TWO DILEMMAS IN VIRTUE ETHICS AND HOW ZHU XI’S NEO-CONFUCIANISM AVOIDS THEM

JOURNAL OF PHILOSOPHICAL RESEARCHVOLUME 36, 2011

I

TWO DILEMMAS IN VIRTUE ETHICS AND HOW ZHU XI’S NEO-CONFUCIANISM AVOIDS THEM

YONG HUANGKUTZTOWN UNIVERSITY OF PENNSYLVANIA

ABSTRACT: Virtue ethics has become an important rival to deontology and consequentialism, the two dominant moral theories in modern Western philosophy. What unites various forms of virtue ethics and distinguishes virtue ethics from its rivals is its emphasis on the primacy of virtue. In this article, I start with an explanation of the primacy of virtue in virtue ethics and two dilemmas, detected by Gary Watson, that virtue ethics faces: (1) virtue ethics may maintain the primacy of virtue and thus leave virtue non-explanatory, or it may attempt to explain virtue in terms of something else and thus render virtue second-ary at most; (2) the explanation of virtue may be objective and thus become morally indeterminate, or it may be normative and thus lack objectivity, merely re-expressing the virtue it intends to explain (Section II). After showing the failure of both clas-sical Aristotelian and contemporary neo-Aristotelian virtue ethics to escape these dilemmas, I turn to the ethical theory of Zhu Xi 朱熹 (1130–1200)—the greatest synthesizer of neo-Confucianism, whose place in Confucianism is comparable to that of Thomas Aquinas in the Christian tradition—to show how it can successfully avoid both dilemmas.

I. INTRODUCTION

n the last couple of decades, there has been an impressive revival of virtue ethics as an alternative to deontology and consequentialism dominating modern and contemporary moral discourse. Virtue ethics, as stated by Michael Slote, one of its most influential advocates, is the “new kid on the block” (Slote 1997, 233). Rosalind Hursthouse, another of its most influential advocates, makes a much more

248 YONG HUANG

optimistic claim just two years later: “it has now acquired full status, recognized as a rival to deontological and utilitarian approaches, as interestingly and challengingly different from either as they are from each other” (Hursthouse 1999, 2). Virtue ethics comes in a variety of forms. While most contemporary virtue ethicists are neo-Aristotelians, there are others who get their primary inspirations from the Stoics, David Hume, Nietzsche, and John Dewey. What is unique about virtue ethics and common to its various forms is its focus on the character of the agent, in contrast to deontology and consequentialism, which focus on the nature of actions, despite the obvious connections between agents and their actions. So while virtue ethics does not always disagree with deontology and consequentialism in moral evalu-ations, even when it issues the same moral approval or disapproval, its judgment is based on something different from those for deontology and consequentialism. Rosalind Hursthouse illustrates this difference well. In her view, the three types of ethics may well agree that I should help someone in need, but they disagree why I should do so: “A utilitarian will emphasize the fact that the consequences of doing so will maximize well-being, a deontologist will emphasize the fact that, in doing so, I will be acting in accordance with a moral rule such as ‘Do unto others as you would be done by’, and a virtue ethicist will emphasize the fact that helping the person would be charitable or benevolent” (Hursthouse 1999, 1). Thus we can say that virtue is primary in virtue ethics. In this article, I start with an explanation of this primacy of virtue in virtue ethics and two dilemmas, detected by Gary Wat-son, that virtue ethics faces: (1) virtue ethics may maintain the primacy of virtue and thus leave virtue non-explanatory, or it may attempt to explain virtue in terms of something else and thus render virtue secondary at most; (2) the explanation of virtue may be objective and thus become morally indeterminate, or it may be normative and thus lack objectivity, merely re-expressing the virtue it intends to explain (section II). After showing the failure of both classical Aristotelian and contemporary neo-Aristotelian virtue ethics to escape these dilemmas (sections III and IV), I turn to the ethical theory of Zhu Xi 朱熹 (1130–1200)—the greatest synthesizer of neo-Confucianism,1 whose place in Confucianism is comparable to that of Thomas Aquinas in the Christian tradition—to show how it can success-fully avoid both dilemmas (Sections V and VI). I close with a brief concluding summary (Section VII).

II. PRIMACY OF VIRTUE IN VIRTUE ETHICS AND ITS TWO RESULTANT DILEMMAS

To say virtue is primary in virtue ethics, of course, means that the virtuous character, in comparison with rightness of action, is primary. This is what Slote means when he claims that “a virtue ethics in the fullest sense must treat aretaic notions (like ‘good’ or ‘excellent’) rather than deontic notions (like ‘morally wrong,’ ‘ought,’ ‘right,’ and ‘obligation’) as primary, and it must put a greater emphasis on the ethical assessment of agents and their inner motives and character traits than it puts on the evaluation of acts and choices” (Slote 1992, 88; emphasis added).

DILEMMAS IN VIRTUE ETHICS AND ZHU XI’S NEO-CONFUCIANISM 249

However, to say that virtue is primary also means two additional things. On the one hand, virtue ethics does not disregard the importance of rightness of action entirely; it only regards it as secondary in comparison to virtue or admirability of character. Hursthouse thus claims that the common characterization of virtue ethics as agent-centered rather than action-centered is somewhat misleading, as it invites the criticism that virtue ethics fails to provide the very much needed action guide. In her view, virtue ethics does provide an action guide, although one differ-ent from those provided by deontology and consequentialism. If consequentialism says that “an action is right iff it promotes the best consequences,” and deontol-ogy says that “an action is right iff it is in accordance with a correct moral rule or principle,” then virtue ethics says that “an action is right iff it is what a virtuous agent would characteristically (i.e., acting in character) do in the circumstance” (Hursthouse 1999, 27–28).

On the other hand, to say that virtue is primary in virtue ethics does not exclude the possibility that other types of ethics can reserve an important role for virtues. It is in this regard that Rosalind Hursthouse deems it important to make a distinction between virtue ethics and virtue theory (Hursthouse 2007, section 1). For example, consequentialism, particularly the so-called motive or character utilitarianism, can allow virtue a prominent place. According to Robert Merrihew Adams, the most important advocate of motive utilitarianism, motives are “principally wants and desires, considered as giving rise or tending to give rise to actions” (Adams 1976, 467). Although he sees the distinction between a desire and a trait of character, he does think that “a desire, if strong, stable, and for a fairly general object . . . may perhaps constitute a trait of character” (Adams 1976, 467). According to his motive utilitarianism, “the morally perfect person . . . would have the most useful desires, and have them in exactly the most useful strengths; he or she would have the most useful among the patterns of motiva-tion that are causally possible for human beings” (Adams 1976, 470). Since what the morally perfect person in motive utilitarianism has is not merely desires or motivations, but patterns of such desires and motivations, such a person can be regarded as a virtuous person. However, motive or character utilitarianism is still different from virtue ethics, because the virtuous character is not intrinsically good but only instrumentally good: it is good only because it is useful to produce the desired consequence.2

Similarly, deontology, particularly the Kantian version, can also have a promi-nent room for virtue. As a matter of fact, some contemporary Kantian scholars even claim, or at least once claimed, that Kant himself is a virtue ethicist. For example, Onora O’Neill, in an article originally published in 1984, focuses on Kant’s idea of maxims, which, according to her, “are underlying principles that make sense of an agent’s varied specific intentions,” and so “can have little to do with the rightness or wrongness of acts of specific types, and much more to do with the underlying moral quality of a life, or aspects of a life . . . . To have maxims of a morally ap-propriate sort would then be a matter of leading a certain sort of life, or being a

250 YONG HUANG

certain sort of person” (O’Neill 1989, 152). While duties indeed occupy a central place in Kant’s ethics, O’Neill argues that they are “duties to act out of certain maxims, that is, to structure our moral lives along certain fundamental lines, or to have certain virtues” (O’Neill 1989, 153). However, we have to keep in mind that maxims in Kant’s ethics are not fundamental. Sometimes one has to refrain from acting on certain maxims, and sometimes one has to act on certain maxims that one is reluctant to accept. It is the moral principle that tells us on which maxims we should act.3 So in a “Postscript” to this article written fifteen years later, O’Neill herself realizes that her earlier characterization of Kant as offering an ethics of virtue is misleading, acknowledging that “Kant’s fundamental notion is that of the morally worthy principle that provides guidelines not only for matters of outward right and obligation, but for good characters and institutions as well” (O’Neill 1989, 162). In short, virtuous characters are good for Kant only within the guidelines of the moral principle.4

Thus, in both deontology and consequentialism, virtues can have very important instrumental values of promoting good consequences or acting on moral principles, while in virtue ethics, because of the primacy of virtue, virtue is intrinsically good. Although there are radical or extreme virtue ethicists, the so-called replacement virtue ethicists (see Anscombe 1958), who regard the notion of the rightness or wrongness of acts as unintelligible and want to replace it with the idea of virtue or vice of agents, most virtue ethicists writing today are reductionists (see State-man 1997, 8). They do not necessarily deny the appropriateness of the notions of the rightness or wrongness of actions that deontologists and consequentialists are concerned with, but they claim that such notions can more or less be reduced to, derived from, or considered as secondary to judgments about the moral character of agents.

So, with regard to virtue, the distinction between virtue ethics, on the one hand, and deontology and consequentialism, on the other, is that virtue is intrinsically good for the former, while it is merely instrumentally good for the latter. In this article, I shall focus on the virtue ethicists’ understanding of virtue. Since virtue is primary in virtue ethics, a natural question to ask is what constitutes a virtue or what makes a character trait virtuous. According to Gary Watson, it is in attempts to answer this question that virtue ethicists encounter the first dilemma: “any eth-ics of virtue that lacks a theory of virtue will be nonexplanatory but any ethics of virtue that has such a theory will collapse into an ethics of outcome” (Watson 1997, 62). What he means is that, on the one hand, if virtue ethicists really want to insist that virtue is primary, they cannot explain it in terms of something else. In this case, however, we will be at a loss about what a virtue is and how it is to be distinguished from its opposite, vice. On the other hand, if virtue ethicists want to provide an account of virtue, they have to explain it in terms of something else. However, as soon as they do that, this “something else,” instead of virtue, becomes primary, and the ethics with such a conception of virtue is no longer a virtue ethics but merely has a theory of virtue.


Representing one important school of contemporary virtue ethics, Michael Slote seems to hold the first horn of the dilemma, as he thinks that there is no such need for an account of virtue. Slote does claim that “traits of character can qualify as virtues through what they enable possessors to do for themselves as well as through what they enable their possessors to do for others, and so we saw . . . that our ordinary employment of the aretaic notion of the virtue gives fundamental evaluative significance to the well-being of self . . . and to the well-being of oth-ers” (Slote 1992, 91). However, he does not mean to claim that everything that enhances the well-being of self and/or of others is virtuous or that enhancement of the well-being of self/other is the criterion to judge whether a character trait is a virtue, although he does think that all virtues do enhance the well-being of self and/or others.5 If so, how do we know whether a particular character trait is virtu-ous or vicious? Slote’s answer is simple: you already know it.6 What he appeals to is our common sense and intuition. For this reason, he prefers to use the term admirable and its opposite, deplorable, instead of good and evil or virtue and vice, for the simple reason that the former does not need an explanation as the latter does: “to say ‘I found his actions (or what he did) good’ is to invite a request for further clarification in a way that use of ‘admirable’ seems not to do” (Slote 1992, 94–95). It is in this sense that Slote regards his virtue ethics as a common-sense virtue ethics: “the virtue ethics I will be describing is largely based on our common-sense ideas and intuitions about what counts as a virtue and, more generally, as admirable. And in order to evoke, or at least not discourage, such intuitions, I think we should try to avoid stretching the usage of terms and, whenever possible, use idiomatic or natural-sounding language in our discussion of foundational issues” (Slote 1992, 94).7 It is interesting to note that Slote appeals to our common sense to determine whether a given person, an action, or a character trait is admirable at the same time when he claims, in his argument against anti-theory in ethics, that “our ordinary intuitive moral thought is not just complex, but subject to paradox and internal incoherence” (Slote 2001, 11; see also Slote 1997, 181).8 Slote does say that his own virtue ethics “gives credence to intuition” only “when and where it doesn’t lead to incoherence and paradox” (Slote 1992, 210). The question is how such incoherence and paradox can be avoided without an account of what constitutes a virtuous or (using his more preferred term) admirable action, person, or character trait.9

In any case, we will not be able find an account of what constitutes a virtue in Slote as he refuses to provide one. If one does not already know whether a given person, action, or character trait is virtuous, then Slote cannot help. This is precisely what Watson means by the first horn of his first dilemma: an ethics of virtue that lacks an account of virtue is non-explanatory. Most other contem-porary virtue ethicists, however, do attempt to provide an account of virtue, and since they are mostly Aristotelians, the account they provide is the Aristotelian one, which appeals to a notion of human flourishing. However, as soon as this account of virtue is provided, virtue is no longer primary: it is needed for a person

252 YONG HUANG

to live a flourishing life. Then, however, human flourishing, rather than virtue, becomes primary in this type of ethics, which therefore collapses into an ethics of outcome: virtue is good only because its outcome (helping its possessor to live a flourishing life) is good. Thus, by avoiding one horn of the dilemma (being non-explanatory), these versions of virtue ethics are caught by the other horn of the same dilemma. By falling into this horn, virtue ethics itself becomes a version of consequentialism, to which it intends to be an alternative, and therefore will be affected with whatever problems that virtue ethicists claim that consequentialism has (see Anscombe 1958).

Moreover, according to Watson, for those virtue ethicists who attempt to provide an account of virtue in terms of human nature, there is a second dilemma:

either the theory’s pivotal account of human nature (or characteristic human life) will be morally indeterminate, or it will not be objectively well founded. At best, an objectively well-founded theory of human nature would support evaluations of the kind that we can make about tigers—that this one is a good or bad specimen, that that behaviour is abnormal. These judgments might be part of a theory of health, the reduction of evil to defect . . . . An objec-tive account of human nature would imply, perhaps, that a good human life must be social in character. This implication will disqualify the sociopath but not the Hell’s Angel. The contrast is revealing, for we tend to regard the sociopath not as evil but as beyond the pale of morality. On the other hand, if we enrich our conception of sociality to exclude Hell’s Angels, the worry is that this conception will no longer ground moral judgment but rather express it. (Watson 1997, 67)

This dilemma, as Watson characterizes it here, can be understood in the context of the broader issue of fact and value: On the one hand, if we treat human nature, in light of which an explanation of virtue is provided, as a fact and attempt to provide a descriptive account of human nature, then such an account is morally indetermi-nate and does not lead to a normative conception of virtue; on the other hand, if we treat human nature as a value and attempt to provide a normative conception of human nature, then it is indeed morally determinant and a conception of virtue can be derived from it. However, such an account of human nature can provide no objective ground for the normative conception of virtue; it simply restates this conception. Thus, as Hursthouse states in her attempt to avoid it, this dilemma re-ally means that “either we speak from the neutral point of view, using a scientific account of human nature—in which case we won’t get very far—or we speak from within an acquired ethical outlook—in which case we will not validate our ethical beliefs, but merely re-express them” (Hursthouse 1999, 193).10

III. THE TWO DILEMMAS IN ARISTOTLE’S FUNCTION ARGUMENT

As most contemporary virtue ethicists today are Aristotelian, it helps to see how Aristotle provides an account of virtue. As is well known, Aristotle provides


this account in his so-called function argument. According to this argument, the good and the “well” of anything that has a function or activity must reside in its unique function. So the good and the “well” of human beings must also reside in the uniquely human function, which is characteristic of human beings or, to use McDowell’s term, it is the business of human beings to perform. Aristotle ascer-tains that this human function is the “active life of the element that has a rational principle” (Aristotle: 1098a3); a good human being is one who performs this unique human function well or excellently, and this excellence of performance of the human function is what he means by virtue. Thus virtue is what makes a person a good human being, a human being who performs rational activities, or lives a rational life, excellently.

This account of virtues, particularly the moral virtues that concern us here, is caught by one horn of the first dilemma: the failure to maintain the primacy of virtue. Aristotle distinguishes between practical reason and intellectual reason and their corresponding virtues, moral virtues and intellectual virtues, in a hierarchical order, with the latter being higher than the former. Thus he claims that “if happiness [eudaimonia] is activity in accordance with virtue, it is reasonable that it should be in accordance with the highest virtue; and this will be that of the best thing in us . . . . The activity of this in accordance with its proper virtue will be perfect happiness. That this activity is contemplative we have already said” (Aristotle, 1177a11–18). It is thus not clear whether a good human being has to be morally virtuous, since one can possibly have a well, if not better, functioning reason in contemplation without being moral. Aristotle does acknowledge that the philosopher, who does contemplation, the highest form of intellectual activity, “perhaps can do so better if he has fellow workers” (Aristotle, 1177a35–1177b1), which might be interpreted as saying that one can do philosophy better if one is morally virtuous toward ones fellow human beings. However, Aristotle also makes it clear that (morally) virtuous activities are unleisurely, while philosophical contemplation depends upon leisure. This could be interpreted as saying that virtuous activities are even detrimental to philosophical contemplation. However, even if moral virtues are necessary for the good functioning of human reason, it is not clear whether the Aristotelian ethics can still be properly regarded as a virtue ethics. As we have seen, virtue ethics regards virtue as primary, which contrasts with other types of ethics that allow virtue only a secondary place in their systems. However, if virtue serves reason in Aristotle, then it is not much different from the virtue that serves the greatest happiness in utilitarianism and the virtue that serves duty in deontology.11 In other words, virtue with such a secondary function will not make the ethics virtue ethics, since virtue in such an ethics does not occupy a primary role.12

Moreover, as noticed by several influential scholars, this account is also caught by one horn of the second dilemma: an objective account of human nature is mor-ally indeterminate.13 While people in the Western philosophical tradition have generally agreed, surprisingly from a Confucian point of view, on Aristotle’s view of rationality as characteristic of human beings, as well as his substantive view

254 YONG HUANG

of moral virtues, there have been serious challenges, yet to be met adequately by contemporary defenders of Aristotle, about whether Aristotle’s virtues can be de-rived from his conception of human nature as rational. Bernard Williams, for one, has consistently doubted its possibility. In an early work, he claims that “if it is a mark of a man to employ intelligence and tools in modifying his environment, it is equally a mark of him to employ intelligence in getting his own way and tools in destroying others” (Williams 1971, 73–74).14 In his major work, Ethics and the Limits of Philosophy, Williams continues to claim that there is a gap between Aristotle’s account of human nature and that of virtue: “it is hard to believe that an account of human nature—if it is not already an ethical theory itself—will ad-equately determine one kind of ethical life as against others” (Williams 1985, 52). Martha Nussbaum tries to respond to Williams’s challenge on behalf of Aristotle. Since in the quoted passage Williams adds the conditional “if it is not already an ethical theory itself,” one of Nussbaum’s goals is to show that Aristotle’s account of human nature is not morally neutral, and, in her analysis of Aristotle’s function argument, she comes to this conclusion:

since no life will count as a good life for us unless it is first of all a life for us, and since a life for us must be a life organized, in some fashion, by practical reason, in which all functions are informed and infused by reason’s organiz-ing activity, then eudaimonia must be sought within the group of such lives, not in a life totally given over to bodily pleasure without reason, not in the sleeper’s life of non-guided digestive functioning. (Nussbaum 1995, 116)

With such an account, a life without the guidance of reason indeed cannot be regarded as a human life, as Nussbaum claims. However, it does not show that a vicious life must be a life without the guidance of reason and so must be not regarded as a human life. So Williams is still not convinced: “As Nussbaum points out, however, the life of a wicked or self-indulgent person is equally a certain kind of life structured by reason; it is also a distinctive kind of human life. So far we still wait for the considerations that may move the idea of a life ‘structured by reason’ in the specific direction of life of moderation” (Williams 1995, 199).15

Williams’s suspicion is shared by John McDowell. In an essay on Aristotle’s eudaimonia, McDowell first imagines a debate between a person, X, who thinks that a human being should exercise certain virtues, including those other-regarding ones, and another person, Y, who thinks that the virtuous life is suitable only for contemptible weaklings and that a real man, who looks out for himself, does not practice those other-regarding virtues. Then McDowell remarks on the passage in which Aristotle develops the so-called function argument: this passage “can be read in such a way that the conclusion is (so far) neutral, as between Aristotle’s own substantive view [of virtues] and, say, a view of eudaimonia corresponding to the position of Y in the dispute” (McDowell 1998, 12). In McDowell’s view,

the thesis that man’s ergon [function] consists in rational activity obviously excludes what might otherwise have been a conceivable view of eudai-monia, namely, a life of unreflective gratification of appetite; in the spirit


of the ergon argument, we might say that that embodies no recognizable conception of a distinctively human kind of excellence. But no other likely candidate is clearly excluded by the eliminative argument for that thesis. (McDowell 1998, 13)

In a different essay, on Aristotle’s moral psychology, McDowell further explains the kinds of life that may be excluded and the kinds of lives that cannot be excluded by Aristotle’s function argument:

In fact there are only two substantive points on which Aristotle suggests that facts about human nature constrain the truth about a good human life . . . . First, a good human life must be an active life of that which has logos (NE 1098a3–4); this excludes, for instance, the ideal of uncontrolled gratification of appetite with which Socrates saddles Callicles, in Plato’s Gorgias. Second, human beings are naturally social (NE 1097b11, 1069b18–19); this excludes a solitary life. Obviously these two points fall a long way short of purporting to afford a validation of Aristotle’s ethic in full. (McDowell 1998, 35–36)

To further illustrate his point that rationality does not lead to virtue, in an essay arguing against Aristotelian naturalism, McDowell imagines a rational wolf. Without reason, the wolf would find it natural for him to play his part in the co-operative activity of hunting with the pack. However, “Having acquired reason, he can contemplate alternatives; he can step back from the natural impulse and direct critical scrutiny at it . . . and frame the question ‘why should I do this?’ . . . wondering whether to idle through the hunt but still grab his share of the prey” (McDowell 1998, 171). In McDowell’s view, even if the wolf by its nature does what virtue might require it to do, the addition of reason may cause it to question its natural behavior. Then, McDowell draws the lesson: “even if we grant that human beings have a naturally based need for the virtues, in a sense parallel to the sense in which wolves have a naturally based need for co-operativeness in their hunting, that need not cut any ice with someone who questions whether virtuous behaviour is genuinely required by reason” (McDowell 1998, 173).

IV. THE TWO DILEMMAS IN HURSTHOUSE’S NEO-ARISTOTELIAN ETHICAL NATURALISM

Aristotle did not claim that he was a virtue ethicist; at least he did not develop his ethics as an alternative to deontology and consequentialism.16 So to claim that his ethics suffers from these two dilemmas is perhaps anachronistic and unfair. In contrast, Rosalind Hursthouse clearly has these two dilemmas in her mind when she develops her neo-Aristotelian virtue ethics as an ethical naturalism. She upholds a conception of human nature, which she claims is both objective and normative and thus can avoid the second dilemma. She does it by an attempt to show that virtues are character traits that are constitutive of human flourishing, and thus she claims that it can avoid the first dilemma as well. In this section, I shall examine whether she is successful in these respects.

256 YONG HUANG

In order to show, as Anscombe, Geach, and Foot did before her, that “‘good’ is an attributive term and does not suddenly change its grammar when we start evaluating ourselves ethically” (Hursthouse 1999, 206; see also 259), Hursthouse, following Philippa Foot, examines how terms such as “good” are applied to plants, animals, and finally humans in the ascending order of the ladder of nature. She seeks to show both the continuity and variations of our applying these same terms to these various things, and that no magic has ever taken place when we climb the ladder. Hursthouse claims that such terms, while normative, are clearly objective when applied to plants and animals, as they are really synonymous with “healthy” and “defective”: “the truth of such evaluations of living things does not depend in any way on my wants, interests, or values, nor indeed on ‘ours’. They are, in the most straightforward sense of the term, ‘objective’” (Hursthouse 1999, 202). So for her, “what goes for ‘good cactus’ and ‘good wolf’ also goes for ‘good human being’, and the adverb(s) ‘morally’ or ‘ethically’ [in contrast to physically, for example] added to ‘good human being’ can do no more than restrict the aspects of human beings to be considered; it cannot change the grammar” (Hursthouse 2002, 52).17 Hursthouse first examines our evaluation of plants. Here, “we evaluate two aspects—parts and operations—in relation to two ends” (Hursthouse 1999, 198). Parts are such things as leaves, roots, and petals, and operations are such activities as growing, taking in water, developing buds, dying back, and setting seed. The two ends are “(1) individual survival through the characteristic life span of such a member of such a species and (2) continuance of the species” (Hursthouse 1999, 198). So a good plant is one that is well fitted with respect to its parts and operations in terms of both its individual survival and the continuance of its species. In other words, a good plant is one whose two aspects serve its two ends well.

Hursthouse then moves to our evaluation of animals, where two additional as-pects and two additional ends are introduced. The first new aspect (the third aspect) is acting or doing in contrast to merely reacting as is found in plants. The second new aspect (the fourth aspect) is a certain psychology of emotions and desires. With these two new aspects, there are not only more complex ways of realizing the two ends; but there are also two new ends to be served. The first new end (the third end) is the “characteristic freedom from pain and characteristic pleasure or enjoyment” (Hursthouse 1999, 199), as animals seek both freedom from pain and enjoyment of pleasure, while plants do not. The second new end (the fourth end), particular in social animals such as wolves and bees, is “the good functioning of the social group.” So now whether an animal, particularly a sophisticated one, is good or defective is to be evaluated with the above four aspects in light of their service to the four ends. This last end, “the good functioning of social group,” characteristic of social animals, is particularly important in view of its relevance to our evaluation of human beings, which are also social animals. This is a function “to enable its members to live well (in the way characteristic of their species); that is, to foster their characteristic individual survival, their characteristic contribution to the continuance of the species and their characteristic freedom from pain and


enjoyment of such things as it is characteristic of their species to enjoy” (Hurst-house 1999, 201). In short, the good functioning of the social group, the fourth end, is to enable its members to better realize the other three ends. To illustrate this, she uses a similar example used by Philippa Foot: “Wolves hunt in packs; a ‘free-rider’ wolf that doesn’t join in the hunt fails to act well and is thereby defec-tive” (Hursthouse 1999, 201).

With such a preparation, Hursthouse is now ready to describe our ethical evaluation of ourselves, human beings. In her view, “if there is any truth in ethical naturalism, our ethical evaluations of ourselves ought to exhibit at least recogniz-ably similar structure to what we find in the botanists’ and ethologists’ evaluation of other living things. More particularly, we would expect the structure of our ethical evaluations of ourselves to resemble that of sophisticated social animals with some differences necessitated by our being not only social but also rational” (Hursthouse 1999, 206). What is unique is that, in our evaluation of ourselves, a significantly new aspect (the fifth) comes onto the scene: rationality. Now we have five aspects, although the first two aspects, parts and merely physical operations, which become the subject matter of human biology and/or medicine, are irrelevant to our ethical evaluations of ourselves (this is what she means, in an earlier quota-tion, when she says that the term “ethical” simply serves to restrict the aspects of human beings to be considered). In this ethical evaluation, the aspects to be evaluated include then those reactions that are not merely physical, i.e., actions, emotions and desires, and our rationality.

It is here that Hursthouse brings the idea of virtue into her ethical naturalism. In her view, “it is primarily in virtue of our actions from reason that we are ethically good or bad human beings” (Hursthouse 1999, 217). However, she also claims that to possess virtue is

not only to be well disposed with respect to actions from reason, but also with respect to emotions and desires . . . . Virtuous action also involves ‘reactions which are not merely physical’ in the perceptions of what is rel-evant in a situation . . . . Hence the concept of virtue emerges as apparently tailor-made to encapsulate a favourable evaluation of just those aspects which, according to the naturalism here outlined, are the ethically relevant ones. (Hursthouse 1999, 208)

It is interesting to see that all aspects that Hursthouse thinks are ethically relevant, other than rationality, are also present in animals. However, why do we not make ethical evaluations of animals and regard them as either morally virtuous or vi-cious? The reason for Hursthouse is that rationality, the unique aspect in human beings, is important not only because it is a new aspect to be evaluated, but also because it affects all other aspects to be evaluated in a human being. For example, action in animals is action from inclination, but action in humans is action from reason. Although we human beings occasionally also act from inclination, it is the action from reason that “makes us good or bad human beings in the ethical sense” (Hursthouse 1999, 207). Similarly, emotions and desires in human beings,

258 YONG HUANG

unlike those in animals, are also ethically relevant aspects and can be regarded as virtuous and vicious, because they are also affected by rationality: “in the person with the virtues, these emotions will be felt on the right occasions, towards the right people or objects, for the right reasons” (Hursthouse 1999, 109). As a matter of fact, Hursthouse claims that full virtue involving feeling emotions is generally impossible without the influence of reason (Hursthouse 1999, 109).

It is important to note that, in animals, particularly social animals, new ends emerge with the appearance of new aspects or capacities. However, in human beings, the appearance of the new capacity of rationality is not accompanied by any additional end. Hursthouse particularly shrugs off Aristotle’s suggestion of contemplation as an additional end for human beings.18 Nevertheless, she still in-sists that the appearance of rationality is enough to register the huge gap between humans and animals. What is characteristic of other beings is largely determined by nature. So “it makes no sense to say that, for example, a male polar bear is a bad/defective polar bear because, far from defending its young, it has to be prevented by their mother from killing them” (Hursthouse 1999, 220), because what they do not do is what they cannot do, and what they ought to do must be what they can do. In contrast, what is characteristic of human beings is to a great extent determined by “our rationality—our free will if you like . . . . Apart from obvious physical constraints and possible psychological constraints, there is no knowing what we can do from what we do do, because we can assess what we do do and at least try to change it” (Hursthouse 1999, 221). So what is the charac-teristic way of being human? Hursthouse claims that it is the rational way, which “is any way that we can rightly see as good, as something we have reason to do. Correspondingly, our characteristic enjoyments are any enjoyments we can rightly see as good, as something we in fact enjoy and that reason can rightly endorse” (Hursthouse 1999, 222).

Despite no new ends emerging with rationality in human beings, Hursthouse claims that her notion of human nature, on the one hand, is still a normative one, as the characteristic way of being human in this sense is not necessarily (and in most cases is not) the way most humans are living their lives, and for that reason, most people are not living the way of life characteristic of being human (Hurst-house 1999, 224); and, on the other hand, it is nevertheless naturalist or objective, because “it is still the case that human beings are ethically good in so far as their ethically relevant aspects foster the four ends appropriate to a social animal, in the way characteristic of the species. And the structure—the appeal to just those four ends—really does constrain, substantially, what I can reasonably maintain is a virtue in human beings” (Hursthouse 1999, 224). So although rationality is unique to human beings, it is not an end in light of which we judge whether a human being is good or not. In this sense, contrary to Christopher Toner’s complaint,19 Hursthouse can respond to McDowell’s question on behalf of the rational wolf about whether it is a good wolf, because Hursthouse, unlike Aristotle, does not regard rationality as the criterion to evaluate whether a human being is good or


not. It is rather something we need to evaluate in light of the four ends: if it serves these ends well, then it is good, and therefore the person who has the rationality is a good person. So the crucial part of Husthouse’s account is the four ends, on the basis of which her account of human nature is indeed both naturalistic and normative, and to that extent it indeed avoids the second dilemma of virtue ethics we discussed above.20

It is no wonder that many critics of Hursthouse target their criticisms to these four ends. David Copp and David Sobel, for example, claim that Hursthouse’s

general point is that what counts as a virtue is still determined by what serves those four ends and this is a real constraint on the view and a real tie with Footian naturalism. But this is difficult to understand. The list of four ends that Hursthouse recommends we use to evaluate plants and animals was developed precisely by generalizing about how, according to Hursthouse, we evaluate the kind of creatures for whom it is the case that nature determines how they ought to be. How can Hursthouse reject the thought that nature determines how humans should be yet think that the same considerations that grounded the four ends in plants and animals also ground the normative status of the four ends for humans? (Copp and Sobel 2004, 540)

So they argue that Hursthouse faces a dilemma: she must either reject the Footian naturalism about the four ends or accept that nature can be normative for us: “if she rejects the idea that nature can be normative with respect to us, as she does, and if she concedes that, for humans, the normatively appropriate way of going on is to act in ways that we can rightly see ourselves as having reason to act, as she does, she must give up the Footian naturalism” (Copp and Sobel 2004, 541).21 A more careful reading of Hursthouse, however, can show that there is no such inconsis-tency. What Hursthouse maintains is that all social animals are good in so far as the relevant aspects foster the four ends in a way characteristic of a species. The difference between non-rational social animals and rational social animals (human beings) is that, in the former, the characteristic ways to foster the four ends are given by nature, while in the latter, it is any way that these rational animals see as right or have reason for. However, whatever ways they have for whatever reasons must be ways to foster the four ends. In short, rationality for Hursthouse does not determine what ends rational animals should serve but only what ways they can find to serve the four ends.

With this defense of Hursthouse, however, we start to doubt whether rationality, as the only thing uniquely human, does register the “huge gap” between human beings and other social animals as Hursthouse claims. Her project of ethical natural-ism attempts to show that an objective conception of human nature is normative. Now, if we ask what human nature is, it seems that we first need to know whether human nature is to be determined by the aspects or the ends. It seems that it can-not be the ends, since in terms of ends there is no difference between humans and other social animals, and human nature presumably is the uniquely human nature.

260 YONG HUANG

We must look for it in the aspects, where only rationality separates humans from other social animals. Rationality must be the uniquely human nature, but rationality in her naturalism is merely a human capacity to foster the four ends that humans share with other social animals. If so, however, the gap, if it can still be called a gap, between human beings and animals is certainly not huge, as the only thing uniquely human, rationality, has to be constrained by the four ends, which humans and animals share.

It is in this sense that we may also start to doubt, while Hursthouse’s ethical naturalism can indeed respond to McDowell’s objection on behalf of a rational wolf, whether it can respond to Watson’s objection on behalf of a social gangster: “can an objective theory really establish that being a gangster is incompatible with being a good human being?” (Watson 1997, 67). Hursthouse claims that it can, and the way she does it is first to validate charity, justice, etc. as virtues and then to estab-lish that gangsters are callous, unjust, etc., in short, not virtuous. For Hursthouse, “these moves together would establish that a gangster was bad qua human being, and thereby unable to live a good human life” (Hursthouse 1999, 228). However, as we have seen, virtue for Hursthouse is the character trait that helps one foster the four ends. The reason Hursthouse can claim that the gangsters lack virtue is apparently that their character traits do not foster the fourth end, the smooth func-tioning of the social group. Since this end is carried over from her discussion of social animals, presumably there is no difference between human beings and other social animals in terms of this end. Now if we look at a good social animal, for example, a wolf that helps in hunting, this animal serves the end of smooth function of the social group. However, the social group that the wolf serves is apparently the particular pack of which it is a member and not the species of wolf. The wolf may well fight wolves from other packs. If this wolf can be regarded as a good wolf simply because it contributes to the smooth functioning of its own group instead of its species, then a gangster may also be regarded as good human being, as this gangster may very well also serve the smooth functioning of his group, the gang of which he is a member. Hursthouse may say that this gangster is perhaps indeed good, but qua gangster, not qua human being. If this is the case, then we cannot say a wolf that helps its pack hunting is a good wolf: it is good only qua a member of its pack and not qua wolf. If a wolf can be regarded as a good wolf simply by helping its pack to function smoothly, while a human being cannot be regarded as a good human being unless he or she helps the human species function smoothly, then the relevant end that humans and other animals serve respectively must be different, a possibility Hursthouse explicitly excludes.

To help Hursthouse fix this problem, several commentators propose that the ends in terms of which we make ethical evaluations of humans cannot be the same as those in terms of which we make evaluations of social animals.22 For them, to add an additional end to human beings corresponding to their acquisition of a new aspect in comparison to animals is just as natural as to add additional ends to animals with their acquisition of new aspects in comparison to plants.23


Christopher W. Gowans, for example, notes the difference between human beings and other social animals we mentioned above. For him, while it is implausible to think that a good human being should only be concerned about members of his or her particular group and not all human beings, “it is implausible to suppose that social animals in general are concerned about all members of their own respec-tive species” (Gowans 2008, 46). Thus in his view, to account for this difference, it is important to show “that human beings, as rational, have as an end (among others) the well-being of all other members of their species” (Gowans 2008, 47). Brad Hooker goes a step further, as he thinks that the human species is still not broad enough, and the end should be further expanded to include all rational beings. Thus he imagines a science-fictional scenario, in which “human beings all have an incurable virus that they foresee will make them all utterly vicious and miserable,” with the result that they entertain themselves by killing off any other intelligent species, one of which is just as smart as but kinder and happier than human beings. In this case, he argues that the survival of human beings should not be considered as one of the ends in light of which we make ethical evalua-tions of human beings, as “all the positive value seems to be on the side of having humans become extinct and the other intelligent species flourish” (Hooker 2002, 35). What he tries to show by using this science fictional scenario is that “what is truly valuable has no necessary connection to the human species” (Hooker 2002, 35). Christine Swanton goes a step still further. In addition to doubting whether all anthropocentric virtues should be understood solely in terms of the four ends, she also challenges whether the ground of all virtues is anthropocentric. What she has in mind is environmental virtues: “on the classic view of deep environmental ethics, the respect we should accord natural objects by virtue of their status as natural is not based on their serving the enjoyment or smooth social functioning of the human species” (Swanton 2003, 92).

It seems that, in order to fix the problem mentioned above and make an ap-propriate response to Watson’s objection on behalf of a gangster, Hursthouse has to adopt something like the suggestions of Gowans, Hooker, and Swanton, so that the fourth end in human beings, instead of the smooth functioning of the social group as in other animals, will be the smooth functioning of the human species, or all rational beings, or the whole environment. A potential problem with this revision of the ends, in light of which we make ethical evaluations of human be-ings as virtuous or not, is that it is not clear whether such a conception of human nature, while certainly normative, is still objective. The virtue of simply carrying the four ends from other social animals over to human beings, as Hursthouse does, is that their objectivity in other social animals is also carried over to human beings. However, when the fourth end is revised as smooth functioning of the human spe-cies, or rational beings, or the environment, its objectivity becomes ambiguous, at least from the Aristotelian perspective, as the truth of ethical evaluation in light of such an end perhaps indeed depends on “my wants, interests, or values,” and they are not, “in the most straightforward sense of the term, ‘objective’” (Hursthouse

262 YONG HUANG

1999, 202). If this is the case, then Hursthouse’s ethical naturalism cannot after all escape the second dilemma.

However, even if we assume that Hursthouse’s ethical naturalism, without any revision of the four ends, can respond to Watson’s objection on behalf of a gangster or has to revise at least one of the four ends in order to respond to this objection and yet can still maintain the objectivity of its account of human nature, in short, even if we assume that it can avoid the second dilemma, there still seems to be a problem. On this account, a character trait is good (virtuous) simply because it fosters the four ends. If this is the case, virtue is no longer primary or intrinsically good. It is good only in the sense that it is conducive to the four ends and therefore is only instrumentally good. It is in this sense that Brad Hooker declares the collapse of virtue ethics into rule-utilitarianism, since Hursthouse’s virtue ethics evaluates a disposition’s being virtuous or vicious with a view to the consequences for well-being (Hooker 2002, 38) and thus falls back into the first dilemma.

Hursthouse’s response to this is that Hooker misunderstood her. While Hooker thought that she uses the four ends to justify a given character trait as a virtue, what she actually does is to use them to explain or to increase our understanding of a given character trait, which we already know independently is a virtue. Thus she claims that “if the criterion were to be regarded as a principle, it would be better to think of it as an explanatory, in contrast to a justificatory one. For, I would say, we are a great deal more certain that charity, justice, honesty, temperance, courage, and so on, are virtues and that people who have those character traits are ethically good human beings than we are about any putative philosophical justification of them” (Hursthouse 2002, 52). This response, however, is not convincing. What Hooker thinks she does and what I consider to be her problem is not that she provides any philosophical justification for virtues, as she does not, but that she tries to provide some naturalistic justification for them, to use the four natural ends to justify a given character trait as a virtue. This seems to be precisely what she does in several places. For example, she claims that “ethical naturalism hopes to validate beliefs about which character traits are virtues by appeal to human nature” (Hursthouse 2002, 193; emphasis added); “I cannot just proceed from some premises about what it is reasonable or rational to do to some conclusion that it is rational to act in such-and-such a way, and hence that a good human being is one who acts that way. I have to consider whether the corresponding character trait . . . would foster or be inimical to those four ends” (Hursthouse 1999, 224); and “the structure—the ap-peal to those four ends—really does constrain, substantially, what I can reasonably maintain is a virtue in human beings” (Hursthouse 1999, 224; emphasis added). I believe it is quite reasonable to say that, at least in such passages, Hursthouse tries to provide a justification for rather than an explanation of a given character trait as virtue. However, it really does not matter that much whether it is a justification or an explanation. Suppose that Hursthouse indeed means to provide an explanation of, rather than justification for, a given character trait as a virtue, and therefore we already know whether a given character trait is a virtue or not. Disregarding how


we could know it without appealing to our intuitions, as Slote does, there is still a question of whether virtue is indeed primary in her ethical naturalism. Although in the ordo cognoscendi (order of cognition), the goodness of certain characters, the explanadum, is prior to the naturalist account of human nature, the explans, in the ordo essendi (the order of existence), the explanans is prior to the explanadum. In other words, although we know that a character trait is a virtue before we know why it is a virtue (before we can explain it), a character trait is virtue only because it serves to realize human nature. If so, virtue is no longer primary. It is good only instrumentally. Thus, after all, Hursthouse’s ethical naturalism is still caught by one of the horns of the first dilemma.

V. ZHU XI’S NEO-CONFUCIANISM AND THE FIRST DILEMMA OF VIRTUE ETHICS

Having examined the inescapability of the two dilemmas in the Aristotelian virtue ethics, I shall examine how Zhu Xi’s neo-Confucianism fares with respect to these two dilemmas in this and the next section respectively. It is true that, just like Ar-istotle, Zhu Xi did not consciously try to develop a virtue ethics, to say nothing of regarding it as an alternative to deontology and consequentialism, and so was of course unaware of the two dilemmas in virtue ethics that we are discussing here. He simply tried to provide an updated and coherent interpretation of Confucian ethics. However, it is my contention that Zhu Xi’s neo-Confucian view of virtue can avoid both dilemmas.

To recall, the first dilemma in virtue ethics is that, on the one hand, if one tries to provide an account of virtue in terms of something else, virtue is no longer pri-mary, and, on the other hand, if one wants to maintain the primacy of virtue and so refrain from explaining it in terms of something else, one lacks a clear idea of what constitutes a virtue. Zhu Xi does choose to provide an account of virtue, and, just like Aristotle, he also appeals to a notion of human nature to do so. However, Zhu Xi’s explanation of virtue does not render virtue as something secondary as is the case in the Aristotelian tradition, or so I shall argue. What makes the dif-ference is that Zhu Xi has a conception of human nature very different from the Aristotelian one. Following the Mencian line of Confucianism, Zhu Xi regards human nature not merely as something that human beings are born with but as something characteristic of human beings, something that distinguishes human beings from other beings. In this sense, he still agrees with Aristotle. The question is what distinguishes human beings from other beings. Philosophers in the Western tradition have almost unanimously accepted the view that it is rationality, although they may disagree on precisely what rationality is.

Given this, it may seem surprising that this idea has hardly occurred to Confu-cian philosophers, including Zhu Xi. For them, it is just a simple truth that what is characteristic of human beings must be their humanness, humaneness, or human-ity, which is precisely what ren 仁, the most important Confucian virtue, means. So for Zhu Xi what distinguishes human beings from beasts is not rationality but

264 YONG HUANG

virtues, particularly the virtue of humanity. Although ren is only one of the four cardinal Confucian virtues, along with yi 義 (rightness), li 禮 (propriety), and zhi 智 (moral wisdom) (and sometimes a fifth one, xin 信 [trustworthiness] is added), Zhu Xi argues that ren is inclusive.24 This is indeed a unique aspect of the Confu-cian account of virtue. Aristotle talks about the unity of the individual virtues, none of which is dominating in the whole system of virtues. In Confucianism in general and in Zhu Xi’s neo-Confucianism in particular, however, other individual virtues are all led by, included in, or reflect some aspects of the dominant virtue, human-ity. Arguing against the Buddhist conception of human nature as empty, Zhu Xi claims that human nature is substantive (shi 實), as its content is nothing but these four cardinal Confucian virtues. Thus, Zhu Xi claims:

human nature is the undistinguished whole of the substance of the ultimate. Originally we cannot say anything about it. However, it entails ten thousand principles, of which the most fundamental are the following four: humanity, rightness, propriety, and wisdom. Confucius did not talk about human nature as much as Mencius did. This is because everyone understood that human nature is good in Confucius’ time. So even though Confucius did not explain it in detail, the idea of the goodness of human nature is there. However, heresies arose everywhere in Mencius’ time, which regarded human nature as not good. Mencius was concerned that the truth was obscured and so he tried to make it clear. (Wenji 58, “Reply 2 to Chen Qizhi,” in Zhu 1996, 2977)

Here Zhu Xi makes it clear that “human nature entails . . . humanity, rightness, propriety, wisdom,” the four Confucian cardinal virtues. In another place, he makes an analogy: “human nature is general. It is like a human body, where humanity is the left hand, propriety is the right hand, rightness is the left foot, and wisdom is the right foot” (Yulei 6, 100).25 Thus, for Zhu Xi, human nature and virtues are identical in the sense that they are constituted by the same things: humanity, right-ness, propriety, and wisdom. In this sense, he often talks about virtuous human nature (dexing 德性) and virtues of human nature (xing zhi de 性之德).26 In other words, human nature is virtuous, and virtues belong to human nature. In his view, it is precisely here that human beings are distinguished from animals. Commenting on Mencius 6a3, Zhu Xi claims,

Human nature is metaphysical (above the form), while the vital force is empirical (below the form). When born, human beings and non-human be-ings all have this principle as well as the vital force. In terms of vital force, there is no distinction between human beings and non-human beings in their having perceptions and ability to move. In terms of principle, however, how can non-human beings be fully endowed with humanity, rightness, propriety, and wisdom? So no human’s nature is not good. Humans are thus the most spirited among ten thousand things. (Mengzi Jizhu 6a3, in Zhu 1994, 457)

Mencius claims that “the difference between humans and animals is very slight. Superior persons preserve it, while inferior people abandon it” (Mencius 4b19). When a student asks what the “little” thing that superior persons preserve and


inferior people abandon is, Zhu Xi replies that it is what makes humans different from animals, not what humans and animals share. While to hear with ears, see with eyes, smell with nose, make noise with mouth, walk with feet, and carry things with hands are what humans and animals share, “what makes humans different from animals is that humans have humanity, rightness, propriety, and wisdom, so that as a son one can have filial love for parents, as a younger brother one can have respect for one’s older brother, and etc. How can animals have them?” (Yulei 57, in Zhu 1997, 1203; citation from this text hereafter will be indicated by the title, volume number, and page number only). Precisely because this is the only distinction between humans and animals, on the one hand, “only when one can preserve this slight difference, can one be distinguished from animals” (Yulei 59, 1240); and, on the other hand, “if one’s heart/mind [whose substance is human nature] is obscured by selfish desires, one has become an animal” (Yulei 57, 1203).

It may occur to us that this distinction between humans and animals is outdated, as contemporary sociobiology and evolutionary psychology have shown convinc-ingly that there is actually continuity between animals and humans in terms of their altruistic behaviors. In this sense, we may think that Zhu Xi’s account of human nature is after all inferior to Hursthouse’s, which makes full use of contemporary scientific materials, particularly in her discussion of the end of smooth function-ing of social groups that all social animals naturally serve. However, while living way before the advent of modern sciences, Zhu Xi was fully aware of the social nature of some animals. For example, he states that “tigers and wolves have love (ren) [between parent and child], bees and ants have rightness [between superior and inferior], jackal and otter know defending their own groups, and pheasant and turtledove know the difference [between male and female]” (Yulei 4, 54). As a mat-ter of fact, Zhu Xi even goes so far to claim that “the nature of human beings and animals are the same, and the difference between them is that they are endowed with different qi (vital forces)” (Yulei 4, 53). In other words, the four qualities that are recognized as virtues in human beings, “humanity,” “rightness,” “propriety,” and “wisdom,” are also present in animals.

In appearance, this is a statement in contradiction with the one quoted above which distinguishes between humans and animals. It seems that when he wants to emphasize the continuity between humans and animals, he holds the view that they have the same nature or principle but only different qi, while when he emphasizes their differences, he seems to think that they have similar qi but different natures or principles.27 When one of his students asked about this seeming contradiction, Zhu Xi explains,

When I say that [ten thousand things] have the same principle but differ-ent qi, I mean that at the very beginning, the same heaven-destined nature goes through all ten thousand things, while the two qi [yin and yang] and five elements [gold, wood, water, fire, and earth] they are endowed with are different from each other in terms of their being clear or muddy and pure or mixed. However, when I say [ten thousand things have the similar

266 YONG HUANG

qi but radically different principles], I speak about the presence of the same principle within the ten thousand things. Although different in terms of be-ing clear or muddy and pure or mixed, they are all qi of yin and yang and five elements and so are similar. However, the principle in the ten thousand things appears radically different in terms of its being dark or bright and open or blocked. (Yulei 4, 52)

Zhu Xi further explains it with an analogy. The principle or nature is like water, while qi is like bowls. With regard to qi, as they are bowls of different colors, dif-ferent bowls are different from each other, and it is in this sense that he claims that human beings and animals have different qi. As they are all bowls of some color, however, they are the same, and it is in this sense that he claims that human beings and animals have the same qi. With regard to principle or nature, since different bowls have the same water, they are all the same, and it is in this sense that he claims that human beings and animals have the same nature. As the same water in bowls of different colors appear also to have different colors, however, they are different, and it is in this sense that he claims that humans and animals have dif-ferent natures (see Yulei 4, 53).

So precisely what is Zhu Xi’s view about the similarity and difference between humans and animals? In his view, the four qualities that are recognized as human virtues are originally present in both humans and animals, just as the same water is present in different bowls; and both humans and animals are endowed with the qi of yin and yang and the five elements, just as different bowls are all made of the same materials. However, the qi that human beings are endowed with are transparent and open, while the qi that animals are endowed with are opaque and blocked, just as different bowls have different colors. This difference in the quality of qi that humans and animals are endowed with also makes a difference to the four qualities present in them, just as the difference in the color of different bowls makes a difference in how the same water they contain looks (water in a black bowl looks black, water in a red bowl looks red, etc.). Since the qi that humans are endowed with are transparent and open, the four qualities in humans are fully extended and therefore recognized as the four virtues; and since the qi that ani-mals are endowed with are opaque and blocked, the four qualities in animals are limited and therefore cannot be regarded as virtues. For example, Zhu Xi states, “ants and bees have only a slight bit of rightness [one of the four qualities] in their behaviors between the leaders and followers; and tigers and wolves have a slight bit of love [ren, another of the four qualities] in their behaviors between parents and children. However, they are unable to extend these qualities. Just like a mirror, there are only a few spots that are clear, while all the rest are dark” (Yulei 4, 52). In contrast, human heart/mind is “vacuous and transparent. [Just like a mirror] there is no place where it is not bright . . . and therefore the four qualities can be fully extended” (Yulei 57, 1202).

It is in this sense that Zhu Xi emphasizes the importance of extension (tui 推), as it is the presence or absence of this ability to fully extend the four qualities that are


naturally present in both humans and animals that sets humans and animals apart; and it is also the presence or absence of this ability that establishes four qualities in human beings, and not in animals, as virtues: human beings can, while animals cannot, extend the four qualities present in them. Thus, Zhu Xi makes a series of important claims in relation to tui. For example, he states that “when manifested, ren (humanity) is love. Shu 恕 (sympathetic/empathetic concern) is to extend the love, while the love is what is extended. Without shu, the love cannot be extended to others, and therefore one cannot be affectionate to one’s parents, humane to other people, and love non-human beings [the three kinds of love in the broad sense that Mencius talks about], but there would be only self-love” (Yulei 95, 2205); when explaining the idea of filial piety and brotherly love as the root of humanity, Zhu Xi states that “once one has filial piety and brotherly love, one should extend it, so that one can love other people and benefit non-human beings” (Yulei 20, 413);28 again, “love is the manifestation of ren. As soon as it is manifested, everything naturally follows: first is love for one’s parents, then for one’s siblings, then for one’s relatives, then for friends, and further extended to people in general” (yulei 119, 2589); indeed, commenting on Mencius’s view about the “slight difference” between humans and animals that superior persons preserve and inferior persons abandon, Zhu Xi claims that, if one’s selfish desires block the original vacuous and transparent heart/mind so that one is unable to extend one’s inborn four qualities, one is no different from animals (see Yulei 57, 1203).

Thus, while Zhu Xi’s metaphysical view of human nature, particularly when coupled with such notions as principles (li) and material forces (qi), may sound unintelligible to Western ears, particularly in this so-called post-metaphysical age, what Zhu Xi wants to emphasize is this simple empirical fact/moral value, which can be disentangled from this metaphysics:29 while other social animals also exhibit a certain degree of altruistic behavior, they cannot extend them beyond their im-mediate group; in contrast, human beings can extend them beyond their immedi-ate group to the human species and even to non-human beings. His metaphysics of li and qi is simply used to explain this empirical fact/moral value. So even if his metaphysical explanation is unintelligible or unacceptable to contemporary Western virtue ethicists, the empirical fact/moral value that it intends to explain is undeniable. Contemporary science may have quibbles with the precise details of the behaviors of the particular animals that Zhu Xi describes, but this is not consequential to our discussion. What is consequential is that this empirical fact/moral value can avoid the difficulty that Hursthouse encounters in her distinction between human beings and animals. As we have seen in the previous section, since Hursthouse thinks that there is no difference between humans and animals in terms of the four ends, particularly the fourth end, the smooth functioning of the social group, she is unable to respond to Watson’s objection on behalf of a social gangster: a gangster may act in exactly the same way a wolf does. Just as a wolf joins its pack in hunting and fighting against other packs, a gangster may also join his gang in robbing and defending his gang against people outside his

268 YONG HUANG

gang. Since Hurthouse regards such a wolf as a good wolf, why cannot we regard such a gangster as a good human being? Now Zhu Xi’s observation of the differ-ence between human beings and animals can avoid this problem, since for him, while it is natural for animals (it is their nature) to have altruistic behaviors only limited within their narrow immediate groups, it is natural for humans (it is their nature) to extend such behaviors to all humans and beyond. With this, the demands made by Gowans, Hooker, and Swanton in their arguments against Hursthouse can be fully met by Zhu Xi. Since such an ability to extend, not rationality, is the unique capacity or (to use Hursthouse’s term) aspect of human beings, a human being who does not exercise this ability is defective, and so is not a characteristic human being.

Now we can take stock of how Zhu Xi’s neo-Confucianism fares with the first dilemma of virtue ethics. As we have seen, on the issue of what constitutes a virtue or what makes a character trait virtuous, instead of relying on our moral intuitions, which leaves the idea of virtue non-explanatory, Zhu Xi proceeds to provide an account of virtue. Moreover, just like Aristotle and Aristotelians, Zhu Xi’s account of virtue is also in the light of human nature. In this sense, he would certainly agree with Hursthouse that “the virtues make their possessor a good human being. (Human beings need the virtues in order to live well, to flourish as human beings, to live a characteristically good, eudaimon, human life)” (Hurst-house 1999, 167). The question now is whether Zhu Xi’s account of virtue in the light of human nature, just as Aristotle’s and the Aristotelians’, fails to maintain the primacy of virtue, so that an ethics based on such a conception of virtue is no longer a virtue ethics but simply a theory of virtue and so is fundamentally indis-tinguishable from other ethical theories that can also leave a significant room for virtue, as we examined in section II. Our answer is negative. Although Zhu Xi also thinks that virtue is good because it makes its possessor a characteristically human being, virtue is not merely an instrumental good, not merely because one cannot live a characteristically human life without virtue or because one cannot abandon virtue as soon as one lives a characteristically human life. This is something that even an Aristotelian would also agree on, and it is in this sense that they claim that virtue has intrinsic value. It is rather because a characteristically human life itself is defined by virtue. For Zhu Xi, it is virtue, not rationality or anything else, that defines human nature. Since human nature is virtuous, and so to be a characteristic human being, one has to be virtuous, as otherwise one will be no different from animals; since Zhu Xi identifies human nature with virtues, to say that one needs virtues in order to become a characteristically human being will not make such an ethics of virtue into an ethics of outcome, as the very outcome one’s virtue helps one to have is to be a person of virtue. This way, Zhu Xi’s neo-Confucianism provides an account of virtue that, while based on a notion of human nature, can maintain the primacy of virtue. It is in this sense that we can claim it avoids the first dilemma of virtue ethics.


VI. ZHU XI’S NEO-CONFUCIANISM AND THE SECOND DILEMMA OF VIRTUE ETHICS

However, there seems to be a reason for caution. In the above, we claim that Zhu Xi provides an account of virtue that escapes the first dilemma of virtue ethics, but how does he manage this? As we have seen, he tries to show that virtue is necessary for one to live a genuine or characteristically human life, which seems to make virtue an instrument for a good life, and then he says that a genuine or characteristically human life is a virtuous life, which restores the primacy of virtue. In other words, he seems to have made a circular argument: virtue is defined by human nature, which itself is defined by virtue. If so, Zhu Xi’s account of virtue does not provide an objective explanation of virtue but simply re-expresses it. As we can recall, this is precisely one horn of the second dilemma of virtue ethics: an account of virtue can be objective and therefore becomes morally indeterminate or normative and therefore lacks objective ground. If Zhu Xi succeeds in avoiding the first dilemma only to be trapped in the second, then the success is a cheap one. So in this section, I shall examine how Zhu Xi’s account of virtue fares with respect to this second dilemma. Since it is clear that his conception of human nature, in light of which he explains virtue, is not morally indeterminate (and so avoids one horn of the second dilemma), our job is then to see whether this account lacks objectivity and thus simply re-expresses what it intends to explain: virtue. In other words, when Zhu Xi claims that human nature is virtuous or that virtues belong to human nature, does he simply make a normative assertion about human nature, or does he also have an objective ground for it?

At first glance, the former seems to be the case. In the passage quoted in the previous section, Zhu Xi says that human nature is metaphysical, an undistin-guished whole, about which we cannot say anything. In the same letter, a little later, Zhu Xi reaffirms this, saying that human nature, when not aroused, is an “undistinguished whole, without sound, smell, and image.” How can we know anything about human nature, to say nothing about its being virtuous? To answer this question, Zhu Xi distinguishes between human nature and human emotions. Humanity, rightness, propriety, and wisdom belong to human nature, while feel-ings of commiseration, shame and dislike, deference and compliance, and right and wrong are their corresponding emotions. Cheng Yi 程頤, from whom Zhu Xi learns a great deal, once said that “from the feeling of commiseration we can know [the nature of] humanity” (Yishu 15, in Cheng and Cheng 2004, 168). Zhu Xi regards what Cheng Yi says here as most insightful, as it “does not say that the feeling of commiseration is humanity, nor does it say that we can talk about humanity independently of commiseration” (Yulei 53, 1151). In Zhu Xi’s view, while we cannot confuse human nature with human emotions, we cannot separate them from each other either: “when there is a particular kind of human nature, there will be a particular kind of human feeling; because there is a particular kind of human feeling, we know that there is a particular kind of human nature. Because today we see that there is such kind of human emotion, we know that originally

270 YONG HUANG

there is such kind of human nature” (Yulei 5, 81). He generalizes what Cheng Yi says about humanity and commiseration and applies it to other aspects of human nature and their corresponding emotions:

The reason we can examine principle is that we can examine it when it is aroused. Everything must have its origin and root. Although the principle of human nature is without shape, its aroused sprouts can be clearly exam-ined. Thus, from the emotion of commiseration, we know that there must be humanity; from the feeling of shame and dislike, we know that there must be rightness, from the feeling of reverence, we know that there must be propriety, and from the feeling of right and wrong, we know that there must be wisdom. If there is no principle within, how can there be sprouts without? Because there are sprouts without, we can tell that there must be principle within. (Wenji 58, “Reply 2 to Chen Qizhi,” in Zhu 1996, 2977)

So although we do not have direct perception of human nature, we can still be sure that human nature is virtuous, because there are good human emotions that we can perceive. In Zhu Xi’s view, human nature is the unaroused state of human emotions, while human emotions are the aroused state of human nature. Human nature is aroused by the things we perceive. For example, “seeing an infant about to fall into a well, one is aroused to have the feeling of alarm and commiseration, which mirrors humanity. Seeing someone stealing, one is aroused to have the feeling of shame and dislike, which mirrors rightness” (Yulei 53, 1152); similarly, passing by a temple or pilgrimage, one is aroused to have the feeling of reverence, which mirrors the human nature of propriety. In short, when things outside excite our sense organs, what is inside, the human nature, will be aroused, and the result is human emotions (see Wenji 58, “Reply 2 to Chen Qizhi,” in Zhu1996, 2977). Zhu Xi thus uses several analogies to explain our knowledge of the goodness of human nature, which we cannot perceive directly, from the goodness of human emotions, which we can and do perceive directly. In one place, he says that it “is just as we can tell that the origin of the water must be clear when we see that the water flowing from it is clear. The four sprouts are feelings, while human nature is principle. What arises is the emotion, while its root is human nature. [To know human nature from human emotions] is like knowing the shape [of a thing] from its shadows” (Yulei 5, 81). In another place, he says that to see human nature from human emotions “is like knowing how a mother must be from our knowledge of how her son is . . . . It is also like the bud of grasses and trees: from the buds above the ground we can tell that there must be roots underground. We do not say that the bud is already the root, nor can we know anything about the root independently from the bud” (Yulei 53, 1151).

In short, we can have indirect knowledge of the goodness of human nature from our direct knowledge of the goodness of human emotions, just as we can have indirect knowledge of the origin of the water from our direct knowledge of the water flowing from it, have indirect knowledge of a thing from our direct knowledge of its shadow, have indirect knowledge of the mother from our direct knowledge of her son, and


have indirect knowledge of the root from our direct knowledge of its sprouts. Thus, it is clear that Zhu Xi’s conception of human nature is neither morally indetermi-nate, as it clearly claims that human nature is virtuous, nor is it non-explanatory or merely subjective, as it is based on solid objective and empirical grounds. Here we may wonder how an objective conception of human nature, a description of what a human is, can be normative, telling us what we should be, since what we should be is based on what we actually are: we must already be what we ought to be. While Zhu Xi thinks that human nature is virtuous, he does acknowledge that human emo-tions can be either good or bad. Thus he claims that “when aroused, human nature is human emotions. There are good emotions as well as bad emotions, although human nature is completely good” (Yulei 5, 82). Thus when Zhu Xi claims that human nature is originally virtuous, he does not mean that all humans are actually already virtuous but that all humans can and ought to be virtuous.

However, this leads to a new question. If from good human emotions we can infer that the human nature from which those emotions arise must be good, can we not equally infer that human nature must be bad from those bad human emotions? The answer is negative. To use Zhu Xi’s analogy again, from the bud sprouting above the ground we can infer that there must be a root underground. However, simply because there is no bud sprouting above ground we cannot say that there must be no root underground; similarly, simply because the bud sprouting above ground is not good we cannot infer that its root underground must be bad. While a bad root, as bad, cannot have a good sprout if at all, a good root does not always have a good sprout: there may be other reasons for there being no bud or for the bud to become bad (the soil in which it grows is not good or it is not properly fertilized, for example). Of course, we cannot push this analogy too far, as it at least does not exclude the possibility that there is no root when there is no bud or the root is bad when the bud is bad. Yet, Zhu Xi categorically excludes the possibility that human nature can be bad. Does this mean that, given the fact that some human emotions are actually bad, and there are actually bad people, Zhu Xi’s claim that human nature is virtuous is after all a metaphysical claim lacking any objective ground?

The answer is again negative. This can be seen from the distinction Zhu Xi makes between bad human beings and animals. Although, as we have seen, he does claim that bad human beings as they are indeed are no different from beasts, Zhu Xi also claims that bad human beings as they ought to be are different from animals. We normally say that bad human beings ought to be virtuous, but we do not say that animals ought to be virtuous, since “ought” implies “can”: bad human beings as a matter of fact can be made virtuous, while animals as a matter of fact cannot. When a student asks why there is difference when everything has the same nature, Zhu Xi responds: “human nature can be either bright or obscured, the nature of non-human beings are all out of balance and blocked. The obscured human nature can be brightened, while the nature that is out of balance and blocked cannot be made clear” (Yulei 4, 51). In this passage, Zhu Xi not only makes distinctions be-tween human beings and non-human beings but also between virtuous and therefore

272 YONG HUANG

genuine human beings and evil and therefore nominal human beings. The difference between human beings and non-human beings is whether their nature is open or blocked. When it is blocked by unbalanced qi, it cannot be made open. That is why social animals cannot extend the four natural qualities, the same natural qualities that become human virtues when fully extended in human beings. The difference between good human beings and bad human beings is whether their nature is bright or dark. It is bright when it is not clouded by one’s selfish desires; it is dark when it is clouded by one’s selfish desires. In the latter case, the “human” being acts no differently from animals, as they have left unused the only unique human capacity, the ability to extend or expand (tui) the four natural qualities to their full scope. However, the identity between non-virtuous people and animals is only in terms of what they actually do and not in terms of what they can do. As Zhu Xi states in this passage, while blocked nature (in animals) cannot be opened, the nature darkened by selfish desires (in non-virtuous persons) can always be brightened. Thus immediately following the passage above, when a student says that “when deeply sinking into the habit of doing immoral things, a person can never return to the right road,” Zhu Xi disagrees, claiming that “everything depends upon the depth of knowledge and the amount of effort” (Yulei 4, 51–52). In other words, those who sink deep in immoral habits can still return to the right track, although they must make more effort than others. As long as they do as stated in one of the greatest Four Confucian Books Zhongyong (Centrality and Commonality): “while other people only need make one effort, they need to make one hundred; while other people only need make ten, they need make one thousand,” then “however benighted they are, people can still be enlightened” (Yulei 4, 59). That is the reason why Zhu Xi repeatedly comments on the following passage from the Great Learn-ing, another of the Four Confucian Books: “inferior persons, when not noticed by others, do all immoral things possible. However, when they see superior persons, they will immediately hide their immoral side and pretend to be moral”; in Zhu Xi’s view, their covering up their immoral behavior indeed shows that they are dishonest (see Yulei 15, 271); that their knowledge is not genuine (Yulei 16, 291); that they are self-deceptive (Yulei 16, 297); and that they are doubly wrong (to do immoral thing is wrong and to cover them up is also wrong) (Yulei 16, 298). But for Zhu Xi, this very fact also shows that even a vicious person also knows what the characteristic way of being human is: the virtuous way (Zhu Xi 1996, 1384). It is in this sense that they are different from beasts.

So when Zhu Xi claims that human nature is virtuous, he does not mean that every individual human being is actually already virtuous. There are people, ac-tually many people, who are not virtuous, and in this sense many people are not characteristically human beings. They are defective human beings. In this sense, while Zhu Xi disagrees with Hursthouse on what is characteristic of human be-ings, he would certainly agree with her when she says that “characteristic” is not a statistical notion. In other words, while it certainly does not mean all humans are going on in the characteristically human way, it also does not necessarily mean that


most human beings are going on in the characteristically human way. To determine whether a way is characteristic of the human species is to see what all humans can do (not actually do) that makes the human species different from other species. So, as Hursthouse states, the notion of “characteristic” is “avowedly normative and is clearly going to yield judgments to the effect that many human beings are not go-ing on ‘in the way characteristic of the species’ and are thereby defective human beings” (Hursthouse 1999, 223).30 It is in this sense that there is no contradiction at all between Zhu Xi’s claim that human nature is virtuous and the acknowledge-ment that some (and actually many) human beings are not virtuous. Zhu Xi only disagrees with Hursthouse on what is characteristic of the human species. While for Hursthouse, the way characteristic of human beings is rationality, for Zhu Xi it is tui, the ability to extend the four natural qualities, the very ability that makes these four natural qualities virtues in human beings.31 Thus, as far as the four natural qualities are not fully extended in a human being, this human being is more or less ethically defective, just as a person who does not have the vision characteristic of the human species is more or less physically defective. More importantly, for Zhu Xi, while a person who is physically defective to a certain degree may never be able to gain the physical ability characteristic of human beings, a person who is ethically defective, however serious it is, can always regain the ethical way characteristic of being human, as long as the person is willing to make a strong enough effort of tui,32 since the ethical way characteristic of human beings, while not a way lived by all or even most human beings, is a way that all human beings can have and ought to have in order to become characteristically human beings. This can also explain why we do not condemn a person for his/her physical defects as we do for his/her ethical defects.

Now once again it is time to take stock. I started this section asking whether Zhu Xi’s notion of human nature, in terms of which he explains virtue, when identified with virtue, merely re-expresses virtue, which it intends to explain, and lacks objective ground. If this is the case, it is caught by one of the horns of the second dilemma, although a horn different from the one by which the Aristotelian notion of human nature is caught. Our discussion in this section shows that this is not the case, since Zhu Xi’s metaphysical conception of human nature as virtuous is based on his empirical observations that, one the one hand, there are virtuous human emotions, and, on the other hand, those human beings who do not exhibit such virtuous human emotions can and ought to have such virtuous emotions, as they are characteristic of human beings, and one’s lack of it indicates one’s defec-tiveness as a human being. In short, Zhu Xi’s conception of human nature is both objective and normative, and therefore succeeds in avoiding the second dilemma.

VII. SUMMARY

In this article, I discuss the two dilemmas of virtue ethics resulting from its em-phasis on the primacy of virtue, as detected by Gary Watson. I argue that, due to the gap between rationality and virtue in Aristotle’s function argument, as pointed

274 YONG HUANG

out by Bernard Williams and John McDowell among others, which has not been adequately bridged by its contemporary defenders, Aristotle’s ethics, if regarded as a virtue ethics in contrast to deontology and consequentionism, suffers from these two dilemmas. I also argue that the sophisticated neo-Aristotelian ethical naturalism developed by Rosalind Hursthouse also fails to avoid these two dilemmas, despite her conscientious attempt to do so. I then turn to Zhu Xi’s neo-Confucianism. I argue that, Zhu Xi, just like Aristotelian philosophers, chooses to proceed with an account of virtue, instead of leaving it unexplained, and does it by appealing to a notion of human nature. However, his account of virtue in terms of human nature avoids the first dilemma, because it has a different conception of human nature. Instead of rationality, Zhu Xi thinks that human nature is virtuous. At the same time, his conception of human nature as virtuous, in terms of which virtue is explained, does not merely re-expresses virtues, as it is based on his objective observations of both the presence of virtuous emotions in human beings and their absence in animals, and thus it also succeeds avoiding the second dilemma.33

ENDNOTES

1. According to the most widely accepted view, there are three main stages of the develop-ment of Confucianism, the classical Confucianism in the Spring and Autumn (722–481 B.C.) and Warring States (403–222 B.C.) periods, neo-Confucianism in the dynasties of Song (906–1279), Yuan (1271–1368), and Ming (1368–1644) (sometimes also including Qing dynasty [1644–1912]), and contemporary Confucianism. Between classical Confucianism and neo-Confucianism, there is a long period of the Buddhist domination in Chinese history. Although not an initiator of the neo-Confucian revival, Zhu Xi has been acknowledged as the singularly most influential figure in this movement, bringing Confucianism back to the dominant position in Chinese society. His commentaries on the Confucian classics later became official textbooks for the imperial civil examinations.

2. Thus, Gary Watson points out that in character utilitarianism, “the value of the outcome of possessing and exercising certain traits is the ultimate standard of all other values. It shares with act utilitarianism the idea that the most fundamental notion is that of a good consequence or state of affairs, namely, human happiness” (Watson 1997, 61).

3. For other problems with such a reading of Kant’s maxims, see Louden 1997, 290–292.

4. Instead of maxims, Robert Louden focuses on Kant’s good will, which in his view, is “a state of character which becomes the basis for all of one’s actions”; from this, Louden infers that “what is fundamentally important in his [Kant’s] ethics is not acts but agents”; and in this sense Kantian ethics is also a virtue ethics, since “Kant defines virtue . . . as ‘fortitude in relation to the forces opposing a moral attitude of will in us.’ The Kantian virtuous agent is thus one who, because of his ‘fortitude’, is able to resist urges and inclinations opposed to the moral law” (Louden 1997, 289). However, Louden himself also acknowledges a problem in reading Kant as a virtue ethicist, for in Kant, “both the good will and virtue are defined in terms of obedience to moral law . . . . Since human virtue is defined in terms of conformity to law and the categorical imperative, it appears now that what is primary in Kantian ethics is not virtue for virtue’s sake but obedience to rules. Virtue is the heart of the


ethical for Kant . . . . But Kantian virtue is itself defined in terms of the supreme principle of morality” (Louden 1997, 290).

5. While this can still lead to a question about how it is the case that all virtuous characters enhance the well-being of self and/or others, to see this can help us avoid attributing the dilemma that Eve Garrard attributes to Slote: “If the claim is that the benevolent [virtuous] agent takes kindness [a virtue] as the reason for action, then she isn’t taking facts about welfare of others as her reason for acting, and it’s no longer clear in what sense she can be said to be aiming at the welfare of others as her goal. If, however, the claim is that the be-nevolent agent takes facts about the welfare of others as reasons for action, then these facts are what make the benevolent action right, and we have lost the dependence of the deontic on the aretaic” (Garrard 2000, 282).

6. It is thus no surprise to find that, when he is going to make a crucial point, Slote very often uses a parenthetical phrase: “check this on yourself” (see, for example, Slote 2001, 12, 35).

7. For this David Copp and David Sobel complain that “Slote must rely covertly on intuitions about right and wrong actions in order to figure out which states of character are admirable” (Copp and Sobel 2004, 519).

8. John Kultgen, for example, notices this irony: “Slote dismisses whole congeries of com-mon sense intuitions . . . . Despite this cavalier treatment of the intuitions of a great number of people, he claims that his view retains a firm grounding in common sense” (Kultgen 1998, 339).

9. One way to do it is perhaps to appeal to the holistic idea of reflective equilibrium: let these conflicting intuitions challenge each other until a reflective equilibrium obtains. This seems to be the way out for Slote.

10. Hursthouse tells us that “Bernard Williams . . . has pinpointed basically the same dilemma. If ‘good human being,’ as it figures in the modern naturalism project is, like ‘good wolf,’ a biological/ethological/scientific concept, then it is objectively all right, but it won’t yield anything much in the way of ethics; it will be largely morally indeterminate . . . . A concept of human nature formed within our ethical outlook may yield us quite a rich hoard—but then, of course, the rich hoard will not be objectively well-founded but the mere re-iteration of the views involved in the ethical outlook” (Hursthouse 2004, 165).

11. Thomas Hurka, for example, asks “the question of how distinctively virtue-ethical a theory is whose central explanatory property is in fact flourishing [eudaimonia] . . . . This ethics would not be at all distinctive if it took the virtues to contribute causally to flourishing, as productive means to a separately existing state of flourishing. But this is not the usual view” (Hurka 2001, 233).

12. For this and other reasons, Santas claims that “the widespread belief that Aristotle had a virtue ethics is false” (see Santas 1997, 281). Contemporary defenders of Aristotelian ethics as a virtue ethics may well respond that virtue in Aristotle is not merely instrumental to but also at least partially, or primarily, or even solely constitutive of human flourishing. However, even if this is indeed the case, Gary Watson rejects it as still an ethics of outcome, similar to ethical perfectionism, as it depends on a theory of ultimate good (see Watson 1997, 63). Watson himself does provide an account of virtue which he claims maintains the explana-tory primacy of virtue without falling back into an ethics of outcome, an account in which “the theory of ultimate good is dependent on the theory of virtue” (Watson 1997, 65). Yet,

276 YONG HUANG

in a lengthy note on whether Aristotle had an ethics of virtue in this sense, with a negative answer, he regrets that “it is somewhat disconcerting not to be able to adduce here a single clear instance of a historically important ethics in the sense I have identified” (Watson 1997, 71n26). As I will try to show in sections IV and V, such an instance can be found in Zhu Xi’s neo-Confucian ethics.

13. It is interesting to note that, while Watson himself believes that he has provided an ac-count of virtue without sacrificing its explanatory primacy so that the first dilemma can be avoided, he is more pessimistic about a solution to the second dilemma, which he discusses only as one of the more troublesome questions of virtue ethics, partially responsible for the resistance to the renewal of interest in Aristotelian ethics (see Watson 1997, 66–68).

14. In response to this challenge, Jiyuan Yu argues that “a good life for human beings must involve the exercise and manifestation of this characteristically human feature [rationality]. If one’s life is dominated by appetite, it ‘appears completely slavish’ and ‘is a life for graz-ing animals’ (NE, 1095b19–20)” (Yu 2007, 66). Appetite is a life without reason, but it is not necessarily against reason. It can and, in humans, should be guided by reason. A life of appetite regulated by reason is obviously not a life for grazing animals but a properly human life, a life of practical reason, although for Aristotle a properly human life is not merely a life of appetite regulated by reason; it also includes the intellectual life. If a life of appetite, broadly understood as bodily life, even when guided by reason, is still not properly human life, then hardly any morally virtuous life, as described by Aristotle, could be counted as a properly human life, as what a virtuous person does to others is, in most cases, to enhance the bodily lives of others. If so, then Aristotle needs to show that no lives of appetite regulated by reason are virtuous practical lives. This is precisely the gap that critics think Aristotle has not bridged.

15. In addition to the function argument, Nussbaum also discusses Aristotle’s argument for the political nature of human beings (see Nussbaum 1995, 102–110), by which Williams is equally unconvinced: “Glaucon and Adeimantus, agreeing that human beings are essentially or typically rational and they essentially or typically live in societies, could still deny that human reason is displayed at its most effective in living according to the restrictive require-ments of society” (Williams 1995, 199).

16. In this sense, MacIntyre may indeed be right that, for Aristotle, there is something “after virtue.” In other words, in order for a particular character trait to be a virtue, we need “some prior account of certain features of social and moral life in terms of which it has to be defined and explained” (MacIntyre 1984, 186). Since this prior account is nothing but the account of the good human life, “the concept of the good life for man is prior to the concept of a virtue” (MacIntyre 1984, 184).

17. This is also partly related to her understanding of ethical naturalism, which is “usually thought of as not only basing ethics in some way on considerations of human nature, but also taking human beings to be part of the natural, biological order of living things” (Hursthouse 1999, 206).

18. Hursthouse thus states: “We might say that the fifth end was the preparation of our souls for the life hereafter, or that it was contemplation—the good functioning of the theoretical intellect. But to adopt the first is to go beyond naturalism towards supernaturalism, and even philosophers have baulked at following Aristotle and endorsing the second” (Hursthouse 1999, 218).


19. For Toner, “it is striking that although Foot and Hursthouse carry out their projects in continual dialogue with McDowell, to my knowledge neither they nor [Michael] Thompson take up the criticisms McDowell offers in this article [‘Two Sorts of Naturalism’]” (Toner 2008, 226).

20. For this reason, it is indeed superfluous to add a fifth end as amended by Allen Thompson, according to whom, “just as good lions should develop their capacity for hunt, good human beings should develop their capacity for reason”; and so the development of a normatively autonomous capacity for practical reason is “a natural end of our natural kind . . . a fifth end for Hursthouse’s botanical and ethological model” (Thompson 2007, 260).

21. Frank Svensson makes a similar objection: “She may choose to give up on the idea that nature cannot be normative with respect to human beings [in which case it is no longer an ethical naturalism] . . . . Or Hursthouse could give up on the idea that four ends of social animals substantially constrain what we can reasonably hold to be a virtue in human beings. But in that case the supposed analogy between evaluation of plants and animals and moral judgments breaks down” (Svensson 2007, 199–201).

22. David Copp and David Sobel even challenge whether Hursthouse’s list of four ends is necessarily superior to competing lists that can be equally scientific, as “there are different ways of approaching scientific study of animal kinds . . . and we think there can be cor-respondingly different conceptions of what makes an animal a good instance of its kind” (Copp and Sobel 2004, 535). For example, they argue that evolutionary biologists, descrip-tive biologists working on the natural history of a species, and veterinarians concerned for animals in the way doctors are supposed to be concerned with humans may very well each provide a very different list of the ends and aspects of an animal from the one Hursthouse provides. However, I think we can at least accept that Hursthouse’s list is one of these objec-tive ones.

23. However, as Gowans points out, if other ends of human beings have analogues in non-human social animals and so can be regarded as natural, “it is hard to see what fact about human nature could play an analogous role in an argument for the claim that our ends should include the well-being of human beings generally” (Gowans 2008, 47).

24. In his essay “On Ren,” Zhu Xi states that “although virtues of the heart/mind are com-prehensive and thorough, with nothing lacking, to summarize it in one word, it is ren. Let me try to explain it. The heart/mind of heaven and earth has four virtues: originating, penetrating, harvesting, and correcting, and the virtue of originating is all inclusive; the movement of the heaven and earth is spring, summer, autumn, and winter, and the life-giving force of the spring is all inclusive. Thus human heart/mind also has four virtues, humanity, rightness, propriety, and wisdom, and humanity is all inclusive” (Zhu Xi 1996, 3545). But precisely in what sense does the virtue of humanity lead and include the other virtues? For Zhu Xi, since humanity is what makes a human being a human being, it “itself is the substance, while propriety is rules according to humanity, rightness is judgment [regarding concrete situations] according to humanity, and wisdom is the distinction [between right and wrong] according to humanity” (Yulei 6, 99).

25. Puzzled by this, a student asks how to explain the primacy of the virtue of humanity. Zhu Xi replies: “This is merely an analogy. But even though one hand does not include the four limbs, still when we talk about hand and foot, hand is before foot, and when we talk about left and right, left is before right” (Yulei 6, 100). So the left hand, which represents

278 YONG HUANG

the virtue of humanity, is still primary. (Note: in Chinese, we always say “hand and foot,” never “foot and hand,” and always “left and right” and never “right and left.”)

26. About the virtuous human nature, Zhu Xi says that “generally speaking, humans’ virtuous nature (dexing) naturally has these four: humanity . . . rightness . . . propriety . . . and wisdom” (Yulei 6, 100); and about virtues of human nature, he states that “humanities, rightness, propriety, and wisdom are all virtues of human nature (xing zhi de)” (Yulei 101, 2324).

27. One way to explain this apparent inconsistency is to think that Zhu Xi holds different views on this issue at different stages in his life, with the later view replacing his earlier views. Chen Lai, a respected Chinese expert on Zhu Xi, does precisely that (see Chen 2000, 124–143). However, this cannot be the case. As shown in the passage I am going to quote right after this, Zhu is clearly aware of the different views he has on this issue and explains, equally clearly, why he thinks they are not contradictory.

28. To avoid misunderstanding, Zhu Xi points out that “filial piety and brotherly love themselves are things one ought to do. They are not things one ought to do only in order to be humane to other people and love non-human beings” (Yulei 20, 414).

29. By this, of course, I do not mean that these two are separable from each other for Zhu Xi. What I mean is that a contemporary Western virtue ethicist does not have to accept Zhu Xi’s metaphysics in order to accept his description of the empirical fact/moral value, just as a non-Christian does not have to accept the Christian belief in God in order to accept its idea of forgiveness, even though for Christians these two are inseparable. As there is a naturalist turn in contemporary virtue ethics, I suspect that many virtue ethicists in the West today may feel content without being metaphysical. However, for those who feel an urge, justifiably to me, to have something similar to what Charles Taylor calls ontological articulation (Taylor 1989, 92), there is a need for the explanation of this empirical fact/moral value: why animals cannot but humans can and should extend their natural altruistic behaviors beyond their immediate groups. If they find Zhu Xi’s “ontological articulation” unacceptable, they may need to provide an alternative one. (As a matter of fact, not all contemporary virtue ethicists found Zhu Xi’s metaphysics appalling. For example, Kenneth Dorter, an Aristotelian virtue ethicist, accepts it as an appropriate explanation, claiming that, while the terminology Zhu Xi uses “sounds alien to modern ears, [ . . . ] the basic ideas are not unfamiliar” [Dorter 2009, 261].) So there can be multiple ontological articulations of the same empirical fact/moral value, each perhaps cherished as the only appropriate one by its respective advocates, forming something similar to John Rawls’s overlapping consensus.

30. This non-statistical notion of “characteristic” as both objective and normative is consistent with Michael Thompson’s conception of a form of life. In Thompson’s view, the form of life of a given species is not falsified by some (even a significant number of) counter examples of members in the species. Those members of the species that do not have the form of life ought to have it. Here, “what merely ‘ought to be’ in the individual we may say really ‘is’ in its form” (Thompson 1995, 295). This view is also consistent with the position David Wong develops on the issue of whether moral reason is internal or external. In contrast to both straightforward internalism and straightforward externalism, Wong adopts a mixed position, which affirms that “while reasons are external with respect to the motivations of the individual agent, they are internal with respect to human nature” (Wong 2006, 188). More precisely, moral reasons, especially in light of Zhu Xi’s view discussed here, are


always internal to human nature, but not always internal to individual human beings: they are internal to virtuous agents, although not to agents who are not yet virtuous.

31. To say that tui is the distinguishing mark of being human for Zhu Xi is not contradic-tory to his characterization of the uniqueness of human beings in terms of principle and vital force: the former is an empirical fact, while the latter is a metaphysical explanation of the fact.

32. For Zhu Xi, only sages do not need to make the effort of tui, but this is only because the four natural qualities in sages are naturally extended: “[the four natural qualities] naturally flow out of sages, nurturing all ten thousand things. But all other people need to make effort to extend [these four qualities] to benefit others” (Yulei 27, 622; see also his letter to Zhang Jingfu in Zhu 1996, 1316).

33. This article originates from a paper presented at the mini-conference on Neo-Confu-cianism and Moral Psychology at APA Pacific 2009 in Vancouver. I want to thank Justin Tiwald, the chair of the organizing committee of the mini-conference for his invitation, Thomas Hurka, the commentator on my paper, for his very helpful and detailed com-ments, and Stephen Angle, Kenneth Dorter, P. J. Ivanhoe, Michael Slote, Richard Stichler, Kwong-loi Shun, Yang Xiao, and many others for their comments at this and a few other occasions when an early version of this paper was presented. I also would like to thank the two anonymous reviewers for this journal for both their positive encouragement and critical comments.

BIBLIOGRAPHY

Adams, Robert Merrihew. 1976. “Motive Utilitarianism.” Journal of Philosophy 73: 467–481.Anscombe, Elizabeth. 1958. “Modern Moral Philosophy.” Philosophy 33: 1–19.Aristotle. 1963. Nicomachean Ethics. In Works of Aristotle, vol. ix. Oxford: Oxford Uni-

versity Press.Chen, Lai 陳來. 2000. A Study of Zhu Xi’s Philosophy 朱熹哲學研究. Shanghai: East China

Normal University.Cheng, Hao 程顥, and Cheng Yi 程頤. 2004. Collections of the Two Chengs 二程集.

Beijing: Zhonghua Shuju.Copp, David, and David Sobel. 2004. “Morality and Virtue: An Assessment of Some Recent

Work in Virtue Ethics.” Ethics 114: 514–554.Dorter, Kenneth. 2009. “Metaphysics and Morality in Neo-Confucianism and Greece:

Zhu Xi, Plato, Aristotle, and Plotinus.” Dao: A Journal of Comparative Philosophy 8: 255–276.

Garrard, Eve. 2000. “Slote on Virtue.” Analysis 60: 280–284.Gowans, Christopher W. 2008. “Virtue and Nature.” Social Philosophy and Policy 25: 28–55.Hooker, Brad. 2002. “The Collapse of Virtue Ethics.” Utilitas 14: 22–40.Hurka, Thomas. 2001. Virtue, Vice, and Value. Oxford: Oxford University Press.Hursthouse, Rosalind. 1999. On Virtue Ethics. Oxford: Oxford University Press.———. 2002. “Virtue Ethics vs. Rule-Consequentialism: A Reply to Brad Hooker.” Utilitas

14.1: 41–53.

280 YONG HUANG

———. 2004. “On the Grounding of the Virtue in Human Nature.” In What is Good for a Human Being?: Human Nature and Value, ed. Jan Szaif. Berlin and New York: Walter de Gruyter.

———. 2007. “Virtue Ethics.” Stanford Encyclopedia of Philosophy (online). http://plato.stanford.edu/entries/ethics-virtue. Accessed on March 1, 2009.

Kultgen, John. 1998. “The Vicissitudes of Common-Sense Virtue Ethics: Part I: From Aristotle to Slote.” The Journal of Value Inquiry 32: 325–341.

Louden, Robert B. 1997. “Kant’s Virtue Ethics.” In Virtue Ethics: A Critical Reader, ed. Daniel Statman. Washington, D.C.: Georgetown University Press.

MacIntyre, Alasdair. 1984. After Virtue. Notre Dame: University of Notre Dame Press.McDowell, John. 1998. Mind, Value, and Reality. Cambridge MA: Harvard University Press.Nussbaum, Martha. 1995. “Aristotle on Human Nature and the Foundation of Ethics.” In

World, Mind, and Ethics: Essays on the Ethical Philosophy of Bernard Williams, ed. J. E. J. Altham and Ross Harrison. Cambridge: Cambridge University Press.

O’Neill, Onora. 1989. “Kant after Virtue.” In her Constructions of Reason: Exploration of Kant’s Practical Philosophy. Cambridge: Cambridge University Press.

Santas, Gerasimos X. 1997. “Does Aristotle Have a Virtue Ethics.” In Virtue Ethics: A Critical Reader, ed. Daniel Statman. Washington, D.C. Georgetown University Press.

Slote, Michael. 1992. From Morals to Virtue. Oxford: Oxford University Press.———. 1997. “Virtue Ethics.” In Marcia W. Baron, Philip Pettit, and Michael Slote, Three

Methods of Ethics: A Debate. Malden MA: Blackwell Publishing.———. 2001. Morals from Motives. Oxford: Oxford University Press.Stateman, Daniel. 1997. “Introduction.” In Virtue Ethics: A Critical Reader, ed. Daniel

Stateman. Washington, D.C.: Georgetown University Press.Svensson, Frans. 2007. “Does Non-Cognitivism Rest on a Mistake?” Utilitas 19: 184–200.Swanton, Christine. 2003. Virtue Ethics: A Pluralist View. Oxford and New York: Oxford

University Press.Taylor, Charles. 1989. The Sources of the Self: The Making of the Modern Identity. Cam-

bridge MA: Harvard University Press.Thompson, Allen. 2007. “Reconciling Themes in Neo-Aristotelian Meta-Ethics.” The Journal

of Value Inquiry 41: 245–263.Thompson, Michael. 1995. “The Representation of Life.” In Virtues and Reasons, ed.

Rosalind Hursthouse, Gavin Lawrence, and Warren Quinn. Oxford: Clarendon Press.Toner, Christopher 2007. “Sorts of Naturalism: Requirements for a Successful Theory.”

Metaphilosophy 39: 220–250.Watson, Gary. 1997. “On the Primacy of Character.” In Virtue Ethics: A Critical Reader,

ed. Daniel Statman. Washington, D.C. Georgetown University Press.Williams, Bernard. 1971. Morality: An Introduction to Ethics. New York: Harper.———. 1985. Ethics and Limits of Philosophy. Cambridge MA: Harvard University Press.———. 1995. “Replies.” In World, Mind, and Ethics: Essays on the Ethical Philosophy

of Bernard Williams, ed. J. E. J. Altham and Ross Harrison. Cambridge: Cambridge University Press.

Wong, David B. 2006. Natural Moralities: A Defense of Pluralistic Relativism. Oxford and New York: Oxford University Press.


Yu, Jiyuan. 2007. The Ethics of Confucius and Aristotle: Mirrors of Virtue. New York and London: Routledge.

Zhu, Xi 朱熹. 1994. Collected Annotations of the Four Books 四書章句集注. Taibei 臺北: Da’an Chubanshe 大安出版社.

———. 1996. Collected Works of Zhu Xi 朱熹集. Chengdu 成都: Sichuan Jiaoyu Chubanshe 四川教育出版社.

———. 1997. Classified Sayings of Master Zhu 朱子語類. Wuhan 武漢: Yuelu Shushe 岳麓書社.

TWO DILEMMAS IN VIRTUE ETHICS AND HOW ZHU XI’S NEO-CONFUCIANISM AVOIDS THEM

Documents