Moral Emotions

University of Toronto

(sousa@chass.utoronto.ca)

Revised text of a paper presented to the meetings of Societas Ethica,  Padova, August 26 1999.
A final version is now published in Ethical Theory and Moral Practice, 4:109-126 (2001)

Three Traditions.

The phrase, "Moral Emotions" suggests two questions.  First, some emotions, such as compassion or sympathy, are generally approved of; others, such as rage or envy, may be thought "nasty". What is the basis of these judgments?  We may also want to ask whether emotions in general can themselves legitimately form the basis of moral judgments. In the following discussion I shall principally be concerned with the second question, but my focus will be on the apparent circularity which arises if emotions are both the object and the basis of moral appraisal.

On the relevance of emotions to ethics, the history of philosophy affords three positions.

The first, which we can roughly identify with the Stoics, denies that emotions have any moral worth: emotions as a group are better stamped out in the service of the good life. This position influenced Kant, and thereby modern philosophy. Its virtue lies in highlighting the importance to ethics of impartiality, but is now widely held to be incredible.

The second view is that some emotions, such as compassion or sympathy, are specifically moral in that they are conducive to moral consciousness and behaviour. The view that there are specifically moral emotions is associated with the tradition of Hume, Hutcheson, and Smith: emotions are regarded as moral primarily in the light of their tendency to motivate other-regarding behaviour. On this view, however, many emotions can continue to be regarded as either amoral or evil, and sorting the good from the bad ones must be done on the basis of some independent criterion.

The third might be traced back to Aristotle's view that moral education involves learning to feel the "right emotion to the right degree at the right time." (EN 1104b13) It views all emotions as intrinsically relevant to ethics rather than identifying a privileged class of emotion. The emphasis in this third tradition is not so much on the motivating power of emotions, but on their intrinsic importance to human life. In this perspective the value of emotions to ethics lies not so much in what emotions can contribute to our moral behaviour, as in their nature as components of the good life, without which the very idea of morality would be pointless. My main purpose today is to defend and extend this third perspective.[1]

The Wild Child's Sense of Justice

Let me begin with a fascinating experiment carried out by the 18th century country doctor, Jean Itard, who undertook to educate a wild child found in the woods of Aveyron. (Itard 1964) Dr. Itard's pupil never succeeded in acquiring even as much language as have some apes in recent times. Despite this, Itard devised an experiment intended to answer this question: Had the child acquired a moral sense? His experiment consisted in "punishing" the child for doing something right. The child reacted with what Itard interpreted as indignation, from which Itard concluded that the child had indeed acquired a moral sense.

Quite apart from the question of its own morality, on which present-day Ethics Committees would no doubt cast a disapproving eye, this experiment raises several interesting questions.

First, it implies that in Itard's mind the emotion could be identified clearly enough despite the absence of language in which to express it. All he had to go on, apparently, was that the boy took his punishments meekly enough when they were "deserved," but reacted with what looked like rage when they were not. The child had evidently learned to discriminate something, but could Itard really be sure of distinguishing the expression of indignation or fury from that of shame, sadness, or frustration? No doubt he could, providing that those emotions and their expressions are natural and pre-linguistic. And if we are to believe the subsequent observations of Darwin and more recently of Paul Ekman (Darwin 1998), he was probably right with respect to some emotions. For Ekman's research seems to confirm that for about half a dozen emotions, including anger and fear, the relevant facial expressions are clearly identifiable across cultures. But it's hard to know how the case of a child's attitude to punishment would generalize to other aspects of justice. Would the child's outrage have generalized to offenses against others than himself, or might it have been mobilized against himself, where he was the one that stood to lose by a just disposition? In the absence of language, it seems impossible to say. The first question, then, can be summed up in this one: how natural is the emotion that Itard thought he detected in the child, and more generally, to what extent are emotions natural phenomena?

Second, granting that Itard did succeed in identifying the emotion in question, must we take it that the emotion was a response to some objective feature of the situation that elicited it-its injustice-or should we rather say that such situations are to be assessed as unjust precisely because of the kind of emotion they elicit? I shall call this the Euthyphro Question.

Naturalness.

On the first question, we can distinguish at least four positions, differing in their increasing insistence on the distance between emotions and nature. All of these have been or might be held, though neither of the extreme positions is at all plausible now. They are summarized in Table 1:
 
 

natural feelings

socially constituted

individual character

libertarian/existentialist

1. On the first view, emotions are purely biological phenomena, and therefore without moral import in themselves. They are arational and amoral, like other natural bodily functions; and while that is quite compatible with their having uses as well as being nuisances, one cannot be praised or blamed for having them, any more than for one's digestion or the colour of one's hair. Nor does having them tell us anything of moral significance about the world. Nowadays this is a view more often attacked than defended, under the general description of the "feelings" view of emotions. (The first modern attack on this view was that of (Bedford 1956).)

2. On the second view, emotions are socially constituted and determined. On the basis of an innate but limited capacity for pleasure and pain, our repertoire of emotions is learned, and tightly controlled by social conventions. On this view, therefore, if emotions are relevant to morality at all, they can only be seen as the instruments of a purely conventional morality. There is little or no room for individual decision or moral autonomy.

3. On the third view, emotions are partly dependent on social environment, but only insofar as the same might be said of individual character. Character certainly depends in part on education, but admits of a high degree of individual variation. On this view emotions are comparable to language: the capacity for emotion and the available repertoire of possible emotions depend heavily on both nature and society, but a large range is left to individual expression.

4. On a fourth view, emotions are entirely chosen. There is no element of determination associated with emotions, either at the level of biology or at the level of social conditioning. They are purely the expression of our individual free-will. This is the view attributed to Sartre and endorsed by (Solomon 1973) in a well-known article.

Both extreme positions are too implausible to be taken seriously. The reason the first is not now credible is that it is incompatible with the thought-dependency of emotions-something that was already very explicit in Aristotle's Rhetoric. The reason the fourth can't be taken seriously is that it is perfectly obvious to anyone not in the grip of a cocaine-inspired fantasy that there is much to be said for the traditional view of emotions as passive. The space of possible views is therefore limited to those boringly moderate middle views that allow for a measure of naturalness in emotions as well as a measure of learned social dependence.

In my view, the difference between them may lie not so much in whether one or the other tells a universal truth, as on the structure of a given society and the kind of upbringing it provides. Societies which tolerate individual critical thought-very roughly, what (Popper 1963) called "Open Societies"-are peculiar in the history of humanity. In nearly all societies surveyed (Murdock 1967), the tyranny of the group forces every individual to conform to rigid social expectations, especially in everything that might remotely affect reproduction, kinship, and sexuality. (Sober and Wilson, 1998, chapter 5). In such older European literature as Icelandic Sagas, the same pattern is evident: their heroes live with the constant preoccupation to defend narrow conceptions of rank, honour, offense and revenge.(Miller 1993, ch. 3).[2]

This picture is no doubt too simple, in that these societies also regularly provide examples of the rare individual who transgresses. But Sober and Wilson stress that the society's grip is particularly powerful in view of the very low cost of the ostracism or social disapproval which generally suffices to discipline the transgressor.

The verdict so far on the question of the naturalness of moral emotions is a mixed one: it's not obvious that emotions don't result from just the right combination of natural dispositions and social conditioning to dovetail with morality. But more needs to be said. Before taking up the Euthyphro problem, it must be noted that from the point of view of our common-sense view of morality three things still seem to be missing from the Wild Child's putative moral sense. These are complexity, some notion of universality, and altruism. All three affect the question of the naturalness of emotions.

Complexity

If there are emotions likely to be pertinent to moral assessment, they cannot be anything so simple as an innate or instinctual disposition. Suppose that among a child's first few words was `justice'. Surely we wouldn't infer that the child had an understanding of the notion of justice, any more than should infer that she understands quantum mechanics if the next words she was heard to utter after `Mama', `No', and `More' were `Bell's inequality'. Any word relating to morality presupposes a whole vocabulary of related concepts. Thus we may cast doubt on Itard's claim to see in the wild child's reaction a genuine moral sense on the ground that his emotion does not seem to belong to a sufficiently complex network of related items-be they emotions, thoughts, principles priorities, preferences, values, or reasons.

We might draw one of two conclusions from this consideration. One is that emotions in themselves are too simple to play a role in the determination of morality. The other is that a certain minimal level of complexity must attend anything we call an emotion proper, if it is to be deemed worthy of claiming some sort of moral status.

Universality.

Universality is the demand of morality that seems least susceptible of being adequately met by the emotions, however benevolent. An emotion like compassion, for example, is typically felt in regard to a particular person, without any serious comparison between that person's plight and that of others who might be compared with them. (Though we sometimes accuse of the "vice of sentimentality" those whose compassion is directed at the current objects of their attention without any thought for others, far more deserving of pity, whom at the moment they don't happen to notice.) Yet it is notable that we speak of a sense of fairness, where fairness is precisely the aspect of justice that is supposed to correct for that kind of negligence of comparisons. To be sure, the implementation of a notion of fairness may turn into an arbitrarily sophisticated task. But even an untutored sense provides a start, and one tutored by successive applications of a method of reflective equilibrium appears indispensable to guide even our most mature deliberations about justice. On the other side, one may well have doubts about the aptitude of the notion of universalizability itself to provide the key to ethics (epitomized by the universalizing policeman's injunction: "Don't just stand there, think what would happen if everyone just stood around: nobody could get by!"). We should not, therefore, too readily abandon the hope that our emotions should contribute a relatively natural element to our grasp of morality.

Altruism

We tend to judge whether someone is altruistic or selfish in terms of motivation. In view of the connection between emotion and motivation, it seems urgent to decide whether we have any intrinsically or essentially altruistic emotions. This seems very difficult to decide. Notoriously inconclusive arguments have raged as to whether people's motivation in helping others, even at apparent cost to themselves, is "ultimately" designed to secure for the agent a superior pay-off in self-satisfaction. The inconclusiveness stems from the fact that even if I genuinely desire the good of another, nothing precludes my experiencing pleasure at the attainment of that good by the object of my concern. This appears to render the matter immune to empirical investigation.

In recent years, however, debate has become more interesting by virtue of acquiring a biological angle. In its biological sense -- a type of behaviour that results in a net loss to the agent's own relative fitness -- altruism presents a paradox for Darwinism. For it seems that any heritable tendency to lower one's relative fitness must necessarily breed itself out of existence. So insofar as morality entails the performance of altruistic acts, its existence requires either that we uncover mechanisms capable of solving the paradox, or that we regard morality as integrally unnatural. There's a long tradition of regarding nature as intrinsically evil and morality as arising from social necessity and by social imposition. Among biologists, a forceful statement of this antinaturalist position is to be found in T. H. Huxley, who wrote:

The propounders of what are called the `ethics of evolution' . . . adduce a number of more or less interesting facts and more or less sound arguments, in favour of the origin of the moral sentiments . . . by a process of evolution . . . but as the immoral sentiments have no less been evolved, there is, so far, as much natural sanction for the one as the other. . . . Let us understand, once for all, that the ethical progress of society depends, not on imitating the cosmic process, still less in running away from it, but in combating it (Huxley and Huxley 1969).

Subjectivism and Objectivism.

I turn now to the second relevant dimension, the one that has its source in the famous Euthyphro problem: is what is pious so-called because the gods love it, or do they love it because it is pious? Adapted to present purposes, this might read: do emotions apprehend antecedently existing facts about value, or are facts about value mere projections of emotions?

The question of subjectivism, like that of naturalness, admits of a spectrum of answers. Here too, the answers traditionally given arrange themselves along a single dimension, which measures the extent to which questions about ethics are dependent upon the condition of the subject. As illustrated in Table 2, we may distinguish four positions: extreme Subjectivism, Foundationalism, Naturalism, Axiologism, and extreme Objectivism.
 
 

Extreme subjectivism (emotivism)

Emotional founda- tionalism (3 variants)

Naturalism

Axiologism

Extreme Ojectivism (Plato)

Emotivism

It was Spinoza who first clearly endorsed the exact contrary of Plato's objectivist position: "We do not want any thing in virtue of the fact that it is good. On the contrary, it is in virtue of the fact that we want it that we judge it to be good." [Ethics III 9] The modern version of that position is the emotivism associated with logical positivism. It is the view that the meaning of "x is wrong/good" is none other than the expression of an un/favourable emotion. As has often been pointed out, however, this uncompromising view suffers from two serious flaws.

The first is that it does not give rise to a coherent semantics when conditionalized. (Williams 1973) For suppose I say: "If torturing small children is wrong, then people who do it should be punished." What is the status of the antecedent? If it has no meaning in isolation except for the expression of my disapproval, then its meaning here poses a dilemma:

Either the antecedent expresses the emotion or it does not.

If not, then since that was held to be its whole meaning, it has none left here at all.

But if it is, then the conditional is short-circuited, and the conclusion follows without the benefit of any minor premise affirming the antecedent..

The second objection can also be framed as a dilemma:

Is it the case that any emotion whatever is equally authoritative as a moral sentiment?

But if not, then what justifies this judgment? We need a principle of selection which will enable us to sort those emotions that have a claim to moral authority from those that do not.

As will become clear, I think this objection can be answered. But the answer cannot be of much help to pure emotivism: it will be made, instead, on behalf of a view I view I have defended under the name of axiologism. First, however, we need to acknowledge another set of intermediate positions, which I'll perhaps slightly misleadingly dub `Emotional Foundationalism'.

Emotional Foundationalism:

This view comes in least three variants.

(i) "Moral sentiments" are psychological motivators of ethical behavior.

The crucial observation here is that whatever argument we encounter for one or another position in ethics, the position will not move us unless it elicits the relevant emotions. The relevance of that observation might be contested: for it might be taken to mean simply that there exists a purely causal condition for acting morally, which has nothing to do with any epistemological conditions. But that answer can only be made to stick if the alternative epistemological foundation can be made clear; and the perennial disputes about the foundations of ethics cast doubt on that hope.

(ii) Moral sentiments are logically grounded in emotions.

If no one ever cared about any outcome, it is difficult to imagine there being a point to moral choice. The existence of emotions, or at least of the primitive or "proto-emotions" of pleasure and pain, as well as whatever emotions are reflected in relations of preference, is necessary for morality to have any point. This type of "foundationalism" neatly encapsulates an intermediate position between subjectivism and objectivism; for the fact that emotions exist is an objective fact; but it is an objective fact about subjects. Clearly, however, it doesn't capture everything about morality, for it "flattens out" the world of value, as it were, into the single dimension of pleasure and pain. A more realistic picture will allow for the multidimensionality of value. (de Sousa forthcoming(b))

(iii) Moral sentiments are developed out of more basic emotional roots.

We cultivate virtues, but could not do so unless we were equipped with certain emotional dispositions. Those dispositions are not yet virtues, and I myself would be inclined to contend that they are not even full-fledged emotions; but they are transformed by appropriate upbringing into moral emotions. This view might be thought to be implicit in Aristotle; but the earliest explicit expression I know of it is in Mencius, who spoke of the "four roots" of virtue. The instinctive sentiment of compassion evoked in every normal human being by the sight of a child in danger constitute the "root" (duan) of benevolence (rén); shame and hatred are the roots of justice (yì); modesty and the feeling of submission are the roots of ritual propriety (lî); and attraction and repulsion are the roots of wisdom (zhi).

One particularly interesting thing about this list is the precise degree to which it is (and is not) intelligible. Despite the cultural distance, it is not incomprehensible. On the other hand, no one word clearly translates "li", and some of the original words used for the "root" emotions themselves lack any strict equivalents in modern Western languages. This suggests that Mencius was largely but not entirely right in postulating natural (and therefore universal) root emotions and their derived virtues. (Had he been entirely right, we couldn't experience any difficulty understanding him.)

Naturalism

The last two variants of Emotional Foundationalism are naturalistic views, in the sense that they view natural facts as sufficient to account for the not only psychological factors such as our motivation to act morally, but for the existence of morality itself.

Something very similar can be found in Hume's view that the passions provide both a psychological and a conceptual foundation for morality. The sentiments of sympathy, love, and pride, in particular, motivate us to serve the interests of others as well as our own. But those sentiments only earn the qualifier "moral" when they have been elaborated an integrated into the moral code of a social group, where stable and universalizable principles replace the inevitably changeable impulses of momentary sentiment.

Recall the problem of discriminating between moral emotions and nasty ones. The naturalistic hypothesis of Mencius or Hume will offer to solve this problem by first borrowing from utilitarianism the idea that such states as pleasure and pain are intrinsically justifying. The naturalistic perspective will then suggest ways of refining the results in terms of the sorts of roles played by different emotions in life. An example of this approach, whatever its shortcomings, is Mill's distinction between higher and lower pleasures.

But Naturalism still suffers from the problem problem of "flattening out" noted above. In its efforts at justification, it can only use a very restricted subset of emotion types. (Classic lists of virtues, including that of Mencius, are similarly short.) The oddity of this result appears when we compare moral emotions to other categories, such as moral acts or moral thoughts. It seems obvious that no short list, indeed no finite list, can be given of moral acts or moral thoughts. I see no a priori justification for this discrepancy. It seems to me to arise from the fact that we think of emotions, we have in mind principally the named emotions which can be listed, and which for the most part can be more or less directly associated with certain biological functions. I think, with (Campbell 1998), that this is a mistake. Campbell has argued, in effect, that there are no practical limits to the number of distinct emotions that can be experienced, any more than there are limits to the number of thoughts one can have. I think this idea becomes particularly persuasive if one thinks of the emotions aroused by esthetic experience. In watching dance or listening to music, for example, it seems obvious that emotions are involved in some crucial way. But if the emotions expressed by music and dance are "the great emotions"-those we can list on demand: anger, fear, love, awe, jealousy, sadness, desire-then they are just those described in kitschy drugstore novels. Bad art is precisely in the business of evoking clichés. But if works of art exist merely to evoke those emotions, and if there are no significant differences between any two instances of "fear", or "anger", and so forth, then why go to all the trouble of making great art? It hardly seems likely that sustains our interest in works of art in all their diversity should have much to do with their representation of emotion.

The Axiological Hypothesis.

I have argued elsewhere that emotions are not only assessable from the moral point of view, but are themselves revealers of value. That is roughly what I mean by the axiological hypothesis. Emotions are viewed as modes of perception, and therefore as giving us access to certain sorts of knowledge. What they give us knowledge of, is the value of things, facts or situations in terms of formal objects determined by the nature of the emotion itself.

The first statement of something like the axiological perspective was probably in Max Scheler's work on Sympathy (Scheler 1973). But it may be said to have been anticipated by the Socratic identification of virtue with knowledge. The axiological hypothesis remains crucially unplatonic, however. Like the versions of naturalist foundationalism discussed a moment ago, it remains an intermediate position, in that what is detected by the apprehension of emotion is not, as in the Platonic view, a transcendent reality. Given the social nature of human life, the reality to which our emotions give us access typically, though surely not exclusively, concern social relations. So the order of reality to which emotions give us access is the relatively objective world of human values. This apparent oxymoron reflects the fact that the realities revealed by emotions are local to certain organisms in certain environments. Thus emotional repertoires can differ, as can the significance of their members; but these differences are not arbitrary. There is no independent access to the world revealed by emotion. In this way the axiological view remains close to what I called the logical foundationalism that looks to the emotions to provide the point of morality. (Recall, however, that it is by no means confined to defining the point of morality exclusively in terms of moral emotions. For the foundational role of emotions, the fact that they constitute the point of morality is prior to the role that any particular emotion plays in promoting it.)

To forestall another possible misunderstanding, it is worth noting that the moderate cognitivism implied by this thesis in no way entails that emotions are infallible. They are no more so than any other mode of perception. The accuracy of our emotions' deliveries, like those of our senses, can be improved: hence the importance of educating the emotions, and hence also the distance between the moral "roots" of the virtues and the virtues themselves.

Nasty Emotions Again

We are still left, however, with the task of sorting nasty emotions from moral emotions. Actually there are two tasks involved here rather than one. The first consists in assessing whether a given emotion is appropriate to its target. To answer that question, we need to look to the triple origin of each emotion in biology, social conditioning, and individual biography. Even the most urgent and the most anti-social emotions, such as rage or murderous jealousy, can be appropriate in all three of these aspects: The sociobiological advantages of jealousy are obvious enough, and so is the social sanction that approves of it (by recognizing a special status to the crime passionnel, for example.) As for the biographical element, we can suppose that something like the Freudian story, given the right circumstances in an infant's life, can construct a certain kind of story, or paradigm scenario, the pattern of which will fit more or less well (and sometimes entirely fail to fit) situations encountered in later life. But that says nothing about the second question, which is whether such emotions are a good thing, or should be tolerated, or can contribute to a thriving moral life. It is in answering that second question that it is worth recalling and extending Huxley's strictures against the naturalistic view.

The fact that a certain mechanism was selected for good functional reasons-which means, because it conferred a comparative reproductive advantage-is of no moral significance whatever, as Huxley saw. But the fact that it was sanctioned by mechanisms of social sanction is not a whit more relevant. In fact, it is not difficult to see that wherever the social force of custom or tradition is invoked to justify anything, whether it be a practice or an emotion, we can be pretty sure that the practice or emotion is either morally neutral-something having to do with dance steps, cooking or costume perhaps-or else that it is positively evil. For if it were not, there would be other reasons to be cited in its favour. Appeals to tradition, like appeals to faith, are the last recourse of those who have already lost a rational argument. Finally, the appeal to biography, like the other two, may certainly evoke a constraint, insofar as the power of our biography to shape us may well grip us as tightly as either of the others. But again it can in no way constitute rational support for the moral worth of an emotion.

What then can provide such support? I have already rejected any transcendent solution: while emotions do indeed reveal something, what they reveal admits of no independent criterion. The answer therefore lies, I suggest, in viewing that fact not as the problem, but as the solution. I call this position Axiological Holism. Consider that in all other empirical domains, the illusions of one sense are corrected by nothing but the evidence of that sense itself, in concert with our other senses, arbitrated by rational thought. We can judge whether an emotion is to be countenanced as part of a good life, and its motivational power acknowledged, only as we test the veracity of perceptual information: by looking at the same thing from different points of view. This means, in effect, by appealing to other perception or to other emotions. Something like the method of reflective equilibrium is commonplace in science as well as in ethics; what is less often noticed is that in the case of ethics, the items that need to come to equilibrium are typically not ordinary empirical judgments, but emotionalresponses. Those emotional responses, I suggest, are best understood as axiological perceptions.

Note that when placed in the holistic context I have been describing, these responses will easily meet two of the three requirements mentioned above: they are obviously complex to a fault, and insofar as they are comprehensive in what they respond to they will be, if not universalized in a strictly Kantian manner, at least relatively insensitive to changes in individual perspective. As for the third condition, that of altruism, we have seen that it is an empirical question to what extent it is promoted by the natural tendencies of our emotional dispositions.

Irony and Biology

It is time to acknowledge the utopian aspect of my proposal. To respond emotionally to everything, and to sum the differently dimensioned vectors of one's incommensurable responses, obviously smacks of an unattainable ideal. But then much the same can surely be said of all the traditional phantasms of moral theory, the Kingdom of Ends, the Original Position, the Greatest Happiness. Still, it's worth mentioning, before concluding, some of the specific difficulties we are likely to encounter, and to suggest an appropriate and realistic attitude we might recommend in the face of these difficulties.

The specific difficulties I have in mind stem from precisely the multiplicity of emotional dimensions that I have been at pains to underline. We can pick out at least four distinct sources of conflict and incoherence among our emotions.

First, from the biological point of view, our most basic emotional capacities are almost certainly relatively independent modules, driven by independent biological needs (Ledoux 1996). There is no reason to think, therefore, that they will always be in harmony, any more than we can hope that the need to flee an enemy will never interfere with a peaceful digestion.

Second, most biologists have tended to agree that group selection can be assigned only a minimal role in the explanation of our emotional dispositions. This entails that selfish motives will frequently conflict with group interests (defined either as the sum of the interests of other members of the group, or in some other way that might appeal to emergent interests). That, in turn, entails that the dispositions fostered by biology are likely to conflict with those nurtured by the group. Indeed, the conflict with society exists at the very heart of an individual's identity. For as all too sadly illustrated by the wild boy of Aveyron, no one can be a fully human individual without having been brought up by other humans in a society of humans. But to be a differentiated individual entails having unique characteristics, and is seldom compatible with avoiding multiple sources of conflict with the dominant culture. In short, individuality both presupposes and conflicts with social norms.

Third, we should not forget the ambivalence not only of most emotions, but of virtues themselves. All virtues are vices, even if not vice-versa: kindness sometimes kills, they say, neatness is most thoroughly practiced by the obsessive-compulsive, and so forth. (And as Amelie Rorty has remarked, "Beware the virtues of your friends, for they will seek occasions for their exercise.")

Finally, there seems to be an intrinsic conflict between the felt objectivity of moral sentiments and the fact that their presence presupposes specific conditions in the course of each particular person's upbringing.[4]

If our emotions are both defendant, judge and jury in moral matters, true moral progress is emotional progress, governed by a regulative ideal of fully comprehensive, adequate emotional response. But the utopian character of this ideal should induce some skepticism as to the correctness of any particular instance of emotional judgment. It should encourage an ironic stance, that is, one in which one always bear in mind the fallibility of the values one holds. (This is quite different from cynicism, which refuses to hold any.)

Conclusion:

In sum, the "Axiological Holism" that I have been advocating offers a number of features which I have suggested reasons for regarding as advantages:

• It stresses the internal link between emotion and value. Indeed, it goes further, since it recognizes no independent access to value.

• It takes account of the biological utility of emotions without implying that biological utility is more than incidental to moral worth.

• It extends the moral sphere to an indefinitely wide range of dimensions of evaluation, as diverse as are the emotions themselves, when thought of as potentially indefinite in number rather than being limited to "named emotions".

• It rather naturally explains the inherently conflicting nature of values, as generated by the likely modular origins of our emotions in evolutionary history.

• Finally, it allows for both the possibility and need of irony for moral progress.

[1]:In this paper, in conformity with the nature of the invitation from Societas Ethica, I have borrowed from my article in Monique Canto-Sperber's Dictionnaire de Philosophie Morale. The positive thesis I stress here, however, is new, though it is adumbrated in two other forthcoming papers, "Learning to be Natural" and "Emotions and moral progress in black-and-white and colour".

[2]: Contemporary examples are all too easily available. Not long ago I witnessed a television interview of several men who had killed a sister or daughter on a matter of "honour"-that is, because the victim had been raped. These men appeared not to suffer the slightest twinge of ambivalence, let alone guilt, about what their sense of "honour" had moved them to do. Such is the power of social conditioning in closed societies, in which, as this example grimly illustrates, what we take to be a common-sense connection between voluntariness and blameworthiness does not exist.

[3]:Note, however, that this needn't be related to the adaptiveness of altruism for groups. One alternative hypothesis worth exploring, is that the altruism entailed by the existence of natural sentiments of sympathy is a side effect of mind reading. I have sketched this hypothesis elsewhere. It is suggested by the possibility of "mirror neurons" discussed by (Gallese and Goldman 1998), and by the hypothesis that our knowledge of others is principally acquired by some sort of simulation, to which the discovery of mirror neurons appears to lend support.

[4]:In (de Sousa 1987) I called this the "Tragedy of Biography," and listed it among other examples of inevitable ways in which the necessary condition for the flourishing of some good is inherently in conflict with that very flourishing.
 
 

REFERENCES

Bedford, E. 1957. Emotions. Proceedings of the Aristotelian Society 57:281-304.

Campbell, S. 1998. Interpreting the Personal: Expression and the Formation of Feeling. Ithaca: Cornell University.

Darwin, C. 1998. Edited with commentary and notes by Paul Ekman. The Expression of the Emotions in Man and Animals. London: Harper Collins.

de Sousa, R. 1987. The Rationality of Emotion. MIT Press <imprint: a Bradford Book>.

____ .Forthcoming (a). Learning to be natural. In Being Humans, ed. N. Roughley. Available at http://sousa.artsci.utoronto.ca/learnat.html.

____ Forthcoming (b). Emotions and moral progress: in black-and-white and colour. Available at http://sousa.artsci.utoronto.ca/nussbaum.html .

Gallese, V., and A. Goldman. 1998. Mirror neurons and the simulation theory of mind-reading. Trends in Cognitive Sciences
2(12):493-501.

Huxley, J., and T. Huxley. 1969. Evolution and Ethics. New York: Kraus.

Itard, J. 1964. Mémoire et rapport sur Victor de l'Aveyron. In L. Malson, Les Enfants Sauvages: Mythe et Realité. Paris: Union Générale d'Éditions.

Ledoux, J. 1996. The Emotional Brain : The Mysterious Underpinnings of Emotional Life. New York: Simon & Schuster.

Miller, W. 1993. Humiliation : And Other Essays on Honor, Social Discomfort and Violence. Ithaca: Cornell University Press.

Nussbaum, M. 1996. Compassion: The basic social emotion. Social Philosophy and Policy 13:27-58.

Popper, Karl. 1963. The Open Society and its Enemies. New York: Harper and Row.

Sartre, J.-P. 1956. Being and Nothingness: An Essay on Phenomenological Ontology. Trans & introd by H. E. Barnes. New York: Philosophical Library.

Scheler, M. 1913. Zur Phänomenologie und Theorie der Sympathie-Gefühle und von Liebe und Hass. Halle.

Sober, E. and Wilson, D. 1998. Unto Others: the Evolution and Psychology of Unselfish Behavior. Cambridge, MA.: Harvard University Press.

Solomon, R. C. 1973. Emotion and choice. Review of Metaphysics 17(20-41).

Williams, B. 1973. Morality and the emotions. In Problems of the Self: Philosophical Papers 1956-1972. Cambridge: Cambridge University Press.