This paper was read to the Societas Ethica at Noordwijk aan Zee, Holland, in 1977, and subsequently published in their Proceedings. It was my first exposition of the ideas I subseqently developed in On Justice. The earlier part of the paper is substantially the same as Chapter I of that book, and the later part indicates work still needing to be done.


Remota iustitia, asked St. Augustine, 1 quid sunt regna nisi magna latrocinia? States without justice are but robber bands enlarged; and earlier Plato had taken it as the mark of the just society that ' it should be free from dissension. Justice is the bond of society, and without it no association of human individuals could subsist. 2 Most thinkers, however, have sought the key to the concept of justice elsewhere, and have construed it in terms of rules or utility or equality; or, more recently, have been concerned simply to observe and record the different usages of the word 'just', without attempting to articulate an account of why the same word should be used in so many different contexts. Justice has in consequence been much misunderstood, and in practice much neglected. This is why our society has become increasingly divided, and why its members have been increasingly alienated from its authority. Authority now is seen as something external, a force to be reckoned with, not a guidance to be reasoned with and accepted. It is a pity. We have been pursuing the wrong political goals - productivity, efficiency, equality - and have neglected the cardinal political virtue of justice, which, together with liberty, is the condition under which I and every man can identify with society, feel at one with it, and accept its ruling as my own. It is therefore justice that we must seek.

We can go some way by linguistic analysis. If we consider carefully how the words 'just' and 'unjust', `fair' and 'unfair' are used, we can see that men and actions and decisions and laws can all be said to be just, and that justice is not the same as expediency, prudence, equality, liberty, generosity, friendliness, mercy or good will, and that since a decision can be said to be both just - in that it was reached by due process by an impartial judge who applied the relevant law - and unjust - in that the law itself was unjust or that the judge was precluded from taking certain relevant factors into consideration, justice itself must be a complicated concept. But if we are to go further, and discover not only the uses of the word but the underlying resemblances in virtue of which the uses form a single family, we need to consider not merely the applications of the word but the arguments we use when issues of justice arise. We cannot know what justice is until we also know why it is a good thing, and by listening to its praises.being sung can come to realise what it must be in order to be argued for in exactly these terms.

According to the Greeks, justice was an a)llo'trion a)gaqo'n (allotrion agathon), the other chap's good. Certainly, questions of justice do not arise except where other people are involved and except where some sorts of good, positive or negative, are in issue. Robinson Crusoe had no opportunity of practising justice or injustice until Man Friday came on the scene. He could be resourceful, brave and temperate on his own, but justice, like compassion and mercy, is an essentially other-regarding virtue. Only when two or more are gathered together, can either exercise justice towards the other, and only then can their arrangements and transactions be acknowledged as just or stigmatized as unjust. But justice is not the other other-regarding virtue. It differs not only from compassion and mercy, for reasons which we shall examine later, but from other political virtues as well. It is not liberty, it is not equality, it is not fraternity. Justice is not liberty or freedom, because liberty and freedom are concerned with the question who shall take the decisions, whereas justice is concerned with how decisions shall be taken, in what frame of mind and with what result. Justice is not equality, because equality is concerned only with results, and not how they are arrived at, and equality is concerned only that people should be treated the same, whereas justice is concerned to consider each individual case on its own merits, treating, if necessary different people differently, as when we punish the guilty and let the innocent go free. Justice is not fraternity, because fraternity is a warm virtues concerned with fellow-feeling, whereas justice is a cold virtue which can be manifested without feeling, and is concerned to emphasize that the other chap is not merely a human being like myself, but a separate individual, with his own point of view and own interests that are distinct from mine.

The distinctness of the other chap becomes important when there is conflict, and one man can get his own way only at the cost of some one else. It is only when somebody's rights or interests are in jeopardy that the unity and coherence of society is under strain; and so it is only then that issues of justice arise. That is to say, it is when injustice is in danger of being done that we become agitated. Injustice wears the trousers. And therefore we should follow the example of Aristotle, and adopt a negative approach, discovering what justice is by considering on what occasions we protest at injustice or unfairness. 3 If we focus our attention only on the positive virtue, we see it as something flat, and without depth or dynamic vigour. We have already seen that justice is a cold virtue. But injustice is something we soon get steamed up about. Although it is not enough to be just or fair, and people and institutions need to be something more if they are to engage our allegiance and affection, nevertheless injustice or unfairness is a fundamental criticism; a man who is unfair is one we want to have as little to do with as possible, and we have little loyalty to any institution or arrangement which we regard as basically unjust. Not only are our emotions a good guide to our intellectual discriminations; but in this case they reveal the underlying thrust of the concept. If I talk only about justice, I am in danger or relapsing into platitudes: it is when I get hot under the collar about some specific piece of unfairness, that my eloquence has an edge to it, and I really know what is getting my goat. And, as we are not beginning to see, the contrast between the mild favour I feel towards fairness and the intense fury unfairness arouses in my breast is symptomatic of a basic asymmetry between justice and injustice which, I shall argue, is crucial to the part they play in our conceptual structure.

'Unfair' is a word of protest. We use it to protest, either on our own behalf or on that of others, when someone has been done down. Both words 'done' and 'down' are important. We cannot entertain the complaint of unfairness, unless there was some agency involved, human or divine. If the dice come down double one in the game of "Monopoly" and I have to move from Park Lane to Mayfair, I may complain of my hard luck but I cannot say that I have been treated unfairly. Equally with the changes and chances of our ordinary life, we cannot complain of unfairness, unless they are due to human actions or we attribute them to God's almighty hand. Moreover, it is not enough merely to be the result of some action or actions: the result must have been foreseen or at least foreseeable by some of the agents involved. The victim of an accident can bewail his misfortune, but not complain of having been unfairly treated, unless one of the agents was negligent and failed to discharge his duty of care towards him, that is, unless somebody failed to manifest sufficient consideration of his, the victim's interests. We need to shift our attention from the bare consequences of actions to the actions themselves. It has been one of our main difficulties in reaching a proper understanding of justice that we are dominated by a wrong view of public life. We think of it as a course of events, not as a system of activities. We consider states of affairs, not actions. Hence we focus on the end-result of activities and processes, and ignore the procedure whereby these results were reached. But with actions, what is done depends partly on the doing of it. Different actions may end in the same state of affairs, but are different nonetheless, and have very different significance for agents and participants. Once we recognize the fact that human beings are primarily agents, not patients, our whole perspective alters. We are concerned as much with what people do as with what gets done, and will be concerned not only with whether the outcome is just, but with whether the agents acted justly in doing what they did. Although other things - laws, particular decisions, general economic arrangements and particular payments - can rightly be described as just or unjust, these are derivative uses deriving from the man in a just frame of mind and the laws he would enact, the decisions he would take, the economic arrangements he would approve of and the particular payments he would make. Justinian and Aquinas were right in characterizing justice as a certain state of mind.

It follows that justice is concerned not only with the consequences of actions but with their significance. It is not so much the injury as the insult that arouses us, and the appropriate reaction to injustice is not simple anger, but rather, indignation. We are indignant, because, as the etymology of the word shows, we have not been regarded as worth anything, as worthy of respect. Unfairness is resented not primarily because of hurt, harm or damage caused to the victim, but as belittling or affronting his worth. Although causality is important for evaluating actions, meaning is more important still, and many perplexing features of justice in general, and in particular of punishment, become intelligible only when actions are construed as a species of communication, like language, and not merely as events within the nexus of cause and effect.

It follows also that justice is concerned with procedures as well as with outcomes, and that the rules of natural justice are properly so called. Rules such as audi alteram partem and that nobody shall be judge in his own cause manifest to any potentially disappointed party our reluctance to decide against him, and our determination that any adverse decision shall be, and shall be seen by him to be, taken only for reasons which are relevant to his case and whose cogency he can appreciate. By keeping to these rules of procedure we show ourselves dispassionately tender-minded towards each and every one concerned, so that nobody can feel slighted or disregarded or believe that he and his interests are of no account.

The word 'down' is also significant. It indicates that justice is neither a mean nor a sort of equality, as Aristotle thought, 4 but is essentially one-way in its concern and is focussed on the individual. There is an essential asymmetry in justice. Good things and bad things are not on a par. Instead of construing it as a concern for the other chap's faring well, the a)llo`trion a)gaqo`n (allotrion agathon) of the Greeks, we should understand it, rather, as an a)llo`trion m*h kako'n (allotrion me kakon), a concern for the other chap's not faring ill. Contrary to the opinion of many, it is not inherently unfair if someone has more than anybody else, but only if others are thereby done out of their due. Generosity is not unjust. Nor is a rational exercise of mercy in appropriate cases. Nor again would Aristotle's dictum a)dikeitai d' ou)deij e(kw`n (adikeitai d'oudeis hekon), one cannot be unjustly treated if it is with one's own consent, 5 be ever true if justice were any sort of equality, concerned to treat all those falling in some class alike, or if it were really a mean, and as much concerned to prevent people being done up as being done down.

The Theory of Games, although concerned only with results and not procedures, provides a useful scheme for formulating this point and contrasting it with classical utilitarianism and the recent theory of John Rawls. In the Theory of Games the different consequences of the different decisions made by the different parties are called "outcomes", and the value for each outcome is called its "pay-off". Classical utilitarianism evaluates outcomes by simply summing the pay-offs, and recommends that we go for the outcome of which the total of the pay-offs is highest; or, in its own terminology, the greatest good of the greatest number. Such a recommendation offends our sense of justice, because it has no regard for any particular individual, and will happily sacrifice his good so long as the total is thereby maximised. It treats people as units, not individuals, and like Caiaphas counsels that one man should die for the people.

Rawls' theory is an improvement. It cannot countenance the worst injustices, because it does not aggregate the pay-offs of each outcome, but considers only the lowest. Each outcome is evaluated by reference to the minimum pay-off it has for any party. And we then seek to maximise the minimum, that is to bring about the outcome which has the highest minimum pay-off. Such a "maximin" policy obviates the worst injustices. If we always consider things from the point of view of the worst off, we cannot do down anyone utterly, for if he will end up worse off than anyone else and than he otherwise would have been, the action will be deemed unjust. Rawls' theory, however, bids us consider only the worst off, and people can suffer injustice without thereby becoming worse off than everybody else. Although it is right to be concerned about the underdog, it is wrong to concentrate on him to the exclusion of every other dog. After all, even overdogs are worthy of respect, and may be in danger of being done out of their due. If we are to do justice, we need to consider things from everyone's point of view, separately, not only that of those in danger of being made worse off than anybody else. Instead of asking just one question, either about the total outcome as the utilitarians do, or about the outcome of the least favoured individual as Rawls recommends, we must ask many questions, one for each individual concerned: and instead of answering it solely by reference to the outcome, we must consider also the way in which the result is reached, and whether it could reasonably have been less unfavourable, for each particular individual, than in fact it was. If for any individual, and not only the worst off, the result is unfavourable, and on further scrutiny we find it is unnecessarily unfavourable, then it is unfair on him.

Some unfavourable decisions are unavoidable. To decide for the plaintiff is to decide against the defendant, and vice versa. Hard decisions have to be taken. We cannot ensure that nobody will ever be hardly done by, but only that those who are being hardly done by are being hardly done by for reasons which even they must acknowledge as weighty. Although a man may be disappointed at a just decision, he should not be angry or indignant. It may run against his interests, and he may wish that it could have been otherwise, but he has to acknowledge the justice of it, and cannot feel that he is being treated without due consideration, or that the decision shows that society has no concern for him, or that he is as nothing in its eyes. Not only was there a manifest reluctance to reach an adverse decision, but it was overborne by reasons into which he can enter as well as any other man. It is because there are reasons into which he can enter, that he can be reconciled to the decision, even though it was adverse, and, as Plato observed, his propensity to anger and righteous indignation is assuaged by the decision's being rational, rather than antagonized by its being adverse. 6 These reasons have to be of a special kind. Not every reason for an action will reconcile those who suffer in consequence. A ruler may act not only arbitrarily out of caprice, but for many reasons which will make his actions intelligible, but not just. Some people have opposed justice to arbitrariness, but, although arbitrary decisions are likely to be unjust, not all non-arbitrary decisions are therefore just. A decision taken for fear or favour is intelligible but not one whose cogency a disappointed litigant could be expected to accept. A government may be guided by raison d'etat, and act entirely rationally, but quite unjustly. It may be highly expedient to hand me over to a neighbouring power, to destroy my house, seize my goods, arrest me, hold me as a hostage, or execute me as a grim warning to others, but none of these actions is just, although all are highly rational.

These reasons will not reconcile the unfortunate man to his hard fate, because although they show why the government acted as it did, they do not show that it had to, - that there was no reasonable alternative to coming down on this one man, or that due weight had been given to his interest in escaping this fate. Sacrifices cannot be called for merely because they are expedient, from the government's point of view, but only if they are inevitable. And even if sacrifices must be made, the individual may still protest that they should not-all fall on him. Justice argues against our picking on one man rather than another. Where there are burdens to be borne, they should be laid on everyone, or on everyone capable of bearing them, or at least on the relatively many shoulders of those best able to bear them. Reasons of state, that is, should not bear immediately on adverse decision-making, but only through the medium of general laws, applicable not to one individual alone, but to all those falling within some general category. The fact that it would be expedient to deprive one man of his liberty is not warrant enough for doing so; but if it is not only expedient but necessary for the safety of the state that people be called up, and there are general laws governing conscription, and these for good reason make young men of good health liable for National Service for a limited period, subject to relatively few and reasonable exemptions, then a person called to the colours can no longer complain that he is being unjustly treated. Although the reasons for his loss of liberty are, basically, ones extraneous to his particular case, there are compelling reasons why the burden of National Service should be imposed in general, and why he should be one of those on whom it is imposed in particular. A person who nonetheless maintains that he ought not to be called up, is reduced to arguing either that the country should not be adequately defended, or that others, rather than he, should bear all the burden of defence. And, at that stage, he, as a member of the community, will acknowledge that these alternatives are not really available, and that, therefore, the arguments about calling him up are inescapable. And then, acknowledging the force of these arguments, he will accept the decision, even though adverse to his own interests, as fair and reasonable all the same.

Just decisions are reasonable, but not every reasonable decision is just. Only these reasons are just that are both inescapable and individualised. They must show why the decision had to be taken, and why it had to be that particular man who was decided against. This is why justice is characteristically backward-looking, in contrast to expediency and utility, which look to the future: only what a man has done or already is constitutes a firm fact about him, and although it may be rational, because effective, to "punish" him for what it is expected he will do, it is unjust, because until he had done it, he may always change his mind and not do it. Only past deeds and present facts are sufficiently actual and unquestionable to provide an adequate basis for an adverse decision. This also is why the rule audi alteram partem is the most important of our rules of natural justice. A just decision must be based on the circumstances of the case, and is unfair if either it is based on some factor that is not relevant or fails to take into account some factor that is, and only by hearing both sides can we elicit all those that are relevant and eliminate any that are not.

Although we can easily agree that certain sorts of reason - fancy, fear, favour or raison d'etat - do not constitute adequate reasons for an adverse decision, we find it difficult to formulate exact criteria of cogency and relevance. The problems are different in corrective and in distributive justice. In corrective - what Aristotle called diorthotic - justice there is no problem of individuations. In adjudication of rights, the sort of cases that are decided by the civil courts, the parties are involved because it is their claims that conflict, and not anybody else's. In criminal cases the accused is involved, because it is he who is alleged to have done wrong; and if he can show that it was not he, but someone else, who did the wrong,. then he is free. What is at issue is not why this particular man should be singled out for adverse treatment, but whether the reasons for meting out such treatment are sufficiently compelling. In the criminal case he can, according to English law, be punished only if he is proved to have committed the crime beyond reasonable doubt. In the civil case the burden of proof is not laid so heavily on the plaintiff, because that would be unfair to him, and the case is decided on a balance of probabilities. In each case there have to be reasons - good reasons - why our reasonable reluctance to decide against anyone had to be overridden. It is normally clear what reasons are relevant on either side. The difficulty is to assess their weight. We are able to eke out our limited agreement with the aid of precedent and convention. Although we cannot all be sure which way the decision should go, we can all be sure that cases should not be treated differently except for good reason, and we therefore lay down, as a further principle of justice, that like cases should be treated alike, and hence that once a decision has been reached, it should be a precedent for subsequent cases, unless they are distinguished from it by reason of some further circumstance. The rule of precedents enables law to grow, introducing greater definition into areas of dispute. Of course, the leading case may have been wrongly decided, with the result that all subsequent ones have to follow suit. This is sometimes a real problem, and may require legislative intervention, but for the most part it does not matter because once people know what the law is, they know where they stand, and can adjust their behaviour accordingly. Often, although not always, definition is enough. Considerations of practical utility apart, it reflects the fact that justice is, as we have seen, very much a matter of communication, where, as in language, meaning is constituted by convention. Convention, whether developed through precedent or established explicitly by legislative enactment, plays a very large part in law, and often it is sufficient that the law be generally known for it to be just. But not always. Since it is intelligible before a case is decided to argue that justice requires it to be decided one way rather than the other, it is intelligible also to say that it was in the event decided the wrong way. Hence it is possible that everyone, apart from the judge himself, should think his decision unjust, although acknowledging it to be a valid decision and a precedent for others. Similarly, laws can be said to be unjust. Although sometimes conventional, they are not always so. It is a convention which side of the road we drive on, but not whether we punish sheep-stealing with death. Although justice is concerned with the meaning of actions, actions acquire their significance not solely by convention, as words do, but also on account of their consequences. Insult, although different from injury, is conceptually linked with it. For those, deep philosophical reasons, the claim of the legal positivists that laws cannot be intelligently stigmatized as unjust is to be rejected. Laws which do people down without compelling reason are unfair. There are many such laws. Their status and claim on our allegiance are questions which lie outside the scope of this paper. Elsewhere I have argued that although we cannot go along with the schoolman and say simply lex iniusta non est lex. we can in extreme cases deny the validity of formally enacted laws, and say lex iniustissima non est lex.

Distributive justice is traditionally taken to deal with the apportionment of benefits and burdens among the members of some society or association. It is important, however, to distinguish the allocation of benefits from the imposition of burdens, and I confine the terms 'distributive justice' to the former, and propose the new name 'contributive justice' for the latter. In each case the basis of apportionment must be relevant facts about the individual. Benefits are distributed according to desert, merit, need, entitlement or status: contributions are called for on the basis of ability, demerit, expected subsequent benefit, agreement or status. Each of these bases is an individualised reason of the right logical type to ground a just decision on, and attempts to make out that only one - say, need or ability - can be the basis of a fair apportionment are misconceived. Rather, each is appropriate in certain contexts, depending on the benefit or burden to be apportioned, the nature of the association, and the way in which the benefit was generated or the purpose for which the burden was incurred. Often there may be conflict between claims based on different bases, and we may have to compromise the claims of desert to accommodate those of need, and vice versa. This does not mean, as many thinkers have alleged, that the concept of distributive justice is incoherent, any more than the fact that in litigation we have to weigh one man's rights against another's and strike a balance means that the justice of the courts is incoherent. Many modern thinkers have attempted, as a simpliste solution, to equate justice with equality. This is a mistake. The concepts are different. If the concepts had been the same, the complaint of the labourers in the vineyard who had borne the burden and the heat of the day (St. Matthews 20, 12) would have been unintelligible. But the Lord of the vineyard did not argue that equal shares were necessarily fair shares, but that an exercise of generosity on his part could give rise to no complaint on the score of justice, so long as nobody was done out of his contracted due. Hence justice is not the same as equality. Sometimes of course it requires equal treatment, namely when the cases are alike. But not all cases are alike, and it is as unfair to ignore relevant differences as to differentiate between cases that are essentially the same. (It is for this reason that justice is affronted when "strict liability" is enacted and people are held responsible and punished for "offences" over which they had no control.) If, when distributing among the members of an association the benefits that arise from their joint activity, we take no account of the part each person played, we are undercutting the concept of responsibility, and denying the importance of' each man's effort and personal contribution to the success of the enterprise. In the limiting case, where only one man is involved, we reckon that he is responsible for what he does, and should generally face the consequences, if they are ill, and, if they are good, enjoy the fruits of his labours. So, too, when people are collaborating, we should in general have regard to what they do. 'I'here is thus a presumption in favour of desert as a base for distribution, even though there are many contexts in which other bases, quite properly, predominate.

Justice is limited. As we have seen, it is not the whole of rationality, nor is it, as Plato maintained, the whole of virtue. There are other political considerations, such as expediency or utility, which may argue in a different direction from justice, and there are other political values, such as liberty, for the sake of which justice should sometimes be compromised. To aim for absolute justice would mean denying everyone any liberty of action - for fear that it might be exercised unjustly - and, in denying liberty, would be doing people down, and would therefore be unjust. Not every issue is, or ought to be, justiciable, not every adverse decision can give rise to even a prima facie charge of injustice. instead of seeking to secure justice everywhere, we should concentrate on avoiding the major injustices, ensuring that nobody is done down needlessly on the things that matter. Moreover, justice, although concerned with the individual, is concerned only in a somewhat external, impersonal way, which distinguishes it from personal morality. There is a certain lack of intimacy in justice, a certain veil of privacy. In treating you justly, I do not enter into all your concerns, and should not seek to take into consideration all the secrets of your heart. If I spoke French or German, I should not tutoyer or dutzen you, but address you only as vous or Sie. This is why justice seems a cold virtue, and fails to be fully fraternal. Although it may require us to imagine ourselves in the other chap's shoes, it does not demand that we should get inside his skin. This also is why justice requires us to exclude not only extraneous considerations but also too intimate ones. Men should be punished for what they have done, not what they think or have in mind to do, and ought not be required to make utterly frank and embarrassing disclosures. We seem to have an "annular" arrangement of relevant considerations for a judicial decision; in saying that justice deals with cases, not persons, we say down two limits, and exclude some considerations as lying outside the one and others as being inside the other.

Justice has often been criticized on this score, especially within the Christian tradition. Judaism had made too much of justice and the law, and we in consequence are very much aware that justice is not the whole of virtue, and that generosity, compassion, sympathy and love are far more important. Justice, because it excludes the innermost counsels of the heart, and address the other as vous, or Sie rather than tu, or du, seems legalistic, cold and impersonal, not spontaneous, warm and friendly. It lacks authenticity and the personal touch that can save a man and make him come alive. Nevertheless, it is a virtue, for it is the bond of society. I can be happy to be one of We, if We are just, because then We will treat Me as well as reasonably possible; and We will be happy to have Me as one of Us, because We know that I, being just, will see things from our point of view, and will not exclude wider considerations from My assessment of the situation, and will not construe everything in terms of My own exclusive self-interest. I can be sure that We will do well by Me, and We can count on My behaving as a member of the community should. There is no sta`sij (stasis), no dissension, in the just society, because it does not matter whether decisions are taken by me or by someone else - they will not conflict, because they are taken in the same frame of mind, whoever it is that takes them - and yet this absence of conflict is achieved not by some self-abnegation, some absorption in a higher whole, but by the acceptance on the part: of each of the existence and legitimate interests of everybody else. If I am treating you justly, it will be all the same whether I am deciding what I should do to you, or you are deciding it, or somebody else is. It is not biassed my way, so my decision will be the same as that of a third party; and above all, it is not biassed against you, and so the decision of a third party will be no less favourable than what you would yourself decide, if you were deciding impartially and not unfairly in your own favour..And so you and I and all of us can live together in harmony and peace, each easily identifying with everyone else, because we all recognise the individuality of each, and respect his interests, and will cherish his interests as he would himself. And thus justice, although a cold virtue, is a great one; for it is what enables the individual to identify with society, and brethren to dwell together in unity.


1. City of God, IV, iv

2. David Hume,.An Enquiry Concerning the Principles of Morals, ed. L. A. Selby-Bigge, Oxford, S 165. p. 206

3. Nicomachean Ethics, V, 1, 5-12; 1129a 17-1129b 12.

4. Nicomachean Ethics, V, 1, 1, 1129a 3-5 and V, 3, I., 1131a 10

5. op. cit. 1136b 6 and 1138a 12

6. Republic IV, 440c.



Return to bibliography

Return to home page