Here is a short answer about the need for comparative philosophy. Progressively longer answers will follow.

Without some knowledge about differences among traditions of ethical thought, it is all too easy to conclude that ‘reason’ requires what amounts to a structural feature of most Western ethical thought. If we are ‘compelled’ to think in a certain way (given our tradition), then it may seem obvious that this is the path of reason.

Further, if the reference points are mainly Western systematic ethics of the last two and a half centuries, it is easy to conclude that the primary task of normative ethical philosophy is to arrive at principles that yield decision procedures. It might seem to follow that any project of transforming oneself into a certain sort of person is secondary to that of arriving at reliable moral decisions.

This paper will argue against all of these tempting conclusions. The first section will lay the groundwork by exploring some important differences among major traditions of ethical thought. The second section will examine the varying roles, in various traditions, of decision procedures and of principles. Part of what will emerge will be the limitations of any ethics that is principally concerned with fundamental principles that underlie major decisions in life. The third section will examine ethical projects of improving the self.

It should not be surprising that much Western ethics centers on a search for acceptable principles or rules, given the importance of the Judeo-Christian tradition that connects us with the foundational model of the Ten Commandments. The influence of this principle-oriented model is reflected in ordinary discourse. An example is that we often refer to a virtuous person as a ‘woman (or man) of principle’ and speak of what should be done as what ‘on principle’ we should do. This influence also finds its way into philosophy. When Kant approaches ethics in terms of maxims (personally adopted principles) that one might or might not be able to will as universal laws, he is very much in touch with this tradition. It enables him to regard it as obvious that principles and putative laws are inherent in moral experience in ways comparable to that in which the familiar structure of space, time, and the categories are inherent in our experience of the world.

In other parts of the world, however, principles and putative laws have not always been assigned such a major role. Some examination of non-Western ethical traditions can undermine the ready assumption that Kant, totally immersed in Western traditions, makes about the centrality in ethics of laws and principles.

The thrust of what follows is not in the direction of designating one or more ethical traditions as totally correct, or for that matter as totally wrong. Rather it is intended to highlight the usefulness of an ethics that is not philosophically insular, and that gives reasonable weight to elements prominent in a number of traditions. I will term this ‘cosmopolitan ethics’. The final section of the paper will examine the advantages of a cosmopolitan ethics.

1. Some Important Differences Among Traditions

Let me begin by distinguishing my concern from other concerns that might seem similar.

Prime examples can be found in Alasdair Macintyre's Whose Justice? Which Rationality (1988) and Three Rival Versions of Moral Inquiry (1990). These are estimable books, but are at some distance from the concerns of this paper. Macintyre places a major emphasis on the adequacy of decision procedures. My present concern though is much more with differences in the kinds of questions that are pursued, and in the methods and assumptions that are then employed.

The differences in questions are often profound. Some of the generalizations that are about to follow bring this out. They however should be read as implicitly qualified. To say that philosophers X are ‘generally’ different from philosophers Y in some respect has to be understood as referring to a frequent difference, and not as claiming that all X philosophers are different from all Y philosophers in that respect. Not all classical Indian or classical Chinese philosophers are alike, nor for that matter are all Kantians, all utilitarians, or all contemporary analytic philosophers.

That said, here is a general claim. The central issues (by and large) in classical Indian philosophy and also classical Chinese philosophy concern ways in which one can modify and develop one's self, and create a path of life that reflects this development.Footnote ¹ In much Western philosophy after Hume, in contrast, decisions in problematic circumstances have been regarded as central, and there has been especial attention to the development of decision procedures.

Two Indian traditions that to a large extent share concerns, although they provide different answers to key questions, are the one that includes the Upanishads and the Bhagavad Gita, on one hand, and that of early Buddhism on the other. The former insists on the fundamental reality of atman, an inner self that is qualitatively identical with Brahman (the entire universe as a divine reality). Atman is impersonal in its nature, and is contrasted with the more superficial individual self (jiva). Buddha denied that there was an atman, and (in some ways like Hume two millennia later) insisted that what we are is a bundle of constantly changing psychic elements. Commentators widely agree that early Buddhist philosophies center on an image of the individual in terms of ‘co-dependent arisings’.Footnote ²

These two Indian traditions converge in some of their recommendations of how to modify one's initial individual nature. But there also are sharp differences. Buddha's value system placed considerable emphasis on the adjustments required in order to be invulnerable to suffering, not as major a feature of the rival tradition.

The important point though, for the purposes of this paper, is that the question of how one could and should modify oneself drew considerable attention in classical Indian philosophy – attention that is not often paralleled in recent Western philosophy.Footnote ³ Both of the major traditions just portrayed regarded (for different reasons) our initial individual nature as a trap, something that it was important to modify. The modified self would result in a different style of behavior. This behavior would involve, in both of the two major traditions of classical Indian philosophy, greater detachment and less worry about the future.

This is not to suggest that there is no concern in these Indian traditions with decisions, or that the Indian philosophies do not provide principles that are held to underlie good decisions. The Bhagavad Gita opens with discussion of a major decision to be made by a central character, the warrior Arjuna. He is about to fight a battle in which it seems likely that he will triumph, and in the process kill relatives of his on the other side. His qualms about this are addressed by his charioteer, the god Krishna, who points out the principle that the Gita plainly regards as decisive. One should behave along the lines of one's assigned role in life. A warrior, in a situation like Arjuna's, ought to fight.

The early Buddhist literature also considers what clearly are regarded as important moral decisions. The most important of these concern the taking of life, whether human or animal life. The strong recommendation against taking life is based on a recommended emotional stance, that of compassion, which in turn is based on Buddha's deconstruction of the ‘self’. If each of us psychically consists of an ever changing bundle of thoughts and experiences, then one could readily imagine oneself as having the experiences of another person, or of the bug that one is about to step on. Compassion dictates that we not kill.

All of this is mentioned as prelude to the contention that, while these discussions of principled decisions have an important role in the two Indian traditions under discussion, they do not have the principal role. Most of the Bhagavad Gita is about how one can transform oneself into a person that can have a life of superlative value, and something similar is true of the early Buddhist literature.

A similar point holds, in an even stronger form, for most of early Chinese philosophy. In the Analects of Confucius, for example, there is not much discussion of specific kinds of decisions. Because of Confucius' ongoing concern with the ethical nature of good government, the two examples that emerge are related to government policies. A persistent question was whether – and under what conditions – a high-minded official should resign his post if he regards the ruler's policies as morally unacceptable. A specific question related to this was about the degree to which raising taxes on the poor peasants, who frequently were barely able to survive, was blameworthy.

Much more of the Analects though is devoted to questions about the development of personal virtue. A large part of Confucius' life was spent in moving around China, looking for a ruler who was willing to implement Confucius' ideas of good government. In the process, Confucius attracted students, who traveled with him and amounted to a sort of extended family. Much of the Analects purports to report conversations between Confucius and these students. Often what was discussed was the project of improving the nature of one's self. The students who came could be described, by and large, as sort-of-good people. Confucius was preoccupied with how a sort-of-good person could be transformed into a deeply (and reliably) good person.

The transition from being sort-of-good to being deeply and reliably good remains arguably a major concern for us. Think of the Milgram Experiment and the Stanford Prison Experiment. In one, most subjects inflicted what they had reason to believe were dangerous high voltage electric shocks on what they thought was another experimental subject. In the other, students assigned the role of guards began to treat brutally students who had been assigned the role of prisoners. One reason why these experiments have semed so disturbing is that one assumes that the majority of subjects who behaved so badly were reasonably decent people in their ordinary lives.

This suggests that to be (in ordinary life) an agreeable and generally cooperative person is not necessarily equivalent to being a really good person. The difference can come out in disorienting circumstances or in ones in which there is pressure from authorities (e.g. during the Nazi occupation of most of Europe or the Chinese cultural revolution). Being a ‘nice’ person turns out not to be equivalent to being a good person. Given this, one needs to consider where the differences lie, and how sort-of-good people can be educated to be really good people. The Analects of Confucius provides complex answers to both of these questions.

A seemingly simple account of how the transition can be accomplished can be found in the mosaic (of passages on various topics) of the Analects. It is ‘Find Inspiration in the Odes, take your place through ritual, and achieve perfection through music.’Footnote ⁴ This simple answer however requires (and gets) considerable elaboration, as we will see in a fuller account of Confucius' view.

The Odes (translated by Arthur Waley and also by Ezra Pound as The Book of Songs) seemed inspirational to the Confucian circle. Clearly there were passages that could be taken as suggestive. A student quotes one of these to Confucius: ‘As if cut, as if polished; as if carved, as if ground’ (Book 1.15). This seems to be taken as an apt metaphor for the refinement involved in becoming a really good person.

On the whole however the Odes would seem to most people today more like texts of folk songs than like moral inspiration. What about ritual? The bare mention of ‘ritual’, as we shall see later, hardly does more than gesture at what Confucius has in mind. Rather little is known about ancient Chinese music, although it seems likely that Confucius would have preferred music that expressed both calm and some joy rather than the counterpart of the Phrygian music that Plato rejected.

The major rival to Confucianism in ancient Chinese philosophy was Daoism (in the older Wade-Giles transliteration, Taoism). Daoism can be characterized broadly as a counter-cultural philosophy that regarded the Confucian emphasis on civic responsibility and on virtue as unnatural. Rigidity of thought was held to be undesirable, and this was true also of being governed by rigid rules in one's conduct. It was thought to be much better to maintain an openness of experience and thought, which allowed for personal spontaneity and a development of life as an extended improvisation.

The Daoist attitude toward virtue is not entirely clear. Some have thought that they were against it. But there are grounds for holding that their view was that ordinary decency was enough.

There is a passage in the Zhuangzi (Chuang-Tzu) that captures this. The power of virtue is sardonically compared to the warmth of the sun, and reference is made to the time when ten suns rose side by side over the earth.Footnote ⁵ The full story of this legend is as follows: The very ancient Chinese had a ten day week, and believed that there were correspondingly ten suns that took turns in moving about the earth.Footnote ⁶ One day the ten suns all rose at once, and the heat of this was about to destroy the earth until nine of the suns were shot out by a superhero archer. In short, what might read at first like an admiring tribute to the power of virtue is in fact an ironic comment on toxic goodness. A standard Daoist complaint about Confucians was that they pushed too hard, in too risky a way, for personal and social improvement, so that in the end the results of what looked exceptionally good were not good at all.

It is impossible to read the Zhuangzi or the eighty-one poems of the Daodejing without a strong sense that daoists were hostile to formulations of decision procedures and to any attempt to isolate justifying principles. Even more strongly than Confucius, their agenda centered on being able to transform oneself into a person who could lead a fulfilled life. The patterns of life recommended by the two philosophies, needless to say, were quite different from one another. They however shared the word ‘dao’ (which means path). Both philosophies saw ethics as primarily concerned with a path of life.

2. Decisions and Principles

Let me first make it clear that by a ‘principle’ I mean a problem-solving formula, much as (say) ‘Thou shalt not kill’ purports to solve the problem of whether you should kill that very irritating neighbor. One question then is whether all normative ethical problems of what we ought to do can be solved by appeal to some principle or principles.

As we look at the roles in ethics of decisions and principles, it is important to resist any automatic all-or-nothing approach to issues. There is a strong case for holding that any reasonable normative ethics will have a role for decisions of various sorts, and sometimes will base decisions on underlying principles. This is not however to concede that decisions and appeals to principles will have a primary or central role in any reasonable normative ethics. It is merely to say that they will have some role.

Decisions of what it is best to do also come in many varieties, and sometimes we might be unsure whether to call what is in some sense a choice ‘a decision’. In much of life we simply carry on, doing the sorts of things that we generally do. Arguably, at any moment we could start to behave differently. If this does not happen, can we say that we made a decision not to behave differently? The answer plausibly could be ‘yes’, if we paused and thought about it. If however there was no pause, and we simply went ahead in our usual way, this does not look like what we normally call a decision. The word ‘choice’ – which is closely related to alternatives being available – might be used, although even this can be debatable.

Many occasions on which we pause and make a decision look removed from the considerations of ethics. For example, whether we should order our usual dessert is normally not a question of ethics. There are some decisions though, in which something significant is at stake (such as possible harm to others, or our chances for fulfillment in life) that we would be more inclined to place in the territory of ethics. How do we decide in these cases?

Clearly there can be many competing answers. An expert casuist might look for similarities between the case at hand and other cases that have been decided in a certain way, thus assimilating ethical thinking (at least in such cases) to legal reasoning. It has been widely noted that much ancient Chinese ethical writing draws on arguments from analogies, a somewhat similar approach.

Alternatively, we could look for underlying ethical principles, which would amount to something a little like ethical software, as a way of making reasonable decisions. We might well hope to find a small number of such principles at a fundamental level. Both Immanuel Kant and Jeremy Bentham offered such underlying ethical principles. We will shortly discuss the appeal of such principles and also their problems. Let us first look at the major options, with regard to the role of ethical principles, in the construction of an ethics.

There are two obvious options, at the extremes. One is to construct an ethics (such as Kantian ethics or utilitarianism) in which underlying principles do all or at least most of the work in arriving at reasonable decisions. At the other extreme would be what has been termed ‘situation ethics’, in which we remain suspicious of any broad principles and instead look at all decisions on a case by case basis.

It is important to realize that these are not the only alternatives. Two great philosophers – Aristotle and Confucius – adopted what might be termed a hybrid ethics. That is, they both seemed to hold that most decisions of what one should do should not be based on any underlying principles, but rather must require a fine sense of the features of the individual case; but they also held that there were some decisions that should be made on the basis of principle.

This requires further explanation. The relevant discussions include Aristotle's examination, in Book 2 of the Nicomachean Ethics, of what is involved in following the mean. There also is a Confucian Doctrine of the Mean (Zhongyong), according to legend written by Confucius' grandson.Footnote ⁷ The most relevant passage in the Analects is 18.8. It refers to a number of famous men who withdrew from public service on moral grounds. In Slingerland's translation, Confucius says, ‘I however am different from all of them in that I have no preconceived ideas of what is permissible and what is not.’ Arthur Waley's translation has ‘As for me, I am different from any of these. I have no “thou shalt” or “thou shalt not”’.Footnote ⁸

Nevertheless, both Confucius and Aristotle did believe that there are some things that one should simply never do. Aristotle's examples are adultery, theft, and murder.Footnote ⁹ Confucius (in 1.5) clearly holds that one should never conscript the labor of the peasants at the wrong times of the year (e.g. during the planting or harvest seasons). Presumably Confucius would have said also that killing innocent people or seizing their money on a whim also is wrong. Confucius’ great 4th c. BCE follower Mengzi (Mencius) also held that to be a good person required that there were some things that one would never do.Footnote ¹⁰

This leaves us with a complicated picture. It is worth emphasizing that Aristotle and the Confucians, even though they hold that some things should never be done, do not accept an account of ethical decision in which rules or principles play a major role. The major role for Aristotle and the Confucians is reserved for people's ability to develop a desirable pattern of life in which they will be at their best.

Here is one reason why a great many (although certainly not all) ethical decisions cannot be viewed simply as matters of principle. It is that nuances which are difficult to articulate can make a major difference to decisions about a particular case. This is a familiar truth in aesthetics. There is a case for saying that it has an important role in ethics.Footnote ¹¹

It has been argued that a major function of reasons in aesthetics is to direct our attention to features of a work that should make a significant difference to our appreciation. This is apart from any ability of ours to be specific and precise about just what it is that should have this weight.Footnote ¹² It is arguable that something like this also fits, in Confucius' and Aristotle's views, the justification of choices that seem to follow ‘the mean’. Rather than invoking general principles, we have to look at the particular case.

We can think that there is something about a case at hand that justifies responding to it in a somewhat different way from the response appropriate to cases that are somewhat similar. This time, we might say, we should do X, even though on those other occasions Y was appropriate. As R. M. Hare argued in Freedom and Reason, this would be incoherent if in fact there were no difference among the cases.Footnote ¹³ But when the difference is subtle, it may be extremely difficult to articulate it in justifying one's different behavior. Any articulation may be unlikely to be at all precise, and we can be left in the position of saying that ‘you should see that this case is different’.

3. Projects of Self-Modification

My suggestion has been that projects of self-modification have not received very much attention in Western philosophy since Hume. We now should look at the nature of such projects. We also can look at their place in earlier Western philosophy.

Self-conscious development of the self was a major topic of Western philosophy from ancient Greek and Roman philosophy through Hume. Both stoicism and epicureanism focus on the values obtainable in the right kind of life, which can be led if one turns out to have the right kinds of interests and also not to have strongly negative responses to unwelcome experiences. Plato and Aristotle also held that through the right kind of education one could achieve a life suffused with rationality. This would involve becoming someone who was wary of specious pleasures and fond of more demanding intellectual activities.

Hume is a special case. Almost every philosophy graduate student knows that Hume denies the existence of the self. That was in Book I of the Treatise. In Book II Hume examines the development of what could be considered a self, placing it in a broad social context in which mirroring fellow minds have a critical role.Footnote ¹⁴

Any reader of Hume's ethical writings knows that he is at least as concerned with understanding our evaluations of people and their virtues as he is with judgments of specific actions. Plainly, in his view, one might want to be someone whose nature is favorably viewed by others. This might well lead to an attempt to become a different kind of person, improving some positive aspects of one's self and eliminating some negative ones.

Unlike the ancient philosophers previously mentioned, Hume does not concentrate on developing a full-fledged and somewhat specific account of the self you should want to have. He is, on the other hand, concerned with the psychological processes involved in wanting to be a different kind of person. The main questions concern how this is possible and how difficult it is.

Hume's view in the Treatise is that such a transformation of self is often difficult, and perhaps sometimes impossible.Footnote ¹⁵ He returns to this subject in his essay ‘The Sceptic’ (1742), asserting that anyone who thinks that philosophical reasoning can have a direct and immediate effect on someone's character is mistaken. Gradual effects though are possible, by modifying over time a person's ‘bent’ of mind and habits.Footnote ¹⁶

In philosophies that recommend change of self in a specific direction, the method of change is usually connected with the view of what the self is like. The Upanishads, as already noted, present a view of a deep level of self (atman). It is to the spiritualized universe as drops of ocean water are to the ocean. To agree with the claims of the Upanishads is not held to amount to enlightenment. Beyond agreement, one needs thoroughly to see the world in their way.

Meditation in which boundaries are erased, and atman is encountered as the nature of everything, is a means to this. Thoroughly to assimilate the early Buddhist view, in which the ‘self’ is a bundle of co-dependent arisings (and desire is folly), also is promoted by meditation. But the two kinds of meditation are very different. Early Buddhist meditation was often devoted to reminding one of facts (including death, disease, and in general the ravages of time) that might be forgotten. Often meditation was in a charnel area, in which one could contemplate skulls and skeletons. Suitably staged meditation also could combat desire. Liz Wilson's Charming Cadavers examines monks' practices of meditating on decayed remains of women, as a way of undermining the attraction of the opposite sex.

The Analects of Confucius provides an unusually complex guide to improving one's self. The basic role of inspirational texts (the Odes), ritual, and music already has been noted. The twenty-first century reader may well find the emphasis on ritual puzzling, especially in that many of us associate ritual with boredom.

It may help to reflect on two facts about what might be thought of as ritual. One is that it usually involves the interaction or two or more people, so that ritual could be thought of as a kind of social dance, providing practice in coordinating with other people. The other is that some very small everyday interactions are in a way ritual. Think of incidents in which one person holds a door open for someone behind her or him, who then says ‘Thank you’. This looks quite tiny. But it can be argued that the accumulation of episodes in which you thank someone for holding the door open, or for some other favor or a present – all of these being in a way ritual – contributes cumulatively to learning the emotion of gratitude.

Many of us are used to a view of ritual in which getting the traditional details right is treated as central. Confucius' circle however had a broader view. The spirit and attitude that a ritual performance involves is all-important. Book 1.12 of the Analects has ‘When it comes to the practice of ritual, it is harmonious ease that is valued.’ The nature of the person performing the ritual must be adjusted to the ritual, so that what emerges is in a way a personal expression. Someone who just ‘goes through the motions’ is, in this view, performing the ritual very badly. The ideal is a ritual performance that comes naturally, and that may include inflections of a personal sort.

The attitude toward music is also a little more complicated than one might think. There is an episode recorded in the Analects (17.4), in which Confucius visits a student of his who has been appointed governor of a town. When Confucius approaches, he hears the sound of music, reflecting the great emphasis (influenced by Confucius) that his student has placed on musical activities. The passage is one of a number in which the relations between Confucius and his students look more informal than one might suppose. In this case, Confucius teases the student, saying ‘Why use an ox-cleaver to kill a chicken?’

There is one further element in the Confucian approach to improving the self that is important. This is a strong sense of personal fallibility. It should be noted that, here again, Confucius and Aristotle have a little in common. In his discussion of the mean Aristotle recommends that we be mindful of any direction in which we tend to go wrong, and compensate for this by going a little in the opposite direction from what seems to us ideal.Footnote ¹⁷

Many passages in the Analects develop the theme that it is important to be aware of what might be faults in oneself, in order to do something about them. Some of this is put in general terms. Confucius is quoted (5.7) as saying ‘I should just give up! I have yet to meet someone who is able to perceive his own faults and then take himself to task inwardly.’

Confucius knows that he himself has faults and limitations, and that he needs to work on himself. The limitations are referred to in 9.8, where he says

In 7.22 we have

A word should be said about Confucius' view of what the satisfactions would be of reaching real goodness. The view is in some respects like Aristotle's, contrasting the security and relative degree of invulnerability connected with ‘internal’ values with the rewards of ‘external’ values. Analects 7.37 points to the contrast in sense of security: ‘The gentleman is self-possessed and relaxed, while the petty man is perpetually full of worry.’ 4.2 has ‘Without Goodness, one cannot remain constant in adversity and cannot enjoy enduring happiness.’ The explanation of this is that in adversity one still has the internal values to hold on to (which helps in remaining constant), and when things are going well the grounds of one's happiness will not be such as to lead to frustrating desires for more and more (which would mar one's enjoyment).

Any Daoist method of self-modification clearly would have to be very different from what Confucius recommends. The great Daoist texts however do not give much detail. To free oneself from being dominated by rules and established categories one would have to undo many of the effects of conventional eduation. Poem 48 of the Daodejing puts this neatly by saying ‘The Way (the enlightened path of life) is gained by daily loss.’Footnote ¹⁸

Immediate experience in all of its detail also should become more important than conceptual analysis. The Zhuangzi (Chuang-Tzu) in effect urges this in saying that to discriminate between alternatives is to fail to see something.Footnote ¹⁹ The emphasis should be on what is there, rather than on conceptual apparatus.

The Zhuangzi (along of course with Buddha's teaching) was a major influence on the development of the Chan Buddhism in China that became Zen Buddhism in Japan. The Zen literature is full of accounts of teaching, which point toward methods of improving a student's self to a point at which it is ‘liberated’. Most of this is designed to undermine any reliance on what might look like literal truth. The most famous example is the question ‘What is the sound of one hand clapping?’, which compels improvisation because there is no alternative. The emphasis on direct experience of detail also connects this literature with the Zhuangzi. One passage describes a Zen practitioner who has (seemingly) completed his training, and is prepared to be a teacher. He visits his old teacher on a rainy day, of course taking off his clogs at the entryway. His teacher then asks him ‘Did you put your umbrella on the right or left side of the clogs?’ The newly minted teacher realizes to his horror that he does not know. He goes back to several more years as a student, in order to realize ‘every-minute Zen’.Footnote ²⁰

4. Some Characteristics of a Cosmopolitan Ethics

It is safe to say that none of the philosophies that have been characterized would disagree with Mengzi's contention that there are some things that never should be done. This suggests that any acceptable ethics will have to have some room for rules or principles. Devotees of principles in ethics might want to push the point further. It can seem plausible to say that a major test of whether a transformed self represents an improvement lies in the principles that the new self respects or violates.

In some cases this is apt. If someone comes to take more seriously the traditional prohibitions against inflicting various forms of harm on innocent victims, this surely is a major improvement of self. Will all forms of improvement of self lend themselves to interpretation in terms of principles? If someone (whose interactions with others have tended to be stilted and cold) develops the harmonious ease that Confucius considers important, is this a matter of improved principles? If someone (along the lines recommended by much Indian philosophy and also Daoism) develops a detachment that leads to an improved focus on difficult situations, is this a case of improved principles? What about someone who, after having been rather lazy and unreflective, manages to adopt Aristotle's formula for eudaemonia and now often exercises intellectual virtues?

No doubt there is some label that can fit, more or less, any form of self-improvement. Along these lines someone who had on occasion been a bit insensitive and also somewhat abrasive, but who eliminates these tendencies, might be spoken of as having accepted the principle ‘Do not be insensitive or abrasive.’ It might be said also that the cases referred to in the last paragraph involve the principles ‘Have a harmonious ease in rituals or other interpersonal communication’, ‘Become more detached and focused’, and ‘Have a high level of eudaemonia’.

In all of this though there is a looseness of fit between improvement and principle. There is certainly a noticeable strain in regarding such things as first and foremost matters of principle. It is the strain of reducing something to a generalization that would not normally be thought of as its root in the way in which ‘Thou shalt not kill’ normally would be the root of a decision not to kill the irritating neighbor.

An example of Wittgenstein's captures this kind of thing nicely. Someone says, ‘All tools serve to modify something’, and goes on to point out how hammers visibly modify the position of nails, saws modify the shape of boards, etc. If it then is claimed that rulers and glue pots are tools, because they modify (respectively) our knowledge of a thing's length and the temperature of the glue, Wittgenstein's response is ‘Would anything be gained by this assimilation of expressions?’Footnote ²¹

The thought process of virtually everyone in the vast majority of cases of murder or theft could be analyzed as having two steps. The first is ‘That would be murder/theft.’ The second is ‘It should not be done/allowed.’

That one should not be insensitive typically does not ground judgment in quite so immediate a way. There are a great many cases in which one thinks that something is unacceptable or a ‘bit off’, and then looks for a word to cover the case. Most cases of insensitivity have this character. The label often functions as a summation of the judgment of the case, and much less often as an algorithm.

Aristotle and Confucius are both clear that a great many decisions about what it would be best to do cannot be viewed as decisions on principle. Many of these cases involve assessment of matters of degree, and appropriate awareness of nuances that may make a difference. Many philosophers have been aware of the difference between the broad moral principles that we teach to children, and that play such an important role in our morality, and the kind of justification that we can give for our judgment of difficult cases. Recall Hare's remark (cited in note xii) about justifications that defy ‘formulation in words at all’. Cases that admit only of such justifications cannot readily be assimilated to those in which we might say that such-and-such on principle is the right choice.

What I am calling a cosmopolitan ethics will treat such hard-to-articulate cases very differently from ones in which we can immediately know on principle that X is wrong (or right). The problems and judgments that a cosmopolitan ethics considers will take a number of forms. The result will not be neat. But it will not risk leaving out matters of importance in the way that much recent ethics has.