his subordinates in terms of “an echo that ying sound and as the shadow that ying the bodily figure.”38 Given certain properties of a hard surface, a sound can produce an echo in only one way. This implies that for the concept of ying, there is only one unique consequence that could occur in a given set of circumstances. In this sense, the causal connection between ying of X and X should also be a necessary one. In light of this observation about ying, we can take Xunzi’s claim “yu is the ying of qing” to mean that yu is a

34 Xunzi 22/63; K 22.5b/3/136. 35 e.g. Xunzi 23/32, 51. 36 e.g. Xunzi 8/50; Xunzi 10/79. 37 e.g. Xunzi 5/60; Xunzi 12/3-4; Xunzi 10/58, 66. 38 Xunzi 16/78.

62 2. Xin and Yu necessary consequence of qing and that there is only one way for yu to occur. This also means that qing and yu come in tandem. As discussed earlier, lu is not about xin choosing the more appropriate feeling (qing) and since yu is necessarily attached to qing, we can further infer that lu is not about xin approving or disapproving yu for Xunzi. Since lu is about xin choosing a course of action and since pursing the objects of yu is one possible course of action, it is the task of xin to decide on whether or not to choose the action of pursuing the objects of yu. Thus, the act of qiu is initiated by xin. This allows us to make better sense of [2]:

ᬵ&wŸ&ಚÏᒮŸ

Although yu cannot be eliminated, when what is pursued is not obtainable, lu zhe would like to moderate pursuit (jie qiu).39

Hence, when Xunzi speaks of jie qiu, he means that when xin decides on a course of action, xin does not excessively choose qiu as the course of action.40 With such an understanding of qiu, let us now return to [1]:

&ಎಚŸᒮ &ಎಚ੪‘Ÿᒮ ੪‘಴

Yu does not depend on whether its obtainment is possible/permissible, but qiu zhe (the one who pursues/one’s pursuing) follows what is possible/permissible (ke). That yu does not depend on whether the obtainment of their objects is possible/permissible is received from Heaven. That the one who pursues follows what is possible/permissible is received from xin.41

Given our previous discussion, we can take [1] to mean that yu is there regardless whether its objects are obtainable or not. However, the pursuit of the objects of yu is initiated by xin and hence follows what is considered to be possible/permissible (ke) by xin.42 If xin considers the act of pursuing the objects of yu as not possible, the

39 Xunzi 22/66; K 22.5b/3/136. 40 See discussion in section 1.4. 41 Xunzi 22/67-8; K 22.5a/3/135. 42 The question as to whether xin would regard an action as not ke is still open-ended at this point. As discussed in Chapter 2, this question concerns the basis of xin’s evaluation, which is a separate

63 2. Xin and Yu pursuit will not be carried out. Since it is entirely up to xin to decide whether the pursuit is to be carried out, this also explains why Xunzi says “order and disorder reside in what is made possible/permissible by xin.”43 On the proposed reading of [1], there is really no conflict or antagonism between ke and yu. The point of [1] is not about whether xin allows or disallows yu to move along, but about whether xin chooses a course of action that will lead to obtaining the objects of yu. Such a reading is consistent with Xunzi’s claim that the existence of yu is irrelevant to any discussion about maintaining order. This also better explains why Xunzi always uses the term “suo ke ” instead of “ke” alone.44 For Xunzi, the key to maintaining order lies in what action xin regards as possible/permissible. Hence, the notion of “suo ke” is tied in with action. This observation does justice to an analogy that is used by Xunzi to explain [1]:

দࡥ)& ॥w॥&‚¹ྞࡥº“പ»š¼ჽ

^º&፪ᬻº¿»À૆Y

All humans follow what is possible/permissible (suo ke) and abandon what is not possible/permissible (suo bu ke)...If there is someone who yu travelling south, cannot be more, and hates travelling north, cannot be less. Surely [this person] would not depart travelling south and walk northward just because it is not possible to exhaust all the southern ground.45

It is more plausible to take “suo ke” to mean the action that xin regards as possible/permissible for it to choose. The passage is saying that as long as the person who likes travelling south regards travelling south as a possible course of action that she can take, she will continue to travel south. Before I propose my own interpretation, let us take a brief look at how ke and yu are generally understood in the literature. Bryan Van Norden understands Xunzi’s idea of moral agency in terms of the distinction between ke and yu. According to Van question that will be dealt with in Chapter 4. Since we are not certain about how xin evaluates things here, I will refer to ke as “possibility/permissibility” in this chapter. 43 Xunzi 22/62; K 22.5a/3/135. 44

This point is also noted by Siu-fu Tang in “The Heart/Mind’s Approval (ke ') and Second-order Evaluation.” 45 Xunzi 22/67-8; K 22.6a/3/136.

64 2. Xin and Yu

Norden, ke is derived from xin whereas yu is part of our nature. Ke can control the movement of yu by motivating actions that have nothing to do with yu or are even contrary to yu because, as Van Norden puts it, “approval simply trumps desire.”46 David Wong argues against the view that ke can override yu in a way that is completely independent of yu because that will imply the source of ke either comes from “the dictates of pure practical reason” or “the perception of irreducible moral properties”.47 Wong does not think Xunzi’s position would allow for these two sources of ke. Although Wong maintains that ke cannot be completely separate from yu for Xunzi, he is still committed to the view that ke can override immediate yu.48 T. C. Kline objects to Wong’s view that only yu can motivate action.49 According to Kline, both yu and ke are motives. Kline writes:

[T]his picture of motivation reveals that Xunzi believes we have two primary sources of motive force in the self. One source is the disposition of our nature that, in responding to our environment, produces desires. The second source springs from the capacity of our xin to both understand and judge possible courses of action, to approve and disapprove.50

Kline makes it clear that ke and yu are two distinct but not separate “motivational mechanisms” for Xunzi.51 Since Kline understands yu and ke as two sources of motivation, he is committed to the view that there could be a potential conflict between ke and yu. Kline maintains that yu can be transformed by xin into a complex motive that “can incorporate a wide range of cognitive descriptions and evaluations as well as rest on sensitive perception and understanding of the nature of external factors.”52 Kline, however, is hesitant to call the transformed, more complex motive “yu” because he would like to maintain that ke is another kind of motive. This raises the question as to which motive is causing action once yu is transformed. Take a superior person who diligently observes ritual practices for example. It is unclear on Kline’s account whether the superior person’s actions are motivated by yu or ke or

46 Van Norden, 2000, p.124. 47 Wong, 2000, p.141. 48 Ibid. 49 Ibid., p.161. 50 Ibid., p.239 51 Ibid., p.160. 52 Ibid., p.161.

65 2. Xin and Yu perhaps both. Moreover, even if yu is eventually brought in line with ke, it would mean that the two moving forces are synchronised with the possibility that they might collide again. Recently, Siu-fu Tang made an effort to break away from the framework that understands ke and yu as antagonistic by arguing that xin’s approval functions as a “second-order” valuation that is at a different level than yu.53 However, Tang also describes yu as capable of “misleading” and “dominating” xin.54 According to Tang, xin has to “gain an active control of heart-mind and put desires under our command.”55 Although Tang acknowledges that “the heart-mind necessarily approves one way or another,” he understands xin as passively approving when yu takes control, as if xin is being forced by yu to approve what it will not approve when it functions properly or ideally. According to Tang, the active evaluations are “proper evaluation” whereas the passive ones are “formed blindly.”56 Tang’s difficulty in breaking away the assumption that yu can sometimes take control of xin reveals his assumption that yu has causal role in action in and of itself. Each of the above accounts offers a different understanding of ke and yu. Nonetheless, the one common view that underlies these interpretations is that both ke and yu have causal roles in action.57 What distinguishes my proposed interpretation from the above interpretations is that the proposed interpretation does not ascribe a causal role in action to yu. Yu is not antithetical to xin because it is not a moving force that can give rise to action. Given our earlier discussion of lu, we can describe xin as taking yu into account when it chooses a course of action. So, when the object of yu is not pursued, it is just that xin has not taken yu to be the dominating factor and hence has not chosen the possibility of pursuing the objects over other possible courses of action. Ke seems to denote the range of admissible options. Pursing the object of yu could be one of the admissible options to be weighed against other options but it is not necessary. It is possible to have a conflict between the options of pursuing the object of yu and another admissible course of action. In other words, there could be a conflict between

53 Tang, 2008b. 54 Ibid., pp.19-21. 55 Ibid., p.22. 56 Ibid., p. 17. 57 This predominant understanding of ke and yu in the literature is also palpable in the frequent usages of terms like “trumps,” “overrides,” “checks,” and “commands” to describe the relation between xin and yu as two competing forces.

66 2. Xin and Yu two possible courses of actions, but not between ke and yu.58 In cases where the objects of yu are pursued, it is xin that has chosen the course of action to pursue. Recall in passage 22.5a, we saw Xunzi talking about how someone will choose death over life because it is not possible/permissible to live:

Y

ࡥ„¡¢ࡥപ›¡¢ࡥྞ „£›ᒮœ&„›&

ࡾ„ࡾ›Y

What human beings yu most is life, what humans hate most is death. Yet there are some who follow life and end up with death. It is not that they do not yu life and yu death. It is not possible/permissible (ke) to live but possible/permissible to die.”59

To the AI interpreters, the above would be a clear instance that shows xin is in conflict with yu. However, on the proposed interpretation, it is simply a case where xin decides on a course of action to pursue death. The conflict is between courses of action xin can choose, that is to pursue life or to pursue death, where it is obviously not possible to choose both at the same time. For someone who decides to die, it is not that her yu loses in the battle with xin. It is just that her xin chooses to pursue death after weighing the possibilities. Hence, it is impossible for ke and yu to conflict for Xunzi, not because xin is absolutely superior to yu, but because it is absurd to describe yu and ke as in conflict. Similarly, it is absurd for one to say that a factor of a decision conflicts with the final outcome of that decision. The two things are simply not in the same category. Suppose a perverse object of yu is pursued. For Xunzi, it would be a case in which xin has somehow included the pursuit of the perverse object of yu as a possible course of action and subsequently chosen this action. In Xunzi’s use of the term, we can say that xin has given a lot of “weight” to yu in deciding what action to choose, but the decision to pursue is one that is entirely made by xin on its own. For this reason, we can only blame xin for making an inappropriate choice. Yu simply falls

58 I intentionally leave open the question as to why xin would or would not choose qiu over other actions because it hinges on the basis of xin’s evaluation. Discussing the basis of xin’s evaluation at this point might frame our interpretation of the relation between the concepts xin and yu with unnecessary assumptions. 59 Xunzi 22/59-60.

67 2. Xin and Yu outside the discussion of moral responsibility. This is not the same as saying xin has failed to suppress the force of yu in causing action or that xin was swayed by it. If Xunzi indeed thinks that xin and yu are both capable of leading one to action, then in a case where the perverse object of yu is pursued, it must be yu that is directly causing the act of pursuing. Although xin would be held responsible for failing to stop yu, yu is ultimately the direct cause of disorder. This would mean that yu is still involved in the category of order or disorder, however small the extent may be. Then, we would have difficulty doing full justice to Xunzi’s view that yu is not in the category of order and disorder.60 If we grant that Xunzi is a serious thinker, it is not sufficient to understand him as just saying that yu is given. He should have assumed also that yu cannot in any way cause action when he says that yu is irrelevant to discussion of causing order and disorder. Hence, interpretations that assume a potential conflict between yu and ke would be working on a categorical mistake that supposes xin and yu are both moving forces that cause action.61 By understanding xin to be the only cause of action, the proposed interpretation allows us to better see why Xunzi stresses the point that xin “on its own (zi ஀) prohibits or causes, renounces or selects, acts or stops.”62 This is because xin is the only source that issues actions. Since all actions must originate from xin, it is impossible that xin is shaped by another source of motion. Thus it is said that xin only “issues orders but does not take orders.”63 The picture of the relation between xin and yu from the Xunzi, then, is that xin can take into account a range of factors, including yu, when it is trying to decide which action to choose. Yu is a factor in xin’s decision. However, it is entirely up to xin to determine whether yu can be the dominating factor for an action. Hence, when Xunzi says yu does not fall into the category of order and disorder, he should mean that the whole discussion of how to maintain order is not applicable to yu because (a) yu is given, (b) yu is a blind response and therefore cannot deliberate upon how to act, and (c) yu cannot cause action. In this sense, it would seem absurd to Xunzi when one tries to argue for any of the following: yu is good, yu is bad, yu is neither good nor bad. Since yu has no causal role in action under Xunzi’s framework, yu cannot be held

60 Xunzi 22/56-7, K 22.5a/3/135. See also discussion in 1.4 for Xunzi’s view on this point. 61 My usage of “categorical mistake” is based on Xunzi’s claim that the issue of having yu and not having yu is not in the same category of order and disorder (Knoblock 22.5a/3/135) and therefore should be distinguished from any other similar usages in contemporary philosophical discourse. 62 Xunzi 21/45-6. K 21.6a/3/105. 63 Ibid.

68 2. Xin and Yu responsible for any moral failures. The idea of blameworthiness or praiseworthiness simply cannot be attributed to yu. The source of failure to pursue what is in accordance with Dao is xin. It is xin that has weighed yu to be the most important at the time of decision and subsequently chosen a course of action that pursues the object of yu. I hope I have by now disambiguated the relation between the concepts xin and yu in the Xunzi and therefore found a point of departure for answering the question of how moral action is possible for Xunzi.64 Let me set aside the issue as to what constitutes moral action for Xunzi and begin with an explication of the AI advocates’ understanding on the issue of attributing responsibility. Recall that, for Xunzi, xin does not begin controlling the pursuit of the objects of yu right away. Under the AI assumption that xin controls the activities of yu, there are a number of possible ways to think about why xin does not intervene to block or redirect the action of yu before moral training. One possible reason xin does not intervene at the outset, under the AI assumption, is that xin is initially weaker than yu. At this stage, xin will be easily swayed and overcome by yu. Moral training is about making xin stronger to combat and control the causl force of yu in action within morally permissible realms. However, given Xunzi has explicitly said that xin is the lord of the body and is not subject to manipulation,65 it is unlikely that the AI interpreters would consciously submit to this view. Even if AI interpreters understand xin as being always stronger than yu, there are still three possible ways for AI interpreters to understand why xin has not controlled the action of yu at the outset. It is possible that xin does not know how to exercise its power over yu. In this regard, moral training is a more technical enterprise that teaches xin how to put itself into use. Or, the AI interpreters could understand xin as not knowing what is morally permissible and what is not. Moral training is a process of educating xin about moral boundaries so that xin can put itself into proper use. The third possible reason is that xin initially endorses the same objects as yu. Moral training would then be about rectifying certain problematic preferences of xin. Among these three possibilities, the first two presuppose a moral quality in xin. So when xin fails to control yu, it is not the case that xin is directly causing the disorder.

64 By “moral,” I only refer to Xunzi’s conception of attributes that are conducive to good order. I am not therefore claiming that Xunzi has a conception of morality that is akin to any of those proposed in the Western traditions. 65 Xunzi 80/21/44-6; K 21.6a/3/105.

69 2. Xin and Yu

Xin is simply not fully functioning. Only the third possibility presupposes a bad moral quality of xin. Nonetheless, we still cannot attribute full responsibility to xin in cases of disorder. This is so because under the AI frameworks, yu alone is capable of causing actions. Even if xin directs yu to pursue its objects, yu is ultimately the agent that is directly causing disorder. It should be clear by now that under AI, we are not able to attribute full responsibility to xin in cases where the pursuit of objects of yu causes disorder. On the proposed interpretation, moral action is not a result of shaping yu into moral yu or training xin to better keep yu in check. Conjoining xin and yu in propositions such as “xin stops yu from qiu” or “xin allows yu to qiu” is probably inconceivable to Xunzi because the two are not in the same category.66 In Xunzi’s construction “A yu X,” the concept of yu denotes the nature of relationship between A and the object of yu, more specifically, it describes the propensity of A towards X, where A may refer to xin or the sense organs. However, action arises from xin and only xin. Hence, any attempt to address the issue of moral action in Xunzi should begin with xin. If moral training or moral transformation is needed at all, it is not because xin does not control or supervise yu properly, but because xin does not correctly assess the importance of one factor against other factors and hence chooses the wrong course of action. A person’s level of moral attainment is reflected by how well her xin makes decisions. And for Xunzi, a good decision is one that is made in accordance with Dao. An implication of understanding xin and yu as concepts of different kinds is the deconstruction of a fictional separation between xin and yu that rests on the assumption that yu and xin are oppositional entities. This opens up the possibility that yu and xin are not distinct entities. Indeed, this impression is intensified by the observation that yu is often discussed as belonging to xin in texts of the same period. In the Lunyu, Confucius describes himself as following the yu of his xin without overstepping the ju (boundaries, lines) at the age of seventy.67 The character ju originally refers to a carpenter’s square and therefore has the connotation of boundaries or limits. Huang Kan ጴࢤ, in his commentary, takes ju to mean “standard

66 See discussion in 1.3 for account of path-blocking and path-re-steering antagonistic interpretations. 67 Lunyu 2.4.

70 2. Xin and Yu

68 (fa Å).” Presumably, Confucius is here referring to certain accepted ethical standards. It is worth noting that Huang Kan draws a parallel between the yu of xin and xing and between ethical standards and practice by saying that the reason

Confucius can freely follow the yu of his xin is that his xing and practice (xi ᒍ) are complete. Huang Kan’s commentary makes three important suggestions about early Confucian understanding of yu. First, xin has yu. This means that xin and yu are not conceived dualistically. Second, there is an initial mismatch between xin’s yu and ethical standards. For Huang Kan, this discord has something to do with one’s xing. Third, the yu of xin can eventually fall into line with ethical standards provided that there is sufficient practice. The Guanzi also talks about how xin has to be empty of yu in order to be stable and properly aligned.69 The “” chapter lists yu alongside “sorrow, happiness, joy, anger” as the things that can make xin lose its form (xing ). 70 This suggests that yu is one of the many states of xin. Furthermore, there is discussion about another xin hiding inside xin. The inner xin is what enables xin to rule the senses and put all things into proper measure71 because it is believed to be where the vital essence lodges.72 With sufficient concentration of vital essence, xin will gain an enlightened knowledge of Dao and thereby make appropriate responses to external circumstances. However, the inner xin will be disturbed and prevented from functioning properly when xin has yu. Similar idea is found in the “Xin Shu I” chapter, which says that the senses will not identify their proper objects when xin is full of

73 yu. It also advocates ridding xin of yu so that shen (numen, spirit) will enter and take abode.74 These observations suggest that, for the author of the Guanzi, xin is the seat of yu. Controlling yu is a matter of xin controlling itself rather than xin controlling a separate entity.

68 Lunyu jijie yishu 1/16b/1-7. 69 Guanzi 49/273/1. 70 Guanzi 49/269/3. 71 Guanzi 49/270/2 and Guanzi 49/270/11.

72 X Guanzi 49/270/14. I follow Rickett’s reading of this line, which emends yin W to yi and translates yi as “power of awareness” (1998 v.2, p.47). According to Rickett, yi “seems to refer to a basic mental capacity that precedes thought (1998 v.2, p.31). 73 Guanzi 36/219/1 and Guanzi 36/220/6. 74

Guanzi 36/219/5. “Shen Y,” a prominent concept in the Guanzi, is often translated as “the Spirit” or “numen.” Here, it suffices our purpose to understand that shen is associated with Dao and that shen gives rise to penetrating insight into all things.

71 2. Xin and Yu

The connection between xin and yu is evident in the Lushi chunqiu. The “Zhong xia” chapter, for example, sets forth certain regimens to which a superior person should adhere in response to the second month of summer. Among them, “maintaining a stable xin” is listed between “withdrawing indulgent yu” and “quieting the various bodily organs”. This suggests that xin is not treated as a distinct and superior entity that sits above yu and the senses. Another observation about the treatment of xin and yu in the Lushi chunqiu is that a certain condition of xin is intimately linked to the presence of yu. The “Chi yue” chapter, for example, says that someone who indulges in yu necessarily has a greedy xin.75 There is one passage in the “Shi yin” chapter that talks about how the senses will not react to their ideal objects when xin is not in a pleasant state, despite it being a characteristic feature of the senses to have yu for their ideal objects.76 It is said that it is only when xin is in a pleasant state that there is yu of the senses. This suggests that the presence of yu is contingent upon xin. In view of this, it is unlikely that yu moves blindly towards its objects of its own accord. Furthermore, there are explicit discussions in the Lushi chunqiu about xin wanting the objects of yu. In another passage of the “Shi yin” chapter, it is said that xin can only be in balance when objects of yu are obtained. Since the attainment of objects of yu is integral to the well-being of xin, it is difficult to disentangle objects of yu from objects of xin.77 The impression that the objects of yu are also objects of xin is intensified by a passage in the “Guan biao” chapter that says, “Actions follow xin, xin follows yu. If someone’s yu lacks proper measure, his xin lacks proper measure.”78 Since the satisfaction of yu is integral to the functioning of xin, it seems implausible that the author of the Lushi chunqiu regards xin and yu as detached entities. A view that suggests xin and yu are not distinct can be found in Chen Da-qi’s interpretation of Xunzi. Chen sees a connection between natural responses and xin in the sense that xin makes natural responses to external stimuli. According to Chen, the arousal of these responses is natural and spontaneous. Although xin can further deliberate and decide on these responses, it is still xin that spontaneously makes responses to these stimuli in the first place. He further uses the relationship between

75 Lushi chunqiu 5/3.3. 76 Lushi chunqiu 5/4.1. 77 Lushi chunqiu 5/4.2. 78 Lushichuqiu 20/8.1.

72 2. Xin and Yu

a 79 a xing Z and xin to show that yu is an activity of xin. Since xing is a component of xin and since yu is a component of xinga , yu is also a component of xin. Xin makes the response of liking (hao) to things it considers as beneficial and subsequently yu that particular thing.80 Chen explicitly argues that the term xin is not used in a narrow sense to refer only to those functions that fall outside xing and yu. Rather, Xunzi uses xin as a generic term that encompasses a range of psychological factors that include xing and yu. Whether it is the xin that knows lia or the xin that yu, it is the same xin.

On Chen’s view, yu is a constituent of xin (಴቗). According to Chen, if one only takes xin to be the part that knows and juxtaposes it with yu, it is not doing justice to the original meaning of Xunzi’s use of terms. 81 Since there are certain dimensions to Xunzi’s concept of xin that I have not explored yet, I will leave discussion of the relation between xin and yu here and resume it in my discussion of xin’s motivation in Chapter 4. Let me also add one final remark on yu. Although yu itself is not morally blameworthy, it does not mean Xunzi cannot consider yu unfavourably. Yu could still be regarded as unfavourable, but not in the sense that yu causes problematic actions or that yu dominates xin. In certain contexts, yu could be directed to very problematic objects such as other people’s property or an elder’s food. Moreover, the proposed interpretation does not preclude the possibility that yu has a role in influencing action.

2.3 Further Issues

Earlier we discussed how moral responsibility is ascribed to xin. The next question is to what extent and in what way is xin held responsible? On AI, the action that is caused by yu is subject to xin’s approval. This presupposes that xin itself is not involved in an act of disorderly behaviour. Although xin would still be held responsible for causing any inappropriate actions, the charge against xin is that xin fails to properly guide yu. This is analogous to saying that xin has given the green light to yu when a red light should have be given and thus, results in a car crash. In

79 Chen 1954: 35-38. 80 Chen first says that xin makes the response of liking (hao) when it encounters beneficial stimuli. Later on when Chen discusses yu, he argues that hao and yu denote the same kind of response for Xunzi. Chen,1954, pp. 33-36. 81 Chen, 1954, p.38.

73 2. Xin and Yu this sense, xin is blamed for giving the wrong signal but not for hitting another car. Yu is the driver that hits the other car. But since yu cannot be the driver on the proposed interpretation, we need to determine who the driver is. An attempt to answer this question requires us to take into consideration the relation between xin and the person in the Xunzi.

Until this point, I have sidestepped discussion of “the person (ren ࡥ)” in my analyses of Xunzi.82 In this section, I will point out the relevance of Xunzi’s conception of the person to our present purpose and its implications on our interpretations of xin and yu. Recall the line in passage 22.5a of the Xunzi that has sparked much discussion about the relation between xin and yu:

[1] &ಎಚŸᒮ &ಎಚ੪‘Ÿᒮ ੪‘಴



Yu does not depend on whether its obtainment is possible/permissible, but qiu zhe (the one who pursues/one’s pursuing) follows what is possible/permissible (ke). That yu does not depend on whether the obtainment of their objects is possible/permissible is received from Heaven. That the one who pursues follows what is possible/permissible is received from xin.83

In Chapter 2, I argued against the assumption of AI that qiu Ÿ is a movement generated solely by yu and pointed out that there is no plausible explanation for the view that the agent of qiu is yu. Nonetheless, it is clear that qiu is a movement towards the object of yu. Our subsequent task is to examine how it is possible for Xunzi to maintain both the claims that qiu is a movement towards the object of yu and that yu is not the agent of qiu. In my translation of the above passage, I intentionally leave open the translation for the term “qiu zhe Ÿᒮ, ” which can be understood either as “the one

82 I am grateful to Karyn Lai for drawing this point to my attention. 83 Xunzi 22/56-62; K 22.5a/3/135.

74 2. Xin and Yu who pursues” or “the pursuing (as in X’s pursuing of something).”84 Although a difference in our reading of “qiu zhe” will not significantly alter the meaning of [1], it does frame the issue of concern differently as we proceed to read the Xunzi. I do not intend to argue against any of the possible readings; rather, my goal is to bring out a hidden caveat in reading “qiu zhe” as “the one who pursues.” The passage that follows [1], passage 22.5b, further elaborates on the idea of limiting qiu. In 22.5b, there are two sentences that bear close resemblance to [1]:

[2] ᬵ&፪Ÿᒮìᡴ፪ᬵ&wŸ&ಚÏᒮŸ

Although the objects of yu cannot be exhausted completely, qiu zhe (the one who pursues/one’s pursuing) would like to come close to exhaust them; although yu cannot be eliminated, when what is pursued is not obtainable, lu zhe (the one who deliberates/one’s deliberating) would like to moderate pursuit.85

Watson finds quotes [1] and [2] so similar that he believes that the sentences in [2] “simply repeat what has been said [in 1]” and suggests that [2] is probably an erroneous insertion.86 Knoblock also thinks that [2] repeats the point made in [1] but more so as an explanatory note.87 As similar as [1] and [2] might seem, I argue that it [2] reveals something subtle in Xunzi’s thought when we consider different implications of our different understandings of qiu zhe. Note that in [2] a differentiation is being made between qiu zhe and lu zhe. Let us unravel some of the complexities involved in our reading of “qiu zhe Ÿᒮ” as “the one who pursues” and “lu zhe Ïᒮ” as “the one who deliberates.” Recall that in section 1.4 we already ruled out the possibility that qiu zhe is yu because the notion of “following (cong)” entails that “the one who pursues” is capable of deliberation and deciding whether to act or not. We are now left with the option of taking qiu zhe to be the person. Only a few sentences after [2] it is said that, “All people (ren ࡥ) follow

84 We can get a glimpse of different understandings of qiu zhe from the different ways scholars have translated the term. Dubs translates qiu zhe as “the person who seeks”; Watson as “the search to satisfy desire”; Knoblock as “what is sought after” (Dubs, p.294; Watson, p.151; K 22.5a/3/135). 85 Xunzi 22/65; K 22.5b/3/136. 86 Watson, p.152. 87 Knoblock, vol. 3, p.136 n.102.

75 2. Xin and Yu what is ke and abandon what is not ke.” This reinforces the impression that qiu zhe refers to the person.88 But here is the caveat. Since [2] is talking about how lu zhe can limit qiu, it would be tempting for us to infer that there are two distinct agents involved here. In this case, the most probable candidate for lu zhe would be xin, for lu is explicitly defined in terms of the activities of xin in connection with inborn feelings.89 Such an understanding of the person as the one who pursues the objects of yu and of xin as the one who deliberates supposes there is a dichotomy between xin and the person. Since it views the person and xin as two separate agents, it falls back into the AI assumptions that yu can cause action and that xin can also move one to act in a way that overrides the action caused by yu. My approach is to keep assumptions at a minimal level. Since I have already shown that the AI assumption (that is, yu can by itself cause action) has little textual support, my strategy is to take what Xunzi has said about qiu and lu at face value and determine whether we can still make sense of his claims. As we have mentioned earlier, a few lines after the passage that says qiu zhe follows ke, we find the line that identifies the person (ren) as following ke. This suggests that the person is the agent of qiu. A survey of other passages makes it clear that when Xunzi speaks of qiu, he mainly refers to the act of the person. For example, it is said that the sage does not qiu to know Heaven, implying it is the person who qiu.90 Another passage talks about the superior person and the petty person having different ways to qiu what they like.91 There is also the passage that says it is because people (ren) lack goodness in themselves that they will qiu from the outside.92 Furthermore, there is a reference to the ancient kings trying to supply the qiu of the people.93 In light of these observations, we can be almost certain that the qiu zhe in [1] and [2] refers to the person.94

As regards to lu, there is one reference to the danger of the person ࡥ who does not lu about the different aspects of a thing:

88 Xunzi 22/67; K 22.6a/3/136. 89 See section 2.1 for detailed discussion. 90 Xunzi 17/10. 91 Xunzi 4/33, K 4.8/1/190. 92 Xunzi 23/33; K 3/155. 93 Xuzni 19/2; K 19.1a/3/55. 94 For further examples of describing the person as the subject of qiu, see Xunzi 23/34 and Xunzi 15/77.

76 2. Xin and Yu

Ñ॥ಶÒ$Ï॥പᒮÑ॥+ಶÒ$Ï॥zᒮ

ÓÔÕÖ$×॥പ੩Ù{&¬᫺¢দࡥഄपÑ

॥&Ï॥പᒮÑ॥+&Ï॥z

See something that can be yu, lu what can be detestable of it. See something that is beneficial, lu repeatedly what can be harmful with it. Weigh (quan) the two together and then carefully calculate it, then decide whether to choose or give up in regards to yu and dislike. Like this, * will regularly avoid falling into a trap. In general, the danger of people is the problem with being partial. If * see what can be yu, then * not lu what can be detestable of it; see what can be beneficial, then not lu what can be harmful with it.95

The above translation is intended to be literal. The asterisk represents where a subject is grammatically missing in English but not specified in the original Chinese text. Given the line that says being partial is a danger to the people (ren), it is reasonable for us to presume that the point of this passage is to advocate that people lu more about the different sides of a thing, suggesting that Xunzi regards the person as the one who lu. It is interesting to note that lu is described in terms of quan Ô, that is, the weighing of the merits and demerits of a thing. If we understand quan as an act of lu, we find another relevant passage that refers to a person’s weighing of options:

Qࡥ਄&ࡾ&ᕸÔࣨÞ&ßOà áࡥࡾჽNNà âࡥ

ࡾჽOãࡥࡾഠ NOÔ&ßåæ ࡥࡾჽççæ പࡥ

ࡾჽåãࡗࡥࡾഠ å炘ᒮੱ࡯ßÔ‚ᬻ˜êᕧຬ{ì

ࡾíîl©ಚ॥ïðటîlòóôჽ& ॥ໟY

Thus, when one acts, one cannot not weigh on some instrument. If the beam of the balance is not correct, a heavier object may rise up so that people will think it light and a lighter object may sink down so people will think it heavy. This is why people become deluded about light and heavy. If the

95 Xunzi 3/45-8; Knoblock 3.13/1/180.

77 2. Xin and Yu

counterweight is not correct, then misfortune may be entrusted to yu and deemed as fortune and fortune may be entrusted to dislike and deemed as misfortune. This is the reason people become deluded about fortune and misfortune. Dao, from the past to the present, has been the correct counterweight…To abandon Dao and rather chooses by own standards from within is like exchanging two for one. How could there be gain? Anyone who would actually exchange the yu that has accumulated for a hundred years for the dislike of a single moment does not understand the count.96

The above passage again takes the person to be the subject who weighs and conveys the importance of weighing on an accurate balance, namely, Dao. This leads Xunzi to say that xin must know Dao:

*Þ¯˜Q಴&ࡾ&˜಴&˜&˜œ˜

What is the balance? I say it is Dao. Hence, xin cannot not know Dao. If xin does not know Dao, then it does not approve Dao and approves what is not Dao.97

Piecing these passages together, Xunzi is basically saying that it is important for one to lu and by that he means weighing options on the accurate balance, Dao. He then concludes that: therefore xin must know Dao. From the way I have presented Xunzi’s position, it is clear there is a jump in his argument, from the claim that the person must lu in accordance with Dao to the claim that xin must know Dao. What assumption about the relation between the person and xin does Xunzi hold that warrants him to make such a jump?98 There are at least three possible explanations. First, xin advises, or even more strongly, dictates a person’s behaviour. When the person lu, she will lu in a way that

96 [ Xunzi 22/71-8; K 22.6b-c/3/137. Note that Watson and Knoblock read the character Z as and understand the last sentence to mean “for the gratification of the moment.” See also Watson p.158. On Knoblock’s and Watson’s reading, Xunzi would mean it is not worth it to exchange something that is highly wanted for a momentary satisfaction. The emphasis is on the disadvantage of not being able to

delay satisfaction. I take the character to be Z because the entire passage is about exchanging likes and dislikes. I understand Xunzi to be talking about the disadvantage of exchanging something that is highly wanted for something that is momentarily disliked. 97 Xunzi 21/29-30 K 21.5a/3/103. 98 I am indebted to Karyn Lai for drawing my attention to this point.

78 2. Xin and Yu is in accordance with the orders issued by xin. This understanding ascribes agency to xin. It sees xin as the agent that carries out the final act and regards xin as an autonomous, animate, and intentional agent who can influence the behaviour of the person. The implication being that there could be a potential tension between xin and the person. Second, xin is identical with the person. When Xunzi talks about how the person lu, he actually means that xin lu. On this understanding, the concept of xin constitutes the person. Hence, when we ascribe agency to the person, we will still be ascribing agency to xin. Ultimately, it is the person that carries out the action. The third position, which is not necessarily incompatible with the second, understands xin as a capacity of the person. So, when Xunzi says the person should lu, he means that the person should exercise the capacity that allows her to valuate the different aspects of a thing. And when he emphasises the importance of xin knowing Dao, he means that the person should know how to exercise this capacity properly. On the third understanding, there is only one agent, that is the person. A failure to act correctly is not an instance of xin deciding wrongly but of the person exercising his capacity in a problematic way. In short, there are at least three ways to construe the relationship: (1) xin is a source of agency that can dictate the person, (2) xin is the person, and (3) xin is a capacity of the person. Paul Goldin seems to have adopted the first view. Goldin understands personhood as being composed of two parts, a person’s initial state (xing Z) and their conduct or artifice (wei ). Xin, which is explicitly identified by Goldin as the agent, pertains to conduct.99 However, this first possibility is unlikely on the proposed interpretation. Since there are descriptions of how the person deliberates and weighs, it is clear that the person has agency. But if the person’s action is not necessarily moved by xin, as Goldin would think, then it is not clear what else could have led the person into action. As we have seen, this additional force that causes action cannot be yu. Moreover, there is no textual evidence that hints at an agency other than xin. Xunzi claims that xin determines action on its own and is never coerced by other’s order. This further intensifies the impression that there is no second agent in Xunzi’s thought. If the person’s action is wholly dictated by agent xin, and if the person also

99 Goldin, p.17.

79 2. Xin and Yu has agency, and given there is no second source that can move one into action, it is difficult to see how xin is merely part of the person. This leaves us with the remaining two possibilities that xin constitutes the person or that xin is a capacity of the person. However, there does not seem to be enough textual evidence for us to adjudicate between the two because these two views are not necessarily incompatible. Xin can be the most defining capacity of the person, such that talking about xin is almost equivalent to talking about the person. Even so, these two views will have a different bearing on the extent to which we can ascribe moral responsibility to xin. If xin constitutes the person, xin will be held responsible for moral failure on the grounds that xin causes and conducts the action. Relating back to the car-crash analogy, this is like saying xin is the one who decides to drive forward and xin is the one who hits the car. If xin is a capacity of the person, then moral failure is a result of the person exercising his capacity, xin, poorly. In this case, the person is the driver, who has exercised the capacity of xin to determine whether she should drive forward. Xin is responsible for making a wrong decision but the person is the one who carries out the action of hitting another car. I am inclined to accept the third view in consideration of Xunzi’s account of the difference between xin and the senses. As we have seen in the previous chapter, xin is discussed along with other bodily senses as being one of them.100 The senses clearly do not belong to xin but belong to the person. Hence, it is more reasonable to understand xin as a capacity of the person just as the eye and the ear are capacities of the person. What allows Xunzi to regard xin as the ruler of the body is the claim that

c xin has the capacity of zhi , which allows it to understand, make decisions, and subsequently cause action. 101 As Xunzi says, zhic is a capacity that is naturally endowed and common to all humans. Both the sages and the petty person share the same capacity of zhic. Because of zhic, xin can collect the evidence supplied by the senses (zheng) and accord a meaning to the things being sensed. According to Xunzi,

100 See section 1.1

^

a a g he i a j hk c ac a de

6 b f

101 `

#_ _ ' _ ' _

Cf. Xunzi 22/418 ( \] and

^ p

kr h c a m n o

q

l ' ]# 4 ) discussed in section 1.1. There is also a reference to how the mouth does not find relish in the taste of food when pain and illness reside in the heart (Xunzi 22/80-1). This understanding of the relation between xin and yu is also consistent with ideas expressed in the Liji, a later Confucian text. It is said that when xin is properly in place, the senses will not be able to appreciate what comes before them. For example, one can sight things but cannot see what it sees, one can hear things but cannot listen, one can eat but does not know the taste (Liji 19/182/20-3).

80 2. Xin and Yu if xin is not in a fitting state to make sense of the data it receives, the person will not be able to taste food when depressed. In another passage in the “Jie Bi” chapter, it is said that:

When [xin] is not [shi], then although black and white are in front of a person’s own eyes, he will not see them, or although the thunder drums are sounding on either side of him, his ears will not hear them.102

Even though the eyes and ears can still sense, without xin causing action, there will not be seeing and hearing. Since xin plays such a defining role in shaping the person, for Xunzi to talk about xin is almost equivalent to talking about the person. Unfortunately, there is no explicit clarification of the relation between xin and ren in the text and there is insufficient textual evidence for me to adjudicate on these three possible views. Some of the positions suggested here could be so alien to Xunzi’s thought that Xunzi himself might not have been aware of or thought through them at all. Instead, he has probably taken the issue for granted. Nevertheless, it is important to be conscious of these intricacies because how we interpret them will yield significantly different understandings of Xunzi’s view on motivation and agency. It is with these issues in mind that I examine the relation between the concepts xin and yu in the Xunzi. The point that it is unclear how Xunzi views the relation between xin and the person serves as qualification to my earlier position that xin is morally responsible for actions. It also calls our attention to the complexities in Xunzi’s thought. Although there is no clear answer to this question, it will not affect our following discussion in a substantial way. It sufficies for the present purpose for us to know that xin is the source of action. Due to the way Xunzi describes xin and the phrases that are suggestive of agency in xin, it is difficult to avoid phrases such as “xin chooses” and “xin decides” in my subsequent discussion of xin. Up to this stage, I have established that xin and yu are two different kinds of concepts for Xunzi. Yu cannot cause action; xin is the sole cause of action. Such a clarification is not only theoretically important in itself, it also lends conceptual clarity to a part of Xunzi’s thinking and hence sets the stage for our later discussion of moral action in Xunzi.

102 Xunzi 21/4-5; K 21.1/3/100.

81 CHAPTER 3

Orientation of Xin

3.1 Natural Inclinations of Xin

In this chapter, I argue that xin has a natural preference for certain objects and is therefore predisposed to cause action to purse them. Nonetheless, xin is capable of modifying the pursuits of these objects by lu, namely, a process of weighing options and deciding what object is to be pursued. I will examine the concept of zhic to show that the reason xin lu is not necessarily enabled by a moral capacity but a bare capacity to know things. Since Xunzi identifies Dao as the proper balance xin should know when it lu, I will also examine the concept of Dao and expose Xunzi’s assumption that there is an objective value attached to each option and Dao is what enables xin to know the actual worth of each option that xin has. The analysis of this chapter will support the conclusion that xin is not dichotomous. The xin that naturally causes problematic behaviours is the same xin that can know Dao and follow Dao, in virtue of the capacities it has. As discussed in the previous chapter, since xin has the ability to lu, one can entertain options and make a choice about whether or not to pursue the objects of yu. This also means that lu can modify qiu:

ᬵ&፪Ÿᒮìᡴ፪ᬵ&wŸ&ಚÏᒮŸ

Although the objects of yu cannot be exhausted completely, the one who pursues (qiu) would like to come close to exhausting them; although yu cannot be eliminated, when what is pursued is not obtainable, one who deliberates (lu) would like to apportion pursuit.1

1 Xunzi 22/65-6; K 22.5b/3/136. 3. Orientation of Xin

We learn from the above passage that qiu is the action that can be subject to lu but it is also suggested that qiu does not necessarily go through the process of lu. In one passage, it is said that:

দࡥഄपÑ॥&Ï॥പᒮ

In general, the danger of people is the problem with being partial. If one sees what can be yu, then not lu what can be detestable of it; when one sees what can be beneficial, then not lu what can be harmful with it.2

We also find another passage that says it is inevitable one will qiu the objects of yu given that yu is inborn:

ࡥ„ྞ&ಚ&vŸ

Humans are born with yu. If yu, but not obtainable, then there cannot be no qiu.3

Since the interpretation I propose only allows xin to cause qiu, this passage is read as implying that xin will cause qiu if it does not lu. Then how can we make sense of the suggestion that qiu will occur anyway, even if xin has not gone through a process of lu? How can we reconcile the claim that qiu is caused by xin with the claim that qiu is not necessarily issued from lu? Granting that Xunzi is not making contradictory claims, Xunzi could have stipulated that xin is multi-dimensional with different, independent functions such as xin can understand (zhic), xin can deliberate about a course of action (lu), and xin can cause actions (shia). In this way, it is possible for xin to cause action even without it deliberating about that particular course of action. This implies that xin is not always in the mode of lu and that xin will by default cause qiu. Let us delve deeper into the implications that xin does not always lu but can cause qiu even without the process of lu. Let us recall the passages that describe xin as naturally having preferred objects:

2 Xunzi 3/47-8; K 3.13/1/180. 3 Xunzi 19/1; K 19.1a/3/55.

83 3. Orientation of Xin

[ᕽ[ੰ[઎಴[+‚{ጳ„ ࡥചZᒮY

The eyes hao (is fond of/likes) colours, ears hao sounds, the mouth hao tastes, xin hao lic…all these are products of humans’ qing xing (inborn feelings given from birth/what is genuine of human nature).4

In another passage, it is said:

Qࡥചੰ[઎@઎) ᄮ[ᒵᒵV)!ᄮ[ᕽ໨#ᕫ%&

')(ᄮ,[‡ÎO᪥*)െᄮ಴[+,-)੅ᄮ

It is the qing (inborn feelings given from birth/what is genuine) of humans that the mouth hao flavours, yet no flavours or aromas are more refined [than those enjoyed by the Son of Heaven]; the ear hao sounds, yet no music is grander; the eye hao color, yet no assemblage of pattern or design, however complex, or of beautiful women is greater; the body hao relaxation, yet no contended ease or period of quietude is more pleasant; and xin hao lic, yet no emolument or salary is more substantial than his.5

According to the above passages, the senses and xin all have their inborn preferences.

c Li + is identified as what xin naturally likes (hao). And since we have discussed how the feeling of fondness, hao, entails yu,6 it can be inferred that xin can also be blindly drawn towards certain things that please it. Lic is usually translated as “profit.” 7 One of the connotations of lic is material possessions or gains. The term is often discussed in relation with concepts related to material possessions such as huo / (goods,

8 9 commodities), cai 0 (valuables, wealth), fu 1 (wealth, abundance), gong 2

4 Xunzi 23/25-26; trans., K 23.2a/3/154. 5 Xunzi 11/78-80; K 11.7b/2/160, modified. The quoted passage says the body prefers

d d 6 9 s pleasure/gratification (yi s). There is another passage that contains the expression “xin yu yi ” (Xunzi 11/46-7; K 11.4/2/156). I will not examine this passage in detail because the expression “xin yu” d

only appears once in the Xunzi. In addition, yi is mostly used in conjunction with “shen t” or “xing j ” to describe the body (e.g. Xunzi 12/5; 12/113; 19/116). 6 See section 2.2. 7 Knoblock vol.3, p.154; Watson, pp.164-5. 8 E.g. Xunzi 4/18.

84 3. Orientation of Xin

10 11 (achievements, good results), and quan Ô (authority, political power). And as we can see in the above passage, one form of lic is identified as grains and remuneration. Nonetheless, the usage of the term is not confined to material gains. In the Xunzi and other pre-Han texts, lic is frequently used to merely refer to that which is beneficial or

c 12 advantageous. Li is often contrasted with hai z (damage, harm) and used verbally to mean “benefit” or “provide favourable condition to something.” For instance, when a ruler lic his people, the people are supposed to be provided with what is beneficial to them;13 iwhen one l c a weapon, the weapon is being sharpened and is therefore put in a more advantageous condition for its purposes.14 For example, promoting lic that is aligned with the world is connected with removing hai of the world, while being fond of small lic is said to injure the state.15 There is also one passage in which caution and carefulness are considered as lic, whereas contest and anger are considered harmful.16 All these usages suggest that the meaning of lic can refer to material possessions as well as more generally to whatever is beneficial. The idea that can be extracted from these observations about lic is that xin is naturally drawn towards lic, that is, towards things that are considered beneficial. In this light, it is not difficult to see why xin causes the act of pursing the objects of yu by default. The objects of yu, as identified in the Xunzi, are mainly concerned with material goods, wealth, and prestige.17 Since these objects of yu are exactly the content of lic, we may infer that xin is naturally drawn towards the objects of yu. And since xin can cause action, it goes without saying that xin is predisposed to cause the act of pursuing the objects of yu. This also explains why Xunzi thinks that external objects can pull xin and make xin tilt:

9 E.g. Xunzi 8/90. 10 E.g. Xunzi 10/8. 11 E.g. Xunzi 4/19, 12/35-6. 12 E.g. Guoyu 1/1/3; Mozi 11/44/11, 15/44/1; Zhuangzi 2/96; 6/232; Xunzi 4/43; 5/24; 18/19. 13 E.g. Zuozhan 2/111; Xunzi 12/32.

14 E.g. Guoyu 10/4/350; Lunyu 15.10.

o v wx yov wz x ~ 

15 B { | } Xunzi 11/19-20. Also compare u in Xunzi 11/123 with in Xunzi 11/132. 16 Xunzi 13/43. 17 For example, it is said that humans all yu to have the entire kingdom, good food, big palaces, achievement, and admiration of the people (Xunzi 11/74-76).

85 3. Orientation of Xin

಴ࡗ{¢Q3ࡾ቗4ࡾ࿾÷)5រࡾ×{œ7óዔ¢}÷8

॥ßùî॥಴ê5&រࡾ7౉቗¢Y

Xin is also like this. Therefore, if you lead it with lia, nurture it with clarity so that no object (wu) can tilt it. Then, it is enough to decide between right and wrong and to resolve any confusions. If small objects (wu) pull it, then its proper position is externally changed. One’s xin internally tilted, then it is not enough to determine even gross patterns.18

Y

Interestingly, similar concerns about a pull from objects can be found in the Mencius, which says:

 &Ú ÷÷ࡕ÷8ఉ¢಴ ÚÚಚ&Ú

&ಚ

The faculties of ear and eye do not think (si) and are obscured (bi) by objects. When one object comes into contact with another object, then they are pulled away. However, the faculty of xin thinks. If it thinks, then it obtains (de) it; if it does not think, it does not obtain it.19

In both passages, we find a shared concern about a pull or distraction coming from objects (wu ÷). In the Xunzi, the term “wu ÷” is frequently used to denote the objects of yu. One recurring idea, for example, is that wu in the world are inadequate to satisfy yu.20 In the Mencius, there is one passage that refers to the “yu of the ear and

21 eye ” and since the above passage talks about how the ear and eye are pulled by wu, I infer that by “wu,” Mencius means specifically the objects of yu. Hence, I understand wu in both passages as referring to the objects of yu. It is evident that both thinkers share a concern about objects of desires but have a different

18 Xunzi 21/56-8; K 21.7b/3/107. 19 Mencius 6A:15. 20 E.g. Xunzi 4/72 (K 4.12/1/195), 10/5 (K 10.1/2/120), 19/3 (K 19.1a/3/55). 21 Mencius 4B:30.

86 3. Orientation of Xin understanding of what is being pulled by the objects.22 Xunzi thinks that xin is what will be led away by external objects, whereas Mencius regards the senses as what will be led away by external objects. Contrary to Xunzi, Mencius distinguishes xin as what is capable of resisting the pull from objects. This contrast helps us to see more clearly Xunzi’s view that xin is attracted by the objects of yu. This also explains why Xunzi says in one place that the ears of a superior person do not listen to lewd music because

23 lewd music makes xin lewd (಴<).

The above observations favour my proposed interpretation over AI because it shows that it is xin rather than the sense organs that is led away by external objects. This opposition to AI is further justified by passages that imply xin does not necessarily lu. In the “Jun Dao” and “Jun Zi” chapters, for example, it is said that the Son of Heaven can zhicu without. l 24 Since only xin has the capacity to zhic, the Son of Heaven must be excercising his xin when he knows (zhic) things. And since it is said that he does not lu when he zhic, we can further infer that lu is not necessarily involved in the operation of xin. In other words, xin can operate at an instinctual or default level where its distinguishing capacity of zhic is still functioning but the capacity has not enabled the activity of lu. At the default level, xin has a predisposition to cause action that pursues the objects of yu. In view of this intepretation, let us recall line [6] of passage 22.5a we considered in Chapter 1:

Q¤਄&੣಴§಴¨቗ᬵ“©प –&੣਄

¤಴«಴¬቗ᬵš©§ ࠿

The reason yu are excessively intense but action does not reach the point is that xin stops it. If what xin makes possible/permits is in line with lia, then although yu are many, how could it harm order? The reason yu does not reach a point as intense as action is that xin causes it. If what xin makes possible/permits loses lia, then although yu are few, how could it stop disorder?

22 Here, I set aside the important question regarding what Mencius means by xin will obtain something when it thinks and will return to this point in section 5.2.1. 23 Xunzi 19/25-6. 24 Xunzi 12/64, 24/2.

87 3. Orientation of Xin

I argued in Chapter 1 that “zhi ” in the above line should refer to action. On the proposed interpretation, this action cannot be caused by yu because yu cannot by itself cause action. It should be clear by now that the action that is being stopped is the action that xin would have undertaken if it were operating on a default level. Since xin can lu, it is also able to put a stop to the action it would naturally cause. In the previous chapter, we also discussed that xin can lu and decide whether or not to cause the act of pursuing. Through moral learning, xin will be able to choose moral actions instead of actions that pursue the objects of yu. This leads to another intricate question. If xin by default causes action to pursue the objects of yu, and if xin does not necessarily lu, what makes xin lu then? While xin will naturally order action to pursue the object of yu, lu can somehow make xin choose not to pursue the object of yu. Then, lu is what sets apart a natural response of xin and a moral response of xin. One possible explanation is that there are two levels or aspects of xin: one is the instinctual level of xin and the other is the moral level of xin. Even if xin is naturally inclined in the direction of objects of yu, it is still possible that there are certain features of xin that are congenial to morality. Tang Jun-yi and Wei Zheng-tong have proposed views similar to this. Tang Jun-yi, for example, argues that there is a moral level of xin that is able to automatically rise above a human level of xin. Lu, according to Tang Jun-yi, is this ability of xin to autonomously rise over its more instinctual side and be in line with Dao.25 Although Tang Jun-yi maintains that there is only one xin in Xunzi, his view implies that there is always a side of xin that is instinctual.26 If Tang Jun-yi is correct, then the antagonism seems to be located in xin itself rather than xin and yu. These two sides could have objectives that pull them in different directions, with the more reflective aspect of xin, which observes Dao, being able to repress a more instinctual side of xin, which aims only at the objects of yu. In the following section, I will examine this issue by paying attention to Xunzi’s understanding of why xin lu, what hinders lu, what enables xin to lu, and the proper goal of lu.

25 Tang, 2005a, pp. 77-78. 26 Tang, 2005a, p. 77.

88 3. Orientation of Xin

3.2 Zhic 

In order to understand why xin would begin to lu, given that it already has its own inborn inclination, we can begin with the passage that contrasts those who lu with those who do not lu, whom I respectively call person A and person B. There is one passage in the “Rong Ru” chapter, which contrasts two people. Person A does not eat the poultry and animals he raises, he does not to have wine, he stores his treasures in secured places, he does not dress in silk, and he does not travel by horse and carriages. Person B, on the other hand, consumes all his food extravagantly and finds himself destitute and in a difficult situation. He ends up starving and becomes a scrawny person, freezing, and begging in a narrow ditch. Xunzi explains this striking difference between the two people in terms of the following:

=(Yœ&@&᪖ÏB$CࡾDQY

It is not that [person A] does not yu. This is because he has a long lu and pays attention to what comes after. He is afraid that he does not have anything to perpetuate his wealth.27

=(Y࡯^ं„࿹HIã&JŠ!ࢩ&B॥$Y

Now [person B] is the kind of person who steals a moment of life and has shallow zhic (knowledge/understanding). Even this he does not zhic. He consumes his food in a very extravagant way and does not pay attention to what comes after.28

It is clearly stated in [1] that the reason person A saves is not that he does not have yu for those things but because he lu. Since person A lu, he worries that there is nothing left for him to consume in the long term. In other words, the reason person A starts to lu is because he wants to better serve his yu in the long run, indicating that he does not begin to lu out of a moral concern. This implies that even though lu is what allows xin to make a moral response, there is no guarantee that lu itself is a moral capacity or a

27 

I follow Wang Niansun and Knoblock in reading “ € ” as excrescent and have therefore omitted the two characters in my translation. 28 Xunzi 4/63-6; K 4.11/1/193-4.

89 3. Orientation of Xin moral aspect of xin. But if lu is not necessarily initiated by a moral concern and if we assume that person B also wants to live an enjoyable life in the long run rather then begging in the drain, then why does not person B lu? At this point, we are still begging the question as to what makes one lu. We can get a clue to this question from [2], which describes person B’s foolish behaviour in terms of him having shallow knowledge (zhic). This hints at a connection between how much zhic a person has and how the person lu. Let us delve deeper into this observation. Zhic is usually translated as “knowledge” or “understanding.”29 In the Xunzi,

c 30 there are numerous occurrences of the binary term “zhi lu Ï.” From passages that link lu to zhic, we find that the relationship between zhic and lu seems to be characterised in both ways. On one occasion, Xunzi condemns those who do not use zhic to lu the right thing, suggesting that it is in virtue of zhic that one can lu.31 Other passages, however, suggest that lu is what leads to zhic. For example, it is said that petty person would find it difficult to zhic uwhen, he l as opposed to the superior person who can easily zhic when he lu.32 And it is even said that the Son of Heaven can zhic uwithout. l 33 These seemingly different statements about the relationship between zhic and lu can perhaps be reconciled if we understand Xunzi as using “zhic” in three different senses. In one sense, zhic [ zhi1] refers to a mere capacity that enables one to understand things. This is the sense of zhic we encountered earlier when Xunzi singles out zhic as the distinguishing capacity of xin that allows xin to form an understanding of things in a way that the senses cannot.34 It is also in this first sense of zhic that Xunzi says humans are all born with zhic35 and that the superior person and the petty person have the same zhic.36 In a second sense, zhic [ zhi 2] refers to knowledge of a specific object. This usage of zhic can be illustrated by the following passage:

29 See for example, K 21.9/3/110, Watson, p.141. 30 “zhic ul ” as a binary term that appears ten times in the Xunzi. e.g. Xunzi 8/58, Xunzi 10/51; Xunzi 18/105. 31 Xunzi 21/84; K 21.9/3/110. 32 Xunzi 4/37; K 4.8/2/190. 33 Xunzi 12/64; 12.7/2/185. 34 See section 1.1. 35 Xunzi 21/36; K 21.5d/3/104. 36 Xunzi 4/32-3; K 4.8/1/190.

90 3. Orientation of Xin

Z&ËᑶQÚÏŸ

Human nature does not zhic rituals and propriety (lib iy a), therefore, [one] thinks and lu to pursue zhic.37

In the above passage, the second “zhic” is used as a noun to refer specifically to knowledge of rituals and propriety. The idea that zhic refers to knowledge of a specific

c object is suggested by the close association of zhi with the concepts ming and tong

38 L in the Xunzi. “Ming” can be literally translated as “bright” or “illumination.” In the classical texts, ming is usually used as an adjective to describe eyesight, in the sense of “clear-sighted.”39 By extension, ming as a noun refers to the ability to see clearly or to a kind of condition that allows one to see clearly. From this we can extract two related points about ming that are relevant to our discussion of zhic. First, the concept of ming involves a relationship between subject and object. It presupposes that there is a subject who sees and an object that is to be seen. Second, a state of ming is one in which one can see clearly. By inference, a state that is not ming is one that prevents one from seeing the object that she can clearly see in state of ming.40 Then, whether or not it is a state of ming determines the state the subject is in to perceive the object, but has no substantial effect on the state of the object. Hence, the concept of ming pertains primiarly to the subject. Let us apply these observations to the Xunzi. In the Xunzi, ming is used as an adjective to describe what Xunzi regards as a favourable condition of zhic. 41 For example, the superior person’s zhic is characterised as ming and his actions as faultless:

|MN༢ੜ፾‘ఈ ¿¤¢

If the superior person learns broadly and each day examines himself, then his zhic will be ming and his actions without fault.42

37

I translate “si ‚” as “thinking” for convenience. See section 2.1 for discussion of si.

38 E.g. Xunzi 3/16-20, 4/36-37, 21/42-44.

ƒ

, L . 39 M

E.g. Lunyu 16.10; Zuozhuan zhonggong + /1488; Xunzi 23/14. 40 This is different from a state that lacks ming, which can be understood as a state in which the subject cannot see clearly. 41 E.g. Xunzi 17/25, 18/105. 42 Xunzi 1/3; K. 1.1/1/135.

91 3. Orientation of Xin

The above passage stands in contrast to the following one that criticises shallow zhic and actions:

¿࿹ᚵྍ፷ྞࡾ፻à¢ࡦࡥ&vี_&v {ࡥ࠲ಶY

To be shallow and thin in zhic and actions that one makes no distinction between crookedness and uprightess so that one is unable to encourage human/benevolent (ren) people and to ming knowledgeable scholars — this is the third way that certainly brings upon destitution. 43

Given our earlier discussion of ming, we can take Xunzi to be saying that there are certain objects of knoweldge that one in state of ming is able to perceive or is supposed to perceive. If the person’s zhic is not ming, then it means that she is in a condition such that her grasping of the object of zhic is prevented. This suggests that zhic is not only a mere capacity to know but also knoweldge of a specific object. “Tong” can be literally translated as “passing from place to place without obstruction.” In the Yijing, for example, tong denotes the smooth flowing of things.44 In the “Nei Ye” chapter of the Guanzi, we find a reference to the idea that vital energies are not tong in the body.45 In this occurrence, “tong” is clearly used in the sense of circulation. The connotation of circulation in “tong’ can also be found in the Lushi chunqiu, which discusses how it is desirable for the blood vessels to be tong.46

Similary, tong in the Xunzi is sometimes contrasted with sai Z (blockage, obstruction) to mean unobstructed movement.47 In its early usages, tong seems to imply that at least two sides are involved, or else there is no point to speak of going through from one place to another. Let us then cnsider the following passage in the Xunzi that relates tong to zhic:

Qࡥ[Åಶჽ\…ࡥྞ[ྞÅ]L

43 Xunzi 5/21-2; K. 5.3/1/205. 44 E.g. Yijing 60, TC/A.11. 45 Guanzi 49/272/10. 46 Lunshi chunqiu 20/5.1. 47 Xunzi 3/44, 4/68.

92 3. Orientation of Xin

Therefore, if a person has neither a teacher or a model and that he zhic, he will certainly become a robber...if a person has both teacher and model and that he zhic, he will quickly tong.48

We learn from the above that there is a difference between the kind of zhic that allows one to tong and the kind that does not. It is also clear that Xunzi favours the former. Putting aside the question of which direction it is, it is reasonable to take Xunzi to mean that zhic moves in a certain direction. It is in this second sense of zhic that Xunzi classifies different kinds of zhic. According to Xunzi, there are the zhic of the sages, of the superior person, of the petty person, and of the servant.49 This further implies that there are different kinds of knowledge and that Xunzi favours certain kinds of zhic over others. Indeed, there is one passage that says “[the sage’s] zhic tong the great Dao,”50 which suggests Dao is the ideal direction towards which zhic should move. This leads us to discuss the third sense of zhic. The third sense of zhic [zhi3] can be understood as a type of zhi2, which refers to Xunzi’s conception of the ideal zhic . If we relate our above discussion back to the example of person B, we can now better understand Xunzi as saying that the reason person B’s behaviour is so reckless has to do with him lacking a certain kind of knowledge. Presumably, this knowledge is what enables xin to lu properly and make sound choices accordingly, as in the case of person A. As mentioned in the previous chapter, lu is about xin weighing possibilities and making a choice. In Xunzi’s view, it is very important for one to know the weight of things so that one will not mistake the light for the heavy or the heavy for the light.51 If one has not obtained knowledge of Dao (zhi3), xin might decide on the object of yu even though such a response would be improper. For this reason, an accurate balance is essential to accurate weighing:

Þ^à¢&_ࡾNO`a^b¢‚ÞᒮఞᕪY

48 Xunzi 8/105-7; 8.11/2/81 49 Xunzi 23/77-8; K23.6d/3/160. 50 Xunzi 31/17. 51 In one passage, for example, Xunzi praises the zhic and lu of the ancient people for not weighing what is light as what is heavy (Xunzi 12/75; K 12.8c/2/186).

93 3. Orientation of Xin

If the beam of balance is genuinely hung, then it is not possible to deceive about lightness or heaviness. The scale is the acme of flatness.52

And it is said that if the balance is not properly aligned, one will be mistaken about what is heavy and what is light:

Þ&ßOà áࡥࡾჽNNà âࡥࡾჽOãࡥࡾഠ NO

Ô&ßåæ ࡥࡾჽççæ പࡥࡾჽåãࡗࡥࡾഠ åç



If the beam of the balance is not correct, a heavier object may rise up so that people will think it light and a lighter object may sink down so people will think it heavy. This is why people become deluded about light and heavy. If the counterweight is not correct, then misfortune may be entrusted to yu and deemed as fortune and good fortune may be entrusted to dislike and deemed as misfortune. This is the reason people become deluded about fortune and misfortune.53

Xunzi unequivocally says that Dao is the correct beam and the correct balance weight, meaning that Dao is the correct balance:

˜ᒮੱ࡯ßÔᬻ˜êᕧຬ&åçæ

Dao, from the past to the present, has been the correct counterweight. If [one] abandons Dao and chooses by own standards from within, then he does not zhic where misfortune and fortune lie.54

*Þ¯˜Q಴&ࡾ&˜಴&˜&˜œ˜

52

Xunzi 19/32-3; K19.2d/3/61. Knoblock translates “ping „” as “equalness” and Watson translates it as “fairness.” I give a literal translation of “ping” as “flatness” because it is used to describe the beam of a scale. When the beam is flat, it means the steelyard is balanced so that the weight of the object being weighed can be obtained. This is consistent with Xunzi’s emphasis on getting the weight of an

object right rather than on being equal or fair. I follow Watson in translating as “acme” (Watson, p.99). 53 Xunzi 22/72-4; K 22.6b/3/137. 54 Xunzi 22/74; K 22.6b/3/137.

94 3. Orientation of Xin

What is the balance? I say it is Dao. Hence, xin cannot not know Dao. If xin does not know Dao, then it does not approve Dao and approves what is not Dao.55

It is worth noting that both quotes draw a connection between Dao and zhic. The general idea is that xin has to know Dao in order to make the correct choices. It should be clear by now that, for Xunzi, Dao should be the proper object of zhic. It is only through knowing Dao that one will be able to lu correctly. Hence, we can conclude that Xunzi’s conception of the ideal zhic is the knowledge of Dao. Although we have differentiated these three senses of zhic in the Xunzi, it is important to note that they are continuous and dynamic rather than independent of one another. Zhic is an inborn capacity that allows one to know things [zhi1]. It is in virtue of the capacity of zhic [zhi1] that one can obtain knowledge of a specific object [zhi2] In light of our observation about the third sense of zhic in Xunzi’s usage, we also learn that Xunzi is making a normative claim about the proper direction of exercising zhic as a capacity in Xunzi’s view, that is, this capacity should be exercised in a way that aims at knowing Dao [zhi3]. Before we can give a coherent account of the role Dao plays in lu, we still have the task of determining the nature of Dao.

3.3 Dao 

Xunzi says Dao is the accurate balance. But is this Dao an objective standard that applies to everyone or is it constructed by humans? Is Dao fixed and valid at all times? Xunzi’s concept of Dao has been a subject of debate in the literature. Recently, Kurtis Hagen has conducted a study on this subject matter and identifies a dominant group of interpretations that he labels as “realism.” Very briefly, in Hagen’s understanding, the main thesis of realism is that Xunzi conceives of reality as having

55 ‡ Xunzi 21/29-30; K 21.5a/3/103. I translate “quan †” and “heng ” respectively as “counterweight”

and “beam of the balance.” The Shuowen explains quan in terms of chui ˆ , which is a heavy metal block that is used as weight units, and heng as a wooden crossbar fixed between horns of oxen. Because of the connotation of wooden crossbar, heng can also be used to refer to the arm of a weighing apparatus. These observations altogether suggest that the kind of balance that Xunzi has in mind is a steelyard balance that comprises of a balance beam and a counterweight that slides along that the beam to counterbalance the load of the object being weighed.

95 3. Orientation of Xin a unique, correct, and fixed structure, both “physical and moral.”56 As a corollary to this view, Xunzi’s Dao is also understood as fixed and absolute. Hagen himself challenges this group of conventional interpretations and argues for a “constructivist” account of Xunzi. On the constructivist interpretation, humans construct social structures, distinctions, and categories in accordance with social values and the given interests of the people.57 According to Hagen, Dao is constructed by humans and therefore is “processional, contextual, and contingent.”58 The framework of debate sketched above is intended to show a tendency in discussion of Dao, that is, Xunzi’s concept of Dao is often discussed concomitantly with Xunzi’s of Dao. For example, when Philip Ivanhoe argues that Xunzi believes the Confucian Way (Dao) is unique and unchanging, he discusses how Xunzi believes that the world did not change much:

[Xunzi] did not believe the world could change as much as we know it does. Though he believed that the rites went through a process of evolution, he believed that this process had reached a conclusion in the rites of the Three Dynasties. Xunzi did not provide an elaborate exposition of the process of the evolution of the Dao, but he clearly believed that the sages had brought the process to a successful conclusion and that the Confucian Way provided the unique solution which would be valid for all times.59

Similarly, when David Nivison talks about how the way of the sages is the only correct set of rules, he identifies social order with the order of the universe:

The set of rules that optimises human satisfaction and minimizes conflict can be worked out essentially in only one way — the way of the sages did it. It thus can be thought of as (we might say) an overflowing into the human social order of the necessity of the order of the universe as a whole.60

56 Hagen, p.17. 57 See Hagen, pp. 13-25 for Hagen’s summary of various realist positions and pp. 30-35 for an outline of his constructivist interpretation. 58 Hagen, p.10. 59 Ivanhoe, 1991, p.318. 60 Nivison, 1996c, p.48.

96 3. Orientation of Xin

The tendency to discuss Xunzi’s Dao along with Xunzi’s metaphysics of Dao can also be found in Hagen’s constructivist account. Shortly after Hagen has stipulated that constructivists can admit of a structured reality, he goes on to argue that Heaven (tian) does not dictate distinctions among things and that Dao is shaped by humans.61 In my discussion, I will avoid this tendency to combine discussion of Xunzi’s Dao with Xunzi’s discussion of reality. This is not to argue for or against a particular view advanced in the literature, for it is entirely possible that the there is a close connection between Dao and reality in the Xunzi. Nonethless, I avoid discussing Xunzi’s metaphysics of Dao in order to maintain a conceptual clarity. These are two different subject matters of inquiry that involve different concerns and call for discussions that are of different nature. My focus is on Dao as a normative concept. In this case, I explore the nature of Dao: can it be conceived as an absolute or objective standard? And, if so, who does it apply to? Even though one might be able to draw implications about Xunzi’s concept of reality from my discussion of Xunzi’s concept of Dao, it should be noted that my claims about Xunzi’s conception of Dao do not amount to claims about Xunzi’s understanding of reality. As discussed earlier, Xunzi calls Dao the correct balance of past and present. The idea that Dao is used for accurate measurement at all times gives rise to the impression that the validity of Dao is unchanging. A similar idea can be found in the following passage:

eࡥᒮࡾఈిᒮQࡾࡥిࡥࡾചిചࡾ#ి#ࡾ'ి2ࡾ˜c፪ੱ࡯

lి#&೽ᬵg~቗Q᣾‘ᣃྍ&ᢎc‘k÷&ഠ

The sage is one who uses himself to measure. He uses people to measure people, uses what is in fact the case/feelings (qinga) to measure what is in fact the case, use categories to measure categories, uses explanation to measure achievements, uses Dao to view everything, the past and the present are one and the same. Categories are not muddled. Although it has been a long time, there is a common lia. Hence, when he faces what is perverse and deviant, he

61 Hagen, pp.32-33.

97 3. Orientation of Xin

will not go astray and when he looks at the various things mixed together, he is not confused (huo).62

The above passage again conveys the point that Dao can be applied to past and present. We can further draw out a few interrelated points about Dao. First, there is a connection between “using qing to measure qinga” and using Dao to view things. Second, there is connection between unvarying Dao and a common lia ࠶. Third, there is a connection between using Dao to view things and the absence of confusions (huo ). In light of these points, let us first consider the meaning of qinga and lia when they are used in relation to Dao in the Xunzi. In Yang Liang’s commentary on the above passage, he takes qinga to refer to the qinga of humans and therefore understands Xunzi as implying that humans have the same likes and dislikes. Yang’s interpretation is supported by Xunzi’s own definition of qinga as inborn feelings of likes and dislikes63 and by the fact that a majority of the usages of qinga in the Xunzi refers to what humans like or dislike.64 At this point, it is also worth noting Kwong-loi Shun’s observations that the term “qinga” in classical Chinese texts is often used to refer to “the facts about a situation” or to “certain deep features that reveal what things of this kind are really like.”65 Such usage can also be found in the Xunzi. For example, one passage refers to the qinga of music as going to the bottom of the root and reaching the limit of changes.66 Another passage presents the qinga of human beings as what all people will do in certain situations, such as dusting their caps when they have just washed their hair.67 Taking Shun’s observation into consideration, we can better understand qinga in the Xunzi as referring to the feelings of likes and dislikes that are characteristic features of human beings in virtue of every human liking and disliking the same things. For example, Xunzi often speaks of human beings having a common qinga such that they all yu the same things, such as wealth and the prestige of the Son of Heaven.68 Since these

62 Xunzi 5/14/35-37; K 5.5/1/207. This passage touches upon many difficult terms. Given the scope of this project, it is not feasible to examine each of the terms in length. Instead, I will only extract some points that shed light on our understanding of Dao. 63 Xunzi 22/3; K 22.1b/3/127. 64 See section 1.1. 65 Shun, 1997b, p.184. See, idem, pp.184-5 for a detailed discussion of the usages of qing in the classical texts. 66 Xunzi 20/34; K 20.3/3/84. 67 Xunzi 3/25-6; K 3.8/1/3. 68 For example, Xunzi 4/72 (K 4.12/1/194-5), Xunzi 11/23-4 (K 11.7b/2/160).

98 3. Orientation of Xin feelings are deep features of human beings, they are presumably difficult to be changed in a substantial way. This explains why Xunzi only talks about modifying

a 69 and adorning (lm) qing , as in straightening a piece of crooked wood. Even though a piece of crooked wood can be straightened, the substance and character of the wood remains unchanged. Hence, we can understand qinga as being used by Xunzi to primarily refer to certain common feelings that are characteristic of human beings. The idea that human beings have the same feelings of like and dislike towards different things merits our attention. Let us recall the passage that says feelings are differentiated by xin:

,ᕽ቗ࡾ0ᒵ1࿾ႚ456ᒵࡾ07ᗇᯄ࿦;ᤇ6઎ࡾੰ0?@

ABCᕓEF6@ࡾG0HIটᅘ၀MNOࡾ,0P'QRSTUV

Wപࡾ಴0

(1) Forms, colors, and designs are differentiated by the eye. Pitch and timbre, bass and treble, modal keys and rhythm, and odd noises are differentiated by the ear. Sweet and bitter, salty and bland, pungent and sour, and distinctive tastes are differentiated by the mouth. Fragrances and stenches, perfumes and rotten odors, putrid and rancid smells, foul and sour orders, and distinctive strange smells are differentiated by the nose. Pain and itching, cold and heat, smoothness and roughness, and lightness and heaviness are differentiated by the body. (2) (Shuo, gu,) pleasure, anger, sorrow, joy, love, dislikes, and yu are differentiated by xin.70

At first sight, it is difficult to pinpoint the similarities drawn between (1) and (2) because the objects that the senses differentiate are external, whereas the feelings that are differentiated by xin are internal. But given Xunzi’s understanding of qinga as what is given, we can understand this passage as saying that there are certain feelings that just exist, independently of xin. Upon contact with stimuli, certain feelings are necessarily aroused. Suppose there is a natural like for good food, then the feeling of

69 Xunzi 23/5-8; K 23.1a-b/3/151. 70 Xunzi 22/17-9 trans., K 22.2/3/129, modified. Commentators differ in their interpretations of the first two characters, shuo and gu. I will not investigate the meaning of these two terms here. My focus is on the succeeding terms that parallel the definition of qing.

99 3. Orientation of Xin like (hao [) will be aroused whenever good food is encountered. Xin can detect and differentiate them but cannot change them. Just as everyone’s month will find sugar to be sweet and salt to be salty, everyone’s xin will also feel pleasure and anger in the same way. It seems inconceivable to Xunzi that the sorrow someone feels in a certain situation will become the happiness another person feels in the exact same situation. This is akin to saying that all humans cannot help but feel a certain way in a given situation. Let us now turn to the concept lia. The character lia is comprised of two radicals: the left side refers to jade and the right side refers to divided field. According to Paul Demiéville, lia originally meant dividing land into lots.71 In its earliest

a a 72 occurrence, li in the Shijing appears in the sentence “wo qiang wo li °ዏ°቗” to mean putting land into order. In the other three occurrences of lia in the Shijing, lia

73 a refers to handling and regulating land. The same expression of “qiang li ዏ ቗” also occurs in the Zuozhuan, in which the ancient kings are said to qiang lia the land in accordance with what is suitable for its products so as to distribute its benefits.74 Here the activity of lia land is understood in terms of allocating land in accordance with its

a produce. Kung Ying-da (opq) makes the observation that li has to do with distinguishing which lot is suitable for which grain.75 It is palpable that there is a nascent sense of making appropriate differentiations in the connotation of lia. According to the Shuowen, lia means polishing jade. As a noun, lia means the veins or the patterns in jade. Lia is often linked to jade in the Warring States texts. As

Wing-tsit Chan notes, both the Yinwenzi r໨| and the Hanfeizi understand jade as a piece of crude stone that has been lia.76 Since one has to follow the veins and properties inherent in the jade in order to polish jade, the association of lia with polishing jade further introduces a sense of putting things into order by taking into account the pattern or structure or a thing. “Lia ” is later evolved to refer to the pattern or structure of a thing. In the Mencius, for example, lia is used to mean the

71 Wing-tsit Chan discusses Professor Demiéville’s view in Chan, 1964, p.123. 72 Shijing 2/6/6.

73 Shijing 3/2/6, 3/3/8.

, . 74 M

Zuozhuan Chenggong ‰ /797. 75 Chan, 1964, p.123. 76 Chan, 1964, p.128. Cf. Yinwenzi Da Dao II/14/15; Hanfeizi 13/66/14, 23/139/5.

100 3. Orientation of Xin harmonious pattern of music.77 And in the Xunzi, lia is once used to refer to the pattern of a face.78 In light of these observations, we can infer two closely related connotations of lia. One is about putting things into order, and the other is about dealing with things as they are. These two connotations are inseparable. If the jade is not dressed, it is only a piece of crude jade. This suggests that the emphasis is not so much on the veins themselves but on the activity of dressing jade. This is in agreement with Tang Jun-yi’s view that the emphasis of the original meaning of lia is more on human activities rather than objective external things.79 However, I would like to add the observation that objective facts of external things are also integral to the concept lia because they provide the basis for the activities of putting things into order. If one cannot distinguish the natural veins of jade, one will not be able to begin the activity of dressing jade. This means that although dressing jade is a human activity, it is not entirely up to humans to decide how this activity is to be carried out. Human activities are conditioned upon certain objective facts of things — facts that will allow one to put things into order. Hence, the concept lia encapsulates an interdependency between what is objective about a thing and the human activities performed on that thing. In the Xunzi, lia is used both as a verb and a noun. When lia is used verbally, it is quite obvious that lia means putting things into order. There are numerous descriptions about how a superior person is often said to lia the myriad things.80 When lia is used as a noun, Knoblock renders lia as “rational principle” and Watson as “reason.” However, it does not seem necessary for us to go so far as to read into lia the meaning of rationality or reason. In fact, we are able to make reasonable sense of Xunzi if we simply apply the meaning of lia in its early usages, that is, as a pattern that allows one to put things into order. As earlier said, the emphasis of lia, even when used as a noun, is not so much on the characteristics of a thing itself as on what enables one to put things into order. This can be observed in the connection between

a a li and order (zhi –). For example, as we have seen, Xunzi maintains that if what xin approves is in accordance with lia, there will be order.81 In addition, Xunzi defines

77 Mencius 5B:1. 78 Xunzi 21/55; K 21.7b/3/107. 79 Tang, 2005a, p.7. 80 E.g. Xunzi 9/66. 81 E.g. Xunzi 22/60-1; K 22.5a/3/135.

101 3. Orientation of Xin

a a 82 a goodness (shan ૽) in terms of li and zhi . The impression that the emphasis of li is on what allows one to put things into order is confirmed by a strong connection between lia and actions. Behaviours that are condemned by Xunzi are described as

83 a “messing up li ࠿቗.” According to Xunzi, actions that are good to li should be established while actions that do not benefit lia should be abandoned.84 In addition, it is said that actions that depart from lia will put one in danger.85 Because of this correlation between lia and orderly actions, the concept lia is also frequently associated with ritual practices. All, except one, occurrences of lia in the “Li Lun” chapter are

a coupled with the term “wen ” to form the binomial expression of wen li ໨ ቗, a concept that is integral to ritual practices.86 Hence, we can infer Xunzi’s point that it is important to observe the patterns of things, which will enable one to perform actions that give rise to order. With the above discussions of qing and lia, we can now turn to the linkages between Dao, qingai, and l a. In the “Jie Bi” chapter, Xunzi explicitly says that “emptiness, unity, and stillness” are what enable one to know Dao.87 It is said that one who has achieved the combined state of “emptiness, unity, and stillness” is able to work out the great lia and put the universe into order:

Y

ûಚ˜Ÿ˜ᒮ*tu*‚tu**!࿾ j÷)&Ñ)

Ñ&v)v¬wx yÑzྦ| ࡯vg}~cj÷॥ച

ੜ–࠿L॥ి᎞‚ƒ j÷’৅!቗ †‡¢YY

One who has not yet attained Dao but is seeking it should be told of emptiness, concentration, and stillness…Emptiness, unity, and stillness are called great

82 Xunzi 23/37; K 23.3a/3/155. 83 E.g. Xunzi 18/107, 23/3. 84 Xunzi 8/31-32. 85 Xunzi 22/79. It should be noted that the term “lia” does not always denote something positive in the Xunzi. In the “Fei Shierzi” chapter, for example, Xunzi harshly criticises those who promote partial knowledge as using statements having li to deceive and mislead ignorant people (Xunzi 6/8). Furthermore, Xunzi warns against the limitless pursuit of knowing the lia of everything (Xunzi 21/78- 82). Nonetheless, lia is mostly used in a positive sense that is associated with giving rise to order. 86 E.g. Xunzi 19/10, 38, 76, 103. 87 K 21.5d/3/104. Cf. Watson, p.131. For convenience, I will follow Knoblock’s translation of xu, yi, and as “emptiness,” “unity,” and “stillness” and will discuss these concepts in length in section 3.4.

102 3. Orientation of Xin

clarity and brightness. Each of the myriad things has a form that is perceptible. Each being perceived can be assigned its proper place. Each having been assigned its place will not lose its proper position. Although a person sits in his own house, yet he can perceive all within the four seas. Although he lives in the present, he can put in its proper place what is remote in space and distant in time. By penetrating into and inspecting the myriad things, he knows the qinga of things. By examining and testing order and disorder, he is fully conversant with their inner laws. By laying out the warp and woof of Heaven and Earth, he tailors the functions of the myriad things. By regulating and distinguishing according the great lia, he encompasses everything in space and time.88

It is worth noting that when the above passage characterises “emptiness, unity, and stillness,” it shifts from speaking of pursuing Dao to speaking of knowing the qing of things and working on the great lia. This suggests that knowing Dao involves knowing qing and lia. In the opening passage of the “Jie Bi” chapter, it is said that “All people

89 suffer from being obscured (bi ) by one bend (qu ྍ) and are shut off from the great lia.”90 While in another passage, Xunzi criticises other thinkers’ understanding for being obscured by one aspect (bi) and says that those who know the bend (qu) see things from one corner of Dao.91 This suggests very strongly that Xunzi is speaking of Dao and great lia as one and the same thing. However, Xunzi’s discussion of Dao and lia is not clear enough for us to pinpoint the nature of their relationship. Some scholars who have highlighted this close association between Dao and lia tend to draw the conclusion too hastily. Wing-tsit Chan, for example, takes Xunzi’s concept of lia to

92 a mean “moral principle.” Zhang li-wen ˆ‰໨, in his studies of the concept li , even

88

Xunzi 21/39-43; trans., K 21.5d-e/3/105, modified. It is possible for “er _” in “xu yi er jing” denotes a kind of sequential order as in it is through xu and yi that jing can be obtained. This does not concern

the subject matter here. My focus is on the relationship between the notions of xu, yi, and jing and the ‹ obtainment of Dao. Following Yang Liang, I take Š to be . 89 According to Shuowen, qu refers to a curved object. Here, it is most likely to be referring to the curvy part of a pattern. 90 Xunzi 21/1; K 21.1/3/100. 91 Xunzi 21/21-24. 92 In Chan’s discussion on Zhuangzi’s usages of lia, Chan notes that Dao is coupled with lia in the Zhuangzi. Chan then quickly concludes that, “for the first time in Chinese history li is equated with ” (Chan, 1964, p.125). It seems quite hasty for Chan to draw such a conclusion simply on the

103 3. Orientation of Xin goes so far to argue that Dao and lia are the same thing for Xunzi, both referring to the principles of things.93 Unfortunately, both Chan and Zhang have not offered elaborate explanation as to why lia has already acquired the connotation of a moral principle other than citing passages in which the terms Dao and lia appear in the same passage.94 Even though the terms Dao and li are sometimes associated in the text, it is unclear whether Xunzi intended to use Dao-lia as a concept. Nonetheless, there is a clear correlation between how well one knows Dao and how well one put things into order. In this sense, the connection between Dao and lia in the Xunzi is definite. This fits well with Xunzi’s view that the superior person, who is guided by Dao, is able to lia Heaven and Earth without uncertainty.95 It also strongly points to the view that lia is something that underlies Dao. Such an observation does not warrant us to make an abstraction and understand lia as the pattern of nature or some ultimate principle.96 All that has been implied is that Dao is dependent upon lia. As mentioned earlier, the concept lia entails the importance of observing the features and patterns of a thing. Since things have their inherent characteristics, one who lia them has to work within the boundaries of these characteristics instead of arbitrarily putting them into order. Thus, the concept lia also implies that there are objectively correct and incorrect ways of ordering things. This impression is further confirmed by Xunzi’s view that one who is guided by lia will have no uncertainty (yic

ዔ). One passage in the “Jie Bi” chapter compares xin to a pan of water. It is said that if the water is still, the mud will sink to the bottom so that the water on top will be clear and bright enough for one to see their beard and eyebrows and discern the ground that the two terms appear together. Similarly, the fact that “Dao” and “li” has appeared together seems to be the basis for Chan to understand Dao-li as a concept and to say that the term Dao-li was already found in Zhuangzi and Xunzi before Hanfeizi (1964, p.126-7). 93 Zhang Li-wen, p.32. 94 E.g. Xunzi 1/41, 22/41. 95 Xunzi 12/28; K12.3/2/179). There is another passage that says the Heaven and Earth is not lia if there is no superior person (Xunzi 9/65-6). This coincides with our earlier discussion about the emphasis of lia being on human activities. However, this is not to deny that things themselves have patterns or specific features. Xunzi seems to think that things have their specific patterns or arrangement that can be identified as lia. But for Xunzi, the meaning for human activities lies in working with these patterns to put things into order. For instance, the natural veins of jade are already there but awaits the activity of dressing jade to put them into order. 96 Wing-tsit Chan, for example, understands “great lia (da li)” as “the great principle,” which he distinguishes from the “principle of the world” (1964, p.126). However, it is unclear whether Xunzi has used “da li” as a singular term to mean an overarching principle that governs all things. It seems that “da (great)” is simply used as an adjective to describe a larger or more comprehensive pattern of things.

104 3. Orientation of Xin pattern (lia) of their face. But if wind stirs up the mud from the bottom, then it is not possible to obtain a general shape of the face. Xunzi then goes on to say:

಴ࡗ{¢Q3ࡾ቗4ࡾ࿾÷)5រࡾ×{œ7óዔ¢

Human xin is also like this. Therefore, guide it with lia, nurture it with clarity so that ojects do not make it tilt. Then, it is enough [for xin] to determine right and wrong and to resolve any suspicion and yic.97

The same idea that lia resolves confusions is framed negatively in the opening passage of the “Jie Bi” chapter, which suggests that people who are shut off from the great lia

c iwill have y ዔ:

ࡥഄ lྍ᫈ !቗‚íዔഠ¢‹m˜eࡥí಴Y

All people suffer from being covered by one bend and are shut from the great lia. If [one has] two yic, then he will be confused. The world does not have two Dao; the sage does not have two xin.98

Yic can be loosely translated as “doubt” or “uncertainty.” In the Xunzi, “yic” is

99 sometimes associated with huo ഠ, which means the state of xin that is adjudicating between two options in an either-or manner. In early texts, the term “yic” is used interchangeably with the term “ning Œ,” which was initially used to refer to ice and later used mean “fixated” or “not moving.”100 It is also worth noting that there is a close connection between yic and action. A person who has yic is caught between two incompatible options and therefore is unable to move forward. Some scholars have recognised the sense of exclusivity that underlies the notion of yic.101 What seems to be overlooked, however, is that yic refers specifically to the exclusivity between what

97 Xunzi 21/56-7; K 21.7b/3/107. 98 Ž 

Xunzi 21/1; K 21.1/3/100. I omit the characters Œ in my quote here because there are many different ways to read this line yet it does not significantly affect our present discussion. 99 E.g. Xunzi 21/49, 22/7-8. 100 See Hao Yixing’s commentary about yic and ning in Wang Xianqian, vol.1, p.251, line 2-4. 101 Kurtis Hagen, for example, translates yic as “conflicting views” (p.26).

105 3. Orientation of Xin

102 is correct and incorrect. This point is raised by Yu Yue ឃ, who interprets the above passage as saying there is only one Dao. “Two (liang í)” refers to what rivals the only one Dao in the world, and “yic” as what confuses people about Dao. It is worth noting that in Yu Yue’s explanation of yic, he draws on a passage in the Guanzi that says it is dangerous for the state to have concubines who yic the wife and ministers who yic the prime minister.103 The point of this passage in Guanzi is not that the concubine and the wife are in conflict but that one cannot confuse a concubine with the wife. It is simply wrong to think that the concubine is the wife. This brings to light an implicit assumption in the connotation of yic, namely, there is only one right option and for this reason, any second option is a wrong option.104 In this sense, yic in the Xunzi is not used as a general term to denote any states of uncertainty that might include one’s being unsure about the time of an event or what will happen in the future. More specifically, yic denotes an uncertainty regarding what is the right thing to do. This understanding of yic makes better sense of Xunzi’s statement in the above passage that there are no two Daos. This also explains why Xunzi values the kind of knowledge (zhic) that that can be used to settle on yia.105 Since the kind of knowledge Xunzi values is the knowledge of Dao [zhi3], Xunzi is basically saying that those who know Dao will not have uncertainty about what is the right thing to do. The implication of this is that in any situation, there is only one right action that is in accordance with Dao. With the above discussion, let us now articulate Xunzi’s concept of Dao. Since we have seen that lia involves working within the factual characteristics of things, and since lia underlies Dao, there should also be an objective element in Dao. Xunzi has faith in the objective lia, which has already laid out the value of things to be considered and believes that the Confucian way will indeed provide one satisfaction that is of a higher value than the satisfaction of yu. It is important to note that the claim that Dao is objective does not amount to the claim that Dao are moral standards that have independent existence. Since the standards that Dao provides for human activities are standards that derived from the confines of factual characteristics of things, Dao is not detached from things in the world. So, even when the sage or the

102 In Wang Xianqian, vol. 2, p.386, lines 2-7. 103 Guanzi 11/178/14. 104 I am grateful to Professor Riegel for helping me think through the concept of yic. 105 E.g. Xunzi 12/105, 15/47.

106 3. Orientation of Xin superior person follows Dao to put things into order, they have to deal with things as they are. Another observation about Xunzi’s Dao is that Dao remain unchanged throughout time. This idea is suggested by the following passage:

ðŽរࡾჽ˜lౣl᝶ංࡾ቗&࠿&&ං

!,û’ࡒ࠿Y

What has remained unchanged through the Hundred Kings is sufficient to be regarded as the connecting thread [guan] of Dao. With each rise and fall, respond with this connecting thread; [lia] the connecting thread, and there will be no disorder. If you do not know the connecting thread, you will not know how to respond to changing circumstances. The main part of this connecting thread has never perished.106

A key term in the above quote is “guan ,” which is usually translated as “the thread” or “connecting thread.”107 While the translation of “guan” itself is literal and appropriate, scholars in their interpreations of the term seem to emphasise the idea that guan is the fundamental tenet or theme of Dao. Edward Machle’s reading of “guan” as the “core” of Dao, for example, is along this line of interpretation.108 Most commentators follow Yang Liang in understanding guan in the above passage to refer to rituals. On this reading, this passage conveys the point that rituals are the fundamental tenet of Dao. However, a careful consideration of “guan” shows that it is more plausible to understand the emphasis of “guan” on connecting parts together. According to the Shuowen, “guan” originally refers to the string that is used to tie coins together. In this usage, guan clearly does not refer to what is fundamental to the coins but only to what connects the coins together. If we extend this observation to other usages of guan in the early texts, we find that the focus of the expression “X guan Y,” where guan is used as a verb, is often on how X continuously runs through

106 Xunzi 17/46-8; K 17.11/3/21. 107 See K 17.11/3/21, Watson, p. 91, Hutton, 2005, p.273. I am grateful to Professor Kwong-loi Shun for prompting me to investigate the concept of guan in the Xunzi. 108 Machle, 1993, p.128.

107 3. Orientation of Xin

Y so that X connects various parts of Y together. One clear example is found in the Guanzi, in which guan is discussed in relation to Dao:

“l”ࡾj÷*˜

Hearing one single word/doctrine so as to guan the ten thousand things is called knowing Dao.109

The author of the above quote is likely to be saying that knowing Dao involves making connections between things such that when one listens to a word or doctrine, she can draw a linkage to other things. It would be absurd if the above is recommending one to regard a word she hears as the fundamental tenet of all things. Another example that shows guan is primarily concerned with making connection comes from the “Quan xue” chapter of the Xunzi. It is said that:

Y

|^&•&–&រࡾჽ Q—ໟࡾژࡾL‚

The genetleman, knowing well that learning that is incomplete and impure does not deserve to be called fine, recites and enumerates his studies that he [guan] them, ponders over them and searches into them that he will fully [tong] them…110

Here, guan is discussed in parallel with tong. Since, as earlier discussed, tong denotes going through without impediment, guan in a similar sense should also refer to building up a connection. Therefore, the above quote is saying that the gentleman should connect what he has studied into a whole and think about his studies until he thoroughly understands them. If my understanding of guan is correct, the guan in the passage about Dao guan should not refer to the core of Dao but to the connecting together the kings’ knowledges of Dao. If this thread remains unchanged and can be applied to respond to changing circumstances, we can infer that Dao for Xunzi is unchanging. This is very different from saying that some standards or some parts of Dao will remain unchanged. Since Dao itself remains unchanged, one can apply the

109 Guanzi 26/156/9. 110 Xunzi 1/46-7; trans., K 1.14/1/142, modified.

108 3. Orientation of Xin same Dao across different situations over time. The proposed understanding of Xunzi’s Dao is also consistent with Xunzi’s criticism of those who make choices by themselves without observing Dao111 and those who argue that there should be a different Dao because circumstances in the past are different from the present.112 It should be noted that the claim that Dao remains unchanged across different situations and over time does not entail the claim that to act in accordance with Dao is to carry out the same set of actions across different situations and over time. Suppose a school has a high standard, that is, all students are required to achieve an A grade in all the exams to pass the course. In order to meet this standard, student X might have to work harder in semester two than she did in semester one because the materials she learned in semester two are more difficult for her and perhaps she was sick and had missed a week’s class. Similarly, if an agent is to act in accordance with Dao, the question of what action she is going to take largely depends on the content of Dao and the circumstances in which she finds herself. This is a different set of questions and is not of concern to the present discussion as it does not affect Xunzi’s claim that Dao is unchanging. All of the above further suggest that, for Xunzi, Dao is objectively valid. This interpretation finds support from the passage that we discussed earlier, which says that an individual who does not weigh options on Dao does not understand the “count” of things:

Ô&ßåæ ࡥࡾჽççæ പࡥࡾჽåãࡗࡥࡾഠ åç

‚˜ᒮੱ࡯ßÔᬻ˜êᕧຬ&åçæ‚ᬻ˜êᕧຬ

{ìࡾíîl©ಚ॥ïðటîlòóôჽ& ॥ໟY

If the counterweight is not correct, then misfortune may be entrusted to yu and deemed as fortune and fortune may be entrusted to dislike and deemed as misfortune. This is the reason people become deluded about fortune and misfortune. Dao, from the past to the present, has been correct counterweight…To abandon Dao and rather choose by own standards from within is like exchanging two for one. How could there be gain? Anyone who

111 Xunzi 22/72-4. 112 Xunzi 5/33; K 5.5/2/207.

109 3. Orientation of Xin

would actually exchange the yu that has accumulated for a hundred years for the dislike of a single moment does not understand the count.113

For Xunzi, those who do not follow Dao do not weigh options in accordance with

Dao and do not really know how to determine the “weight units ໟ” that each thing carries. The reference to numerical value gives the impressions that everything has a value attached. And since Xunzi condemns those who choose by their own standards on the basis of what their feelings of like and dislike of the moment are, we can infer that there is an objective value of things that is not influenced by an individual’s likes and dislikes of the moment. Dao, then, is what allows one to clearly see how much things are in fact worth. As a corollary to the view that Dao is not detached from how things are, Dao is also not detached from humans but incorporates the features that are characteristics of human beings (qinga), which includes characteristic human feelings. This is also why Xunzi thinks that Dao is applicable to everyone. In one passage, the sages are said to have knowledge of Dao so that they can differentiate the qinga of myriad things and lia what is so and not so of things:

*!eᒮL‘!˜ං&Yᡝ‘j÷ചZᒮ!˜ᒮࡾš

›£j÷ചZᒮࡾ቗&੩Ù

Those who are called great sages are persons whose knowledge has accessed the great Dao, who respond to changes but without exhaustion, and who differentiate the qinga and nature of myriad things. The great Dao is the reason for changes and transformations and the subsequent completion of things. Qinga and nature are the reason they can distinguish what is so and not so of things and what is to be selected and what is to be abandoned.114

Recall that Dao is the balance on which one decides what is to be selected and abandoned, while, in the above passage, it is said that one selects and abandons things

113 Xunzi 22/73-8; K 22.6b-c/3/137. 114 a

Xunzi 31/17-8; K 31.2/3/31. Although this passage contains the character li ‹, it is not evident that it is used in the sense that concerns our present discussion. It is used a verb to mean “to distinguish” or “to sort out.”

110 3. Orientation of Xin because of qinga and nature. This connection further supports the interpretation that knowing Dao involves knowing the characteristic features of things. The understanding that a regard for characteristically human feelings is embedded in Dao allows us to make better sense of Xunzi’s view of finding satisfaction on the basis of Dao. One of the recurring themes in the Xunzi is that Dao- conforming actions actually entail greater degrees of pleasant feelings, rather than requiring one to suppress or overcome their feelings. According to Xunzi, Dao should be the way for people to obtain the objects of yu.115 It is said that all human beings have security and honour as their objects of yu and yet, only the superior person can get constant security and honour. In another passage, it is explicitly said that great

116 security (!Î) and great honour (!x) is consequent upon the attainment of Dao.

Moreover, Xunzi draws an analogy between taking delight in the Dao of the sage and liking fine meat and grains. According to Xunzi, a person who had never encountered fine meat and grains before would consider these as “strange things” when they first saw them. But after they have tasted them, they would realise they are indeed pleasant to the senses and would from then on choose their new food and abandon the food they used to have.117 Here, Xunzi seems to take for granted the fact that abiding by Dao can facilitate harmonious obtainment of the objects of yu. The belief that a person can find greater satisfaction by doing the right thing is not alien to the trend of thought in Xunzi’s intellectual context, at least not when we put aside different thinkers’ conceptions of what constitutes right actions. According to the author of Lushi chunqiu, for example, succeeding in lia is the basis for obtaining

118 the four yu. In the Mencius, it is said that a person who is not worthy (xian œ) will not be able to find pleasure even when he has luxuries such as grand towers, ponds, and birds.119 My interpretation can also make better sense of Xunzi’s statement that “Dao is neither the Dao of Heaven nor that of Earth. Dao is the basis that guides people. It is the basis that guides the superior person.”120 Since Dao is a standard derived from certain characteristic features of human beings, humans who want to flourish may turn to Dao for guidance.

115 Xunzi 22/67-8; K 22.6a/3/136. 116 Xunzi 11/1; K 11.1a/2/149. 117 Xunzi 4/52-5; K 4.10/1.192. 118 Lushi chunqiu 5/4.2. 119 Mencius 1A:2. 120 Xunzi 8/24; K 8.3/2/71.

111 3. Orientation of Xin

The observation that following Dao entails greater satisfaction brings to light important assumptions Xunzi has made about Dao. Referring back to the previous passage, Xunzi says that if the balance is not properly aligned, one will deem misfortune for what they like, and vice versa. In another passage, Xunzi describes the lowest kind of people, the common people (G), as those who do not know the value of things:

༢ᢣຬ ÷&ž ÷Ÿ&rࡋ ჽß಴ ¡ã*ో

ࡥ¢

Day after day when they choose and select things, they do not know what is valuable. They follow things like streams of water, not knowing where to return. The wu zao (five senses/five passions)121 are in the centre. Xin follows and becomes corrupted. If one is like this, then he can be called a common person.122

The above passage reiterates the idea that there is a mismatch between what a person regards as valuable when one does not know Dao and what is in fact valuable. We can extract two assumptions. First, it is according to Dao that individual preferences are brought in line with what is in fact preferable. Second, since things have objective values, there is one best option prescribed by Dao in every situation. Laying bare these two assumptions help us to answer the question we begun with: How does one’s knowledge of Dao affect lu? We now come to see that, for Xunzi, how well one knows Dao determines how well one understands the value of things. For one who has a full grasp of Dao, the issue of choice does not even arise because they automatically know what is the right thing to do. This interpretation is consistent with Xunzi’s view that a person who knows Dao does not lu.123 Since she already knows the right option, she no longer has to choose between options in light of her feelings.

121 Interpreters differ in their understanding of wu zao. According to Yang Liang, wu zao could refer to the five viscera, with xin being one of them. In this sense, the passage is saying that although the senses appear to be functioning properly, xin is led astray by external objects. He also adds that wu zao could refer to the five feelings. Liang Qi-xiong favours the former, while Hao Yi-xing and Wang Nian-sun favour the latter. For our present purpose, it should suffice to note that what xin chooses is not necessarily align with what is in fact valuable. 122 Xunzi 31/8-9; K 31.2/3/260. 123 See discussion in section 3.2.

112 3. Orientation of Xin

In Chapter 1, we briefly discussed that the notion of “ke ” has both the connotations of “possibility” and “permissibility.” It should be clear by now that the differences between these two meanings diminish when xin is in line with Dao: what Dao does not permit will automatically be considered as an impossible option and what Dao permits will be considered as the only possible option. In the next section, I will discuss how it is possible for xin to know Dao.

3.4 Xin and Dao

Let us begin with the negative question of what prevents xin from knowing Dao. As mentioned earlier, Xunzi identifies bi  (obsession, partial knowledge) as the problem that prevents one from seeing the great lia and Dao.124 In another passage, it is said that the sages who dispel bi are able to uphold the correct balance. Presumably, the correct balance refers to Dao.125 We can infer from these that bi is why xin does not know Dao. The meaning of bi is variously interpreted in the literature. Yang Liang understands bi as one’s being stuck in a corner as if there is an object obstructing us from getting to the bright side or to the light.126 Burton Watson translates bi as “obsession.”127 Although Knoblock also agrees that bi refers to what obsesses us, he chooses to render bi as “blindness.”128 The difference between Watson’s and Knoblock’s understanding is subtle yet significant. Knoblock’s interpretation of bi focuses on the state of ignorance, which is due to a lack of knowledge. Watson’s interpretation of bi as “obsession” may be understood in two senses. In one sense, one’s being obsessed could be entirely due to feelings and have nothing to do with knowledge. One could well have a comprehensive knowledge of something but still be overwhelmed by feelings. For instance, a doctor could be well aware of the harmful effects of smoking. But since she is overwhelmed by her intense feelings for cigarettes, she cannot resist smoking. In another sense, one’s being obsessed could be due to inappropriately overwhelming feelings that result from a lack of knowledge.

124 Xunzi 21/1. 125 Xunzi 21/29-30. 126 See Wang Xian-qian, vol. 2, p.386. 127 Watson, p.125. 128 Knoblock 21/3/100.

113 3. Orientation of Xin

We can imagine a teenager who is overwhelmed by her intense feelings for cigarettes, but the reason she has such intense feelings is that she does know the harmful effects of smoking. While the former sense of “obsession” stresses only the role feelings play in bi, the second sense is compatible with Knoblock’s translation in that it understands bi as arising from both overwhelming feelings and a lack of knowledge. Given our previous discussions, Xunzi is most likely to be using bi in the second sense of obsession. It is clear that bi concerns a lack of knowledge of Dao. More precisely, bi refers to a partial knowledge of Dao. This can be inferred from the following two passages:

ჽപჽ£ჽ¤ჽ}ჽᡴჽMჽ࿹ჽੱჽ࡯ჽ

Yu is bi; the feeling of dislike is bi. The beginning is bi; the end is bi. What is remote is bi. What is near is bi. Broadness is bi; shallowness is bi. The past is bi. The present is bi.129

j÷ჽ˜ll÷ჽj÷l¥ᒮჽl÷lᕧࡾჽ˜

All the myriad things are one part of Dao. One object is one part of the myriad things. The fool operates on one part of one object and he thinks that he knows Dao.130

“Yu and dislike,” “beginning and end,”and “remote and near” etc. are all opposite terms that can be conceived as the extremes of a continuum. Xunzi seems to be saying that those who are bi are fixated on one end of the continuum and thereby lack the knowledge of a bigger picture of things. As a corollary, when bi is used in relation to Dao, it refers to a partial knowledge of Dao. But Xunzi seems to be saying more than this. In the text, bi is often linked to huo ഠ (uncertainty, puzzlement). This also means that when one lacks knowledge of Dao, she will not know which one is the right option to choose. There is a reference to tyrant Zhou’s bi for the attractive women Tanji and Feilian, and it is said that his heart is being huo.131 Recall Xunzi’s position that when xin does not know Dao, “misfortune may be entrusted to yu and

129 Xunzi 21/6-7; K 21.1/3/100. 130 Xunzi 17/50-1; K 7.12/3/22. 131 Xunxi 21/9; K 21.2/3/101.

114 3. Orientation of Xin deemed as fortune and good fortune may be entrusted to dislike and deemed as misfortune.”132 This means that xin would go by personal likes and dislikes to value things when it does not know Dao. And since xin is predisposed to like the objects of yu, it will assign more value to the objects of yu and in doing so, miscalculate the actual value of things. This is different from Chen Deng-yuan’s interpretation that understands Zhou’s bi as his xin being covered by the object and therefore not functioning.133 On my interpretation, what is problematic about bi is not that xin is dysfunctional, but that xin would be guided personal likes and dislikes because it does not have the knowledge of Dao. In other words, the kind of bi that Xunzi criticises is due to both affective and intellectual errors, with the former originating from the latter. Hence, the process of dispelling bi and teaching xin about Dao is a process of using knowledge of Dao to give correct guidance to xin, rather than restoring the function of xin.

e According to Xunzi, xu yi jing tl* are the means to know Dao. Since bi is what obstructs xin from knowing Dao, we can further understand xu yie jing as means to dispel bi. In light of this connection, I will build on the ideas that are extracted from my analysis of bi to inform my analysis of xu yie jing. Since Xunzi is likely to be intentionally redefining these three concepts that are already popular by his time,134 it is important for us to confine our understanding to Xunzi’s definitions, which are as follows:

Y

಴û’&ᕤྞ*t‚&ࡾఉᕤz੪*tY

Xin never stops storing; yet there is what is called xu. Not allowing what has been stored to interfere with what is about to receive is called xu.

~òÓíྞ*l&ࡾ^lzãl*u

[Xin] knows two aspects of things at the same time; yet there is what is called yie. Not allowing that “one” to harm this “one” is called yie.

132 Xunzi 22/73-4; K 22.6b-e/3/137. 133 Chen Deng-yuan, p.161. 134 See Appendix for discussions on the usages of xu and jing in other early texts.

115 3. Orientation of Xin

಴û’&਄ྞ**&ࡾ§৔࠿**

Xin never stops moving; yet there is what is called jing. Not allowing dreams and fantasies to bring knowledge into disorder is called jing.135

“Xu” could carry the multiple connotations of vacuity, vacancy, and deficiency.136 The term “xu” is frequently used in conjunction with kong (empty, unfilled)”137 and

138 contrasted with ying ፝ (surplus, filled) and shi ª (substantial, real). Note that ying is not just about being filled up but also about being abundant and sufficient. Similarly, shi refers to what is real or substantive. Both the concepts ying and shi require the presence of particular elements. Then, when xu is contrasted with ying and shi, xu should mean the absence of certain elements. This speculation is supported by textual evidence that uses xu to denote political or bodily weakness, implying the lack of certain elements in governance or physical health.139 With this connotation of xu in mind, we can better understand Xunzi’s concept of xu, which is not a kind of absolute emptiness; the focus is more on the absence of what has been stored. Verbally, the term “xu” means making room in xin available for use or occupation, presumably for taking in new knowledge. Xunzi’s concept of yie is commonly translated as “unity.” Bryan Van Norden translated yie as “single-minded.” There is a subtle but substantial difference between these two usages. Unity implies that things are brought together to become one; single-mindedness can mean devotion to one and only one purpose. If we look at Xunzi’s definition of yie, it explicitly says that a person can know more than one thing but it is important not to let one impair the other. In this sense, quite opposite to “unity,” Xunzi’s concept of yi emphasises the notion of keeping a certain seperation between things. More specifically, he means that certain knowledge should not be tangled with other knowledge and that there is only one kind of knowledge, that is, knowledge of Dao, to which one should direct full attention. The term “jing” has the connotations of being still and free from disturbances. As Kwong-loi Shun notes, jing, in the sense of opposite to dong, can mean complete

135 Xunzi 21/34-39; K 21.5d/3/104. 136 Xu might also take on the connotation of weakness or falsity. However, given Xunzi’s definition, it is unlikely that Xunzi’s concept of xu has these two connotations. 137 E.g., Lushi chunqiu 18/6.3; Guoyu 19/9/555; Mengzi 7B:12. 138 E.g., Yijing 23, 41, 55; Lunyu 7:26, Guanzi 26/158/14. 139 E.g. Liushichunqiu 3/5.2, 8/4.2, 18/6.3.

116 3. Orientation of Xin stillness. When jing is paired with qing, it can mean free from disturbances.140 In light of this observation, we can see that Xunzi’s definition of jing focuses more on the absence of disturbances rather than complete stillness. Dreams and miscellaneous thoughts are thought to be what disturb xin. Hence, Xunzi adopts the notion of jing to stress the point that erroneous thoughts should not confuse xin from knowing Dao. Common across the notions of xu, yie , and ying is the emphasis on not letting certain thoughts disturb zhic. If we recall the three different senses in which Xunzi uses “zhic, we can better understand Xunzi as saying how one can exercise her inborn capacity (zhi1) to know things (zhi2) in a way that is in line with Dao (zhi3). The key is not to let zhi2 to confuse one with zhi3. In this sense, we can understand xu yie jing as practices enabled by xin’s capacity to zhic. Note that Xunzi’s concepts of xu, yie, and ying only solve the question as to how xin can know Dao but leaves open the question as to why xin knows Dao. The latter will be addressed in Chapter 4. We can now return to the question of what makes xin lu. As my analysis in this chapter shows, xin naturally wants to pursue the objects of yu. The activity of lu itself is not necessarily activated by a moral concern. Any person who wants to better satisfy his wants and interests and has the knowledge that there are certain things that cannot be concurrently achieved would start to lu and make choices. But such kind of knowledge is not necessarily knowledge of Dao. This is precisely why Xunzi worries. If one’s knowledge is not in line with Dao, her xin will be guided by its natural inclination to misjudge the value of things and therefore make wrong choices. It is only when one’s capacity is realised in a specific way, namely, when the capacity is developed in the direction of Dao that a proper choice can be made. This also allows us to see that there is no dichotomy between an affective xin and a cognitive xin or between a human xin and a moral xin. It is the same xin making decisions all along. It is in virtue of the inborn capacity to zhic that xin can lu. The natural response and the modified response are not issued by two parts of xin with one trying to surpass another. The two-heart view is employing very misguided vocabularies that understand xin and yu as antagonistic or sees an emotive xin as opposed to a moral xin. A moral response of xin is made in virtue of xin’s capacity to zhic being realised in a specific way, that is, when Dao is the object of zhic.

140 Shun, 2006, pp.198-9. See Shun’s article for a detailed analysis of the concepts xu and jing.

117 3. Orientation of Xin

Hence, the difference between making a good decision and a bad decision lies in whether xin is guided by Dao, rather than in whether a moral aspect of xin has overcome a problematic aspect of xin. This is consistent with Xunzi’s view that xin is the ruler of the body and that xin never stops moving in virtue of the fact that xin inherently has zhic.141 On this point, I find myself in agreement with Xu Fu-guan. According to Xu, xin does not issue morality. Rather, xin only issues actions, which can happen to be in accord with ethical ideal.142 In the next chapter, I will build on this understanding of xin to examine Xunzi’s view on xin’s basis for judgements.

141 See section 3.2.

˜ ™

8 A–— A–—

’“” B

142 `

U • _  ‘ Hsu, p.240.

118 CHAPTER 4

Basis for Xin’s Judgement

4.1 The Issue

This chapter raises the question: why xin abides by Dao? I approach this question by way of another question: is there a change in the basis for xin’s judgement when xin complies with Dao? I will discuss five major groups of interpretations in order to make explicit the key points of debate that complicate what is at issue. I attempt an interpretation that resolves such disagreements and, at the same time, keeps modification to Xunzi’s thought to a minimum. I will argue that xin does not necessarily change the basis of judgement when it complies with Dao, but there could be a change in the basis provided that xin receives the right training. What enables Xunzi to hold this view is his assumption that it is in fact the case that only following Dao provides satisfaction. To recapitulate, I explicated in the previous chapter Xunzi’s view that xin has inborn preferences and, for this reason, xin by default values and chooses the objects that it naturally prefers. Xunzi thinks that for every decision there is an objective value. This objective value is rooted in Dao. In virtue of xin’s capacity to zhic , it is possible for xin to know Dao, which is also to correctly judge the objective value of things and choose the right object. We learn from this that there is an initial mismatch between xin’s judgements and Dao. We also learn that xin’s judgements could be in line with Dao. Suppose X is in fact heavier than Y. When there is a mismatch, xin could have judged that Y is heavier than X; but once xin is in accordance with Dao, xin will judge that X is heavier than Y. An important implication of this view is that when xin’s judgements are in line with Dao, xin necessarily values things in a way that is in accordance with Dao. What is unclear, though, is why xin’s judgements are brought in line with Dao. We can frame our question in terms of Xunzi’s scales metaphor. We learn that xin is originally an inaccurate scale and that this scale can be adjusted to become an accurate scale through knowing Dao. What we do not know is why such a process of adjustment takes place. Since xin’s knowledge of Dao is what 4. Basis for Xin’s Judgement enables it to be in accordance with Dao, the question at issue hinges on our understanding on why xin knows Dao. Although we know that xin’s capacity to zhic is what enables it to know Dao, it is still unclear why xin would start exercising its capacity of zhic to know Dao, given that xin does not necessarily exercise its capacity of zhic in a way that takes Dao as its object.1 How we understand xin’s reason for knowing Dao, in turn, will determine how we understand the connection between xin’s knowing of Dao and xin’s abiding by Dao as an end in itself. The questions I am now raising are different from the ones I dealt with in the previous chapter because I am no longer asking how it is possible for xin to abide by Dao, but rather why xin would abide by Dao. The difference between these two questions is analogous to the difference between asking how it is possible for someone to learn French, on one hand, and asking why someone would learn French, on the other. The former hinges on the capacity that enables one to learn French, while the latter on the rationale behind one’s learning of French. It is questions of the latter kind that concern our present discussion. Indeed, the questions I raise here are by no means novel, but ones that have been widely explored by scholars. However, in studies of Xunzi, the issue of why xin abides by Dao is often discussed along with Xunzi’s view on human nature, with the intention to use the former question to make sense of Xunzi’s view of human nature. One difficulty with this common approach is that scholars tend to discuss two open-ended issues in tandem. It is often unclear whether the issue of why xin abides by Dao is the subject matter of investigation guided by Xunzi’s view on human nature or the other way around. In view of this problem, my strategy is to address these two issues separately. To give a focus to my present discussion, let me hereby clarify the objective of my investigation at the outset. The object of inquiry in this chapter is confined to the issue of why, for Xunzi, xin, which does not initially abide by Dao comes to abide by Dao. My goal is neither to defend nor challenge Xunzi’s view on human nature. I will put aside Xunzi’s claim that human nature is bad and base my interpreation on the results of my earlier discussion. My strategy is to bring to light the assumptions Xunzi would need to adopt in order for him to remain consistent in his thought. For this reason, it is possible that Xunzi himself might not be interested in certain problems I explore or fully aware of the implications I draw from my analysis of the text.

1 For example, the content of the petty person’s zhic could be very different from the content of the superior person’s zhic. See section 3.1 for discussion on zhic.

120 4. Basis for Xin’s Judgement

We can begin our inquiry into these questions by way of a more fundamental question: what is the basis for xin’s judgement? By “basis,” I refer to xin’s motivation when it makes judgements. Earlier in section 2.1, I drew a distinction between the reason (X) for xin’s choice and the reason xin has X as its object of choice. It is with the latter that the present discussion concerns. We saw that it is possible for xin to choose an object O because xin knows that it is the correct option in accordance with Dao. However, we are still not sure at this point as to why xin would have “choosing what in accordance with Dao” as the reason for its judgements. Since I would like to avoid imposing on Xunzi’s thought the framework of desires and moral motivation found in contemporary philosophical discourse, I will use the term “basis” to refer to the motivating reason xin has in abiding by Dao. The basis of xin’s judgements is worth examining because even when xin’s judgements are in line with Dao, it is not necessarily the case that xin likes Dao or regards Dao as an end. It is possible that xin has some prudential interests that can only be fulfilled by knowing Dao. So, even when xin makes judgements in accordance with Dao, it could still be the case that xin is judging things on the basis of wanting to satisfy personal interests. Similarly, even when xin’s judgements are not in fact in accordance with Dao, xin could still be making judgements on the grounds that its own judgements are in accordance with Dao. It is possible that xin has certain dispositions structured in a way that direct it towards following Dao. In this case, even when xin is initially not in line with Dao, it could still be judging things on the basis of its inclination to follow Dao. It is just that xin gets the content of Dao wrong because it does not know Dao. In my discussion of the basis for xin’s judgements, I will consider the only two available possibilities for us to understand the basis for xin’s judgement according to Xunzi. I will discuss this in detail in section 4.3. At this point, let me briefly introduce and label these two bases to facilitate my subsequent discussion. One possible basis for xin’s judgements is xin’s goal to maximise satisfaction, which I call L (self-interested choices prompted by consideration of lic). Another possible basis for xin’s judgements is xin’s goal to abide by Dao, which I call Y (Dao-abiding choices prompted by consideration of yia). These two bases are not necessarily exclusive of each other. L and Y can co-exist; L and Y can be competing bases when they co-exist; L and Y can preclude the other. L can be an initial basis whereas Y is an acquired basis; Y can be transformed from L. In order to bring to light the complexities involved in the question at issue, I will set up five possible

121 4. Basis for Xin’s Judgement interpretations of the role L or/and Y play in xin’s abiding by Dao. Although my discussion of these interpretations is framed in terms of the relation between the basis for xin’s judgements and xin’s abiding by Dao, they are based on various positions advanced in the literature. One interpretation I will not elaborate on here, but is widely held in the literature, is one that which understands yu and xin as two sources of motivation and consequently understands xin’s complying with Dao as a result of xin overriding yu. As we have already discussed in Chapter 2, this conception of xin and yu is misguided because it takes xin’s approval and yu to be on the same level as potentially competing causal roles in action. For Xunzi, xin’s approval and yu are in two different categories. Since yu itself cannot cause action and since only xin can decide on the objects of action and cause action, the issue does not arise of whether there is another yu that constitutes a basis that competes with the basis for xin’s judgements. This, however, does not deny the possibility that yu can be a basis for xin’s judgement. There is no contradiction in saying that yu cannot cause actions and that xin has the goal to satisfy yu. Since I have already separated the question of how xin and yu are related from the question of what the basis for xin’s judgements is, I will continue to examine interpretations of the latter issue while bracketing their views on the former issue.2 In order to avoid imposing any AI presuppositions about xin and yu on our discussion of the basis for xin’s judgements, I will avoid the common expression “desire-satisfaction” and use the expression “personal satisfaction” instead. I will further justify my use of this term in section 4.3. For convenience, I will occasionally refer to Dao as “moral standards,” actions that abide by Dao as “moral actions,” judgements that are in accordance with Dao “moral judgements,” and the process of training xin to abide by Dao as “moral training.” Again, I use the term “morality” to refer only to what is in accordanc with Dao in Xunzi’s view.

4.2 Five Interpretations

The first interpretation understands xin as making evaluations on a basis that seeks to maximise satisfation (L) and that this basis remains unchanged even when xin’s

2 See section 1.2 for the distinction between these two questions.

122 4. Basis for Xin’s Judgement judgement abides by Dao. On this view, maximisation of personal satisfaction is always the goal of xin. Since xin has the capacity to calculate which action will yield optimum satisfaction, xin can learn from moral training that observing the standards of Dao is the only way to maximise satisfaction in the long run. On this view, xin abides by Dao for prudential reasons. This does not necessarily mean xin takes delight in following Dao itself nor that xin has the goal to abide by Dao. For xin, Dao is instrumentally good because it optimises personal satisfaction. However, this does not preclude the fact that this is also the objective, right thing to do. David Nivison and Lee Yearley offer accounts of xin that are similar to the first interpretation.3 The second interpretation also acknowledges that xin initially adopts L but maintains that xin has certain innate dispositions which, with proper guidance and development, can eventually adopt Y. On this view, Dao is inherently congenial to xin and xin has a nascent knowledge of Dao. If xin effectively focuses its natural liking for Dao and expands its nascent knowledge of Dao, it will judge things in accordance with Dao and follow Dao. Xin starts off observing L because self-regarding interests initially obscure xin’s liking for and knowledge of Dao. Through moral training, xin’s liking for and knowledge of Dao can be expanded and entrenched against self- regarding interests so that L no longer dominates Y. This interpretation implies that xin inherently has L and Y and they are two competing bases for xin’s judgement. Tang Jun-yi and Eric Hutton have proposed interpretations along this line.4 The third interpretation also holds that xin has inherent predispositions that direct it towards abiding by Dao but maintains that xin has no nascent knowledge of Dao. Once xin recognises that Dao is in actuality the Dao it naturally likes, it will automatically follow Dao. So, even though xin always wants to be in line with Dao, the content of xin’s knowledge can be wrong. Moral training is required to constantly develop and refine the knowledge of xin until it becomes completely in line with Dao. This allows for the reading that Y is the only basis for xin’s judgement throughout. Even though xin observes Y, it does not inherently know Dao and will therefore mistake Dao for many other things. This means that when xin does not know what

3 According to Nivison, for example, Xunzi is essentially concerned with the optimum satisfaction of human desire. Since xin is the faculty of intelligence, xin will choose courses of action that best satisfies the total set of desires. When people make a bad choice, “they are simply confused: in a sense they have failed to add and subtract, and take one immediate satisfaction in place of many long-term satisfactions.” See Nivison’s “Can Virtue Be Self-Taught?” and “Xunzi on ‘Human Nature’” in 1996c. See Yearley, 1980. 4 See Hutton, 2000, especially pp.220-5. See Tang, 2005a, Chapter 4 and 2005b, Chapter 2.

123 4. Basis for Xin’s Judgement

Dao is, even if it observes Y, it will make wrong judgements. The content of Y could be the wrong content that xin mistakes for Dao. The point of moral training, then, is to make xin know Dao so that it can refine or correct the content of Y. In this sense, the reason xin abides by Dao is that it removes whatever obscures its knowledge of Dao so that it can properly define the content of Y. On this interpretation, Dao is intrinsically good for xin. Yang Yunru ±« and David Wong hold positions similar to this.5 The fourth interpretation regards L as the original basis for xin’s judgement and maintains that L can be transformed into Y. Like the first interpretation, this interpretation also understands xin as initially abiding by Dao for prudential reasons. Nonetheless, it also goes further to maintain that xin can come to appreciate Dao through its being aligned with Dao and therefore adopts Y. This interpretation does not ascribe to xin any predispositions that direct it towards Dao. It maintains that it is through moral training that xin finds delight in Dao and takes Dao as an end in itself. I formulate this interpretation on the views advanced by P. J. Ivanhoe and T. C. Kline.6 The fifth interpretation does not suppose that there is any connection between xin’s following of Dao and xin’s basis of judgement. It maintains that L is always the basis of judgement but xin does not have to judge on its own basis. When xin’s judgement is in accordance with Dao, it does not mean that xin is adopting a new basis Y nor that its judgement is made on L. Rather, quite simply, xin is brought into accordance with Dao through moral training and therefore disregards its own basis for judgements. This means that there is no basis in xin that supports xin’s abiding by Dao. This view is different from the first interpretation. On the first interpretation, the reason xin abides by Dao is that it knows Dao provides satisfaction and hence is compatible with its basis L. On the fifth interpretation, the judgement of xin does not have anything to do with its basis L. The question of whether or not it is compatible with L becomes irrelevant. For this interpretation to hold, the source that makes xin

5 According to Yang Yunru, there are two aspects of xin, one “material” and one “spiritual” (Yang, p.52). Yang has not elaborated on this distinction but it is implied that the “spiritual” aspect of xin is what enables xin to abide by Dao because elsewhere, Yang understands xin’s lu in terms of “a positive and brave spirit,” “of wanting the self to choose good” (Yang, p.46). Yang’s view is similar to Tang Jun-yi’s view in the sense that Yang also identifies these two aspects of xin in terms of a “human xin” and “a xin of Dao.” Nonetheless, Yang maintains that these two aspects belong to one xin (Yang, p.194). David Wong takes Xunzi to be implying that there are certain natural feelings in one’s nature that are congenial to morality. See Wong, “Xunzi on Moral Motivation.” 6 See Ivanhoe, “Human Nature and Moral Understanding in Xunzi,” see especially pp.173-174; Kline, “Moral Agency and Motivation in the Xunzi.”

124 4. Basis for Xin’s Judgement comply with Dao must completely be external. There must be some force that comes from moral training that can make xin disregard its own basis and judge things in accordance with Dao. This interpretation is perhaps the least favoured in the literature but this line of thought can be traced in the writings of Xu Fu-guan and Wei Zheng-

7 tong ¬­L.

We can flesh out four key points that stand at the centre of debate in these five interpretations:

1. Whether there is one basis or two bases for xin’s judgement; 2. Whether xin has to adopt Y in order for it to make judgements that abide by Dao; 3. If xin adopts Y, whether xin has to have a preference for abiding by Dao in order for it to adopt Y; 4. If xin changes its basis from L to Y, why xin adopts Y.

It can be argued, as in the case of the first interpretation, that there is only one basis for xin’s judgement, namely, the goal of seeking satisfaction. Hence, xin does not have to adopt Y in order for it to make moral judgements. According to the third and fifth interpretations, there is just one unchanging basis for xin’s judgements. Contrary to the first interpretation, the third interpretation takes Y to be the basis xin is observing. The fifth interpretation takes L to be the unchanging basis for xin but sees no connection between xin’s basis and xin’s abiding by Dao because it understands xin as being brought into accordance with Dao by external forces that come from moral training. The second and fourth interpretations endorse the claim that it is possible for xin to make moral judgement on L but maintain that it is also possible for xin to make moral judgements on Y. In this sense, there could be two bases for xin’s judgements, L and Y. The introduction of Y into the picture further begs the question as to why xin would change its basis given that it begins with L. The second interpretation ascribes to xin a predisposition to adopt Y. The fourth and fifth interpretations do not ascribe such predisposition to Y but understand xin’s acquisition of Y as entirely a result of moral training. The difference between the fourth and fifth

7 Both Xu Fu-guan and Wei Zhengtong seem to think that xin is passively brought into accordance with Dao by moral training. See Xu, p.246; Wei, p.71.

125 4. Basis for Xin’s Judgement interpretations is that while the former maintains that xin’s adoption of Y is constituted by xin finding satisfaction in doing so, the latter does not postulate a connection between xin’s adoption of Y and xin’s finding satisfaction in adopting Y. Note that, on the fourth interpretation, the kind of satisfaction xin obtains from adopting Y is transformed from L, suggesting that the two kinds of satisfaction are not necessarily independent of each other. This is very different from the second interpretation, which understands the kind of satisfaction xin finds in adopting Y as stemming from its natural disposition to find satisfaction in abiding by Dao. Hence, the kind of satisfaction obtained from xin’s adopting of Y is altogether different and independent of the kind of satisfaction obtained from adopting L. The difference between the second and fourth interpretations tells us that even if xin finds satisfaction in adopting Y, the nature of this satisfaction can be understood differently. Since the second and fourth interpretations maintain that xin can both adopt L and Y, they also need to deal with the question whether xin will revert from adopting Y to adopting L. And since the second interpretation understands that both L and Y can exist simultaneously in xin, this further raises the question as to whether L and Y are competing bases or whether xin’s adopting of Y precludes the possibility of xin’s adopting of L. It is in light of these four questions, which are not fully addressed in the above five interpretations, that I propose my own interpretation.

4.3 Two Bases

Before I propose my interpretation of why xin abides by Dao for Xunzi, let me address two points of debate: (1) whether there is one basis or two for xin’s judgements and (2) whether xin has to adopt Y in order for it to make judgements that abide by Dao. I will argue that there can be two bases for xin’s judgements and that xin has to adopt Y to make genuine moral judgements. Recall that we established in the previous chapter that xin has a predisposed

c inclination to li + , namely, things that are considered as beneficial. And the things that are often taken by xin to be beneficial are what satisfy the agent’s feelings.8 Xunzi sees a problem with xin making choices in a way that is guided by feelings, mostly yu, because it will fail to see the objective value of things. For Xunzi, it is

8 See section 3.1.

126 4. Basis for Xin’s Judgement through knowing Dao that xin can correctly judge the objective value of things and make the right choice, instead of being guided by feelings. Complying with what is required by Dao might even mean that xin has to choose what is naturally the most detestable, such as death. Now, let us turn our attention to the difference between lic

a iand y ᑶ.

In the Xunzi, lic is often juxtaposed with yia.i When l c is contrasted with yia, the emphasis is usually on which one should be the underlying consideration or motivation of an action. In the following passage, for example, Xunzi distinguishes actions made on considerations of lic from actions made on considerations of yia:

ß+ჽ*®ßᑶჽ*¿

Acting on consideration of lic is called shib (business); acting on consideration of yia is called xingb (action).9

Another observation about the difference between lic iand y a concerns the differences in actions one takes. When one is motivated by lic, one will do anything; on the contrary, when one follows yia, there are certain things that one will not do regardless. The following two passages stand in stark contrast to each other:

¯+­&5ᗆ{*}ࡥ¢

Where only lic is, there is nothing that does not tilt [him/his xin]. Then this can be called a petty person.10

ᑶ­&5 Ô&B॥+ᕺ±ᕸ&ჽ²³O›ฆᑶ&µ{_

|ৼ

Where yia is, [one/one’s xin] is not tilted by power, does not care about [his/its] lic, does not change standpoint even when offered the whole state, uphold yia

9 Xunzi 22/4-5; K 22.1b/3/127. I follow the common interpretations in taking the term “zheng” to mean “considerations” or “decisions” and the term “xingb” in this sentence to mean good actions rather than any actions in general. 10 Xunzi 3/44; K 3.11/1/180.

127 4. Basis for Xin’s Judgement

and does not bend even though [he/his xin] takes death seriously. This is the courage of scholarly superior person.11

Xunzi describes the difference between someone who follows lic and someone who

a b follows yi in terms of “qing 5 (tilt).” Moreover, when Xunzi describes someone who abides by yia, he sometimes describe the person as having a xin that does not tilt,12 as opposed to his description of a petty person having a xin that tilts.13 We can then infer that, in the above passages, “qingb” refers to whether or not the xin of the superior person or the petty person is tilted. This further suggests that whether one is being motivated by lici or y a correlates with whether one’s xin is tilted. Let us consider what else Xunzi has said about a “tilted xin.” In one passage, Xunzi relates one’s not

a making decisions and choices on si · and being able to correctly identifying right and wrong to their not having a tilted xin. The passage states:

&¸ྞࡾ·7ຬ&¸ྞࡾ·੩ᕸ…¹º෇¼ೂ5ࣿ಴{¿

¯{œ¿¯œ…

Not daring to make choices and decisions on the basis of sia; not daring to select and offer on the basis of sia…firm and resolute, correct in intention, and not having a xin that tilts, saying that what is right is right and that what is wrong is wrong…14

In other words, sia should be the basis for a tilted xin to make decisions and choices. The term “sia” is usually translated as “private” or “personal.” In the text, the concept

a of si is frequently contrasted with gong À (public, impartial) to convey the point that

11

Xunzi 4/19-20; K 4.4/1/188. I understand that term “quan † ” in this content to refer to political or

social power held by authority and influential people. Xunzi adopts this sense of quan when he uses the

x š † compound expression “quan li † ” and “quan shi ” (e.g. Xunzi 1; Xunzi 22). The usage of quan in this passage is therefore different from the sense of weighing options we discussed earlier.

12 E.g. Xunzi 12/109, 13/27.

d› œ

13 6 } Xunzi 3/18 _ . 14

Xunzi 13/25-6; K13.4/2/201. My translation of zhi  is only tentative.

128 4. Basis for Xin’s Judgement one cannot place what is personal before what is at stake for the public.15 Sia is also used in conjunction with yu  to refer to feelings that pertain exclusively to the self.

The emphasis of sia in Xunzi’s usage is not so much on what belongs to a person or what is not shared with others, but more on what benefits or favours the self in a way that disregards others. This also explains why Xunzi thinks that one who relies on sia

16 17 is partial in their view of things and cannot see things in a proper light ᫈. In light of these observations, the basis for a tilted xin to make choices and decisions should be one that is only concerned with what favours the self, namely, lic. In this sense, when lic is juxtaposed with yia, its emphasis is specifically on benefits that pertains only to the self, rather than benefits in general. What is the basis for a xin that does not tilt then? According to Xunzi, a xin that does not tilt is able to correctly identify right and wrong.18 Since it is only when xin is in accordance with Dao that one can disregard personal feelings and identify what is in fact right and wrong, the basis for a rectified xin is most likely to be Dao. And since Dao has incorporated what is characteristic of human beings as a kind, it fits with the idea that a rectified xin can take the whole or public (gong) into consideration.19 If I am right in understanding Dao as the basis for a xin that does not tilt, and given the xin that does not tilt follows yia, we can infer that yia involves disregarding benefits that pertain only to the self and abide by what is in accordance with Dao. That the notion of yia requires one to disregard benefits that pertain only to

a the self is consistent with the idea that the superior person uses “gong yi À ᑶ” to

a 20 trump “si yu ·.” The requirement that one must be in accordance with Dao is supported by Xunzi’s view that the sages create rituals and yia in order to bring people in line with Dao.21 With the understanding that yia, for Xunzi, is being in accordance with Dao, we can make sense of the quality of firmness displayed by someone who follows yia. As we have already seen, a superior person who follows yia will not change his standpoint even when he is being offered immense benefits such as wealth

15 E.g. Xunzi 8/121, Xunzi 13/45. 16 Xunzi 21/3-4. 17 E.g. Xunzi 3/40-1, Xunzi 12/86. 18 Xunzi 13/27 (K 13.4/2/201); Xunzi 21/27 (K 21.7a/3/107). 19 See discussion in section 3.3. 20 E.g. Xunzi 2/49, 12/59. 21 Xunzi 23/6-8 (K 23.1b/3/152); Cf. Xunzi 8/24 (K 8.3/2/71).

129 4. Basis for Xin’s Judgement and power. In the Xunzi, a person who relies on yia is often described as firm and does not bend.22 This is so because a person who relies on yia abides by Dao. As discussed before, when one acts in accordance with Dao, there is only one right option in every situation. For this reason, unlike those who follow lic and can therefore do anything to pursue what benefits them, there are certain things that those who follow yia simply will not do. It should be noted that there is a distinction between one’s firmness in abiding by Dao as an end in itself and one’s rigidity in action. It is exactly because of this firmness in abiding by Dao as an end in itself that one has to be flexible in action. For example, a filial son is required to disobey his father where appropriate because following yia is more important than following his father.23 Hence, what Xunzi finds problematic is not one’s being flexible in actions as such but one’s being flexible in a way that is unsupported by yia.

c Another interesting observation is that, for Xunzi, whether one follows li +

a or yi ᑶ is determined by xin. In one passage, it is implied that xin is the reason one only sees lic and does not know yia:

GG…OG !"ড$… 24' 6&ן6 Humans are originally born as petty person. Without teachers and models, they will only see lic….Now how can the mouth and stomach of a person know libi y a?...without teachers and models, a person’s xin is just like his mouth and stomach.

According to the above passage, it is when xin is akin to the senses that one only sees lic and does not know yia. This suggests that xin can determine whether one follows lic or yia . To Xunzi, if xin does not exercise its capcity to zhi, the capacity that distinguishes it from the senses, then it is just like the senses. Also, there is the reference to a superior person making his xin sincere when following yia.25 Furthermore, Xunzi draws a distinction between honour that results from one’s

22 E.g. Xunzi 13/27, Xunzi 30/8-9. 23 See Xunzi 29/3-6 for the three circumstances in which a son should disobey his father. 24 Xunzi 4/50-2; K 4.10/1/192. 25 Xunzi 3/27. For a contrast with petty person’s lack of sincerity from within, see Xunzi 27/84.

130 4. Basis for Xin’s Judgement

a following yi (ᑶx) and honour that results from one’s having certain social and political influence (Áx). He also distinguishes between disgrace that results from

a one’s failure to follow yi (ᑶᡠ) and disgrace that results from one’s being subject to certain social punishments (Áᡠ). According to Xunzi, the kind of honour or disgrace

a that concerns yi comes from within (zhong ¨), whereas the kind of honour or

26 disgrace that concerns social recognition comes from without (wai ù). Since xin is understood in the classical texts as what is within, I assume that Xunzi’s usage of “zhong” also refers to xin and, accordingly, infer that xin determines whether or not one follow yia.27 I have shown that xin can either judge on the basis of personal benefits or on what is in accordance with Dao. I have also shown that xin itself determines which basis is to be adopted. Hence, I understand Xunzi to be implying that there can be two bases for xin’s judgements, which are incompatible with and exclusive of each other. One basis is concerned solely with obtaining benefits for oneself and having no regard for others. It is in this sense that I call it the basis of seeking personal satisfaction (L). The other basis is concerned with abiding by Dao and disregarding benefits that pertain only to the self. Hence, I call it the basis of following Dao (Y). In light of this distinction, it is clear that L is the basis xin naturally adopts whereas Y is acquired from moral training, for Xunzi explicitly says that it is of human nature to prefer lic and that yia is created by the sages and can only be acquired from moral training.28 Since Y is an acquired basis, it is unlikely that Xunzi would think there is a predisposed preference for Dao. The difference between L and Y also allows us to draw a distinction between apparent moral judgements and genuine moral judgements. Suppose one’s xin makes judgements that are in accordance with Dao. If these judgements are made solely on considerations of what benefits the self (for example, following rituals makes me happy in the long term), it is difficult to imagine how

26 Xunzi 18/104-7; K 18.9/3/46-7. The view proposed here is very different from Cua’s understanding of xin and shame. In Cua’s view, Xunzi has not made a connection between xin and shame (2005, p.222). 27

See appendix for discussion of the connection between zhong ž and xin in other classical texts. 28 Xunzi 23/1-5. See discussion in section 5.3 concerning difficulties with understanding what Xunzi means by humans have yia. At this point, it suffices to note that xin can adopt two different bases.

131 4. Basis for Xin’s Judgement

Xunzi would regard such actions as genuine acts of abiding by Dao because the very notion of abiding by Dao as an end in itself requires one to disregard benefits that only pertain to the self. Hence, only judgements derived from Y can be called genuine Dao-abiding judgements. This further explains why the fact that humans have yu does not bother Xunzi.29 As long as xin adopts Y, xin will not make decisions in light of personal feelings. It is worth noting that even if xin has acquired Y, it does not seem to guarantee that xin will always adopt Y. This is suggested by Xunzi’s emphasis on placing yia beforei l c30 andi using y a ito control l c.31 In one passage, it is said that lic andi y a are the two things humans have. While good rulers can make their people’s preference for yia trump their preference for lic , bad rulers do the opposite.32 If the adoption of Y does not preclude the possibility of adopting L, this raises the question as to why there is a possibility for xin to revert to L. In the following, I will propose an interpretation of why xin adopts Y. My strategy is to take the results of my analysis as given and make a conjecture about what else Xunzi could have implied in order for us to make sense of his claim that xin can shift from initially adopting L to adopting Y.

4.4 Proposed Interpretation

We learned that xin originally adopts L and that it has no innate preference or knowledge of Dao. Xin is capable of acquiring Y because it is capable of knowing Dao. The knowledge of Dao is acquired from moral training. This suggests that moral training has some role to play in creating a new basis, Y, in xin. We can lay out five premises about why xin adopts Y in the following:

(1) Xin has the innate goal of seeking satisfaction; (2) Xin has no innate preference or knowledge of Dao; (3) Xin can know Dao through moral training; (4) Xin knows Dao; (5) Therefore, xin adopts Y.

29 Xunzi 22/55-62. 30 Xunzi 4/22. 31 Xunzi 18/57-8 32 Xunzi 27/65-8.

132 4. Basis for Xin’s Judgement

Once these premises are laid out, it becomes clear that there is still a gap between (4) and (5) that needs to be filled. Why does knowledge of Dao entail xin’s adoption of Y? It is difficult to see why Xunzi would think that knowledge of Dao necessarily entails xin’s abiding by Dao as an end in itself. It cannot be the case that xin naturally prefers Dao so that xin abides by Dao once it knows Dao. It also cannot be the case that xin abides by Dao because xin knows that abiding by Dao is the way to achieve long-term satisfaction. It is also unlikely that moral training can forcibly bring xin to abide by Dao because Xunzi explicitly says, “Xin cannot be forced to change its

33 intentions Â.” It seems that the only way we can make sense of this shift from xin’s adopting of L to its adopting of Y is for us to add the following three premises (premises 5, 6, and 7):

1. Xin has the innate goal of seeking satisfaction; 2. Xin has no innate knowledge of Dao; 3. Xin can know Dao from moral training; 4. Xin knows Dao; 5. It is a fact that the only way to obtain satisfaction is to abide by Dao (i.e. adopting Y) as an end in itself; 6. Xin has no innate knowledge of 5; 7. Xin knows 5; 8. Therefore, xin adopts Y as an end in itself.

Premises (5) to (7) are the three assumptions I believe Xunzi would need to adopt in order for him to fill the gap between xin’s knowing of Dao and xin’s adopting of Y. 34 The key assumption I ascribe to Xunzi is (5): it is a fact that abiding by Dao as an end in itself is the only way to obtain satisfaction. If xin knows that abiding by Dao as an end in itself, that is, adopting Y, in fact provides satisfaction but does not know that adopting Y is in fact the only way to obtain satisfaction, xin could be merely abiding by Dao to fulfil its own goal of seeking satisfaction. In this sense, abiding by Dao is a means for xin to achieve the goal of seeking satisfaction. Hence, even if xin makes

33 Xunzi 21/46; K 21.6a/3/105. 34 I am indebted to an anonymous examiner whose comments significantly help me to articulate these premises more clearly.

133 4. Basis for Xin’s Judgement judgements that are in accordance with Dao, they will be apparent moral judgements made on the basis of L. But when xin knows that abiding by Dao as an end in itself is in fact the only way to obtain satisfaction, xin necessarily takes on the goal of abiding by Dao. Given its goal now is to abide by Dao and since it has already obtained knowledge of Dao from moral training, xin adopts Y and makes genuine moral judgements that are in accordance with Dao. Let me explain in detail why xin’s knowledge of the fact that abiding by Dao as an end in itself is the only way to provide satisfaction entails xin’s taking on the goal of abiding by Dao without having to attribute to xin an innate preference for Dao. Imagine that I am travelling in a foreign city and I want to have French food for dinner. Suppose I do not know the fact that Café D is the only French restaurant in town. I start looking for dining options in the travel guides and discover Café D, a restaurant is highly recommended by food critics for its French cuisine. I follow the directions given in the guidebooks and find Café D. I see its décor and look at its menu but I am still not sure if this is the restaurant I want. I continue to stroll around to see if there are some other options until I learn that Café D is the only French restaurant in town.35 The moment I know that Café D is the only French available is also the moment I decide that Café D is where I want to go. I go to Café D and meet two other tourists, Anne and John. Anne tells me that the goal of her trip is to go to Café D because the place has special meaning for her. But since she has never been to this city before and since she has no information about the restaurant with her, she has run into many wrong places before she finally finds Café D. John, on the other hand, just wants some French food for dinner (like me). He walks by Café D and realises it is a French restaurant, so he enters Café D. Obviously, John is the only one among the three of us who does not have the goal of going to Café D. To John, going to Café D is simply a means to achieve his goal of having French food. It is correct to say that, at one point, both Anne and I have the goal of going to Café D. Nevertheless, there is a significant difference in terms of why we take on this goal of going to Café D. I initially just had the goal of going to a French restaurant, whereas Anne has always had the goal of going to Café D. It was only later when I knew that Café D is the only French restaurant available that I decided I wanted to go to Café D. Our difference is that an extra step is involved

35 See section 4.5 for a discussion of the problem related to this point.

134 4. Basis for Xin’s Judgement in my taking on the goal of going to Café D. Not only do I need to know where Café D is, I also need to know that Café D is the only restaurant that fits my goal. Prior to acquiring this knowledge, the option of going to Café D is only contingent to me. At the moment when I acquire the knowledge that Café D is the only restaurant that fits my goal, there is a shift in the meaning of Café D to me. Provided that my initial goal of going to a French restaurant does not change, going to Café D is no longer a means to fulfil my goal but it itself becomes a goal of mine. We can represent how knowing Café D is the only French restaurant bears on the meaning of going to Café D for me as follows:

Before knowing that Café D is the only French restaurant: I have the goal of going to a French restaurant. I know that Café D is a French restaurant. Going to Café D is a means to achieve my goal.

After knowing that Café D is the only French restaurant: I have the goal of going to a French restaurant. I know that Café D is the one and only French restaurant. In this case, the goal of going to a French restaurant is identical to the goal of going to Café D. Going to Café D is the goal.

Let us now return to Xunzi. The reason xin abides by Dao, when it knows that abiding by Dao as an end in itself is the only way to obtain satisfaction, is analogous to the reason I go to Café D. This is because abiding by Dao is not an inherent goal of xin, but can be a goal of xin as long as xin obtains the knowledge that only following Dao ensures satisfaction. On this view, we need not suppose that xin has predisposed inclinations towards following Dao. We only need to suppose that, for Xunzi, xin has the inherent goal of seeking satisfaction that does not change and it is a fact that only abiding by Dao as an end in itself provides satisfaction.36

36 It should be noted that it is not entirely impossible that, under the proposed interpretation, a person does not choose A even when she knows that A is weightier. This is when she knows how to achieve her goal but she decided not to. In that case, the person will be acting against her goal to seek satisfaction and is irrational. I do not think Xunzi conceives of this possibility and I will not discuss this in detail.

135 4. Basis for Xin’s Judgement

I anticipate two objections. The first is that xin does not have the inherent goal of abiding by Dao. Someone might object by arguing that, according to the account I offered, Dao is in fact the inherent goal of xin after all. This is analogous to how a bystander may say Café D is in fact the restaurant I am seeking, given the goal I have and the facts about the state of the world. My reply is that the bystander is making a prescriptive claim about what goal I should have instead of an explanatory claim about what goal I have. Likewise, we should not confuse an investigation of what goal xin should have with the investigation of what goal xin has. The proposed account deals with the latter question and argues that xin inherently has the goal of seeking satisfaction and it is through acquiring the knowledge that abiding by Dao as an end itself is in fact the only way to obtain satisfaction that it takes on the goal of abiding by Dao. The second possible objection is that if xin’s abiding by Dao is consequent upon xin’s having the goal of seeking satisfaction, then xin is essentially abiding by Dao on the basis of L. Such an objection misses the point of what it means to take abiding by Dao as a goal. As discussed earlier, the very notion of abiding by Dao requires xin to disregard considerations about personal satisfaction. As a corollary, when xin takes on the goal of abiding by Dao, xin has to commit itself to making judgements in accordance with Dao and discard considerations about what it itself finds satisfying. Hence, if xin has the goal of abiding by Dao, xin necessarily adopts Y. Since L and Y are incompatible with and exclusive of each other, it is impossible for xin to adopt L if it has the goal of abiding by Dao. When xin adopts Y, it will put yia ibefore l c. Although satisfaction will follow from xin acting in accordance with Dao, the very notion of acting in accordance with Dao means that one cannot act on considerations that relate to seeking personal satisfaction. If this understanding is correct, then, for Xunzi, a moral act has to be performed for two reasons, because the agent is motivated to perform the act and because the agent is obligated to perform it. On the proposed interpretation, we do not need to assume the presence in xin a natural preference for Dao in order for it to adopt Y. For example, I can come to believe that Café D is the only French restaurant and given the goal I had, I will take Café D to be my goal. It is not because I initially like Café D that I take on this goal. In a similar vein, xin takes on the goal of abiding by Dao because it knows that it is

136 4. Basis for Xin’s Judgement the only way to achieve its goal, not because it has a preference for abiding by Dao. Xin adopts Y because it has the goal of abiding by Dao, not because xin has a natural preference for adopting Y. Xin’s adoption of Dao is only a matter of xin’s knowing a fact about Dao. This implies that xin’s compliance with Dao is not contingent upon xin having a preference for Dao. Even though the reason xin adopts Y does not involve xin having a prior preference for abiding by Dao, it does not mean that xin cannot find satisfaction from abiding by Dao as an end in itself. Recall that in section 3.3 we discussed how abiding by Dao entails a “greater” kind of satisfaction. Since Dao has already incorporated characteristics of human feelings, one who acts in accordance with Dao will surely find satisfaction in a guided and appropriately balanced way.37 It is probably against this background that Xunzi makes the distinction between satisfaction (e.g. honour) and “great” satisfaction (e.g. great honour). The former kind of satisfaction is obtained from the basis of personal satisfaction, whereas the latter kind is obtained from abiding by Dao and disregarding concerns about what satisfies oneself. In this light, we can better understand Xunzi’s distinction between “satisfaction” and “great satisfaction” in terms of the different bases from which the satisfaction is derived. Hence, performing moral actions as ends in themselves is a precondition of being “greatly” satisfied. In this sense, Xunzi does not seem to think that there is a clash between doing something because it is obligated and doing something because it makes one feel good. This is because Xunzi assumes that what is morally obligated is what is in fact satisfying. However, there is a difference when it comes to the reason that motivates xin. There is a significant difference for Xunzi between xin’s judgements that are derived from L and those that are derived from Y. The requirement for there to be a change in xin’s basis for judgement further implies that it is not possible for xin to have knowledge of Dao but fail to be motivated to abide by Dao. If xin is not motivated to abide by Dao as an end in itself, then it only means that xin has not really acquired knowledge of Dao. The proposed view also allows us to understand that there are two senses of “knowing” involved in Xunzi’s concept of “knowing Dao.” One is concerned with xin’s knowledge of the content of these moral standards themselves and the other is concerned with xin’s knowledge of the status of these moral standards. It also implies that xin obtains these two types of knowledge in the same manner. Knowing the fact

37 It is also most likely in this sense that Xunzi speaks of “dao yu” and “jie yu” in K 22.5a.

137 4. Basis for Xin’s Judgement that only abiding by Dao provides satisfaction is just like knowing facts about the content of Dao. With proper moral guidance and training, xin can know both the status and the content of Dao. These two senses of “knowing” do not presuppose each other. Just like I might know the address of Café D but not know that it is the only French restaurant or I might know Café D is the only French restaurant but not know its address, so too xin can know what the moral standards are but not know that abiding by these standards is the only way to obtain satisfaction or xin can know the latter but not the former. We can say that knowledge of the content of Dao facilitates the obtainment of knowledge of the status of Dao. The more information xin gathers about Dao, the more likely xin is going to know the status of Dao. This allows us to further infer that knowledge of the content of Dao is a necessary but insufficient condition for adopting Y. The knowledge of the status of Dao, on the other hand, is a both a necessary and sufficient condition for adopting Y. It is in this sense that I understand xin’s knowing the status of Dao as the sufficient condition that guarantees moral motivation. The obtainment of knowledge of the content of Dao is necessary for moral action but is not sufficient. The gap between knowing the content of Dao and acting in accordance with Dao has to be filled by xin’s knowledge of the status of Dao. So, on the proposed interpretation, it is possible for xin to know the content of Dao but does not abide by Dao but it is not possible for xin to know that only following Dao provides satisfaction but still does not adopt Y. This interpretation is supported by the passage that says knowing the teachings of sages is not as good as acting them out.38 This implies that there is still a gap between knowing Dao and acting in accordance with Dao. In light of the proposed interpretation, we can better understand this gap in terms of the gap between one’s knowing the content of Dao and one’s knowing that only following Dao provides satisfaction. This is supported by the observation that the “zhic” in these expressions all refer to the words of sages, which are presumably about the content of Dao. The issue is complicated by the consideration that the second type of knowing involves a belief on the part of xin. This speculation is not ungrounded if we make note of some hints in the text that suggest xin’s being taught about moral facts is not the same as xin’s grasping of moral facts. Consider the following passage:

38 Xunzi 8/102.

138 4. Basis for Xin’s Judgement

಴ࡗ{¢Q3ࡾ቗4ࡾ࿾÷)5រࡾ×{œ7óዔ¢

}÷8॥ßùî॥಴ê5&រࡾ7౉቗¢

Xin is also like this. Therefore, if you lead it with lia, nurture it with clarity so that no object (wu) can tilt it, then it is enough to decide between right and wrong and to resolve any confusions. If small objects (wu) pull it, then its proper position is externally changed. One’s xin internally tilted, then it is not enough to determine even gross patterns.39

The first half of the passage says that when xin is guided by lia, objects cannot tilt xin and hence one can resolve uncertainty (yic) about what Dao requires.40 The second half says that if external objects attract xin, then xin will be tilted and therefore cannot even determine the most basic lia.41 Xunzi seems to be making two suggestions here:

1. When xin is guided by lia, objects cannot tilt xin and one can decide on right and wrong; 2. When xin is tilted, even the basic li cannot be grasped.

In this sense, there is an apparent tension in Xunzi’s thought. It is said that a tilted xin will prevent a grasp of li. But in order to make xin untilted, we need the guidance of lia. The apparent circularity can be resolved if Xunzi adopts one additional position:

3. For lia to guide xin is not equal to xin’s grasping of lia.

In this regard, Xunzi can either maintain that (a) even when xin does not grasp lia, it can continue to be guided by lia and shaped into a rectified xin; or, (b) it is ultimately up to xin to grasp lia. We can try to resolve this difficulty by turning to the concepts of xu yie jing. Recall that it is through xu yie jing that xin can know Dao. As discussed in sectioin 3.4, the emphasis of these three notions are on prescriptive practices that aim at

39 Xunzi 21/56-8; K 21.7b/3/107. 40 See Chapter 3 discussion of yic. 41

The Song editions of the Xunzi have Ÿ‹ . Here I follow Wang Xianqian’s edition and take it to be

‹ .

139 4. Basis for Xin’s Judgement bringing xin in line with Dao, rather than descriptive statements about mental states that are achieved through moral cultivation. In virtue of xin’s capacity to zhic, xin can be trained to xu, yie and jing so that knowledge about Dao can be externally drilled into it. Xunzi must have to attribute a power to moral training, which is derived from Dao, because Dao is indeed satisfying. Thus, it does not require the presence of voluntarism. The moral training can just straighten a tilted xin. Once it has been rectified, it can pereceive lia and follow Dao. This explains why Xunzi says that moral training can never stop, given humans have this predisposed inclinations to problematic objects.42 It explains why, as we will see towards the end of Section 4.3, there is the possibility for xin to revert to L despite it has adopted Y. Xin has to fixate its orienation on the goal of abiding of Dao. Hence, all Xunzi has to assume is that xin has a bare capacity to zhic that does not have any content. Being moral itself is a capacity and humans clearly do not have this capacity when they are born. Zhic is the capacity that enables one to be moral. In other words, zhic is a capacity that xin has and can be used to develop the moral capacity that humans do not yet possess. An example will help to illustrate the difference between these two capacities: swimming itself is a capacity; learning how to swim is another capacity. If we say that humans are born with the capacity to learn to swim, it means that we have the capacity to move our arms and legs in a certain way in water. But this capacity itself does not have a goal that can be developed in different ways. The key is to develop our capacity to move our arms and legs in a way that takes swimming as our goal. Similarly for Xunzi, humans are born with the capacity of zhic but it itself is a capacity that can be developed in many different ways. Hence, it is in virtue of this bare capacity – zhic – that xin can become moral. Xunzi’s own analogy explains this view. A tree is naturally a plant that has certain features and characteristics. Some of these characteristics can be manipulated by external forces and can be created into a cup, which is another capacity. When it is being made into a cup, it is no longer a plant that has branches and stem but a container. Although the tree and the cup might share some characteristics, they are different. Similarly, xin is born with a capacity that has nothing to do with moral capacity but can be manipulated by external forces to become a moral capacity. So, as long as Xunzi believes that Dao is knowable and that xin has the capacity to know, Xunzi is able to

42 Xunzi 1/1.

140 4. Basis for Xin’s Judgement maintain that it is possible for xin to abide by Dao. It is also presupposed in this view that xin has a capacity to disregard self-regarding interests and follow moral standards. Nonetheless, such a capacity is different from moral capacity in that it is directed towards a particular direction.

4.5 A Remaining Difficulty

In the previous section, I tried to extract all the possible assumptions Xunzi could have adopted in order for him to maintain his position xin can change the basis for its judgement from L to Y. However, I am afraid the question of why xin abides by Dao is still not satisfactorily answered. A critical question that has not been answered is how xin comes to know that abiding by Dao is the only way to obtain satisfaction. In order for someone to internalise knowledge of Dao, she has to believe that it is the only way that she can find satisfaction. Unlike the first sense of knowing, which can be measured by how many facts I know about Dao, the second sense of knowing is an either/or type of knowing. Either I know it or I do not. I cannot say I know a little bit that only Dao provides satisfaction or I know a lot. There must be a qualitative change at some point. It seems more difficult for us to understand how xin acquires the second type of knowledge. For example, my mathematics teacher could teach me the easiest method to solve a particular calculus problem. As a result of her teaching, I know how to apply the method to solve the problem. Nonetheless, even though it is written everywhere in textbooks that it is the easiest method, I still do not sincerely believe that it is the easiest method to solve the problem. I might, out of curiosity, continue to try new methods that I believe might be even easier; or I could continue to use the method I used before, secretly believing my method is easier than the standard textbook method. The question is how I will come to believe that the method my teacher taught me is in fact the easiest method. How many teachers and textbooks are enough to convince me? How many other methods do I have to try? How long will I keep using my old method until one day I concede that my teacher’s method is indeed the easiest? In the way I presented the problem, it seems that something additional is required in the second type of knowing, namely, a belief on the part of xin. What the teachers need to do is not just to make xin know what is permissible and impermissible in accordance to Dao, but also the fact that only abiding by Dao provides satisfaction. However, could it be possible that even when

141 4. Basis for Xin’s Judgement xin has been repeatedly taught of the fact that only abiding by Dao provides satisfaction, xin still does not believe that abiding by Dao is in fact the only way to obtain satisfaction? It is unclear how Xunzi would answer this because it is unclear how xin comes to know that abiding by Dao as an end in itself is in fact the only way to obtain satisfaction. Xunzi could mean that xin comes to believe by way of information accumulation. For example, I might come to believe that Café D is the only French restaurant in town by reading many guidebooks and asking many people. Similarly, xin can come to believe that only abiding by Dao provides satisfaction by repeatedly gathering this idea from moral training until it comes to one point, at which xin decides that it has obtained enough information for it to believe that only abiding by Dao provides satisfaction. I can also take accumulation to mean that the person keeps practising what she has been taught, and with enough repetition her xin knows from practising that only xin provides satisfaction. So, xin can draw the inductive inference that since X numbers of sources say only following Dao provides satisfaction or since X numbers of Dao acts provide satisfaction, then all Dao acts must provide me satisfaction. However, since it is based on experience, there is always a danger for xin to doubt whether it is indeed the case that following Dao is the only way to obtain satisfaction. Xunzi’s understanding is further complicated by the constrast he draws between zhic that is shallow and zhic that is ming.43 This suggests that the knowledge of Dao can be quantified and that it comes in incremental degrees. On this view, the more fully exposed xin is to relevant facts about Dao, the more likely xin is going to believe that only following Dao provides satisfaction and the more likely that this belief will be entrenched. Nonetheless, there remains the question as to how much accumulation is needed in order for xin to believe the fact that only following Dao provides satisfaction. Alternatively, Xunzi could mean that with sufficient accumulation of knowledge of the content of Dao, there will be a qualitative change in xin’s knowledge of Dao such that it comes to see the fact that following Dao is the only way to obtain satisfaction. Again, it is unclear at what point this qualitative change will occur.44

43 Xunzi 5/21-2; K. 5.3/1/205. See discussion in section 3.3. 44

See section 5.3 for a connection between this point and our understanding of transforming nature ( ¡

").

142 4. Basis for Xin’s Judgement

The implication of adding the requirement for belief on the part of xin is that xin is actively involved in motivating a person to act in a way that is in accordance with Dao. There are facts of the matter about what we have reason to do that are prior to and independent of an agent’s deliberations. Xin has the capacity to reflect on these objective facts and is potentially answerable to these moral facts; but what gives these moral facts their normative force depends on the agent’s understanding. So, on one hand, these facts are independent of xin but it is only when xin has internalised them that xin will be motivated to observe them. In this sense, the acquisition of moral beliefs involves both bringing xin to fit with external facts and bringing external facts to fit with xin. Moral belief works on some existing capacities and dispositions of xin to produce moral motivation. These dispositions do not have to be moral dispositions as such but dispositions in a neutral sense, dispositions that allow xin to internalise a moral fact. In Xunzi’s view, it suffices to say that the dispositions on which moral beliefs build are xin’s inborn capacity to zhic and the disposition to seek satisfaction. The answer to this question largely depends on the way we understand how xin comes to know that only abiding by Dao as an end in itself is satisfying. If xin acquires this knowledge just from external training, then the possibility of reversion seems to suggest that xin can forgot or maybe abandon the knowledge it acquires. If xin comes to believe this fact through accumulating information about Dao, then an instance of reversion is probable when xin doubts its belief. If xin comes to acquire its belief because of a qualitative change, then it could mean that xin can abandon what it once believed. Since we cannot determine why xin comes to know that only abiding by Dao as an end in itself provides satisfaction, I will leave this question open-ended. In this chapter, I have shown that xin is predisposed to seek satisfaction and has no innate knowledge or preference for Dao. It is in virtue of xin’s capacity to zhic

 that xin can know the content of Dao from moral training and that abiding by Dao as an end in itself is indeed the only way to obtain satisfaction. When xin acquires such knowledge of Dao, xin will take abiding by Dao as an end in itself and make judgements that are in accordance with Dao.

143 CHAPTER 5

Xin and Xinga

5.1 Xunzi on Xinga

a This chapter examines Xunzi’s position on xing , commonly translated as “human nature.” My focus is on using my analysis of Xunzi’s concept of xin to suggest another way of understanding Xunzi’s claims about xinga. I will also discuss the difference between Xunzi’s and Mencius’s understanding of xin in hope of shedding light on their different views of xinga. My earlier discussion of Xunzi’s xin has prepared the ground for me to argue in

a this section that the emphasis of Xunzi’s concept of xing  is on the inherent characteristics and natural tendencies of xin. The “Xing e” chapter of the Xunzi is devoted to arguing that human nature is bad.1 The opening line of the chapter can be roughly translated as:

ࡥപ॥૽ᒮ

The xinga (nature) of human beings is e (bad); its shan (goodness) is a result of wei (artifice).2

1 Some scholars have doubted the extent to which Xunzi is committed to the claim that human nature is bad. Dan Robins, for example, has conducted a detailed analysis of the “Xing e” chapter and argued that not only is the claim not very important to “Xing e,” Xunzi’s position on human nature has even changed

later on. Robins also argues in length that most occurrences of the refrain: “If we look at it this way, then it

¤ ¥ ¦¥ § ª ¨ ©

is obvious that people’s xing is bad; when they are good, it is because of artifice ¢ £ «¬ ­® ¯ ” are interpolations (Robins, 2001-2). My focus is not so much on the importance of this claim in Xunzi’s entire system of thought; rather I focus on explaining what could Xunzi have meant when he says human nature is bad. I take for granted that Xunzi at some point did argue that human nature is bad. 2 Xunzi 23/1; K 23.1a/3/150. 5. Xin and Xinga

a Let us first briefly examine what Xunzi means by “xing ،,” “wei ,” “e പ,” and “shan

૽.” In the following, I provide literal translations of Xunzi’s definitions of these four terms:

 ࡾ ᒮ ઉංᕧ 

That which is so from birth is called xinga. That which is generated out of the harmony of xinga, merges with the quintessential, is aroused and responds, is not worked at but is so is called xinga.3

ᒮ

Xinga is what brought to fruition by Heaven.4

ࡥᒮ

What cannot be learned, cannot be worked at, and yet resides in humans is called xinga.5

Wei, a concept that contrasts with xing, is defined as:

಴ ჽ਄"ᄮ ᒍᄮ%&

Xin deliberates (lu) and because of this, one is able to put into action called wei. Deliberations accumulate and one is able to practice repeatedly and is later completed is called wei.6

3 ± ² ³

Xunzi 22/2-3; K 22.1b/3/127. My translation for ° is only tentative as it involves highly complicated concepts and it is not possible to fully understand its meaning without a thorough investigation of their relevant usages in the Zhuangzi and the Gaunzi. For our present purpose, it suffices to understand that Xunzi is here referring to spontaneous reactions that do not require effort. Cf. Graham, 1989, p.245. 4 Xunzi 22/63; K 22.5b/3/136. 5 Xunzi 23/12; K 23.1c/3/152. 6 Xunzi 22/4; K 22.1b/3/127.

145 5. Xin and Xinga

 &ࡥᒮ

What can be learned and then one is able to do it, what can be worked at and then can be achieved by human beings is called wei.7

Xunzi understands e and shan in terms of the following:

দੱ࡯* ૽ᒮ+቗ఞ. പᒮ/ᬛ೽࠿

In every case, in the past and present, what the world has called shan is what is correct, put into order, impartial, and orderly. What is called e is being partial, dangerous, contrary to the right, and disorderly.8

In light of the above discussions, we can return to the claim, “Human xinga is e, its shan is a result of wei” and understand it as saying that the nature of human beings is innately and naturally partial, dangerous, contrary to the right, and disorderly. It is because of xin’s deliberations, learning and making effort that human beings can be correct, put into order, become impartial and orderly. Put in this way, it becomes clear that this description about human xinga fits with our earlier discussion about xin. In the previous chapters, we saw that, for Xunzi, xin is predisposed to seek satisfaction. Since xin has no innate knowledge or preference for Dao, xin is guided by its own basis and only values what satisfies the self. Hence, the judgements xin makes prior to deliberations and learning are bound to be wrong, partial, and cause disorder because only judgements that are made in accordance with Dao are objectively right judgements. Nonetheless, since xin’s has the capacity zhic, it can step back from feelings and lu. It is also because of zhic that xin can,

7 Xunzi 23/13; K 23.1c/3/152. 8 Xunzi 23/37; K 23.3a/3/155. In what follows, I will use the term “bad” to refer to e and “good” to shan to facilitate discussion. Dubs, Watson, and Knoblock all translate “e” as “evil.” Other scholars such as A. C. Graham, P. J. Ivanhoe, and Eric Hutton take “e” to mean “bad” (Graham, 1989, p.248; Ivanhoe, 2000, p.29; Hutton, 2001, p.298). As A. C. Graham points out, “[Xunzi’s] doctrine has important differences from the Western idea of Original Sin, of which one is too easily reminded when for example his slogan is translated ‘Man’s nature is evil’. He does not of course have the Christian conception of a corrupted nature in permanent rebellion, successfully kept under only with the help of divine grace” (Graham, 1989, p.248). In order to avoid framing Xunzi’s view of human nature with unnecessary Judeo-Christian presuppositions, I will translate “e” as “bad.” Similarly, when I translate shan as “good,” I only refer to Xunzi’s own definition of shan.

146 5. Xin and Xinga from learning and practising, know the content of Dao and that abiding by Dao as an end in itself is indeed the only way to obtain satisfaction. Once xin acquires such knowledge of Dao, xin will make Dao its basis for judgements and hence make correct judgements that are impartial and conducive to good order. This suggests that when Xunzi speaks of the xinga of human beings being bad, he means that the xin of human beings is bad, to be precise. This is different from saying that the reason human nature is evil is that giving inborn yu its way will result in a series of negative consequences. For now, I will spell out in some detail the difference between my understanding and the common understanding of Xunzi’s concept of xinga. As far as I can tell, few scholars have proposed that Xunzi’s claim that human xinga is bad” refers to a bad xin. This is probably because Xunzi has included xin’s deliberation in his definition of wei. So, even when some scholars notice that there is a problematic side of xin, they tend to explain it away on the ground that xin also has a positive side that enables human beings to transform their xinga.9 Since xin is generally understood as outside the scope of xinga, the common understanding is that Xunzi’s

a a concept of xing includes only qing ച and yu 4 and excludes xin. Tang Jun-yi, for example, argues that the concepts xinga and xin are separated for Xunzi. Tang explicitly argues, “What belongs to xin is not xinga.”10 Benjamin Schwartz similarly understands Xunzi’s xinga as referring to the “realm of the passions and desires” but not the tendency of xin.11 More recently, Aaron Stalnaker echoes the view that Xunzi’s concept of xinga excludes xin and its capacities.12 The understanding that xinga and xin are separate for Xunzi could lead to two readings of Xunzi’s xinga. One reading runs along the line that says since Xunzi frequently discussed xinga in relation to qinga and yu, Xunzi’s claim that human xinga is bad can be read as saying yu is bad. This reading is problematic because it

9 Consider Tang’s differentiation between “human xin” and “Dao xin” discussed in section 3.1. 10 Tang, 2005b, p.31, 35; Tang. 2005a, p.78 11 Schwartz, p.293. 12 Stalnaker wrote in Overcoming Our Evil (2006) that, “For Xunzi, xing does not exhaust our human

capabilities; in particular, it leaves out the human heart/mind (xin ´ ) and its capacities to respond to the world and direct action…” (pp.58-9). Later, in the article “The Mencius-Xunzi Debate in Early Confucian ” (2008), Stalnaker seems to have somewhat softened his stance and argued that, “For Xunzi, the xing does not exhaust our capabilities and potentials; in particular it has little to do with the human

heart/mind (xin ´) and its capacities to respond to the world and direct action…” (p.92).

147 5. Xin and Xinga understands yu as bad in itself and inherent in xinga. As discussed in Chapter 2, Xunzi clearly expresses his view in passage 22.5a that yu is not bad in itself. A second reading focuses on the consequence of following one’s xinga and takes Xunzi’s claim that human nature is bad to be saying that letting yu go wayward without the control of xin is bad.13 Since the underlying assumption of this reading is that yu has causal role in action in and of itself and since I have shown (in Chapter 2) that this is an unfounded assumption, this reading is also rejected. Indeed, xinga is largely tied in with qinga and yu in the Xunzi.14 However, this should not prevent us from understanding that xin is also included in the scope of xinga. While Xunzi says that the xinga of human beings has inborn dislikes and yu of senses, he also says that xinga has an inborn preference for lic.15 This is consistent with our earlier discussion that xin has a predisposed preference for lic. Moreover, it should be noted that Xunzi’s emphasis is not so much on these inborn likes and dislikes as such but rather on the problematic consequences of following along (shun 5) these inborn likes and dislikes, such as strife and contention, lewdness and disorder. Hence, Xunzi sees it

b a a 16 necessary to have teachers and li yi 6ᑶ to transform xing . As we discussed in

Chapter 4, for Xunzi, the point of moral training, that is, learning from teachers and observing lib yia is to help xin know Dao so that it can discard its inborn basis for judgements, that is, personal satisfaction, and adopt a new basis, that is, Dao. This observation reinforces the point that Xunzi’s claim about humans having problematic xinga is grounded on his view that humans have problematic xin. There is one quote in

a d a 17 d Xunzi that says: “Qing is the zhi 8 of xing ,” where “zhi ” is usually translated as

“substance” or “stuff.”18 Although these are suitable translations of “zhid,” in this context, it might create the impression that qinga is the sole constituent of xinga. It is reasonable to understand qinga as an underlying constituent that supports xinga, but it will be far-

13 See, for example, Xu Fu-guan, 1969, p.238.

14 ¶ If we look at the definitions of xing, they are all intertwined with µ and . E.g. Xunzi 22/63-4, 23/1-3, 23/17-18, 23/29-31. 15 Xunzi 23/26; K 23.1b/3/23. 16 Xunzi 23/5-7; K 23.1b/3/23. 17 Xunzi 22/63; K 22.5b/3/136. 18 See for example, Watson, p.155; K 22.5b/3/136; Graham, 1989, p.245; Hutton, 2001, p.297.

148 5. Xin and Xinga fetched to understand it as the sole constituent of xinga and thereby infer that the scope of xinga is restricted to qinga. Kwong-loi Shun is among the few who suggests that Xunzi’s xinga could include xin. According to Shun, “Even if [Xunzi] had allowed xinga to include the capacities of [xin], xinga would still be evil in exactly the same sense – unless [xin] exercises control over the emotions and desires, strife and disorder will result.”19 In light of our earlier discussion of xin, it is more plausible to understand Xunzi including xin in the scope of xinga and that his claim “xinga is bad” is a directed at the predisposed inclinations of xin to seek only what satisfies personal feelings. Xunzi is saying that if xin follows its natural inclinations, it will surely cause actions that will bring about disorder. And since xin has no preference for and innate knowledge of Dao, there is nothing inherent in xin that will enable it to make impartial and right decisions that will lead to order. It is through learning and practising — wei — that xin can acquire knowledge about Dao and hence shift from following its natural inclinations to following Dao. Just because xin is capable of deliberating and making the right decisions does not mean that there is no problematic predisposition inherent in xin.20 Although Xunzi regards xin’s deliberation belongs to wei, it is the inborn capacity zhic that allows xin to deliberate. And as discussed before, zhic is just a bare capacity to know, without necessarily having a predisposed orientation. One possible objection to my explanation is that if xin is capable of being good and if xin is part of xinga, then Xunzi would contradict his own claim that xinga is bad. I have in mind something like D. C. Lau’s assessment of Xunzi. According to Lau, the weakness in Xunzi’s theory of xinga is that he fails to include the capacity of xin in xinga. If xin is part of xinga, then goodness is also part of xinga. However, Xunzi cannot take this position because that would contradict Xunzi’s claim that xinga is bad.21 In reply, I again refer to the difference between xin’s having the capacity to become good and the capacity of being good. Zhic is a mere capacity to know things and does not have content or predetermined direction. It enables one to be good but it itself is not the capacity of being good itself. Without acquiring knowledge about Dao from external training, there is nothing inherent in xin that tends towards goodness. If there is no external intervention, it

19 Shun, 1997, p.228. 20 See section 3.1. 21 Lau, 1953, p.561.

149 5. Xin and Xinga is inconceivable that xin will by itself take on the course of following Dao. My reply might then lead one to question the extent to which Xunzi can really claim that human beings can be good, given xin is part of xingb and that it does not have inborn goodness. The reason I anticipate this challenge is that in the literature there is a general hostility or unfavourable attitude towards Xunzi’s idea that goodness can be acquired. In Lau’s view, for example, acquired goodness “does not possess the same significance as it would if it were part of original human nature.”22 This hesitation to accept acquired goodness as having the same status as “true” goodness often stems from the understanding that xin has no substantial attachment to goodness. Lao Si-guang and Xu Fu-guan, for example, think that xin is not actively present in the process of acquiring goodness but only passively reflects objective knowledge.23 Heiner Roetz suggests that it is on utilitarian grounds that xin signs up to be good.24 Common across these views is the worry that one’s xin is not attached to goodness. To address this challenge, it is important for us to first differentiate acquired goodness and invented goodness. Lau, for example, equates goodness with acquired goodness. I agree with Lau that invented goodness is based on social conventions and customs.25 However, when Xunzi says xinga can be good, he clearly does not only mean that xin can abide by social conventions. For Xunzi, a truly good person is someone who abides by Dao. Since xin has no innate knowledge of Dao, it is through acquiring knowledge from external training that xin can abide by Dao. It is most likely in this sense that Xunzi understands goodness as acquired. Just because goodness is acquired through training does not mean that xin cannot attach to Dao. As we saw previously, xin has to internalise the value of Dao and adopt Dao as the basis for judgements in order to genuinely abide by Dao. Although it may be for self-interested consideration that xin takes on the goal of abiding by Dao, the very goal of abiding by Dao requires xin to change its basis for judgements of Dao. This means that for a person who has truly acquired goodness, her xin must be appropriately aligned with and motivated by Dao.26 On the basis of my proposed understanding of Xunzi’s xin, I now proceed to argue that

22 Ibid. Xu Fu-guan also makes a similar point 23 Lao, pp.256-7; Xu, pp. 258-260. 24 Roetz, p.226. 25 Lau, 1953, p.561 n1. 26 See discussion in section 4.4.

150 5. Xin and Xinga the difference between Xunzi’s and Mencius’s concept of xinga lies in their different understanding of xin.27

5.2 Xunzi and Mencius

Critical to our understanding of the difference between Xunzi and Mencius is the observation that with Xunzi’s statement that “xinga is bad,” he is trying to convey that xin is bad. In the literature, the similarities between Xunzi’s and Mencius’s views on human nature tend to be more readily identified than their differences, albeit Xunzi and Mencius have flatly stated two seemingly contradictory claims about human xinga. It is commonly agreed that both thinkers belive that humans can be good, that humans can be bad, and that making effort in the Confucian way is important. However, these similarities do not entail that there are no substantial differences between these two thinkers’ view on xinga, however. Let me begin with a brief evaluation of the common approaches to this issue in the literature. The traditional way of explaining Xunzi’s and Mencius’s views on xinga focuses on their different usages of the term “xinga.” The common explanation is that because they use “xinga” differently, there is no substantial difference between the two thinkers’ views on xinga or that Xunzi’s attack against Mencius is off the point. Chen Da-qi’s study of Xunzi is an example that focuses on the different meanings of the term “xinga” and concludes that there is no substantial disagreement between Xunzi and Mencius. In an earlier article (1953), Chen acknowledges that Xunzi’s and Mencius’s theories of xinga oppose each other but maintains that their differences are not significant. According to Chen, Xunzi includes sensual desires and excluded xin in xinga, whereas Mencius includes xin and excluded sensual desires in xinga. After all, they both think that humans can be either good or bad and that it is important to make efforts to become good.28 In a later article (l967), Chen further argues that not only are the two theories not opposed to each other, they are in general compatible and similar. Chen goes further to argue that since the scope of Xunzi’s and Mencius’s xinga is different, the two thinkers’ conception

27 I am grateful to an anonymous examiner whose comments prompt me to reframe my discussion. 28 Chen, 1953.

151 5. Xin and Xinga of xinga denotes very different things. According to Chen, it is like comparing a human to a horse:

For instance, one says: humans stand straight on two feet. Another person argues that horses do not stand straight on two feet. These two judgements certainly do not oppose to each other. Mencius’s theory of human xinga is good and Xunzi’s theory of human xinga is bad is similar to the above example. The real meaning of their theories do not oppose each other.29

Chen then goes on to argue that Mencius’ and Xunzi’s theories are indeed very similar because they both understand ren and yia as paramount and that sensual desires are bad.30 According to Chen, Xunzi and Mencius also have a shared view of the origin of goodness. In Chen’s view, Xunzi should also submit to the claim that goodness, to some extent, originates in xinga. Since human nature is the raw material and since there is no way to create goodness or transform something into goodness without the raw material, then goodness must arise from xinga.31 Likewise, Mencius cannot say that goodness comes entirely from xinga because good sprouts await human effort to cultivate them.32 Chen draws an analogy between the two thinkers’ theories of xinga and their conception of gardens. According to Chen, both Mencius and Xunzi conceive of a garden. The garden of Mencius and the garden of Xunzi could appear the same. Where they differ is on how the garden initially appears. For Mencius, the garden was initially a beautiful thriving garden; for Xunzi, the garden was initially a wild, deserted garden.33 This, in Chen’s view, is only a trivial difference between Mencius and Xunzi. But if our focus is on Mencius’ and Xunzi’s theories of xinga, how precisely the initial state of the two gardens differ is of crucial importance to us. If Mencius’ garden is a beautiful one to begin with, then what is required is a matter of maintenance and nourishment. For Xunzi, on the other hand, the effort required is a matter of renovation and creation. They differ on the reason why effort is needed and how this effort is to be executed.

29 Chen, 1967, p.403 30 Ibid., pp.403-406. 31 Ibid., pp.408-410. 32 Ibid., p.406-7. 33 Ibid., p. 410-1.

152 5. Xin and Xinga

As for other approaches that address the differences between Mencius’s and Xunzi’s concept of xinga but do not propose that the two thinkers are similar, they tend to conclude that Xunzi’s attack on Mencius misses the point. For example, A. C. Graham notes that Xunzi’s criticism against Mencius is “not quite to the point.” Mencius includes what is learned in xinga but Xunzi does not;34 Mencius does not include indulging in spontaneous desires as part of xinga but Xunzi does.35 Xu Fu-guan similarly argues that the content of xinga for Xunzi and for Mencius is different. Since with Xunzi has included yu of the senses in xinga but Mencius has not, there is “no point” for Xunzi to launch his criticism against Mencius on the issue of xinga.36 It is unfortunate that the differences between Xunzi’s and Mencius’s conception of xinga are often dismissed quite hastily. Dennis Ahern has also examined the meaning of the term “xinga” for Xunzi and Mencius and found that the two thinkers have very different definitions of xinga.37 According to Ahern:

There is an equivocation in the use of the word “xinga” and thus one does not capture the difference between the views of Xunzi and Mencius by claiming that they had differing opinions about the goodness or evil of man’s nature or by claiming that they include in or exclude from human nature different items.38

Let us put aside Ahern’s subsequent evaluation of Xunzi’s and Mencius’s position and consider the point that there can still be differences between two thinkers even when the term xinga has different meanings for them. Granting that the scope of xinga in Xunzi and Mencius is different, it does not entail that they are talking about completely different things. If we invite two people to describe their understanding of object X, person A might give a description that corresponds to a cat while person B might give a description that corresponds to a dog. We cannot casually dismiss the difference between A and B on the grounds that they are describing different things. As long as X is their common object

34 Graham, 1989, p.250. See also Graham, 1967, pp.256-7.

35 Graham, 1967, p.244.

¦¸ ¹ º§ º »¼½ ¾ ÂÃ Ä ÄÇÈ ÉÊË Ì§ º¬ ¾Í ÎÏÐ Ñ ½ ÒÎÏ

¿ÀÁ ÅÆ Á Á 36 ¨

· ÓÔ (Xu, 1969, p.238). 37 Ahern, 1980, pp.178-9. 38 Ibid., p.181.

153 5. Xin and Xinga of consideration, then their descriptive statements, albeit different, could suggest something very important about their understanding of X. We can infer from A’s statements that X in A’s view is a cat-like thing, while X in B’s view is a dog-like thing. If our goal is to approximate how Mencius and Xunzi have understood xinga, then their different descriptions of xinga indeed say something very important about their different understanding of xinga. It is equally problematic to focus on how the two theories are similar and thereby overlook and dismiss important differences between them. In order to illuminate the approach that is frequently taken in studies of Mencius’s and Xunzi’s ideas, let me draw an analogy.39 Suppose we are discussing the different views of two people, A and B, on marriage and discrimination against homosexual people. To have a detailed grasp of the differences and similarities of their views, we need to investigate their positions on (a) marriage, (b) discrimination against homosexuals, and (c) possible links in their views between marriage and discrimination. It would not help our understanding of their views if we only investigated (b) in order to understand the full extent of their disagreement. For example: if both A and B hold exactly the same views on discrimination against homosexuals, it would be inappropriate for us to conclude therefore that they share the same views about (a) and (c). Suppose both A and B agree that opposition to same-sex marriage constitutes discrimination against homosexual couples. As it happens case, A and B do have different views of marriage. A understands the concept of marriage as the formal union between a man and a woman and therefore thinks that the institution of marriage discriminates against homosexual couples. B, on the other hand, understands the concept of marriage as about the formal union between two people and hence thinks that the institution of marriage does not particularly discriminate against homosexual couples. This shows that A and B can still have very different conceptions of marriage even when they agree on what constitutes discrimination against homosexuals. Hence, we cannot conclude that they have a similar understanding of marriage based on their similar conceptions of what constitutes discrimination against homosexuals. It is also noteworthy that A’s and B’s conceptions of marriage are not contradictory and can even be consistent. Nonetheless, they are still substantially different. This further shows that we cannot

39 I am indebted to Karyn Lai for helping me set up this analogy.

154 5. Xin and Xinga trivialise the differences between two thinkers just because their definitions are not contradictory or consistent. If what is at issue is A’s and B’s conception of marriage, then our focus should be on why A thinks that marriage is between opposite sex couples while A thinks it also includes same-sex couples, rather than their understanding of what constitutes discrimination against homosexuals. Similarly, Xunzi and Mencius could have very different conceptions of xinga and it will be problematic to argue that they have similar conceptions of human nature on the grounds that they have agreement on what constitutes goodness or that their understanding does not contradict each other. Laying bare their agreements about other issues and the logical consistencies of their claims is certainly important as it provides leads that enable us to probe into the substantive differences between the two thinkers’ conceptions of xinga. However, it becomes problematic when their agreements about other issues or the consistencies in their views misguide one into thinking that their very conceptions of xinga are also similar. This questionable approach is what is under scrutiny here. Many scholars, in order to establish some connection between Mencius’s and Xunzi’s views, are keen to identify similarities in their views. Then, on the basis of these similarities (for example, A and B on (c)), they go on to suggest that there might be other similarities (for example, A and B on (a)). Some approaches, for example, focus on the “goodness” element and suggest that Mencius and Xunzi hold similar views. From this, they conclude that Mencius and Xunzi have similar views of xinga. We can see from the above analogy that this is problematic. In what follows, I discuss the differences between Xunzi’s and Mencius’s concept of xinga from the perspective of their different understanding of xin. Given Mencius’s concept of xinga is itself a vast topic, I will not be able to provide a comprehensive account of his theory of xinga. Hence, I limit my discussion to Xunzi’s and Mencius’s different conceptions of xin in order to highlight an area of substantive difference between their views on xinga. Even so, I might not be able to fully represent Mencius’s position on xin in this thesis.

155 5. Xin and Xinga

5.2.1 Mencius on Xin

In Mencius 6A:6, Mencius famously explains his claim that xinga is good by identifying four aspects of xin. According to Mencius, all humans have the xin of compassion 9ᬜ, the xin of respect ;<, the xin of shame =പ, and the xin of right and

40 wrong >?. In 2A:2, these four xins are identified by Mencius as the sprouts or beginnings (duan ১) of the four cardinal ethical ideals: ren I, yia  li , b ড , and zhie ޡ. Mencius gives the example that human beings all have alarmed and distressed xin upon seeing a child about to fall into a well. Another related example says that people will sweat upon seeing the remains of their parents being eaten by foxes and sucked by flies. Mencius notes:

@A ?ჽࡥAB಴CDEFGHIJKึ

The sweating was not for others to see. The innermost xin arrives at the face. They went home for baskets and spades and bury it [the remains of their parents].41

We can infer from the above examples these four xins for Mencius are akin to sentiments, which can cause changes in bodily states and naturally prompt one to act. This further suggests that xin has predisposed inclinations that are directed towards goodness. Hence, moral cultivation should be grounded in this affective aspect of xin. In addition, it implies that these predisposed sentiments can be easily translated into action and do not require much effort. This is supported by Mencius’s assertion that the reason a king fails to be a true king is not that he is not able to (bun eng  ) but that the king does not do (bu wei

ჽ) what a true king does. For Mencius, for a king to claim that he cannot show kindness to the people is like someone who can lift up something that weighs hundreds of kilograms but claims he cannot lift up a feather. The king in this case is simply not using

40 I follow D. C. Lau’s translations of Mencius’s four descriptions of xin (Lau, 2003, p.125). 41 Mencius 3A:5.

156 5. Xin and Xinga the strength he has.42 However, the fact that there are people who just do not practice goodness implies that something could have prevented one from exercising xin. I will return to this passage. For now, let us consider how it is possible that one is not prompted to be good even when the person’s xin has predisposed inclinations towards goodness. One clue to understanding what prevents the proper manifestation of xin lies in Mencius’s understanding of the relationship between xin and the body. In 6A:7, Mencius discusses the similarity between xin and the bodily senses. Mencius observes that the sense organs have common preferences in things. For example, everyone’s ears are fond of the music of Shi Kuang MN and everyone’s eyes are fond of the beauty of Zi Du O

P. Just as how the senses have similar tastes, so too Mencius claims that everyone’s xin has a common preference for yia.43 If we understand the analogy between the senses and xin in a strict sense, then we can hold Mencius to mean that xin is blindly drawn towards yia if there is no intervention. However, the analogy is probably not meant to be taken in a strict sense. It is suggested that even though xin naturally prefers yia, xin is not blindly drawn towards yia as the way the senses prefer their proper objects. Mencius speaks of the difference between xin and the senses in terms of a big part opposing a small part in 6A:15:

QFRSTDUUࡕUWXఉZ಴RWSSWಚSWಚ

\ ᕸ^ᒮ_`a॥bᒮW॥cᒮd e \ჽbࡥఉZ

The organs of hearing and sight are unable to think (si) and can be bi (beclouded, obscured) by external things. When one thing meets with another thing, they are led away by it. But the organ of xin thinks. If it does not si, then it does not obtain. This is what Heaven has given me. By first establishing the great part, the small part cannot duo (seize, snatch). This is being a great person.

42 Mencius 1A:7. 43 See section 5.2.2 for detailed discussion.

157 5. Xin and Xinga

A cursory reading of the above passage tells us that xin, the great part, can rule over the

b bodily senses, the small part. Presumably, the capacity to think (si S) is what enables xin to be the greater part.44 Since the sense organs do not have the capacity to sib, they will be obscured (bi T) and led astray (yin X) by external objects. By contrast, xin can sib and therefore it can rule over the senses. One implication we can draw from this is that there can be a potential rivalry between xin and the senses. What is it to be obtained when xin sib then? In 6A:6, we again find a connection

b between si and obtaining (de ಚ):

ࡦᑶ6ཅ?hij^ ^஭ྞ dSQZmnoWಚpWq

Ren, yi, li, zhi are not externally welded to me. I originally have it, just that I do not si. If pursue, then de it; if put aside, then lost it.

We can infer from the above passage that yia is one of the objects of de. If Xunzi thinks that xin has to sib in order to obtain yia, then it seems that the response of xin towards yia is not as immediate and direct as the response of the senses to their proper objects. Although xin will obtain yia if it sib, the picture seems to be more complicated than that. Apparently xin does not necessarily sib and there are at least two possible explanations for why it is the case. One explanation is that, for Mencius, xin can be distracted by external objects. In this case, it is in virtue of sib that xin can guard against such distractions so that it can obtain what it really likes. Alternatively, it could be the case that xin is not always in an activated state and sib is the step that activates xin to seek what it is supposed to obtain. These two positions are compatible with each other. Indeed, there is textual evidence for both positions. In 2A:2, Mencius identifies

f zhi ೂ as the commander of qi s (vital energies) while qi is also said to be capable of

44 I am aware that si is a very important in Confucian moral cultivation. The English translation “think” is far from adequate but since it does not significantly affect our current discussion, I will follow the general translation of si as “think” for convenience.

158 5. Xin and Xinga moving zhif. Ideally, when zhif leads qi, one will have an unmoved heart.45 Zhif is usually translated as the “will” but in the context of Mencius, it does not denote a kind of power that pertains only to the mental realm. Rather, it is primarily concerned with xin being oriented towards a goal.46 This idea about xin being focused on its proper goal is also expressed in 6A:9, wherein Mencius explains why the “sprout” of a king is not growing by an analogy with learning chess.47 According to Mencius, learning chess itself should be as simple as nurturing a sprout. However, a student is not going to obtain (de) the art of the game if his xin is not focused enough to bring about zhif. Mencius’s emphasis on ensuring xin is focused on its goal, along with his understanding that vital energies can move zhif, suggests that xin can be distracted from orienting towards its proper goal. And what distracts xin is most likely to be desires that arise from interaction with external objects. Hence, Mencius maintains that there is no better way to nurture xin than by reducing bodily desires (7B:35). Granting that Mencius does think xin can be distracted in its pursuit of yia, Mencius clearly does not want to say that xin itself has bad inclinations. In 7B:21, Mencius draws an analogy between an unused trail overgrown by grass and an unused xin stuffed with grass. There are also references describing those who let the body take charge of things as having abandoned (fang t) their xin (6A:8, 6A:11). Common across these passages is Mencius’s insistence that xin itself is good. In order to make coherent sense of Mencius’s claim that xin can be distracted from its proper goal and the claim that xin itself is good, we can understand Mencius as saying that xin is in an inert state when it is distracted. This leads to the observation that, for Mencius, although xin is the seat of moral exertion, it is not necessarily in a ruling position because it is not always active. Hence, the key to moral cultivation is to ensure xin is active so that it is in the ruling position to control the senses and pursue what it should be pursuing. This explanation also allows us to make better sense of Mencius’ famous dialogue with King Xuan,

45 Mencius 2A:2 is a passage that has been extensively explored by scholars. Here, I will only focus on how zhif relates to xin and affects xin. See, for example, Riegel 1980, Nivison 1996c, Shun 1997b, Alan Chan 2002. 46 Shun, 2004, p.186. 47

In this passage, Mencius uses the term “meng Õ,” which can be literally translated as “sprout.” Presumably, it is another metaphorical way of referring to the ethical predispositions of xin.

159 5. Xin and Xinga described in 1A:7. According to Mencius, it is not that the King is not able to be kind (bu neng  ). Rather, the king is simply not acting out his kindness (bu wei ჽ). In

Mencius’s view, the King has all the resources ready in his xin to pursue goodness. But since his xin is deactivated, his xin is simply not pursuing goodness. Another interesting observation that can be made from 1A:7 is Mencius’s assumption that xin itself is the standard for making judgements. In an attempt to make the King aware that he is misguided in thinking that he can extend his bounty to animals but not to his people, Mencius urges:

u %vwxి %v᪖{Uጳ ಴ჽ}~ి

Weigh, then weight can be known; Measure, then length can be known. It is so with all things, even more so with xin. Your Majesty, please measure it.

This reveals Mencius’s assumption that a simple act of weighing will reveal the value of things. In 6A:10, it is said that even if one likes both fish and bear’s paw, one will choose bear’s paw over fish when the two cannot be chosen at the same time because bear’s paw is more valuable. Likewise, when life and yia cannot be chosen simultaneously, one will certainly choose yia over life. Mencius then concludes that those who choose material

a 48 gains at the expense of violating yi have lost their original (ben €) xin. It is worth nothing that Mencius takes xin to be the basis of choosing and thinks that one will not be mistaken about the value of things when she chooses in accordance with her xin. As soon as one weighs the two options against her xin, she will know which is the right option to choose. Problems occur at the level when one does not know that the two options are in conflict.

48 I understand “ben xin” to be referring to the xin that is original because it is consistent with the idea in 4B:12 that a great person does not lose her infant-like xin. Mencius’s emphasis is on retaining the initial state of xin.

160 5. Xin and Xinga

5.2.2 Differences Between Xunzi’s and Mencius’s Xin

On the proposed interpretation, the main difference in Xunzi’s and Mencius’s conception of xinga lies in their different opinions about xin.49 They disagree on whether xinga is good or bad because they disagree on whether xin is good or bad. While both thinkers have discussed xinga in terms of the preferences of xin, it is evident that they have very different understanding of the natural preferences of xin. According to Xunzi:



F‚ᕽQ‚„ੰ‚઎಴‚‡ˆ>ጳ Dࡥചᒮ 

The eyes like colours, ears likes sounds, the mouth likes tastes, xin likes lic…all these are products of humans’ qing xinga.50

Mencius also discusses xinga in relation to people’s xin having a common preference for yia. In 6A:7, Mencius observes that just as all human senses have certain common

a preferences, so too the xin of all humans also has a common preference for yi ᑶ:

ੰD઎ྞ‰ᒯ ‹Œ_ಚ^ੰ ᒯᒮ ŽੰD઎ ॥ᕸࡥᗆ‘

’ᕸ^‰“ W*”ᒯጳ•‹ŒD઎ ੰD઎ ྞ‰ᒯᄮ m቗

ᑶೲ^಴˜™šೲ^ੰ

Mouth has the same preferences in taste. Yi Ya was the first to discover what my mouth prefers. If it were the case that the xinga of mouth to taste differ among humans like hounds and horses differ for me in kind, then why all mouths in the

49 Indeed Xu Fu-guan has made the observation that Xunzi has included both yu and the capacities of faculties in xinga, which means that the capacities of xin are also included in xinga. But Xu also thinks that Xunzi has not included the capacities of xin in the claim that xinga is bad. Hence, Xu argues that there are contradictory elements in Xunzi’s theory of xinga (Xu, p.255). 50 Xunzi 23/25-6; K 23.2a/3/154.

161 5. Xin and Xinga

world follow the taste of Yi Ya…Hence, lia yia please my xin in the same way as 51 fine meats please my mouth. 

As the above passage shows, Xunzi and Mencius have both included the preferences of xin in their conceptions of xinga. Their differences lie in their understanding of what xin naturally prefers and of what makes it possible for xin to find pleasure in goodness. To help us see their differences more clearly, we can first sort out two implications of Mencius’s view as stated in 6A:7:

(1) xin naturally prefers yia and this preference for yia is akin to taste preference; (2) Xin has the capacity to distinguish and enjoy yia just like the sense organs have the capacity to distinguish and enjoy their proper objects.

Xunzi disagrees with (1). Xunzi argues that xin naturally prefers lici, not y a. Recall that lic for Xunzi only concerns what benefits the self and is juxtaposed with yia, which requires one to abide by standards and disregard considerations about what benefits the self. In Xunzi’s own words, xin naturally tilts. It is through learning and knowing Dao that xin can be shaped into a rectified position so that it is not attracted by lic but rather follows yia.52 This imagery of altering and setting xin straight is very different from the xin that is growing and expanding on Mencius’s picture. To be precise, Xunzi’s view differs from that of Mencius in two ways. First, he does not regard xin as having any natural

51 a a

Cf. Lau 2004, p. 127. Mencius actually says that xin prefers “li yi Ö × .” The appearance of the term “lia” in this context is ambiguous. Commentators and translators tend to understand both that lia iand y a are used to denote an ethical ideal or standard exhorted by Mencius. Zhao Qi understands lia in connection with Dao such that a person of lia is someone who has obtained dao. Zhu Xi understands lia as that which resides in a thing and yia as that which is being acted out. James Legge follows along Zhu Xi’s interpretation and translates lia as “principles of [human] nature.” Other translators like W.A.C.H. Dobson and D.C. Lau translate lia as “the underlying principle and the essential Justice” and “reason and rightness” respectively. Recently, Bryan Van Norden translate lia as “order” with reference to the neo-Confucian Cheng Yi’s understanding of liai and y a. However, there seems to be little evidence in the Mencius that suggests lia is also regarded by Mencius as an ethical ideal like yia. 6A:7 is the only passage that has the two terms “lia yia” appearing together. Moreover, aside from this passage, Mencius has not treated lia as a philosophical term elsewhere in the text. Passages 5B:1 and 7B:19 are the only two other passages in which the term lia appears, though lia is not highlighted as a topic of discussion. In 5B:1, lia is used to mean harmonious musical pattern. And in 7B:19, lia has the meaning of what is favourable or smooth. None of these two usages, however, suggests that lia had already acquired the connotations of reason or principle in Mencius’s usage. 52 See discussion in section 3.1 concerning Xunzi’s usage of lic.

162 5. Xin and Xinga preference for yia. Second, he believes that xin has a natural preference for what is opposed to yia. This observation allows us to understand the substance of Xunzi’s disagreement with Mencius. If we accept that yia, for both thinkers, is an attribute of moral goodness, we can better appreciate Xunzi’s claim that xinga is bad is not merely arguing against the case that xinga is good. If Xunzi were simply arguing that it is not the case that xinga is good, Xunzi could still be committed to the claims that xinga is neither good nor bad (xin prefers neither lic yior a ) or that xinga can be good or bad (xin prefers both li and yia). However, Xunzi specifically wants to say that xinga is bad because xin in its natural state only prefers lic. In relation to (2), Xunzi argues that xin does not have the capacity that specialises in distinguishing and taking delight in yia. All Xunzi can concede is that there is a distinguishing capacity of xin that makes the enjoyment of yia possible but the capacity itself does not have any predisposed content or direction of development. It is true that on my interpretation proposed in section 5.4, Xunzi might also think that it is possible for xin to eventually find enjoyment in yia. However, the reason xin can take delight in yia is not due to some distinguishing capacity that has yia as its content or goal. As earlier discussed, lia is the only basis xin inherently has. It is because xin knows Dao that xin changes its predisposed basis to that of yia. Hence, what allows it to enjoy yia is the capacity zhic — an inborn capacity that does not have any content or predisposed direction for development.53 This also explains why there are numerous passages in the Xunzi that say xin is juggling between goals and is therefore uncertain about what is right (huo ഠ). On the other hand, there is no such reference in the Mencius because the distinguishing capacity of xin already has yia as its content or its development is directed towards yia. Because of this distinguishing capacity, xin’s fondness of yia springs naturally from what is inherent in xin. Xunzi’s xin does not have this capacity with predisposed content or direction of development and hence, its fondness for yia has to be acquired externally from learning. In its natural state, Xunzi’s xin would take delight in a lot of bad things in which Mencius’s xin does not take delight. Even if xin eventually comes to prefer yia, it is not because it is naturally satisfying to xin but because Dao is indeed satisfying. For

53 See section 3.2 for three senses of zhic.

163 5. Xin and Xinga

Mencius, xin’s preference for yia is the key to goodness and it occurs prior to carrying out yia acts. For Xunzi, xin’s knoweldge of Dao is the key to goodness and xin’s preference for yia, if there is any, occurs after xin has carried out yia acts. In this light, we see that Xunzi would agree with Mencius in saying that goodness is pleasant to xin but dispute Mencius’s view that xin naturally finds yia pleasing. The former concerns the nature of goodness, whereas the latter concerns the nature of xin. It is on the latter that the two thinkers differ. If my interpretation of Xunzi and Mencius is correct, we can summarise the differences between their views on xin as follows:

a. For Xunzi, xin has a predisposed inclination to only seek what satisfies the self and has no natural tendency to appreciate goodness, or seek it. If the natural inclination of Xunzi’s xin is followed without external intervention, undesirable consequences will ensue. Mencius, on the other hand, believes that xin is predisposed to appreciate goodness and has an inbuilt orienation towards goodness. If the natural inclination of Mencius’s xin is followed without external intervention, goodness will ensue. This also makes sense of the observation that the scope of xinga for Xunzi and Mencius are seemingly different. Xunzi has included in xinga the problematic inclination of xin to satisfy yu; Mencius does not understand xin as having such problematic inclination to satisfy yu but rather understands yu as what obstructs the natural movement of xin.

b. Since Xunzi has included problematic inclinations in xin, he sees external intervention that drills knowledge about Dao into xin as essential to the rectification and transformation of xin. The role learning plays for Xunzi’s xin is to prevent xin from following its natural inclinations. Since Mencius regards xin as thoroughly good, he understands external intervention in the sense of the presence of external objects that will thwart the natural course of development of xin. The role learning plays for Mencius’s xin is to facilitate the growth and expansion of xin’s natural inclinations.

164 5. Xin and Xinga

c. When Xunzi talks about dispelling bi T, his emphasis is on making xin know

Dao so as to ensure that xin is no longer guided by its natural inclinations. By contrast, the reason Mencius sees bi as problematic is that external objects can distract xin from following its natural inclinations. For Xunzi, it is because xin is directed towards badness that it is bi. For Mencius, it is because xin is bi that xin is not directing towards goodness. d. Xunzi thinks that xin is always in an active state and is always in a ruling position. There is no conflict between xin and the senses (or sensuous desires) on Xunzi’s picture. Mencius’s xin is not necessarily functioning or functioning properly. This gives rise to the possibility that the senses (or sensuous desires) can seize the authority of xin. e. Like Mencius, Xunzi thinks that xin does not necessarily weigh options before it makes decisions. Once it gains a comprehensive view of the consequences of its decisions, it will start weighing options. Unlike Mencius, Xunzi thinks that even when it weighs, the right choice will not be made if it is weighed on what xin naturally likes. Xunzi is adamant that Dao has to be the standard for making choices. Mencius, on the other hand, thinks that the problem occurs only at the level when one does not know that she has to weigh options. Once the two options are weighed on xin, the right decision will automatically be made. For Mencius, xin has to be the standard for making choices. f. The reason xin can be good, in Xunzi’s view, is that xin has the distinguishing capacity zhic that makes knowing Dao possible, although the capacity itself is not content-specific or has predetermined direction towards Dao. On Mencius’s view, xin has a distinguishing capacity that already has goodness as its content and the development of which is directed towards goodness. Hence, xin is already good in itself.

165 5. Xin and Xinga

5.3 Four Complications

The above clarification helps to tackle two complications in Xunzi’s view. In the “Wang

a Ba ~œ” chapter, Xunzi makes the famous claim that human beings have yi , a claim that has led many scholars to perceive a tension in Xunzi’s thought.54 According to Xunzi:

žྞsŸ ¡ྞ Ÿv¢ᇿྞvŸᑶࡥྞsྞ ྞvࡗ¥ྞᑶm

¦ჽ*§ …ࡥ ᑱ© ᑱ ࡥ”ࡾ ᑱnিি”ࡾ «nᑶ

Water and fire have vital energies but have no life. Grass and tress have life but have to zhic. Birds and beasts have zhic but have no yia. Humans have vital energies, life, zhic,i and also y a…Humans are able to form into societies but [animals] cannot. How are humans able to form into societies? I say: fen. How are humans able to fen? I say: yia.55

David Nivison understands the above passage as saying that yia is the distinguishing property of humans and therefore, yia is an aspect of xinga.56 Nivison’s move from identifying yia as a distinguishing property to identifying yia as an aspect of xinga is perhaps too quick. Xunzi seems to have deliberately drawn a distinction between characteristics given at birth and distinctive characteristics of humans. In a passage in the “Fei Xiang ෤” chapter, Xunzi points out that humans have certain common characteristics, such as wanting food when hungry and wanting warmth when cold, that even the sage Yu and the tyrant Zhou share. For Xunzi, these characteristics are given from birth and do not require any effort to acquire. But what makes someone really a

57 human being is the ability to draw distinction — bian ᡝ. Xunzi contrasts these common characteristics given at birth and the characteristic that distinguishes humans

54 See, for example, Tang, 1974; Nivison’s “Xunzi on ‘Human Nature’” in 1996c. 55 Xunzi 9/69-71; K 9.16b/2/103-4. 56 Nivison, 1996c, pp.206, pp.211-3. 57 Xunzi 5/23-28; K 5.4/2/206.

166 5. Xin and Xinga from other beings, suggesting strongly that the latter is not given at birth and requires effort to develop. If so, by Xunzi’s definition, this distinctive characteristic of humans does not reside in xinga. Even though there is another reference that says the senses’ ability to bian their proper objects is given from birth, Xunzi is probably not using the term “bian” in the same sense in this context. When Xunzi identifies bian as the distinctive characteristic of humans, he says that what makes a human a human rather than a wild animal is that humans have the ability to bian. Since Xunzi would probably agree that an animal also has senses to bian their proper objects, his usage of “bian” here is likely to be different from his usage when he talks about the senses’ ability to bian. When Xunzi talks about bian as the distinctive characteristic of humans, his emphasis is clearly on the ability to draw some ethical or social distinctions, such as the proper affection between father and son and the proper separation between man and woman, that can only be developed by learning. Thus, it is unlikely that Xunzi has included distinctive characteristic of humans in his concept of xinga.

Furthermore, we observe that there is a close connection between bian and fen ি, where “fen” is usually understood as the allocation of social divisions. For example, it is said that it is most important for one to exercise the ability to bian in order to fen.58 Elsewhere, it is said that only someone who is most able in bian can fen the world.59 We can infer from this linkage that the ability to bian is what enables humans to fen. And since Xunzi regards yia as what enables humans to fen, we can infer that bian and yia are intimately related. Given our understanding of yia, Xunzi probably considers yia as the proper basis for bian. This explains why Xunzi seems to think that both yia and bian are characteristics that make humans distinctive. The finding that the ability to bian is acquired through effort is also consistent with the proposed interpretation that yia is acquired through effort. Hence, Xunzi’s claim that humans have yia does not amount to the claim that yia resides in human xinga. Xunzi is saying that yia is an attribute that only belongs to human beings as a kind. This, however, does not entail that every single human being has yia. If one does not make efforts to acquire yia, then, to Xunzi, she is not much different from

58 Ibid. 59 Xunzi 18/24-25; K 18.2/3/35.

167 5. Xin and Xinga wild animals (qin shou ¢ᇿ). It is clear from the text that Xunzi thinks one can be a wild animal, as evidenced by Xunzi’s condemnation of those who abandon the goal of learning as wild animals60 and of those who do not follow yia as acting as if they are wild animals.61 Since xin has the capacity to zhic, it becomes possible for one to acquire yia and therefore become truly human. However, since zhic does not have yia as its predisposed content or direction, it is also possible that a human being is never able to become a genuine human being. This also helps to make sense of the distinction Xunzi draws

62 between those capable of knowing (ࡾv) and those capable of doing (ࡾ ).

My proposed interpretation seeks to offer a solution to a complication in Xunzi’s thought concerning the sage in the Xunzi. One of the perplexing questions often raised by scholars concerns how sages could be sage. Interpreters tend to perceive a tension in Xunzi’s thought in the following way: Xunzi thinks that humans can only be good. Given the sages themselves are already sages, they cannot learn from anyone beside themselves. If sages have never had a teacher, then how did sages become sages? The main proposal to solve this quandary says that there is already something built into xinga as an “innately positive trait”63 and the sages work on this innate moral sense to develop full-blown goodness. But as we just saw, Xunzi does not take yia to be innate in humans. But if the sages do not have predisposed traits that are congenial to goodness to work on, how do sages become sages? There are three possible ways of understanding how sages grasped Dao: (1) sages’ xin by default are not tilted by objects, (2) sages obtained some esoteric knowledge of Dao, or (3) sages obtained some exoteric knowledge of Dao. Given our previous discussion, (1) and (2) are unlikely because Xunzi is adamant that sages and lay people are the same. Xunzi explicity says that wei is the sole reason why a sage is a sage64 and that it is possible for everyone to become a sage.65 In Xunzi’s view, the sages and

60 Xunzi 1/28; K 1.8/1/139. 61 Xunzi 4/34, 19/20. 62 Xunzi 23/63-4. Cf. See Shun, 1997, pp. 222-231 and Kim-Chong Chong, 2007, pp.67-72 for discussions of the distinction between ke and neng. 63 Munro, 1996, p.198. 64 Xunzi 23/29. 65 I am indebted to Professor Riegel for the comments he made on this passage in one of our reading sessions on the Xunzi.

168 5. Xin and Xinga common people share the same nature and yu. So sages’ xin should also be naturally drawn to objects of yu. And if a common person submits himself to learning, he too will become a sage.66 We can reframe the question in terms of whether moral training is about learning a paradigm, learning from the sage, or learning something objective, the Dao. When Xunzi says one should learn “the way of sages,” is the focus on the word “way” or the word “sage”? For the former, the object of inquiry is the way itself, whereas the latter focuses on whatever the sages do. This may be explained more clearly with an example. Suppose that there is an ideal way to swim such that swimming in this ideal way can minimise water resistance and maximise speed. Call it I. For Xunzi, this I is objectively

a there, as part of li ቗. It is simply a matter of whether or not humans can recognise it and subsequently act in accordance with it. For Xunzi, whoever obtains I is a practitioner of I. Let us also suppose that John has found this ideal way I. Then students of swimming should aim at obtaining the content of I. In this case, a practitioner of I would be the most eligible figures to teach students of swimming how to acquire I and in this case it happens to be John. Is Xunzi saying that a learner should learn to match her knowledge with John or a learner should learn from John? If it is the former, even someone who had never met John or heard about the teachings of John has the possibility of becoming a perfect swimmer. She might figure out how her body works best in water through her own accumulated experiences and self-practice. If she were successful, then both this self- directed learner and John would be practising the same movements of swimming, that is, I. If it is the latter, then the person is doomed and it is unclear how John would come to be the perfect swimmer in the first place because John cannot learn from John’s swimming style. Applying this to the sage issue, Xunzi seems to be saying that there is an ethical ideal (I) and there are some practitioners of I. When common people try to learn about the content of I from someone, they should learn from the practitioners of I. Xunzi labels these practitioners of I as “sages.” It would be more precise to say that the common people should aim at matching their knowledge and actions with those of the sages, rather

66 Xunzi 23/61-9.

169 5. Xin and Xinga than common people should aim at learning from the sages. This gives rise to the possibility for someone who has been never exposed to the teachings of sages to obtain the knowledge and actions that sages have. They might figure out the ethical ideals from their own life experiences. Although the sages themselves become paradigms, the paradigm comes after grasping of the objective way. The first sage should be a self- learner who observes and learns about lia and hence is able to act morally. It is possible that some situations just happen to arise in a sage’s life so that the first sage accidentally started performing Dao-abiding acts. Since following Dao is in fact satisfying, the sage eventually grasped knoweldge of Dao. Similar understanding is found in Schwartz, who argues that: “The sages may have set out from a self-regarding impulse to escape from wretchedness, but through their enormous intellectual exertions and insights they achieved ccomplete success in internalising within themselves the spirit of ‘li and righteousness.’”67 Even though self-learning is possible, Xunzi would maintain that it is extremely unlikely for one to successfully teach oneself about morality because that requires both the right circumstances and a high level of intelligence. The way of the sages is a short cut to know lia because the sages have already detected lia and know the right way to weigh things. Since Dao remains unchanged through time,68 and since the sages have already figured out the Dao, people can simply rely on the manual provided by the sage to learn about morality. Should one decide to learn without the manual handed down by sages, it is still possible for her to be morally cultivated, though Xunzi might remain sceptical of such a likelihood. Xunzi maintains that it is rare for someone to know Dao without the guidance of teachers, just as it is possible to walk across the world but no one has been able to do it.69 Finally, let me turn to two complications in Xunzi that cannot be satisfactorily solved by the proposed interpretations of Xunzi’s concept of xin. The first one concerns

a Xunzi’s concept of hua xing ­, which is commonly understood as the transformation of human nature. According to Xunzi, the teachings of sages have a transforming effect

67 Schwartz, p.300. 68 Xunzi 22/74, 5/35-6. 69 Xunzi 23/73.

170 5. Xin and Xinga on the individual. It is said that if the practices and the environment is right and effort is accumulated in learning, the bad nature can be transformed (hua ­) and the basic

70 substance can be moved (yi ®). This transformation can be understood as a transformation from one’s only wanting to gain whatever is beneficial for oneself to not even wanting to see what is contrary to the teaching of sages71 and deferring food to others.72 My account of Xunzi’s xin can make sense of how such a transformation is possible. As discussed before, once xin adopts Y as its basis, it will disregard considerations of what benefits the self and follow what it is accordance Dao. However, it is unclear (a) at what point hua occurs and (b) whether hua entails a permanent change in xinga. I cannot provide a definitve answer to (a) because it is unclear at which point xin finally knows that following Dao is the only way to obtain satisfaction.73 Although Xunzi stresses the importance of accumulating learning and practice, it is unclear how much accumulation is sufficient. The difficulty with (b) is associated with my earlier point that there always seems to be a possibility, in Xunzi’s view, for xin to revert to adopting L after it has adopted Y.74 This tension in Xunzi’s thinking is reflected in his two seemingly inconsistent remarks about hua. In one place, Xunzi defines hua as “a change in appearance but not in what is real.”75 Elsewhere, Xunzi describes a person who has been hua as not returning to the original state.76 While the former seems to imply that there can be a permanent change in one’s xinga, the latter suggests that xin will always have problematic inclinations. Since there is insufficient evidence that allows us to reach a definitive answer, I will leave my discussion of this problem open-ended. Another difficulty in understanding Xunzi’s xinga concerns Xunzi’s claim that

“humans desire to be good 4ჽ૽.” According to Xunzi:

70 Xunzi 8/110-1. 71 Xunzi 1/47. 72 Xunzi 23/18. 73 See section 4.5. 74 See section 4.3. 75 Xunzi 22/28. 76 Xunzi 3/34.

171 5. Xin and Xinga

দࡥ4ჽ૽ᒮჽപ @ᚵ°੅പ°²ᇬ°౤µ°¶·°§¸ŸBᒮಶ

oDi

As a general rule, the fact that humans desire to be good is because xinga is bad. Those with very little wish to have much, those who are ugly wish to be beautiful…whatever one does not have from within will certainly seek from without.77 If Xunzi were to think that human beings want to be good, then he would be committing to the claim that xin inherently has Y.78 Since Xunzi cannot be saying xin inherently has Y on the proposed interpretation, then we are left with the option of understanding the above claim as a descriptive claim. In this case, Xunzi is only describing from an outsider’s perspective that the reason humans desire goodness is that they have a lack of it. Similarly, an outsider may describe the reason I desire to go to Café D is that I want French food and Café D is the only French restaurant.79 However, such a descriptive claim is independent of how the agent feels herself. Just like how I had no desire to go to Café D before I learnt that it is the only French restaurant, human beings as agents themselves do not have desire to be good before they learn that it is only by following Dao that they can obtain satisfaction.

77 Xunzi 23/32-4; K 23.2b/3/154. 78 See discussion of the second and third interpretation in section 4.2. 79 See section 4.3 for discussion on the two perspectives used to describe xin’s reason for abiding by Dao.

172 Conclusion It is worth noting that, in studies of Xunzi, discussion of Xunzi’s concept of xin tends to place asymmetrical emphasis on the cognitive side of xin. In earlier English language studies of Xunzi, the term “xin” is usually translated as “mind.”1 Benjamin Schwartz, for example, argues that, “In Xunzi it would be far more appropriate to translate the world xin simply as ‘mind,’ since it by no means comes into the world directed by innate moral propensities.”2 This is probably the view that is also assumed by Antonio Cua. In the article “Ethical Significance of Shame,” Cua explicitly points out that Mencius’s usage of “xin” should be more appropriately rendered as “heart and mind” because it has both affective and cognitive aspects.3 However, as Cua moves on to discuss Xunzi in the same article, he does not make the same remarks about Xunzi’s xin but continues to discuss it simply as “mind.” Since the early 1990s, there has been a growing trend in English-language studies of pre-Qin Chinese philosophy to render the term “xin” as “heart/mind.”4 Recently, scholars are also starting to render Xunzi’s “xin” as “heart/mind.”5 However, a satisfactory explanation is seldom provided to justify such a translation. Although Kurtis Hagen translates “xin” as the “heart-mind” and acknowledges that xin can be influenced by feelings, Hagen remarks that, “If one were forced to choose, in the case of Xunzi, the choice would have to be the ‘mind.’”6 It is not difficult to gather the impression that scholars still understand the cognitive aspect as that which underlies Xunzi’s concept of xin. Such an asymmetrical emphasis on the cognitive aspect of xin may be partly explained by an antagonistic understanding of xin and yu in Xunzi. The assumption that both xin and yu have causal roles in action serves as one of the major bases for AI interpreters to understand xin as what overcomes yu, where yu is understood as the source of disorder. Antagnoisitic interpretation of this kind is called into question in Chapter 1. In Chapter 2, I argued that only xin can cause action and that yu is a concept that is in a

1 For example, Dubs, Watson, Wing-tsit Chan and Knoblock all translate “xin” in the Xunzi as “mind.” One exception is Graham, who translates xin as “heart” (1989, pp. 251-5). 2 Schwartz, p.293. 3 Cua, p.219. 4 See, for example, Yearley 1990, Shun 1997, Geaney 2002, Slingerland 2003. 5 See, for example, Stalnaker 2006, Chong 2007, Hagen 2007. 6 Hagen, p.160. Conclusion different conceptual category from xin. In Chapter 3, we saw that xin itself is the source and direct cause of disorder because it has a predisposed tendency to only seek what satisfies the self. This allows us to see that Xunzi’s xin, too, is guided by feelings and will always be guided by feelings if xin does not know Dao. Xunzi is targeting this affective aspect of xin when he claims that xinga is bad. But since xin has the capacity to know and stand back from feelings when it makes choices (Chapter 4), it can also change its reason for judgement from that of pursuing personal satisfaction to that of following Dao (Chapter 5). This cognitive aspect is, in Xunzi’s view, what allows humans to be good. Since the affective and cognitive aspects are equally emphasised in Xunzi’s concept of xin, it is more appropriate to render the term “xin” in the Xunzi as “heart/mind.” I raised the question of whether xin is discussed as equivalent to the person in Chapter 2. As previously discussed, Xunzi seems to write as if xin is a capacity of the person and sometimes as if xin is itself the agent that is equivalent to the person. Given the textual evidence we have, I am not able to provide a definitive answer to this question and have left it open-ended. Throughout my discussion of xin, I cannot avoid moving back and forth in discussing xin as if it itself is an agent and sometimes as if xin is a capacity of the person.7 I have also not solved satisfactorily answered the question of whether yu is a state of xin. Towards the end of my discussion of the relation between xin and yu, I raised the possibility that xin and yu are not distinct entities for Xunzi.8 Recall the following passage:

º»ᕽ቗ࡾF¼ᒵ¾࿾ႚÁÂÃᒵࡾQ¼Äᗇᯄ࿦ÈᤇÃ઎ࡾੰ¼ÊËÌÍ

ÎᕓÐÑÃËࡾÒ¼ÓÔটᅘ၀Øwxࡾº»¼ÙÚmÛÜÝÞßàപ4ࡾ

಴¼

Forms, colors, and designs are differentiated by the eye. Pitch and timbre, bass and treble, modal keys and rhythm, and odd noises are differentiated by the ear. Sweet and bitter, salty and bland, pungent and sour, and distinctive tastes are differentiated by the mouth. Fragrances and stenches, perfumes and rotten odors,

7 See section 2.3. 8 See section 2.2.

174 Conclusion

putrid and rancid smells, foul and sour orders, and distinctive strange smells are differentiated by the nose. Pain and itching, cold and heat, smoothness and roughness, and lightness and heaviness are differentiated by the body. (Shuo, gu,) pleasure, anger, sorrow, joy, like, dislikes, and yu are differentiated by xin.9

Since this passage asserts that feelings are differentiated by xin, and since yu is defined as a response of feelings, Xunzi seems to get very close at this point to saying that yu is issued from xin. On the other hand, Xunzi makes the point that senses have their own yu:

@ࡥചF4ᎫᕽQ4Ꭻᒵੰ4Ꭻ઎Ò4ᎫË಴4Ꭻâ\ࡋᎫᒮࡥച ಶ

ä å

It is the qinga of human beings that eyes yu the fullest (qi) of colours, ears yu the fullest of sounds, mouth yu the fullest of tastes, nose yu the fullest of smells, xin yu the fullest of relaxation. These five qi are the inevitable thing of the qinga of human beings.10

Could it be the case, then, that yu is a state of xin so that yu cannot exist independently of xin? This is a question I must leave open-ended in the present study. In the past five chapters, I dealt with a local question that concerns the nature of xin for Xunzi and put aside the task of assessing Xunzi’s argument. In doing so, I have explicated the assumptions Xunzi has adopted or would have to adopt in order to maintain the coherency in his argument. It is likely that some of the assumptions or questions I posed might not have occured to Xunzi and for this reason, it could be difficult at times to arrive at a solution that does justice to Xunzi’s view.

9 Xunzi 22/17-9; K 22.2d/3/129. 10 Xunzi 46-7; trans., K 11.4/2/156, modified.

175 Appendix

Situating Xunzi’s Concept of Xin in Context

This section examines the concept of xin in some of the major classical texts up to and around the time of Xunzi (except the Mencius, which we already discussed in Chapter 5). I hope to provide a general sense of how Xunzi’s discussion of xin can be understood as an engagement with a movement of thought concerning xin and also to trace some ambiguities in Xunzi’s understanding of xin to their source in other texts before and around his time. My approach is expository rather than argumentative. In the following discussion, I will conduct a general examination of the concept of xin in the pre-Warring States and Warring States texts. Rather than giving a coherent account of the concept of xin in each of the texts discussed, I limit myself to the task of highlighting certain features of xin so as to identify a broader picture of xin that will be helpful for our understanding of Xunzi’s engagement with this trend of thought. I also intentionally expose a lot of ambiguities about xin that are pertinent to the ambiguities we found in Xunzi’s notion of xin.

The Concept Xin in Pre-Warring States Texts

It is generally accepted that, for early Chinese thinkers, xin is tangible and understood as one of the organs. The Chinese character xin is a pictograph that represents the anatomical organ of the heart.1 In the classical texts, xin is often listed alongside the other bodily senses, that is, mouth, ear, eyes, and the body as an additional organ and is described as behaving like the senses.2 In this regard, we can understand xin as the organ of the heart. This suggests that xin is generally considered by early thinkers as one of the sense functions and is integrated with other senses in the body. In the pre-Warring States text, the concept of xin was barely discussed. In the Confucian tradition, xin had never been discussed as a proper concept before Mencius. In

1 Oshima, p.68. 2 Geaney, Chapter 3. Appendix: Situating Xunzi’s Concept of Xin in Context the Lunyu, the character xin appears in only five passages. None of these passages discusses xin as a philosophical concept or uses the term “xin” in a technical sense. In the Daoist tradition, “xin” appears in only six passages in the Laozi.3 When it comes to the Warring States period, there is a significant increase in discussions about xin.4 The major texts that pay noticeably more attention to the concept of xin and treat it as a subject of inquiry are the Mencius, the Zhuangzi, the Guanzi, and the Xunzi. The intensive discussions of xin in the Warring States period have somehow overshadowed scholarly attention on the conception of xin in the pre-Warring States period and the latter is often briefly described as offering very little information about xin. However, if we piece together the fragmented occurrences of xin in these pre-Warring texts, we can gather some nascent ideas about xin that is later shared and developed by latter States thinkers: xin comprises both cognitive and affective aspects, there is an intimate connection between xin and affective states, and xin is the locus of all actions.

5 Shijing æ᎞ (Book of Songs)

Xin in the Shijing is predominantly linked to the issuance of feelings, suggesting that xin is the locus of affective activities. For example, there are numerous instances where xin is characterised as sad6 and in a state of prolonged anxiety.7 These affective feelings of xin could be very intense, such that xin is sometimes described as if it were an agent that can at different times be drunk,8 mournful,9 burnt,10 and aches as if it is suffering from illness.11 Releasing the tensions of xin is depicted as one of the ways to

3 There are debates about the dating of most parts of the Laozi. I follow the usual interpretation and take it to be before the Zhuangzi. See Liu Xiagoan, 2005, pp.241-246 for a detailed discussion of the dating of the Laozi. 4 I am indebted to Professor Chen Guying for alerting me to this point. 5 In light of the consideration that the Shijing is composed by various authors across a wide span of time (c. 1000- c. 600 BCE), and that it is unlikely for all the individual authors to share the exact same view of xin, my analysis only aims to reconstruct a picture of xin as it is presented in the text. 6 E.g Shijing 1/6/1, 1/13/1. 7 E.g. Shijing 1/3/1. 8 Shijing 1/6/1. 9 Shijing 1/13/4. 10 Shijing 3/3/4. 11 Shijing 1/5/8; 2/5/9.

177 Appendix: Situating Xunzi’s Concept of Xin in Context resolve these problematic states of xin.12 These observations about how xin can be preoccupied with feelings show the active engagement of xin with affective activities. These feelings of xin are often a result of a person’s encounter with the external world, such as a girl’s xin becoming unsettled upon seeing someone she loves. The idea seems to be that xin can interact with external things and become affected by them. It is unclear whether it is the person or xin that is interacting with the external objects and whether xin is identical to the person. There is also an indication that these affective activities can make xin unsettled and restless. When feelings of distress or longing arise, for instance, xin would be described as in toil and agitation.13 This further points to a correlation between the affective states of xin and the functioning of xin, the inference being that xin is the seat of affect. It is already suggested in the Shijing that activities of xin guided by affective states are not desirable. The movement of xin guided by affective feelings is often portrayed negatively in terms of xin is being stirred14 or as having its melody interrupted.15 The view that xin may sometimes negatively affected by feelings is possibly built on the assumption that the more desirable condition of xin is calm and still. This can be observed in passages that describe conditions favourable to xin in terms of calmness and stillness. For example, when a girl does not see the man she yearns for, her xin is said to be worried. Once she sees him, her xin will be at rest,16 at ease,17 and be stilled.18 Similarly, when a superior person takes up office, the people’s xin will be brought to peace.19 These observations suggest the affective activities of xin can disrupt its restful state of xin, and that it is desirable for xin to be at rest. However, it is unclear whether the authors mean that xin is disrupting the person or that xin itself is brought to disruption. At times, the authors of Shijing discuss xin as if they are discussing the person. Although xin can be affectively stimulated by feelings, there seems to be no necessary link between the affective state of xin and action. In one song, a woman speaks

12 Shijing 1/2/3; 2/2/9. 13 E.g. Shijing 1/12/7. 14 Shijing 2/5/5. 15 Shijing 1/11/3. 16 Shijing 2/3/2. 17 Shijing 2/5/5. 18 Shijing 1/2/3. 19 Shijing 2/4/7.

178 Appendix: Situating Xunzi’s Concept of Xin in Context of having reluctance in her xin to leave her husband but still walks away from him.20 Similarly, although the xin of those who bid farewell to their loved ones is described as having an unpleasant feeling, nothing is done to stop their departure.21 This means that how one behaves does not always reflect what xin wants or how xin feels. The observation that there could be a discrepancy between the affective state of xin and the behaviour of a person hints at the presence of competing affective states or a cognitive component that holds xin back. Another place that is suggestive of cognitive functions of xin is found in a song describing King Wu’s uncertainty about going to war. In this song, King Wu is persuaded not to have two xins about going to war because a higher being is on the King’s side. The war referred to is the well-known battle of Mu Ye, in which King Wu defeated the tyrant King Zhou of Shang.22 Historical records show that the feudal lords had indeed attempted to persuade King Wu to attack King Zhou about two years before this song was written. But King Wu replied at that time that he did not know the decree of Heaven and hence did not invade King Zhou. This could also explain why “the presence of a higher being” is offered as a reason to dispel the uncertainty of King Wu. In light of this historical background, the expression “two xins” can be taken to mean that xin is caught between two options or the person is undecided about a certain issue. This further suggests that there is cognitive faculty that can entertain the different wishes of xin and adjudicate between options. Although these occurrences of xin strongly hint at the presence of a cognitive faculty, it is unclear whether such a cognitive faculty resides in xin. There is no explicit discussion in the Shijing about the cognitive capacity of xin. There is a reference stating that what is stored in xin cannot be forgotten, suggesting that xin has the capacity of memory.23 But since what is being stored in xin in this case is a girl’s affection, it is still possible to understand this storage as concerned with the persistence of certain affective states rather than something that is remembered. Xin is also depicted as capable of formulating intentions. One song refers to the “progressive xin” of the people to mean

20 Shijing 1/3/10. 21 Shijing 1/3/19. 22 Shijing 3/1/2. 23 Shijing 2/8/4.

179 Appendix: Situating Xunzi’s Concept of Xin in Context that people want to make progress with their lives.24 In this context, it appears that the concept of xin is used to allude to people’s intentions. However, it is also possible that it simply means what people want or desire in a way that is not necessarily formulated on a conscious or cognitive level. The strongest evidence for a cognitive component of xin probably comes from the emphatic association between xin and governance in the Shijing. It is explicitly stated in “Jie Nan Shan” that the song is written with the aim to examine the xin of an incompetent ruler.25 The assumption here is that a ruler’s xin is reflected in the way one governs. For instance, King Cheng is said to be exerting his xin in his effort at enlarging the foundations laid by his predecessors.26 By contrast, the grand master Yin is criticised as not guarding against his xin for his inattentiveness to governance matters.27 This correlation between how one exercises xin and the kind of measures one adopts leads to the inference that xin is capable of thinking and forming judgments. The judgements made by rulers’ xin will determine the way they go about governing the world.28 However, it is also possible to understand this correlation as between action and the corresponding affective states. In such a case, if one has a xin that cares about people, then when exercised, the agent will take actions that show care for people. This line of thought can be traced in a song that describes the deeds of meritorious officers in terms of

29 them having enlarged their virtuous ಮ xin. Hence, the linkage between xin and its exercise of cognitive capacities is a tentative one in this case. Setting aside the question as to whether xin has a cognitive capacity, it should be clear from the above discussion that, for the authors of Shijing, xin guides action and that what xin wants affects how a person behaves. However, since we are unable to determine the locus of cognition or whether xin is at all a cognitive capacity in the Shijing, we are still unable to fully explore how the Shijing understands the interplay among affective capacity, cognitive capacity, and action. However, we can be relatively certain that the affective activities of xin do not always take full control of things. A noteworthy song in

24 Shijing 3/3/3. 25 Shijing 2/4/7. 26 Shijing 4/1/6. 27 Shijing 2/4/7. 28 Schwartz adopts this line of interpretation (1985, p.185). 29 Shijing 4/2/3.

180 Appendix: Situating Xunzi’s Concept of Xin in Context this regard refers to the firmness of xin. It is said that xin cannot be rolled about like a stone or rolled up like a mat.30 Such a description stands in stark contrast to the xin that can be overwhelmed by feelings that we saw earlier. If intense feelings could render xin as drunk or ill, and if xin could also be so firm that it cannot be rolled, then there may perhaps exist an additional component that somehow immunises xin against the sway of affective feelings. Related to this point is the notion of guarding against xin. The authors of Shijing obviously think it is important for xin to be unguarded and associate with good behaviours. As discussed before, the incompetent master Yin is said to be not alert to or guarding against his xin,31 whereas the highly respected King Wen is said to be watchful of xin.32 This idea of alertness implies the presence of a conscious dimension that oversees and manages the affective activities of xin. However, it is not clear whether xin has the cognitive capacity to govern itself or if there is something extrinsic to xin that does the guarding. Due to ambiguities in the text and insufficient information, we are unable to definitively locate this cognitive dimension in xin and conclude that the xin, as it is presented in the Shijing, also has the faculty of cognition.

33 Yijing ‹᎞ (Book of Changes)

In the Yijing, we see some themes about xin that are similar to what we have seen in the Shijing. Again, xin is treated as the seat of affective states and is thought to be capable of being moved and responding to things.34 There are descriptions of xin having sorrow35 and of xin being pleased or not pleased.36 We also find similar imageries, such

30 Shijing 1/3/1. 31 Shijing 2/4/7. 32 Shijing 3/1/2. 33 There are controversies over the dating of Yijing. Sima Qian wrote that Confucius composed the Tuan, Xi

Ci, Xiang, Shuo Gua, and Wen Yan (Shiji 47). Ou-yang Xiu disputed the claim that Confucius is the author ÙÚ

of the Yijing. Scholars of the Gu shi bian Ø movement even go so far to argue that the Yijing was mainly composed in the Han dynasty. I follow the more commonly accepted view that dates the Yijing to sometime between the late Spring and Autumn period and the Warring States period. Cf. Shaughnessy, 1993. 34 E.g. Yijing 31. 35 E.g. Yijing 48. 36 Yijing 56; Yijing TC/B.12.

181 Appendix: Situating Xunzi’s Concept of Xin in Context

37 as xin being fumigated é, which suggest xin is overwhelmed by intense feelings. There is one reference to the cleansing ê of xin in the Yijing. It is said that the sages could use the messages in the hexagrams to cleanse their xin.38 We can infer from this notion of cleansing that xin could be tainted and that xin could be subject to rectification. The need for rectification further suggests that xin is the seat of moral exertion and that there is a conception of an ideal xin. This impression is supported by the emphasis placed on maintaining an ethical xin. One hexagram describes someone who is virtuous as having the xin that wants to benefit others.39 It adds that having such a xin that wants to benefit

f others is auspicious because it will lead to the attainment of the goal of xin (zhi ೂ), where the notion of zhif supposes that xin has a focus. This further suggests that there is a connection between being ethical and xin obtaining what it wants. Similarly, another hexagram links being humble to xin obtaining something.40 Although we learn that the goals of xin could be obtained by being ethical, it is not entirely clear what it is actually obtained. For this reason, we still cannot determine whether the goals of xin are ethical goals. Another observation is that the obtainment of the goals of xin is consistently portrayed as beneficial and advantageous to the person. We can extract from this the belief that obtaining the goal of xin will yield benefits to the person. This also enables us

41 to make sense of the hexagram that interprets not having a constant ë xin as ominous.

The absence of a constant xin is probably an indication that xin is not fully oriented to its goals and is thereby preventing a person from receiving certain benefits. This means that even though there is a goal for xin, xin does not necessarily move toward its goal.

37 Yijing 52. 38 Yijing TC/A.11. 39 Yijing 42. 40 Yijing 15. 41 Yijing 42.

182 Appendix: Situating Xunzi’s Concept of Xin in Context

One possible factor that stops xin from achieving its goals is when there are barriers,

c 42 c c yi ዔ, in xin. When the “Xi Ci” refers to yi in xin, it says that the presence of yi in xin will lead to digression in words.43 It is inviting for us to understand this lack of clarity in speech as a result of a lack of clarity of thought and therefore read into xin a cognitive dimension. But since it is unclear in this case whether a lack of clarity in speech mirrors the status of xin, we cannot determine whether yi involves unclear thinking. In the Yijing, the presence of yic often entails a halt in a certain course of action. The “Xiao Xu” hexagram, for example, warns a superior person not to march forward and says that there should be yic,44 while the “Tui” hexagram explains the fortune of being calm and joyous in terms of actions are not yic.45 iHence, y c can also be understood as an obstructive element to a course of action. In this sense, having yic in xin should mean xin’s course of action is obstructed. This contrasts with the hexagram that says having an unobstructed

46 ࡙ xin will result in action that is of high value.

As is in the Shijing, the cognitive aspect of xin is hinted at in the Yijing but there are insufficient grounds for a definitive conclusion. For example, the connection between xin and discourse is implied in the claims that words will digress when xin is blocked47 and that people will say the same words when they have the same xin.48 However, it could be understood as saying that what one says is prompted by an affective state of xin.

49 b Elsewhere, there are references to wishes (yuan î) of xin and the intention (yi ï) of xin50. Since these terms are only mentioned without further elaboration, it is not entirely obvious that yuan and yib are formulated in a reflective way, for they might well be affectively driven. In one hexagram, it is suggested that the xin of Heaven and Earth can

42 It is not entirely clear whether the term “yi” in the Yijing has the connotation of doubt or uncertainty that

emerges in later texts. It is probably used interchangeably in classical terms with “ning Û ” meaning “to fixate” or “to freeze,” suggesting that when xin has uncertainty one will not be able to proceed with things. 43 Yijing TC/B.12. 44 Yijing 9. 45 Yijing 58. 46 Yijing 29. 47 Yijing TC/B.12. 48 Yijing TC/A.8. 49 Yijing 11, 61. 50 Yijing 36.

183 Appendix: Situating Xunzi’s Concept of Xin in Context be inferred from their movements.51 Although xin could be denoting the intention of Heaven, it is also possible that xin in this occurrence refers to Heaven’s affection for its people. Due to these ambiguities and little evidence, no definitive conclusion can be drawn about whether xin in the Yijing is also a cognitive faculty.

Finally, it should be noted that both the Shijing and the Yijing use “zhong xin” B಴ interchangeably with “xin.”52 Since “zhong” means “inside” or “middle,” we get a sense that zhong xin has something to do with an inner part. James Legge’s translation of zhong xin as “the centre of xin” proposes a reading of xin that has different areas of specialities. Again, it is unclear what role this connotation of the interiority plays in the concept of xin in the Shijing and Yijing. The term “zhong xin” appears again in the Warring States texts, where the term is treated as a concept and given a more elaborate account.

The Lunyu

The term xin appears only in five passages of the Lunyu. In Lunyu 2.2, Confucius claims that he followed what his xin desires (yu 4) at the age of seventy. Since it is said that xin has desires, we can infer that Confucius considers xin as capable of making responses that are guided by affective states. Lunyu 14.39 tells of a man who judges from the way Confucius plays music that “Confucius has xin,” meaning that Confucius is worried about something. This not only highlights the affective aspect of xin, it also hints at xin’s capacity to think. The cognitive and volitional capacities of xin are also touched upon in Lunyu 20.1, which says that all decisions reside in the xin of the king. It also says that good governance will result in the return of the people’s xin. Although xin in this context refers to the support and love of the people, the fact that good governance is a precondition of gaining back the people’s xin suggests that the people’s love or support for their king rest upon certain endorsements or approvals that come from xin. The cognitive aspect is highlighted when Confucius criticises those who indulge in eating

51 Yijing 24. 52 E.g. Shijing 1.6.65, 1.10.123; Yijing 11, TC/B. 12.

184 Appendix: Situating Xunzi’s Concept of Xin in Context without ever using their xin.53 It is arguable that by “not using xin,” Confucius simply means “not feeling what one is doing.” Since Confucius is criticising activities that are merely biologically driven and lack awareness, it is more likely that Confucius regards the exercise of xin as involving activities of a reflective kind. Then, “not using xin” could be an indication that a cognitive or intellectual capacity is not being exercised. Although there is insufficient information to adjudicate between the two, it becomes increasingly obvious that there is an element of consciousness or alertness to xin. From the way Confucius unambiguously values using xin over indulging in eating, we get a sense that Confucius is thinking of the importance of being mindful or of xin being involved in thoughtful acts. Still, even if the cognitive aspect of xin is implied in these passages, Confucius does not seem to have given any serious treatment to it. Hence, there is no elaborate account of the intellectual qualities of xin. In Lunyu 6.7, Confucius describes ’s xin as not departing from ren, the all- encompassing ethical ideal for Confucius. The notion of wei ð (departing from, going against) suggests a potential discrepancy between what is expected and what is in fact the case. If xin could depart from ren, it means xin is not necessarily constituted by ren. So, while the connection between xin and being ethical is obvious, the connection between xin and ethical ideal is unclear. For this reason, we cannot determine whether, in the case of the Lunyu, xin is the seat of ethical ideals or whether xin abides by an extrinsic ethical ideal.

The Laozi ᒪO

Xin is also not a prominent concept in the Laozi, appearing in only six passages. Due to the cryptic nature of the Laozi, it is not clear what the term “xin” means in these six passages. One thing that we can be relatively certain of is that, similar to the early Confucian texts, the Laozi also makes references to an affective xin that can go mad.54

53 Lunyu 17.22. 54 Laozi 12.

185 Appendix: Situating Xunzi’s Concept of Xin in Context

One passage that deserves particular attention draws a connection between xin and yu.55 In earlier discussion, we saw how xin, in other texts, is portrayed as being unsettled because it has uncertainty or has yu. The Laozi seems to have drawn a link between xin responding to the objects of yu and xin being unsettled or confused ࠿:

Not seeing what can be yu cause people’s xin to be not in disorder. Hence, the governing measures of the sages: empty [xu] their xin, fill their bellies,

weaken their zhi ೂ, and strengthen their bones. [Sages] constantly make people

have no understanding and no yu.56

It is implied in the first line that seeing what can be yu can bring disorder to xin. The inference is that xin responds to the object of yu. The passage then proposes emptying ò xin as a way of eliminating people’s yu. This strongly suggests yu is what should be emptied from xin. Since a xin without yu will not react to the objects of yu, it is consistent with the idea expressed in the first line. Laozi 55 speaks of xin as being capable of influencing qi (vital energies). Since qi is intimately linked to bodily states, it follows that xin is capable of causing bodily changes. This is an explicit discussion in pre-Warring States texts about the relationship between xin and the body, and more significantly, about how xin is capable of guiding the body.

Warring States Texts

When it comes to the Warring States period, especially after Mencius, there is a noticeable increase in more systematic and sophisticated discussions about xin’s role in guiding ethical action. Such an increase is most likely a product of the interactions between intellectual and historical forces. Intellectually, this increase in discussion can be understood as a natural continuation and development of thought. These Warring States

55 Laozi 3. 56 Laozi 3.

186 Appendix: Situating Xunzi’s Concept of Xin in Context thinkers were trying to elaborate on and refine certain concepts that their predecessors touched upon but did not fully explore. In this regard, there is no abrupt discontinuity from the earlier conceptions of xin. At the same time, we should also look for the presence of elements peculiar to the historical context, which acted as a catalyst for such intellectual development. The Warring States period was a time of great social unrest. As the name suggests, war was the defining characteristic of this period. The feudal lords’ insatiable appetite for land resulted in successive and pervasive invasions, conquests, deceits, and assassinations. We can get a glimpse of this period from the introduction of Liu Xiang’s Zhanguoce:

Merge with the big and annex the small…the end of what is ethical…greedy and without shame…armies never rest with the rise of cheating and deception.57

Being witness to these catastrophic events, Warring States thinkers were eager to find a way of proper governance that will lead to order. But before they can offer a solution, they need to first determine why humans engage in such disorderly and malevolent behaviour in the first place. Since it was assumed xin is the source of agency, their answers are often embedded in their views about what xin is and how xin works. Accordingly, their proposals for achieving order and stability are derived from their conceptions of xin. As we have already seen, this is the case for Mencius and Xunzi. Due to the scope of this thesis and the increasing depth of discussion in the Warring States, I will not be able to provide an exhaustive examination of the concept of xin in the Warring States texts. Instead, I focus on examining the Zhuangzi and the Guanzi because the concepts of xin discussed in these two texts are highly relevant to Xunzi’s thinking.

The Zhuangzi58

57 Zhanguoce, preface. 58 I am aware of the difficulties associated with dating and assigning authorship to the Zhuangzi. Fung Yu- lan and Zhang Hengshou, for example, proposed that the outer chapters were composed during early Han. However, I follow Liu Xiaogan’s view that the Zhuangzi is a late Warring States to pre-Qin text (see Liu 1994: 47-82 for detailed discussion of dating the outer and miscellaneous chapters). Although there are controversies about the authorship of the outer (Ch. 8-22) and miscellaneous chapters (23-33), many scholars consider them to have contained strands of Huang-Lao thought (see, for example, Schwartz 1985: 216 and Liu 1994: 121-134). This ambiguities concerning authorship will not substantially affect my

187 Appendix: Situating Xunzi’s Concept of Xin in Context

The Zhuangzi presents a complicated view of xin. In one instance, xin seems to be connected with the physical body in that it changes as the body changes.59 In another passage, however, xin is unhurt by one’s bodily death.60 In view of this discrepancy, let us examine the references to xin in detail. In this passage, xin changes with the body, we find reference to the cheng xin &಴:

óôõö÷øùúᄮûüᕸჽýþጳÚa॥ྞᄮ>ጳྞჽᕠa॥

ᕠរࡾ፻. ॥፻ჽᕠa॥ྞúᄮ…॥º­॥಴ᕸ 

bÞa…

[i] The hundred joints, the nine openings, the six organs, all come together and exist here [as my body]. But which part should I feel closest to? I should delight in all parts, you say? But there must be one I ought to favor more. If not, are they all of them mere servants? But if they are all servants, then how can they keep order among themselves? Or do they take turns being lord and servant? It would seem as though there must be some True Lord among them. But whether I succeed in discovering his identity or not, it neither adds to nor detracts from his Truth...His body decays, his xin follows it — can you deny that this is a great sorrow?

@ᬙ॥&಴Mü ¥ŸMa &a಴ྞ>?>࡯༢ ឃᕪ

discussion as long as I have settled the dating issue of the Zhuangzi. As my concern is not about how Xunzi responds to Zhuangzi but some of the main ideas that are already circulating around the intellectual scene in Xunzi’s time. Since I take for granted that the Zhuangzi is a late Warring States to pre-Qin text and since Xunzi is acquainted with the major movements of thought at his time, I further assume that Xunzi is aware of the main ideas posited in the outer and miscellaneous chapters of the Zhangzi. That said, the unsettled issue about the dating also forces me to examine the ideas in the Zhuangzi in an asymmetrical way. I will primarily focus on the inner chapters and, where necessary, turn to the outer and miscellaneous chapters for information that I believe can supplement our understanding of ideas found in the inner chapters. 59 Zhuangzi 2/56. 60 Zhuangzi 6/274.

188 Appendix: Situating Xunzi’s Concept of Xin in Context

[ii] If someone follows his cheng (completed) xin and makes it his master (shi), then who is without a master?

ಶvࡼ಴ᕧ੩ᒮྞᒮᕸྞᄮ

[iii] Why is it necessary for zhi (knowledge) to replace a xin that chooses by itself in order for someone to have a master? Even a fool has a master.

&a಴ྞ>?>࡯༢ ឃᕪ >ࡾŸྞჽྞŸྞჽྞᬵྞ¥

vû ¥”૆

[iv] To say that one has not yet had a cheng xin but already had a conception of yes and no is to say that one sets off for Yue today and arrived there yesterday. This is to claim that what doesn't exist exists. Even the holy Yu is not able to understand what doesn't exist exists, how can I alone understand?61

A number of observations can be made from the above passage. First, xin is regarded as a bodily organ or one of the senses. Commentators generally identify xin as one of the six viscera (in [i]). In addition, there is reference to how xin could be in a leadership role and take charge of the other bodily organs or senses. The question “how can organs or senses order themselves without a master” sounds more of a rhetorical question, yet it does suggest that there was a common belief at Zhuangzi’s time that there is a master organ that leads and orders other organs. Zhuangzi is attacking the claim that xin should be the leader. This does not preclude him from saying that xin is capable of leading. The idea that xin can lead is reinforced in another place when a fictitious Confucius disapprovingly says that Yan Hui is still making xin his master.62 I will return to this point later.

61 Zhuangzi 2/55-56; consulted Watson 1968, pp.38-9. While this seems to suggest that xin is only part of the person, there is another reference to the correlation between someone losing himself and his xin being (figuratively) dead, which suggests that the idea of the self diminishes upon the demise of xin (Zhuangzi 2/43-5). I will return to this point later. 62 Zhuangzi 4/145.

189 Appendix: Situating Xunzi’s Concept of Xin in Context

More specifically, it is the cheng (completed, formed) xin that leads ([ii]). What it means for cheng xin to lead is to follow its judgment about things. This indicates that xin is deemed capable of making judgements ([iv]). It also raises the question as to whether Zhuangzi thinks that some other aspect of xin can also lead. Furthermore, it is implied in (iii) that xin makes discriminations of its own accord and judges according to its own standard. This is consistent with the observation that the term “xin” is also used narrowly to denote “own opinion.”63 Although cheng xin is capable of leading, it is noteworthy that its leadership is compared to that of shi (teacher, master) rather than of a ruler as in the Xunzi. The emphasis of shi seems to be more on the content of opinions, that is, whether it can really point people to a path that has to be deemed reliable in order for others to follow. By analogy, the status of cheng xin could be contingent because xin would be stripped of its title as master once its guidance is no longer regarded as reliable. In other words, the judgements made by cheng xin do not necessarily dictate one’s action. Further support for this is provided by another passage that says one does not listen to xin but listens to qi, which will be examined later. Xin’s ability to calculate and plan things is related to the ability to make discriminations and judgements. One passage depicts xin as being in strife:

bv᪤᪤cv᪥᪥…༢ࡾ಴ᒮᒮ ᒮ…

Big understanding is idle; small understanding is busy…Every day they use their xin in strife, sometimes sluggish, sometimes sly, sometimes secretive.64

Such a state of being in strife is characterised in terms of sluggishness, slyness, and secretiveness. This means that xin has capacity or a person is using the capacity of xin to respond to things, formulate conceptions, and think about how it should operate. All this points to an active dimension of xin. This active dimension of xin is cast in a negative light in the Zhuangzi, as evidenced by the subsequent descriptions of the decay brought about by being fixated in a stance and the death of xin:

63 E.g. Zhuangzi 4/165. 64 Zhuangzi 2/51.

190 Appendix: Situating Xunzi’s Concept of Xin in Context

ᡴ!಴"Žದ$ ÜÝÞß%&൐(â)*ßহòᙳ&-༢.፻ࡼa/

"v॥ 0

And when their [xin] draw near to death, nothing can restore them to the light. Joy, anger, grief, delight, worry, regret, fickleness, inflexibility, modesty, wilfulness, candor, insolence — music from empty holes, mushrooms springing up in dampness, day and night replacing each other before us, and no one knows where they sprout from.65

It is worth noting that the death of xin is mentioned alongside the constant changes of emotional states. In another passage, we find that the ethical ideal de ಮ is defined in terms of xin not being moved by affective states:

ᕧ॥಴ᒮÞß‹1a/v॥”2ᗆ3ಮᕪ

When serving one’s own xin: sadness and joy do not change and move it, know that it cannot be otherwise and securely set in it as it is destiny. This is the utmost of de.66

We can infer from the above that an ideal xin is one that is not disturbed by emotional states. That ideal is not to be moved by affective states tells us that xin can be moved and that xin has the capacity to respond to emotions. In light of this observation, the passage that relates the death of xin to the changes of affective states is most likely suggesting that xin will eventually come to exhaustion if it is constantly moved by emotions.67 Although it is said that xin can be affected by emotional states, we also find descriptions in the Zhuangzi of xin being unaffected in instances where xin is described as idle68 and roaming effortlessly.69 Perhaps there is a connection between xin roaming

65 Zhuangzi 2/51; trans., Watson 1968, pp.37-38, modified. 66 Zhuangzi 4/155. 67 Zhuangzi 2/51. 68 Zhuangzi 6/259.

191 Appendix: Situating Xunzi’s Concept of Xin in Context effortlessly and flowing freely. In another instance, xin being influenced  by thoughts about ruling the world is contrasted with xin flowing freely with things and having no room for selfish thoughts.70 This suggests that an activated state has something to do with one’s being preoccupied with thoughts that have to do with the self. If I am correct on this, then Zhuangzi seems to be specifically attacking personal or private opinions.

The concept of chang (constant, unchanging) xin 4಴ encapsulates the idea that having an unaffected xin is ideal. In a hypothetical dialogue between Chang Ji and Confucius, Chang Ji asks how a man who has lost his foot still manages to be a prominent teacher.71 Confucius’s explanation is that such a person is able to use his xin to see all things as one so that his xin can roam in harmony and is unchanged by life and death. Following Confucius’s remark that such a person can reflect (jian 5) in still water,

Chang Ji then concluded that this person has obtained chang xin.72 It is clear that this chang xin that is unaffected by bodily changes is different from the cheng xin that can die. Another clue we have about this chang xin is that it is like still water. Bearing close resemblance to the idea of reflecting in still water is a passage that explicitly likens the idealised xin to a mirror:

ᕪࡥ6಴ᗆᨎ89ංøm :Uप

Utmost person uses xin like mirror: does not go after, does not welcome, respond but does not store so that he is able to win over things but not harmed.73

What is peculiar about a mirror is that it has to be still and clear in order to reflect things accurately. We can infer from this that the Zhuangzi has drawn a distinction between being responsive and being active. It is believed that xin can respond without moving or rather the ideal response is one that does not involve one’s own value judgements.

69 Zhuangzi 7/294. 70 Zhuangzi 7/293. 71 Zhuangzi 5/192. 72 Zhuangzi 5/187-193. 73 Zhuangzi 7/307.

192 Appendix: Situating Xunzi’s Concept of Xin in Context

The notion of clarity implied in the mirror imagery further means that xin should be free of impurities when it responds, where “impurities” are most likely a metaphor for emotions and personal opinions. This suggests that a less than ideal xin can hold some impurities and that these impurities can be removed. This idea of maintaining a clean xin can also be found in the passage about the fasting of xin:

ᗆ<ೂŸ„ࡾQ„ࡾ಴Ÿ„ࡾ಴„ࡾs„=DQ಴=D>

s ᒮòಎUᒮ @Aᬫòòᒮ಴C

Unify zhif (will). Do not listen with the ear but listen with xin; do not listen with xin but listen with vital energies. Listening stops at ear; xin stops at tallying/correspondence. As for vital energies, it is xu and awaits all things. Dao gathers in xu alone. Xu is the fasting of xin.74

According to the above passage, the ideal case is that in which xin does not provide guidance, as it is said that one should listen only to qi, which is defined in terms of xu. uHence, is the xideal condition xin should be in when it receives things. In the above quote, xu is defined in terms of “xin zhai.” A literal translation of the expression “xin zhai” is “fasting of xin (xin zhai).” When applied in this context, it means that xin should not be taking in certain things. It is noteworthy that the early use of “zhai (fasting)” was closely related to cleaning and the removal of whatever that considered harmful. Mencius 4B:25, for example, contrasts an unclean beautiful woman with a ugly person who zhai and bathes. The former is said to repel people whereas the latter is fit enough to offer sacrifices to a higher deity. In Mozi, zhai is again associated with bathing and clean food.75 In view of the relationship between zhai and being clean, we can infer that the fasting of xin is about making xin clean. What exactly are the things that chang xin should not be taking in? It is said that the ideal xin stops at tallying. This echoes the mirror metaphor, which accurately reflects what is in front of it. In other words, it should not go

74 Zhuangzi 4/147. 75 E.g. Mozi 12/50/2, 27/123/6.

193 Appendix: Situating Xunzi’s Concept of Xin in Context further than a mere registration of what it encounters.76 Thus, the fasting of xin is concerned with the emptying of conceptions and judgements in xin, while these personal or private judgements are guided by affective states. It is also noteworthy that the state of the fasting xin is later characterised by Yan Hui as not having himself. Similarly, another passage says that the reason someone can make his xin like dead ashes is that he has lost himself.77 This further suggests that the idea of the self is constituted by a xin that judges. As the above discussion shows, xin has the capacity to respond to emotions, reflect upon its operations, make discriminations and formulate conceptions. There seem to be two levels of xin in the Zhuangzi: the cheng xin and the chang xin. The cheng xin may also be understood as a xin that makes personal judgements based on personal feelings, whereas the chang xin is unmoved by affective states and receives Dao. It is apparent in the Zhuangzi that the chang xin enjoys a higher degree of autonomy. As opposed to the cheng xin that is helplessly moved by emotions and giving thought to things all the time, the chang xin is immune to the sway of emotions and can somehow refrain from the operations in which cheng xin engages. It should be noted that even the ideal xin functions as a mirror, meaning xin inevitably respond to things. It is unclear whether we can ascribe both cheng xin and the chang xin to the same conception of xin, or whether they denote two separate conceptions of xin in the Zhuangzi. If it is the former case, then it means xin for the Zhuangzi can refrain from its own operations and cleanse whatever preconceptions or conceptions it has off itself. Since there are seemingly two modes of xin, the connection between xin and Dao is really contingent upon the state xin is in. It is said that a xin that tries to know beyond what it is supposed to know brings harm to Dao,78 whereas a xin that is empty collects Dao.79 Another ambiguity in the Zhuangzi concerns the relationship between xin and emotions. We noted earlier that cheng xin is troubled by emotional states whereas chang xin is still and clear. It is also said that once one goes with the timing and flow of things, joy and sadness cannot enter.80 Not allowing feelings to enter is defined as “loosening of

76 Zhuangzi 4/147. 77 Zhuangzi 2/43-45. 78 Zhuangzi 6/229. 79 Zhuangzi 4/147. 80 Zhuangzi 3/12.

194 Appendix: Situating Xunzi’s Concept of Xin in Context the bonds DE.”81 Elsewhere, it is said that children’s love for their parents cannot be loosened from xin.82 This suggests that some feelings are naturally tied to xin. Nonetheless, it is also suggested that these feelings can be unbound from one’s xin. There are passages that suggest an authentic person, someone whose way of living is regarded as desirable, has feelings that are appropriate to the context:

॥಴ೂ॥FG॥HI࿝ KLM KNÜÝOPQᕸUྞR"v॥S

[Authentic person’s] xin sets direction, his face is calm, his forehead is broad. He is bitterly cold like autumn, warm like spring, his joy and anger pass through the four seasons. He goes along with what is appropriate for things and does know its limit.83

It is unclear what it is meant by joy and anger passing through the four seasons. Both Guo

Xiang TU and Cheng Xuan Ying &ህᗌ in their commentaries take it to mean that by not having xin, one is able to respond to whatever is encountered as it is. This seems to imply that inappropriate emotions are tied to cheng xin. They are harmful because they spring from judgements that do not necessarily reflect things as they are. By emptying all preconceptions and judgements, xin accurately reflects things as they are and therefore does not succumb to emotions. Hence, the above passage seems to mean that one will be attuned to things in the world by not having cheng xin. However, the Zhuangzi has not clearly explained why that would be the case. What is clear is that emotional states can cause harm.84 Nonetheless, it is unclear whether the Zhuangzi believes that emotional states are necessarily bad or that one should avoid having feelings at all. It could be saying that emotional states are harmful when they damage one’s body. In that sense, there is nothing bad about these emotional states in and of themselves if they do not damage the body.

81 Zhuangzi 6/260. 82 Zhuangzi 4/155. 83 Zhuangzi 6/230-231. 84 Zhuangzi 6/221.

195 Appendix: Situating Xunzi’s Concept of Xin in Context

The Guanzi

The Guanzi85 also articulates the differences between the bodily senses and xin. The “Xin Shu I” discusses the relationship between xin and the bodily senses in terms of the relationship between a ruler and his officials: xin occupies the ruling position and manages the bodily senses while each of the bodily senses runs its own area of responsibility. Although xin is likened to a ruler, it is explicitly said that ordering the senses is about not interfering in the activities of the bodily senses (wuwei).”86 If ordering of the sense does not depend on xin actually doing something to the senses, it is likely that the activities of xin itself must entail the status or the functioning of the senses. On one hand, xin could be interfered by the senses;87 on the other, it is also said that the senses would also be ordered once xin is ordered.88 So although xin can be confused by the senses, it is more likely to be a misjudgment on xin’s part. This also explains why an ideal person’s xin is described as not being disturbed by the input of bodily senses. It is said that if xin is placed properly, the bodily senses will also function properly; if xin has yu, then the bodily senses will not even detect their proper objects.89 Because of the superior position xin is in and its capability to provide a guiding function to the bodily senses, the Guanzi designates the matter of ordering to xin.90 More importantly, it raises

85 For the Guanzi, I will only focus on the four chapters that are commonly identified as the “Xin Shu (Art

of Mind ´ Ü)” texts (Ch. 36, 37, 38, and 49). One reason for this concerns the ambiguities involved in the authorship and dating of the Guanzi. Since it is generally agreed among scholars that the Guanzi could not be attributed to a single author and that the text could have been composed sometime between the Warring States Period and the Han, it is unlikely that the entire Guanzi text is relevant to our studies of Xunzi (see Rickett, 1985, pp.3-47). But since it is also generally agreed that the core chapters of Guanzi originated around the time of the Jixia Academy, there are reasonable grounds to speculate that Xunzi was familiar with some of the main doctrines proposed in the Guanzi. One set of these core chapters is the “Xin Shu” chapters. Although the precise dating of these four chapters is still debated among scholars, they share the basic point of view that the “Xin Shu” chapters had already come into shape in Xunzi’s time. Another

indication that Xunzi was probably aware of the ideas in the “Xin Shu” is that a lot of the technical terms

ß

Á ÝÞ employed in Xunzi’s discussion of xin, such as “xin shu ´ Ü” and “xu yi er jing ”can also be found in the “Xin Shu” chapters. There are arguments that say the “Nei Ye” chapter predates Laozi or originated around the same time as the Laozi; however, I take for granted the three dominant views that dates the “Xin Shu I,” “Xin Shu II,” “Nei Ye,” and “Bai Xin” chapters in the Warring States Period (see Wu 1986). 86 Guanzi 36/220/7. 87 Guanzi 49/270/13. 88 Guanzi 37/222/8. 89 Guanzi 36/219/1-3, 36/220/4-7. 90 Guanzi 37/223/4.

196 Appendix: Situating Xunzi’s Concept of Xin in Context the point that xin can have yu and the problems associated with xin having yu. The reason it is problematic for xin to have yu is not only that it hinders bodily senses but more importantly, it means xin is displaced from the proper status that will allow it to obtain Dao. It is clear in the Guanzi that there is a specific status xin should be in such that it can function properly, as evidenced by the numerous references to the importance of having a secure 2 91 and fixated X xin,92 which altogether convey the idea that xin should be locked in a particular position. The merit of a fixated xin is that it could provide a lodging place p for the vital essence :

X಴Bˆࡾჽp ᒮsᒮ sAY YSSYv

With a fixated xin positioned in zhong (inside, middle)…[xin] can serve as a lodging place for the vital essence (jing). The vital essence is the essence of vital energies. When vital energies are channeled, there is life; when there is life, there

b c 93 is thinking i(s ); when there is thinking, there is understanding (zhi .v)

It is said that a fixated xin should be in zhong, where the term “zhong” means “inside” or “middle.” As mentioned earlier, there was already a tendency to associate xin with zhong in the Shijing and the Yijing. However, these pre-Warring texts offer little explanation of the connection between xin and zhong. When it comes to the “Xin Shu” chapters of the Guanzi, the idea of “zhong” features even more prominently. An overwhelming majority of the references to an ideal xin would describe it as situated in “zhong.” But unlike the Shijing and the Yijing, which uses “xin” and “zhong xin” interchangeably, the Guanzi seems to associate “zhong” with a particular kind of xin only. This is suggested by the passage that says there is another xin inside xin:

91 Guanzi 37/223/3, 49/270/12. 92 E.g. Guanzi 49/270/2, 49/271/11. 93 Guanzi 49/270/2-3. Zhic (understanding, knowledge) is an important and complicated concept in the Guanzi. I follow the majority view that jing is the essence of Dao. The ideal zhic is concerned with enlightened understanding of Dao, which will allow one to discern patterns of things and therefore make appropriate responses to circumstances.

197 Appendix: Situating Xunzi’s Concept of Xin in Context

಴B੡ྞ಴ïࡾ_[ï %ºº %SS %v

Inside xin there is another xin. Awareness (yif) comes before words. After awareness come forms, and after forms comes thinking. After thinking comes understanding.94

The description of “inner xin” parallels the description of the xin that allows vital essence to take abode. This strongly suggests that the inner xin is the lodge for vital essence, thinking and understanding. Since the inner xin is the locus of thinking and understanding, the inner xin should have a cognitive capacity that allows it to operate these conscious activities. This impression is supported by passages that say having an ordered xin will lead to well-ordered country, well-ordered words and well-ordered affairs.95 Since inner xin is regarded as the locus of intellect and understanding, how inner xin is exercised will certainly impact on the decisions one makes. Hence, a majority of the discussions about maintaining an ideal xin is said to be happening “inside.” It might be inviting for us to understand this inner xin as the cognitive xin and the outer xin as the affective xin, but a close examination of how the inner xin can be disturbed suggests that the inner xin also includes affective components. It is important to note that the inner xin could be affected. One passage states:

಴ࡾø಴಴B੡ྞ಴ᄮ….ಶ࠿࠿Y!

Xin stores xin. Inside xin there is another xin…if [it is] not ordered, there must be disorder; if there is disorder, there must be death. 96

Underlying this remark is a palpable worry about xin not being ordered. An indication of xin not being ordered is when xin is not still:

94 Ó

Guanzi 37/224/1-2. Liu Ji, Wang Niansun, and Zhang Peilun emended the two characters à to but Ó even so the meaning of yif remains unclear in this context. Since the meaning of yif does not bear on my analysis here, I will follow the common translation of yif as “awareness” (e.g. Rickett, 1998 v. 2, p.63; Roth, p.72). A similar passage also appears in the “Nei Ye” chapter (270/49/14-5), which Rickett takes to be corrupt. 95 Guanzi 223/37/3-4. 96 Guanzi 49/270/14-16 .

198 Appendix: Situating Xunzi’s Concept of Xin in Context

B\಴.

When the inner is not still, xin is not ordered.97

Given that a disordered xin is one that is not still, then the task of ordering xin should largely be about stilling xin, in other words, about removing whatever that moves xin. One passage says,

]൓ßÜÝ4‡಴YI_

If [one is] able to eliminate sorrow, happiness, anger, yu, and benefit seeking, xin will revert to calmness.98

Clearly, yu is one of the things that can move xin. This suggests that inner xin is not immune to the interferences of affective xin. The following passage about the relationship between eliminating yu and stilling xin further elucidates the point that inner xin is a locus of both cognitive and affective activities:

]4W``W\Z\WW `Z WaaWZˆmnbWc

If [people] eliminate yu, then they will be spacious. Being spacious, they will be quiet. Being quiet, they will [grasp] vital essence. Grasping vital essence, then they will be alone. Being alone, they will be illuminated. Being illuminated, they will [grasp] heavenly spirit…Hence it is said that the heavenly sprit does not position in what is not clean.99

As the above passage shows, eliminating yu will result in stillness. It should be noted that being spacious (xuan ӻ) is depicted as a bridging state between eliminating yu and being

97 Guanzi 49/270/7.

98 á

Guanzi 49/269/4. The meaning of character á is clear. Haloun and Riegel emend (flawless states) to â (a state of equilibrium). See Rickett, 1998 v.2, p.49. Either way, the point here is that xin will be in a desirable state as a result of eliminating certain affective states. 99 Guanzi 36/220/12-13.

199 Appendix: Situating Xunzi’s Concept of Xin in Context still. Pertinent to the notion of spaciousness is the image of xin as a palace d.100

Elsewhere, it is said that the reason xin is called a palace is that it can provide lodging

101 place for wisdom ཅ. Then the above passage seems to be saying, in a figurative sense, that yu occupies xin and hence leaves no room for the proper exercise of cognitive capacity. At first glance, this seems to be insinuating that the affective and cognitive activities are exclusive. A caveat is the last line of the passage, which suggests that the reason shen does take abode is that xin is tainted by unclean yu. It is generally agreed among scholars that shen here refers to that which illuminates a heighted understanding of Dao. In light of this observation, the author(s) of the Guanzi seem to be saying that yu is hindering a specific way of cognitive functioning, namely, the kind that can only be achieved through the presence of shen. Then, it is not that the affective activities will preclude cognitive activities as a whole but a proper understanding of Dao. Hence, the emphasis of emptying yu is essentially on how it can lead to the entering of shen instead of how it can enable xin to exercise its cognitive capacity. It is said that:

ò॥48ep

Empty (xu) [xin’s] yu, then shen will enter the lodge.102

And xu is further described in terms of the absence of planning, pursuing, and deliberation:



òᒮŸø ˆŸøWfZŸoŸfWŸŸWIgòZ

Emptiness (xu) is no storing up…if there is no storing up, what is there to plan? If there is no pursuing (qiu) and no planning, there is no worried deliberation (lu). If there is no deliberation, then [one] will return to emptiness.103

è

å æ ç ë ä 100 ä é ê

ã Guanzi 36/221/9. 101 Ibid. 102 Guanzi 36/219/5. 103 Guanzi 36/220/15.

200 Appendix: Situating Xunzi’s Concept of Xin in Context

These two passages show that the presence of yu in xin is not just compatible with xin’s exercising of cognitive capacity, it is even implied that yu will inevitably move xin to engage in a range of cognitive activities, such as purposeful thinking. Then, the reason yu needs to be eliminated has more to do with preventing xin from obtaining a specific kind of understanding, rather than understanding in general. It is sufficient for our purpose here to simply understand it as an illuminated understanding of things. This explains why yu is considered unclean, for yu is not just an occupant but also a source of disturbance. Since it is assumed that affective states will entail certain cognitive states of inner xin and since it is believed that certain affective states need to be removed before inner xin can obtain the appropriate kind of cognitive state, the inner xin cannot simply be the locus of cognitive activities. There is no clear evidence that the reason the author(s) of the “Xin Shu” chapters introduce the concept of inner xin is for drawing a rupture between the cognitive and affective domains of xin. Rather, it seems that the concept is introduced to distinguish the physical xin from the mental xin. However, an ambiguity remains. Even if affective activities do not preclude cognitive activities, it is clear that the author(s) of the “Xin Shu” chapters only value illuminated understanding, which can only be achieved with the elimination of yu. How can yu be eliminated then? The removal of these affective states certainly cannot depend upon the faculty of understanding. Since the attainment of Dao comes after the removal of certain affective states, it is unlikely that an understanding of Dao would help xin to eliminate yu. And since the inappropriate kind of cognitive activities such as planning and deliberating are said to be resulting from yu, they are also unlikely to be what help yu to be eliminated. The closet answer in the text is something like xin cleanses itself. This is suggested by the opening passage of “Nei Ye,” which describes xin as an autonomous entity that xin by itself infuses, fills, grows, and completes.104 Granted that this is correct, it is still unclear how and why xin would decide not to want what it wants.

* * * The scope of my investigation does not allow me to analyse how Xunzi has responded to each thinker’s conception of xin but I hope I have at least provided a sense

104 Guanzi 49/269/3.

201 Appendix: Situating Xunzi’s Concept of Xin in Context of the intellectual framework within which Xunzi’s concept of xin is developed. It should be clear by now that Xunzi’s discussion of xin can be seen as an engagement with problem areas about xin that likewise concern other thinkers. (This, however, should not be taken to mean that Xunzi’s solution to the problem of xin is not unique.) There are two problem areas that are especially relevant to our discussion. First, it is generally agreed among pre-Qin thinkers that xin is the locus of both affective and cognitive activities and that xin, in its natural state, is guided by its affective states. A point of controversy occurs when they probe into questions such as whether the affective states that naturally guide xin hinder or facilitate ethical pursuit or whether these affective states stem from xin. These questions are not necessarily openly debated or consciously treated by by these early thinkers but are substantive in the sense that how these two questions are answered determines how their concepts of xin are construed, especially their understandings of the way in which the cognitive capcacity of xin is or should be exercised. Second, it is a common assumption in pre-Qin thinking that xin constitutes the person. This can imply that xin is regarded as the equivalent of person or that xin is the most important capacity of the person that speaking of xin is equivalent to speaking of the person. Although this relation is not fully sorted out by early thinkers, it is clear they share the view that xin is the source of action. For this reason, even since the pre-Warring States texts, xin is intimately linked to ethical action and assumes a guiding role in ethical action. Hence, the training or cultivation of xin has become one of the major concerns for early thinkers. Although this concern is formulated differently, the shared belief is that it is only when xin is cultivated or trained that ethical actions are possible.

202

Character list

Bi T Qi s

a Bian ᡝ Qing ച

b Cong • Qing j

De ಚ Qiu o e പ Ren ࡦ

Fen ি Shan ૽

a Gong h Shi Ž

b Hua ­ Shi 

Jie i Sia 

b Jing \ Si S

Ke  Tong O

Lu  Wei 

Luan ࠿ Wu U

a a Li ቗ Xing 

b b Li 6 Xing «

c Li ‡ Xu ò

a Lu  Yi ᑶ

b Ming a Yi ï

c Neng Yi ዔ

203

Yie <

f Yi ï

Yin X

Ying ං

Yu 4

Zhai C

a Zhi .

Zhib =

c Zhi v

d Zhi 8

e Zhi ཅ

f Zhi ೂ

g Zhi œ

Zi ஀

204

Bibliography

(Year in square brackets indicates year of first printing. References follow the year of reprint.)

Ahern, Dennis M. 1980. “An Equivocation in Confucian Philosophy.” Journal of Chinese Philosophy 7:175-185.

Ames, Roger T. 1991. “The Mencian Conception of Ren xing ࡥ: Does it Mean‘Human

Nature’?” In Henry Rosemont, ed., Chinese Texts and Philosophical Contexts: Essays Dedicated to Angus C. Graham, pp.143-175. La Salle: Open Court.

Ames, Roger T. and Henry Rosemont, Jr. 1998. The Analects of Confucius: A Philosophical Translation. New York: The Random House Publishing Group.

Behuniak, James. 2000. “ Nivison and the ‘Problem’ in Xunzi’s Ethics.” Philosophy East and West 50, no. 1: 97-110.

Bloom, Irene. 1994. “Mencian Arguments on Human Nature (Jen-hsing).” Philosophy East and West 44, no.1: 19-53. _____. 1997. “Human Nature and Biological Nature in Mencius.”Philosophy East and West 47, no.1: 21-32.

Boltz, William G. 1986. “Reviews of Books: Guanzi: Political, Economic, and Philosophical Essays from Early China. Volume I: Chapters I, 1-XI, 34 and XX, 64-XXI, 65-66 by W. Allyn Rickett.” Journal of the American Oriental Society 106, no. 4: 843-846.

Chan, Alan K.L. 2002. “A Matter of Taste: Qi (Vital Energy) and the Tending of the Heart (Xin) in Mencius 2A2.” In Alan K.L. Chan, ed., Mencius: Contexts and interpretations. Honolulu: University of Hawai’i Press.

205

Chan, Wing-tsit. 1963. A Source Book in Chinese Philosophy. Princeton: Princeton University Press.

_____. 1964. “The Evolution of the Neo-Confucian Concept Li ቗ as Principle.” Tsing Hua

Journal of Chinese Studies 4:123-149.

Chen Da-qi ᫶bl. 1953. Mengzi xingshanshuo yu xunzi xingeshuo de bijiao yanjiu mO૽

ÚnoOപÚጱqᠰst. Taibei: Zhongyang wenwu gongyingshe.

_____. 1954. Xunzi xueshuo oOÚ. Taibei: Zhongyang wenwu gongyingshe.

_____. 1987. Chenbainian xiansheng wenji vol. 1: kong meng xun xueshuo ᫶óట_ ໨ᬫw

྘. Taibei: Taiwan shangwu yinshuguan.

_____. 1987 [1964]. “Kongzi de yuwangguan yOጱ~€.” In Chenbainian xiansheng wenji

vol. 1: kong meng xun xueshuo ᫶óట_ ໨ᬫw

Taiwan shangwu yinshuguan.

_____. 1987 [1958]. “Mengzi yi li xueshuo de tantao mOᑶ‡Úጱฯ‚.” In Chenbainian

xiansheng wenji vol. 1: kong meng xun xueshuo ᫶óట_ ໨ᬫw

271-302. Taibei: Taiwan shangwu yinshuguan.

_____. 1987 [1970]. “Yantao renxing shan e wenti de jige xianjue tiaojian s‚ࡥ૽പƒ„ጱ

†_ঊˆࢉ.” In Chenbainian xiansheng wenji vol. 1: kong meng xun xueshuo ᫶óట_

໨ᬫw

_____. 1987 [1963]. “Kong meng xun sanzi suoshuo renxing de fenxi ymo࠲O Úࡥጱ

ি‹.” In Chenbainian xiansheng wenji vol. 1: kong meng xun xueshuo ᫶óట_ ໨ᬫw

206

_____. 1987 [1967]. “Mengzi xing shan shuo yu xunzi xing e shuo de buxiangdichu mO૽Ú

ᕸoOപÚጱ፻ᆗ.” In Chenbainian xiansheng wenji vol. 1: kong meng xun

xueshuo ᫶óట_ ໨ᬫw

yinshuguan.

_____. 1987 [1961]. “Meng xun xueshuo suo yi yiqu de yice moÚ ࡾ¼ផጱᕏ.” In

Chenbainian xiansheng wenji vol. 1: kong meng xun xueshuo ᫶óట_ ໨ᬫw

oÚ, pp.412-436. Taibei: Taiwan shangwu yinshuguan.

_____. 1987 [1971]. “Meng xun erzi suojian ren de techang yu qi zhongxin zhuzhang mo‘O

’ࡥጱᆜ᪖ᕸ॥B಴”•.” In Chenbainian xiansheng wenji vol. 1: kong meng xun

xueshuo ᫶óట_ ໨ᬫw

yinshuguan.

Chen, Deng-yuan ᫶–—. 1935. “Xunzi zhexue oO૝.” In Minguo congshu vol.4 ™š੯ྒྷ

wPᏕ. Shanghai: Shanghai shudian.

Chen, Gu-ying ᫶žං. 2007 [1983]. Zhuangzi jinzhu jinyi ᗳO࡯ ࡯¡. Beijing: Zhonghua

shuju.

Chen, Zheng-yang ᫶¢๕. 2005. “A Study of ‘Ming’ in Mencius and Chuang Tzu mOᕸᗳO

‘3’¤st.” Alethela 8:135-158.

Chong, Kim-chong. 2003. “Xunzi’s Systematic Critique of Mencius.” Philosophy East and West 53, no.2: 215-233. _____. 2007. Early Confucian Ethics: Concepts and Arguments. Chicago: Open Court.

207

_____. 2008. “Xunzi and the Essentialist Mode of Thinking on Human Nature.” Journal of Chinese Philosophy 35, no.1: 63-78.

Creel, Herrlee. 1931. “Confucius and Hsun-tzu.” Journal of the American Oriental Society 51, no.1: 23-32. _____. 1953. Chinese Thought: from Confucius to Mao Tse-tung. Chicago: The University of Chicago. _____. 1974 [1929]. Sinism: A Study of the Evolution of the Chinese World-view. Westport: Hyperion Press.

Cua, Antonio S. 1982. “Morality and Human Nature.” Philosophy East and West 32, no.3: 279- 294. _____. 1985. Ethical Argumentation: A Study in Hsün Tzu’s Moral . Honolulu: University of Hawaii Press. _____. 2003. “The Ethical Significance of Shame: Insights of Aristotle and Xunzi.” Philosophy East and West 53, no.2: 147-202 _____. 2005a. Human Nature, Ritual, and History: Studies in Xunzi and Chinese Philosophy. Washington, D.C.: The Catholic University of America Press. _____. 2005b [1976]. “Philosophy of Human Nature.” In Human Nature, Ritual, and History: Studies in Xunzi and Chinese Philosophy. Washington, D.C.: The Catholic University of America Press. _____. 2005c [1978]. “Dimensions of Li (Propriety).” In Human Nature, Ritual, and History: Studies in Xunzi and Chinese Philosophy. Washington, D.C.: The Catholic University of America Press. _____. 2005d [1984]. “Ethical Uses of the Past.” In Human Nature, Ritual, and History: Studies in Xunzi and Chinese Philosophy. Washington, D.C.: The Catholic University of America Press. _____. 2005f [1999]. “Xin (Mind/Heart) and Moral Failure: Notes on an Aspect of Mencius’s .” In Human Nature, Ritual, and History: Studies in Xunzi and Chinese Philosophy. Washington, D.C.: The Catholic University of America Press.

208

Du, Guoxiang ¥š¦. 1962. “Xunzi cong Song Yin Huang-Lao xuepai jieshoule shenme oO

•§¨ᯰᒪªา੪ࡃ®¯?” In Du Guoxiang wenji ¥š¦໨ᬫ. Beijing: Renmin

chubanshe.

Duan, Yu-cai °ሊ². Shuowen Jiezi Zhu Ú໨E³´. Reprinted from the original (1815).

Shanghai: Shanghai Guji Chubanshe.

Dubs, Homer H. 1930. “ ‘Nature’ in the Teaching of Confucius.” Journal of the American Society 50: 233-237. _____. 1956. “Mencius and Sun-dz on Human Nature.” Philosophy East and West 6, no.3: 213- 222. _____. 1966a [1927]. Hsuntze: The Moulder of Ancient Confucianism. Taibei: Cheng-Wen Publishing Company. _____. trans. 1966b [1928]. The Works of Hsuntze. Taibei: Cheng-Wen Publishing Company.

Duyvendak, J. J. L. 1924. “Hsun-tzu on the .” Toung pao 23: 221-254. _____. 1932. “Notes on Dubs’ Translation of Hsun-tzu.” T’oung Pao, Second Series 29, no. 1/3: 1-42.

Eno, Robert. 1990. The Confucian Creation of Heaven: Philosophy and the Defense of Ritual Mastery. New York: State University of New York Press.

Feng, You-lan. 2007 [1998]. Zhongguo zhexue shi xinbian Bš૝੹¶Ꮥ. Beijing: Renmin

chubanshe.

Fraser, Chris. 2006. “Zhuangzi, Xunzi, and the Paradoxical Nature of Education.” Journal of Chinese Philosophy 33, no.4:529-542.

209

Fu, Charles Wei-hsun. 1983. “The Mencian Theory of Mind (Hsin) and Nature (Hsing): A Modern, Philosophical Approach.” Journal of Chinese Philosophy 10, no.4: 385-410.

Fu, Si-nian ऊ¸ట. 2006 [1938]. Xing ming guxun bianzheng 3ੱ¹º». Guilin: Guangxi

shifan daxue chubanshe.

Geaney, Jane. 2002. On the Epistemology of the Senses in Early Chinese Thought. Honolulu: University of Hawai’i Press.

Goldin, Paul Rakita. 1999. Rituals of the Way: The Philosophy of Xunzi. Chicago: Open Court.

Graham, A. C. 1967. “ The Background of the Mencian Theory of Human Nature.” Ching-hua hsueh-pao (Tsing Hua Journal of Chinese Studies) 6 :215-271. _____. 1985. Reason and Spontaneity. London: Curzon Press. _____. 1989. Disputers of the Tao: Philosophical Argument in Ancient China. La Salle, Ill: Open Court. _____. 1992 [1958]. Two Chinese Philosophers: The Metaphysics of the Brothers Cheng. La Salle, Ill.: Open Court. _____. 2001 [1981]. Chuang-Tzu: The Inner Chapters. Indianapolis: Hackett Publishing Company.

Guo, Mo-ruo T¼ᗆ. 2008 [1945]. Shi pipan shu ½ප¿ྒྷ. Beijing: Zhongguo huaqiao

chubanshe.

Guo, Qing-fan T്ᛶ. 2007 [1961]. Zhuangzi jishi ᗳOᬫÂ. Beijing: Zhonghua shuju.

Hagen, Kurtis. 2007. The Philosophy of Xunzi: A Reconstruction. Chicago: Open Court.

Han, Yu. 1968. Han Changli ji ÃÄÅᬫ. Taibei: Taiwan shang wu yin shu guan.

210

Hansen, Chad. 1992. A Daoist Theory of Chinese Thought: A Philosophical Interpretation. Oxford: Oxford University Press.

Hu, Shi ᓺ . 2005 [1931]. Zhongguo zhexueshi dagang Bš૝੹bᎰ. Beijing: Tuanjie

chubanshe.

Hutton, Eric. 2000. “Does Xunzi Have a Consistent Theory of Human Nature?” In T. C. Kline III and Philip J. Ivanhoe, eds., Virtue, Nature, and Moral Agency in the Xunzi, pp. 220-236. Indianapolis: Hackett Publishing Company. _____. 2005 [2001]. “Xunzi: Introduction and Translation.” In Philip J. Ivanhoe and Bryan W. Van Norden, eds., Readings in Classical Chinese Philosophy second edition, pp. 255-309. Indianapolis: Hackett Publishing Company. Ihara, Craig K. 1991. “David Wong on Emotions in Mencius.” Philosophy East and West 41.1: 45-53.

Ivanhoe, Philip J. 1991. “A Happy Symmetry: Xunzi’s Ethical Thought.” Journal of the American Academy of Religion 59.2: 309-322. _____. 1994. “Human Nature and Moral Understanding in Xunzi.” International Philosophy Quarterly 34.2:167-175. _____. 2000a. Confucian Moral Self Cultivation. Indianapolis: Hackett Publishing Company. _____. 2000b [1994]. “Human Nature and Moral Understanding in the Xunzi.” In T. C. Kline III and Philip J. Ivanhoe, eds., Virtue, Nature, and Moral Agency in the Xunzi, pp. 237-249. Indianapolis: Hackett Publishing Company.

Ivanhoe, Philip J. and Bryan W. Van Norden, eds. 2005 [2001]. Readings in Classical Chinese Philosophy, 2nd ed. Indianapolis: Hackett Publishing Company.

Kline III, T. C. 2000. “Moral Agency and Motivation in the Xunzi.” In T.C. Kline III and Philip J. Ivanhoe, eds., Virtue, Nature, and Moral Agency in the Xunzi, pp. 135-175. Indianapolis: Hackett Publishing Company.

211

_____. 2006. “The Therapy of Desire in Early Confucianism: Xunzi.” Dao: A Journal of Comparative Philosophy 5.2:235-246.

Kline III, T. C. and Philip J. Ivanhoe, eds. 2000. Virtue, Nature, and Moral Agency in the Xunzi. Indianapolis: Hackett Publishing Company.

Knoblock, John. 1982/3. “The Chronology of Xunzi’s Works.” Early China 8: 28-52. _____. 1988, 1990, 1994. Xunzi: A Translation and Study of the Complete Works, 3 vols. Stanford: Stanford University Press.

Knoblock, John and Jeffrey Riegel. 2000. The Annals of Lu Buwei. Stanford: Stanford University Press.

Kubo, Ai ÈÉà. 1977. “Xunzi zengzhu oOÊ´. ” In Yan Ling-feng ୿ÌÍ, ed., Xunzi

jicheng oOᬫ& 43, 44. Taibei: Cheng-Wen Publishing Company.

Lai, Karyn. 2006. "Li in the Analects: Training in Moral Competence and the Question of Flexibility." Journal of Chinese Philosophy 56.1: 69-83.

Lao, Si-guang ÎSÏ. 2005. Xinbian zhongguo zhexue shi vol.1 ¶ᏕBš૝੹<ਮ. Guilin:

Guangxi shifan daxue chubanshe.

Lau, D.C. 1953. “Theories of Human Nature in Mencius and Shyuntzyy.” Bulletin of the School of Oriental and African Studies 15, no.3: 541-565. _____. trans. 1963. Lao Tzu . London: Penguin Books. _____. trans. 2002 [1979]. Confucius: The Analects. Hong Kong: The Chinese University Press. _____. trans. 2003 [1970]. Mencius. London: Penguin Books.

Lee, Jang-hee. 2005. Xunzi and Early Chinese Naturalism. New York: State University of New York Press.

212

Legge, James. trans. 1967. The Book of Poetry. New York: Paragon Book Reprint Corporation. _____. trans. 1985. The Works of Mencius. New York: Dover Publications.

Li, Di-sheng ÑÒ . 1979. Xunzi jishi oOᬫÂ. Taibei: Taiwan xuesheng shuju.

Liang, Qi-chao ÓÔ᝾. 1959. Intellectual Trends in the Ch’ing Period (Qingdai xueshu gailun)

࿾ࡼzÖ¤. trans. by Immanuel. C.Y. Hsu. Cambridge: Harvard University Press.

_____. 1960. “Lun zhongguo xueshu sixiang bianqian zhi dashi ¤BšzSശ&ᢢb਋.”

In Yin bing shi wenji ÚÛÜ໨ᬫ. Taibei: Taiwan zhonghua shuju.

_____. 1968. History of Chinese Political Thought During the Early Tsin Period. trans. by L.T., Chen. Taibei: Cheng-Wen Publishing Company.

Liang, Qi-xiong ÓÔᬩ. 1983. Xunzi jianshi oOÞÂ. Beijing: Zhonghua shuju.

Liao, Ming-chun ౝàN. 1994. “Xunzi xin tan oO¶ฯ.” In Dalu diqu boshi lunwen congkan

b᫻âãäå¤໨੯ৃ Taipei: Wenjin chubanshe.

Liu, Shu-hsien ৖è_. 1996. “Some Reflections on Mencius’ Views of Mind-heart and Human

Nature.” Kwong-loi Shun trans. Philosophy East and West 46, no.2: 143-164. _____. 2003. “Li: Principle, Pattern, Reason.” In Encyclopedia of Chinese Philosophy. ed. by Cua, Antonio S. New York: Routledge.

Liu, Xiao-gan ৖éê. 2005 [1997]. Laozi: niandai xinkao yu sixiang xinquan ᒪOëటࡼ¶ᒬ

ᕸSശ¶í. Taibei: Dongda tongshu.

213

_____. 2006. Laozi gujin: wuzhong duikan yu xiping yinlun ᒪOੱ࡯: ࡋîïਇᕸ‹ñX¤.

Beijing: Zhongguo shehui kexue chubanshe.

Loewe, Michael. 1993. “ Hsun Tzu.” In Michael, Loewe. ed., Early Chinese Text: A Bibliographical Guide. Berkeley: University of California.

Long, Yu-chun òóô. 1987. Xunzi lunji oO¤ᬫ. Taibei: Taiwan xuesheng shuju.

Lu, Martin. 1983. “Was Mencius a True Successor of Confucius?” Philosophy East and West 33, no.1: 79-86.

Lunyu jijie yishu ¤õᬫEᑶö. In Sibu yaoji zhushu congkan Pᣩøù ö੯ৃ. 1998 ed.

Beijing: Zhonghua shuju. References are to the volume, page, and line numbers.

Machle, Edward J. 1993. Nature and Heaven in the Xunzi: A Study of the Tian Lun. Albany: SUNY.

Mei, Y. P. 1951. “Hsun-tzu on Terminology.” Philosophy East and West 1, no. 2: 51-66.

Munro, Donald J. 1996. “A Villain in the Xunzi.” In Philip J. Ivanhoe, ed., Chinese Language, Thought and Culture: Nivison and His Critics, pp. 193-201. La Salle, Ill: Open Court. _____. 2001 [1969]. The Concept of Man in Early China. Michigan: Center for Chinese Studies, The University of Michigan.

Nivison, David S. 1996a. “Critique of Donald Munro, ‘A Villain in the Xunzi.’” In Philip J. Ivanhoe, ed., Chinese Language, Thought and Culture: Nivison and His Critics, pp. 320-323. La Salle, Ill: Open Court. _____. 1996b. “Critique of David B. Wong, ‘Xunzi on Moral Motivation.’” In Philip J. Ivanhoe, ed., Chinese Language, Thought and Culture: Nivison and His Critics, pp. 323-331. La Salle, Ill: Open Court.

214

_____. 1996c. The Ways of Confucianism: Investigations in Chinese Philosophy, ed., Bryan W. Van Norden. Chicago: Open Court. _____. 1999. “The Classical Philosophical Writings.” In The Cambridge History of Ancient China: from the Origins of Civilization to 221 B.C. Cambridge: Cambridge University Press. _____. 2000 [1991]. “Xunzi and Zhuangzi.” In T. C. Kline III and Philip J. Ivanhoe, eds., Virtue, Nature, and Moral Agency in the Xunzi, pp. 176-187. Indianapolis: Hackett Publishing Company.

Ogyu, Sorai ú ûࣹ. 1977. “Du Xunzi ýoO.” In Xunzi jicheng oOᬫ& 40. ed. by Yan

Ling-feng ୿ÌÍ. Taibei: Cheng-Wen Publishing Company.

Oshima, Harold H. 1983. “ A Metaphorical Analysis of the Concept of Mind in the Chuang-tzu.” In Experimental Essays on Chuang-tzu. ed. by Mair, Victor H. Honolulu: University of Hawaii Press.

Puett, Michael. 1997. “Nature and Artifice: Debates in Late Warring States China Concerning The Creation of Culture.” Harvard Journal of Asiatic Studies 57, no.2: 471-518. _____. 2004. “The Ethcis of Responding Properly: The Notion of Qing in Early Chinese Thought.” In Halvor Eifring, ed., Love and Emotions in Traditional Chinese Literature, pp. 37-68. Boston: Brill.

Qian, Mu þ. 2001 [1935]. Xian qin zhuzi xi nian _Oట. Beijing: Shangwu yinshu

guan.

Rickett, W. Allyn. 1965. Kuan-Tzu: A Repository of Early Chinese Thought. Hong Kong: Hong Kong University Press. _____. 1998. Guanzi: Political, Economic, and Philosophical Essays from Early China. (vol.1 and vol.2.) New Jersey: Princeton University Press.

215

Riegel, Jeffrey. 1980. “Reflections on an unmoved Mind: An Analysis of Mencius 2A2.” Journal of the American Academy of Religion, 47.3S: 433-457. _____. “The Zuozhuan Lecture of Physician He on Passion and Disease.” Unpublished paper.

Robins, Dan. 2001. “The Development of Xunzi’s Theory of Xing, Reconstructed on the Basis

of Textual Analysis of Xunzi 23, ‘Xing E’ പ (Xing is Bad).” Early China 26:99-158.

Roetz, Heiner. 1993. Confucian Ethics of the Axial Age: A Reconstruction under the Aspect of the Breakthrough toward Postconventional Thinking. New York: State University of New York Press.

Roth, Harold D. 1999. Original Tao: Inward Training (nei-yeh) and the Foundations of Taoist Mysticism. New York: Columbia University Press.

Sato, Masayuki. 2000. “The Development of pre-Qin Conceptual Terms and their Incorporation into Xunzi’s Thought.” In Jan A.M. De Meyer and Peter M. Engelfriet, eds., Linked Faiths: Essays on Chinese Religions and Traditional Culture in Honour of Kristofer Schipper, pp. 18-40. Boston: Brill. _____. 2003. The Confucian Quest for Order: The Origin and Formation of the Political Thought of Xunzi. Boston: Brill.

Schofer, Jonathan W. 2000 [1993]. “Virtues in Xunzi’s Thought.” In T. C. Kline III and Philip, J. Ivanhoe, eds., Virtue, Nature, and Moral Agency in the Xunzi, pp. 69-88. Indianapolis: Hackett Publishing Company.

Schwartz, Benjamin I. 1985. The World of Thought in Ancient China. Cambridge: Harvard University Press.

Shaughnessy, Edward L. 1993. “I Ching.” In Early Chinese Text: A Bibliographical Guide, ed., Michael Loewe. Berkeley: University of California.

216

Shiji ੹. Compiled by Sima Qian. Beijing: Zhonghua shuju 1959.

Shun, Kwong-loi. 1991. “Review of A. S. Cua, Ethical Argumentation: A Study in Hsun Tzu’s Moral Epistemology.” Philosophy East and West 41.1: 111-117. _____. 1997a. “Mencius on Jen-hsing” Philosophy East and West 47.1:1-20. _____. 1997b. Mencius and Early Chinese Thought. Stanford: Stanford University Press. _____. 2002. “Mencius, Xunzi, and Daizhen: A Study of the Mengzi ziyi shuzheng.” In Alan K.L. Chan, ed., Mencius: Contexts and Interpretations, pp. 216-241. Honolulu: University of Hawai’i Press. _____. 2004. “Conception of the Person in Early Confucian Thought.” In Kwong-loi Shun and David B. Wong, eds., Confucian Ethics: A Comparative Study of Self, Autonomy, and Community, pp. 183-199. Cambridge: Cambridge University Press. _____. 2006. “‘Purity’ in Confucian Thought: Zhu Xi on Xu, Jing, and Wu.” In Kim-chong Chong and Yuli Liu, eds., Conceptions of Virtue: East and West, pp. 195-211. Singapore: Marshall Cavendish Academic. _____. 2009. “Studying Confucian and Comparative Ethics: Methodological Reflections.” Journal of Chinese Philosophy 36.3:455-478.

Sivin, Nathan. 1995. “The Myth of the Naturalists.” In Medicine, Philosophy and Religion in Ancient Chinai. Great Britain: Variorum, 1995.

Slingerland, Edward. 2003. Effortless Action: Wu-wei as Conceptual Metaphor and Spiritual Ideal in Early China. Oxford: Oxford University Press.

Stalnaker, Aaron. 2003. “Aspects of Xunzi’s Engagement with Early Daoism.” Philosophy East and West 53.1:87-129. _____. 2006. Overcoming Our Evil: Human Nature and Spiritual Exercises in Xunzi and Augustine. Washington, D.C.: Georgetown University Press. _____. 2008. “The Mencius-Xunzi Debate in Early Confucian Ethics.” In Jeffrey L. Richey, ed., Teaching Confucianism, pp. 85-107. Oxford: Oxford University Press.

217

Tan, Ming-ran. 2008. “A Reevaluation of Xunzi’s Moral Theory from the Aspect of Mind.” Journal of Chinese Philosophy 35.1: 121-138.

Tang, Jun-yi . 2005 [1966]. Zhongguo zhexue yuanlun: daolun pian Bš૝¤: ¤

. Beijing: Zhongguo shehui kexue chubanshe.

_____. 2005 [1968]. Zhongguo zhexue yuanlun: yuanxing pian Bš૝¤:  . Beijing:

Zhongguo shehui kexue chubanshe.

Tang, Siu-fu c . 2008. “Xunzi zhong ‘xing’ yu ‘wei’ de duochong jiegou oOB ‘

ᕸጱx. Taiwan guoli daxue zhexue pinglun 36:1-28.

_____. “The Heart-Mind’s Approval (ke ) and Second-order Evaluation.” Unpublished paper.

Van Norden, Bryan W. 2000 [1992]. “Mengzi and Xunzi: Two Views of Human Agency.” In Virtue, Nature, and Moral Agency in the Xunzi, eds., T. C. III Kline and Philip J. Ivanhoe. Indianapolis: Hackett Publishing Company. _____. trans., 2008. Mengzi: With Selections from Traditional Commentaries. Indianapolis: Hackett Publishing Company.

Waley, Arthur. 1964 [1938]. The Analects of Confucius. London: George allen & unwin Ltd.

Wang, Xian-qian ~_. 1988. Xunzi jijie oOᬫE. 2 vols. Beijing: Zhonghua shuju.

Wang, Ling-kang ~Ìొ. 2003. “Studies on Xunzi in the English Speaking World.” The

National Cheng-chi University Philosophical Journal 11:1-37.

Watson, Burton. trans., 1968. The Complete Works of Chuang Tzu. New York: Columbia University Press.

218

_____. 2003 [1963]. Xunzi: Basic Writings. New York: Columbia University Press.

Wei, Zheng-tong ¢O. 1992 [1966]. Xunzi yu gudai zhexue oOᕸੱࡼ૝. Taibei: Taiwan

shangwu yinshuguan.

Weng, Hui-mei ᒂ². 1988. Xunzi lun ren yanjiu oO¤ࡥst. Taibei: Zhengzhong shuju.

Wilhelm, Richard. 1968. The I Ching or Book of Changes: the Richard Wilhelm Translation third edition, trans., Cary F, Baynes. London: Routledge and Kegan Paul.

Wong, David B. 1991. “Is There a Distinction between Reason and Emotion in Mencius?” Philosophy East and West 41, no. 1: 31-44. _____. 1996. “Xunzi on Moral Motivation.” In T. C. Kline III and Philip, J. Ivanhoe, eds., Virtue, Nature, and Moral Agency in the Xunzi, pp. 135-154. Indianapolis: Hackett Publishing Company.

Wu, Kuang Ï. 1985. Huang-Lao zhi xue tong lun ᯰᒪO¤. Hangzhou. Zhejiang:

Zhejiang renmin chubanshe. (A comprehensive study of the evolution of Huang-Lao tradition.)

Xia, Zhen-tao ኹ᫹. 1979. Lun xunzi de zhexue sixiang ¤oOጱ૝Sശ. Shanghai:

Shanghai renmin chubanshe.

Xiong, Gong-zhe ᅃh૝. 1977. Xunzi jinzhu jinyi oO࡯ ࡯¡. Taibei: Taiwan shangwu

yinshuguan.

_____. 1980. Xunqing xuean o . Taibei: Taiwan shangwu yinshuguan.

219

Xu, Fu-guan ದ€. 1969. Zhongguo renxinglun shi Bšࡥ¤੹:_ . Taibei: Taiwan

shangwu yinshuguan.

_____. 2004. Zhongguo sixiangshi lunji BšSശ੹¤ᬫ. Shanghai: Shanghai shudian

chubanshe.

Xu, Yuan-gao —!. 2002. Guoyu jijie šõᬫE. Beijing: Zhonghua shuju.

Xunzi xinzhu oO¶ . Beijing daxue zhexuexi "#b૝$. 1983. Taibei: Liren shuju.

Xunzi yinde oOXಚ [A Concordance to Hsun Tzu]. 1950. Harvard-Yenching Institute

Sinological Index Series, supp. 22. Beijing: Yenching University Press.

Yang, Bo-jun %&'. 1960. Mengzi yizhu mO¡ . Beijing: Zhonghua shuju.

_____. 1980. Lunyu yizhu ¤õ¡ . Beijing: Zhonghua shuju.

_____. 2006 [1981]. Chunqiu zuozhuan zhu NLఃद . Beijing: Zhonghua shuju.

Yang, Da-ying %bᕉ 1935. “Mengzi xueshuo yanjiu mOÚst.” In Minguo congshu vol.4

™š੯ྒྷ 010 wPᏕ. Shanghai: Shanghai shudian.

_____. 1936. Xunzi xueshuo yanjiu oOÚst. Shanghai: Zhonghua shuju.

Yang, Liu-qiao %+,. 1985. Xunzi guyi oO-¡. Jinan: Qilu shushe.

Yang, Yun-ru %. 1935. “Xunzi yanjiu oOst.” In Minguo congshu vol.4 ™š੯ྒྷwP

Ꮥ. Shanghai: Shanghai shudian.

220

Ye, Shao-jun /01. 1930. Xunzi oO. Taibei: Shangwu yinshu guan.

Yearley, Lee. 1980. “Hsun Tzu on the Mind: His Attempted Synthesis of Confucianism and Taoism.” The Journal of Asian Studies 39.3: 465-480.

Yinwenzi. All references are to the volume, page, and line numbers in the Zhuzi Jicehng edition.

Zhang, Jue •2. 1995. Xunzi yizhu oO¡ . Shanghai: Shanghai guji chubanshe.

Zhang, Li-wen •`໨. 1991. Li ቗. Beijing: Zhongguo renmin daxue chubanshe.

Zhang, Shi-tong 3æ‰. 1974. Xunzi jianzhu oOÞ . Shanghai: Shanghai renmin chubanshe.

Zhu, Xi 4ᅠ. 1968. Jin si lu ji zhu ᡴSᧁᬫ´. In Guo xue jib en cong shu š{€੯ྒྷ.

Taibei: Taiwan shang wu yin shu guan.

Zhuzi Jicheng Oᬫ&. 1954. Beijing: Zhonghua shuju.

221