A STRIKING FEATURE of moral and political argument in the modern world is the extent to which it is innovators, radicals, and revolutionaries who revive old doctrines, while their conservative and reactionary opponents are the inventors of new ones. So the contract theorists and the believers in natural rights in the seventeenth century were reviving features of medieval doctrines, while the doctrine of the divine right of kings was essentially a sixteenth- and seventeenth-century invention. So also, at the time of the French Revolution, it is Tom Paine who revives Locke, and it is Burke who invents a quite new form of the appeal to tradition. Paine is not a source of philosophical argument in himself; his importance lies in the way in which he, and more especially, his French associates helped to force the moral traditions of the English ruling oligarchy away from the doctrine of natural rights. How did they do this?

The danger of all appeals to general principles on one’s own behalf on a particular occasion is that one renders oneself liable to have the same principles invoked against one on some subsequent occasion. Precisely this is what happened to the English ruling class; the principles of 1688 were invoked against them by the Americans in 1776 and by the revolutionaries against their French colleagues in 1789. It was this fact which underpinned Tom Paine’s appeal to the rights of man; and it was this fact that Richard Price, whom we have already noticed in his role as a believer in the rational intuition of moral first principles, emphasized in his sermon at the dissenting meetinghouse in Old Jewry in November, 1789. Price emphasized the assertion in 1689 of the right to choose and to dismiss sovereigns, and above all, of the right to frame anew the constitution, and reiterated the correctness of this assertion. In so doing, he played his part in goading both to fury and to reply Edmund Burke. Burke’s attitude to the mass of men is well conveyed by his phrase “a swinish multitude”; his attitude to the rights of man is entirely coherent with this. He denies, first as a matter of history, that the Whig Revolution of 1688 did involve the kind of assertion of rights that Price claimed. The displacement of James II was due to a fear lest his critics should weaken the throne and the hereditary principle; hence the preference for the next line in succession, even though it was the German line of Hanover, in order that that principle might not be discredited. But Burke was not merely concerned with history. Not only was 1689 not an appeal to natural rights, there are no such rights. They are metaphysical fictions.

Burke says of the writers of the French Revolution that they are “so taken up with their theories about the rights of man, that they have totally forgotten his nature.” By nature Burke does not mean a state prior to a social contract, but society as it is, and above all, as it has grown to be. Theoretically based plans for the reform of society are violations of a divinely ordained history of social growth, so that Burke can speak of social development as “the known march of the ordinary Providence of God.” Established institutions are thus rated as high by Burke as they are low by Rousseau. Both invoke “nature,” but while for Rousseau nature is contrasted with society, for Burke nature includes society. Burke however does not view nature simply as all that is; for if nature were all-inclusive, one could not war against it, as revolutionaries do. Nature is in fact equated by Burke with certain established norms and procedures, including the procedure of relying on prevailing habit rather than on argument. “Politics ought to be adjusted, not to human reasonings, but to human nature, of which the reason is but a part, and by no means the greatest part.”59 This is not just a doctrine about politics, but about the moral life in general. Hence Burke’s defense of what he calls “prejudice.” “Prejudice is of ready application in the emergency; it previously engages the mind in a steady course of wisdom and virtue, and does not leave the man hesitating in the moment of decision, skeptical, puzzled, and unresolved. Prejudice renders a man’s virtue his habit, and not a series of unconnected acts.”60

Burke’s positions are of importance, if only because of their subsequent influence. The assessment of them confronts one initial difficulty, namely that if Burke is right, rational argument upon these topics is misplaced. Hence by even venturing to argue with him we appear to presuppose the truth of what we are trying to establish. But this difficulty is not in fact ours, but Burke’s. For to deny the possibility of rational argument playing the role of arbiter means that in advancing one’s views one cannot be appealing to any criterion by which they may be established. But if this is so, then not only can one not argue in one’s own favor, but one has made it difficult to understand what it could mean to call one’s views “true” or “false.” For the application of these predicates always involves an appeal to some criterion. Suppose, however, that at this point we look to Burke’s practice of arguing rather than to his principle of condemning argument. We shall then find in his arguments two mistakes, both diagnosed by William Godwin, the anarchist, in what is in effect a reply to Burke’s Reflections on the Revolution in France, Political Justice.

First of all, Burke confuses society and the state. He assimilates particular forms of political institution to institutions in general. From premises which assert merely the need for stable and established social arrangements he tries to derive the conclusion that Louis XVI’s head should not be cut off. The roots of this confusion are more interesting than the confusion itself. Burke understands that appeal to moral and other norms presupposes an established form of social life. He tries to picture the revolutionary theorist as a man who wishes to destroy the very social life which is necessary to give meaning to the norms in the name of which he intends to carry out his act of destruction. But in so doing, he equates the notion of an established form of social life with the notion of an established set of institutional arrangements. In fact, the institutions of a society may well be at odds with its norms. To maintain these institutions may be fatally destructive. Burke never noticed the fact that revolutions are extremely difficult to make. Theorists become revolutionaries only when their theories are able to articulate a deep dissatisfaction which the theorists did not invent. And at this point it is the refusal to destroy and recreate social institutions which is destructive of social life itself. The true nihilists in history were all kings: Charles I, Louis XVI, and Tsar Nicholas. The revolutionaries in their societies had to save social life from their rulers’ destructive maintenance of the existing order.

Secondly, Burke’s defense of prejudice and habit against reflective criticism rests on an inadequate analysis of the notion of following a rule. I may in my conduct follow and abide by rules which I have never made explicit; breaches of such rules may shame or shock me without my articulating any formula adequately expressive of the rule. But such unreflective behavior is as much rule governed as is the behavior of the man who consciously invokes an expressly formulated maxim. And it is clearly the kind of behavior which Burke wants to exalt; we use words like habit and prejudice to bring out not that such conduct is not governed by rules, but that our attitude to the rules is unreflecrive. Burke is right to suppose that the moral life would be destroyed by our reflecting upon our rules of conduct prior to each and every action. Action must for the most part rely on our habitual dispositions to do this rather than that. But if for this reason, reflection can only be occasional, the importance of such occasions is heightened, not lessened. Because we are right not to be continually re-scrutinizing our principles, it does not follow that we are wrong ever to scrutinize them. So Godwin speaks rightly of a need to articulate and examine them, “to cast away the coat of prejudice and to leave nothing but the naked reason.”

Godwin, who was married to the mother of female emancipation, Mary Wollstonecraft, and who was the father of Shelley’s second wife, was the prototype of the innovating moralist in the modern world. The kind of abuse which was later to be hurled at a Bertrand Russell or a Wilhelm Reich was showered on Godwin. De Quincey recalled in his reminiscences that “most people felt of Mr. Godwin with the same alienation and horror as of a ghoul, or a bloodless vampyre, or the monster created by Frankenstein.”

Godwin in fact was a humane and sensitive man who applied himself to the classical problem of eighteenth-century moral theory. He accepted from Hume the view that we are moved to action by feelings not by reason and from Locke the view that it is reason which discerns moral distinctions. He thus has a more complex position than most eighteenth-century writers. Our feelings move us to action, but they will only move us to right action if we have a clear and rational view of the facts. Such a view includes taking into account the consequences of our actions; it must also include the application to them of principles such as that of impartiality, of not making exceptions to general rules in our own or in anyone else’s favor. Godwin’s view that there are rational moral principles of an inescapable kind is never developed with sufficient clarity. But he is as much as anyone since Aristotle the father of the notion that at the foundation of morals lies the principle that if morality is to be argued about at all, then the onus of justification lies upon those who propose to treat men differently. The very process of moral argument presupposes the principle that everyone is to be treated the same until reason to the contrary is shown. This principle is formal in the sense that it does not prescribe how in fact anyone is to be treated. But it has important practical consequences. For it forces into the open the justification of treating people differently because of their age, sex, intelligence, or color. Equality in its most minimal form is embodied in a society in which this is the case.

Godwin himself extended the scope of principles of reason far beyond this. He thought that reason showed me that there is more value in the happiness of a number of men than in that of one, and that this is true irrespective of whether that one is myself, my friend or relative, or a total stranger. Hence I ought to prefer the general happiness to my own. To the rejoinder that if I do so, it is only because I am so psychologically constituted that I will feel unhappier if I disregard the general happiness of others than if I disregard my own, Godwin’s reply is that the pain which I feel at disregarding the unhappiness of others is felt only because I recognize that I ought to be benevolent. The pain cannot be the reason for my action, for it is only because I have quite a different sort of reason for it that I am liable to feel the characteristic pain. It is only in the light of my rational principles, for example, that my own actions will inspire in me satisfaction or guilt.

If men have within themselves rational principles prescribing the general good, why do men disregard that good? Godwin’s answer is that we are corrupted by the social environment, and above all, by the influence of government. For government claims an authority which belongs only to right reason. And right reason is only grasped by individuals, assisted by the rational persuasion of other individuals. The hope for man lies in the perfectability of human nature. Godwin’s belief is that the influences of social and governmental forms can be overcome and replaced by a free community of rational beings in which it is the opinions of those who are informed and objective that will carry weight.

Godwin is a figure curiously akin to and curiously at odds with Bentham. Where Godwin is utopian in his political proposals, Bentham is the careful reformer, anxious to escape accusations of utopianism by being prepared to suggest the exact size of the beds to be used in prisons or the precise reforms needed in the laws of evidence. Where Godwin believes human nature to be committed au fond to disinterestedness, Bentham believes that private interest always needs to be weighted and guided if it is to serve public interest. Yet Bentham’s criterion of the greatest happiness of the greatest number is essentially the same as Godwin’s; both were in limited sympathy with the French Revolution; both represent the future rather than the past. It can be put like this: if one takes the stock of characteristically modern liberal clichés and banalities, one is in a world of which both Godwin and Bentham are ancestors. For both, society is nothing but a collection of individuals; for both, the good of individuals is a matter of their happiness; for both, that happiness can be summed and calculated. In Godwin the notions of good and evil still retain some of their traditional force; in Bentham they are to be redefined in terms of pleasure and pain.

Bentham’s thesis was not of course that words such as good and right were or had been used by most people to mean “productive of the greatest happiness of the greatest number” or some equivalent phrase. It is not even the case that Bentham always propounded the same thesis. Sometimes he seems to be concerned not with the meaning of terms in the moral vocabulary, but only with the statement of a moral-and political-criterion. Sometimes he does indeed offer us a definition, but in the form of a proposal rather than of an elucidation. He says in effect that we may define good and right in terms of the concept of the greatest happiness of the greatest number or we may not; but that unless we do, shall talk nonsense. And sometimes he seems not to distinguish these tasks. Nor for his purposes does he need to distinguish them. For his central proposal amounts to the contention that the only rational and consistent criterion available for the guidance of action is the assessment of the pleasurable and painful consequences of any particular action, and that the meaning of evaluative expressions can only be understood in this context. There is no alternative rational criterion for at least two kinds of reason.

The first is that theories, such as those based on a belief in natural law or natural rights, which suppose that there are rights, duties, and obligations apart from and prior to those embodied in positive law are thought by Bentham to rest on a basis of logical error. For they are, on his view, the product of a belief that words like duty and obligation are names which have a sense and a reference quite independent of their use in any particular context. Bentham’s own logical views on this point are a mixture of truth and error. On the one hand, he grasped correctly that only in the context of a sentence, does a naming, describing, or referring expression have meaning-a point that was to be made a commonplace only by Frege and Wittgenstein. On the other hand, it is in no way clear that adherents of natural-law and natural-rights theories are necessarily committed to the logical error of supposing otherwise. A more serious criticism of such theories is intimately connected with one of Bentham’s most important motives in attacking them. Suppose that anyone asserts that men possess natural rights or are bound by natural laws: invite him, then, to make a list of such rights or laws. It is notorious that adherents of such theories offer lists which differ in substance from each other. Is there, then, any criterion for the correct inclusion of an item on such a test? Bentham’s conviction that there is not was directed in the first instance at the reactionary sanctification of the legal and penal status quo that Blackstone, in his Commentaries on the Law of England, accomplished by the use of the theory of natural law. But Bentham was completely impartial in the application of his skeptical doubts, and in spite of his sympathy for the American Revolution and for at least the initial phases of the French Revolution, he is trenchant in his criticism of the revolutionary doctrine of the rights of man, a doctrine which he declares to be nonsense, and in his criticism of the doctrine of imprescriptible natural rights-“nonsense on stilts.”

If, then, a first reason for holding that only the principle of utility, the principle of the greatest happiness of the greatest number, furnishes us with a criterion for action is the alleged logical impossibility of any metaphysical theory of morals, a second is the foundation laid for the principle in human psychology. Men are made so that they are placed under the dominion of “two sovereign masters,” pain and pleasure. Bentham’s psychology, whose source is in Hartley, is mechanical and associationist. We cannot but pursue pleasure and flee pain, and the association of the prospect of either with something else will draw us to or repel us from whatever pleasure or pain is associated with. Bentham takes it for granted that pleasure and pain are correlative terms, and that both are equally simple and unitary concepts. He gives fifty-eight synonyms for pleasure, and his logical sophistication about naming on other occasions does not prevent him from behaving as if happiness, enjoyment, and pleasure all name or characterize the same sensation. Different sources of pleasure can be measured and compared in respect of the intensity and duration of the sensation derived from them, the certainty or otherwise of having the sensation, and the propinquity or remoteness of the pleasure. In choosing between alternatives, quantity of pleasure is the only criterion: “Quantity of pleasure being equal, pushpin is as good as poetry.”61 Moreover, in summing up the pleasures of a number of people everybody is to count for one and nobody as more than one.

If each individual is in fact moved by the prospects of his own pleasure or pain, what becomes of altruism? Bentham’s thought is not entirely coherent here: on the one hand, in his political and legislative proposals he recognizes the conflict between public and private interest and the need for molding human nature. His wish to construct a society in which a man’s pursuit of his private pleasure and his pursuit of the greatest happiness of the greatest number will coincide clearly rests on the assumption that society is not at present so organized. But elsewhere, and especially in the Deontology, Bentham implicitly identifies the greatest happiness of the individual with that to be found in the pursuit of the greatest happiness of the greatest number. The only motive for obeying the rules necessary to social life is the pleasure to be found in obedience or the pain resulting from disobeying them.

There is no problem which Benthamite utilitarianism raises which was not raised within the utilitarian tradition itself, and the burden of these problems fell upon John Stuart Mill. His father, James Mill, was an enthusiastic collaborator of Bentham’s, himself a psychologist in the tradition of Hartley, who once wrote that he aspired to make the human mind as plain as the road from St. Paul’s to Charing Cross. This spirit of self-confidence was scarcely inherited by his son. In late adolescence, after an education which had laid adult burdens upon him from the earliest possible age, he turned from his absorption in schemes of social reform to inquire whether, if all such projects were to be accomplished, this would render him happy. The sinking heart with which he answered, No presaged a nervous breakdown from which he was rescued to an important extent by the poetry of Wordsworth and Coleridge. But it was to be significant for more than Mill’s personal life that the coincidence between private happiness and that of the greatest number should have failed the utilitarians themselves so early. Mill’s whole tenor of thought is that of a utilitarian who cannot avoid any of the difficulties which this doctrine raises, but who cannot conceive of abandoning his doctrine either. What are the difficulties?

First, Mill abandons the view that the comparison between pleasures is or can be purely quantitative. He introduces a qualitative distinction between “higher” and “lower” pleasures. The higher pleasures are to be preferred: better Socrates dissatisfied than a fool satisfied. How can we be sure of this? Only he who has experienced both is qualified to judge, and only the wise man who prefers the Socratic classification has this experience. Yet here doubt necessarily arises: how could a Mill know what it was like to be a satisfied fool, any more than the fool could know what it was like to be Mill? The point of this question extends further than to cast doubt upon a single contention of Mill’s. For what it brings out is that Mill is still engaged, as Bentham was, in trying to bring all the objects and goals of human desire under a single concept, that of pleasure, and trying to show them as all commensurable with each other in a single scale of evaluation. Moreover, he, like Bentham, treats pleasure as a unitary concept.

He is able to do so because the concept of pleasure has tended to degenerate, just as the concept of duty has. I have already suggested that in the case of duty, a highly specific concept associated with the notion of the duties of an office holder evaporates into a generalized notion of “what a man ought to do.” So pleasure as the concept of one specific kind of goal is transformed into the concept of any goal at all. Both hedonists and puritans contribute to the history of this degeneration. Hedonists, who begin by commending pleasure, against other goals, then become defensive and insist that they are not merely commending wine, women, and song, but also the higher pleasures, such as reading the Critique of Pure Reason. Puritans insist that they are not against pleasure as such, but only low or false pleasures. They, too, are for true and lasting pleasures, such as only Zion’s children know. So concepts like “pleasure” and “happiness” are stretched and extended in all directions until they are used simply to name whatever men aim at. By this extension they become useless for evaluative and moral purposes. For in evaluation, and especially in moral evaluation, we are not only engaged in grading and in choosing between alternative objects which we already desire; we are also engaged in grading and choosing between the cultivation of alternative dispositions and desires. The injunction “Pursue happiness!” when happiness has been given the broad, undifferentiated sense which Bentham and Mill give to it is merely the injunction “Try to achieve what you desire.” But as to any question about rival objects of desire, or about alternative and competing desires, this injunction is silent and empty. And this is equally true whether the happiness which I am to cultivate is to be my own or that of the greatest number.

Mill, faced with the objection that there are many cases in which one cannot assess which out of the alternative possible courses of action will produce the greatest happiness of the greatest number, asserts that utilitarianism enjoins no more than that in cases where one can so assess the consequences of action one ought to use the principle of utility as a criterion. But this concession is more deadly than he perceives: for he is forced to allow implicitly that there are other evaluative criteria. What they are and what their relationship to the principle of utility may be he never makes clear. But we may accept Mill’s concession in the spirit in which it is offered, if we recognize that when utilitarians speak of the greatest happiness they are often in practice speaking of a quite specific goal for action rather than of the generalized concept of their theoretical appetites. This goal is that of the public welfare, and it is a goal peculiarly relevant to those areas of life in which Bentham was especially interested. Prisons and hospitals, penal codes and constitutional processes-in these areas it is possible to ask and to answer adequately, even if only crudely, the question of how many people’s lot will be bettered, how many people’s lot will be worsened, by such and such a measure. For we have obvious and established criteria for faring well or ill in these areas. Will ill-health be increased or diminished? Will the attaching of this rather than that penalty to this crime diminish or increase the frequency of its occurrence? Even in these cases there are choices to be made on which no version of the principle of utility can guide us: an example is the choice between devoting resources to health services or devoting them to penal reform. But it is necessary to emphasize that the utilitarian advocacy of the criterion of public happiness is not only a mistake. That it seems so obviously the criterion to be considered in certain areas of life is something we owe to Bentham and Mill.

The concept of happiness is, however, morally dangerous in another way; for we are by now well aware of the malleability of human beings, of the fact that they can be conditioned in a variety of ways into the acceptance of, and satisfaction with, almost anything. That men are happy with their lot never entails that their lot is what it ought to be. For the question can always be raised of how great the price is that is being paid for the happiness. So the concept of the greatest happiness of the greatest number could be used to defend any paternalistic or totalitarian society in which the price paid for happiness is the freedom of the individuals in that society to make their own choices. Freedom and happiness can in certain circumstances be radically incompatible values. We can trace one legitimate offspring of utilitarianism for whom freedom was sacrificed to happiness in the history of Fabian socialism. For Fabianism socialism was a matter of schemes of reform initiated from above by the enlightened few for the welfare of the unenlightened many. Fabianism stands at the opposite pole in the history of socialism from the revolutionary democracy of Rosa Luxemburg or the I.W.W., for whom socialism consisted in workers becoming free from the domination of others, and owners and directors of their work and their lives.

Moreover, the concept of the greatest happiness of the greatest number is only applicable with any kind of moral legitimacy in a society in which it is assumed that nonutilitarian norms of decent behavior are upheld. The concept of the public happiness has obviously legitimate application in a society where the consensus is that the public happiness consists in more and better hospitals and schools; but what application has it in a society where the public happiness is found by the public itself to consist in the mass murder of Jews? If in a society of twelve people, ten are sadists who will get great pleasure from torturing the remaining two, does the principle of utility enjoin that the two should be tortured? Nothing could have been further from the thought of Bentham and Mill. But this only makes it clearer that they are not consistent utilitarians, that they rely on an implicit appeal to other norms, which they covertly use to define the greatest happiness.

It is this sievelike nature of the utilitarian concept of pleasure or happiness which makes Mill’s proof of the principle of utility so unimpressive. Mill’s proof runs as follows. He begins by allowing that in any strict sense, proof on matters concerning ultimate ends is not to be obtained. But, nonetheless, we may adduce considerations capable of influencing the intellect. The argument then proceeds from the assertion that just as the only way to show that something is visible is to show that men can see it, so the only way to show that something is desirable is to show that men desire it. But all men desire pleasure. So pleasure is universally desirable. Mill has no problem about the transition from the desire for my own pleasure to that for the greatest happiness of the greatest number, which he makes by means of the bald assertion that the pleasure of another is naturally pleasurable to me. When Mill comes to show that only pleasure is desired his method is to take apparent alternative goals and show that originally they are desired for the pleasure which accompanies them, and only secondarily do they become desired for their own sake. This method of argument is of course necessarily ineffective. If anything, it shows that there are goals other than pleasure. But criticism of Mill has centered on that part of his argument where he passes from the assertion that pleasure is desired to the assertion that it is desirable. What Mill’s critics, beginning with G. E. Moore, have in effect said is that Mill illegitimately tries to deduce the conclusion that pleasure ought to be desired from the premise that it is in fact desired. But this, so it is alleged, is necessarily a fallacious inference. For an is cannot by itself entail an ought. One does not have to enter on any general discussion of fact and value to deal with such critics. They are of course right that the inference in question is fallacious if it is intended as an entailment. But they are simply mistaken in their reading of Mill.

For what Mill says about proof makes it clear that he does not intend to use the assertion that all men do in fact desire pleasure as a premise which entails the conclusion that they ought to desire it. What the form of his argument is, is not perhaps entirely clear. But one way of reading him, more consonant with the text of Utilitarianism, would be this. He treats the thesis that all men desire pleasure as a factual assertion which guarantees the success of an ad hominem appeal to anyone who denies his conclusion. If anyone denies that pleasure is desirable, then we can ask him, But don’t you desire it? and we know in advance that he must answer yes, and consequently must admit that pleasure is desirable. But this reading of Mill, and indeed any reading, has to interpret him as treating the assertion that all men desire pleasure as a contingent factual assertion. Now it can only be such if pleasure is being treated as the name of one possible object of desire among others; for if it is simply an expression equivalent to “whatever men desire,” then the assertion is a vacuous tautology and will not serve Mill’s argumentative purposes. Yet if pleasure is the name of one specific object of desire (the wine, women, and song sense)-as it often is-then it is certainly false that all men desire it (puritans do not) or that it is the only desired goal. It is thus on the haziness of his central concept that Mill founders and not on the transition from is to ought.

In the course of the previous discussion another difficulty has come into view. Clearly, even on the best and most charitable interpretation of the concept of the greatest happiness of the greatest number, there are occasions where its use as a criterion would lead us to recommend courses of action which conflict sharply with what ordinarily we think we ought to do. A typical case was propounded by a later critic of utiliarianism, E. F. Carritt. The hanging of an innocent man may well redound to the public happiness if certain conditions are satisfied: that he is publicly believed, although not by us, his would-be executioners, to be guilty of murder, let us say, and that his execution will act as a deterrent, preventing the deaths of sundry innocent people in the future. Surely on a utilitarian view, we ought therefore to hang him. There are two possible types of utilitarian response to this criticism. The first is simply to deny that there is anything abhorrent in the situation. Certainly, a tough-minded utilitarian might say that this is the sort of thing that we ought sometimes to do. There is nothing philosophically criticizable in this response when it is taken in isolation from the rest of the case against utilitarianism. But when this response is combined with the protean utilitarian concept of pleasure one understands its danger. For by allowing the principle of utility to override our existing principles-such as that a man ought not to be hanged for a crime which he has not committed-we remove one more barrier to using the concept of the general happiness to license any enormity. That it can be so used has been amply demonstrated in this century; in particular the high-minded are apt to use totalitarianism as a justification to excuse their responsibility for involvement in the large-scale crimes of their societies, such as Auschwitz or Hiroshima. But, it may be objected, this is surely a moral and not a philosophical objection to utilitarianism. To which the reply is plain: utilitarianism which appears under the pretext of offering a criterion, among other things, for distinguishing good and evil, is in fact offering us a revision of those concepts, such that if we accepted it, we could allow that no action, however vile, was evil in itself or prohibited as such. For all actions are to be assessed in terms of their consequences, and if the consequences of an action are going to be productive of the general happiness, then that action, whether it is the execution of the innocent or the murder or rape of children, would be justified. Thus utilitarianism is a revisionary analysis of our attitudes and concepts; and it is relevant to ask whether it would preserve what we value in those attitudes and concepts.

A second type of response, that of Mill himself, is to argue that utilitarianism, rightly understood, does not license actions which we would ordinarily abhor. So Mill argues that only the maintenance of an impartial system of justice, in which innocent and guilty receive their deserts, could serve the general happiness; and more generally, he argues that to allow exceptions to generally beneficial rules is to weaken their authority and so is always to have harmful consequences. Later utilitarians have also argued that the principle of utility is not in all cases a criterion for judging of particular actions; rather, it is often a criterion for judging of principles. This contention has been argued in its most sophisticated form in terms of a distinction between two logically distinct types of rule: summary rules, which are logically subsequent to the actions which they prescribe or prohibit; and rules of practice, which define classes of action and are logically prior to the actions in question. An example of the first type of rule would be one forbidding walking on the grass. The actions, walking or not walking on the grass, are logically prior to any rule about so walking. An example of the second type of rule would be that which specifies the ways in which a batsman may be out at cricket. The concept of “being out” and the associated actions are specifiable only in terms of the rules defining the practices which constitute the game of cricket. The first type of rule may be represented as a summary or generalization about what is enjoined or prohibited in terms of some general criterion on many particular occasions. The second type of rule cannot be so understood. Its application on particular occasions must-logically must-be subsequent to its general formulation. It has been argued that if we apply this distinction to the problem posed for utilitarianism, we see that it is only in the case of the former type of rule that the problem can arise, but that in this case it is easily soluble. If on many particular occasions we find that doing or refraining from doing some particular action is productive of the greatest happiness, then we may summarize our discovery in a general rule prescribing or prohibiting that action. If subsequently we find a case where to do what the rule enjoins would not be productive of the greatest happiness, then we need have no hesitation about abandoning the rule for this occasion, because the rule has no force or authority except that which derives from the greatest happiness principle. But this only applies to the first type of rule.

The second type of rule constitutes or partly constitutes a practice which as a whole and in the long run may be justified by appeal to the greatest happiness principle; but one cannot ask for a particular rule to be set aside because on a particular occasion its application violates that principle. For the rule is adhered to because of its connection with the practice, not because directly and in itself it promotes the greatest happiness principle. Thus it is logically inappropriate to ask whether a particular rule in a game should be waived on a particular occasion because its application violates the greatest happiness principle; and it is logically inappropriate to ask for the waiving of a particular rule of justice on a particular occasion because the application of that rule violates the greatest happiness principle. It is a whole system of justice which stands or falls at the bar of the principle of utility, and not the detail of particular cases. So the hanging of the innocent man on a particular occasion to secure a particular deterrent is not sanctioned by a utilitarian justification of justice after all. It is the whole practice of justice with its systematic protection of innocence, and nothing less than that, which receives a utilitarian justification.

Will this defense suffice? Does it succeed in showing that utilitarianism is compatible with our ordinary belief in justice? What it ignores is the fact that we often do waive, and regard ourselves as justified in waiving, principles of justice in the interests of human happiness. So someone may fail to report a crime or fail to punish a criminal because of the effects on his family. The fact that justice is a systematic body of practices, justifiable as a whole in utilitarian terms, is not incompatible with there being clashes between particular applications of the principles of justice and the application of the greatest happiness principle. We then have to decide what weight to give to the principles of justice, and we should not have to make such a decision if it were entirely a matter of applying a single ultimate principle. The value we set upon justice is not, therefore, entirely derived from our adherence to the principle of utility.

Thus the attempt to shore up utilitarianism in this way is itself a misconceived attempt to give a false unity to our values. That such an attempt should be made is easily understood. The individualism of modern society and the increasingly rapid and disruptive rate of social change brings about a situation in which for increasing numbers there is no over-all shape to the moral life but only a set of apparently arbitrary principles inherited from a variety of sources. In such circumstances the need for a public criterion for use in settling moral and evaluative disagreements and conflicts becomes ever more urgent and ever more difficult to meet. The utilitarian criterion, which appears to embody the liberal ideal of happiness, is apparently without rivals, and the fact that the concept of happiness which it embodies is so amorphous and so adaptable makes it not less but more welcome to those who look for a court of appeal on evaluative questions which they can be assured will decide in their own favor.

No philosopher expressed the moral situation of nineteenth-century England-and to some extent we are all still in the nineteenth century-better than Henry Sidgwick. Sidgwick is a touching figure whose defects are usually the defects of his age. He was preoccupied with the loss of his own Christian faith in a way that is foreign to us. His moral psychology is crude because the psychology of his time was crude. And in his moral philosophy he mirrors his age also. For Sidgwick the history of moral philosophy in the preceding century had centered on the clash between utilitarianism and what he called intuitionism. By this he meant the doctrine that moral first principles are intuitively known, the doctrine of Price, and earlier, of Locke. Within utilitarianism further there is the argument about the relationship between the pursuit of my own happiness and the pursuit of the greatest happiness of the greatest number. Sidgwick painstakingly examined all the possible ways of assimilating intuitionism to utilitarianism, or of bridging the gap between the goals of private and public happiness. But in the end there remain three distinct sources of morality. Sidgwick’s account of the methods of ethics misses questions beyond those which he explicitly discusses. The background to his account is the moral consciousness of his day, taken as given. Philosophy appears as essentially a clarifying rather than a critical activity. In this respect Sidgwick’s is a ghost that haunts much recent writing. In his acceptance of the utilitarian consciousness of his own age he contrasts sharply with his contemporaries T. H. Green and F. H. Bradley.

Green and Bradley are often classed together as Oxford idealists; it is important however to remember that the classing of them together in this way is the work of their later critics. They themselves worked independently, and the similarities in their writings are the result of the similarity of their self-set tasks. Both were keen students of Kant and Hegel; both wish to find materials in Kant and Hegel with which to carry through a criticism of Hume and Mill. Both draw on Greek philosophy as well as on German. But Green was perhaps influenced by Rousseau as much as by any other author, while there is little trace of Rousseau in Bradley. And Green’s philosophical preoccupations were intimately related to his commitments to social and educational reform whereas Bradley was a philosophical recluse.

Both Green and Bradley break with the individualism of utilitarianism. The utilitarian picture of society is of a collection of individuals, each with his own determinate desires and his consequent goals. The shared aims and norms of society are a product of the compromises and agreements of individuals: the public good is a sum total of private goods. Both Green and Bradley break with this picture, whether in its utilitarian or its social contract forms. Both recognize that the individual discovers his aims and his desires from within a set of rule-governed relationships to others. He finds himself through, he identifies himself by means of, a set of relationships through which goals are partly specified for him. The individual then has his choices to make; he can appraise his own desires in a variety of ways. But his nature, including his desires, is not presocial.

If this argument were pursued, it would have to press the question of the relationship of morality to the social framework more seriously and in detail. Both Green and Bradley, however, place the individual not merely in a social, but in a metaphysical context. Or rather, they appear to perform social analysis in a highly metaphysical style. To make clear what this means, it is necessary to follow through the key themes of each in turn. Bradley, for example, poses the question, Why should I be moral?, a question which he uses as the title of one of his Ethical Studies, only to reply that, as it stands, the question is improper. For it suggests that there is an end beyond morality, to which the exercise of moral virtue is only a means. But from within the moral consciousness we can discern that morality does have an end, an end not beyond morality but constituted by morality itself in its highest achievement, the realization of the self as a whole. I realize my self as a whole through actions which express the stirring of the self to be something better and higher than it is already, so that the principles to which I aspire to conform come to be the principles expressed in my actual behavior. In any situation of choice between alternatives, I realize my self, first, insofar as I am aware of myself independently of the two alternatives and confronting them; and second, in self-consciously choosing one alternative and identifying myself with it, whereby I bring the whole self into being in concrete form. This Bradley calls “the concrete universal,” the judgment of universal import made concrete in the realized activity of the concrete individual.

The self develops to the point at which it realizes itself completely by identifying itself as a part in an infinite whole and so transcending its own finite bounds. “The difficulty is being limited and so not a whole, how to extend myself so as to be a whole? The answer is, be a member in a whole. Here your private self, your finitude ceases as such to exist; it becomes the function of an organism. You must be, not a mere piece of, but a member in, a whole; and as this, must know and will yourself.”

What is the whole in which the individual self must realize itself? We get, not a completely coherent answer to this question, but at least part of an answer in a later chapter of Ethical Studies, “My Station and Its Duties.” Bradley had already, in previous essays, attacked the view that the end which the moral consciousness places before us can be either pleasure or duty for its own sake. His grounds for breaking with Benthamite utilitarianism and with Kantianism are partly different and partly the same. He argues, for example, that pleasure supervenes upon a desired end, and so cannot be the end; and he argues that on the Kantian view duty is proposed as an end for a self which is constituted by desires and inclinations such that duty can have no interest for it, cannot be an end for it. But in both cases he argues that the end proposed is too general and abstract; the formulas of Kant and Bentham alike try to bring the multifarious ends which men in different circumstances and at different times pursue under a single characterization, and in so doing, they present a formula which is in effect contentless. Because it includes anything which a man might pursue, it identifies nothing which he must pursue, if he is to be true to the deliverances of his moral consciousness.

The end which Bradley lays down is that of finding my station and carrying out its duties. These duties will be specific and concrete. Bradley allows that I may have some choice of what station in life to fill; but once I have chosen some station, the question of what duties attach to it is not a matter of choice. That this is so is of some importance, for it is only insofar as the end is an objective end, and not one chosen by me, that I can hope to realize my individuality through it. What Bradley means by this is not entirely clear, but he is partly making the substantial point that any criteria by which I am to judge of my own moral progress must be criteria whose authority derives from something other than my own choices. For if my own choice is all that is authoritative, I am in the end playing an arbitrary self-enclosed game, a variety of spiritual patience in which if the cards will not come out the first time, I can, if I choose, allow myself a indefinitely large number of reshufflings. Moreover, to fill my station in life, I can utilize every part of my nature; the Kantian divide between duty and inclination is overcome.

What Bradley is presupposing rather than asserting here is that the moral vocabulary can only be given a coherent sense in the context of a form of social life with well-defined roles and functions, and one, moreover, in which men live out the substance of their lives in terms of those roles and functions. But is there such a society any more? Sociologists have often emphasized the difference between a modern individualist society in which a man’s life and status can be distinct from his various roles and functions and earlier more integrated forms of society in which a man might fill his station in life in much the way that Bradley envisages. That Bradley is able not to raise this type of question is perhaps due to his ability to pass into a metaphysical style of speech in which it is the nature of reality as such, and nothing less, that guarantees his thesis about morality.

This is less true, but it is still true, of T. H. Green. Green is more self-consciously aware that his moral views require a certain kind of society. But his metaphysical mode enables him to pass from the view that society ought to be the locus of a rational general will of a Rousseauesque kind to the view that at bottom this is what society really is. Green is more socially aware than Bradley because of his own political involvement. He came on the philosophical scene as an educator in a period when liberal young men of the ruling class, morally earnest as a result of their training in evangelical homes and by Arnoldian schoolmasters, who could not imbibe the romantic Toryism of Disraeli, were looking for a frame that would lend meaning to their lives. Green’s Balliol pupils carried into the civil service, the church, politics, into the cabinet itself-one of them was a Liberal prime minister-a belief that liberal individualism could be overcome within a liberal framework. Green was the apostle of state intervention in matters of social welfare and of education; he was able to be so because he could see in the state an embodiment of that higher self the realization of which is our moral aim.

Green’s Prolegomena to Ethics rests its argument on an extended analysis of human nature, designed to show that human existence is not wholly explicable in terms of the laws of nature. Reflection on the purposive and self-conscious character of human existence reveals to us the awareness of ourselves as intelligent beings, and members of a society of intelligent beings, whose final satisfaction cannot be anything merely physical or perishable. What, then, is the human good? We know it only in part, because our faculties for realizing it are themselves only partly realized. But the contemporary moral consciousness is a record of our highest achievement of it to date. Kant was right in thinking that the one unconditional good is the good will; but wrong in his too abstract characterization of it. The good will is manifested in the desire to transcend the existing moral consciousness in the creation of a greater good; and every expression of the good will is the creation of a form of life specifiable along the lines of “the Greek classification on the virtues.” The good will is defined as “the will to know what is true, to make what is beautiful, to endure pain and fear, to resist the allurements of pleasure, in the interest of some form of human society.”

Green’s specification of the good in terms of a form of social life, even if his own specification is a highly abstract one, enables him at least to avoid the individualist puzzles over egoism and altruism. “The idea of a true good does not admit of the distinction between good for self and good for others,” precisely because it consists of a form of social life in which different individuals play out their parts. The individual finds his good through a form of life which exists prior to himself.

Yet is Green describing at this point what actually happens? Clearly not. Is he specifying an ideal state of affairs which ought to be brought into existence? Only partly, for he believes the ideal to be implicit in the actual. Like Bradley, he makes it clear that the moral vocabulary cannot be understood except against the background of a certain kind of social life; like Bradley, his metaphysical style enables him to evade the question of the relation between that form of social life and social life as it actually is lived out in nineteenth-century western Europe. But at least Bradley and Green force these questions upon us. Their immediate twentieth-century successors were to write as if morality, and with it, moral philosophy existed apart from all specific social forms.


MODERN MORAL PHILOSOPHY opens on a quietly apocalyptic note. Moral philosophers, it is explained, have hitherto failed to answer the questions which they posed satisfactorily, because they have failed to be clear about the questions themselves. In particular they have failed to distinguish between the questions, What kind of actions ought we to perform? and, What kind of things ought to exist for their own sake? The distinction is made at last, or so it is proclaimed, in the preface to G. E. Moore’s Principia Ethica. The implication is that the problems will now be solved. The date is 1903.

The answer to the question, What kind of actions ought we to perform? is those “which will cause more good to exist in the universe than any possible kind of alternative. We are thus brought to ask what states of affairs are good, what kind of things ought to exist for their own sake. Moore takes it that the things which ought to exist for their own sake are those which we call intrinsically good. How do we know what is intrinsically good? The answer is that we cannot fail to recognize the property of intrinsic goodness when confronted with it. Propositions concerning what is intrinsically good-as contrasted with what is good only because it is a means to something intrinsically good-are susceptible neither of proof nor of disproof. This is because good is the name of a simple, unanalyzable property, which Moore calls “non-natural” because it cannot be identified with any natural property. Moore holds that good is indefinable, partly in virtue of an analogy which he propounds between good and yellow, and partly by reason of an argument about the consequences of holding good to be definable. But both the analogy and the argument depend in part on the curious sense which he assigns to definition. To define, he holds, is to break up a complex whole into its constituent parts. So the definition of horse will be a statement to the effect that it has four legs, a head, a heart, a liver, etc., all of them arranged in definite relations to one another. (Moore recognizes other senses of definition, but deliberately puts them on one side.) Now if this is what is meant by definition, it is not difficult to agree that good is indefinable, but this sense of definition is so idiosyncratic that nothing has been gained. Moore also tries to reinforce his case by an appeal to what we allegedly must recognize when we hold a given notion “before” our minds. He says that if we consider good and, let us say, pleasant, or any other notion with which we might be tempted to confuse good, we can see that we “have two different notions before our minds.” This technique of holding one’s concepts up to the light, as it were, is reinforced by Moore’s method of calm assertion. More unwarranted and unwarrantable assertions are perhaps made in Principia Ethica than in any other single book of moral philosophy, but they are made with such well-mannered, although slightly browbeating certitude, that it seems almost gross to disagree. But what, then, is Moore’s case?

Moore originally rests his analogy between yellow and good on his notion of definition. “Yellow and good, we say, are not complex: they are nations out of which definitions are composed and with which the power of further defining comes.” Moreover, just as we cannot identify the meaning of yellow with the physical properties of the light which produces the effect of seeing yellow, so we cannot identify the meaning of good with the particular natural properties which are associated with good. It might be the case that anything “good” was pleasant, just as all yellow light is of a certain wavelength, but it does not follow that good means what pleasant means, any more than it follows that yellow means the same as “light of a certain wave length.”

Moore’s one genuine argument is used to show that good cannot be the name of any complex whole. Of any such whole, however defined, we can always significantly ask whether it is itself good. This argument can be deployed not only against the attempt to define good as the name of a complex, but also against the attempt to define it at all. Suppose that I do identify good with pleasant. My mistake can be exhibited by showing that I can always significantly ask of pleasure or of anything pleasant, Is it good? But if good named the same property that pleasant names, to ask, Is what is pleasant good? would be equivalent to asking, Is what is pleasant pleasant?-that is, it would be vacuously tautologous.

Moore frames this argument in order to refute hedonists, whom he conceives to hold two incompatible positions: they hold that pleasure is good, indeed the good, in a significant, nontautological sense; and they claim to demonstrate this by urging that good means nothing other than what pleasant means. But the first position requires that “pleasure is good” be taken to be analytic. Yet it cannot be both. So the hedonist position collapses. But, of course, it only collapses for those hedonists unwise enough to attempt to hold both these positions.

The philosophers whom Moore chiefly criticizes are J. S. Mill and Herbert Spencer. In the case of Mill, Moore’s criticisms are misdirected, if only because he reads into Mill a definition of good as meaning pleasant, whereas all that Mill at the most says is that pleasure provides us with our only criterion of goodness. It is by now almost a commonplace to recognize that Moore misrepresented Mill; it is a measure of the extent to which contemporary philosophers read Mill, but do not read Spencer, that it goes unrecognized that Moore also misrepresented Spencer. Moore accuses Spencer of having thought that good meant the same as “more evolved.” Spencer’s however was a far more complex, if quite implausible, position. Spencer held, first, that human society has evolved, just as the human species evolved, and indeed that the evolution of species and of society can be placed on a single continuous scale. Secondly, he believed that the higher a society is upon this scale the more ideal its morality; and thirdly, that conduct tends more and more toward the end of preserving life, it being assumed that in life there is, especially as one ascends toward the ideal, more pleasure than pain. As with Mill, Spencer may in unguarded moments have given the impression that he was defining the moral vocabulary. But the real Herbert Spencer is as far from being Moore’s straw man as is the real J. S. Mill.62

To the doctrine that good was the name of a natural property Moore gave the name “the naturalistic fallacy.” For Moore this fallacy is committed in the course of any attempt to treat good as the name of a property identifiable under any other description. Good cannot mean “commanded by God,” any more than it can mean pleasant, and for the same reasons the expression “the naturalistic fallacy” has since been adopted by the adherents of the view that one cannot logically derive an ought from an is; but although this latter doctrine is a consequence of Moore’s, it is not identical with it.

Is good, then, the name of a simple, unanalyzable property? To the doctrine that it is, there are at least two conclusive objections. The first is that we can only use the name of a simple property intelligibly where we are acquainted with some standard example of the property by reference to which we are to recognize whether it is present or absent in other cases. In the case of a simple property like yellow we can use standard examples of the color to recognize other cases of yellow. But how could having learned to recognize a good friend help us to recognize a good watch? Yet if Moore is right, the same simple property is present in both cases. To this, a disciple of Moore might reply that we are confusing the question by our example. A good watch is not “intrinsically” good. But how, then, do we recognize the intrinsically good? The only answer Moore offers is that we just do. Or put this point another way: Moore’s account could only reach the level of intelligibility if it were supplemented by an account of how the meaning of good is learned, and an account of the relation between learning it in connection with some cases, and knowing how to apply it in others.

The second objection is that Moore’s account leaves it entirely unexplained and inexplicable why something’s being good should ever furnish us with a reason for action. The analogy with yellow is as much a difficulty for his thesis at this point as it is an aid to him elsewhere. One can imagine a connoisseur with a special taste for yellow objects to whom something’s being yellow would furnish him with a reason for acquiring it; but something’s being “good” can hardly be supposed to furnish a reason for action only to those with a connoisseur’s interest in goodness. Any account of good that is to be adequate must connect it intimately with action, and explain why to call something good is always to provide a reason for acting in respect of it in one way rather than another.

That it does connect good with action is the chief virtue of the other seminal moral philosophy of the twentieth century, that of John Dewey. For Dewey the chief trap in all epistemology is the tendency to abstract our knowledge both from the methods by which we acquired it and from the uses to which we may put it. We only acquired whatever knowledge we have now because we had certain purposes, and the point of that knowledge is for us inseparable from our future purposes. All reason is practical reason. Moral knowledge is not a separate branch of knowledge; it is simply the knowledge we have-in physics, biology, history, or what you will-considered in relation to those purposes. To characterize something as good is to say that it will provide us with satisfaction in our purposes. As means or as end? As both, and Dewey is concerned to emphasize what he takes to be the interrelated character of good-as-a-means and good-as-an-end. We are as far as it is possible to be from Moore’s concept of the “intrinsically good” with its sharp separation of means and ends. Dewey concentrates on the agent, while Moore concentrates on the spectator. Dewey almost obliterates the distinction between fact and value, between is and ought, while Moore emphasizes it. Dewey thinks that in making choices we are guided by considerations which we express in statements of an ordinary empirical kind, statements which presuppose the direction of the agent’s purposes and interests, but which do not differ from, which in fact are the statements of our empirical studies. That Dewey has not been more influential, particularly in England, is perhaps explicable by the fact that he so seldom attends explicitly to the problem which has been at the center of Anglo-Saxon moral philosophy in this century, that of the meaning of moral predicates And where Dewey did exert a major influence it was indirectly, in a discussion that stemmed from Moore.

Moore’s immediate heirs were of two kinds. There were those who carried on moral philosophy of the same type as Moore’s, the so-called intuitionists, such as Prichard, Ross, and Carritt. It ought to be emphasized both that these writers did not in fact acquire their views from Moore, but independently, and that the value of their writings is not only a matter of how cogently they presented their own views. Carritt, for example, will be remembered for his power as a critic of utilitarianism (as well as for his writings on aesthetics). This particular succession of writers was ushered in by a text as dramatic in its way as Principia Ethica: a paper by H. A. Prichard in 1912, entitled “Does Moral Philosophy Rest upon a Mistake?” Prichard takes the task which moral philosophy has set itself to be that of providing a reason or justification for holding that something which we take to be our duty is indeed our duty. But he argues that the demand for such a reason or justification is utterly misconceived. In defense of this position he offers in effect two sorts of reason. I may try to justify the view that something is my duty by showing that it is to my interest or would lead to my happiness. But if this is what provides me with a reason, then I am not treating whatever I take my duty to be as a duty at all. For what I do because it is to my interest, I thereby do not do as a duty. Or I may try to justify the view that something is my duty by showing that to perform it would be to produce some good. But-so Prichard says-that something is good does not entail that it is obligatory on me to bring it about. This first kind of argument starts from a list of what Prichard presumably takes to be the only possible types of alleged justification. His second consists in appeals to that of which we are all alleged to be conscious. The apprehension of duty is said to be immediate and unquestionable, and therefore not to be supported by reasons.

Prichard’s outstanding characteristic, and one which he shares with Moore as well as with other intuitionists, is the treatment of good, right, duty, obligatory, and the rest of the moral vocabulary as though it was a coinage of permanently fixed values and simple scrutiny. It is doubtless because of this that the proportion of assertion to argument is so high in Prichard. In other intuitionist writers, such as Sir David Ross, who holds that we have independent intuitions of “rightness” and “goodness,” the standards of argument are much higher. But all intuitionist writers suffer from one difficulty: they are, on their own view, telling us only about what we all know already. That they sometimes disagree about what it is that we all know already only makes them less boring at the cost of making them even less convincing.

The two most powerful critics of intuitionism were R. G. Collingwood and A. J. Ayer. Collingwood, whose attack extended to many other recent writers in ethics, attacked them for their lack of historical sense, for their tendency to treat Plato, Kant, and themselves as contributors to a single discussion with a single subject matter and a permanent and unchanging vocabulary. They are, he says in his Autobiography, like men who translate the Greek word τριήρης by steamship, and when it is pointed out to them that the characteristics which Greek writers assign to the τριήρης are not at all the characteristics of a steamship, they reply that this just shows what odd and mistaken ideas about steamships Greek writers held. We ought rather, according to Collingwood in the Autobiography, to understand moral and other concepts in terms of a developing historical sequence. What this might entail I shall consider later in this chapter.

Ayer’s critique of intuitionism has quite different roots. In Language, Truth, and Logic he revived some of Hume’s positions, but did so in the context of a logical-positivist theory of knowledge. So moral judgments are understood in terms of a threefold classification of judgments: logical, factual, and emotive. In the first class come the truths of logic and mathematics, which are held to be analytic; in the second come the empirically verifiable or falsifiable truths of the sciences and of common-sense knowledge of fact. The third class necessarily appears as a residual category, a rag-bag to which whatever is not logic or science is consigned. Both ethics and theology find themselves in this category, a fact in itself sufficient to make us suspicious of the classification. For on the face of it, statements about the intentions and deeds of an omnipotent being and judgments about duty or about what is good do not obviously belong together. We can, however, easily detach the emotive theory of moral judgment from this dubious classification; all that we need retain from it is the contrast between the factual and the emotive. In this form the most powerful exponent of emotivism has been C. L. Stevenson. Stevenson’s writing exhibits many influences, above all, those of both Moore and Dewey, and his position can perhaps be most easily expounded by returning to Moore.

I said earlier that Moore had two kinds of heirs, the first of whom were the intuitionists. What the intuitionists continue is the philosophical appeal to what we all are alleged to recognize in moral matters. But Moore himself was, above all, anxious to clear up the philosophical confusions over the concept of goodness so that he could proceed to a second task, that of saying which things are, in fact, good. In his chapter on conduct he makes it clear that right action is valuable only as a means to what is good. In his chapter on The Ideal, Moore tells us what is good. “Once the meaning of the question is clearly understood the answer to it, in its main outlines, appears to be so obvious, that it runs the risk of seeming to be a platitude. By far the most valuable things which we know or can imagine, are certain states of consciousness which may be roughly described as the pleasures of human intercourse and the enjoyment of beautiful objects.” J. M. Keynes has described for us how this view of the supremacy of personal relationships and of the beautiful broke upon the generation immediately younger than Moore with all the force of a revelation. Almost half a century later Keynes could write: “I see no reason to shift from the fundamental intuitions of Principia Ethica; though they are much too few and too narrow to fit actual experience. That they furnish a justification of experience wholly independent of outside events has become an added comfort, even though one cannot live today secure in the undisturbed individualism which was the extraordinary achievement of the early Edwardian days.” It all depended, of course, on who “one” was and to which social class one belonged. The values which Moore exalts belong to the realm of private rather than public life; and, supremely important as they are, they exclude all the values connected with intellectual inquiry and with work. Moore’s values are those of a protected leisure, though it is in what he excludes rather than in what he does value that the parochial and classbound character of his attitudes appears. It is worth commenting on this feature of Moore’s views simply to emphasize the fact that they are not, as he apparently supposed, beyond controversy. For Moore combines highly controversial moral views with an appeal to the evidence of simple recognition in order to establish them. Keynes, in the memoir, quoted earlier, My Early Beliefs, gives us a penetrating account of the consequent behavior of Moore’s disciples. They would compare alternative possible situations and solemnly inquire in which there was most good, inspecting each in turn and comparing them. They would then announce what they “saw.”

In an extremely homogeneous group, like that of Moore’s immediate disciples, the congruence between what different people “see” is likely to be fairly high. But the arrival of D. H. Lawrence on the scene, who reacted against the attitudes of this group with all the passion at his disposal, might have made them aware that if challenged on their valuations, their own position allowed them no use for argument, but only for reinspection and reassertion. Since there is in fact no simple, nonnatural property which good names, the whole process is merely an elaborate game of bluff. And it would not be unfair to remark that what this group did was to invoke Moore’s philosophical theory in order to endow their own expressions of attitude with an authority which those expressions would not otherwise have possessed.

“But if there is no such property as Moore supposes, then all they can be doing is to express their own feelings.” Perhaps in some such reaction to Moore lies one of the seeds of emotivism. Moore himself staked everything on the appeal to objectivity. In an argument which he used in an essay written after Principia Ethica he contended that moral judgments cannot be reports of our feelings, for if they were, two men who uttered apparently contradictory judgments upon a moral issue would not in fact be disagreeing. A man who said, “You ought to do X” would no more be disagreeing with a man who said, “You ought not to do X” than would someone who said, “I stacked my hay yesterday” be disagreeing with someone else who said, “I didn’t stack my hay yesterday.” To which argument Stevenson’s reply was that two men who disagree on a moral question need not be involved in any factual disagreement; they need only be involved in disagreement on the facts of the case; the issue between them at this level can be settled by an empirical inquiry. But they may further disagree in their attitudes; and this disagreement can only be resolved by one party changing his attitude. The primary function of moral words, according to Stevenson, is to redirect the attitudes of others so that they accord more fully with our own.

In concentrating on the dynamic function of moral words, Stevenson shows the influence of Dewey, and it is through Stevenson mainly that Dewey influenced later moral philosophers.

Moral words are able to have this dynamic function of which Stevenson speaks because they are emotive. “The emotive meaning of a word is the tendency of a word, arising through the history of its usage to produce (result from) affective responses in people.” Ayer, in his version of the emotive theory, concentrated upon my expression of my own feelings and attitudes; Stevenson, in his, concentrates upon my attempt to influence your feelings and attitudes. As to what the key moral words mean, Stevenson offers two models, stressing in each case that the nature of emotive meaning is such that we cannot hope to arrive at more than a rough approximation. His first model is one in which “This is good” is elucidated as roughly equivalent to “I like this. Do so as well.” In his second model he attends to those expressions which embody what he names “persuasive definitions.” Such expressions have a descriptive meaning, and they associate with that meaning an emotive one; we can thus always analyze them into two component elements. Two men in controversy may, for example, use just in such a way that each associates different descriptive meanings with the emotive element in the meaning of that word.

Stevenson’s view of moral expressions leads to a number of other positions. It follows from his view, for example, that for good and for other evaluative expressions no complete definition in descriptive terms can ever be given. Thus Stevenson agrees with Moore that good cannot function as the name of a natural (empirically descriptive) property. The facts are logically divorced from the evaluations for Stevenson as much as for Moore. Secondly, Stevenson commits himself to the view that philosophical ethics is a morally neutral activity. The doctrines that we hold about the meaning of moral expressions cannot commit us to any particular moral view. Clearly the emotive theory itself, if true, does appear, at least on the surface, to be morally neutral. For presumably we can use emotive words to commend any class of actions whatsoever. Moreover, if Stevenson is right, evaluative disagreement may always be interminable. There is no limit to the possibilities of disagreement, and there is and can be no set of procedures for the resolution of disagreements. It is not surprising that this should be a consequence of Stevenson’s position, since he himself initially laid it down as one of the prerequisites for a successful theory that it should provide for disagreement to be interminable. Finally, on Stevenson’s view, the reasons which we cite to support our evaluative, and more specifically, our moral judgments cannot stand in any logical relationship to the conclusions which we derive from them. They can only be psychological reinforcements. It follows that words like because and therefore do not function as they do in other parts of discourse.

The difficulties which can be raised about emotivism are of several different kinds. The notion of “emotive meaning” is itself not clear. What makes certain statements guides to, or directives of, action is not that they have any meaning over and above a factual or descriptive one. It is that their utterance on a specific occasion has import for, or relevance to, the speaker or hearer’s interests, desires, or needs. “The White House is on fire” does not have any more or less meaning when uttered in a news broadcast in London than it does when uttered as a warning to the President in bed, but its function as a guide to action is quite different. Emotivism, that is, does not attend sufficiently to the distinction between the meaning of a statement which remains constant between different uses, and the variety of uses to which one and the same statement can be put. (Of course, meaning and possible range of use are intimately related; but they are not the same.)

Moreover, not only does Stevenson tend to conflate meaning and use, but, the primary use which he assigns to moral expressions is not, and cannot be, their primary use. For the use to which he attends is the second-person use in which we try to move other people to adopt our own views. Stevenson’s examples all picture a thoroughly unpleasant world in which everyone is always trying to get at everyone else. But in fact one is only in a position to try to convert others to one’s own moral views when one has formed views of one’s own; yet none of those uses of moral language which are necessary to the formation and expression of one’s views with an eye to one’s own actions figure in Stevenson’s initial account.

Finally, one can justifiably complain of the emotive theory not only that it is mistaken, but also that it is opaque. For its proponents seek to elucidate moral expressions in terms of the notions of attitudes and feelings, and it is relevant to ask for further characterization of the attitudes and feelings in question. How, for example, are we to identify these attitudes and feelings so that we may distinguish them from other attitudes and feelings? Emotivist writers are, in fact, largely silent on this point; but the suspicion is strong that they would be compelled to characterize the attitudes and feelings under discussion as just those attitudes and feelings which are given their definitive expression in acts of moral judgment. Yet if this is so, the whole theory is imprisoned in uninformative circularity.

Nonetheless, some of its central features are preserved by its immediate successors. The moral neutrality of philosophical analysis, the logical gap between fact and value, the interminality of disagreement all remain upon the scene. What is altered in later writers is the attention paid to two intimately related topics, the question of the criteria which are employed in calling things, acts, or people good or bad, and the question in the nature of moral reasoning. If I call something good or commend it in some other way, I can always be asked upon what criterion I am relying. If I say that I ought to do something, I can always be asked, And what if you do not? and, On account of what ought you to do it? What is the relation between my answers to these questions and my beliefs as to what is good and as to what I ought to do?

One systematic answer to these questions is to be found in R. M. Hare’s The Language of Morals, and his views are added to and further elucidated in Freedom and Reason. Hare specifies the nature of moral language by means of an initial distinction between prescriptive and descriptive language. Prescriptive language is imperatival, in that it tells us to do this or that. It is itself subdivided into two classes, that comprising imperatives in the ordinary sense, and that comprising properly evaluative expressions. All value judgments are practical, but in different ways. Ought sentences, for example, if they are genuinely evaluative, entail imperatives addressed to anyone in the relevant situation, and anyone here includes the person who utters the sentence. The criterion of uttering the ought sentence sincerely is that, on the relevant occasion and if the speaker can, he does in fact act in obedience to the imperative entailed by the ought which he utters to himself. Good, by contrast, is used to commend; to call X good is to say that it is the kind of X we should choose if we wanted an X. The criteria which I employ in calling something good are criteria which, if I am engaged in genuine evaluations, I have chosen, and which I endorse by my very use of them. Evaluative expressions and moral rules are thus both expressions of the agent’s fundamental choices. But the role of choice in Hare’s prescriptivism is far clearer and far less objectionable than the role of attitudes or feelings was in emotivism. Unlike the latter, it does not preclude the use of argument in morals.

Hare was, in fact, a pioneer in the logical investigation of imperatives. He pointed out that in imperatival discourse, conclusions can follow from premises in a perfectly straightforward way, violating none of the ordinary rules of entailment. Because and therefore carry their usual meanings, and genuine moral argument is possible. But, so Hare further holds, the meaning of evaluative prescriptive expressions is such that no evaluative or prescriptive conclusion can follow from premises which do not include at least one evaluative or prescriptive premise. In other words, Hare reiterates the thesis that no ought follows merely from is. So far as the doctrine of The Language of Morals goes, it seemed to follow that the pattern of moral argument is a transition from a moral major premise and a factual minor premise to a moral conclusion. Wherever I appear to pass from fact to value (“I ought to help this man because he is starving”), there is a gap in the argument, a concealed major premise (“I ought to help the starving”). This major premise itself may figure as the conclusion of some other syllogism, but at some point the chain of reasoning must terminate in a principle which I cannot justify by further argument, but to which I must simply commit myself by choice. Once more it seems to follow, as it did with emotivism, that on matters of ultimate principle, assertion cannot be met by argument but only by counterassertion.

In Freedom and Reason, Hare argued that this was not entailed by his view; that the universalizability of moral judgments provides an argumentative weapon against those who hold unacceptable moral principles. For of a man who holds, for example, that other men ought to be treated in certain unpleasant ways merely because their skins are black, we can always ask, Are you then prepared to allow that you should be treated in the same way if your skin were black? And Hare believes that only a minority, whom he denominates fanatics, would be prepared to accept the consequences of replying Yes to this. This last contention is a question of fact on which I believe that recent social history does not bear Hare out. But I do not want to quarrel with this part of Hare’s view so much as to emphasize that it still remains true on Hare’s view that, as a matter of logic and of the concepts involved, what I call good and what I hold I ought to do depend upon my choice of fundamental evaluations, and that there is no logical limit to what evaluations I may choose. In other words, Hare’s prescriptivism is, in the end, a reissue of the view that behind my moral evaluations there is not and cannot be any greater authority than that of my own choices. To understand evaluative concepts is to understand that our use of these concepts does not of itself commit us to any particular set of moral beliefs. The criteria for true belief in matters of fact are independent of our choices; but our evaluations are governed by no criteria but those which we ourselves choose to impose upon them. This is a repetition of Kant’s view of the moral subject as lawgiver; but it makes him an arbitrary sovereign who is the author of the law that he utters, and who constitutes it law by uttering it in the form of a universal prescription.

An ambiguity in Hare’s whole enterprise, an ambiguity pointed out by Mary Warnock,63 becomes important here. When Hare characterizes evaluation and prescription, is he in fact defining these terms in such a way as to protect his thesis against possible counterexamples? If we produce an example of ought which does not entail a first-person imperative, or an example of good in which the criteria are not a matter of choice, will Hare be able to reply that these are simply nonprescriptive and nonevaluative uses of ought and good? Hare certainly recognizes that there are some nonprescriptive and nonevaluative uses. But if he has simply legislated so that evaluation and prescription shall be what he says they are, why should we assent to his legislation? If he is not legislating, then we must have the class of evaluative and prescriptive expressions delimited for us independently of Hare’s characterization, in a way that Hare himself never delimits it. He seems indeed to rely on an almost intuitive understanding of what is to be included or left out of the class of evaluative expressions.

Why is this important? It is important partly because Philippa Foot64 and Peter Geach65 have challenged Hare with primafacie convincing counterexamples. Philippa Foot’s attention has been concentrated on evaluative expressions connected with the virtues and vices, such as rude and courageous; Geach’s on good and evil. The criteria for the correct application of rude and courageous are, so Mrs. Foot argues, factual. If certain factual conditions are fulfilled, this is sufficient to show that these epithets apply and their application could only be withheld by someone who failed to understand their meaning. So if a man at a concert spits in the face of an acquaintance whom he knows slightly and who has done nothing hostile to him, then he is certainly rude. Equally, if a man with a reasonable prospect of saving the lives of others by sacrificing his own does sacrifice his own life, he is certainly courageous. But in each of these cases, when we show that the necessary and sufficient conditions apply to justify the epithet, we could rewrite what we say so that the necessary and sufficient conditions appear as premises which in virtue of the meaning of rude or courageous entail the conclusion “So he was rude” or “So he was courageous.” But if any conclusions are evaluative, these are. Thus some factual premises do appear to entail evaluative conclusions.

Equally, it is clear that in many cases at least where I call something or someone good, the appropriate criteria are determined by the kind of case it is and are not open to choice. The criteria for calling something “a good X” depend, as Geach has pointed out, on the nature of X. “A good watch,” “a good farmer,” “a good horse,” are cases in point. But what about “a good man”? Here surely, it might be argued, we do use a variety of criteria and we have to choose between them. Here surely, an argument like Hare’s is the convincing one. I do not want to pursue this as yet unfinished argument further; I want rather to inquire what sort of argument it is, and why it arises.

It is important to see that a whole range of interconnected differences of view are involved here. On the one side, it is held that facts can never entail evaluations, that philosophical inquiry is neutral between evaluations, that the only authority which moral views possess is that which we as individual agents give to them. This view is the final conceptualization of the individualism which has had recurrent mention in this history: the individual becomes his own final authority in the most extreme possible sense. On the alternative view, to understand our central evaluative and moral concepts is to recognize that there are certain criteria we cannot but acknowledge. The authority of those standards is one that we have to recognize, but of which we are in no way the originators. Philosophical inquiry, which reveals this, is therefore not morally neutral. And factual premises do on occasion entail evaluative conclusions.

Each view systematically insulates itself from the other by its choice of examples. And neither will allow that the issue between them could be settled by an empirical inquiry into the way in which evaluative concepts are actually used. For each is quite prepared to allow that the ordinary usage in morals may on occasion be confused, or indeed perverted, through the influence of misleading philosophical theory. Perhaps, however, this controversy is one that cannot be settled, and perhaps the reason why it cannot be settled can be seen if we try to place in historical perspective the concepts which generate it. But before we can do that we must consider certain very unsophisticated points which locate this controversy as one not merely for philosophers, but for all contemporary moral agents.

Emotivism and prescriptivism initially alienate us because their explanations of evaluative language in terms of the notions of feelings, liking, choice, and imperatives leave us asking why there should exist any specifically evaluative language over and above the ordinary language of feelings, liking, choice, and imperatives. When I say, “You ought to do this,” or when I say, “This is good,” I want to protest that I say more and other than, “You or anyone else-do this!” or, “I like this. Do so as well.” For if that is what I mean, that is what I could and would say. If that is what I do say, then certainly what I say will have no authority but that which I confer upon it by uttering it. My attitudes and my imperatives have authority for me just because they are mine. But when I invoke words such as ought and good I at least seek to appeal to a standard which has other and more authority. If I use these words to you, I seek to appeal to you in the name of those standards and not in my own name. Yet even though this may be what I seek to do, it does not necessarily follow that I succeed. Under what conditions might I succeed? Under what conditions must I fail?

Suppose a society of the kind which I tried to characterize when I discussed Greek society, in which the form of life presupposes agreement on ends. Here there are agreed criteria for the use of good, not only when we speak of “good horse” and “good farmer” but also when we speak of “good man.” In this society there is a recognized list of virtues, an established set of moral rules, an institutionalized connection between obedience to rules, the practice of virtues, and the attainment of ends. In such a society the contrast between evaluative language and the language of liking or of choice will be quite clear. I may tell you what I like or choose, and I may tell you what you ought to do; but the second makes a claim upon you which the first does not. You may disregard what you ought to do through annoyance or negligence; but you cannot use the moral vocabulary and consistently deny the force of ought, and you cannot remain within the social commerce of the community, and abandon the moral vocabulary.

Is moral criticism in such a society impossible? By no means; but it must proceed by an extension of, and not by a total break with, the established moral vocabulary. Does this mean that the authority of the morality does not extend beyond the community whose social practices are in question? One is tempted to reply, Does the authority of arithmetical rules extend beyond the community in which the practice of counting is established? This is intended as a genuine, and not as a rhetorical question, which deserves a fuller answer; but at the least, to connect rules and social practice in this way is not obviously to give moral rules less of a hold on us than mathematical, except that no society could advance far without the same type of simple counting, whereas there can be wide variations in the social practice to which moral rules are relevant.

In discussing Greek society, I suggested what might happen when such a well-integrated form of moral life broke down. In our society the acids of individualism have for four centuries eaten into our moral structures, for both good and ill. But not only this: we live with the inheritance of not only one, but of a number of well-integrated moralities. Aristotelianism, primitive Christian simplicity, the puritan ethic, the aristocratic ethic of consumption, and the traditions of democracy and socialism have all left their mark upon our moral vocabulary. Within each of these moralities there is a proposed end or ends, a set of rules, a list of virtues. But the ends, the rules, the virtues, differ. For Aristotelianism, to sell all you have and give to the poor would be absurd and meanspirited; for primitive Christianity, the great-souled man is unlikely to pass through that eye of the needle which is the gateway to heaven. A conservative Catholicism would treat obedience to established authority as a virtue; a democratic socialism such as Marx’s labels the same attitude servility and sees it as the worst of vices. For puritanism, thrift is a major virtue, laziness a major vice; for the traditional aristocrat, thrift is a vice; and so on.

It follows that we are liable to find two kinds of people in our society: those who speak from within one of these surviving moralities, and those who stand outside all of them. Between the adherents of rival moralities and between the adherents of one morality and the adherents of none there exists no court of appeal, no impersonal neutral standard. For those who speak from within a given morality, the connection between fact and valuation is established in virtue of the meanings of the words they use. To those who speak from without, those who speak from within appear merely to be uttering imperatives which express their own liking and their private choices. The controversy between emotivism and prescriptivism on the one hand and their critics on the other thus expresses the fundamental moral situation of our own society.

We can in the history of moral philosophy situate certain writers usefully in terms of this account. Kant, for example, stands at the point at which the loss of moral unity means that morality can be specified only in terms of the form of its rules, and not of any end which the rules may serve. Hence his attempt to derive the content of moral rules from their form.

Kant also stands at the point at which moral rules and the goals of human life have become divorced to such a degree that it appears both that the connection between abiding by the rules and achieving the goals is merely a contingent one and that, if this is so, it is intolerable. It is Kant’s grasp of the former point, as well as of the vagueness about goals which had led to the notion of happiness becoming vague and indefinite, that leads him to enjoin us to seek not to be happy, but to be deserving of happiness. It is his grasp of the latter point that leads him to invoke God as a power that will crown virtue with happiness after all. Kant seeks to hold together an earlier and a later view of morals; the tension between them is apparent.

Eighteenth-century English moralists and nineteenth-century utilitarians write from within a society in which individualism has conquered. Hence they present the social order not as a framework within which the individual has to live out his moral life, but as the mere sum of individual wills and interests. A crude moral psychology makes of moral rules instructions as to effective means for gaining the ends of private satisfaction. Hegel, Green, and to a lesser extent, Bradley are not only critics of this view of morals, they try to specify the type of community within which the moral vocabulary, can have a specific and distinctive set of uses. But the philosophical analysis of the necessary form of such community is no substitute for the deed of re-creating it; and their natural successors are the emotivists and the prescriptivists, who give us a false account of what authentic moral discourse was, but a true account of the impoverished meanings which evaluative expressions have come to have in a society where a moral vocabulary is increasingly emptied of content. Marx resembles Hegel and the English idealists in seeing a communal framework as presupposed by morality; unlike them, he sees that it no longer exists; and he proceeds to characterize the whole situation as one in which moralizing can no longer play a genuine role in settling social differences. It can only be an attempt to invoke an authority which no longer exists and to mask the sanctions of social coercion.

All this of course does not entail that the traditional moral vocabulary cannot still be used. It does entail that we cannot expect to find in our society a single set of moral concepts, a shared interpretation of the vocabulary. Conceptual conflict is endemic in our situation, because of the depth of our moral conflicts. Each of us therefore has to choose both with whom we wish to be morally bound and by what ends, rules, and virtues we wish to be guided. These two choices are inextricably linked. In choosing to regard this end or that virtue highly, I make certain moral relationships with some other people, and other moral relationships with others impossible. Speaking from within my own moral vocabulary, I shall find myself bound by the criteria embodied in it. These criteria will be shared with those who speak the same moral language. And I must adopt some moral vocabulary if I am to have any social relationships. For without rules, without the cultivation of virtues, I cannot share ends with anyone else. I am doomed to social solipsism. Yet I must choose for myself with whom I am to be morally bound. I must choose between alternative forms of social and moral practice. Not that I stand morally naked until I have chosen. For our social past determines that each of us has some vocabulary with which to frame and to make his choice. Nor can I look to human nature as a neutral standard, asking which form of social and moral life will give to it the most adequate expression. For each form of life carries with it its own picture of human nature. The choice of a form of life and the choice of a view of human nature go together.

To this view, each side in the contemporary philosophical controversy will reply in its own terms. The emotivists and prescriptivists will stress the role of choice in my account. Their critics will stress the way that the agent must come to the act of choice with an already existing evaluative vocabulary. Each will try by their choice of examples to redefine their opponent’s case away. And the same attempt is already being made in other controversies elsewhere. Indeed, it is a reinforcement for the view that this philosophical controversy is an expression of our social and moral situation that it should have occurred in quite a different context in the arguments that have proceeded in France between Catholic moralists, Stalinists, Marxists, and Sartrian existentialists.

For both Catholics and Stalinists the moral vocabulary is defined in terms of certain alleged facts. Each has their own characteristic list of virtues. For Sartre, by contrast, at least for the Sartre of the immediate postwar period, to live within a readymade moral vocabulary is necessarily an abdication of responsibility, an act of bad faith. Authentic existence is to be found only in a self-conscious awareness of an absolute freedom of choice. Kierkegaard’s view of the act of choice is detached from its theological context, and made by Sartre the basis for political as well as for moral decision. Sartre does not locate the source of the necessity of the act of choice in the moral history of our society, any more than Kierkegaard did. He locates it in the nature of man: a conscious being, être-pour-soi, differs from a thing, être-en-soi, in his freedom and his consciousness of freedom. Hence men’s characteristic experiences of anxiety before the gulf of the unmade future, and their characteristic attempts to pretend that they are not responsible. Thus Sartre locates the basis of his moral view in a metaphysics of human nature, just as much as the Catholic or the Marxist does.

Like Sartre, the prescriptivist and emotivist do not trace the source of the necessity of choice, or of taking up one’s own attitudes, to the moral history of our society. They ascribe it to the nature of moral concepts as such. And in so doing, like Sartre, they try to absolutize their own individualist morality, and that of the age, by means of an appeal to concepts, just as much as their critics try to absolutize their own moralities by means of an appeal to conceptual considerations. But these attempts could only succeed if moral concepts were indeed timeless and unhistorical, and if there were only one available set of moral concepts. One virtue of the history of moral philosophy is that it shows us that this is not true and that moral concepts themselves have a history. To understand this is to be liberated from any false absolutist claims.