The ideas of right and wrong conduct are, as we have seen, those with which ethics is generally supposed to be most concerned. This view, which is unduly narrow, is fostered by the use of the one word good, both for the sort of conduct which is right, and for the sort of things which ought to exist on account of their intrinsic value. This double use of the word good is very confusing, and tends greatly to obscure the distinction of ends and means. I shall therefore speak of right actions, not of good actions, confining the word good to the sense explained in Section II. (§ 12 ¶ 1)

The word right is very ambiguous, and it is by no means easy to distinguish the various meanings which it has in common parlance. Owing to the variety of these meanings, adherence to any one necessarily involves us in apparent paradoxes when we use it in a context which suggests one of the other meanings. This is the usual result of precision of language; but so long as the paradoxes are merely verbal, they do not give rise to more than verbal objections. (§ 12 ¶ 2)

In judging of conduct we find at the outset two widely divergent methods, of which one is advocated by some moralists, the other by others, while both are practised by those who have no ethical theory. One of these methods, which is that advocated by the utilitarians, judges the rightness of an act by relation to the goodness or badness of its consequences. The other method, advocated by intuitionists, judges by the approval or disapproval of the moral sense or conscience. I believe that it is necessary to combine both theories in order to get a complete account of right and wrong. There is, I think, one sense in which a man does right when he does what will probably have the best consequences, and another in which he does right when he follows the dictates of his conscience, whatever the probable consequences may be. (There are many other senses which we may give to the word right, but these two seem to be the most important.) Let us begin by considering the second of these senses. (§ 12 ¶ 3)

The question we have to ask ourselves is: What do we mean by the dictates of the moral sense? If these are to afford a definition of right conduct, we cannot say that they consist in judging that such andsuch acts are right, for that would make our definition circular. We shall have to say that the moral sense consists in a certain specific emotion of approval towards an act, and that an act is to be called right when the agent, at the moment of action, feels this emotion of approval towards the action which he decides to perform. There is certainly a sense in which a man ought to perform any act which he approves, and to abstain from any act which he disapproves; and it seems also undeniable that there are emotions which may be called approval and disapproval. Thus this theory, whether adequate or not, must be allowed to contain a part of the truth. (§ 13 ¶ 1)

It is, however, fairly evident that there are other meanings of right conduct, and that, though there is an emotion of approval, there is also a judgment of approval, which may or may not be true. For we certainly hold that a man who has done an action which his conscience approved may have been mistaken, and that in some sense his conscience ought not to have approved his action. But this would be impossible if nothing were involved except an emotion. To be mistaken implies a judgment; and thus we must admit that there is such a thing as a judgment of approval. If this were not the case we could not reason with a man as to what is right; what he approves would be necessarily right for him to do, and there could be no argument against his approval. We do in fact hold that when one man approves of a certain act, while another disapproves, one of them is mistaken, which would not be the case with a mere emotion. If one man likes oysters and another dislikes them, we do not say that either of them is mistaken. (§ 13 ¶ 2)

Thus there is a judgment of approval, and this must consist of a judgment that an act is, in a new sense, right. The judgment of approval is not merely the judgment that we feel the emotion of approval, for then another who disapproved would not necessarily hold our judgment of approval to be mistaken. Thus in order to give a meaning to the judgment of approval, it is necessary to admit a sense of right other than approved. In this sense, when we approve an act we judge that it is right, and we may be mistaken in so judging. This new sense is objective, in the sense that it does not depend upon the opinions and feelings of the agent. Thus a man who obeys the dictates of his conscience is not always acting rightly in the objective sense. When a man does what his conscience approves, he does what he believes to be objectively right, but not necessarily what is objectively right. We need, therefore, some other criterion than the moral sense for judging what is objectively right. (§ 13 ¶ 3)

It is in defining objective rightness that the consequences of an action become relevant. Some moralists, it is true, deny the dependence upon consequences; but that is to be attributed, I think, to confusion with the subjective sense. When people argue as to whether such and such an action is right, they always adduce the consequences which it has or may be expected to have. A statesman who has to decide what is the right policy, or a teacher who has to decide what is the right education, will be expected to consider what policy or what education is likely to have the best results. Whenever a question is at all complicated, and cannot be settled by following some simple rule, such as thou shalt not steal, or thou shalt not bear false witness, it is at once evident that the decision cannot be made except by consideration of consequences. (§ 14 ¶ 1)

But even when the decision can be made by a simple precept, such as not to lie or not to steal, the justification of the precept is found only by consideration of consequences. A code such as the Decalogue, it must be admitted, can hardly be true without exception if the goodness or badness of consequences is what determines the rightness or wrongness of actions; for in so complex a world it is unlikely that obedience to the Decalogue will always produce better consequences than disobedience. Yet it is a suspicious circumstance that breaches of those of the Ten Commandments which people still hold it a duty to obey do, as a matter of fact, have bad consequences in the vast majority of instances, and would not be considered wrong in a case in which it was fairly certain that their consequences would be good. This latter fact is concealed by a question-begging addition of moral overtones to words. Thus, e.g., thou shalt do no murder would be an important precept if it were interpreted, as Tolstoy interprets it, to mean thou shalt not take human life. But it is not so interpreted; on the contrary, some taking of human life is called justifiable homicide. Thus murder comes to mean unjustifiable homicide; and it is a mere tautology to say Thou shalt do no unjustifiable homicide. That this should be announed from Sinai would be as fruitless as Hamlet’s report of the ghost’s message: There’s ne’er a villain, dwelling in all Denmark, but he’s an arrant knave. As a matter of fact, people do make a certain classifiation of homicides, and decide that certain kinds are justifiable and certain others unjustifiable. But there are many doubtful cases: tyrannicide, capital punishment, killing in war, killing in self-defence, killing in defence of others, are some of these. And if a decision is sought, it is sought usually by considering whether the consequences of actions belonging to these classes are on the whole good or bad. Thus the importane of precepts such as the Ten Commandments lies in the fact that they give simple rules, obedience to which will in almost all cases have better consequences than disobedience; and the justification of the rules is not wholly independent of consequences. (§ 14 ¶ 2)

In common language the received code of moral rules is usually presupposed, and an action is only called immoral when it infringes one of these rules. Whatever does not infringe them is regarded as permissible, so that on most of the occasions of life no one course of action is marked out as alone right. If a man adopts a course of action which, though not contrary to the received code, will probably have bad consequences, he is called unwise rather than immoral. Now, according to the distinction we have made between objective and subjective rightness, a man may well act in a way which is objectively wrong without doing what is subjectively wrong, i.e. what his conscience disapproves. An act (roughly speaking, I shall return to this point presently) is immoral when a man’s conscience disapproves it, but is judged only unwise or injudicious when his conscience disapproves it, but is judged only unwise or injudicious when his conscience approves it, although we judge that it will probably have bad consequences. Now the usual moral code is supposed, in common language, to be admitted by every man’s conscience, so that when he infringes it, his action is not merely injudicious, but immoral; on the other hand, where the code is silent, we regard an unfortunate action as objectively but not subjectively wrong, i.e. as injudicious, but not immoral. The acceptance of a moral code has the great advantage that, in so far as its rules are objetively right, it tends to harmonize objective and subjective rightness. Thus it tends to cover all frequent cases, leaving only the rarer ones to the individual judgment of the agent. Hence when new sorts of cases become common, the moral code soon comes to deal with them; thus each profession has its own code concerning cases common in the profession, though not outside it. But the moral code is never itself ultimate; it is based upon an estimate of probable consequences, and is essentially a method of leading men’s judgment to approve what is objectively right and disapprove what is objectively wrong. And when once a fairly correct code is accepted, the exceptions to it become very much fewer than they would otherwise be, because one of the consequences of admitting exceptions is to weaken the code, and this consequence is usually bad enough to outweigh the good resulting from admitting such and such an exception. This argument, however, works in the opposite direction with a grossly incorrect code; and it is to be observed that most conventional codes embody some degree of unwarrantable selfishness, individual, professional, or national, and are thus in certain respects worthy of detestation. (§ 15 ¶ 1)

What is objectively right, then, is in some way dependent on consequences. The most natural supposition to start from would be that the objectively right act, under any circumstances, is the one which will have the best consequences. We will define this as the most fortunate act. The most fortunate act, then, is the one which will produce the greatest excess of good over evil, or the least excess of evil over good (for there may be situations in which every possible act will have consequences that are on the whole bad). But we cannot maintain that the most fortunate act is always the one which is objectively right, in the sense that it is what a wise man will hold that he ought to do. For it may happen that the act which will in fact prove the most fortunate is likely, according to all the evidence at our disposal, to be less fortunate than some other. In such a case, it will be, at least in one sense, objectively wrong to go against the evidence, in spite of the actual good result of our doing so. There have certainly been some men who have done so much harm that it would have been fortunate for the world if their nurses had killed them in infancy. But if their nurses had done so their action would not have been objectively right, beause the probability was that it would not have the best effects. Hene it would seem we must take account of probability in judging of objective rightness; let us then consider whether we can say that the objectively right act is the one which will probably be most fortunate. I shall define this as the wisest act. (§ 16 ¶ 1)

The wisest act, then, is that one which, when account is taken of all available data, gives us the greatest expectation of good on the balance, or the least expectation of evil on the balance. There is, of course, a difficulty as to what are to be considered available data; but broadly we can distinguish, in any given state of knowledge, things capable of being forseen from things which are unpredictable. I suppose account to be taken of the general body of current knowledge, in fact the sort of consideration which people expect when they ask for legal or medical advice. There is no doubt this brings us nearer to what is objectively right than we were when we were considering the actually most fortunate act. For one thing, it justifies the unavoidable limitation to not very distant consequences, which is almost always necessary if a practical decision is to be reached. For the likelihood of error in calculating distant consequences is so great that their contribution to the probable good or evil is very small, though their contribution to the actual good or evil is likely to be much greater than that of the nearer consequences. And it seems evident that what it is quite impossible to know cannot be relevant in judging as to what conduct is right. If, as is possible, a cataclysm is going to destroy life on this planet this day week, many acts otherwise useful will prove to have been wasted labour, for example, the preparation of next year's Nautical Almanac; but since we have no reason to expect such a cataclysm, the rightness or wrongness of acts is plainly to be estimated without regard to it. (§ 16 ¶ 2)

One apparent objection at once suggests itself to the definition. Very few acts are of sufficient importance to justify such elaborate and careful consideration as is required for forming an opinion as to whether they are the wisest. Indeed, the least important decisions are often those which it would be hardest to make on purely reasonable grounds. A man who debates on each day which of two ways of taking exercise is likely to prove most beneficial is considered absurd; the question is at once difficult and unimportant, and is therefore not worth spending time over. But although it is true that unimportant decisions ought not to be made with excessive care, there is danger of confusion if this is regarded as an objection to our definition of objective rightness. For the act which, in the case supposed, is objectively wrong is the act of deliberation. And the deliberation is condemned by our definition, for it is very unlikely that there is no more beneficial way of spending time than in debating trivial points of conduct. Thus, although the wisest act is the one which, after complete investigation, appears likely to give the most fortunate results, yet the complete investigation required to show that it is the wisest act is only itself wise in the case of very important decisions. This is only an elaborate way of saying that a wise man will not waste time on unimportant details. Hence this apparent objection can be answered. (§ 17 ¶ 1)

One further addition is required for the definition of the objectively right act, namely, that it must be possible. Among the acts whose consequences are to be considered we must not include such as are either physically impossible to perform or impossible for the agent to think of. This last condition introduces difficulties connected with determinism, which are discussed in Section IV. Ignoring these difficulties, we may say that the objectively right act is that one which, of all that are possible, will probably have the best consequences. (§ 18 ¶ 1)

We must now return to the consideration of subjective rightness, with a view to distinguishing conduct which is merely mistaken from conduct which is immoral or blameworthy. We here require a new sense of ought, which it is by no means easy to define. In the objective sense, a man ought to do what is objectively right. But in the subjective sense, which we have now to examine, he sometimes ought to do what is objectively wrong. For example, we saw that it is often objectively right to give less consideration to an unimportant question of conduct than would be required for forming a trustworthy judgment as to what is objectively right. Now it seems plain that if we have given to such a question the amount and kind of consideration which is objectively right, and we then do what appears to us objectively right, our action is, in some sense, subjectively right, although it may be objectively wrong. Our action could certainly not be called a sin, and might even be highly virtuous, in spite of its objective wrongness. It is these notions of what is sinful and what is virtuous that we have now to consider. (§ 19 ¶ 1)

The first suggestion that naturally occurs is that an act is subjectively right when it is judged by the agent to be objectively right, and subjectively wrong when it is judged to be objectively wrong. I do not mean that it is subjectively right when the agent judges that it is the act which, of all that are possible, will probably have the best results; for the agent may not accept our above account of objective rightness. I mean merely that it is the one towards which he has the judgment of approval. A man may judge an act to be right without judging that its consequences will be probably the best possible; I only contend that, when he truly judges it to be right, then its consequences will probably be the best possible. But his judgment as to what is objectively right may err, not only by a wrong estimate of probable consequences, or by failing to think of an act which he might have thought of, but also by a wrong theory as to what constitutes objective rightness. In other words, the definition I gave of objective rightness is not meant as an analysis of the meaning of the word, but as a mark which in fact attaches to all objectively right actions and to no others. (§ 20 ¶ 1)

We are to consider then the suggestion that an act is moral when the agent approves it and immoral when he disapproves it; using moral to mean subjectively right and immoral to mean subjectively wrong. This suggestion, it is plain, will not stand without much modification. In the first place, we often hold it immoral to approve some things and disapprove others, unless there are special circumstances to excuse such approval or disapproval. In the second place, unreflecting acts, in which there is no judgment either of approval or disapproval, are often moral or immoral. For both these reasons the suggested definition must be regarded as inadequate. (§ 20 ¶ 2)

The doctrine that an act is never immoral when the agent thinks it right has the drawback (or the advantage) that it excuses almost all the acts which would be commonly condemned. Very few people deliberately do what, at the moment, they believe to be wrong; usually they first argue themselves into a belief that what they wish to do is right. They decide that it is their duty to teach so-and-so a lesson, that their rights have been grossly infringed that if they take no revenge there will be an encouragement to injustice, that without a moderate indulgence in pleasure a character cannot develop in the best way, and so on and so on. Yet we do not cease to blame them on that account. Of course it may be said that a belief produced by a course of self-deception is not a genuine belief, and that the people who invent such excuses for themselves know all the while that the truth is the other way. Up to a point this is no doubt true, though I doubt if it is always true. There are, however, other cases of mistaken judgment as to what is right, where the judgment is certainly genuine, and yet we blame the agent. These are cases of thoughtlessness, where a man remembers consequences to himself, but forgets consequences to others. In such a case he may judge correctly and honestly on all the data that he remembers, yet if he were a better man he would remember more data. Most of the actions commonly condemned as selfish probably come under this head. Hence we must admit that an act may be immoral, even if the agent quite genuinely judges that it is right. (§ 21 ¶ 1)

Unreflecting acts, again, in which there is no judgment as to right or wrong, are often praised or blamed. Acts of generosity, for example, are more admired when they are impulsive than when they result from reflection. I cannot think of any act which is more blamed when it is impulsive than when it is deliberate; but certainly many impulsive acts are blamed—for example, such as spring from an impulse of malice or cruelty (§ 21 ¶ 2)

In all these cases where reflection is absent, and also in the case of inadequate reflection, it may be said that blame does not belong properly to the act, but rather to the character revealed by the act, or, if to some acts, then to those previous deliberate acts by which the character has been produced which has resulted in the present act. The cases of self-deception would then be dismissed on the ground that the self-deceiver never really believes what he wishes to believe. We could then retain our original definition, that a moral act is one which the agent judges to be right, while an immoral one is one which he judges to be wrong. But I do not think this would accord with what most people really mean. I rather think that a moral act should be defined as one which the agent would have judged to be right if he had considered the question candidly and with due care; if, that is to say, he had examined the data before him with a view to discovering what was right, and not with a view to proving such and such a course to be right. If an act is unimportant, and at the same time not obviously less right than some obvious alternative, we shall consider it neither moral nor immoral; for in such a case the act does not deserve careful consideration. The amount of care which a decision deserves depends upon its importance and difficulty; in the case of a statesman advocating a new policy, for example, years of deliberation may sometimes be necessary to excuse him from the charge of levity. But with less important acts, it is usually right to decide even when further reflection might show the present decision to be erroneous. Thus there is a certain amount of reflection appropriate to various acts, while some right acts are best when they spring from impulse (though these are such as reflection would approve). We may therefore say that an act is moral when it is one which the agent would judge to be right after an appropriate amount of candid thought, or, in the case of acts which are best when they are unreflecting, after the amount and kind of thought requisite to form a first opinion. An act is immoral when the agent would judge it to be wrong after an appropriate amount of reflection. It is neither moral nor immoral when it is unimportant and a small amount of reflection would not suffice to show whether it was right or wrong. (§ 22 ¶ 1)

We may now sum up or discussion of right and wrong. When a man asks himself: What ought I to do? he is asking what conduct is right in an objective sense. He cannot mean: What ought a person to do who holds my views as to what a person ought to do? for his views as to what a person ought to do are what will constitute his answer to the question What ought I to do? But the onlooker, who thinks that the man has answered this question wrongly, may nevertheless hold that, in acting upon his answer, the manwas acting rightly in a second, subjective, sense. This second sort of right action we call moral action. We held that an action is moral when the agent would judge it to be right after an appropriate amount of candid thought, or after a small amount in the case of acts which are best when they are unreflecting; the appropriate amount of thought being dependent upon the difficulty and importance of the decision. And we held that an action is right when, of all that are possible, it is the one which will probably have the best results. There are many other meanings of right; but these seem to be the meanings required for answering the questions: What ought I to do? and What acts are immoral? (§ 23 ¶ 1)