You are here

Chapter V. Subjectivism

1. There are stages or degrees of subjectivism in ethics. Westermarck stands both logically and historically between the moderate subjectivism of Hume and the extreme subjectivism of the emotivists. Hume held (not quite consistently) that when I call an action right, I mean that people generally have a certain feeling about it. Westermarck held that when I call it right, I mean that I have a certain feeling about it. The emotivists hold that when I call it right, I do not mean to assert anything at all; I am merely expressing a feeling. In discussing these theories, it will be well to reserve the name ‘subjectivism’ for Westermarck's middle position. Although it may be very briefly stated, its force and richness can hardly be understood apart from two facts which will come out in the course of our discussion, and which serve to explain it on its negative and positive sides respectively. One is that it is an ethics of protest and reaction—of reaction against the rationalism and objectivism that we found in the ethics of Sidgwick. The other is that it is rooted in anthropology and is offered as the only theory consistent with all the newly known facts about the diversity of ethical practices. The statement of subjectivism in Westermarck's Ethical Relativity and his application of it to the whole range of human conduct in The Origin and Development of the Moral Ideas form together the most impressive defense of the position that it has yet received.

2. Since Westermarck is in reaction against objectivism, it will repay us to see at the outset what it is that he is concerned to deny. All objectivists agree that when I say ‘that act is right’, I am not merely expressing an attitude of my own; I am asserting that some character belongs to the act or the agent. But what sort of character is this? It is objective; yes, but in what sense? To say that it is objective means that it is independent of something subjective. But this is too vague. Independent of what? Is the rightness that belongs to an act independent of my thought, or of my feeling, or of my very existence itself? Or is it independent of what men generally think, or of what they feel, or of their existence? Or is it perhaps independent of consciousness as such, human, diabolic, or divine?

(1) There have been objectivists who held this latter view. G. E. Moore once held that there would still be good and bad in the world even if every kind of consciousness were wiped out; the world would still be beautiful with no one to see it, and this beauty would be better than ugliness, even if there were no one to think so. Westermarck did not, to my knowledge, discuss this view; but he would certainly have rejected it, since he rejected other forms of objectivism that were far less extreme.

Is the objectivism he is chiefly opposing, then, the view (2) that moral values are independent of human consciousness? At times it seems so. The second sentence of his Ethical Relativity runs as follows: ‘The supposed objectivity of moral values, as understood in this treatise, implies that they have a real existence apart from any reference to a human mind, that what is said to be good or bad, right or wrong, cannot be reduced merely to what people think to be good or bad, right or wrong.’1 This is an unfortunate beginning. For Westermarck seems to be placing side by side, as having the same meaning, two forms of objectivism that are a world apart, and this suggests at the outset a lack of clearness in his own mind. According to the first of these forms, that of which moral values are independent is the human mind as such. According to the second, it is merely people's opinions about them. Now if it is really the first form that he is attacking, his subjectivism is without novelty or point, for the very persons he is most concerned to criticize would themselves accept it. Bentham, Mill, and Sidgwick, for example, would all accept it. They all held that, so far as we know, moral values have no existence apart from the human mind. Right acts, they said, depend for their rightness on effects in the way of pleasure and pain; only those beings who can feel such effects and act in the light of them are capable of right or wrong action. And the only beings we know who can feel and act in this way are men. If men were to disappear, moral values too, for all we know, would disappear with them. Thus if objectivism means independence of the human mind, Westermarck is preaching to the already converted, and his polemic is pointless. It seems hardly credible that he should not have seen this. I think we must say therefore that the singling out of this form of objectivism as what he wished to attack was a mere slip of the pen or the mind.

(3) Is it, then, the other form here mentioned that he is attacking? According to this form, ‘what is said to be good or bad, right or wrong, cannot be reduced merely to what people think to be good or bad, right or wrong’. Westermarck, as we have seen, seems to take this second form as equivalent to the first, but in truth it is extremely different. All three of the philosophers just mentioned, who would deny objectivism in the first sense, would accept it in the second as obviously true. They would say that while, so far as we know, right and wrong do depend on the human mind, they certainly do not depend on what people think about them. If an act causes as much happiness as any act that could have been done instead, then it is right, even if every human mind thinks it wrong; if it does not, then it is wrong, even if every human mind thinks it right. Right and wrong are thus independent of human opinion about them. Is this the objectivism that Westermarck wants to deny?

We have just heard him say that it is. But here again I think he is misstating his aim. It is not really human opinion on which right and wrong in the end depend, for rightness and wrongness in his sense may precede any formed opinion, and even when the opinion is formed, it is itself dependent on something else. Consider the first point. It seems almost certain that men were repelled by some actions, for example wanton attacks on themselves, long before they began to reflect on moral issues and make explicit moral judgments. Would Westermarck say that among such men right and wrong conduct had begun to appear? He would unquestionably. But if rightness and wrongness can thus exist before moral opinion is formulated, they can hardly be said to depend on such opinion. But again, even when the opinion is formulated, Westermarck would hold emphatically that it only reflects something else, which supplies its underlying ground. What is this something else? It is, of course, feeling or emotion.

(4) But whose feeling or emotion? That of the community? Readers have often supposed so, and since nearly all the practices that Westermarck describes in such instructive detail are group mores, it is natural enough to suppose so. But this would be a mistake. Westermarck rejects it explicitly.2 To accept such a view would be to say that when a reformer denounces the practices of the community as evil, his denunciation reflects and expresses the feeling of the group, whereas it seems clear that the feeling he expresses is the very opposite of the group's. Moral appraisals that are sincere express the feeling of the judge himself. Here we come to the end of the line. The objectivism Westermarck is attacking is one that denies the dependence of moral values on individual feeling. The subjectivism he advocates would make the moral qualities of conduct depend on the feelings of the individual persons who contemplate it.

3. Of course this view is at odds with common opinion. To most of us there is something shocking in the suggestion that slavery was really right among the Greeks merely because they felt as they did about it, or that if Hitler had so destroyed opposition that everyone left alive admired him, this would prove his conduct to have been right. Most of us believe that there are some things, for example wholesale and gratuitous slaughter, that are simply wrong, and that if some persons think or feel otherwise, what this shows is not that the act is right but that the persons are perverse or stupid. Westermarck frankly admitted that he had the plain man against him. He admitted, too, that he had the majority of philosophers against him. But he seems to have been as little moved by the one consideration as by the other, and this for the reason that the two modes of opposition are in the end the same; he thought that in ethical matters the philosophers had taken their lead from common sense. The two main questions of ethics are often said to be, What is the meaning of right? and What is the meaning of good? Philosophers have commonly held that the way to answer these questions is to look into our own mind and see what in fact we mean by these terms. And when we do look into our mind, we find that we mean by them something objective, something that actions possess regardless of our attitudes toward them. With this, philosophers suppose that they have proved that right and good really are objective. But this, says Westermarck, is confusion. They have confused what we mean with the truth of what we mean. They have confused the analysis of an assertion with the proof that in the nature of things there is something answering to this assertion. And these are wholly different. Men everywhere may believe in objectivity, they may imply it in all their moral judgments, and yet they may suffer from a general delusion. This is indeed what Westermarck holds. ‘If, as I maintain, the objective validity of all moral valuation is an illusion, and the proposition “this is good” is meant to imply such validity, it must always be false.’3

4. But if plain men and philosophers alike are deluded in this matter, there must be strong and widespread influences at work to account for the delusion. So there are, says Westermarck. The chief factors are three, and they are not hard to discover.

(a) First, there is ‘a very general tendency to assign objectivity to our subjective experience’.4 If we feel threatened by the prospect of rain, we call the dark clouds threatening; if the lightning is a source of danger, we call it dangerous; and these words come to stand for supposed qualities in the object. We experience a sensation of blue when we look at a delphinium, or a piercing sound when we hear a train whistle, or an impression of beauty when we see a rainbow in a waterfall, and proceed to talk of the flower as blue, the whistle as noisy, and the bow as beautiful, though reflection makes it improbable that colour or sound or beauty ever exists apart from the experience of it. If these things are true, we can never argue from the common belief that a quality belongs in an object to the conclusion that it really does belong there, or even that there is such a quality. It is perfectly possible to explain our moral judgments without supposing any moral qualities whatever. We dislike pain; things that cause pain are therefore repellent to us; this repellingness we assign to the thing and call it bad. People act so as to pain us; we call them also bad and their acts wrong. If we go by analogy, the subjectivist explanation here is just as plausible as the objectivist.

(b) Again, there is a tendency to take impressions as objective when they are like those of others. My toothache is private because there is no evidence that, when I am having it, other people have anything like it. But if, when I report that snow is white, everybody else confirms it; if, when I say a rainbow is beautiful, everyone understands and agrees; and if, when I say that lying is wrong, everyone again understands and agrees, I assume that whiteness, beauty, and wrongness are characters in the objects and are presenting themselves alike to all of us. And it must be remembered that, however widely communities may differ about moral rules, within the limits of a given community practically everyone does agree about them. ‘Society is the school in which we learn to distinguish between right and wrong. The headmaster is custom, and the lessons are the same for all the members of the community.’5 The simplest explanation of our agreeing about such things, is to say that just as we all agree about the squareness of a box because it really is square, so we agree about the wrongness of an act because it really is wrong. Westermarck regards the two cases as totally unlike to the discerning eye, but deceptively similar to the undiscerning one.

(c) Finally, there is the immense influence of authority. Parents, teachers, the Church, the Bible, public opinion, the sages of the race, all combine to impress on us that moral principles are true. ‘From our earliest childhood we are taught that certain acts are right and that others are wrong.’6 To question this would be almost impious. We are extremely suggestible, and in a region where nonconformity is very rare, we are bound also to be uncritical. We come to regard the moral law, as Kant did, with the sort of awe that is given to the starry heaven. Conscience seems to ratify its absolute authority. Religion further assures us that it is the expression of the divine will. Though Westermarck was convinced that this whole notion of moral principles as true and objective was an illusion, he thought that, when backed by a weight of authority so overwhelming, this or almost any other illusion could be imprinted upon docile minds.

5. ‘I maintain…’ writes Westermarck, ‘that the qualities assigned to the subjects of moral judgments really are generalizations derived from approval or disapproval felt with regard to certain modes of conduct, that they are tendencies to feel one or the other of these emotions interpreted as qualities, as dynamic tendencies, in the phenomena which gave rise to the emotion.’7 ‘Those who first established the use of these and all other moral concepts felt disapproval or approval and expressed in the concepts their tendency to feel such an emotion in the given circumstances. This is what may be called the intrinsic meaning of the terms.’8 The unsophisticated man does not know all this; he ascribes rightness to acts very much as he does beauty to the rose; but the sophisticated man sees that this is all one has a right to assert. For he sees, as the plain man does not, how moral concepts arise. He sees that they are home-made articles, as truly fashioned out of the stuff of emotion as the concept of fearfulness, and that there is as little presumption in the one case as in the other that any such quality exists. Hence the sophisticated man will correct the plain man's illusion. When he calls an act of courage or justice good, he will know what he is doing. He will understand that the moral quality lies not in the act but in his own feeling, and that all he can sensibly mean is that the act gives rise to approving feelings in his mind.

6. An interesting consequence follows from this reinterpretation of the moral judgment. We have seen that as the plain man makes it, this judgment is always false. But as the sophisticated man makes it, it turns out to be all but infallible. When he says that generosity is good, he means that generosity tends to arouse in him favourable feelings, and if it does arouse such feelings, the judgment is true.9 Now he is in a better position to know his own feelings than anyone else in the world. Since his moral judgments are in essence reports of these feelings, and it is hardly possible for him to mistake favourable for unfavourable feelings in his own mind, it is hardly possible for his judgments to be mistaken.

7. Another odd consequence follows from this view, namely that there is no such thing as a science of ethics in the traditional sense. Ethics has sought to tell us what ends are really good, that is, what all men should try to gain, and what sort of conduct is really right, that is, what men generally ought to do. We now see that, as commonly meant, no such assertions can be true; there are no possible answers to questions of this kind, since the answers, if they came, would be nonsense. I can say, of course, that all men ought to be honest, but this turns out to mean only that honesty wakes favourable feelings in me. Or my statement may be a descriptive generalization to the effect that all men have such feelings about honesty. But then this is not an ethical statement at all; it says nothing about what is good or right or obligatory; it is a statement of psychological fact. Once I have seen clearly how my value concepts and feelings are related, I can see that a general ethical statement is a contradiction in terms. The claim it makes to universal validity is a claim that in terms of its ‘intrinsic meaning’ can be valid only for me. An ethics that knows its business will therefore give up moral legislation for the description of fact; it will become an adjunct, though a useful adjunct, of psychology and anthropology. Ethics as a normative science, pursuing the summum bonum with its butterfly nets of speculation, formulating solemnly ‘the whole duty of man’, codifying the universal rules of human conduct, will have to go.

8. We have now seen what forms the predicate of the moral judgment, that which is asserted, and it turns out to be a tendency to arouse in us certain feelings. What now of the subject? What is it that we are asserting of? We have been talking as if it were an act, but ‘act’ is a treacherous term. If A robs B, is the act that we disapprove the mere physical process of going through B's pockets, or this plus A's intention to do so, or both plus his motive, i.e. that which moved him to form and execute the intention, or all of these plus the character out of which the motive sprang? Westermarck argues, convincingly I think, that it is all four of these things, and that their importance increases as we go back along the line. Certainly the physical act alone cannot carry the burden of odium; if these identical motions were made by an automaton or a hypnotized man, we should not feel about them as we do. So we must include the conscious intention of the man who does them, the will to bring them about. But we cannot stop here either. If he were compelled to do them under threats upon his life, he would still be willing them, but his motive—to save his own life—would make his ‘act’ extremely different from that of a footpad. We must go on, then, beyond the motive to the character, the set of habitual motives that governs the life. A sudden impulsive breaking over of the traces is one thing; conduct that springs from deliberate and persistently repeated choice is another, and we feel far more strongly about it. Indeed in Westermarck's view it is character that carries the chief weight of praise or blame. There is a well known injunction to the effect that our righteous indignation should be reserved for the sin and should spare the sinner. Westermarck considers this unnatural and impracticable, for the reason that moral emotions are in the nature of the case ‘reactionary attitudes towards living beings’.10 We do, to be sure, pass judgments on what a man does, but we do so in the light of what he intended to do, what moved him to do it, and, above all, whether it expresses the man himself—whether it springs from the settled bent of his will.

9. Granting now that some acts arouse favourable feelings and some unfavourable, why the difference? If we see a hold-up in the street, and then, after a horrified moment, see a cameraman in the offing grinding out what is obviously a scene for a motion picture, our mood changes abruptly; why? Westermarck would answer; because one is harmless, the other not. Or, more formally, there is a wide difference in the two situations in their tendency to cause pleasure and pain. Moral approval is ‘a kindly attitude of mind towards a cause of pleasure’; moral disapproval is ‘a hostile attitude of mind towards a cause of pain: we approve a person who causes pleasure and condemn one who causes pain’.11 It would have been better, perhaps, to say ‘one who is believed to cause pleasure,’ etc; do not persons often approve what causes an excess of pain, a declaration of war, for example, because they think it will cause the opposite, and disapprove conduct that causes a surplus of pleasure, for example the stern discipline of a parent, because it appears to be more productive of pain? Probably Westermarck would accept this. He points out, indeed, that many a line of conduct has been approved or condemned not because of its actual consequences, but because of theological teaching about it; ‘the view that celibacy is the height of moral perfection is simply the outcome of some specific religious and magical beliefs of a rather primitive character’.12 There are of course many other questionable ways in which conduct may acquire an aura of harmfulness or the opposite. But the point of ethical importance is that approval or condemnation is dealt out to such conduct in the light of its assumed productiveness of pleasure or pain.

Pleasure or pain to whom? To the person himself who passes the judgment, to others, or to both? On so fundamental a question, Westermarck might well have been at more trouble to be clear. On the one hand he denies emphatically that the moral judgment is based merely on the consequences of behaviour to oneself; it is disinterested in its very essence.13 The ground of his moral condemnation is that the act tends to cause pain as such, not pain to A rather than B, or to himself rather than another. On the other hand, he holds with equal definiteness that it is our own pleasure or pain on which these judgments are grounded; ‘our retributive emotions are, of course, always reactions against pain or pleasure felt by ourselves’.14 Sympathy and altruistic sentiment may make us spare the feelings of other people, and hence if we feel satisfaction or resentment at an action, this may well be because other people's pleasure or pain has excited our own. But Westermarck seems to take it as obvious that whatever the remote causes of our moral feelings and judgments may be, their proximate cause and ground lie always in our own pleasure or pain.

10. It is natural that if moral concepts reflect emotions, moral judgments should not be a matter of black and white, but of degree. The degree of wrongness, for example, will depend on the intensity of our indignation. This again will depend on further factors such as the amount of pain caused, and whether those affected are near or far from us. Westermarck thinks this implication of his theory an effective argument for it. For truth, he says, is not a matter of degree, and if in our judgment we were really ascribing a predicate, wrong, to the act, that predicate must either belong to it or not, and the judgment would be wholly true or wholly false. But he insists that this is not the way we think about wrongness; it is always wrong in degree, the degree answering precisely to the intensity of our feeling. Hence the accord between fact and theory is most happy.15 (Unfortunately, the argument would seem to prove too much. It would equally show that brightness, loudness, and heat are not qualities that we assert, because they too are always experienced in degree; whereas it seems clear enough that they are such qualities.)

11. His theory here comes close to some other well-known theories, notably those of Mill and Hume, and it may be well to mark precisely the divergence. According to Mill, to judge an act wrong was to say that it produced a smaller net pleasure, or a greater net pain, than some other act that might have been done instead; and the truth of the judgment depended on a chain of hedonic consequences that took into account everyone affected by the act through a possibly endless series. According to Hume, it was not the consequences, but men's attitude towards the consequences that made an act wrong; to judge it wrong was to say that in virtue of producing less pleasant consequences than some alternative, it aroused in those who contemplated it a feeling of disapprobation. The truth of the judgment thus depended on whether most people did in fact feel about it in this way, and as Broad has remarked, the issue could in principle be settled by statistics. Hume's is thus a feeling-theory, and he is obviously closer to Westermarck than Mill is. Where does Westermarck differ from these predecessors? He would hold that to judge an action wrong is not to say with Mill that it produces less pleasure, nor with Hume that in consequence of this it arouses general disapprobation, but that in consequence of a pain in me of which I believe it to be the cause, it tends to arouse in me the feeling of disapproval. The truth of the judgment thus depends not on how people generally feel, but on how I feel; ‘if I am right,’ says Westermarck, ‘in my assertion that the moral concepts intrinsically express a tendency to feel a moral emotion of either approval or disapproval, it is obvious that a judgment which contains such a concept may be said to be true if the person who pronounces it actually has a tendency to feel the emotion in question with reference to the subject of the judgment’.16

12. Since moral concepts thus depend on moral emotions, it is important to see what these emotions are. Their place among emotions generally will be made clearer by the diagram on facing page. First we divide emotions generally into retributive and non-retributive. ‘Retributive’ in this usage has no special connection with punishment or the requital of evil; emotions are retributive when they are a response either to an injury suffered or to some service rendered. Since all moral emotions fall in this first class, we may dismiss the non-retributive emotions from further notice. Retributive emotions fall naturally into two classes, negative and positive, those that consist in some form of resentment for injury suffered and those that consist in ‘kindly emotion toward a cause of pleasure’. But these retributive emotions are not always moral emotions. Anger clearly is not; there is no moral complexion about a mere burst of anger at a blow or an insult. On the other hand, indignation toward the cruelty or treachery that is practised on another has the moral tone unmistakably. The same distinction may be made within those retributive emotions that are ‘kindly’. If a man feels grateful to someone who has just pulled him out of the river, there is not necessarily anything moral about that. But, if seeing a courageous rescue by another, he feels an appreciative glow, that again has the moral character. What is it that marks these retributive emotions as moral? Westermarck replies that it is impartiality. If the emotion is aroused merely because pleasure or pain has been occasioned to oneself as opposed to someone else, it is not moral, as we have seen; if it is disinterested, it is. Of course it is not always easy to tell whether one's emotion is moral or not. The best test lies in an imaginary change of relationship;17 if, and so far as one would feel the same way if the insult were addressed not to oneself but to another, and the person pulled from the river were not oneself but a stranger, one is feeling genuine moral emotions. These emotions bear special names. What one feels in the first case is the emotion of disapproval, in the second, approval or approbation. These terms, so often used indiscriminately of feelings and judgments, are strictly reserved by Westermarck for the two retributive moral emotions. And these two emotions are in his view all-important, for if an action arouses approbation, it is ipso facto right; if it arouses disapproval, wrong.

13. From the two root terms, approval and disapproval, both of them names of emotions, Westermarck derives the other main concepts of ethics. But the two are not of equal importance. The notion of resentment determines a far larger range of ethical terms and judgments than the notion of kindliness. Take the meanings of ‘right’ and ‘wrong’. One would naturally expect that what makes an action right is its arousing kindly emotions, just as what makes it wrong is its arousing moral resentment. This turns out to be untrue; resentment is the basis of both concepts. As regards wrong the argument is already clear, but consider right. Do we in practice confine the term to those acts that arouse in us benevolent or kindly feeling? Westermarck says no. The truth of the matter is rather this: moral rules are generally customs, and customs come in time to have something of the force of laws, that is, of commands by the community to its members. Disobedience to such commands is felt as a flouting of the community's will and naturally wakens resentment. But in mere conformity to custom there is nothing to excite feeling one way or the other; it is the norm, the established practice, what everybody expects. A man who abuses his parents is running against the current of custom; his conduct will be called wrong, and such wrongness implies felt reprobation. But when we say that the man who treats his parents in the normal way is acting rightly, do we have any corresponding glow of active approval? No. ‘Right’ then does not mean approved; it means merely not disapproved. It follows that ‘right’ and ‘wrong’ are not really correlative terms. ‘Just’ and ‘unjust’ are dealt with by Westermarck in similar fashion.

The terms ‘ought’ and ‘duty’ must also be redefined. What is it that makes an act one's duty, that which one ought to do? Objectivist writers have often found this in the perception that one act rather than others tends to the general good. ‘Strictly speaking, however,’ writes Westermarck, ‘“ought” and “duty” only express the tendency of an act's omission to call forth moral disapproval and say nothing about the consequences of its performance.’18 Thus ‘duty’ is like ‘right’ in drawing its substance from the capacious well of human resentment. It cannot derive this from kindly feeling, for too plainly we do not always think of our duty with a glow of positive emotion. Hence if it is to have any ethical meaning at all, this must come from moral antipathy. Antipathy to what? Not to the duty itself, or it would presumably never get done, but to the non-performance of duty. ‘Every “ought”—judgment contains implicitly a prohibition of that which ought not to be done.’19 To say that anything is one's duty is thus to report an unfavourable feeling about the failure to do it.

Though hostile feeling is the most fertile of the moral emotions, not all ethical terms owe their meaning to it. ‘Good’, ‘virtuous’, ‘meritorious’, reflect the emotion of approval. ‘Good’ is properly applied not to an act but to a person, and it implies a kindly feeling towards him as a cause of pleasure.20 The virtues, such as temperance and courage, are merely dispositions to be good in particular ways. When we call a man meritorious, we are expressing the approbation we feel for the good man, but also something more; we are saying that he ought to be approved, which means that if his goodness passed unrecognized, this would meet with our disapproval. It is worth noting that in Westermarck's theory another term gets a meaning from kindly emotion which in many ethical theories would be set down as meaningless. This is ‘super-obligatory’. If duty means only that whose omission would arouse indignation, to go beyond one's duty becomes both possible and easy; we all do things from time to time which we should not be blamed for omitting, even perhaps by ourselves. When we do these, we are going beyond our duty in Westermarck's sense; our action is super-obligatory.

14. We now have the main features of Westermarck's theory before us. Unfortunately, the most persuasive of his arguments for it is one that cannot be presented with its full force in any summary statement, because that force depends on a vast mass of inductive evidence. The argument is based on the facts of moral diversity. It is not, of course, new. That men differ widely in their moral convictions and practices was perfectly well known to those who have held objectivist theories; Sidgwick was much interested in ‘history studied as inductive sociology’;21 Hegel's attention was arrested by the variety of cultures, as indeed by nearly everything else; Kant was a student of anthropology; Plato must have known his Herodotus; and the conclusion so strongly suggested by the diversity of men's moral judgment was no doubt carefully weighed by all of them.

That conclusion is that in morals there is nothing objective to know. If there were, it is argued, there would be some agreement on moral issues, whereas in fact there is anarchic diversity. Men do not differ about the multiplication table; they all agree that two and three make five; and if they do, it is because the laws that govern the relation of numbers are universal and independent of the will. Nor, except on the speculative side, is there any substantial difference about chemistry and physics, for there is always an arbiter standing ready, to whom passing disputes may be taken, an arbiter whose decisions are final, namely nature. If we study such agreement, we shall find that what makes it possible is always an independent law or structure through comparison with which our personal deviations may be cancelled. Now if agreement depends on objectivity, general disagreement strongly suggests the lack of it. And in ethics such disagreement is precisely what we find. Indeed we find something like chaos. There is no custom, however sacred, and no principle, however compelling, that is not somewhere put aside as plainly unsound. How are we to explain this except by saying that the objective principles that we find in other sciences do not exist in ethics? If they did, surely there would be some signs of order, some suggestion of an agreed-upon set of principles, emerging from the chaos.

To this the objectivists have replied that the diversity can be explained in a much simpler and less violent way. This is the way suggested to the then youthful Westermarck by Sidgwick in a conversation to which we have referred.22 Why not account for the diversity by the very simple hypothesis that some men are stupid in moral matters and others more discerning? Indeed if agreement is to be made out in science itself, it would seem to be only by taking advantage of this sort of explanation; we tacitly rule out the dissenting votes of savages on mathematics and chemistry as not worth including in our poll. If we did include them, we should have a diversity very similar to that in ethics. But it is clear how we should go about it to deal with that diversity. No one would question a judgment of Newton on the ground that some Andaman Islander shook his head over it. We should say that if the Islander failed to see, that was not because Newton's insight was false, or indeed anything less than self-evident to a mind qualified to judge, but because the Islander was stupid or untrained. And if we are entitled to say that in science, why not in ethics? The civilized man sees clearly, or thinks he does, that to rob a stranger is wrong, and that to deprive a man of rights because he is of another colour or country is unjust. To most savages such views are absurd. Does this show that fairness is really no better than unfairness? Is it not far more sensible to say that fairness really is better, but that the savage has not reached the point where he can see it? To Sidgwick, perhaps the clearest-headed of all moralists, one could in this manner hold firmly to the objectivity of moral law while admitting to the full the diversity of moral judgment.

15. How does Westermarck reply? He replies by denying the parallel between scientific and ethical diversity. In science all who apply themselves impartially to the subject recognize certain persons as experts, and these experts on the whole agree. In ethics the experts themselves disagree. The only way one can get agreement is to set up one's own insights as authoritative and admit to the rank of expert only those who see eye to eye with oneself, a method that remains unconvincing in spite of its wide popularity. Short of this high-handed procedure, one can only concede, says Westermarck, that there is not a single proposition that qualified moralists have regarded as self-evident which is not held by moralists of equal standing to be untrue. He is ready, as one would expect, to supply chapter and verse in profusion. ‘Some “moral specialists” say it is an axiom that I ought not to prefer my own lesser good to the greater good of another; whilst others not only deny the self-evidence, but thoroughly disagree with the contents of this proposition. According to Sidgwick the assertion that pleasure is the only rational ultimate end of action is an object of intuition; according to Moore, the untruth of this proposition is self-evident. The latter finds it self-evident that good cannot be defined; but others, who have no smaller claim to the epithet “moral specialists”, are of the very contrary opinion.’23 These are not bones of minor contention at the bottom of the cultural scale; they are divergences at the very top, and concern matters of the first moment. The theory, then, that would explain the variety of ethical insights as approximations by minds of differing level to a common objective truth does not cover the known facts.

Is there any theory that does? One only, Westermarck holds. It runs as follows: Assume that what all men primarily want is to gain pleasure and avoid pain, and that approval and disapproval themselves consist of emotion. Assume finally that men's perception of the causal connections that actually hold is wavering and uncertain, so that it is likely to be distorted by fear and chance association, by superstition and desire. You will then have a theory that, without any need for objective moral truths, covers the facts completely.

16. As he develops this theory in his great work, two features gradually come to light that must deeply affect our estimate of it. The first is the somewhat surprising disclosure that the supposedly enormous divergences are, after all, less moral than intellectual. Westermarck goes down the list of the more important mores and finds in one after another that what men chiefly differ about is not ends but means, not the importance, for example, of the happiness, strength, and health of oneself and one's immediate family, but about the means by which these are to be achieved. And he recognizes that a divergence about means is not a moral divergence at all. Take a well known case in which the difference does seem to be moral and also to be so wide that it is beyond all bridging. Among ourselves kindness to parents is approved and demanded. On the other hand, Westermarck notes more than a dozen peoples among whom approval is actually given to the killing of parents.24 The moral difference here seems deep and shocking. But as the circumstances unroll, this difference steadily fades. Among tribes where physical strength is all-important and weakness is misery, where the food will not go round, where the group must keep on the march, and where, above all, it is firmly believed that men retain in the next life the mental and physical faculties they take with them at death, ‘what appears to most of us an atrocious practice may really be an act of kindness, and is commonly approved of, or even insisted upon, by the old people themselves’.25 To us the shocking thing about such behaviour is the indifference it seems to show to the life and happiness of those who are normally most dear. But no such indifference is necessarily involved; in some of the cases ‘the son will kiss and weep over his aged father as he prepares him for the grave.…’26 No doubt as one goes down the cultural scale, there is a falling off of imagination and sympathy just as of other capacities of mind. But if a savage, in doing what he does, thinks he is serving the best interest of his parent and the community, it is idle to say that he is living in a wholly alien moral world. He is living in the same world because seeking substantially the same ends as ourselves, though in ways imposed upon him by a harder economy and a more uncritical religion. Thus by Westermarck's implicit admission, the very research which casual readers suppose to have demonstrated moral chaos shows that there is an impressive unity and stability in man's moral nature, and that if we have regard to the grounds on which its approbation is given we shall find in its verdicts a smaller variation than in his intellectual beliefs.

17. The second feature is more important still; and here appears what seems to me an almost incredible omission in Westermarck's ethical system. To approve something, he holds, is to feel a favouring emotion toward it as a cause of pleasure; to disapprove it is to feel an opposite emotion toward it as a cause of pain. Pleasure and pain are thus the grounds, and the prospect of them the cause, of all moral emotion. It is not because acts by themselves have any value for us that they arc approved or disapproved; it is because of what they lead to, because they are means to something else; and this something else which alone gives importance to the means is pleasure or pain. These and these only are what in the end we prize or disprize; pleasure is what all men everywhere want and always have wanted; pain, what they have sought to avoid. It would seem clear, then, that the question whether an act is objectively right or wrong would depend on whether pleasure and pain themselves have objective value. If, when we say ‘pleasure is good’, we are saying something objectively true, true independently of what we may be feeling, then the judgment ‘this act is right’ is also objectively true. Hence it is of the first importance for Westermarck to show that the judgments we pass upon pleasure and pain are as lacking in objective truth as those we pass upon acts.

Now what I find almost incredible in Westermarck's treatment is that on this issue, so clearly fundamental to his whole construction, he has nothing to say. He passes it over as if the problem had never occurred to him; What he would say about it if he did raise and discuss it one can only conjecture. But he would need to say something very convincing, for the view that pleasure and pain are themselves without objective value is far from plausible. Let us try to think of the most probable and plausible things that he might say.

He might dismiss the problem by observing that ‘pleasure is good’ and ‘pain is evil’ say nothing at all, that ‘good’ means pleasant and ‘bad’ means painful, and that such statements as ‘pleasure is pleasant’ and ‘pain is painful’ would obviously not be worth making. I will not linger over this suggestion because I think it has come out clearly enough in the long debate about hedonism that we do not mean by ‘good’ simply pleasant, nor by ‘bad’ painful. Two remarks will perhaps serve. First, as has been shown by G. E. Moore, if ‘good’ meant simply pleasant, the question whether pleasure is good ought to be meaningless, whereas in fact it is not. Secondly, ‘good’, even when used of intrinsic goods, has a far wider application than ‘pleasant’; we call pleasure good, but we say the same of knowledge, love, and justice, when it seems clear that it is not their pleasantness that we are concerned to stress. One can hardly cut off discussion of the value of pleasure and pain by dismissing the problem as meaningless.

But Westermarck might dismiss the problem in another way. He might say that he was dealing with moral judgments alone, and that judgments about the value of pleasure and pain were not moral. Verbally, this would be correct. In his view, it is always on persons or acts, or more accurately on both together, that moral judgments are passed; pleasure and pain are obviously neither. Nevertheless to take this line would be to evade the real problem. Suppose we agree that when we call conduct right, our judgment is moral, and that when we call pleasure good, our judgment is non-moral; still on Westermarck's theory it can only be because we accept the latter that we ever accept the former; unless we took pleasure to be good, we should not approve acts that produced it. Hence it would be odd to discuss elaborately whether our judgments on acts were objective, and refuse to discuss, as irrelevant, the objectivity of the judgments we pass on the consequences of those acts. We may call such judgments on consequences non-moral if we like, but if the force of our moral judgments depends on them, then they are moral in effect, whatever we choose to call them.

Very well, suppose Westermarck did raise the question whether the judgments of value that we pass on pleasure and pain are objectively true or not; what would he say? One would perhaps expect him to say that they should be dealt with in the same way as moral approvals. If we follow this clue, the statement ‘pleasure is good’ will mean ‘pleasure is something that tends to arouse in me an approving emotion in virtue of its being thought of as a cause of pleasure’; ‘pain is bad’ will mean ‘pain is something that tends to arouse in me a disapproving emotion in virtue of its being thought of as a cause of pain’. Clearly this will not do. If we approve of pleasure merely because it is a cause of later pleasure, we shall approve of this later pleasure only because it is a cause of a still later pleasure, and we shall have no ground for approving any pleasure at all until we have completed an endless series, that is, never. On the other hand, if one stops anywhere along the line with a pleasure conceived as self-justifying, then there is precisely the same reason for stopping at the first term, and making our approval of that pleasure depend on something in the pleasure itself.

‘Pleasure is good’ is now to mean ‘pleasure causes in me or tends to cause an emotion of approval’, and ‘pain is bad’ to mean ‘pain causes in me or tends to cause an emotion of disapproval’. Now that pleasure does normally cause favouring, and pain hostile, emotion may well be true. But that what a sophisticated person means by calling them good or bad is some fact of causation seems to me plainly false. Historians tell us that in the fifteenth century, many thousands of women were burnt at the stake as witches. When we think what this meant, even in a single case, we say that the suffering involved must have been horrible. Is it our intention, in such a remark, to say something about the character of the suffering or merely to report its causal action on ourselves? On the theory before us it is the latter. On this theory we are asserting a chain of causation leading from certain events in the fifteenth century to occurrences in our own minds. If this is really what we are doing, then the question whether our judgment is true, the question whether the experience of being burned alive in the fifteenth century was good or bad, will depend on the effect which that event has on us now. It seems to me perfectly clear that this is not our meaning. By way of test we may put to ourselves the question whether we mean to say of the suffering something that would still be true even if we ourselves were not alive. We should surely answer that we do, that whether we were alive or not could not make the slightest difference to the goodness or badness of events that occurred in the fifteenth century. If so, it cannot be the action of those events upon ourselves that we are primarily asserting. The badness asserted is something we take to have belonged to the suffering at the time when it occurred.

The case is perhaps clearer still of judgments of the future. Suppose a man makes a will in which he leaves a substantial amount to a small grandson of whom he is fond. If asked to explain this, he might say, ‘I want the boy to be happy’. If pressed as to why he wanted him to be happy rather than miserable, he would probably say, ‘Don't be silly; it is surely a good thing to be happy; one doesn't need to argue about that.’ If it were then said to him that the boy's future happiness could not possibly have any value, because its value depended on its causal action on himself and he would be dead before it came about, he would regard this as stuff and nonsense. Probably most of us would agree. What he is aiming at and talking about is plainly a future good, that is, something that will be good in the future. And if this is an intelligible aim at all, the goodness cannot possibly consist in a causal action running from the future to the past.

18. What we have here essentially is old-fashioned egoism in a new guise. Pleasure is good, and pain bad, but only in their relation to me. To say that the suffering of someone else, which does not affect me, is bad is meaningless, since badness consists in effects on me; to call such suffering bad would be like saying of an armless man that he is awkward in shaking hands. Similarly, if other people's happiness does not in some way act on me, there is literally no sense in my calling it good. One may define goodness and badness in this fashion if one pleases; but certainly neither the plain man nor the reflective man would agree that it catches his meaning, since it would commit him in an obvious way to what he would reject as completely unreasonable. Suppose the view were to be generally accepted. Then, just as I should attach no value to other people's pleasure or pain except as it affected mine, so they would attach no value to mine except as it affected theirs. And I do not believe that anyone thinks of good and bad in a way that would require everyone else to regard his happiness or misery as something in itself of no account. To be sure there is nothing formally inconsistent in this sort of egoism. ‘I can find nothing irrational or inconsistent’, writes Westermarck, ‘in the proposition that it is right for everybody to be an egoist, myself as well as others.’27 There is no contradiction in your calling good what I call valueless if what each of us means by these terms is merely a relation to his own feelings. But when Westermarck goes on to say that in such egoism there is nothing irrational, we begin to feel uneasy. Why? Perhaps we may put it this way: if I were to ask myself why I regarded a certain state of mind in myself, for example a severe pain, as bad, while regarding as good or neutral an exactly similar pain in another, I could see, if I were honest, that its special relation to me had nothing to do with its goodness or badness, that what made it bad was its character, the sort of experience it was, its being an intense pain, not the accident of its happening to one man or to another. The unreasonableness lies in making essential to goodness and badness a relation that is plainly not essential. The trouble with such egoism is not that it is inconsistent, but—if one may say so—that it is stupid.

19. If Westermarck is to escape such clashes with our plain meaning, he must move on into another position. He must say that the value of pleasure and pain depends not on their relation to me, but on their relation to whoever feels them. When I call my happiness good, I still mean that I have a favourable feeling toward it, but when I say that your happiness is good, I mean, not that I have such a feeling toward it, but that you have. The misery of the women burned as witches was really bad if they felt in a certain way about it, whether the people around them felt this way or not, and regardless of whether you and I ever lived. This is a far more plausible view than the last, but still it will not serve. For, first, though pain may be made worse by our emotional attitude toward it, one can hardly believe that its evil is wholly created by this attitude. And this is still harder to believe of the pain of animals. I am sure that if my dog were struck by a car and his leg broken, he would suffer intense pain, and I am sure that the occurrence of this pain would be an evil thing. But it seems extremely questionable whether the pain would become bad only when and because the dog took up an unfortunate emotional attitude toward it, and that unless he did, there would be nothing bad about it at all. Pain exists, and is surely bad, at psychical levels at which this sort of analysis would be highly artificial.

Secondly, even if the analysis were adopted, Westermarck could not consistently accept it. It would destroy his subjectivism. For the goodness and badness of consequences no longer depends on what I feel about them or even whether I exist. I approve of an act, he says, in virtue of the pleasure it produces. But the goodness of this pleasure does not depend on me; it is now admitted to have value, whatever my attitude may be, if only it is welcomed by the person who experiences it. Such independence of my feeling is, as we have seen, what Westermarck means by objectivity. On such a view, not only are good and bad objective; we can lay down those universal moral laws which are one of his chief aversions, for example, ‘Always so act as to produce the greatest total good.’ This will now mean almost exactly what Mill meant by it, and it will give a yardstick of right conduct that can be applied in the same sense everywhere and always.

Let us sum up what we have been saying. According to Westermarck, the rightness of an act depends on its approval, and its approval turns on its believed conduciveness to pleasure or pain. We have seized on this latter point, and held it to be implicit in Westermarck's argument that the justification of the approval rests on the truth of the belief. If an act does really conduce to the largest net pleasure, must he not, on his own implicit premises, call it objectively right? That some acts do really conduce to pleasure or pain is clear enough. But might he not save subjectivism by pointing out that even if they do, the pleasure, and pain they produce are themselves without objective value? We have asked whether there is any way in which he could consistently maintain this view, and we have reviewed the only ways that seem plausible. None of these was at all convincing. Our conclusion is that Westermarck leaves the objectivity of moral judgments very much where it was. It is still open to us to say, for example, that some acts are objectively right because they do in fact produce results that are objectively good.

20. He seems indeed to admit this and even to make it, in a half conscious way, the basis of his whole treatment of morals. When conflicting customs are equally approved by the groups that practice them, he certainly does not put them all on a level; he thinks that some are better than others, and he evidently considers his judgment that they are better to have a ground beyond his own feeling. On this point Rashdall remarks: ‘Professor Westermarck, though he is generous to the savage, and perhaps underrates the difference between his morality and ours, never exhibits the slightest real doubt that his own ideal of life, when it is different from that of the savage, is the higher of the two.’28 We have seen that in his view the morality of the savages who kill their parents is not as remote from our own as appears; but even so, it seems clear that he would not place it on the same level. These people were misled by superstition as to the true consequences of their conduct; we are not; and with our release from that superstition we can see things as they are and determine our conduct in the light of truth rather than illusion. If happiness is, as he holds, what everyone wants, is it reasonable to say that conduct which produces this is objectively no better than conduct that does not? Verbally this is what he is maintaining; the fact is that he abandons it within the first three pages of his great book. He writes:

‘in every society the traditional notions as to what is good or bad, obligatory or indifferent, are commonly accepted by the majority of people without further reflection. By tracing them to their source it will be found that not a few of these notions have their origin in sentimental likings and antipathies, to which a scrutinizing and enlightened judge can attach little importance; whilst, on the other hand, he must account blameable many an act and omission which public opinion, out of thoughtlessness, treats with indifference.’29

One who could write this does not really believe that the emotional approval of a line of conduct closes the question of its rightness. He clearly believes that the approval which springs from ‘sentimental likings and antipathies’ is less likely to be sound than that of ‘a scrutinizing and enlightened judge’ who is able to engage in ‘further reflection’. And what is the point of such reflecting if not to reveal the truth about the act, in the sense of its actual consequences, the contribution it really makes to human welfare? Thus it seems to have been Westermarck's conviction, whatever his theory required, that conduct dictated by these correctly seen consequences was better—really and objectively better—than conduct chosen in ignorance of them. A main purpose of his book is to enable us to achieve this more rational conduct, a purpose, we may add, that he excellently fulfilled.

Westermarck is of course quite aware of this fine of criticism, and reminds us that it has been made against him by G. E. Moore. ‘He says it is commonly believed that some moral rules exhibit a higher morality than others, and asks what I could mean by saying that A's morality is higher than B's. He himself gives the answer: I could only mean that “A's morality is my morality, and B's not”.’30 In spite of this frank avowal, it is hard to believe that this is all Westermarck means. He argues that an advantage of his own view is that it would lead to greater social tolerance and more general self-criticism.31 These things then possess value. But is Westermarck really saying that they have value only in the sense that he has a certain emotion toward them, whose absence would render the judgment false? Would his judgment no longer be true if he suffered a paralytic stroke? Surely he means more in saying such things than his theory allows him to mean. As Arthur Murphy has remarked:

‘the relativist is committed in advance to a refusal to deal with moral issues on a moral basis. He dare not make moral judgments except by disguising them as statements of existing matter of fact. Since the whole point of adducing the variety of codes and customs as a ground for tolerance is to provide a reason for one course of conduct and a basis for condemnation of its intolerant opposite, the result of this procedure can only be confusion.’32

21. It is time to turn from Westermarck's implicit to his explicit emphasis. He says little, as we have seen, about judgments of intrinsic value: he says a great deal about judgments of right and wrong; his main point about these is of course that they are statements of our own emotion. We do not feel approving emotions toward conduct because we see it to be right; we judge it to be right because we feel these emotions toward it; indeed to the self-critical mind, the judgment is simply the statement that the conduct arouses (or tends to arouse) emotions. This gives us Westermarck's views on two very important matters. It tells us first what he thinks about the psychology of moral judgment, about the respective parts played in it by thought and feeling; and it tells us, secondly, what he thinks is the meaning of the judgment, what a reflective mind means to assert in calling an act right or wrong. The two topics are very closely related. But it will be convenient to consider them separately and to begin with the first.

Westermarck's theory of moral judgment is often called an emotional theory. But we have seen that he never accepted the view held by the more extreme positivists that such judgment expresses nothing but emotion. It is not an exclamation merely; it is genuine judgment, with its own truth or falsity. Still, as made by the plain man, he holds that its concepts are bogus concepts, like those of the ‘frightfulness’ of a storm or the ‘frowning’ aspect of a cliff. Though they pretend to be, and are commonly taken to be, notions of attributes in the object, they are really reports of how we feel. The sophisticated man will see this. There will still be an intellectual element in his judgment, but it will be confined to two things, first an apprehension, usually implicit and vague, of the act as a cause of pleasure or pain, second, an apprehension, which is the essence of the judgment, that it tends to arouse approving or hostile emotion in his own mind. Westermarck's case, as we have noted, is curiously incomplete about the first of these elements. But on the second he is explicit and emphatic. He holds that while the judgment of approval is not reducible to a mere exclamation like ‘Cheers!’, it is reducible to a statement that I feel, or tend to feel, as I should if I did so exclaim. The intellectual element has not quite vanished. But it approximates the vanishing point, since it derives from the emotional element the whole of its distinctive meaning.

Now we must surely agree that feeling is bound up with moral judgment in manifold and intimate ways.33 When we have granted this, have we conceded Westermarck's point that the moral judgment is essentially an expression of feeling? No, I think not, and he finds it impossible himself to adhere to it consistently. He cannot fit one of the most important of ethical judgments into an emotional theory. He recognizes that while the judgment that something is wrong involves hostile emotion, the judgment that something is right seems to involve no corresponding emotion of approval; ‘… we regard it as perfectly right to pay a debt and to keep a promise, or to refrain from killing, robbing, or lying, though these acts or forbearances have no tendency whatever to evoke in us an emotion of moral approval’.34 What are we to make of this? We have been told that to use moral concepts is to assert ‘tendencies to feel either moral approval or disapproval with reference to that of which those concepts are predicated’.35 We are now told that the judgment of right asserts no such feelings or tendencies to feel. It follows that the judgment of right is not a moral judgment at all. Since everyday right acts excite no emotion in us, one way or the other, our moral judgments about them must, on the emotional theory, have no content. I find this very hard to accept. It is not easy, of course, to be sure what we do mean by ‘right’, but it is difficult to believe that we mean nothing by it. Most people would think it better to start by assuming that a term so long regarded as fundamental did mean something, and would say that if a theory had no room for it, so much the worse for the theory. Westermarck is prepared to drop the concept rather than rewrite his theory to suit its actual meaning.

He does not, indeed, think it necessary to call the judgment wholly meaningless. He attempts to supply it with emotional meaning by connecting it with disapproval. When we say that an act is right, we are saying that the failure to do it would excite such disapproval, and since disapproval is an emotion, the term after all draws from emotion such significance as it has. I do not think this defence will serve. It still denies to the judgment of right any distinctive meaning. It holds that when we say an act is right, we really mean that some other act is wrong. ‘Chamberlain's course at Munich was right’ means ‘Chamberlain would have been wrong in not taking that course’. But these judgments, as commonly made, do not mean the same thing. This is clear from the fact that the first can be true while the second is false. When a man says that Chamberlain was right, he is certainly not committing himself to the view that all alternative courses would have been wrong; it is often the case, so far as we can see, that any one of a variety of courses would be equally right. The attempt, then, to reduce the judgment of right to an expression of disapprobation for not acting in a given way does violence to our meaning. And unless this violent course is taken, the term remains without any meaning at all.

22. The term ‘just’ is dealt with in much the same way as ‘right’. ‘When we style an act “just”, in the strict sense of the term, we point out that an undue preference would have been shown someone by its omission.’36 The criticism just made applies here again; justice disappears as a moral concept. But there is a further difficulty. If ‘just’ means what Westermarck here says that it means, it is certainly not an expression of emotion merely, either positive or negative. It connotes a complex set of relations, connecting one man with others, and the agent is conceived as giving each of these others his due. This complex relationship belongs to the essence of justice. But it is evidently not included in any emotion that we may assert ourselves to be having. If it is replied that ‘just’ has no emotional meaning of its own, but draws one from ‘unjust’, the difficulty remains. For ‘unjust’ involves a relationship equally complicated and equally far removed from any mere feeling. In his concern to exalt the emotional element in the judgment at the expense of the intellectual, Westermarck seems here to have poured out the baby with the bath.

Another line of reply may be offered; it may be said that the work of intellect or cognition is limited to defining the object toward which the emotion is felt, and that the judgment declares this emotion. But this would imply that the characterization of the act as being in the above sense just is one thing and the judgment another, that the first precedes the second as the condition of its occurrence. In our actual thought there is no such division or precedence. When we call an act just or unjust, the set of apprehended relations is ascribed in the judgment itself and forms an essential part of its content. Indeed if this judgment were reduced to the expression of an emotion, everything that could distinguish the judgment of justice from the more general judgment of rightness would have disappeared.

This fact that the content of justice cannot be supplied by emotion raises a further doubt. We have seen that for Westermarck the distinctive note of moral emotion is its impartiality. Suppose, for example a careless driver rams my motor-car from behind, injures it badly, and makes off before I can discover his identity. Mere anger at what he has done to me is not a moral emotion. The emotion is moral only so far as it is what would have been felt toward any driver in similar circumstances, excluding any feeling arising from the fact that it was I in particular who was the victim. Now Westermarck is surely right in implying that the moral judgment is a universal or as-such judgment, that what it pronounces wrong is not merely X's injuring my car by his careless driving, but anyone's injuring anyone's car in this way. The judgment is then impartial and impersonal. But does it make sense to say that the emotion is impartial and impersonal? To split an emotion up into a personal part, directed against a certain man, and an impersonal part, directed against this sort of man and act, would prove no easy business either in practice or in theory. But even if it were done, is it the emotion that would be impartial? Impartiality involves applying a principle; it implies a distinction between those features on the one hand that are essential and universal and, on the other, the irrelevant claims of particular persons; and one would have thought that if anything is the special function of intelligence and beyond the power of emotion, this is. No doubt when intelligence thus comes into play, emotion is modified to suit. But emotion itself cannot distinguish the universal, or recognize the species of a genus, or distinguish the essential from the irrelevant. Hence when the emotional theory makes impartiality the distinctive trait of moral judgment, it is implicitly giving up its case. Impartiality belongs to intelligence, not to feeling.

23. This refusal to give intelligence its due part has also distorted Westermarck's account of the judgment of duty. ‘Strictly speaking,’ he writes, ‘“ought” and “duty” only express the tendency of an act's omission to call forth moral disapproval…’37 When the inveterate toper, then, judges that he ought not to visit the bar again, he is really judging that his doing so would excite his own moral indignation. I do not find this convincing. I should have thought it not only a possibility, but a rather common occurrence, that he should see that he ought not to indulge himself further, and yet at the same time know that he is going to, and feel no indignation against himself for doing so. The issue is a somewhat treacherous one because it is no doubt true that in most cases where we see something to be our duty, we should disapprove ourselves for failing to do it, and one must resort to exceptional cases to dissociate the two factors. That they do not necessarily go together seems clear from the instance given. But even if they always did go together, they would still not be identical. One could still see that in calling something my duty, I did not mean merely that I should be indignant over my failure to do it. If a kindly fate should provide us with an introspective and articulate toper, I think he would agree that in saying it was his duty to stop indulging himself, he was not merely saying that he would be indignant with himself if he did not. He would say that he knew quite well he ought to stop, though he was very far from certain whether he would be indignant with himself if he did not, that his belief that he ought to stop would not be shown false merely by his failure to become angry with himself, and that even if he did become angry with himself, this would be because he saw that he ought to stop. The perception is the cause of the anger, not the anger of the perception. Westermarck has got the cart before the horse; he has made the emotional consequence of a judgment into the substance of the judgment itself.

We have been commenting on the place that Westermarck assigns in moral appraisals to the cognitive and the emotional factors. It is time to turn to a somewhat different kind of consideration. Suppose we assume that moral judgment is as largely an affair of emotion as he holds it is; do the implications of such an analysis fit in with the rest of what we know, or seem to know, about right and wrong? We have seen that, on the surface, they obviously do not. But we must examine this apparent conflict more closely.

The conflict resolves itself into a single large issue, which crops up in different forms. This issue is not, as it is sometimes said to be, whether moral judgments are objectively true or simply true; ‘objective’ here adds nothing. Westermarck does not deny their truth. On his theory, if I call an act wrong, what I mean is that I tend to feel reprobation for it; if I do feel this, the judgment is true; if not, false. No; the issue is not over the truth of moral judgments but over the objectivity of moral values. Westermarck maintains that these are subjective, that there is nothing good or bad, right or wrong, apart from my feelings about it. It is clear that most men believe the contrary. They are persuaded, justly or not, that these terms answer to some character in the object itself that is independent of our feelings. How large a revision would be required in their ways of thinking about morality if this view were admitted?

24. Since we have just been discussing duty, let us go on with that from this fresh point of view. We have seen that what ‘ought’ means for Westermarck is at variance with our ordinary meaning. But may it not be that though the connotation differs, the denotation will remain the same? It is notorious that philosophers may differ widely in what they profess to mean by such terms as ‘life’, ‘cause’ and ‘right’, and yet agree all along the line in their applications of them. Granting that Westermarck's definition of ‘right’ is different from that of most of us, would his view make any significant difference to the range of actions we take as obligatory?

It would make a very great difference. All that would be needed to remove an act from the list of duties would be to change one's feelings about it. Suppose I have agreed to pay a certain rent on the first day of each month; we should certainly say that if I were able to pay it I ought to pay it, and that my having a headache, or being in love, or feeling a dislike for my creditor, or having an aversion, however intense, to parting with my money, made no difference to its being my duty, since that does not depend on how I feel. On the theory before us, this is not true. I can erase the obligation merely by a change of feeling. All that makes it an obligation lies in me; it lies in the moral aversion I feel for the non-payment of the debt, and disappears as I get over this aversion. Is it replied that there is no danger of my getting over it, since I can see that I ought not to try? But all this means is that I have another aversion to trying to get over the first aversion; there is no objective ground for either; and the cure for both is the same, namely to get myself to feel in a different way.

‘Not so fast,’ it may be said; ‘Westermarck has given us such a ground, for moral aversion is not in his view arbitrary; it is felt against an act “as a cause of pain”. My conviction that non-payment should be avoided thus rests upon a judgment which is, or may be, independently true.’ At this stage such a reply is mere confusion. If it really is productiveness of pain, conceived as bad in itself, that makes an act wrong, that is an intelligible view, but then, as we have seen, there is no subjectivism about it. What makes an act right for Westermarck is not the pleasure or pain it causes, but the emotional attitude which the apprehension of such consequences happens to arouse in me. I may normally be moved to sympathy by the prospect of my creditor's pain, and, as a result, moved to indignation by an act that would produce such pain; but apart from the way I happen to feel, there is no duty to have either emotion, and once I see this clearly, no reason remains why I should not try to outgrow them both. It may be replied, once more, that the payment of rent belongs to a large network of practices which must be maintained as a whole if my happiness is to be assured. But this is continued confusion. Not only may the non-payment, with its rich and immediate return, mean far more to my happiness than payment ever would through its roundabout influence on order generally, but even if payment did have such an effect, this would impose no sort of obligation on me. For obligatoriness, we repeat, does not rest on conduciveness to anyone's good, either society's or my own; it rests on my feelings; and if I happen not to feel disapproval for an act, there is no duty to avoid it. The ‘ought’ has really gone. It has completely vanished into ‘is’. And since the ‘is’ has to do with my emotions, the range of my duties will depend on my emotional self-control. If I can ‘condition’ my emotions in a certain way, I shall not only believe that it is no longer my duty to pay the rent; it will in fact cease to be my duty. This certainly opens vistas to tired saints. Let them become sufficiently expert in applied psychology and their ‘moral holiday’ can begin. It can begin, furthermore, without the slightest detraction from their saintliness.

One can foresee the protest that will arise here. ‘You find this implication in Westermarck only because you have left out his express proviso about what makes emotion moral. You speak of the self-disapprobation bound up with non-payment as if it were a whim that would evaporate as soon as one sees its inconvenience. But since this emotion has nothing to do with any special advantage to oneself, it is not likely to be exorcized merely by finding it somewhat troublesome. It is the sort of emotion one feels toward the evasion of debts as such. And however easy it might be to tone down the condemnation of oneself for such a practice, one cannot wave away one's condemnation of others. That, at least, is deep-seated and firm. And as long as it remains, the sense of duty will remain.’ Does this not meet the criticism completely?

On the contrary, it gives away the case. What is apprehended when the act is thus impartially viewed is either something objectively wrong or it is not. If it is—and we have seen that Westermarck really though inconsistently implies that it is—then the wrongness clearly does not lie in my emotion, nor does the judgment primarily express emotion; it expresses an insight or a belief. On the other hand, if in the act thus viewed there is nothing objectively wrong, then my antipathy to it is groundless, and there is no more reason why I should not treat it as an inconvenience to be got rid of than any other emotion that proves a nuisance. Let us remember that for Westermarck, while speaking in the official robes of the subjectivist, there is nothing wrong whatever about rent-dodging as such apart from my emotions about it, nor, these emotions again apart, is there any obligation to feel about it in one way rather than another. Once I see this, why should I cling to a quite ungrounded antipathy which stands in the way of my desires? I had supposed duty to be a stern taskmaster imposing obligations upon my reluctant will in the name of an impersonal reason. I now see that the taskmaster is a creation of my own, and that the reason he invokes is only a wraith like himself. If I can exorcize both together by adopting a new attitude, why in the world should I not?

To the person who replies, ‘Exactly: why not indeed?’, I think the plain man's answer is the best one: ‘Because you can see, if you look at the matter straight, that your duty to pay the rent does not depend on how you feel, but on an agreement you have made and on the bad results of breaking it.’ To be sure, the emotion of disapproval is a natural and suitable attitude toward a wrong act, but its very suitability implies that the act has a character of its own that makes this suitable. To hold the emotion constitutive of the rightness is to make the tail wag the dog. Was it not Mr Milquetoast who said to his boy, ‘Son, go down cellar and bring up some wood.’ ‘I won't,’ said the boy, ‘I don't want to.’ ‘Then don't,’ was the reply; ‘I'll be obeyed.’ For duty to capitulate to the changes of feeling strikes one as a performance of about equal dignity.

25. We have been considering the strange shift that would take place in the content of duty if the emotionalist theory were accepted. There would be, if anything, a still stranger shift in the list of right and wrong acts generally. This follows from the reflection that it is all but impossible, on the theory, to make a mistake about right and wrong. Acts which, as we should commonly say, are approved mistakenly will be really right, and acts mistakenly disapproved will be really wrong, wrong in the only sense in which any acts are. Most people would say that Booth was mistaken in judging it right to assassinate Lincoln. On the theory before us, he was making no mistake whatever. If he felt an emotion of approval toward the act, it was not only right but laudable. Some people who are inclined to subjectivism deceive themselves thoroughly on this point by failing to see what the theory commits them to. ‘Surely,’ they say, ‘if Booth felt as he did, the act was right to him.’ What they fail to see is that the theory they are defending makes this defence irrelevant. For most of us there is merciful distinction between what is objectively right and what is subjectively right; even if we were to do something accounted monstrously wrong, like murdering a great and good man, it could always be said that if we did what we believed to be right, our action was right subjectively. But it is the essence of the theory before us to render this distinction meaningless. An act that is subjectively right, in the sense of being sincerely approved, is right, simply and without qualification.

26. We may conclude by pointing out two or three more ways in which subjectivism commits us to paradoxes. (1) It makes it meaningless to ask whether what we approve of is right, or what we disapprove is wrong. A man may, because of parental or religious teaching, disapprove most strongly of contraception, let us say, or euthanasia. If all he means by calling these things wrong is that he has an unfavourable feeling toward them, then in calling them wrong he is clearly correct, and that should end the matter. He knows that he has the feeling and since this is all the judgment affirms, to ask whether it might be mistaken would be meaningless. Yet it is plainly not meaningless. In spite of his conviction that these things are wrong, he may be willing to listen to argument about them, and come to think that his earlier judgment was mistaken. How would the subjectivist deal with this conviction of error? He would say that the new judgment—that the conduct in question is right—is true, because it reports correctly the favourable attitude now taken. This, however, casts no criticism upon the truth of the former judgment, since that too reported correctly what was felt. This is certainly a paradox. Most men would say simply that they had come to think their earlier belief mistaken, and surely this is the natural way to put it. Yet on subjectivist assumptions, it is a meaningless thing to say. Their earlier judgment reported correctly how they felt; they know it to be correct; they still know it. How then could it have been mistaken? Nevertheless, it was.

(2) Again, we attach great importance to consistency in moral judgments, an importance that is hardly intelligible on the subjectivist view. A man who yesterday struck a white man with his car and gave him every attention, while today, when he strikes a black man in the same circumstances, he pays him no attention whatever, would be criticized for inconsistency. But there is no inconsistency for the subjectivist. Two differing emotions are simply two events, and events are not inconsistent with each other. Neither are the implied judgments. The judgment ‘I ought to help the white man’ is quite consistent with ‘I need not help the black man’, for these judgments are merely the reports of successive states of feeling, both of which are true. ‘Emotional moods vary: and if either an individual or a community were angry with a certain piece of conduct today and not angry with precisely similar conduct tomorrow, nobody would have any right to reproach them with inconsistency.’38 But of course we do so reproach them. And if this reproach is justified, it can only be because the moral judgments concerned are more than reports of feeling. They are inconsistent because they mean to say something about conduct, and to say incompatible things.

(3) Then there is the consideration that G. E. Moore found so decisive against subjectivism.39 We do mean at times to contradict each other about moral matters, but if subjectivism is true, we cannot even if we try. You may say that an income tax is robbery, and I may maintain that it is not. Each of us may suppose that he is contradicting the other; each of us would certainly say that he means to take exception to what the other has said. But with the worst intention in the world, we cannot bring the contradiction off. All that you are doing is reporting—quite correctly—your antipathy to this tax. All that I am doing is reporting—equally correctly—my lack of antipathy to it. Both judgments are true, and two true judgments can never conflict. The only way I could contradict you would be to say that you do not really feel what your judgment says you feel, and that, besides being impertinent and almost certainly false, is not an ethical difference at all, but one about psychological fact. So long as we confine ourselves to judgments of value, conflicts of opinion are impossible, since we are never talking about the same thing, but you about your feelings and I about mine. Moore thought this absurd. He held it to be far more certain that we really do mean to deny others’ opinion at times on moral issues than that the subjectivist analysis of ethical judgment was correct. I agree.

But at the time Moore wrote, modern emotivism had not yet shown its head. Assuming that moral judgments really were judgments, it seemed absurd to say that they never conflicted with each other. But what if they were not judgments at all? What if they were pseudo-judgments, mere exclamations masking under the guise of assertions? Moore had not thought of that. And if they really were such, then the consequence he alleged against Westermarck as an intolerable paradox—that our moral judgments can never conflict—will be a perfectly natural one. They will say or assert nothing; and if they assert nothing, they obviously deny nothing. Whether this emotivist analysis is nearer the truth than the subjectivist one we shall soon examine.

From the book: