I WILL first discuss the meaning of ‘right’, and I will begin by some discussion of the evolutionary type of view. On historical grounds it is justifiable to treat this school as a whole, though it has in fact not limited itself to one type of definition of ethical terms. It has at times tended to define ‘right’ by reference to consequences—to the promotion of life—and at times to define it by reference to the approval of the community; and at times it has tended to define it in yet a third way, which does not belong either to the attitude type or to the consequence type—to define right conduct as evolved conduct.
The evolutionary and sociological school of thought has on the whole shown little if any awareness of the distinction between two questions which are logically entirely different. One is the question as to the meaning of such terms as ‘right’ or ‘obligatory’, as to what it is that we intend to say about conduct when we describe it as right or obligatory. The other is the question what is the other characteristic, or what are the other characteristics, in virtue of which we describe conduct as having the characteristic of being right or obligatory. The method usually followed by this school is to pass under review a variety of types of act that are commonly called right; to find, or argue, that they have some characteristic in common, e.g. that of being comparatively highly evolved; and then to assume that that is what ‘right’ or ‘obligatory’ means. But it is clear that, assuming the review of instances to be adequate, and the discovery of a common characteristic to be correct, two possibilities remain. The common characteristic thus discovered may be what we mean by ‘right’; or it may be a characteristic on which rightness is consequent but which is itself different from rightness; not the essence of rightness, but its ground. Now what we are considering at present is views as to the meaning of rightness, and it is surely obvious that the supposition that ‘right’ means ‘comparatively evolved’ is not one that can be seriously entertained. If we ask ourselves what ‘more evolved’ means, we shall find in it, I think, two main elements: (1) that conduct so described comes, in time, after a process of evolution of more or less duration, and (2) that it has a characteristic which usually emerges in the course of evolution, that of being complex, in comparison with the simple activities which appear in an early stage of evolution. And it is surely clear that neither temporal posteriority nor complexity, nor the union of the two, is that which we mean to refer to when we use the term ‘right’ or ‘obligatory’. Even if it be true that there is a perfect correspondence between the characteristic of being right and that of being more evolved, such that neither is ever found without the other, there is really no resemblance between the characteristic which we have in mind when we say ‘right’ or ‘obligatory’ and that which we have in mind when we say ‘more evolved’. The claim to have found the meaning of rightness must be completely rejected. We are left, however, with two possibilities as to the underlying intention of those who carelessly claim in this way to have discovered the meaning of rightness. (1) They may be thinking that some actions genuinely have the characteristic of rightness, and that they have discovered another quality or characteristic on which this always depends. In this case theirs will not be at bottom a naturalistic account of rightness, and there will be nothing in their view which on general grounds need be rejected by those who hold that rightness is an indefinable characteristic which certain actions have and others have not. It will simply be a matter for inquiry in detail whether the characteristic of evolvedness is in fact that whose presence in an action renders the action right. Now to a large extent the coincidence between evolvedness and rightness which is claimed to exist may be admitted to exist. Let me anticipate the view I will later put forward, so far as to say that one of the main grounds on which we regard actions as right is their presumed tendency to bring as much good as is possible into being for all men, or for all sentient beings. Let us accept the general truth of the evolutionary account which describes man as having evolved from the condition of an animal seeking food and safety for itself and for its offspring, but without any interest in any wider community, to his present condition, in which many men consciously seek the good of their whole class or community, and some consciously seek the good of the whole human race. If this be true, then there will be a tendency for the acts that come later in the course of evolution to envisage and to promote a wider good than those that come earlier. And if ‘more evolved’ means, not merely coming later in the course of evolution, but sharing in the general characteristics that accompany the course of evolution, the more evolved acts will tend to have, in comparison with the less evolved, the characteristic of promoting a wider range of goods, and will on that account tend more to be right acts.
But even if we admit that the characteristics of being highly evolved, of tending to promote the maximum good, and of being right tend to a large extent to go together, we must surely recognize a closer relation between the last two characteristics than between the first and the third. It will be not because of the merely historical fact that they come later in the course of evolution, but because they share in a characteristic common to the later stages of evolution, the characteristic of being promotive of a wider good, that acts will tend to be right. The evolutionary account will be encouraging, in so far as it gives us reason to believe that, as time has gone on, actions envisaging a wider good have tended to become commoner, and that the same process may be expected to continue; but it will not have thrown any light on the question what makes right acts right, any more than on the question what rightness is. No particular position in the temporal process, be it late or early, has as such any tendency to make an act right.
(2) So far I have been dealing with one of two alternatives, that the evolutionary account recognizes the existence of the attribute of rightness and is looking for its ground. But another alternative remains. May the upshot of the evolutionary account be, not that obligatoriness just means evolvedness, nor that evolvedness is the ground of obligatoriness, but that there is no such thing as obligatoriness; that there is nothing in reality answering to the meaning which we have in mind when we use the word obligatory, the only distinction that remains being that between less and more evolved acts? This was certainly not Herbert Spencer's intention; there never was a more serious moralist, one more persuaded that there is an objective difference between right and wrong. Nor again was it the intention of Huxley, who in his famous Romanes Lecture urged that we are under an obligation to reverse to a large extent the tendency of the evolutionary process. Yet there is no doubt that in many minds the study of evolutionary science has tended to produce scepticism about the difference between right and wrong. That has come about in this way. Before the evolutionary theory had been put forward, the question as to the origin of the idea of obligation had hardly arisen. So long as man was thought to be a species standing in no historical continuity with the lower animals, it was possible to regard the idea of obligation as an original and permanent endowment of the human race. But we cannot now refuse to accept the view that the human race is historically continuous with lower animal species, which we cannot credit with having had any idea of obligation. The question of the origin of the idea becomes a pressing one, and the position is often adopted that either the idea of obligation must be a complex one, a sort of amalgam put together by a sort of mental chemistry out of simpler elements which we can ascribe to our animal ancestors, or that if it is not this, it is a fanciful and illegitimate invention belonging to a fairly late stage in evolution.
To this the answer has often been made, that the question of origin has no logical connexion with that of validity, and that in particular the validity of the idea of obligation cannot be impaired by any problem as to how the idea arose. This answer I cannot accept in its entirety. I venture to quote again from my earlier book.
‘An inquiry into the origin of a judgement may have the effect of establishing its validity. Take, for instance, the judgement that the angles of a triangle are equal to two right angles. We find that the historical origin of this judgement lies in certain pre-existing judgements which are its premisses, plus the exercise of a certain activity of inferring. Now if we find that these pre-existing judgements were really instances of knowing, and that the inferring was also really knowing—was the apprehension of a necessary connexion—our inquiry into the origin of the judgement in question will have established its validity. On the other hand, if any one can show that A holds actions of type B to be wrong simply because (for instance) he knows such actions to be forbidden by the society he lives in, he shows that A has no real reason for believing that such actions have the specific quality of wrongness, since between being forbidden by the community and being wrong there is no necessary connexion. He does not, indeed, show the belief to be untrue, but he shows that A has no sufficient reason for holding it true; and in this sense he undermines its validity.’1
The question is, whether evolutionary theories have done this. Now the solid fact with which we start is that we now have the thought that certain acts are right and others wrong. And it seems clear—I will attempt to argue for this in more detail later2—that by right and wrong we do not mean ‘commanded, or forbidden, by the community’. The question then is, how did we come by this thought? Two answers may be given. One is that by an exercise of fancy we came to believe, without justification, that certain actions have this character. The other is that, having up to a certain time recognized only naturalistic characteristics of actions, such as that they conduced to survival or that they were commanded by the community, the human mind, when it had reached a certain degree of maturity, became able to detect in actions the non-naturalistic characteristic of rightness. Whichever account is given, we are crediting the human mind with having made at some time a new departure—either the fanciful invention of a new idea, the idea of rightness, or the detection of a hitherto undetected characteristic of actions. In either case a breach of continuity is involved, and that involved in the former case is certainly no more easy to understand than that involved in the latter. We are perfectly familiar with the fact that within the limits of a single life a mind may pass from a state in which it is quite incapable of forming certain ideas or of making certain judgements, to one in which it is capable of doing this, and we do not doubt the truth of our mature judgements because we were earlier incapable of making them; we do not for that reason treat them as mere plays of fancy. We recognize that the truths in question—say, mathematical truths—were there all the time to be apprehended, but that a certain degree of mental maturity was necessary for their apprehension. And if this be so within the limits of a single life, it is only natural to suppose that the growth and ripening of mind from generation to generation which has taken place in the history of evolution was similarly the necessary condition for the apprehension of certain truths which were there all the time to be apprehended.
There is, however, another way in which the work of the evolutionary school tends to produce in some minds a scepticism about moral principles. The studies of the comparative sociologists have revealed very clearly the great variety of views on moral questions that exists in different societies, and this sometimes leads to the belief that there are no objective moral principles, but only a variety of codes adopted by various communities. And this scepticism is reinforced by the fact that even within the same society and as between people of approximately the same degree of mental development very different views are often held on moral questions. This might seem to make the position about moral truth very different from that with regard to mathematical truth. Mathematical truths are accepted unanimously by those who have reached sufficient mental maturity; but mental maturity is no sufficient guarantee of agreement on moral questions.
Yet on examination the diversity of opinion on moral questions is found to rest not on disagreement about fundamental moral principles, but partly on differences in the circumstances of different societies, and partly on different views which people hold, not on moral questions but on questions of fact. Professor Taylor has pointed out3 that the approval of the blood-feud by some societies and its condemnation by others is explicable by the simple fact that in an early and unsettled state of society, where there is no proper provision for the public punishment of murderers, private vengeance is the only way of securing respect for life, while in a more settled state of society this is better left to the arm of the law. That is an example of the fact that an actual change of circumstances may make that wrong which once was right (or at least more wrong that which was less wrong), without any variety of fundamental moral principles being involved. Or again, where there is no variation in the outward circumstances, there may be a difference of view on some non-moral question which leads to a difference of view on a secondary moral question, while the same fundamental moral principle may be accepted by both parties. To quote two more of Professor Taylor's examples,4 the difference between those who think vaccination right and those who think it wrong turns largely on a difference of opinion on the question of fact whether vaccination does or does not prevent smallpox, while both parties accept the principle that parents should try to save their children from disease. And the difference of opinion between fox-hunters and those who condemn fox-hunting turns largely on a difference of view as to the comparative intensity of the pain of the fox and the enjoyment of his hunters.
The more we examine differences of opinion on the media axiomata of morals, the more we shall find them not to depend on divergence on fundamental moral questions, but either on the different circumstances of the differing parties or on the different opinions they hold on ordinary matters of fact. Now where a difference of circumstances causes one type of act to be judged right in one state of society and another in another, no doubt is cast on either moral judgement by the fact that if they are stated abstractly, without reference to the difference of circumstances, they contradict one another. And where different acts are judged right owing to a difference of opinion on a question of fact, it is not on the fundamental moral judgement that is accepted by both parties, but on the opinions about ordinary fact which form their minor premisses, that doubt is cast.
Yet it would be a mistake to regard the differences of opinion on moral questions as due entirely to these two causes. For while all men are probably at bottom agreed in thinking we ought to produce as much that is good as we can, and agreed also as to the goodness of certain things—virtue, intelligent thought, and happiness—there is a real difference of opinion as to the comparative worth of different goods; and this is a difference not on an ordinary matter of fact but on a moral question. It is in this region—in the comparative valuation of things that are agreed to be valuable—that the source of many of our differences of moral opinion is to be found.
Yet even if we admit the existence of great differences of opinion in this region, that does not really justify the conclusion that in this region there is not an objective truth to be known, any more than difference of opinion about ordinary matters of fact or about scientific questions justifies the conclusion that there there is no objective truth. The very fact of difference of opinion is itself evidence of the persisting confidence of all of us that there is an objective truth. To find a difference of opinion between ourselves and others, or between our own ages and previous ages, should weaken perhaps our confidence in our own opinions, but not weaken our confidence that there is some opinion that would be true. And I think we can go farther than this. We may feel doubt about many scientific doctrines which hold the field at any given moment. We may think that a new theory may for a time displace an older theory which was nearer the truth. But we do not really doubt that, in the main, science progresses towards a truer and truer view of the nature of the physical universe. We do not seriously question that we are nearer the truth about the physical universe than were the Greeks, or the men of the Middle Ages, or indeed the men of any century previous to our own. Individual vagaries of opinion may for a time prevail, but there is going on all the time the steady work of men whose purpose is to discover the truth, and the truth is there to be discovered. It is only natural then that mankind should, though with many set-backs, progress in the main steadily towards truer views. The same is in a measure true of moral questions. There is going on all the time a steady devotion, on the part of many people, to the task of discovering moral truth. Individual interests may often draw us aside from the truth, to over-estimate goods that will be enjoyed by ourselves or our class or our country, and to under-estimate those that will be enjoyed by other individuals or classes or countries. Probably the search for moral truth has been more affected by selfish interests than the search for scientific truth. But the moral facts are there, and disinterested thinking about them is always going on; and in the end the facts tend to prevail and to win our assent. And this progress goes on not only within individual lives. Each individual does not start where his predecessors started. He absorbs the new discoveries that have been made in the previous generation. Nor is the progress always gradual. Every now and then there arises in the course of history a genius who discovers some great moral truth which only needs to be proclaimed to be generally recognized; and all who come under his influence find their whole moral insight lifted to a higher plane. In the main, then, we need not doubt that man progresses fairly steadily towards moral truth as he does towards scientific.
There occur, also, periods in which mankind appears to sink to a lower moral plane, in which old moral standards are given up and moral lawlessness sets in. But these need not disquiet us too much. What is questioned in such periods is not, as a rule, the fundamental principles of morality but the media axiomata, the rules for which no a priori evidence can be claimed but which rest partly on circumstances that have ceased to exist, and partly on opinions about ordinary fact that have been given up. No doubt many people whose nature it is to escape from all moral restrictions will turn such a period of questioning to their own account. But the questioning itself is often perfectly sincere, and springs from a desire to get down to bedrock in morality; and this is all to the good. Moral codes that will not survive such questioning do not deserve to survive it, and those that do deserve to survive it will do so. Magna est Veritas et praevalet. Acquiescence in moral codes merely because they are accepted by the society in which one lives spells death to progress in moral insight. The honest questioning of old rules, when it has led to their abandonment, has led to the discovery of new ones which have usually been much more rigorous, demanding an inner morality which is harder to practise than outward conformity to a code. Thus the periods in which the old tables of the law are broken form no real exception to the general progress in moral insight which we are justified in believing to have taken place.
These considerations have led us rather far afield from the study of evolutionary theories of morality, though they arose legitimately out of it. I return to repeat that evolutionary theories do not seem to have offered us anything that can be accepted as a definition of ‘right’ or ‘obligatory’.
From this short study of evolutionary attempts at defining the meaning of ‘right’, I will turn to the group of theories which attempt to define the rightness of action by reference to the attitude adopted towards the action by some mind or minds. As we have seen, some evolutionary moralists have themselves tended to adopt such a theory; they have identified ‘right’ with ‘approved by the community’. But it will be well to examine this view, apart from its connexion with evolutionary studies, as one among the group of mental-attitude theories.
Professor Broad divides attitude theories into the two varieties of private and public. ‘If, e.g., a man holds that a “right” action means an action which evokes in him a certain kind of emotion when he contemplates it, he is a Private Psychological Naturalist. If he holds that a “right” action means one which evokes a certain kind of emotion in all or most men, or in all or most Englishmen, or in all or most Etonians, he is a Public Psychological Naturalist.’5 These are clearly two possible forms of psychological naturalism. But it seems to me clear that they do not exhaust the possible varieties of such a view. For, suppose we confine ourselves to views which define rightness by some feeling or emotion roused by the contemplation of a given act; a third possibility besides the two recognized by Professor Broad remains. I might hold that I mean by a right act not one which arouses approval in me when I contemplate it, or one which arouses approval in all or most of some class of beings, but one which arouses in the agent, when he contemplates himself as doing it, a certain specific emotion, the kind of emotion which we certainly in fact feel when we think ourselves bound to do an act. This view will be neither private nor public in Professor Broad's sense; not private because the person to whose emotion reference is made is not the judger but the agent; not public, because no reference is made to the whole or the majority of any class of beings.
Let us consider first what may be called the private reaction view. This is the view that by calling an act right I mean that it awakes in me the emotion of approval. The theory has some plausibility, because the thought that an action is right and the feeling of approval always go together. We never judge an action right without experiencing the feeling of approval, nor vice versa. But the theory is open to serious objections. For (1) to begin with, ‘approval’ is much too wide a term. We approve of many things to which we do not ascribe the character of being obligatory or morally right—e.g. works of art. We shall have to equate rightness not with being an object of approval, but with being an object of a specific kind of approval, which we feel towards right actions but do not feel towards, for instance, works of art. Now I do not doubt that such a feeling exists, that our emotional reaction towards a right action is different from our emotional reaction towards anything else. But when I consider this emotion, it appears to me that it is not just a feeling which arises in us, we know not why, when we contemplate a right action. It seems to presuppose some insight into the nature of the action, as, for instance, that it is an action likely to redound to the general good, or a fulfilment of promise. It seems to be an intellectual emotion, presupposing the thought that the action is right, and right as being of a certain recognized character. And if this contention is correct, if the emotion of moral approval presupposes the thought that the action is right, it follows that we cannot mean by calling the action right that it awakes this emotion, since in order to have the emotion we must already be thinking of the action as right.
(2) Is it not clear that when we call an action right we mean essentially that it stands in a certain relation not to a spectator considered as capable of emotion in contemplating it, but to an agent considered as an agent? An action is never obligatory in the abstract; it is obligatory on a particular person in particular circumstances. To say that an act is obligatory is only another way of saying that a particular person ought to behave in a particular way. It is surely quite clear that what we have in mind when we call another person's action right is not any relation which it has to us as emotional beings, but a relation which it has to him as an active being.
(3) Another objection appears as soon as we consider the question of time. If I judge that Brutus did wrong in assassinating Caesar, I certainly do not think that his act first acquired its wrongness when I began to experience disapproval of it, or will cease to be wrong when I have ceased to do so.
And (4) this view does away with the possibility of difference of opinion on the rightness of acts. If all I mean when I say ‘action A is right’ is ‘I have a feeling of approval towards it’, and all you mean when you say ‘it is wrong’ is that you have a feeling of disapproval towards it, then we are not disagreeing; for what you say is perfectly compatible with what I say. I may well be approving and you disapproving of the same act. But it is surely perfectly clear that when I say an action is right and you say it is wrong we mean to be making incompatible statements about it. We might try to get over the objection by supposing that I am judging it to be right in certain respects, and that you are judging it to be wrong in certain other respects; and if this is all we are doing, our statements are of course compatible. But if I am judging it to be right on the whole and you are judging it to be wrong on the whole, we are certainly making statements each of which means to contradict the other. And this they could not do if each only stated the personal reaction of an individual to the act. We should then no more be contradicting one another than we do if you say you like jazz music and I say I don't.
In view of these fatal objections to the private view, shall we take refuge in a public reaction view? Shall we say that when we pronounce an action to be right, what we mean is that all or most men, or all or most members of some class of men, react to the act with a feeling of approval?
The first three of the objections to the private reaction view apply equally to the public reaction view. The feeling of approval presupposes the judgement of rightness; rightness evidently stands for the relation of an act not to any man or body of men as emotional reagents, but for the relation of it to a possible doer; and rightness is not held to belong to an act only when some man or body of men is having any sort of reaction to it. The fourth objection will not apply. If I say ‘all or most so-and-so's have a feeling of approval towards this act’, and if you say ‘all or most so-and-so's have a feeling of disapproval towards it’, we certainly are contradicting one another, and a real place is left for difference of opinion on the rightness of acts. But then a new objection (4a) makes its appearance, from which the private reaction view was free. I never judge an action right except when I have the feeling of approval towards it, but I obviously may judge it right when I am not thinking that the whole or a majority of any set of people have such a feeling. Indeed, I may judge an act right when I think no one but myself approves of it, or wrong when I think no one but myself disapproves of it. That is, in fact, what happens with every moral reformer when he enunciates a new moral principle or denies an old one. Are we to suppose that when Wilberforce began to denounce the slave-trade as wrong, what he meant was that a majority of Englishmen or even a majority of the Clapham sect had a feeling of disapproval towards it? The first alternative he could not have thought true, and the second was equally certainly not what he meant to assert. What he meant to assert was that the slave-trade was wrong, however any body of men reacted to it.
But we still have on our hands the third reaction view. According to this, to say that an action is obligatory on so-and-so means ‘so-and-so has a certain feeling in face of it’. Here we seem in one respect to be nearer to the truth; for, an obligation being an obligation resting on an individual agent, the state of his mind is much more likely to be relevant to a particular act's being obligatory on him than the state of mind of any observer or body of observers. We may later6 see grounds for believing, not indeed that rightness consists in the arousal of any state in his mind, but that the state of his opinion about the facts of the case, rather than the actual facts of the case, is what makes a particular act right for him. But this is very different from saying that the presence of the emotion of obligation in him is the rightness of the act. And a little consideration will show that this cannot be so. For as of moral approval, so of the emotion of obligation, we must say that it is not a blind feeling that arises in us, we know not why, on contemplating a possible act. It is an intellectual emotion which arises only when we judge the act to have a certain character, say that of producing a maximum of good, and to be on that account obligatory. And if the emotion of obligation presupposes the judgement that we are obliged, our being obliged cannot consist in our having the emotion of obligation. It is surely quite plain that the thought that I am obliged to do a certain act is not the thought that the contemplation of a certain possible act affects me in a certain way as an emotional being, but the thought that the act itself is related to me in a certain way as an active being.
We have now fairly considered all the reaction views, and have found that none of them gives even a plausible account of the essence of obligation. Are the causal views in any better position? One of them is the view that for an act to be right means just that it is, of all the acts possible for an agent at a certain time, that which will procure for him most pleasure. This is one form of egoistic Hedonism. It is not the only possible form; for an egoistic hedonist might quite well hold right to be an indefinable notion, or one definable in some other way, but hold that what makes acts to have the characteristic of rightness is their having the other characteristic of tending to produce most pleasure for the agent. With that variety of Hedonism we are not at present concerned. One consideration will sufficiently refute the view we are considering. No one will have the slightest difficulty in remembering instances in which he has thought of some act as his duty, without in the least thinking of it as likely to bring him more pleasure than any other would. And if the two thoughts are not even necessarily found together, still less can it be pretended that they are but one thought.
The suggestion that right means ‘such as to produce most pleasure for all human beings’ is in one respect not so remote from the truth as the view last considered. The thought of an action as contributing to the pleasure of others is far more closely associated with the thought of its rightness, than is the thought of it as contributing to the agent's own pleasure. But it must surely be admitted that there are other things than pleasure which we think it our duty to promote for other people—the improvement of their character and of their intelligence; and even if this were not so, even if pleasure were the only thing we deemed it our duty to produce, it is clear that the proposition, ‘right action is that which produces most pleasure for humanity’, is not an analytic proposition in which we unfold what we mean by ‘right’, but a synthetic proposition in which we express the view that the characteristic in actions which entails their having the characteristic of rightness is their tendency to produce a maximum of pleasure.
I have dealt rather summarily with these theories of the meaning of right, partly because the subject has been treated very fully and well by Professor Moore, partly because I suspect that the issue is not a very live one. It requires only a very little attention to what is in our minds when we use the word ‘right’, to see that none of these suggested meanings is really in our minds. To each of the theories I have discussed there corresponds a pair of views, each of which is more plausible than the views I have considered. Take any of the characteristics that have been put forward as giving the essence of rightness. It would, in the first place, be more plausible to say it must be granted that the mention of this characteristic does not state the meaning of right. Right is perhaps to be defined otherwise, or perhaps it is indefinable. But this is the characteristic that makes right acts right.’ It would be, for instance, far more plausible to say that contribution to the general happiness is what makes an act right, than to say that it is its being right. And in fact the clearest-headed of the hedonists have defined their view thus. Sidgwick, for example, regards rightness as an indefinable notion, and there are indications that Mill did the same.7 And any one can see that this is much more plausible than the alternative. Thus, answering to each of the psychological theories as to the essence of rightness, there will be a view as to the ground of rightness. And these views we shall have to consider when we come to that part of our inquiry.8
There is, however, another possibility. It may be said, ‘Let us grant that none of these psychological characteristics is what we mean by rightness. These characteristics, however, are those which the actions we call right really have. Rightness is only a characteristic which we fancy some actions to have; and so is wrongness.’ I have already, in dealing with evolutionary views, indicated the main reasons which have led some people to adopt this type of view, and shown, I hope, their insufficiency.9 But to point out the insufficiency of the reasons which have led people to adopt a view is not to prove the view untrue. One can urge that one has oneself a clear conviction that certain acts are right and others wrong; and if we believe in the fundamental identity of human nature in all men, it is very hard to suppose that any sane person is totally without such a conviction, which it is clear that nearly all men have. But if any one denies that he has it, I cannot prove to him that he is wrong in his denial. As a rule, however, he will betray himself in one way. He is quite likely, when blamed for some act, to say that he is convinced that there is no such thing as right or wrong. But when his interests are attacked or threatened by some one else, his reaction usually convinces one that he thinks there are things that other people ought not to do; and if so he can hardly refuse to admit that there are things that he himself ought not to do. The denial of any distinction between right and wrong can usually be seen to be a disingenuous excuse for doing as one pleases. But if any one is prepared to exempt others as well as himself from moral obligation, I do not think he can be argued out of his view.
In fact, however, such an attitude is extremely uncommon, and much that passes for it is not really of this nature. It is not unusual to deplore the present age as one in which the sense of moral distinctions is weakening, possibly to its final disappearance. I do not for a moment believe this to be true. Whether the practice of morality is weakening is a different question, with which I am not at present concerned; though there, too, one may be allowed to express the opinion that we are at least as much better in some respects than our ancestors, as in others we are worse. What I am concerned with is the question whether the recognition that there is a right and a wrong is weakening. Take, for instance, one particular question about which the complaint is often made—the relations between the sexes. Undoubtedly there has been a growth of opinion in favour of a relaxation of the code which has hitherto governed Christian countries. But the advocates of relaxation are just as much moralists as their opponents. Both alike think that there is some right way of arranging the relations between the sexes. And even if some go so far as to say that all rules for individual behaviour in this matter ought to be abolished, they say they ought to be abolished, i.e. that legislators ought to abolish certain laws and that public opinion ought not to visit certain acts with its displeasure. No one says ‘it does not matter what we do about the question; there is no right or wrong about it at all’. In fact the difference that divides us is not a difference on the question whether there is a right and a wrong, but a difference on the questions, ‘What are the characteristics of acts which make them right or wrong?’ and ‘How far do certain types of act in fact possess these characteristics?’ The first is a question for ethics, and is probably its main problem. The second is a question for applied ethics or casuistry.
It is on our finding of the true answers to these two types of question that our attainment of a true solution of the problem of the sexes depends. Much will depend, obviously, on the nature of our answer to the first question. If, for instance, the only characteristic that makes acts right is their tendency to produce pleasure, that might point to one solution of the sex question; if a tendency to promote certain qualities of personality is also a ground of rightness, and still more if it is a more important ground, that might point to quite a different solution. But again, even if it were to be agreed that Hedonism is wrong and that there are other goods than pleasure, we should still have to ask what kind of relation between the sexes best tends to promote those qualities of personality. It is by showing that a strict code of sexual behaviour in fact secures for actions the characteristics which in fact make actions right, that a strict code is to be defended; and I do not think that it need fear the results of such an inquiry.
Before leaving the naturalistic attempts to define ethical terms, I ought to take some account of a way of thinking which has affinities with these attepmts, which has come rather rapidly to the front in the last few years. It has not, I think, made as yet any wide popular appeal, but it has attracted many of our younger philosophers, and bids fair to attract more. This is the way of thinking represented by the Vienna school of positivists, who found their inspiration in the teaching of Mr. Wittgenstein but have developed and modified his views. I take as representatives of the school Mr. Carnap and Mr. Ayer—the Philosophy and Logical Syntax of the former, and the Language, Truth, and Logic of the latter.
Their original inspiration goes a good deal farther back than Wittgenstein; in fact they may be regarded as having reverted to the views of Hume. Mr. Carnap10 quotes with approval Hume's famous words:11
‘It seems to me, that the only objects of the abstract sciences or of demonstration are quantity and number.… All other enquiries of men regard only matter of fact and existence; and these are evidently incapable of demonstration.… When we run over libraries, persuaded of these principles, what havoc must we make? If we take in our hand any volume; of divinity or school metaphysics, for instance; let us ask, Does it contain any abstract reasoning concerning quantity or number? No. Does it contain any experimental reasoning concerning matter of fact and existence? No. Commit it then to the flames: for it can contain nothing but sophistry and illusion.’
‘We agree’, says Mr. Carnap,12 ‘with this view of Hume, which says—translated into our terminology—that only the propositions of mathematics and empirical science have sense, and that all other propositions are without sense.’
Stated more definitely, the view is that all significant propositions are either a priori, in which case they are purely tautologous, or else are empirical hypotheses reached by reasoning from observation and having a meaning only because they can be tested by further observation. It has long been matter of agreement among philosophers that at least the vast majority of the propositions of the natural sciences are empirical hypotheses, not known to be true but rendered probable by experience and capable of being rendered more, or less, probable by further experience. Furthermore, the positivists, if I may take Mr. Ayer as typical, adopt the more sensible of two possible views with regard to the testability by further experience which is required in order to make an empirical proposition significant. He sees that no empirical general proposition can ever be completely tested by further experience, in the sense of being completely proved, and adopts therefore the more temperate view that the sort of testing of which an empirical proposition must be capable, if it is to have sense, is a testing which will render the proposition more or less probable.
Mr. Ayer sees that ethical judgements in which we pronounce something to be good or bad, or right or wrong, offer at first sight a difficulty to the positivist's theory that ‘all synthetic propositions are empirical hypotheses’. Ethical judgements seem to be synthetic, but ‘they cannot with any show of justice be represented as hypotheses, which are used to predict the course of our sensations’.13 He rejects two well-known attempts that have been made to exhibit ethical judgements as empirical hypotheses—viz. the subjectivist theory which defines the rightness of actions, and the goodness of ends, in terms of the feeling of approval which a certain person, or group of people, has towards them, and the utilitarian theory, which defines rightness and goodness in terms of pleasure. These he rejects on grounds similar to some of those on which I have already rejected them;14 and it is unnecessary to repeat these. He is careful to point out that he is not ‘denying that it is possible to invent a language in which all ethical symbols are definable in non-ethical terms, or even that it is desirable to invent such a language and adopt it in place of our own; what we are denying is that the suggested reduction of ethical to non-ethical statements is consistent with the conventions of our actual language’.15 Clearly, then, his own theory, which he offers in place of Subjectivism and Utilitarianism, is meant to be consistent with the conventions of our actual language, i.e. to be an account not of what we ought to be saying, but of what we mean when we actually do say that so-and-so is right, or is good.
If Subjectivism and Utilitarianism are rejected, as they are by the positivists, it might seem that the conclusion to be drawn is that ‘right’ and ‘good’, and their opposites, are terms which cannot be defined naturalistically, and that judgements in which we use them as predicates are a priori judgements, judgements in which we express not the results of observation but a direct insight. But the positivists cannot accept this view, since they have committed themselves to the view that all a priori judgements are pure tautologies and that only empirical hypotheses have factual content, and since, as they admit, it is clear that when we say that something is right or good, we are not uttering a tautology. Holding, then, that all judgements that have meaning are either empirical hypotheses or tautologies, and that ethical judgements do not belong to either of these types, what are they to say about them? The positivists cut the knot by saying that ethical judgements, or rather, those most important ethical judgements whose predicate is ‘right’ or ‘good’, are not judgements at all, that in them nothing whatever is asserted. There is a minor difference between the positivists as to what such ‘pseudo-judgements’ are. Mr. Carnap16 says roundly that they are all commands—that to say ‘so-and-so is right, or good’ is to say ‘do so-and-so’; Mr. Ayer17 distinguishes ‘actual ethical judgements’ from ‘exhortations to moral virtue’, i.e. from commands.
The theory that all judgements with the predicate ‘right’ or ‘good’ are commands has evidently very little plausibility. The only moral judgements of which it could with any plausibility be maintained that they are commands are those in which one person says to another ‘you ought to do so-and-so’. A command is an attempt to induce some one to behave as one wishes him to behave, either by the mere use of authoritative or vehement language, or by this coupled with the intimation that disobedience will be punished. And there is no doubt that such words as ‘you ought to do so-and-so’ may be used as one means of so inducing a person to behave in a certain way. But if we are to do justice to the meaning of ‘right’ or ‘ought’, we must take account also of such modes of speech as ‘he ought to do so-and-so’, ‘you ought to have done so-and-so’, ‘if this and that had been the case, you ought to have done so-and-so’, ‘if this and that were the case, you ought to do so-and-so’, ‘I ought to do so-and-so’. Where the judgement of obligation has reference either to a third person, not the person addressed, or to the past, or to an unfulfilled past condition, or to a future treated as merely possible, or to the speaker himself, there is no plausibility in describing the judgement as a command. But it is easy to see that ‘ought’ means the same in all these cases, and that if in some of them it does not express a command;, it does not do so in any. And if the form of words ‘you ought to do so-and-so’ may be used as a way of inducing the person addressed to behave in a particular way, that does not in the least imply that the apparent statement is really not a statement, but a command. What distinguishes its meaning from that of the genuine ‘do so-and-so’ is that one is suggesting to the person addressed a reason for doing so-and-so, viz. that it is right. The attempt to induce the person addressed to behave in a particular way is a separable accompaniment of the thought that the act is right, and cannot for a moment be accepted as the meaning of the words ‘you ought to do so-and-so’.
While Mr. Ayer avoids the crude view that all ethical judgements are really commands, he agrees with Mr. Carnap that, whereas all judgements proper have two characteristics—that of expressing a state of mind and that of asserting something,18 ethical judgements assert nothing, and are mere expressions of a state of mind in which we are liking certain kinds of conduct and wishing others to behave accordingly. Mr. Ayer's choice of an example is rather unfortunate. ‘If I say to some one’, he remarks,19 ‘“You acted wrongly in stealing that money”, I am not stating anything more than if I had simply said, “You stole that money”.’ There is some plausibility in this, simply because the word ‘steal’ already connotes wrongful action, and therefore the addition of the word ‘wrongly’ may at least plausibly be said to add nothing. But let us avoid the use of a question-begging term. Let us take the example, ‘In saying that which you did not believe you acted wrongly’. It cannot, I am sure, with any plausibility be maintained that, in saying that, I am asserting no more than that you have said that which you did not believe. I am quite definitely meaning to characterize your action further in a certain way. The judgement, we are told, merely expresses my personal dislike and disapproval of the action; but when this is said it is forgotten that whatever be true of dislike, it is impossible to disapprove without thinking that what you disapprove is worthy of disapproval.
This denial that when we use such terms as ‘right’ or ‘good’ we mean (as opposed to expressing) anything at all is not, I think, the product of disinterested reflection on such judgements. It is the product of a preconceived theory about judgements in general, viz. of the theory that judgements which are both synthetic and a priori, i.e. are neither tautologous nor empirical, are impossible. For that thesis no genuine proof is ever offered by its supporters; we are simply told repeatedly that it is manifestly true. In considering whether it is true or not, it may be useful to consider the straits to which its supporters are reduced in dealing with certain classes of judgements. I will take first judgements in which we appear to make universal synthetic a priori statements. The instance Mr. Ayer takes20 is the judgement ‘a material thing cannot be in two places at once’. This, says Mr. Ayer, is not a statement expressing any knowledge about the nature of things, but a statement about the use of language. ‘It is necessary only because we happen to use the relevant words in a particular way. There is no logical reason why we should not so alter our definitions that the sentence “A thing cannot be in two places at once” comes to express a self-contradiction instead of a necessary truth.’ There is, of course, a truth at the bottom of this contention. There is no necessary connexion between any of our words and the meanings in which we use them. The meaning which we express by ‘can’ might have been expressed by ‘cannot’ (if ‘cannot’ had not already come to have the meaning which it in fact has); and if in saying ‘a thing cannot be in two places at once’ we had meant what we do mean by saying ‘a thing can be in two places at once’, our statement would have been as obviously untrue as, with the existing usage of ‘cannot’, it is obviously true. Before we can discuss the meaning of any proposition, we must be satisfied that we are using our words in the same sense; but when we have satisfied ourselves of that, the question remains whether the things we are talking about have the connexions which they are alleged to have; and the fact that with different conventional meanings of words the statement ‘a material thing cannot be in two places at once’ might have been untrue throws no light on the question whether with the existing meanings of words it is not both true and necessary and synthetic.
I will take as a second example the positivistic view of statements about the past. Mr. Ayer thinks, in accordance with his general view that all non-tautologous propositions are empirical hypotheses stating what experiences may be expected in the future, that statements which are expressed as statements about the past are really ‘rules for the prediction of those “historical” experiences which are commonly said to verify them’;21 i.e. the statement that the battle of Hastings was fought in 1066 is really the statement that any one who makes the necessary investigation will get certain experiences which will satisfy him that the battle was fought in that year. As against this it seems to me sufficient to say that a statement about the past is a statement about the past and not about the future-though no doubt a statement about the past may involve consequences about the future, and a statement about the future may involve consequences about the past. Mr. Ayer claims that those he is criticizing are assuming that the past is objectively there to be corresponded to, an assumption which he regards as objectionable. But it is surely clear that, if his opponents’ view involved that, his own theory would equally involve that the future is objectively there to be corresponded to, and that is open to at least as much objection. There are reasons for thinking that only the present and neither the past nor the future are real, and reasons for thinking that only the present and the past are real,22 but it is difficult to think of any ground on which it could be maintained that the present and the future are real, but not the past.
The positivistic theory simply falls into a confusion, amounting to an explicit identification, between what a statement means to assert and the evidence which would lead one to believe in its truth. And this confusion is very far-reaching in its effect on the views of the positivists. Their objection to recognizing ethical judgements as genuine assertions arises from the fact, long known to moral philosophers, that ethical judgements cannot be verified by any sensible experience, coupled with the view of the positivists that the only synthetic judgements that have meaning, i.e. that are genuine judgements, are those that are verifiable. Sometimes, indeed, they go so far as to say that the meaning of a synthetic judgement is its verification. Now the verification, or partial verification, of a general statement may be achieved in either of two ways. If the proposition states that every A has the attribute B, we may effect a partial verification by producing, one after another, instances in which, by the use of the senses, particular A's are perceived to have the attribute B. And in this case the facts which form the evidence for the proposition are the very facts which (and others like them) are summed up in the proposition itself. In this case the meaning of the proposition may loosely be identified with the facts which verify it. But even then there must from the nature of things be some difference between that which is verified and that by which it is verified; and there is in fact a difference between the general statement that all A's are B, and the sum of the particular statements ‘this A is B’, ‘this second A is B’, and so on.
It may, however, well be the case that we cannot by the use of the senses perceive directly that any A is B. In such a case we may be able to verify the statement by discovering by sensuous experience A's which have the attribute C, which we already know to imply B. In such a case the facts which verify the proposition are entirely distinct from the meaning of the proposition. Instances in which statements about the past are verified can obviously be only of the latter order. And the existence of this type of case shows that statements can have a meaning completely different from the facts which verify them.
Sometimes, however, the positivists adopt a view less crude than that which identifies the meaning of a statement with its verification, and content themselves with saying that no statement can have meaning unless it is verifiable, or at least partly verifiable. As against this form of the theory it seems to me enough to refer to Dr. Ewing's convincing refutation of it.23 He shows inter alia that the positivists could not ‘establish the truth of their view even in a single case merely by sense-experience. For how can we ever know by sense-experience that there is not a part of the meaning of a statement that we cannot verify? The fact that we do not have any sense-experience of such a part proves nothing, since the point at issue is whether there is something in what we mean beyond sense-experience; and how can we know by sense-experience that there is not?’24
If it cannot be verified by sense-experience that even the meaning of a single statement is entirely exhausted by what can be verified by sense-experience, still less, of course, can the general theory that all statements are meaningless unless they are thus verifiable be itself verified. Thus if the theory is true, the sentence which states the theory must be meaningless, since it is an unverifiable statement. ‘But a sentence cannot possibly be true and meaningless. Therefore the sentence in question cannot’ (i.e. cannot, on the basis of the theory itself) ‘be true, but must be either meaningless or false.’25
I conclude, then, that the latest attempt to discredit ethics is not successful. Indeed, there is one of the arguments put forward by the positivists which seems to me to provide, when reflected on, an argument in favour not only of the view that our ethical judgements are genuine judgements, but of the view that there are fundamental ethical judgements for which general agreement may be claimed. Mr. Ayer remarks26 that, while his theory escapes many of the objections brought against subjectivistic theories in ethics, there is one which it does not escape. This is the argument27 that such theories would make it impossible to argue about questions of value, which nevertheless we undoubtedly do. He admits that his own theory also would make it impossible to argue about questions of value; as he holds that such sentences as ‘thrift is a virtue’ and ‘thrift is a vice’ do not express propositions at all, he clearly cannot hold that they express incompatible propositions. If, then, he is to resist the argument in question, he must simply deny that in fact we ever do dispute about questions of value; for if we did dispute about things which on his theory we cannot dispute about, his theory would clearly be untrue. He boldly adopts the course to which he is logically forced, and denies that we ever do dispute about questions of value. And he justifies this by saying that apparent disputes about questions of value are really disputes about questions of fact.
‘When some one disagrees with us about the moral value of a certain action or type of action, we do admittedly resort to argument in order to win him over to our way of thinking. But we do not attempt to show by our arguments that he has the “wrong” ethical feeling towards a situation whose nature he has correctly apprehended. What we attempt to show is that he is mistaken about the facts of the case. We argue that he has misconceived the agent's motive: or that he has misjudged the effects of the action, or its probable effects in view of the agent's knowledge; or that he has failed to take into account the special circumstances in which the agent was placed.… We do this in the hope that we have only to get our opponent to agree with us about the nature of the empirical facts for him to adopt the same moral attitude towards them as we do. And as the people with whom we argue have generally received the same moral education as ourselves, and live in the same social order, our expectation is usually justified. But if our opponent happens to have undergone a different process of moral “conditioning” from ourselves, so that, even when he acknowledges all the facts, he still disagrees with us about the moral value of the actions under discussion, then we abandon the attempt to convince him by argument. We say that it is impossible to argue with him because he has a distorted or undeveloped moral sense; which signifies merely that he employs a different set of values from our own.… It is because argument fails us when we come to deal with pure questions of value, as distinct from questions of fact, that we finally resort to mere abuse.’28
It is perfectly true that, when we differ on a question of right or wrong, or of goodness or badness, it is by consideration of questions of fact—of the precise nature of the consequences or of the probable consequences, or of the motives involved—that we try to remove the difference of opinion on the moral question. And in doing so we betray the conviction that if we could get down to agreement about the facts of the case, we should find ourselves in agreement on the moral question; or in other words, that though we may differ in our moral judgements on some complicated case, we agree in our fundamental judgements as to what kinds of consequences ought to be aimed at and what kinds of motive are good. The more Mr. Ayer emphasizes this element in our discussion of moral questions, the more he pays tribute to the strength of this conviction; for unless we thought that if we could agree on the factual nature of the act we should probably agree on its rightness or wrongness, there would be no point in trying to reach agreement about its factual nature. And in the great majority of cases we find this confidence confirmed, by finding that we agree in our moral judgements when we agree about the facts. But no doubt we sometimes fail to find agreement even then. We do not find, however, as Mr. Ayer claims, that no subject of dispute remains. We find, indeed, that there is no room for further argument; when we have come to some premiss which to us seems axiomatic, and which the other person denies, we can argue no further. But we do not find that all difference of opinion has vanished, and that we are left only with different feelings, one liking certain consequences or motives and another disliking them. We find ourselves still saying ‘this is good’, and the person with whom we are speaking still saying ‘this is bad’. And it is not by showing that argument ceases, but by showing that difference of opinion ceases, that Mr. Ayer could escape from Professor Moore's argument.
But indeed our adoption of the very practice which Mr. Ayer here describes is enough to refute his account of the nature of what are commonly called ethical judgements. He denies that they are judgements; he says they are mere expressions of liking or dislike. If that were all they are, why argue at all? What should we be trying to prove? Is A arguing to prove that he likes the given act, and B to prove that he dislikes it? Clearly not. A does not doubt that B dislikes it, nor B that A likes it; and if they did doubt, they would adopt quite different means of convincing one another, e.g. A by consistently seeking to do similar acts and B by consistently avoiding them. What they are attempting to do by the process Mr. Ayer describes is to convince each other that the liking, or the dislike, is justified, in other words that the act has a character that deserves to be liked or disliked, is good or is bad.
The Right and the Good, 14.
Mind, xxxv (1926), 289.
Five Types of Ethical Theory, 259.
Cf. pp. 9–10.
p. 35; cf. Ayer, p. 56.
Enquiry concerning Human Understanding, ad fin.
Supra, ch. 2.
Cf. pp. 254–5, infra.
Ayer, op cit., 158.
Cf. Broad, Scientific Thought, 66.
Mind, xlvi (1937), 347–64.
Professor Moore's argument, in Philosophical Studies, 333–4.