First published Thu Jan 10, 2008; substantive revision Fri Sep 14, 2018

The tradition of Chinese ethical thought is centrally concerned with questions about how one ought to live: what goes into a worthwhile life, how to weigh duties toward family versus duties toward strangers, whether human nature is predisposed to be morally good or bad, how one ought to relate to the non-human world, the extent to which one ought to become involved in reforming the larger social and political structures of one’s society, and how one ought to conduct oneself when in a position of influence or power. The personal, social, and political are often intertwined in Chinese approaches to the subject. Anyone who wants to draw from the range of important traditions of thought on this subject needs to look seriously at the Chinese tradition. The canonical texts of that tradition have been memorized by schoolchildren in Asian societies for hundreds of years, and at the same time have served as objects of sophisticated and rigorous analysis by scholars and theoreticians rooted in widely variant traditions and approaches. This article will introduce ethical issues raised by some of the most influential texts in Confucianism, Mohism, Daoism, Legalism, and Chinese Buddhism.

In the Analects 13.18, the Governor of She tells Confucius of a Straight Body who reported his father to the authorities for stealing a sheep. Confucius (Kongzi, best known in the West under his latinized name, lived in the 6th and 5th century B.C.E) replies that in his village, uprightness lies in fathers and sons covering up for each other. In the Euthyphro, Socrates encounters Euthyphro (whose name can be translated as “Straight thinker”), reputed for his religious knowledge and on his way to bring charges against his father for murder. The conversation between Socrates and Euthyphro leads to a theoretical inquiry in which various proposed answers as to piety’s ousia (essence) are probed and ultimately found unsatisfactory, but in which no answer to the piety or impiety of Euthyphro’s action is given. The contrast between these two stories highlights one of the distinctive features of Chinese ethics in general: its respect for the practical problem. The practical problem discussed by Confucius and Socrates is arguably a universal one: the conflict between loyalty owed to a family member and duty to uphold public justice within the larger community. Confucius’s response is one dimension of a characteristically Chinese respect for the practical problem. The nature of the problem demands a practical response. However, another dimension of a reflective respect for the practical problem is to maintain a certain humility in the face of a really hard problem. It is to be skeptical that highly abstract theories will provide a response that is true to the complexities of that problem. A tradition exemplifying such respect will contain influential works that will not pretend to have resolved recurring tensions within the moral life such as those identified in the Analects and the Euthyphro.

Confucius gives an immediate practical answer in 13.18, but the reader and commentators have been left to weave together the various remarks about filiality (or as it is often called, “filial piety”) so as to present a rationale for that answer. These remarks quite often concern rather particular matters, as is the matter of turning in one’s father for stealing a sheep, and the implications for more general issues are ambiguous. Do fathers and sons cover up for each other on all occasions, no matter how serious, and if there is a cover-up, is there also an attempt to compensate the victim of the wrongdoing? The particularity of these passages is tied up with the emphasis on praxis. What is sought and what is discussed is often the answer to a particular practical problem, and the resulting particularity of the remarks invites multiple interpretations. The sayings often are presented as emerging from conversations between Confucius and his students or various personages with official positions, or among Confucius’s students. One passage (11.22) portrays Confucius as having tailored his advice according to the character of the particular student: he urges one student to ask father and elder brother for advice before practicing something he has learnt, while he urges the other to immediately practice; the reason is that the first has so much energy that he needs to be kept back, while the second is retiring and needs to be urged forward. With this passage in mind, we might then wonder whether the apparent tension between remarks made in connection with a concept is to be understood in terms of the differences between the individuals addressed or the context of the conversation.

All texts that have become canonical within a tradition, of course, are subject to multiple interpretations, but Chinese texts invite them. They invite them by articulating themes that stay relatively close to the pre-theoretical experience that gives rise to the practical problems of moral life (see Kupperman, 1999 on the role of experience in Chinese philosophy). The pre-theoretical is not experience that is a pure given or unconceptualized, nor is it necessarily experience that is universal in its significance and intelligibility across different traditions of thought and culture. This attention to pre-theoretical experience also leads to differences in format and discursive form: dialogues and stories are more suited for appealing to and evoking the kind of pre-theoretical experience that inspires parts of the text. By contrast, much Western philosophy has gone with Plato in taking the route of increasing abstraction from pre-theoretical experience.

The contrast is not meant to imply that Chinese philosophy fails to give rise to theoretical reflection. Theoretical reflection of great significance arises in the Mozi, Mencius, Hanfeizi, Xunzi, and Zhuangzi, but there is more frequent interplay between the theorizing and references to pre-theoretical experience. In Chinese texts there are suggestions for theorizing about this experience, but the suggestions often indicate several different and fruitful directions for theorizing to go further. These directions may seem incompatible, and they may or may not be so in the end, but the tensions between these directions are real. The result is a fruitful ambiguity that poses a problematic. Pre-theoretical experience poses a practical problem. Apparently incompatible solutions to problems are partially theorized in the text, but the apparent incompatibility is not removed. Much of the value of these texts lies in their leaving the tensions in place with enough theory given to stimulate thinking within a certain broadly defined approach. There is more than enough for the sophisticated theorist to try to interpret or to reconstruct a more defined position as an extension of that broadly defined approach. At the same time, the problematic is partly framed with the language of pre-theoretical experience in the form of dialogue and story, making the texts accessible to a much broader range of readers than is usually the case with philosophy texts. The following sections present some of the major kinds of problematic that appear in the major schools of Chinese ethical thought.

A common way to understand Confucian ethics is that it is a virtue ethic. For some scholars this will be an obvious, uncontroversial truth. For others, it is a misconstrual that imposes contentious Western assumptions on Confucianism about what it is to be a person and what an ethics should be about. In light of this controversy, it is important to specify the sense in which it is relatively uncontroversial to claim that virtues constitute a major focus of attention in these texts. Virtues in the relevant sense are qualities or traits that persons could have and that are appropriate objects of aspiration to realize. These virtues go into the conception of an ideal of a kind of person that one aspires to be. Given this rather broad sense of “virtue,” it is unobjectionable to say that Confucian texts discuss ethics primarily in terms of virtues and corresponding ideals of the person.

What makes the characterization of Confucianism as a virtue ethic controversial are more specific, narrower senses of “virtue” employed in Western philosophical theories. Tiwald (2018) distinguishes between something like the broad sense of virtue and a philosophical usage that confers on qualities or traits of character explanatory priority over right action and promoting good consequences. Virtue ethics in this sense is a competitor to rule deontological and consequentialist theories. There simply is not enough discussion in the Confucian texts, especially in the classical period, that is addressed to the kind of questions these Western theories seek to answer.

There are other narrower senses of “virtue” that are clearly mischaracterizations when applied to Confucian ethics. Virtues might be supposed to be qualities that people have or can have in isolation from others with whom they interact or from their communities, societies, or culture. Such atomistic virtues could make up ideals of the person that in turn can be specified or realized in social isolation. Further, virtues might be supposed to be identifiable through generalizations that hold true in every case, such that the ways these traits are concretely manifested in conduct do not vary across context or situation. Prominent and influential critics of the “virtue” characterization of Confucian ethics--Roger Ames (2011) and Robert Neville (2016)--seem to be supposing that the term is loaded with such controversial presuppositions.

As will become clear in subsequent discussion here, one can employ virtue language with the appropriate qualifiers and at the same time acknowledge much of what the critics claim as insights of Confucian ethics: e.g., that the process of realizing the virtues characteristically takes place in relationship to others--those to whom one has responsibilities as a son or daughter or mother or father, for example--and that it can be part of one’s very identity to be a particular person’s son or daughter, mother or father. It is part of the Confucian vision of a life befitting human beings that it is a life of relationships marked by mutual care and respect, that one achieves fullest personhood that way. One achieves this in a manner that is particular to one’s circumstances, including the particular others with whom one most interacts. None of this is inconsistent with virtue characterizations in the broad sense (for an alternative role-ethic characterization of Confucian ethics that incorporates these insights in a different way, see Ames, 2011).

The most frequently discussed ideal is that of the junzi. The Chinese word originally meant “prince’s son,” but in the Analects it refers to ethical nobility. The first English translations rendered it as “gentleman,” but Ames and Rosemont (1998) have usefully suggested “exemplary person.” Among the traits connected to ethical nobility are filiality, a respect for and dedication to the performance of traditional ritual forms of conduct, and the ability to judge what the right thing to do is in the given situation. These traits are virtues in the sense that they are necessary for following the dao, the way human beings ought to live their lives. As Yu (2007) points out, the dao plays the kind of role in ancient Chinese ethics that is analogous to the role played by eudaimonia or flourishing, in ancient Greek ethics. The junzi is the ethical exemplar with the virtues making it possible to follow the dao.

Besides the concepts of dao and junzi, the concept of ren is a unifying theme in the Analects. Before Confucius’s time, the concept of ren referred to the aristocracy of bloodlines, meaning something like the strong and handsome appearance of an aristocrat. But in the Analects the concept is of a moral excellence that anyone has the potential to achieve. Various translations have been given of ren. Many translations attempt to convey the idea of complete ethical virtue, connoting a comprehensive state of ethical excellence. In a number of places in the Analects the ren person is treated as equivalent to the junzi, indicating that ren has the meaning of complete or comprehensive moral excellence, lacking no particular virtue but having them all. However, ren in some places in the Analects is treated as one virtue among others such as wisdom and courage. In the narrower sense of being one virtue among others, it is explained in 12.22 in terms of caring for others. It is in light of these passages that other translators, such as D.C. Lau, 1970a, use ‘benevolence’ to translate ren. However, others have tried to more explicitly convey the sense of ‘ren’ in the comprehensive sense of all-encompassing moral virtue through use of the translation ‘Good’ or ‘Goodness’ (see Waley, 1938, 1989; Slingerland, 2003). It is possible that the sense of ren as particular virtue and the sense of comprehensive excellence are related in that attitudes such as care and respect for others may be a pervasive aspect of different forms of moral excellence, e.g., such attitudes may be expressed in ritual performance, as discussed below, or in right or appropriate action according to the context. But this suggestion is speculative, and because the very nature of ren remains so elusive, it shall be here referred to simply as‘ren’.

Why is the central virtue discussed in such an elusive fashion in the Analects? The answer may lie in the role that pre-theoretical experience plays in Chinese philosophy. Tan (2005) has pointed to the number and vividness of the persons in the Analects who serve as moral exemplars. She suggests that the text invites us to exercise our imaginations in envisioning what these people might have been like and what we ourselves might become in trying to emulate them. Use of the imagination, she points out, draws our attention to the particularities of virtue and engages our emotions and desires. Amy Olberding (2008, 2012) develops the notion of exemplarism into a Confucian epistemology, according to which we get much of our important knowledge by encountering the relevant objects or persons. Upon initial contact, we may have little general knowledge of the qualities that make them so compelling to us, but we are motivated to further investigate. Confucius treated as exemplars legendary figures from the early days of the Zhou dynasty, such as the Duke of Zhou and Kings Wu and Wen. Confucius served as an exemplar to his students, perhaps of the virtue of ren, though he never claimed the virtue for himself. Book Ten of the Analects displays what might appear to be an obsessive concern with the way Confucius greeted persons in everyday life, e.g., if he saw they were dressed in mourning dress, he would take on a solemn appearance or lean forward on the stanchion of his carriage. Such concern becomes much more comprehensible if Confucius is being treated as an exemplar of virtue from which the students are trying to learn. The focus of Book Ten and elsewhere in the Analects also suggests that the primary locus of virtue is to be found in how people treat each other in the fabric of everyday life and not in the dramatic moral dilemmas so much discussed in contemporary Western moral philosophy.

Analects 1.15 likens the project of cultivating one’s character to crafting something fine from raw material: cutting bone, carving a piece of horn, polishing or grinding a piece of jade. The chapter also stresses the importance of li (the rites, ritual) in this project. In the Analects ritual includes ceremonies of ancestor worship, the burial of parents, and the rules governing respectful and appropriate behavior between parents and children. Later the word came to cover a broad range of customs and practices that spelled out courteous and respectful behavior of many different kinds. Engaging in ritual, learning to perform it properly and with the right attitudes of respect while performing it, is to engage in a kind of cutting and carving and polishing and grinding of the self. One of the most distinctive marks of Confucian ethics is the centrality of ritual performance in the ethical cultivation of character. For example, while Aristotelian habituation generally corresponds to the Confucian cultivation of character, there is no comparable emphasis in Aristotle on the role of ritual performance in this process of character transformation. Yet Confucians will say that any complete description of self-cultivation must include a role for the culturally established customs that spell out what it means to express respect for another person in various social contexts. Just how that role is conceived in the Analects is one of the central interpretive puzzles concerning the Analects. The interpretive question of how li is central to self-cultivation is posed in particular about its relation to the chief virtue of ren.

In the Analects 3.3 the Master said, “A man who is not ren—what has he to do with ritual?” The implication is that ritual is a means of cultivating and expressing a ren that is already there, at least in a raw or unrefined state. This implication about the role of ritual is consistent with passages of the Analects in which Confucius shows flexibility on the question of whether to follow established ritual practice. 9.3 shows him accepting the contemporary practice of wearing a cheaper silk ceremonial cap rather than the traditional linen cap. 9.3 also shows Confucius rejecting the contemporary practice of bowing after one ascends the stairs leading up to the ruler’s dais, and maintaining the traditional practice of bowing before one ascends the stairs. The implication is that the contemporary practice expresses the wrong attitude toward the ruler—presumptuousness in assuming permission to ascend. 9.3 suggests that it is something like the right attitude that is cultivated and expressed by ritual. Kwong-loi Shun (1993) has called this kind of understanding of ritual the “instrumental” interpretation.

However, in other places of the Analects, ritual seems to take on a more central role in the achievement of ren. Indeed, it seems to be presented as the key. A very common translation of 12.1 has Confucius telling his favorite student Yan Hui that “Restraining yourself and returning to the rites constitutes ren. If for one day you managed to restrain yourself and return to the rites, in this way you could lead the entire world back to ren. The key to achieving ren lies within yourself—how could it come from others?” (translation from Slingerland, 2003, though see Li, 2007, for a different translation of the word wei usually translated as ‘constitutes’, with different implications for the question of the relation between li and ren). Such passages have given rise to the “definitionalist” interpretation, as Shun calls it, which makes li definitive of the whole of ren. Obviously the instrumental and definitional interpretations cannot both be true.

One possibility for resolving this tension is to construe Confucius’ remarks as directed towards a particular student and informed by his conception of what sort of advice that student needs to hear given his strengths and weaknesses. Perhaps Confucius believes that Yan Hui should be focusing on disciplining himself through observing the rites, but his advice should not be be taken as an intended generalization about the relationship between ren and li. Perhaps the remarks that suggest more of an instrumentalist construal of the relationship are similarly context and audience bound. Such interpretation, of course, leaves open the question of what that relationship is, or indeed, whether Confucius ever had in mind a generalization about the relationship that informed his remarks.

Some have argued that such serious conflicts within the text constitute reasons for thinking that the Analects is an accretive text, i.e., composed of layers added at different times by different people with conflicting views. To some extent, viewing the Analects as accretive is nothing new, but Bruce and A. Takeo Brooks (1998, 2000) have taken that view very far by identifying Book 4 (and only part of it, for that matter) as the most reflective of the historical Kongzi’s views, and the other books as stemming from Confucius’s students and members of his family. The different books, and, sometimes, individual passages within the books, represent different time periods, people, with different agendas who are responding to different conditions, and often putting forward incompatible strands of Confucianism. The Brooks suggest that the parts of the Analects most directly associated with the historical Confucius and his disciples are the parts that feature ren as the pre-eminent virtue and that de-emphasize the role of ritual. The parts that are due to another trend in Confucianism, headed by Confucius’s descendants, are the parts that elevate ritual as the key to ren. The Brooks’s theory of the Analects has drawn appreciation and disagreement (e.g., see Slingerland, 2000 for both). It threatens to dislodge the assumption that underlies the dominant mode of interpreting the Analects, which is that the text, or most of it, reflects the coherent thought of one person.

One response to this interpretive challenge is to acknowledge the real possibility that different sets of passages are the products of different thinkers, but also to hold that these different people, even if they have different pragmatic and political agendas (a factor that the Brooks tend to emphasize), might also have had different and philosophically substantial perspectives on common problems. One of those problems might indeed have been the relation between ren and li, and at least part of the explanation of why different and potentially conflicting things are said about that relation is that the relation is a difficult one to figure out and that different thinkers addressing that common problem might reasonably have arrived at different things to say. Whether these different things are ultimately irreconcilable remains an open question. One might take a constructive attitude to these differences, ask what good philosophical reasons could motivate the different approaches, and ask whether there is a way of reconciling what all the good reasons entail.

Kwong-loi Shun’s approach exemplifies such a reconciling strategy. He holds that on the one hand, a particular set of ritual forms are the conventions that a community has evolved, and without such forms attitudes such as respect or reverence cannot be made intelligible or expressed (the truth behind the definitionalist interpretation). In this sense, li constitutes ren within or for a given community. On the other hand, different communities may have different conventions that express respect or reverence, and moreover any given community may revise its conventions in piecemeal though not wholesale fashion (the truth behind the instrumentalist interpretation).

Chenyang Li (2007) proposes a different approach based on a different reading of the word ‘wei’ used in 12.1 and often translated as ‘constitutes’ to render the crucial line, “Restraining yourself and returning to the rites constitutes ren.” Li notes that a common meaning of the word is ‘make’ or ‘result in.’ The relation between li and ren need not be construed as either definitional or constitutive, nor need it be construed as purely instrumental. Li proposes that li functions something like a cultural grammar where ren is like mastery of the culture. Mastery of a language entails mastery of its grammar but not vice versa.

Both Shun and Li are striving to capture a way in which ren does not reduce to li but also a way in which li is more than purely instrumental to the realization of ren. There are good philosophical reasons for this move. Consider the reasons for resisting the reduction of ren to li. As indicated above, 9.3 suggests that the attitudes of respect and reverence that are expressed by ritual forms are not reducible to any particular set of such forms, and Shun has a point in arguing that such attitudes could be expressed by different sets of such forms as established by different communities. In studying the cultures of other communities, we recognize that certain customs are meant to signify respect, even if we do not share these customs, just as we recognize that something that does not signify disrespect in our culture does indeed so signify in another culture. The fact that we can distinguish the attitude from the ritual forms that we use to express them allows us to consider alternative ritual forms that could express the same attitude. Ceremonial caps that are made of more economical material are acceptable, perhaps, because wearing such caps rather than the material ones need not affect the spirit of the ceremony. By contrast, bowing after one ascends the stairs constitutes an unacceptable change in attitude. To maintain that particular ritual forms do not define the respect and reverence they are intended to express is not to underestimate their importance for cultivating and strengthening these attitudes. Acting in ways that express respect given the conventionally established meanings of accepted ritual forms helps to strengthen the agent’s disposition to have respect. The ethical development of character does involve strengthening some emotional dispositions over others. We strengthen dispositions by acting on them. By providing conventionally established, symbolic ways to express respect for others, ritual forms give participants ways to act on and therefore to strengthen the right dispositions. The cultivating function of observing ritual highlights the distinctive practical focus of Confucian ethics. It is every bit as concerned with how to acquire the right sort of character as it is with what the right sort is.

On the other hand, there is good reason to resist the reduction of li simply to the role of expressing and cultivating a set of attitudes and emotional dispositions. In his influential interpretation (1972) of the Analects, Herbert Fingarette construes ritual performance as an end in itself, as beautiful and dignified, open and shared participation in ceremonies that celebrate human community. Ritual performance, internalized so that it becomes second nature, such that it is gracefully and spontaneously performed, is a crucial constituent of a fully realized human life. There are nonconventional dimensions of what it is to show respect, such as providing food for one’s parents (see Analects 2.7), but the particular way the agent does this will be deeply influenced by custom. Indeed, custom specifies what is a respectful way of serving food. On the Confucian view, doing so in a graceful and whole-hearted fashion as spelled out by the customs of one’s community is part of what it is to live a fully human life.

Ritual constitutes an important part of what ren is, and hence it is not merely an instrument for refining the substance of ren. At the same time it is not the whole of ren. Consider that part of ren that involves attitudinal dispositions. Attitude is not reducible to ritual form even if acting on that form can cultivate and sustain attitude. Moreover, 7.30 emphasizes the connection between desire for ren and its achievement (“If I simply desire ren, I find that it is already there”). The achievement of ren is of course a difficult and long journey, and so 7.30 implies that coming to truly desire it lies at the heart of that achievement. The multifaceted nature of ren emerges in Book 12, where Confucius is portrayed as giving different descriptions of ren. In 12.1, as already noted, he says that ritual makes for ren. But then in 12.2, he says that ren involves comporting oneself in public as if one were receiving an important guest and in the management of the common people behaving as if one were overseeing a great sacrifice (the duty to be respectful toward others). 12.2 also associates ren with shu or “sympathetic understanding,” not imposing on others what you yourself do not desire. Here the emphasis is not so much on ritual or not exclusively anyway, but on the attitudes one displays toward others, and on the ability to understand what others want or do not want based on projecting oneself into their situation. In 12.3, when asked about ren, Confucius says that ren people are hesitant to speak (suggesting that such people take extreme care not to have their words exceed their actions). And then in 12.22, when asked about ren, Confucius says that it is to care for people. Such diverse characterizations are appropriate if ren is complete ethical virtue or comprehensive excellence that includes many dimensions, including but not reducing to the kinds of excellence associated with li.

If we take the relevant passages on li and ren as forming a whole in which a coherent view is embedded, there is a pretty good case for regarding the observance of ritual propriety as a constituent of ren as well as crucial for instrumentally realizing some other dimensions of ren. But it does not exhaust the substance of ren. If the text is as radically accretive as the Brooks maintain, then the proposed construal of the relation is more of a reconstruction of what the best philosophical position might be on the nature of the relation. The reconstructive possibility should not be disturbing as long as we recognize it for what it is. Thinkers within a complex and vigorous tradition frequently re-interpret, expand, develop, revise, and even reject some of what one has inherited from the past. The fact that the Analects itself might be a product of this kind of engagement might usefully be taken as encouragement for its present students to engage with the text in the same way.

The Confucian position on the importance of li in ethical cultivation is interesting and distinctive in its own right, and this is partly because Confucianism hews close to a kind of pre-theoretical experience of the moral life that might otherwise get obscured by a more purely theoretical approach to ethics. If we look at everyday experience of the moral life, we see that much of the substance of ethically significant attitudes such as respect is in fact given by cultural norms and practices, and learning a morality must involve learning these norms and practices. Children learn what their behavior means to others, and what it should mean, by learning how to greet each other, make requests, and answer requests, all in a respectful manner. Much of our everyday experience of moral socialization lies in the absorption of or teaching to others of customs that are conventionally established to mean respect, gratitude, and other ethically significant attitudes (see Olberding, 2016). So construed, Confucian ethics provides an alternative to understanding the nature of the moral life that is different from an understanding that is primarily based on abstract principles, even abstract principles that require respect for each person. This is why there is significant resonance between Confucianism and communitarian philosophies such as those defended by Alasdair MacIntyre (1984, 1989) and Michael Walzer (1983). One of the distinctive marks of communitarianism is the theme that much of the substance of a morality is given not in abstract principles of the sort typically defended in modern Western philosophy but in a society’s specific customs and practices. In the Analects, the ambiguous relation between ren and li poses the problematic of how we are to understand the relation between cultural norms and practices on the one hand and that part of morality that appears to transcend any particular set of norms and practices. The Analects suggests a large role for culture, but on the reading suggested here, not a definitional role. There is much room for theoretical elaboration on the nature of that role.

Furthermore, in understanding why Confucians take a life of ritual practice to be partly constitutive of a fully human life, one must understand the aesthetic dimension of their notion of a fully human life. Such a life is lived as a beautiful and graceful coordinated interaction with others according to conventionally established forms that express mutual respect. A good part of the value attached to the fully human life lies in the aesthetic dimensions of a “dance” (Ihara, 2004) one performs with others. To better understand why the moral and the aesthetic cannot be cleanly separated in Confucian ethics, consider that a graceful and whole-hearted expression of respect can be beautiful precisely because it reflects the extent that the agent has made this moral attitude part of her second nature. The beauty has a moral dimension. Both these themes—the importance of contextualized moral judgment and aesthetic value of human interaction according to custom and tradition—offer opportunities for practitioners of, say, Anglo-American moral philosophy to reflect on what their approaches to the moral life might miss.

Consider ren in its meaning as the particular virtue of caring for others and li in its aspect as the valued human dance. These values are the basis for characterizing Confucian ethics as a relational ethic, meaning that it is in part distinguished by its placement of relationships at the center of a well-lived life (see Ames, 2011). Confucian ethics are often taken to stand in contrast to ethics that place individual autonomy and freedom to choose how to live. While there is much that is true about this contrast, it must be carefully described so as to differentiate it from some other contrasts. For example, the value of individual autonomy usually includes several different dimensions that do not necessarily accompany one another: (1) prioritizing of individual interests over group or collective interests when these conflict; (2) giving moral permission to the individual to choose from a significantly wide range (within certain moral boundaries) of ways to live; and (3) emphasizing the importance of living according to one’s own understanding of what is right and good even if others do not see it the same way.

Confucian ethics in significant part, though not in all parts, accepts autonomy in the sense of (3) (see Shun, 2004; and Brindley, 2010). Confucius is often depicted in the Analects as emphasizing the importance of cultivating one’s own character even when others do not recognize or appreciate one’s efforts (e.g., 4.14) and of acting independently of what is conventionally approved or disapproved (e.g., 5.1). The texts associated with Mencius (Mengzi, best known in the West under his Latinized name, lived in the 4th century B.C.E.) and Xunzi (4th and 3rd centuries B.C.E.), the most pivotal thinkers in the classical Confucian tradition after Confucius, both articulate the necessity to speak up when one believes the ruler one is serving is on a wrong course of action (e.g., Mencius 1A3 and Xunzi 29.2). On the other hand, none of these classical thinkers argue for the necessity of protecting a frank subordinate from a ruler who is made angry by criticism, and it could be argued that Confucianism does not fully endorse autonomy in sense (3) without endorsing such protection for those who wish to engage in moral criticism of the powerful.

Most interpretations present Confucian ethics as rejecting (2). There is a way for human beings to live, a comprehensive human good to be realized, and there can be no choosing between significantly different ways of life that are equally acceptable from a moral perspective (an important exception to this kind of interpretation is provided by Hall and Ames, 1987, who interpret Confucius’s dao as a human invention, collective and individual). On the other hand, Confucian ethics de-emphasizes legal coercion as a method for guiding people along the way and instead an puts the emphasis on moral exhortation and inspiration by way of example (see, most famously, 2.3 of the Analects, which emphasizes the necessity of a ruler’s guiding his people by instilling in them a sense of shame rather than by the threat of external punishment). While a Confucian might believe in a single correct way for human beings, she might endorse a significant degree of latitude for people to learn from their own mistakes and by way of example from others (see Chan, 1999).

Confucian ethics does not accept (1), but not because it subordinates individual interests to group or collective interests (for criticism of the rather common interpretation of Confucianism as prioritizing the group over the individual, see Hall and Ames 1998). Rather, there is a different conception of the relationship between individual and group interests. The best illustration of this different conception is a story to be found in the Mencius that concerns sage-king Shun. When Shun wanted to marry, he knew that his father, influenced by his stepmother, would not allow him to marry. In this difficult situation, Shun decided to marry without telling his father, even though he is renowned for his filiality. Mencius in fact defends the filiality of Shun’s act in 5A2. He observes that Shun knew that he would not have been allowed to marry if he told his father. This would have resulted in bitterness toward his parents, and that is why he did not tell them. The implication of this version of Shun’s reason is that filiality means preserving an emotionally viable relationship with one’s parents, and in the case at hand Shun judged that it would have been worse for the relationship to have asked permission to marry. The conception of the relation between individual and group interests embodied in this story is not one of subordination of one to the other but about the mutual dependence between the individual and the group. The individual depends on the group and must make the group’s interests part of his or her own interests, but, on the other side of the equation, the group depends on the individual and must make that individual’s interests part of the group’s interests. Shun’s welfare depends on his family and therefore must make his family’s interests part of his own (he resolves to do what is necessary to preserve his relationship to his parents), but his family’s welfare depends on Shun, and therefore it must recognize his interests to constitute part of its welfare (the family must recognize that it is damaging itself in requiring Shun to deny himself the most part important of human relationships).

The way that Confucianism conceives of the relationship between the individual and the group, as well as the way it is typically misconceived, is reflected in its notion of harmony or he. A typical misconception of harmony as a Confucian value is that it involves agreement and conformity with the views of others. In Analects 13.23, however, Confucius says that the junzi pursues harmony rather than sameness, while the small person does the opposite. The pre-Confucian thinker Yan Ying expresses a similar idea about harmony in likening it to a soup made with meat or fish, the strong flavor of which must be balanced and complemented by other ingredients such as vinegar, sauce, salt, and plum (Zuozhuan, Duke Zhao, 20th year; see Legge 1960 translation). The metaphor is meant to convey the idea that a ruler will not seek only ministers who agree with him but will seek to reconcile a diversity of viewpoints from his ministers. Reconciliation not only involves acceptance of difference but also tension and conflict (Li 2014) that are brought into a productive equilibrium. Moreover, that equilibrium is dynamic and is continuously created and re-created.

Along with the emphasis on li, the centrality of filiality is one of the most distinctive characteristics of Confucian ethics. The Analects 2.6 says to give parents no cause for anxiety other than illness, whereas 2.7, as mentioned earlier, emphasizes the need for the material support of parents to be carried out in a respectful manner. 2.8 emphasizes that it is the expression on one’s face that is filial and not just taking on the burden of work or letting elders partake of the wine and food before others.

Is obedience to parents always required of the filial child? What if the child believes that parents are wrong and their wishes run contrary to what is right or to ren? In those cases where one thinks them wrong, what is one to do? The Analects 2.5 portrays Confucius as saying, “Do not disobey,” but when queried further as to his meaning, he explains obedience in terms of conformance to the rites for burying and sacrificing to deceased parents. In 4.18 Confucius says that when one disagrees with one’s parents, one should remonstrate with them gently. Most translations of what follows have Confucius concluding that if parents are not persuaded, one should not oppose them (e.g., Lau, 1979; Slingerland, 2003; Waley, 1938), but it is possible to read the spare and ambiguously worded passage as requiring instead that one not abandon one’s purpose in respectfully trying to change one’s parents’ minds (Legge, 1971; see Huang, 2013, 133-37 for a survey of the different interpretations and an argument for the persistence-in-remonstrating translation). In other Confucian texts, the question of whether obedience is required has received different answers in the Confucian tradition. Chapters 1 and 2 of the Record of Ritual (Legge, 1967, vol. 1) say that one must obey if one fails to persuade one’s parent. On the other hand, Xunzi declares that following the requirements of morality rather than the wishes of one’s father is part of the highest standard of conduct (29.1 of the Xunzi; for a translation see Knoblock, 1988–94; or Hutton 2014) and moreover that if following the course of action mandated by one’s father would bring disgrace to the family and not following it would bring honor, then not following is to act morally (29.2 of the Xunzi). Xunzi’s position is supported in part by the distinction between service to parents and obedience to them. It might very well fail to be of service to parents if following their wishes is to bring moral disgrace to them and the family.

Another ethical issue arising from the strong Confucian emphasis on filiality concerns possible conflicts between loyalty to parents and loyalty to the ruler or public justice. Consider again Analects 13.18, in which Confucius says that uprightness is found in sons and fathers covering up for each other. In this case, at least, loyalty to parents or to children takes precedence over loyalty to ruler or to public justice. This precedence is one implication of the Confucian doctrine of care with distinctions (“care with distinctions” is the usual translation, but perhaps “care with distinctions” is less misleading because it covers both the emotionally freighted attitude toward kin and a more distanced attitude toward strangers). Though all people are owed moral concern, some are owed more than others, according to the agent’s relationship to them.

To introduce other kinds of problematic treated by Confucian thinkers, it is necessary to identify a pivotal critic of Confucianism in the classical period. Mozi (probably 5th century B.C.E), who possibly was once a student of Confucianism, came to reject that teaching, partly on the grounds that the Confucian emphasis on ritual and musical performance was a wasteful expenditure of resources that could otherwise be used to meet the basic needs of the many (Mozi, chapters 25, 32; see Watson, 1967 for a translation). A related criticism in the text of the Mozi is that tradition does not hold normative authority simply because it is tradition, for there was a time when the practice in question was not tradition but new (chapter 39). If a practice has no authority when it is new, it has no authority at any subsequent time simply because it is getting older.

Mozi also argued that exclusive concern for one’s own (oneself, one’s family, one’s state) is at the root of all destructive conflict (chapter 16). Exclusive concern for the self causes the strong to rob the weak. Exclusive concern for one’s family causes great families to wreak havoc on lesser families (it is not difficult to see how this thought might apply to the idea of protecting one’s own, even if they have committed serious crimes against others outside the family). Exclusive concern toward one’s state causes great states to attack small states. Mozi advocated the doctrine of jian ai as a remedy. “Ai” usually means “love” or “affection,” but for Mozi it probably meant an emotionally cooler form of concern. “Jian” usually means “inclusive.” One possible translation of “jian ai” is “inclusive care” (see Fraser, 2016). Other translations render “jian” as “impartial,” as in “impartial care” (see Ivanhoe 2005). The latter translation conveys clearly the understanding that Confucians, at least since Mencius, have had of jian ai, which is that it requires one to have equal concern for everyone regardless of one’s relationship to them. The next section explains how Mencius incorporated this understanding of jian ai in his criticism of it and in his development of Confucian ethics.

The substantial following that Mohism gained in the classical period forced a response from Confucians (see Hansen, 1992, and Van Norden 2007, for a discussion of Mozi’s pivotal impact on the Chinese tradition). They responded on two subjects: first, they had to address what is required by way of concern for all people and how to reconcile such concern with the greater concern for some that the Confucian doctrine of care with distinctions requires; second, they had to address the question of what kinds of concern are motivationally possible for human beings, partly in response to the Mohist argument that it is not difficult to act on jian ai (which they came to interpret as being contrary to care with distinctions), and partly in response to others who were skeptical about the possibility of acting on any kind of genuinely other-regarding concern. Mencius, in the text purporting to be a record of his teachings, explicitly sets himself to the task of defending Confucianism not only against Mohism but the teachings of Yang Zhu. Yang’s teachings seemed to Mencius to sit on the opposite end of the spectrum from Mohism (there is no surviving text purporting to articulate and defend Yangism). According to Mencius’s characterization, Yang Zhu criticized both Mohism and Confucianism for asking people to sacrifice themselves for others. Yang Zhu on this view was an ethical egoist: i.e., one who holds that it is always right to promote one’s own welfare. Mencius positioned Confucianism as the occupying the correct mean between the extremes of having concern only for oneself on the one hand and having an equal degree of concern for everyone.

Mencius 1A7 purports to be an account of a conversation between Mencius and King Xuan, the ruler of a Chinese state. Mencius is attempting to persuade the king to adopt the Confucian dao or way of ruling. The king wonders whether he really can be the kind of king Mencius is advocating, and Mencius replies by asking whether the following story he has heard about the king is true. The story is that the king saw an ox being led to slaughter for a ritual sacrifice. The king decided to spare the ox and substituted a lamb for the ritual sacrifice. Thinking back on that occasion, the king recalls that it was the look in the ox’s eyes, like that of an innocent man being led to execution, that led him to substitute the lamb. Mencius then comments that this story demonstrates the king’s capability to become a true king, and that all he has to do is to extend the sort of compassion he showed the ox to his own people. If he can care for an ox, he can care for his subjects. To say that he can care for an ox but not for his people is like saying “my strength is sufficient to lift heavy weight, but not enough to lift a feather” (translation adapted from Lau, 1970a) His failure to act on behalf of his people is due simply to his not acting, not to an inability to act. What the king has to do, suggests Mencius, is to treat the aged in his family as aged, and then extend it to the aged in other families; treat his young ones as young ones, and extend it to the young ones of others; then you can turn the whole world in the palm of his hand.

The passage demonstrates one characteristic of the text that is pertinent to Mencius’s response to Mohism. In contrast to the Analects, the ruler’s duties to care for his people are more frequently discussed and play a more prominent role in the conception of a ruler’s moral excellence. Mencius is portrayed in this text as very much engaged in getting the kings of Chinese states to stop mistreating their subjects, to stop drafting their subjects into their wars of territorial expansion, and to avoid overtaxing them to finance their wars and lavish projects. At the same time, Mencius’s assertion that the king is able to extend the kind of concern he showed the ox toward his own people is a reply to those who advocate Yangism on the grounds that acting for one’s own sake is natural. Mencius holds that natural compassion is a part of human nature. The task of moral self-cultivation is the task of “extending” what is natural. What is natural, or at least more so, is properly acting toward the aged and the young in one’s family and then extending that to the aged and the young in other families.

Extension is necessary because natural compassion is uneven compared to where it ought to extend. King Xuan may find it natural to have compassion for an innocent man about to be executed or a terrified ox about to be slaughtered, but not toward all his subjects when he is focusing on the benefits that a war of territorial expansion might bring him. This story of Mencius, the King, and the ox is rich material for reflection on the nature of moral development. It seems plausible that development must begin with something that is of the right nature to be shaped into the moral virtues, and also plausible that what we begin with is not as it fully should be. The questions posed by the story is what the natural basis of morality is and how further development occurs. Mencius’s theory of the “four duan” addresses these questions. “Duan” literally means “tip of something” and is often translated as “beginnings” in this context.

What are the four beginnings of morality? In 2A6 human nature (ren xing) it is said that no person is devoid of a heart (the word for heart in Chinese stands for the seat of thinking and feeling, hence often translated as “the mind”) sensitive to the suffering of others, and to illustrate this beginning, Mencius asks us to suppose that a man were suddenly to see a young child about to fall into a well. Such a man would certainly be moved to compassion, not because he wanted to get in the good graces of the parents, nor because he wished to win the praise of his fellow villagers or friends, nor yet because he disliked the cry of the child. This natural compassion can develop into the virtue of ren (in Mencius, ren is more often a particular virtue that concerns caring and hence is often translated as “benevolence”). A second beginning is the heart that feels shame in certain situations, e.g., in 6A10, Mencius says that if rice and soup are offered after being trampled upon, even a beggar would disdain them. Under the right conditions, innate shame develops into the virtue of yi or righteousness—being able to do the right thing. The third beginning is the heart that feels courtesy,e.g., the younger sometimes instinctively knows to respect and be courteous to the older. Under the right conditions, courtesy develops into li, which as a virtue consists in the observance of the rites or the virtue of ritual propriety. And finally, there is the heart that has a sense of right and wrong (shi/fei, the thing to do or not to do). Under the right conditions, this sense of approval and disapproval develops into wisdom, which includes having a grasp of the spirit behind moral rules so that one knows how to be flexible in applying them.

It is important to note that Mencian beginnings of morality are not just blind feelings or primitive urges to act in certain ways, but contain within them certain intuitive judgments about what is right and wrong, what is to be disdained and what is deferential, respectful behavior. In the example of the beggar who does not accept food that has been trampled upon, it seems that Mencius is suggesting we have an original, unlearned sense that allows us to judge the sort of respect that is due to ourselves as human beings. Similarly, in suggesting that we have an unlearned sense of deference, Mencius is suggesting that we have an unlearned sense of what is due to others such as elders and our parents. Mencius’s theory tallies with some of the more recent theories of emotion that point toward the intertwining of cognitive and affective dimensions (the theory does not necessarily imply, however, that the affective amounts to nothing more than the cognitive, as shall be discussed later).

The Mencius contains different metaphors that convey a view of human nature as the basis for moral development. On one metaphor, used in a debate with rival philosopher Gaozi in 6A2, the inborn goodness of human nature is like the tendency of water to flow downward. The metaphor implies that human beings develop virtues in the absence of abnormal interference such as water being damned up or struck so that it splashes upward. On the other way of conceiving ethical development, the four beginnings are more like barley sprouts that need nurture analogous to sun, water, and fertile soil (6A7). That these two conceptions are significantly different can be seen through the recognition that “growing” conditions for the sprouts are not necessarily provided in the normal course of affairs (see Wong, 2015b).

In some passages, extension is characterized as a matter of simply preserving or not losing what is given to one at birth (4B12, 4B19, 4B28, 6A10, 6A11), and such passages accord with the water metaphor in suggesting that moral development happens in the absence of abnormal interference. In other places, the thinking seems to be more in accord with the sprout metaphor and identifies conditions for moral development that go well beyond noninterference: kings are held responsible for providing for their subjects a constant means of livelihood (1A7) that enables them to support parents and nurture wife and children; kings must also ensure the appropriate moral education about filiality, about the duties that rulers and subjects owe to each other and about respect for the elder. Mencius furthermore recognized natural predispositions other than the four beginnings that could potentially lead human beings astray. He mentions the desires of the senses in this regard (6A15). This is why Mencius places responsibility on everyone to si (reflect on, turn over in one’s mind) (6A14, 6A15) the manifestations of the four beginnings. With such reflection, human beings can recognize that virtue takes precedence over satisfaction of potentially conflicting desires and feelings (e.g., the priority of righteousness over the desire for life if one cannot have both), but lack of reflection will stunt moral development (6A9). If the deprivation of nourishing conditions is severe enough, the sprouts can be killed off (6A8). Thus, while Mencius is often characterized superficially by his saying that human nature is good (6A6), he means (at least when his thinking is guided by the sprout metaphor) that it contains predispositions to feel and act in morally appropriate ways and to make intuitive normative judgments that can with the right nurturing conditions give human beings guidance as to the proper emphasis to be given to the desires of the senses (see Shun, 1997; Van Norden, 2004, 2007).

It is not surprising that there should be the kind of ambiguity expressed by the juxtaposition of the water and sprout metaphors in Mencius. A very common contemporary conception of the innate comes very close to the implications of the water metaphor, i.e., that which develops under normal conditions. On the other hand, we are also capable of recognizing that other things develop under a narrower or much more contingent (not necessarily realized in the normal course of affairs) set of conditions. A barley sprout develops only if human beings plant it in the right kind of soil and put effort into cultivating it. Yet it seems intuitively correct to say that its direction of growth is innate. If the conditions for growth are realized, it will become a barley plant, not a corn plant. Contemporary thinking about the innate bases of morality also shows this range of thinking. Claims that morality is constrained by an innate universal grammar (e.g., Hauser, 2008; Mikhail, 2011) seem closer to the idea that the moral (or its underlying universal structure) develops under normal conditions; other conceptions acknowledge more of a role for contingent factors (Nichols, 2004; Haidt and Bjorklund 2008). The ambiguity in Mencius’ thought, then, anticipates contemporary swings in thinking about the relative roles of what human beings are born with and what they acquire through learning, experience and culture.

Much of what is fascinating in Mencius lies in his explorations of how moral learning takes place and how this learning might also interact with emotion. Consider now in combination the theme that the cognitive and affective go into the constitution of emotion and the theme that the emotional beginnings of morality can be extended through provision of the right kind of nurture. What is necessary for extension? Is cognitive extension, i.e., more moral knowledge, sufficient? The answer to this question depends on the nature of the intertwining between the cognitive and affective in emotions. Consider again the story of King Xuan and the ox. Mencius expresses confidence in King Xuan’s ability to have compassion for his people, based on his act of compassion for the ox. Here the question of whether cognitive extension is sufficient emerges in the concrete. Was it sufficient for Mencius to have reminded the king that he has even more of a reason to spare his people from suffering than he had to spare the ox from suffering (more reason because Mencius clearly ranks the interests of animals below those of human beings, and because for him there is a good moral reason for the performance of ritual sacrifices)? Logical consistency alone cannot be expected to provide motivation, as David Nivison has pointed out (1996), but then what is Mencius trying to do with the King if not move him through logic?

Nowhere in the Mencius is there enough said to point to a definitive interpretation on this matter, but various reconstructions of possible positions can be given. Perhaps the King’s innate nature contains all the motivation he needs, and all that Mencius is doing is reminding him that he has the motivation to spare his people. This interpretation seems roughly in accord with Mencius’ likening moral development to water flowing downward: it will proceed unless interfered with. Perhaps the King’s nature needs some degree of transformation that starts with the sort of compassion he can feel for a terrified ox or an innocent man about to be executed and then expands the scope of that compassion to more of its appropriate objects. This interpretation seems roughly in accord with Mencius’ likening moral development to the growth of sprouts that need the appropriate water, soil and cultivation efforts. (See Im, 1999, Ivanhoe, 2002, Shun, 1997, Wong, 2002, Van Norden, 2007, and McRae 2011 for a range of different possible positions that could be attributed to Mencius).

What seems philosophically fruitful about the Xuan and ox story is that it portrays an attempt at moral teaching of the kind that actually occurs in the moral life, and the ambiguity that it presents to the reader is fruitful precisely because it is not a completely theorized story. We are not told exactly what Mencius is trying to do with the King in terms of a theory of the nature of emotions and the relation between the cognitive and affective. Rather, we are led to reflect on the most plausible possibilities in trying to arrive at a reconstruction of what might have been meant by the text, as well as what might be the most illuminating position on its own merits. The story is particularly intriguing for those philosophers who believe in the possibility that learning can influence emotion at the same time that emotion can help learning become motivationally efficacious. At one moment in the conversation between King Xuan and Mencius, the King remarks that Mencius has helped him to understand that he was truly motivated by compassion to spare the ox. As Mencius was talking, he says, he felt the stirrings of compassion again. It is at this point that Mencius reminds the King of the suffering of his people. Perhaps what Mencius was trying to do was to expand the scope of the King’s feeling for those who suffer. He may know in the abstract that he has a duty to advance the welfare of his people, but he may not feel it very much, and by stimulating the King to re-experience the feeling of compassion he had for the ox, Mencius is perhaps trying to get that feeling to expand to include his subjects. Mencius is reminding the King of his duty precisely when he has primed the King to assimilate it emotionally. In any case, this is a plausible picture of how the cognitive and affective can interact in emotionally development and both be necessary for effective moral motivation (see Wong, 2015a).

If we recall Mencius’s saying in 6A10 that even a beggar would refuse food that has been trampled upon, we may infer that something like a concept of human dignity is implied. In virtue of the four duan, we are entitled to respect. What is of further interest, however, is the idea that the source of our dignity is also the source of our responsibilities to each other and to ourselves. We have the responsibility of reflecting on the duan, and if we abuse them and kill them off, we have made ourselves unworthy of the respect to which we were originally entitled. These implications of the Mencian idea about the source of human dignity are not so much highlighted in the Western tradition. The idea that we might render ourselves unworthy of the basic respect to which we are originally entitled might appear harsh, but what should be kept in mind is that others share in the responsibility to develop our duan. Most importantly our family shares in that responsibility, but as Mencius also emphasizes, the king has responsibility for putting in the conditions for material security that enable most people to reflect on and develop their duan. In fact Mencius holds responsible the king who does not provide a constant means of livelihood for the crimes his people commit in order to survive and feed their families. It is those who have adequate material security and moral education and yet who still fail to reflect who must bear the fuller share of responsibility for their wrongdoing.

What about the priority of filial loyalty over loyalty to the larger community? Mencius’ defense of this priority draws from his theory of human nature as containing not only the beginnings of affective motivations for being moral but also intuitive judgments about what is right and about what deserves the feeling of shame. His question to a Mohist, Yizi, is how Yizi can justify providing his deceased parents a special burial when the Mohist prescriptions are for a plain burial for anyone. Yizi’s reply is to quote from the Book of History: the sage-kings treated all their subjects as if they were their new-born children. Yizi’s interpretation of this saying is that there should be no distinctions in one’s concern for people, though the practice of it may begin with one’s parents (how concern without distinctions is compatible with special burial for one’s parents is unclear, and there is no further clarification in 3A5, but the issue involves the question of what Mohists meant by jian ai, which is addressed in section 3 below). Mencius’s counter-reply is to ask whether Yizi really holds that a person loves his elder brother’s son no more than his neighbor’s baby. This is not just an assertion about what people tend to feel but also an assertion about what people intuitively hold to be right to feel and to do. Then Mencius makes a puzzling remark to the effect that Yizi is singling out a special feature in a certain case: “when a new-born babe creeps toward a well, it is not its fault.” This last part of Mencius’s response is puzzling because Yizi did not say anything about a baby and a well. One possibility is that Yizi may have obliquely referred to Mencius’s claim that all have the original and unlearned feeling of distress at seeing a child about to fall into a well. In other words, Yizi might have been challenging Mencius by asking, “Does not your own postulated unlearned compassion require us to regard that child the same way, regardless of whose child it is?” This way of taking Yizi helps makes sense of Mencius’s reply. First, he points out what he takes to be the indisputably greater affection one feels for elder brother’s son over one’s neighbor’s baby. Mencius grants that we all respond to a child about to fall into the well with alarm and distress, and it doesn’t matter whose child it is. However, one cannot infer from this one particular situation that we ought to have equal concern for everyone in all situations. The case of the child about to fall into the well has a particular feature that makes it relevant to treat it as one would any child. That feature seems to be innocence.

The Mencian position is premised on the principle that it is right to treat all people alike only when the ways they are alike are the most ethically relevant features of the situation. We should do the same thing only when the similarities between two cases are the most ethically relevant features of the situation. Mencius believes that in many instances, the presence or absence of a family relationship to a person is the most relevant feature (in deciding which children to give gifts, the fact that one child is one’s elder brother’s son and the other child is one’s neighbor’s child may be the most relevant feature). In other types of situations, such as a child about to fall into a well, it is the innocence that children share that is the most relevant feature. That is why it is proper to feel alarm or distress toward any child in that situation. The implied application of this idea to the sage-kings’ treatment of the people is that these kings treated all people alike insofar as they did not deserve the harm about to befall them.

Two issues arise from this response to Mohism as he understood it (whether he understood it correctly will be addressed in section 3). One issue is whether Mencius has sufficient warrant to trust the kinds of intuitive judgments he attributes to human nature. Mencius holds that the beginnings of morality are sent by Heaven, but in the absence of such a metaphysical warrant, can these intuitive judgments be accepted, particularly the ones that underwrite care with distinctions? Doubt about the metaphysical warrant may not doom Mencius’s response to Mohism, however, if one holds that all normative theories ultimately depend on intuitive judgments and if one has no good reason to be skeptical about these judgments. Thus one might hold that whether or not there is a metaphysical warrant, there is a great deal of plausibility to the intuitive judgment about owing parents more concern because they are the source of one’s life and nurturance. Of course one might also hold, as Mencius appears to hold, that people are owed concern in virtue of their being human, and the possibility for conflict of duties arises from these different sources of concern. The second issue is how the Mencius text deals with conflicts of the sort exemplified by the sheep-stealing case in the Analects.

The text contains themes embodying the theme of filial loyalty, and as in the Analects, such loyalty takes precedence over public justice. 7A35 tells a story about the sage-emperor Shun that illustrates this theme. Because Shun was renowned for his filiality, Mencius is asked what Shun would have done if his father killed a man. Mencius replies that Shun could not stop the judge from apprehending his father because the judge had the legal authority to act. But then, Mencius says, Shun would have abdicated and fled with his father to the edge of the sea. 5A2 and 5A3 describe the way that Shun dealt with his half-brother Xiang’s conspiring with his father and stepmother to kill him. He enfeoffed Xiang because all he could do as a brother is to love him. At the same time, Shun appointed officials to administer the fief and to collect taxes and tributes, to protect the people of Youbi from Xiang’s potentially abusive ruling. That is why some called Shun’s act a banishment of Xiang. However, the Shun stories exhibit a complexity that differentiates them from the story of the sheep-stealing coverup in the Analects. Though filial loyalty is clearly given a priority in each story, there is in Shun’s actions an acknowledgment of the other value that comes into conflict with filial loyalty. Though Shun ultimately gives priority to filial loyalty in the case of his father, his first action acknowledges the value of public justice by declining to interfere with the judge while he is king. While Shun declines to punish his half-brother, he protects the people of Xiang’s new fiefdom.

These Shun stories illustrate that an agent’s response to a situation in which important values come into conflict need not be a strict choice between honoring one value and wholly denying the other. While some sort of priority might have to be set in the end, there are also ways to acknowledge the value that is subordinated, but how exactly that is to be done seems very much a matter of judgment in the particular situation at hand. The Shun stories are an expression of the Confucian theme that rightness cannot be judged on the basis of exceptionless general principles but a matter of judgment in the particular situation. It is difficult to see how this theme can be taught except by the way it is done in the Mencius: through exemplars of how it is done, and where the situation is presented through some kind of narrative.

The characteristic form of reasoning in Mencius is analogical reasoning (see Lau, 1970b; Wong, 2002). Starting from what seems true in one case and “extending” similar conclusions to another case that has similar conclusions. The trick in doing analogical reasoning correctly, as suggested earlier, is to extend the similar conclusions only when the two cases share ethically relevant and decisive features. The Mencius 4A17 shows a similar concern for treating like cases alike. Mencius grants that to save the life of one’s drowning sister-in-law, one of course suspends the customary rule of propriety prohibiting the touching of man and woman when they are giving and receiving. Another philosopher proposes to apply this idea of suspending the usual rules of propriety to save something else from drowning—the entire Empire! Mencius replies that one saves one’s sister-in-law with one’s hand but cannot save the Empire from drowning in chaos and corruption with one’s hand. The Empire can only be pulled out by the Way. Mencius is rejecting the analogy between compromising on ritual propriety to save the country and compromising on propriety to save one’s sister-in-law. There is a relevant dissimilarity between the case of the drowning sister-in-law and saving the country: one cannot save the Empire through compromises of ritual propriety, but instead by following the Way, which itself involves following ritual propriety.

So what do we do when we confront a problematic case in the present and we do not automatically know what the right thing to do is? Mencius believes we can rely on past cases in which we have made reliable judgments about, for example, what is right and shameful. These reliable judgments made in past cases serve as paradigms or exemplars of correct ethical judgment. In encountering new problem situations, we determine what sort of ethical reaction to the new situation is correct by asking which of the cases in which we’ve had paradigm judgments are relevantly similar. We then determine what reactions to the new situations would be sufficiently similar to the relevant paradigm judgments. Analogical reasoning is careful attention and comparing to a concrete paradigm. The pool of paradigm ethical judgments we have not only includes cases from our own personal experience, but also include the experience of others, especially those who serve as models of wise judgment. The stories of sage-king Shun in the Mencius text seem to give us such paradigms. Shun’s judgments on what to do about conflicts between filial loyalty and public justice are perhaps meant to serve as paradigm judgments. The conception of moral reasoning found in the Mencius offers important material for reflection on the process of moral judgment, especially for those who have come to reject the simple model of judgment as deduction from premises including a general moral principle and a description of the conditions that make the principle applicable to the situation at hand. The Mencian picture includes general moral considerations or values that bear on the situation at hand, such as the importance of family loyalty and public justice, but the picture also suggests that judgment in difficult situations includes finding a way to adequate recognize and realize the values in play. “Finding a way” seems much more a matter of imagination and ingenuity rather than deduction, but the Mencian picture also suggests that we can be guided by exemplars of wise judgment. Identifying the relevant similarities and dissimilarities between these exemplars and one’s present situation seems a matter of perception and close attention rather than deduction from principle.

In the Xing E (“[Human] Nature is Bad”) chapter, Xunzi explicitly opposes his position on human nature to Mencius’s. He asserts that far from being good, human nature is bad because it includes a love of profit, envy and hatred, and desires of the eyes and ears that lead to violence and anarchy. To avoid these consequences of indulging our spontaneous desires and impulses, it takes wei (conscious activity or deliberate effort), models and teaching, and guidance through observing ritual and yi (standards of righteousness). Through such efforts, natural emotions and desires are transformed as a crooked piece of wood is steamed and then straightened upon a press frame. All rituals and standards of righteousness are sheng (generated, produced) by the sages. These are generated from the conscious activity of the sages and not from their original nature. Just as the vessel made by a potter is generated from his conscious activity and not his original nature, so the sages accumulated their thoughts and ideas and made a practice of conscious activity and precedents, thereby generating rituals and standards of righteousness.

Part of Xunzi’s argument against Mencius is that human nature is not what is produced by conscious activity but rather that which is already there in human beings independently of conscious activity. Since it is clear that human beings are not already good but must work at it, it is clear that human nature cannot be good. When Mencius is attributed the water-metaphor view of the human inclination toward goodness, Xunzi’s criticism has a point. Becoming good does not seem to be merely a matter of not interfering with what will unfold in normal circumstances. However, when Mencius is attributed the sprout-metaphor view, the differences between him and Xunzi are more subtle. On the sprout-metaphor view, effort and reflection must be put into the project of extending the sprouts to where they should be. It might be thought that one of the real differences between Mencius and Xunzi is that the former believes the necessary effort lies in growing or extending what lies in human nature, whereas the latter beieves that the effort lies in remaking and reshaping what lies in human nature. Perhaps one believes that we can go “with the grain” of what we are born with, and the other believes we must go “against the grain.”

Each thinker emphasizes one of these opposing directions, but it is a credit to the subtlety and power of their views that each also takes into account the direction that the other emphasizes. Mencius acknowledges that moral development is hindered when a person pays more attention to the “small” parts of the self that include desires for sensual and material satisfaction and fails to use the heart-mind to reflect on the great parts that have normative priority. In the chapter on rituals, Xunzi identifies natural and powerful emotional dispositions such as love of one’s own kind that rituals must give expression to and that seem to form more of a positive basis for moral development. Such natural love is expressed in love for parents and intense grief upon their deaths, which must be given appropriate expression in mourning and burial rituals. Thus Mencius acknowledges that there are natural parts of the self that must be disciplined and held in check while Xunzi acknowledges that there are natural parts that are largely congenial to morality in the sense that they are the natural basis for taking great satisfaction and contentment in virtue once one has gotten the self-aggrandizing desires and emotions under control.

Another disagreement between Mencius and Xunzi has to do with Mencius’ claim that human nature contains moral predispositions. As indicated earlier, such Mencian predispositions appear to contain moral intuitions (e.g., about what is shameful and right or wrong). On one plausible interpretation of Mencius, morality is part of the order imparted to the world by tian or heaven. By contrast, Xunzi seems to rule out the existence of natural predispositions with moral content when he claims that the sage kings generated ritual principles and precepts of moral duty. One natural interpretation of “generated” is “created” or “invented.” On these interpretations of each thinker, the contrast between Mencius and Xunzi exemplifies the contrast between a robust moral realism that has moral properties such as rightness existing independently of human invention and a constructivist position that makes moral properties dependent on human invention.

The interpretation of Xunzi as a constructivist does not necessarily commit him to a denial of the objectivity of morality or to the denial that there is a single objectively correct morality. It is possible to see Xunzi as a constructivist about morality but also as an objectivist (see Nivison, 1991). On the constructivist interpretation, Xunzi holds a functional conception of morality, according to which it is invented to harmonize the interests of individuals and to constrain and transform the heedless pursuit of short-term gratification for the sake of promoting the long-term interests of the individual and the group. Ritual principles and moral precepts are invented to accomplish such a function, and human nature constrains which of the possible principles and precepts are better or worse for accomplishing that function. Xunzi’s point about the mourning rituals prescribed by Confucians being suited to the nature of human love for one’s parents is a case in point.

Xunzi’s functional theory of morality bears added interest for those exploring the possibilities of a naturalistic approach to morality. One fairly common interpretation of Xunzi’s conception of tian or heaven is that it is an order-giving force in the cosmos that is neutral to whatever human beings have come to regard as right and good. In fact, a translation that better conveys such a meaning for ‘tian’ is “nature,” which is the translation given by Knoblock’s valued translation of the Xunzi. Textual passages that support this interpretation stress that tian operates according to patterns that remain constant no matter what human beings do or whether they appeal to it for good fortune (chapter 17). It is the proper task of human beings to understand what these patterns are in order to take advantage of them (e.g., so that they may know to plow in the spring, weed in the summer, harvest in the fall, and store in the winter).

Such a view of the difference between Xunzi and Mencius, however, depends on interpretations that been disputed in favor of alternative interpretations. Roger Ames (1991, 2002) defends an interpretation of Mencius that gives the greatest role in shaping the direction of moral development to human “creative social intelligence” rather than tian conceived as a force operating independently of human beings. For a contrasting view, see Irene Bloom (1994, 1997, 2002), who, sometimes in response to Ames, defends a greater role for biology in her interpretation of Mencius while also leaving an important role for culture. The Xunzi text is also susceptible to very different interpretations, partly because of the originality of its synthesis of several streams of thought: Confucianism, Daoism, Mohism, and the Jixia Academy. When Xunzi asserts that tian is unresponsive to human supplication and ritual sacrifice, it looks as if he might be drawing from Daoism (see below), but when he refers to the tian-given faculties human beings should exercise to solve the problem of conflict, he might be interpreted as implying that tian conferred these faculties upon human beings for the purpose of solving the problem of conflict and realizing fulfilling human lives together (see Eno 1990; and Machle, 1993 for an exploration of the rich interpretive possibilities regarding Xunzi’s conception of tian). While it might still be possible to interpret Xunzi as a constructivist on the origin of morality, this alternative interpretation might suggest that Xunzi’s tian had a blueprint it intended human beings to fulfill. Under alternative interpretations of Mencius and Xunzi, then, the differences do not disappear, but they might form even more subtle contrasts (Wong, 2016).

Even some of the theoretical difficulties that Xunzi has are instructive. In pressing his case against Mencius for the badness of human nature, he stresses the self-serving drives of human nature. Unlike Hobbes, he does not accept that human beings are inevitably motivated by self-interest, and he does not try to base adherence to moral norms on the basis of self-interest alone. This arguably is a promising move, given the heavy criticism that can be directed against the Hobbesian project and subsequent attempts to carry it out its basic idea (see Gauthier, 1986 for such an attempt; see Vallentyne, 1991 for criticism). Xunzi rather argues that the problems created by unrestrained self-interest point to the need to transform human motivation. People can come to love moral virtue and the rites for their own sakes, and this is necessary, on Xunzi’s view, for a stable solution to the problem of conflict between self-interested individuals. At times, Xunzi suggests that the intellect can override the desires arising from the natural emotions, but it remains unclear as to how self-regarding motivations can become a love of virtue and the rites simply because the intellect approves of them. The parts of Xunzi asserting a more complex picture of human motivation suggest a solution. If human beings are capable of genuine compassion and concern for others, as the chapter on rites suggests, then the ritual principles and moral precepts invented by the sage kings have some motivational leverage for the birth of a love of virtue and rites. Such a solution draws from what are arguably some of the most plausible positions of Mencius: that human beings are capable of altruism and compassion even if they are motivated much of the time by self-interest; and that moral transformation is a matter of cultivating and extending a motivational substance that is congenial to morality (see Wong, 2000).

Mencius and Xunzi, then, offer sophisticated theories that expand the range of possible ways of understanding moral knowledge, motivation, and the nature of morality itself. Mencius presents an interesting conception of the way that we reason by analogy from intuitive judgments and also a plausible conception of innate predispositions that are compatible with a major role for learning and upbringing in the development of character and virtue. Those who are more naturalistically inclined in their approach to morality (at least insofar as this involves resisting the idea of a transcendent source of moral properties) may find the interpretation of Xunzi as offering a functional conception of morality appealing, especially if it allows for a degree of objectivity regarding the content of morality.

In recent years, Gilbert Harman (1998–99, 1999–2000) and John Doris (2002) have pointed to the influence of situations over attitude and behavior as a problem for virtue ethics. Citing empirical work in social psychology, Harman and Doris claim that the extensive and surprising influence of situational factors undermines the commonsense idea that people possess stable character traits that explain what they do. Some of the classic psychological studies used in this argument appear to show that ordinary respectable American citizens will administer dangerous electrical shocks to an innocent person when urged to do so by an experimenter in a lab coat (Milgram 1974), and that being late for an appointment is the most influential factor in whether a seminary student will stop and help someone who seems to be falling ill, even if the appointment is to attend a lecture on the Good Samaritan (Darley and Batson 1973). Such studies pose a problem not only for the commonsense conception of character traits, but also for virtue ethics, which appear to assume the possibility of achieving stable character traits that are virtues. Perhaps human beings are inevitably creatures who are influenced by the situation in which they act and not by any characterlogical dispositions they bring with them to the situation. If so, it appears that the ideal of attaining virtues is misguided.

There are good reasons to expect Confucianism to offer some distinctive resources for dealing with this problem. First, as pointed out in 2.3, Confucians appreciate the relational nature of human life: who we are as persons very much includes our social context: the people with whom we are in relationship and our institutions and practices. So they are very much in a position to appreciate situational influences on how human beings think, feel, and act. Second, they appear to hold something like a conception of virtues as stable character traits that are resistant to undue situational influences. As noted in 2.3, and this pertains to the challenge posed by the Milgram study, the Confucians emphasize the importance of living according to one’s own understanding of what is right and good even if others do not see it the same way. Third, as noted at the beginning of this entry, Chinese philosophy in general is distinguished by a focus on the practical. This is illustrated in the Confucian case by the tradition of scholar-officials who not only wrote about and taught the importance of the ethical to the political life, but strove to enact this importance in their own careers. As a consequence, they were very much concerned with specifying in practical terms how one could go about cultivating the virtues in oneself. Fourth, and this is very much in response to the combination of the previous points, they describe a long and arduous program of ethical training to inculcate the virtues.

As Edward Slingerland (2011) has put it, Confucianism is in a good position to appreciate the “high bar” challenge of situational influence to the project of cultivating the virtues in oneself and others. In response to this challenge, their program of ethical training includes study of the classics (after the ancient period, the classics came to include, of course, the Analects and the Mencius), memorized and rehearsed until they become fully internalized and embedded in the unconscious patterns of thought that are so powerful in shaping what we do in everyday life (see Slingerland 2009). This is one characteristic pattern of Confucian self-cultivation: one consciously, deliberately and assiduously undertakes a program that inculcates dispositions to have ethically appropriate emotional responses and patterns of conduct. The intent is to make the dispositions for these responses reliable and resistant to undue situational influence.

Furthermore, the Confucians very much appreciated the power of models to inspire, to make one want to transcend one’s present self. The psychologist Jonathan Haidt (2003) has given empirical evidence for an emotion he calls “elevation,” which is something like awe and admiration upon contemplating the morally admirable. The Analects, in fact, has been read as a record of how a group of men gathered around a teacher with the power to elevate, and as a record of how this group created a culture in which goals of self-transformation were treated as collaborative projects. These people not only discussed the nature of self-cultivation but enacted it as a relational process in which they supported one another, reinforced their common goals, and served as checks on each other in case they went off the path, the dao. They were each other’s situational influences. See Sarkissian, 2010 for the argument that Confucius shows how one can turn the power of situations on people’s attitudes and behavior toward positive ends; if situations can influence people, one can through small details of comportment and attitude be a situational influence on others that tilts things toward a better course.

Training in ritual, li, takes on another dimension of importance in light of the situationist problem. As noted in section 2.2 Confucian rituals help to express attitudes of respect and reverence for others that can exist independently of the rituals themselves, but rituals provide conventionally established, symbolic ways to express these attitudes toward others. Ritual forms, therefore, give participants manifold and (just as importantly) regularly recurring ways to act on and therefore to strengthen the right attitudes and behavioral dispositions. Given the renewed appreciation in contemporary psychology for the power of emotions to influence attitude and behavior, the resource offered by ritual training should not be ignored by anyone concerned about the problem of how to resist undue situational influence.

Finally, Confucianism points to the possibility that individuals, under the right circumstances and encouragement, can enhance their reflective control of their own emotions and impulses. Mencius’ conversation with King Xuan can be conceived as an attempt to get the king to nourish his moral sprouts by reflecting on them, to become aware of what his moral emotions are (such as compassion) and to take action to grow them. It should be noted that contemporary psychology is exploring some possible venues for the regulation of one’s emotions and impulses. See Walter Mischel’s by-now classic study (1989) of children who are able to defer gratification for greater reward in the future (here’s one marshmallow; if you can wait fifteen minutes before you eat it you can have another one). It turns out the effective delayers use strategies of diverting their attentional focus from the marshmallow sitting in front of them. Projects are underway to teach children these strategies. See Lieberman (2011) and Creswell (2007) for studies indicating that meditation focused on cultivating compassion in oneself can be effective through enhancing one’s ability to identify and gain better control of one’s emotions.

Finally, in considering why robust character traits that could qualify as virtues are so rare, we should consider the perspective that very much informs the self-cultivation projects of Confucius and his students. They were very much aware of the lack of virtue as a social and political condition and not merely as an individual condition that just happened to be widespread (Hutton 2006 makes this point). There is a reason why Confucius and Mencius after him sought to have kings adopt their teachings. If in fact the achievement of robust virtues requires long and hard training, supported and guided by others who have taken similar paths before, and if as Mencius 1A7 holds, people cannot engage in such training until they have the material security that enables them to take their minds off the sheer task of survival, then it is no mystery at all why there are no such traits in societies structured to achieve very different goals. Ironically, the situationist psychological experiments do not take into account this underlying relational factor that might deeply influence the ability of people to form robust virtues, and neither do the philosophical critics of virtue ethics who rely on the situationist experimental evidence.

Zhu Xi (1130–1200) reinterpreted ethical themes inherited from the classical thinkers and grounded them in a cosmology and metaphysics that had absorbed the influence of Buddhism, particularly as it transformed in its interaction with Daoism when entering China (see the chapters on Zhu Xi and Wang Yang Ming in Ivanhoe, 1993 for the neo-Confucian reaction Buddhism and Daoism). Zhu established the Confucian canon that served as a basis for the Chinese public service examination, including the Analects and Mencius, along with the Great Learning (Da Xue) and Doctrine of the Mean (Zhong Yong). In fact, he had his greatest influence through the commentaries he wrote on these texts (see Gardner, 2003 for a discussion of the influence of Zhu Xi’s reading of the Analects). Zhu affirmed the Mencian theme that human nature is good, with greater emphasis on that vein of thought in the Mencius that stresses that goodness is internal to human beings and will develop in the absence of interference. This reading of Mencius is unsurprising given the influence of Buddhism on the Neo-Confucians, and it meant the demotion of Xunzi within the influential Neo-Confucian reading of the tradition. Much of Zhu’s metaphysics centers on the relation between li (in this case not ritual but principle or pattern or the fit and coherence between things) and qi (the material force or energy stuff from which objects emerge and return at the end of their existence). How Zhu Xi conceived this relation is a matter of interpretive debate. Some view him as holding a dualistic metaphysics in analogy to the way that Plato’s distinction between the Forms and the sensible world is often taken to embody a metaphysical dualism (Fung, 1948, chapter 25). However, others interpret Zhu’s li not as ontologically prior to qi but rather as being a pattern or deep structure that is immanent within and expressed by qi and delineates the range and possibilities of qi’s transformations (Graham, 1986; Thompson, 1988; Angle and Tiwald, 2017; Liu, 2018). Others have noted that li appears to have both subjective and objective aspects: it lays down the lines along which everything moves in a way that is independent of personal desires; but on the other hand, it is related to the pattern of one’s profoundest responses to things (see Angle, 2009, for an attempt to reconcile these aspects). With regard to qi, Zhu Xi held that even though goodness is within human nature, individuals differed with respect to their native endowment of energy stuff, and that this, together with differences in their family and social circumstances, affected the development of their good natures (Angle and Tiwald, 2017; Liu, 2018).

Zhu Xi saw one’s self-cultivation as a matter of apprehending the li of one’s own mind, largely through meditation practice, and, at the same time, investigating the li or patterns of things not only as revealed in texts such as the Analects but as embodied in concrete situations, including the patterns in relationships between persons. Both kinds of activities must be conducted with jing, which in Zhu’s thought means respectful attention. Zhu is sometimes characterized as a kind of scholastic, but he emphasized study of the texts in conjunction with acting, with observing li in external situations and relationships, and realizing the correspondence between the li of one’s own mind and the li of texts and of situations and relationships. Apprehending li in a concrete situation in order to respond appropriately to it was not a simple matter of absorbing generalizations from texts and applying it to the situation, but rather a matter of bringing to bear a mind that has been cultivated by meditation and by study of the texts and by observing and acting in previous situations. Such a mind can take into account relevant ethical considerations and is disciplined in attending to the situation (see the chapter on Zhu Xi in Ivanhoe, 1993; and Gardner, 1990).

The other Neo-Confucian whose influence rivals that of Zhu Xi is Wang Yang Ming (1472–1529). Wang saw Zhu’s emphasis on the investigation of patterns in external things as overly scholastic and leading to abstract speculation rather than practical guidance. He rejected what he saw to be the intellectualization of personal realization, and identified the mind with li (xin ji li or mind is pattern or principle). This means that the dispositions to judge properly the appropriate action in various situations constitute the mind’s original pure state. Li is not to be sought as a pattern residing in an independently existing external world but is embodied in judgments of the mind. This seems to commit Wang to an identification of the world with the experienced world and to a denial of a mind-independent world, but an alternative interpretation is that the structure of the world is mind-dependent while the world itself is not (Liu, 2018). Wang’s version of the Mencian theme that human nature is good is even more innatist than Zhu Xi’s (see Ivanhoe, 1990, for a comparison of Mencius and Wang Yang Ming). Original goodness does not need completion through learning about the external world. Then why aren’t all people fully good? Why are some very bad? Wang’s answer is that selfish desires cloud the sun of the complete and perfect moral mind, and that the task of human beings is to eliminate selfish desires and recover that mind (Chan, 1963, sections 21, 62).

One of Wang’s better-known themes is the unity of knowledge and action. There can be no gap between knowing what to do and doing it. Genuine knowledge is necessarily practical. Selfish desires and emotions get in the way of achieving genuine knowledge. One way of understanding this identification is to take knowledge as a knowing how to act that is expressed in acting. One expresses one’s knowing how to ride a bicycle by riding, not by articulating propositions about how to ride a 