Dai Zhen and David Hume

戴震與休木

A notable feature of pre-modern culture is that a lot of creative writing is presented in the form of commentary on certain specified texts. The following paper attempts to show how this feature can account for significant differences between Chinese and Western philosophy. To illustrate the argument a comparison is made of two roughly contemporary works, the On Original Goodness by Dai Zhen and the Enquiry concerning the Principles of Morals by David Hume. Both works were written to give a metaphysical basis to an empirical morality. Both were revised after their initial composition. In the case of Dai Zhen’s work this resulted in the addition of a considerable number of quotations from the classics. Hume also quotes classical sources but whereas Dai Zhen understands his purpose as making manifest the genuine meaning of past authors, Hume believes that constant recourse to ancient authors is superfluous for establishing moral truth. It is with respect to this dividing line between modern and pre-modern culture, as defined by J.B. Henderson in Scripture, Canon and Commentary, that the two authors differ. Nonetheless closer inspection reveals that elements of pre-modernity lie in Hume’s work, whilst, conversely, elements of modernity are not lacking in Dai Zhen.

徵之古訓 zheng zhi guxun (Yuan Shan 3, p 16)

The distinctiveness of Chinese philosophy and culture means that the work of comparative culture is generally undertaken as a dialogue between China and the West tout court. The present paper attempts to show that this is not the only factor to be taken into account. At least in the field of philosophy the division between modern and pre-modern philosophy is as, if not more, important. Crudely we may say that a pre-modern philosopher believed some text or set of texts contains the essential truth about the world. The task of the sage is to be a good exegete. The post-Cartesian philosopher, however, believes that truth can only be established by a direct appeal to reason. Ancient texts serve only to confirm results obtained by reason and based on principles that are supposedly universally valid.

1 Introduction: Basic philosophy of Dai Zhen and David Hume

Dai Zhen was born in Xiu Ning , Anhui安徽. He studied phonology. In 1773 he was appointed editor of the Siku Quanshu四庫全書. His essay On Original Goodness (Yuan Shan原善) was written in 1765, when he was forty. The first version comprised three short essays with no textual references to the Confucian Classics. Between 1768 and 1774 these three works were expanded by incorporating the Discussion of Human Nature after Reading the Great Commentary to the Changes (Du Yi Xici Lun Xing讀易繫辭論性) and the Discussion of Human Nature after Reading the Mencius (Du Mengzi Lun Xing讀孟子論性). He also wrote a preface showing how he intends to expose the true teachings of the Classics (Jing zhi daxun 經之大訓).

His major work Mengzi Ziyi Shuzheng 孟子字義疏證was composed under the title Prefatory Words (Xu Yan 序言) in 1768 and renamed in 1774. It was not published until after 1777. According to Professor Cheng Chung-ying, Dai Zhen had two objectives in mind: to criticize the Neo-Confucian notion of li 理; and to reconstruct the classical Confucian notion of shan 善.

Dai Zhen was initially taught according to the tradition of Zhu Xi 朱熹 and the Song Ru儒 but, according to Hu Shi胡適, he came under the influence of Yan Yuan顏元 and Li Gong李塨, thanks to Cheng Tingzuo程廷祚 (1691-1767).

Following Li Gong and Cheng Tingzuo, Dai Zhen criticises the Song Ru for locating the principle, li理, of virtue and human nature outside feelings and desires, which they denounce. Hence, in Dai's view, the Song Ru lack a basis for their ethical opinions.

Dai Zhen contended that the Song misreading was owing to Buddhist influence. Neo-Confucianists sought a state free from all desires and considered nature as bad. Dai Zhen wished to return to Mencius孟子 and affirm the goodness of human nature with all its natural inclinations and capacities. The Neo-Confucian separation of man into a good part that has the principles of righteousness and reason, yi-li義理, and a bad part that has natural endowment, qi-zhi氣質, is more like the theory of Xunzi荀子, than that of Mencius. Dai Zhen sought to restore the original meaning of Confucian terms as found in the Analects論語, Mencius, Li Ji禮記 and Zhou Yi周易.

Hume's underlying philosophical programme can only be understood in the light of his first Enquiry on Human Understanding. The chief idea that Hume wished to express in the first Enquiry is that our use of the term 'cause' does not mean that there is a metaphysical entity that ensures necessity but rather that experience leads us to form a habit whereby we universalise from known experience to affirm rules valid outwith our past experience. Where such a habit is engrained there we may say that x causes y.

In moral actions Hume again stresses the social context in which human behaviour is found. Habit teaches us which actions benefit society as a whole and thus enables us to form moral norms. It is thus habit which leads us to infer causality, the basis for all scientific reasoning in Hume’s view, just as habit it is which impells us to draw up moral norms. In both cases experience is fundamental for without it we should not know what human social life is like and thus could not form habits based on observation.

Dai Zhen’s philosophy is notable for stating the inseparability of li, principle, and qi, matter/stuff. In contrast to the Cheng Brothers and Zhu Xi and in unison with Li Gong, Dai asserts that li is within qi; li is contemporaneous with qi rather than prior to it. Li is universally present and is simply that in things which must be so and cannot be changed.

Ju fan tiandi, renwu, shiwei, qiu qi biran bu ke yi, li zhi mingxian ye. 舉凡天地、人物、事為、求其必然不可易，理至明顯也。（孟子字義疏證卷上理）
Song-Ming Li Xue founded its distinction between the physical and metaphysical on a phrase in the Book of Changes, Dai Zhen argues that the interpretation of the relevant phrase is faulty both in thought and in terms of grammar. Thus where Zhu Xi divides the universe into Dao道 above and Vessel (Qi器) below, Dai Zhen argues that the division applies to things as such, below, and the movement of qi氣, above: Qi hua zhi yu pinwu, ze xing er shang-xia zhi fen ye.
氣化之於品物，則形而上下之分也。（孟子字義疏證卷中天道）
The original text of the Great Appendix reads as follows:
What is above form wei zhi Dao; what is below form wei zhi Qi (Vessel).
形而上者謂之道；形而下者謂之器

Where the two seem to differ the most is in their methodological approach and terminology. Hume writes in a discursive essay style in which he aims to present his own opinions. He draws on historical examples as quoted in the classics but does not become engaged in exposition of the classics. Dai Zhen employs a terminology that is determined by the whole weight of Chinese tradition. He believes that he is expounding the real meaning of the classics, the Mencius in particular, in contrast to those scholars of the Song-Ming period, whose works are marred by false dualist ideas that come from Buddhism and Daoism.

4. Moreover, a long period of time separates us from the ancient sages, therefore people accept what they hear without discriminating true and false
以今之去古聖哲既遠，治經之士，莫能綜貫，習所見聞，積非成是。
Turning to the text proper we can note in the first place the use of explicit quotations from the classics. The most quoted text is the Li Ji, then the Mencius, the Changes and its Appendices, the Shang Shu尚書, Analects, Shi Jing詩經 and one quotation each from the Zuo Zhuan左傳 and Da Dai Li Ji大戴禮記. This listing may be rather misleading in places. The chapters of the present Li Ji were not originally all in one volume and by the time of Dai Zhen, the Zhong Yong中庸 and Da Xue大學 had an independent existence. Similarly scholars were always conscious of a distinction to be drawn between the text of the Changes and its Appendices, traditionally ascribed to Confucius. If this is taken into account then the ranking of quotations is Zhong Yong, Mencius, Great Appendix to the Changes繫辭大傳. It is immediately apparent that the list of texts, originally not large, is further restricted by Dai Zhen, who virtually creates his own canon within the canon. Moreover, the choice of quotations is not spread evenly throughout the three chapters. Concentrating only on texts cited three times or more, we find that the first chapter cites the Great Appendix five times and the Zhong Yong thrice. The second chapter quotes the Mencius eight times, quoting only chapters 6A, 7A and 7B of the classic. Whilst the third chapter cites the Zhong Yong eleven times, Analects and Shang Shu five times each, Shi Jing four times, Mencius (1A, 3A) and Changes three times each. In other words the second chapter concentrates on the Mencius, the first refers to both the Zhong Yong and the Great Appendix, whilst the third chapter draws more frequently on a broader range of materials.

Looking at the role of the quotations in more detail we may note that in Chapter One quotations are used to break up what would be a highly dense text in which key terms are defined in a systematic way. The quotations are all used to show links between the metaphysical and natural orders and so show that morality is conformity to both, which are in fact one. In places this means heavy glossing of the original. Thus the last quotation which describes the way of Heaven (Qian) as "easy" and the way of Earth (Kun) as simple, enables Dai Zhen to argue that the author was thinking of the principle of benevolence (ren仁) with respect to heaven and of order and reason (tiao-li條理) and wisdom (zhi智) with respect to earth. In places it leads to using one text to illuminate another (Da Dai Li Ji used to interpret the Zhong Yong) but normally it relies on expounding a few hints in each passage. The chosen quotations are well-known and not obscure.

In Chapter Two the main argument is a justification of the Mencian position on human nature. Other quotations simply provide some basic terminology: hun-po魂魄, shen-ling神靈, and a list of moral virtues.

Yu Yingshi余英時makes the point that Dai's addition of the quotations was motivated by a desire to justify his exegetical technique in the face of an intellectual climate that valued exegesis. He wanted to prove that his reasonings were not just empty thoughts but were rooted in texts. Yu makes clear that Dai Zhen's main protagonists were his fellow exegetes whose interest failed to extend to the philosophical meaning of the texts they were reading. Using Isaiah Berlin's distinction of the fox and the hedgehog, Yu says that while Dai was engaged in the scavenging task of the fox, he was really a hedgehog, bent on discovering the underlying philosophical enterprise.

We may classify Hume's use of quotations into three categories: Neutral quotations, Critical quotations and Laudatory quotations.

Apart from these quotations, however, one must also note the choice of terminology. Chinese thinkers rarely compose new terms but draw on the stock of existing ones, imbuing them with a new meaning. Dai Zhen singles out shan [善 goodness], ren [仁 benevolence], li [禮 propriety] and yi [義 righteousness].

Whatever the original status of the first Two Books, the Da Xue and the Zhong Yong., post-Han readers knew them as chapters of the Li Ji. The term shan occurs 96 times. Of these occurrences eleven are found in the Zhong Yong and ten in the Da Xue, a very high ratio compared to the rest of the Li Ji. Only the Xue Ji pian學記篇 with 13 instances has more uses of the term. In the Analects the same term is found 43 times, notably in the seventh and thirteenth pian. The Mencius is considerably longer than the Analects and so we are not surprised to find shan used 112 times, with five or less occurrences in pian 1A, 1B, 2B, 3B, 5A and with ten or more instances in pian 2A, 6A, 7A and 7B. There is no perfect match between these latter four pian and those Dai Zhen quotes but there is a general similarity. Dai Zhen quotes most frequently from the Zhong Yong and the later pian of the Mencius. It is precisely in these texts that the notion of goodness is expounded. Thus we may conclude that the texts he quotes are indeed the ones that provided his special vocabulary. In other words, even in the original draft of the manuscript, which is said not to have contained quotations, the influence of the texts that were later cited would have been present.

The next characteristic feature of Dai's argument is his equating goodness with propriety, benevolence and righteousness. As well as this tripartite listing Dai integrates the more familiar four-part listing of the Mencius: benevolence, righteousness, propriety and wisdom. As in the Mencius all moral principles are generated from benevolence and wisdom. Moreover, it is wisdom which enables one to understand righteousness and principle and not exhibit one's subjectivity. In other words complete virtue can be defined in terms of the three virtues with zhi [智 wisdom] playing a role in the prevention of selfishness. Thus it is that Dai Zhen can be faithful to Mencius whilst operating under a different paradigm.

Dai Zhen's reading of benevolence in terms of the productive force of heaven as portrayed in the Great Appendix is dependent on a Song tradition, found in Zhu Xi for instance. Where Dai Zhen differs is in his reduction of principle from the main heavenly principle to a simple matter of reason tiao-li條理. Thus his contribution lies not in new vocabulary but in a rearrangement of known vocabulary. This is one aspect which defines his work as pre-modern.

The choice of selfishness [si私] and benightedness [bi蔽 ] as the two causes of all wrong means that the sensible man will do no wrong. Morality is an intellectual issue. That selfishness was seen as a fault is nothing new in Chinese philosophy. Cheng Yi程頤 defined selfishness as doing something for one's own personal wishes (Reflection on Things at Hand, p 275). However, more normally in the Song period faults are said to be due to desires (Reflection on Things at Hand, p 274) or neglect (Reflection on Things at Hand, p 277). Benightedness is perhaps not singled out with the same insistence before Dai Zhen. Doubtless he needs the term to explain how a person can lack wisdom and hence fall into error.

In comparing Dai Zhen and his contemporaries, notably Zhang Xuechang章學誠, Yu Yingshi points out that Dai does not throw commentary out of the window. Whereas Zhang Dai "attempts to escape from the tyranny of the commentarial tradition by advocating a direct relationship with the text of the canon itself" (Campbell,D., The Body of the Way is without edges: Zhang Dai (1597-?1684) and his Four Books Epiphanies, Conference Paper CSAA: Adelaide 1997)and David Hume seeks to avoid all appeal to authority except that of Cicero, Dai Zhen wishes to maintain the integrity of both text and tradition. His return to the genuine meaning of Mencius is evidenced by his use of that classic, but his keeping to tradition is also shown by his acceptance that reading Mencius can be done in the light of the metaphysics that developed from the Great Appendix. Where he castigates Song metaphysics is in its reading of the text of the Great Appendix resulting in a dualism of Principle and Matter. In other words Song metaphysicians are not guilty of constructing a metaphysical system as such, but for poor exegesis.

David Hume spent much of his later life writing history books whilst Dai Zhen earned his living from editing texts. These two activities are in accordance with their underlying philosophical direction. Hume wanted to look at human behaviour because it was precisely there that habits were formed and hence norms could be established. Dai Zhen believed that the norms of human conduct could be found first of all in ancient texts. This distinction should not be seen as a tale of two cities, Beijing and Edinburgh, but rather as a tale of two epochs: pre-modern and modern.

Post-Cartesian thought has conditioned us to seeking truth on grounds established by reason on its own rights. Even when contemporary thought challenges the limits of the rational enterprise it still remains sure of its own ability to affirm its authority by itself. Hume is no exception to this goal. Yet even Hume admits a fondness for Cicero and thus demonstrates that the modern philosopher can never be wholly free of his/her pre-modern setting. Dai Zhen speaks from a pre-modern setting and claims only to be presenting the real meaning of an ancient text. Yet in so doing he has constructed a cognitive theory of morality that is systematic to a degree Mencius himself would doubtless never have imagined. The pre-modern philosopher is thus able to create a new system of thought even though it comes with the label of an ancient classic.

Appendix 1: Quotations in Yuan Shan

(page references to Cheng Chung-ying’s edition)

Li Ji禮記 (20 quotations)

Mencius孟子(13 quotations)

Zhou Yi周易 (9 quotations)

Shang Shu尚書

Shi jing詩經

Analects論語

Da Dai Li Ji大戴禮記

Zuo Zhuan左傳

[1] Ch 2: p 85

[3] Mencius 1A, 3A

Appendix 3: Summary of argument, showing use of quotations

A: The Argument

• A1 Tripartite structure of Goodness

The rest of the chapter uses quotations to fill in the argument.

1. That man's nature consists of the desires of the eyes, ears. This doctrine leads to a counter-emphasis on the spiritual at the expense of natural desires.
2. That man's nature consists of perception, jue覺. Again this leads to a counter-emphasis on spiritual aspects and a view of human nature lacking in content.
3. That man's nature is to have reason while desires are credited with leading to evil (Xunzi).