W. D. Ross, The Right and the Good, 1930.

APPENDIX I

RIGHTS

A general discussion of right or duty would hardly be complete without some discussion, even if only a brief one, of the closely related subject of rights. It is commonly said that rights and duties are correlative, and it is worth while to inquire whether and, if at all, in what sense this is true. The statement may stand for any one, or any combination, of the following logically independent statements:

  1. A right of A against B implies a duty of B to A.
  2. A duty of B toA implies a right of A against B.
  3. A right of A against B implies a duty of A to B.
  4. A duty of A to B implies a right of A against B.

What is asserted in (1) is that A's having a right to have a certain individual act done to him by B implies a duty for B to do that act to A; (2) asserts the converse implication; what is meant by (3) is that A's having a right to have a certain act done to him by B implies a duty for A to do another act to B, which act may be either a similar act (as where the right of having the truth told to one implies the duty of telling the truth) or a different sort of act (as where the right to obedience implies the duty of governing well); (4) asserts the converse implication.

Of these four propositions the first appears to be unquestionably true; a right in one being against another is a right to treat or be treated by that other in a certain way, and this plainly implies a duty for the other to behave in a certain way. But there is a certain consideration which throws doubt on the other three propositions. This arises from the fact that we have duties to animals and to infants. The latter case is complicated by the fact that infants, while they are not (so we commonly believe) actual moral agents, are potential moral agents, so that the duty of parents, for instance, to support them may be said to be counterbalanced by a duty which is not incumbent on the infants at the time but will be incumbent on them later, to obey and care for their parents. We had better therefore take the [49] less complicated case of animals, which we commonly suppose not to be even potential moral agents.

It may of course be denied that we have duties to animals. The view held by some writers is that we have duties concerning animals but not to them, the theory being that we have a duty to behave humanely to our fellow men, and that we should behave humanely to animals simply for fear of creating a disposition in ourselves which will make us tend to be cruel to our fellow men. Professor D. G. Ritchie, for instance, implies that we have not a duty to animals except in a sense like that in which the owner of an historic house may be said to have a duty to the house.1 Now the latter sense is, I suppose, purely metaphorical. We may in a fanciful mood think of a noble house as if it were a conscious being having feelings which we are bound to respect. But we do not really think that it has them. I suppose that the duty of the owner of an historic house is essentially a duty to his contemporaries and to posterity; and he may also think it is a duty to his ancestors. On the other hand, if we think we ought to behave in a certain way to animals, it is out of consideration primarily for their feelings that we think we ought to behave so; we do not think of them merely as a practising-ground for virtue. It is because we think their pain a bad thing that we think we should not gratuitously cause it. And I suppose that to say we have a duty to so-and-so is the same thing as to say that we have a duty, grounded on facts relating to them, to behave in a certain way towards them.

Now if we have a duty to animals, and they have not a duty to us (which seems clear, since they are not moral agents), the first and last of our four propositions cannot both be true, since (4) implies that a duty of men to animals involves a right of men against animals, and (1) implies that this involves a duty of animals to men, and therefore (4) and (1) together imply that a duty of men to animals involves a duty of animals to men. And since the first proposition is clearly true, the fourth must be false; it cannot be true that a duty of A to B necessarily involves a right of A against B. Similarly, the second and third propositions cannot both be true; for (2) and (3) taken together [50] imply that a duty of men to animals involves a duty of animals to men. But here it is not so clear which of the two propositions is true; for it is not clear whether we should say that though we have a duty to animals they have no right against us, or that though they have a right against us they have no duty to us. If we take the first view, we are implying that in order to have rights, just as much as in order to have duties, it is necessary to be a moral agent. If we take the second view, we are implying that while only moral agents have duties, the possession of a nature capable of feeling pleasure and pain is all that is needed in order to have rights. It is not at all clear which is the true view. On the whole, since we mean by a right something that can be justly claimed, we should probably say that animals have not rights, not because the claim to humane treatment would not be just if it were made, but because they cannot make it. But the doubt which we here find about the application of the term 'rights' is characteristic of the term. There are other ways too in which its application is doubtful. Even if we hold that it is our duty not merely to do what is just to others but to promote their welfare beyond what justice requires, it is not at all clear that we should say they have a right to beneficent treatment over and above what is just. We have a tendency to think that not every duty incumbent on one person involves a right in another.

This characteristic of our way of thinking about rights has been fastened upon by theory. Green, for instance, divides the whole region of duty into three parts: (1) moral duties which involve no rights on the other side, (2) obligations involving such rights, both obligations and rights being included in the jus naturae and being such as should be legally recognized, (3) legal obligations involving legal rights on the other side.2 He describes the rights in class (2) -- what I will for brevity call moral rights -- as sharing with legal rights the characteristic of depending for their existence on some form of general recognition. The recognition in the latter case consists in the making of a law; in the former it consists simply in a general state of public opinion. Now it is plainly wrong to describe either [51] legal or moral rights as depending for their existence on their recognition, for to recognize a thing (in the sense in which 'recognize' is here used) is to recognize it as existing already. The promulgation of a law is not the recognition of a legal tight, but the creation of it, though it may imply the recognition of an already existing moral right. And to make the existence of a moral right depend on its being recognized is equally mistaken. It would imply that slaves, for instance, acquired the moral right to be free only at the moment when a majority of mankind, or of some particular community, formed the opinion that they ought to be free, i.e. when the particular person whose conversion to this view changed a minority into a majority changed his mind. Such a view, of course, cannot be consistently maintained, and we find Green implying in successive sections that social recognition is indispensable to the existence of rights,3 and that the slave has a right to citizenship though this right is not recognized by society.4 In the latter passage we see the true Green, the passionate lover of liberty, reacting against the theory of the previous page. Some may think that slavery is not wrong; but every one will admit that there are certain forms of treatment of others which are wrong and which the sufferer has the right to have removed, whether this right is recognized by society or not.

There is, however, to be found in Green another view which is less clearly false. According to this, the existence of a right is made to depend not on the recognition of it but on the recognition of a power in the person in question to seek an end common to all the citizens of a community.5 This avoids the patent error of making the existence of a right depend on its being recognized to exist. Yet like the former view it makes a moral right depend not on the nature of a given person and his relations to his fellows, but on what people think about them, i.e. on what a majority of the community think about them. But [52] though the existence of legal rights depends on the degree of enlightenment of the community, the existence of moral rights plainly does not, but on the nature and relations of the persons concerned.

Green's theory seems to have arisen as follows. He starts his historical survey with Hobbes and Spinoza, both of whom identify right with power. A legal right may be identified with a certain kind of power; it is the power of getting certain things not by one's own brute force but by the aid of the law. Green seems to have tried to get a theory of moral rights by making a similar amendment of the bare identification of right with power; and he accordingly identifies them with tlhe power of getting certain things not by one's own brute force nor by the aid of the law but by the aid of public opinion; instead of saying, what is surely evident, that a moral right is not a power at all. Yet there are elements in his account which point to a truer theory; e.g. 'a "right" is an ideal attribution ("ideal" in the sense of not being sensibly verifiable)'.6 Now whether a given society recognizes a particular right is, I take it, sensibly verifiable in the sense in which Green here insists that a right is not. What is not sensibly verifiable is whether the society is justified in recognizing the right, and this depends on whether the right is there antecedently to society's recognition of it. Thus the insistence that a right is not sensibly verifiable points to an objective theory of rights; but unfortunately Green follows this clue no farther.

If we eliminate the possibility of holding that animals have rights, by saying that only that which has a moral nature can have a right, our main doubt with regard to the correlation of rights and duties is on the question whether there is a right to beneficence. It is obvious that a man has a right to just treatment, and it is commonly agreed that he has a right to have promises made to him fulfilled; it is less generally agreed that he has a right to beneficent treatment, even when it is admitted that it is our duty to treat him beneficently.

Some would even say that to treat others beneficently is to go beyond our duty. But probably this statement rests on a [53] mere confusion. We usually oppose justice to benevolence. But while treating a man justly is commonly understood to mean doing certain things to him (paying our debts to him, and the like), irrespective of the spirit in which we do them, treating him benevolently obviously means doing certain things to him from goodwill. And it is rightly felt that there is a great difference between the two things, and it is found natural to say that the one implies, and the other does not, a right on the other side, and (by some people) even to say that the one is a duty and the other is not. But if we will distinguish between doing what is just and doing it in the spirit of justice, and between doing what is beneficent and doing it in the spirit of beneficence, then (in accordance with the principle that it is always acts, and not acts from a certain motive, that are our duty) it is clear that it is not our duty to act in the spirit of justice, any more than in the spirit of beneficence, and that it is our duty to do what is beneficent, as it is our duty to do what is just.

If we are clear on this point, our main objection to saying that the other person has a right to beneficence disappears. I do not say that our whole objection disappears; for there hangs about the notion of a 'right' the notion of its being not only something which one person should in decency respect but also something which the other person can in decency claim, and we feel that there is something indecent in the making of a claim to beneficence.

These doubts about the application of the term 'right' appear to spring from the fact that 'right' (the noun) does not stand for a purely moral notion. It began, I suppose, by standing for a legal notion, and its usage has broadened out so as to include certain things that cannot be claimed at law; but its usage has not yet broadened out so much as to become completely correlative to duty. Once we start on the process of broadening it out, however, there seems to be no secure resting-place short of this.

Returning now to the four propositions about the correlativity of duties and rights, it seems that with regard to the second proposition, 'A duty of B to A implies a right of A [54] against B' (which has latterly been the subject of our discussion), we should say (1) that this is not true when A is not a moral agent, and (2) that it is true when A is a moral agent (even if the duty be the duty of beneficent action). And since our only doubt about the third proposition, 'A right of A against B implies a duty of A to B', arises from our doubt whether animals have not rights, if we agree that animals have not rights we need not doubt the truth of this proposition. It is this proposition, above all, that has been maintained by those who have insisted on the correlativity of rights and duties; for this was maintained essentially against the belief that men have 'natural rights' in a state of nature in which they have no duties.

A further problem, however, awaits us, viz. whether a failure to do one's duty involves a corresponding loss of right. Or rather, as we have found the meaning of 'rights' more doubtful than that of 'duties', it will be more profitable to omit any reference to rights, and put our question in the form, 'if A fails in his duty to B, does that put an end to B's duty to A?' In some cases we seem to be clear that this is so. If a tradesman sends me goods inferior to those I chose in his shop, I am not morally, any more than legally, bound to pay him the full price; I may return the goods and pay nothing, or (with his consent) keep them and pay a lower price. And in general any duty arising out of a contract is cancelled by non-fulfilment of the corresponding duty on the other side. In other cases we are not so clear. It is not so generally agreed, for instance, that if A tells lies to B, B is justified in telling lies to A. Two blacks, we say in such a case, do not make a white. Yet the peculiar stringency of the duty of veracity seems to spring from an implicit understanding that language shall be used to convey the real opinions of the speakers, and it would seem that a failure to carry out the understanding on one side makes it no longer binding on the other; and we should have small patience with an habitual liar who insisted on strict veracity in others. It must be admitted that a man who has deceived me has destroyed what would have been the main reason for its being my duty to tell him the truth. But we should probably hesitate to say that by his breach of the implicit understanding my duty [55] to tell him the truth has been entirely destroyed, as by the tradesman's breach of contract my duty to pay him has been destroyed. Various reasons help to account for this. For one thing, it is likely that by deceiving a liar I may indirectly deceive innocent people; for another, the consequences for my own character are likely to be particularly dangerous. But the main reason probably lies elsewhere. Before the contract was made between my tradesman and me, there was no duty incumbent on me of paying him this sum of money. I had a general duty to promote the good of all men, but there was no obvious reason for supposing that this could be best done by transferring this sum of money to him. But even before the implicit undertaking to tell the truth was established I had a duty not to tell lies, since to tell lies is prima facie to do a positive injury to another person. Since this duty does not rest on contract, it is not abolished by the breach of contract, and therefore while a person who has been deceived by another is justified in refusing to answer his questions, he is not justified in telling him lies. Yet that this forms only a small part of the stringency of the duty of truthfulness may be inferred from the leniency with which we should judge deceit, in a case in which no implicit undertaking to tell the truth has been established, e.g. when a civilized man deceives a savage whom he has just met for the first time, or vice versa, or when one of two savages belonging to different tribes deceives the other. Deceit is much more venial in such a case, because the offender has no reason to suppose that the other is not deceiving, or going to deceive, him.

Taking, then, the obvious division between duties arising out of contract and those that arise otherwise, we must say that while the former are cancelled by breach of the contract on the other side, the latter are not cancelled by the bad behaviour of the other person. It would also seem, from a consideration of our actual moral judgements, that the former type of duty is the more stringent of the two.

Now the distinction between the rights corresponding to duties that arise out of contract, and the rights corresponding to other duties, may be quite suitably expressed as a distinction [56] between contractual and natural rights, and the notion of natural rights as a distinct class may thus be vindicated, if it be cut free from the belief which has been so often bound up with it, that there are rights in a state of nature, i.e. in a state in which there are no duties. Such a belief is made possible for Hobbes only by a complete confusion between rights and powers, amounting to an express identification of the two.

APPENDIX II

PUNISHMENT

In connexion with the discussion of rights it is proper to consider a question which has always interested and usually puzzled moralists, and which forms a crucial example for the testing of moral theories -- the question of punishment. A utilitarian theory, whether of the hedonistic or of the 'ideal' kind, if it justifies punishment at all, is bound to justify it solely on the ground of the effects it produces. The suffering of pain by the person who is punished is thought to be in itself a bad thing, and the bringing of this bad thing into the world is held to need justification, and to receive it only from the fact that the effects are likely to be so much better than those that would follow from his non-punishment as to outweigh the evil of his pain. The effects usually pointed to are those of deterrence and of reformation. In principle, then, the punishment of a guilty person is treated by utilitarians as not different in kind from the imposition of inconvenience, say by quarantine regulations, on innocent individuals for the good of the community. Or again, if a state found to be prevalent some injury to itself or to its members that had not been legislated against, and proceeded to punish the offenders, its action would in principle be justified by utilitarians in the same way as its punishment of offenders against the law is justified by them, viz. by the good of the community. No doubt the state would have greater difficulty in justifying its action, for such action would produce bad consequences which the punishment of law-breakers does not. But the difference would be only in degree. Nay more, a [57] government which found some offence against the law prevalent, and in its inability to find the offenders punished innocent people on the strength of manufactured evidence, would still be able to justify its action on the same general principle as before.

Plain men, and even perhaps most people who have reflected on moral questions, are likely to revolt against a theory which involves such consequences, and to exclaim that there is all the difference in the world between such action and the punishment of offenders against the law. They feel the injustice of such action by the state, and are ready to say, in the words imputed to them by Mr. Bradley: 'Punishment is punishment, only when it is deserved. We pay the penalty because we owe it, and for no other reason; and if punishment is inflicted for any other reason whatever than because it is merited by wrong, it is a gross immorality, a crying injustice, an abominable crime, and not what it pretends to be. We may have regard for whatever considerations we please -- our own convenience, the good of society, the benefit of the offender; we are fools, and worse, if we fail to do so. Having once the right to punish, we may modify the punishment according to the useful and the pleasant; but these are external to the matter, they cannot give us a right to punish, and nothing can do that but criminal desert.'7

There is one form of utilitarian view which differs in an important respect from that above ascribed to utilitarians. Professor Moore admits the possibility, which follows from his doctrine of organic unities, that punishment may not need to be justified merely by its after-effects. He points out8 that it may well be the case that though crime is one bad thing and pain another, the union of the two in the same person may be a less evil than crime unpunished, and might even be a positive good. And to this extent, while remaining perfectly consistent with his own type of utilitarianism, he joins hands with intuitionists, most of whom, at any rate, would probably hold that the combination of crime and punishment is a lesser evil than unpunished crime.

Most intuitionists would perhaps take the view that there is a fundamental and underivative duty to reward the virtuous [58] and to punish the vicious. I am inclined to diverge from this view. Two things seem to me to be clear: that we have a prima facie duty to do this, and that a state of affairs in which the good are happy and the bad unhappy is better than one in which the good are unhappy and the bad happy. Now if the first of these is an underivative fact, the two facts are logically unconnected. For it can be an underivative fact only if the intuitionist view is true, and if that view is true the superiority of the one state of affairs over the other cannot follow from the duty of producing it, since on the intuitionist view there are duties other than the duty of producing good. But an intuitionist may with propriety perform the reverse derivation; he may derive the prima facie duty of reward and punishment from the superiority of the state of affairs produced, since he may -- and, as I think, must -- admit that if a state of affairs is better than its alternatives there is a prima facie duty to produce it if we can. The duty of reward and punishment seems to me to be in this way derivative. It can be subsumed under the duty of producing as much good as we can; though it must be remembered that the good to be produced in this case is very different from the other goods we recognize (say virtue, knowledge, and pleasure), consisting as it does in a certain relative arrangement of virtue, vice, pleasure, and pain.

But if we hold that there is this duty, it must be admitted that it is one which it is very difficult for us to see our way to performing, since we know so little about the degrees of virtue and vice, and of happiness and unhappiness, as they occur in our fellow men. And in particular there are two grave objections to holding that the principle of punishing the vicious, for the sake of doing so, is that on which the state should proceed in its bestowal of punishments.

(1) What we perceive to be good is a condition of things in which the total pleasure enjoyed by each person in his life as a whole is proportional to his virtue similarly taken as a whole. Now it is by no means clear that we should help to bring about this end by punishing particular offences in proportion to their moral badness. Any attempt to bring about such a state of affairs should take account of the whole character of the persons [59] involved, as manifested in their life taken as a whole, and of the happiness enjoyed by them throughout their life taken as a whole, and it should similarly take account of the virtue taken as a whole, and of the happiness taken as a whole, of each of the other members of the community, and should seek to bring about the required adjustments. In the absence of such a view of the whole facts, the criminals that a retributive theory of state punishment would call on us to punish for the sake of doing so may well be persons who are more sinned against than sinning, and may be, quite apart from our intervention, already enjoying less happiness than a perfectly fair distribution would allow them. The offences which the state legislates against are only a small part of the wrong acts which are being done every day, and a system which punishes not all wrong acts, but only those which have been forbidden by law, and does not attempt to reward all good acts -- such an occasional and almost haphazard system of intervention does not hold out any good hope of promoting the perfect proportionment of happiness to virtue. Nor would it be in the least practicable for the state to attempt the thorough review of the merit and the happiness of all its members, which alone would afford a good hope of securing this end.

(2) Even if it were practicable, it is by no means clear that it is the business of the state to aim at this end. Such a view belongs, I think, to an outworn view of the state, one which identifies the state with the whole organization of the community. In contrast to this, we have come to look upon the state as the organization of the community for a particular purpose, that of the protection of the most important rights of individuals, those without which a reasonably secure and comfortable life is impossible; and to leave the promotion of other good ends to the efforts of individuals and of other organizations, such as churches, trade unions, learned and artistic societies, clubs. Now it cannot, I think, be maintained that the apportionment of happiness to merit is one of the essential conditions to the living of a reasonably secure and comfortable life. Life has gone on for centuries being lived with reasonable security and comfort though states have never achieved or [60] even attempted with any degree of resolution to effect this apportionment. And in fact for the state to make such an attempt would seriously interfere with its discharge of its proper work. Its proper work is that of protecting rights. Now rights are (as we have seen) rights to be treated in certain ways and not to be treated in certain ways from certain motives; what the state has to take account of, therefore, is not morally bad actions, but wrong acts, and it has to take account of them in such a way as to diminish the chance of their repetition. And this attempt would only be interfered with if the state were at the same time trying to effect a proportionment of happiness to moral worth in its members. The latter task, involving as it would a complete review of the merit and happiness of all its members, would involve leaving the punishment for each offence undetermined by law, and to be determined in the light of all the circumstances of each case; and punishment so completely undetermined in advance would be quite ineffective as a protector of rights.

But to hold that the state has no duty of retributive punishment is not necessarily to adopt a utilitarian view of punishment. It seems possible to give an account of the matter which retains elements in punishment other than that of expediency, without asserting that the state has any duty properly denned as the duty of punishing moral guilt. The essential duty of the state is to protect the most fundamental rights of individuals. Now, rights of any human being are correlative to duties incumbent on the owner of rights, or, to put it otherwise, to rights owned by those against whom he has rights; and the main element in any one's right to life or liberty or property is extinguished by his failure to respect the corresponding right in others.9 There is thus a distinction in kind which we all in fact recognize, but which utilitarianism cannot admit, between the punishment of a person who has invaded the rights of others and the infliction of pain or restraint on one who has not. The state ought, in its effort to maintain the rights of innocent persons, to take what steps are necessary to prevent violations of these rights; and the offender, by violating the life or liberty or property of [61] another, has lost his own right to have his life, liberty, or property respected, so that the state has no prima facie duty to spare him, as it has a prima facie duty to spare the innocent. It is morally at liberty to injure him as he has injured others, or to inflict any lesser injury on him, or to spare him, exactly as consideration both of the good of the community and of his own good requires. If, on the other hand, a man has respected the rights of others, there is a strong and distinctive objection to the state's inflicting any penalty on him with a view to the good of the community or even to his own good. The interests of the society may sometimes be so deeply involved as to make it right to punish an innocent man 'that the whole nation perish not'. But then the prima facie duty of consulting the general interest has proved more obligatory than the perfectly distinct prima facie duty of respecting the rights of those who have respected the rights of others.

This is, I believe, how most thoughtful people feel about the affixing of penalties to the invasion of the rights of others. They may have lost any sense they or their ancestors had that the state should inflict retributive punishment for the sake of doing so, but they feel that there is nevertheless a difference of kind between the community's right to punish people for offences against others, and any right it may have to inconvenience or injure innocent people in the public interest. This arises simply from the fact that the state has a prima facie duty not to do the latter and no such duty not to do the former.

We can, I think, help ourselves towards an understanding of the problem by distinguishing two stages which are not usually kept apart in discussions of it. The infliction of punishment by the state does not, or should not, come like a bolt from the blue. It is preceded by the making of a law in which a penalty is affixed to a crime; or by the custom of the community and the decisions of judges a common law gradually grows up in which a penalty is so affixed. We must, I think, distinguish this stage, that of the affixing of the penalty, from that of its infliction, and we may ask on what principles the state or its officials should act at each stage.

At the earlier stage a large place must be left for [62] considerations of expediency. We do not claim that laws should be made against all moral offences, or even against all offences by men against their neighbours. Legislators should consider such questions as whether a given law would be enforced if it were made, and whether a certain type of offence is important enough to make it worth while to set the elaborate machinery of the law at work against it, or is better left to be punished by the injured person or by public opinion. But even at this stage there is one respect in which the notion of justice, as something quite distinct from expediency, plays a part in our thoughts about the matter. We feel sure that if a law is framed against a certain type of offence the punishment should be proportional to the offence. However strong the temptation to commit a certain type of offence may be, and however severe the punishment would therefore have to be in order to be a successful deterrent, we feel certain that it is unjust that very severe penalties should be affixed to very slight offences. It is difficult, no doubt, to define the nature of the relation which the punishment should bear to the crime. We do not see any direct moral relation to exist between wrong-doing and suffering so that we may say directly, such and such an offence deserves so much suffering, neither more nor less. But we do think that the injury to be inflicted on the offender should be not much greater than that which he has inflicted on another. Ideally, from this point of view, it should be no greater. For he has lost his prima facie rights to life, liberty, or property, only in so far as these rested on an explicit or implicit undertaking to respect the corresponding rights in others, and in so far as he has failed to respect those rights. But laws must be stated in general terms, to cover a variety of cases, and they cannot in advance affix punishments which shall never be greater than the injury inflicted by the wrong­doer. We are therefore content with an approximation to what is precisely just. At the same time we recognize that this, while it is a prima facie duty, is not the only prima facie duty of the legislator; and that, as in the selection of offences to be legislated against, so in the fixing of the penalty, he must consider expediency, and may make the penalty more or less severe as it [63] dictates. His action should, in fact, be guided by regard to the prima facie duty of injuring wrong-doers only to the extent that they have injured others, and also to the prima facie duty of promoting the general interest. And I think that we quite clearly recognize these as distinct and specifically different elements in the moral situation. To say this is not to adopt a compromise between the intuitionist and the utilitarian view; for it can fairly be claimed that one of the duties we apprehend intuitively is that of promoting the general interest so far as we can.

When the law has been promulgated and an offence against it committed, a new set of considerations emerges. The administrator of the law has not to consider what is the just punishment for the offence, nor what is the expedient punishment, except when the law has allowed a scale of penalties within which he can choose. When that is the case, he has still to have regard to the same considerations as arose at the earlier stage. But that, when the penalty fixed by law is determinate, this and no other should be inflicted, and that, when a scale of penalties is allowed, no penalty above or below the scale should be inflicted, depends on prima facie duty that did not come in at the earlier stage, viz. that of fidelity to promise. Directly, the law is not a promise: it is a threat to the guilty, and a threat is not a promise. The one is an undertaking to do or give to the promisee something mutually understood to be advantageous to him; the other, an announcement of intention to do to him something mutually understood to be disadvantageous to him. Punishment is sometimes justified on the ground that to fail to punish is to break faith with the offender. It is said that he has a right to be punished, and that not to punish him is not to treat him with due respect as a moral agent responsible for his actions, but as if he could not have helped doing them. This is, however, not a point of view likely to be adopted by a criminal who escapes punishment, and seems to be a somewhat artificial way of looking at the matter, and to ignore the difference between a threat and a promise.

But while the law is not a promise to the criminal, it is a promise to the injured person and his friends, and to society. [64] It promises to the former, in certain cases, compensation, and always the satisfaction of knowing that the offender has not gone scot-free, and it promises to the latter this satisfaction and the degree of protection against further offences which punishment gives. At the same time the whole system of law is a promise to the members of the community that if they do not commit any of the prohibited acts they will not be punished.

Thus to our sense that prima facie the state has a right to punish the guilty, over and above the right which it has, in the last resort, of inflicting injury on any of its members when the public interest sufficiently demands it, there is added the sense that promises should prima facie be kept; and it is the combination of these considerations that accounts for the moral satisfaction that is felt by the community when the guilty are punished, and the moral indignation that is felt when the guilty are not punished, and still more when the innocent are. There may be cases in which the prima facie duty of punishing the guilty, and even that of not punishing the innocent, may have to give way to that of promoting the public interest. But these are not cases of a wider expediency overriding a narrower, but of one prima facie duty being more obligatory than two others different in kind from it and from one another.


Notes

1 Natural Rights, 108.

2 Principles of Political Obligation, §§ 10, 11.

3 'A claim to which reality is given by social recognition, and thus implicitly a right (§ 139). Cf. 'This recognition of a power, in some way or other, as that which should be, is always necessary to render it a right' (§ 23). 'Rights are made by recognition. There is no right "but thinking makes it so" ' (§ 136).

4 § 140 implies that the slave's right to citizenship is founded on his possessing a common human consciousness with the citizens of the state.

5 Cf. e.g. §§ 25, 26.

6 § 38.

7 Ethical Studies, ed. 2, 26-7.

8 Principia Ethica, 214.

9 Cf. pp. 54-5.