Reading: Rawls and Nozick.

Some of the questions here are familiar from the dispute between Rawls and Nozick. However, the issue is much wider, and of course much more ancient, and evokes the beginnings of political philosophy in Plato’s Republic. For many thinkers justice is the root concept of political philosophy, and the pursuit of justice the final goal of political life. Whether this is so depends on the meaning of ‘justice’, and in particular on whether that term’s primary application is to a distribution of goods or to the procedures involved in arriving at that distribution. (Distributive versus procedural justice.)

The emphasis of the discussion has shifted since the decline of socialism as an idea, and its effective collapse as a reality. People are no longer confident that there is a way of planning society so as to secure some desirable distribution of goods among its members; and they are more aware than they were of the negative side-effects of planning – in terms of the loss of freedom, the erosion of responsibility, and the rewarding of fecklessness. On the other hand, there is still an important streak of post-socialist (‘social market’) thinking, especially in Europe, which argues that, while we must accept the market mechanism as an indispensable epistemological tool, it stands in need of rectification, not only because it is inadequate to convey all the information needed for economic decisions (John O’Neill, The Market; Ethics, Knowledge and Politics, 1998), but also because it fails to answer our intuitions concerning desert , intuitions which are vital to the experience of society (David Miller, Principles of Social Justice, 2001). Through the concept of ‘social justice’, the old socialist concerns have been kept alive, and given a new impetus in the criticism of distribution patterns in modern economies. Moreover, debates over the validity and extent of the welfare state are rooted in the concept of justice, and – in the arena of real politics – the old socialist ideal of equality of outcomes is still invoked, in the name of justice, by liberals, and decried, in the name of justice, by conservatives. To disentangle these debates is, not surprisingly, a strenuous and also delicate task. There is a lot of resentment out there, and in the vicinity of resentment it is hard to keep cool.

First, for reference only, my dictionary entry on justice:

justice. *Plato’s Republic begins with the question, ‘What is justice?’, and with the famous refutation of the view that justice is the *interest of those with *power: in other words, it argues for the non-identity of rights and powers. (For the persistence of the doctrine refuted by Plato, see *rights and powers.) Since then, Plato’s question has lain at the heart of all moral, political and legal philosophy, and is often con­sidered to be the single most important question in political thought. *Aristotle, who thought of justice as the true subject-matter of political philosophy, and of its execution as a major purpose, perhaps the major purpose, of the * polis (a view also expressed by Aeschylus in the Oresteia), made a famous distinction between *dis­tributive and commutative justice, the first being concerned with the distribution of goods among a class, the second with the treatment of the individual in particular transactions (e.g. when punishing him). In the second case justice involves giving to someone what he deserves, or else what he has a *right to receive. In the first case it is, according to Aristotle, a matter of ‘treating equals equally’. These are not necessarily the same idea (although it could be said that, in the abstract, equals do have a right to be, and deserve to be, treated equally). If I have to distribute a cake among five starving people, then they may each deserve the whole cake, but it may be unjust not to divide it.

The Aristotelian distinction survives in various forms, although it would now be very unusual for a philosopher to think that there are really two concepts of justice rather than two applications of a single idea. Contemporary use of the term *social justice makes extensive use of the distribu­tive conception, while, in regulating actions between people, the commutative conception – justice as right or desert – seems to be prevalent. Many of the problems arise because the two may enter into conflict, as when Robin Hood acts unjustly (by taking what he has no right to take) in order to bring about social justice (through *redistribution). Is this conflict real, or would deeper reflection resolve it? Theories divide over their answer to that question. Those who take patterns of distribution as their model (among whom the most influential recent theorist is *Rawls) might find themselves resolving this issue only to admit that Robin Hood acts justly (provided perhaps that he dignifies himself with the title of *sover­eign and redistributes in the name of the *state); those who regard the respect for individual rights as the central idea in justice (such as *Nozick) might resolve the issue only to find themselves condoning distributions of goods so unequal as to be very oddly described as just. The import­ant questions seem to be these:

(i) Is justice primarily the attribute of an act, of a person, or of a state of affairs? If of an act, then states of affairs (e.g. ‘distributions’ of goods) are just or unjust only to the extent that they are the outcome of just or unjust acts. If of a person, then rules of just action must be subordinate to a conception of human *virtue (e.g. an action is just if it would be recommended by the ‘impartial judge’ or the ‘just person’) – a conception which vastly complicates the discussion, and probably undercuts the desire of many modern philosophers to establish simple laws of justice. If of a state of affairs, then we might consider it unjust that one person is born more advantaged than another, even though human agency has nothing to do with it. Aristotle seems to have favoured the second idea; much dispute between socialists and liberals in modern politics comes from socialists being attached to the third of those ideas, and liberals (*Hayek, for example) to the first.

(ii) Is justice a forward-looking or a backward-looking conception? In other words, does it look to the results of an action, or to its antecedents? The connec­tion with desert and right suggests the latter, and this is at the root of the conflict – which many philosophers have thought to exist – between justice and *utility, itself cited as one of the principal failings of *utilitarianism. But some theories of social justice seem to construe justice as a forward-looking conception, nevertheless.

(iii) Is justice a constitutive or a procedural concept? That is, is the differ­ence between the just and the unjust to be found in the nature of the action, or in the manner of its execution? Study of the law has sometimes led to the opinion that laws themselves are never either just or unjust, but that the process of their application may be one or the other, since it is only when legal rights have been determined that we can speak of ‘treating equals equally’.

(iv) What is it, to treat equals equally? What is the relevant respect in which people are to be compared? If there is no respect, then the result of this maxim is *egalitarianism; if the respect is that of legal rights, and there is no natural limit on those (but only the limits embodied in *positive law), then rights may be unequally determined, and a slave may be justly treated as a slave.

(v) Is there, in other words, *natural justice? And are there *natural rights?

(vi) If so, does that solve the question, Why be just? See *value, *virtue.

In all those questions the problem of the objectivity of the concept of justice is paramount. There is a further problem suggested by them: the nature of the ‘sentiment of justice’ (cf. *Hume). Is there a real fact of human nature here, or could that sentiment be educated away? It always seems as though political systems that override the sense of justice thereby render themselves precarious. Is this because that sense is at the root of sentiments of *allegiance? Or is there a genuine conception of *political obliga­tion that requires justice to be preserved? (Cf. *piety.)


Many of the disputes over justice concern the idea of equality. Is this fundamental to justice, and if so equality in what respect? The sense of a connection between the two ideas is very ancient (see Aristotle, as reported above), and is enshrined in the use of the term ‘equity’ in law, to denote certain procedures designed to do justice in the particular case, when the law itself seems harsh or oppressive. But there is no doubt that liberals, in the American sense, have very different ideas about equality from those entertained by conservatives. The idea that we are all ‘equal before the law’ has its origin in the Roman concept of citizenship – though non-citizens were not included in this equality, and women and children had only a distorted form of it. Nevertheless, the idea of equality before the law has been taken by many conservatives as their model of equality in all its desirable forms – all the forms that it might be right and feasible to pursue at the political level. And this equality can be spelled out in terms of rights and duties: we have the same rights in a court of law, and fall under the same laws, which we must obey in the same way. This is taken for granted in modern systems of law (though until recently, and perhaps even now in special cases, Britain retains some vestige of the old feudal doctrine, of a right to ‘trial by one’s peers’. This may have been functional once, but is so no longer).

Equality before the law can be generalized, to become ‘equality before the judge’. In other words, whenever my fortune depends upon another’s judgement, I should be given the same treatment as anyone else in my situation – the same rights, the same duties, the same opportunities. What the outcome will be depends on other circumstances, and there is no reason to think that equal treatment at every point where judgement might affect the outcome will produce equal outcomes. Obviously it will not.

Problems here: are we speaking always of judgement in the same sense? Perhaps not. It is only in certain circumstances that judgement has the impartiality of law. What about judging people for a job? You, the employer, have a serious interest in the outcome; you cannot regard this as a purely judicial affair. Nevertheless, there is a distinction between judging people on their merits (relevant to the job), and judging them on features that are irrelevant to their ability to do the job – sex, race and so on. So even in this case, where there is no question of overall impartiality, there is the equivalent of impartiality – in other words, being guided only by considerations relevant to the decision to be made.

Some conservatives (Hayek for example) think that equality before the law, and whatever equalities can be generalised from it, are the only equalities that are politically pursuable. They are pursuable because they can be directly acted upon by all of us, simply by treating others impartially whenever judgement is required, and whenever their fortune depends upon a judgement. However, there are other equalities and inequalities that are exemplifed among people, and which could perhaps be affected by political action. People are unequal in all the following respects:

Wealth, looks, health, strength, intelligence, moral worth, will power, attachment to life, social advantages, imagination, talent, social power.

Some of those inequalities can be rectified, at least in part, by political action. And conservatives differ radically from liberals over the question whether we should try to rectify them. Moreover, there is a great difference between rectifying inequalities by reducing the assets of those who have most of some good, and rectifying them by increasing the assets of those who have least. (You can abolish material inequality tomorrow, by confiscating everything from everybody: the Maoist method. Most liberals would not recommend such a policy.)

The position taken by most conservatives is that there are certain natural inequalities – of looks, strength, intellect, energy and talent – that we interfere with at our peril. (The peril of Brave New World.) It is not the business of politics to rectify these even if it could be done. And the onus is on the critic to say what is wrong with them: it is surely not unjust that I don’t have ‘this man’s art or that man’s scope’. If we allow these inequalities, then equality before judgement will produce unequal distributions of every other good, including those that exist only as a result of social cooperation, such as wealth and social power. In response to socialist and liberal complaints that the resulting inequalities are intolerable, one response is to ask ‘why?’. If people have enough, what is wrong with the fact that some have more?

Answers to that question may be sociological: inequality breeds resentment, which breeds social unrest, and so on. (But does inequality breeds resentment as much as an enforced equality – i.e., an equality which depends upon confiscating the goods of the enterprising and the successful? The empirical record here is far from clear.) All such sociological responses are of no real significance from the philosophical point of view. And conservatives tend to respond to them in the following way: resentment comes when people compare their fate with that of others. But when people belong to different spheres, and find fulfilment in the sphere that is theirs, these comparisons, and their invidious consequences, are minimised. The ‘social question’ is not about providing equality – which is both impossible and not a cure; it is about defusing resentment by generating a diversity of rewards. This area of social psychology was not unfamiliar to Aristotle; but it of course came to the fore in recent times with Nietzsche’s theory of ressentiment, and the response to that and similar theories from Max Scheler (Ressentiment) and Helmut Schoeck (Envy).

Sociological answers are of less significance for us than answers in terms of justice. Egalitarians may answer the question given above by saying that inequalities of the kind described are unjust, or prima facie unjust. The latter position (slightly more subtle) attempts to establish an onus of proof. Inequality is unjust until justified: it is for you to show how this inequality does not offend against the laws of justice. To which the conservative response is: not so, it is for you to show how it is unjust. The onus of proof lies on the egalitarian, since I (the conservative) have assumed no unnatural processes, no coercion, no wrongdoing, in arriving at the conclusion that social inequality is the inevitable result of natural inequality. It is certainly true that the question ‘What is wrong with inequality?’ is not sufficiently addressed by egalitarians. Rawls shocked the left-liberal consensus by espousing the Difference Principle, which allows (socially derived) inequalities when these benefit the worse off. But conservatives are even more shocking, in allowing that it is an argument in favour of inequalities that they benefit the better off . Of course, not a conclusive argument, and one that is easily upset if the condition of the worse off is made intolerable. But an argument all the same. Until we have developed a satisfactory concept of justice this whole issue remains pretty well undecidable.

A problem is created by Aristotle’s definition of justice as ‘treating equals equally’. This looks like a kind of egalitrianism; but it should be read in the context of a general view of society, according to which people are not equal in any respect relevant to their social advancement. To treat equals equally will leave the structural inequalities in place.

Social Justice.

The debate over social justice has roots in the 19th century, and centres on the following question: are we entitled, and ought we, to consider the social product (i.e. the sum total of goods, services and advantages produced in a society) as produced by the cooperation of all members of society, so that each is entitled to a share of it? And if so, what is the just distribution among the membership? Advocates of ‘social justice’ as an idea say yes to the first of those questions; but not all of them are socialists, and not all of them would put forward an egalitarian answer to the second. (Rawls, for instance, answers yes to the first question, while being loosely inegalitarian in answer to the second.)

The conservative response to the social justice idea is complex. Some just deny the major premise. The social product, they argue, is not the product of society, but the by-product of countless individual transactions, each of which can be adjudicated from the point of view of justice, but the totality of which – because unintended, and unagreed – cannot be considered in such a way. That would be the Hayekian view. Entitlements, on this view, depend upon the histories that create them: and each person’s history is different.

Others argue that the idea of social justice is trying to import a distributive idea of justice into a situation where considerations of justice, if they arise, are procedural. This is an argument that we can take forward, using Nozick’s idea of a ‘justice-preserving transfer’, which he likens to a ‘truth-preserving inference’. If property is justly held in position A, and is then transferred to position B by a ‘justice-preserving transfer’, it is justly held in position B. This argument can be refined, by arguing that the concept of justice is inherently procedural, so that questions of ‘original acquisition’ are outside its reach. All we know is whether advantages, of whatever kind, have been justly transferred.

It is clear, as Aristotle pointed out, that we raise questions of justice both when there is a good to be distributed among those with a claim to it, and where there is some action or course of actions which demand some kind of response, such as punishment or reward. In both cases questions of desert may arise; but they arise differently. In distributions we may tie desert closely to need, rather than to contribution. This is the socialist move that conservatives most object to, since it licences massive state intrusion into property rights, and so makes ‘social justice’ into an instrument of in justice towards individuals.

Is the major premise coherent? If people enter an agreement to do something, on condition that the profits are shared, then it is clear that relations of desert arise and must be satisfied if there is to be justice. But when people exist together in a society of strangers, making variable contributions to an economic order whose overall product is the unintended consequence of agreements concerning something else, it is not at all clear that any uniform relations of desert arise, in relation to that product. And if they do arise, then justice would surely require that we take each person’s contribution into account. How do you measure that contribution?

If we are to allow entitlements to be defined across a whole society, then what is the criterion of membership? The social product of the USA derives from economic activity with inputs in many other countries. Ought their citizens to be included in the distribution? Of course some say yes, here: but the important point is that theories of ‘society’ which make the ‘social product’ into a manageable idea are really theories of the state: societies are simply states in their social aspect, and there is, built into the idea of social justice, an implied statism which tends of its own accord in the socialist and ‘planned economy’ direction.

This last point raises questions about ‘global justice’. Who is best able to deal with these question: the conservative who believes that local attachments take precedence over universal principles, or the liberal, whose universalist philosophy makes no room for local attachments?

Rawls and Nozick.

This is a familiar controversy, and I have put the essential fragments on the blackboard. The reason for revisiting it, is that it shows standard moves in the liberal-conservative dispute. The moves go as follows:

Liberal: the good society is a voluntary association, which we choose by contracting together, or would choose in imaginary circumstances which undo those aspects of history that constrain our free choice. What we would choose in those circumstances defines the obligations between us. These obligations would be obligations of justice – concerning what we owe to each other, and therefore what we each are owed.

Conservative: the good society is a voluntary association, but only up to a point. Custom and tradition tie us to our context and curtail our choice. What makes a society good is not that we would choose it in some hypothetical context in which the burdens of history have been discounted. What makes it good is that we can exercise our freedom, realize our potential, within a settled framework that sustains agreements and creates relations of trust. That framework is what history provides. Our obligations to others arise in two ways – obligations of piety to members of the family and others bound to us by unchosen ties; obligations of justice to those to whom we are joined by agreement and contract. We don’t have obligations of justice to those strangers with whom we have never contracted. Obligations of charity maybe: but those are like gifts, and it would be unjust to compel them.

Liberal: your well-being depends upon others with whom you did not contract directly, but whose labour is essential to maintain the social order without which your contracts would be empty and unenforceable. You can understand the extent of their claim on you only by some thought experiment which establishes what ideal social arrangement you would choose by way of including them. Their claims, so derived, will be claims of justice, since they will be based on an understanding of what is owed to them.

Conservative: that is to take the concept of justice out of the only context in which we can understand it, which is the context provided by actual relations between individuals, in which conceptions of desert and right emerge from human agreements and from the use and abuse of our fellows.

Liberal: OK, so don’t call it justice. Maybe you should call it gratitude.

In addition to such endless arguments, we have the very important argument of Nozick’s that ‘liberty upsets patterns’ – the example of Wilt Chamberlain, and the questions that example raises concerning ‘just original acquisition’. This issue is quite independent of the details of the Rawlsian theory – the original position, lexical ordering of principles, the effect of the Difference Principle, and so on – though all these can be questioned from the conservative viewpoint, along with the view that we are to discount our several ‘conceptions of the good’ in making our choice in the original position. The crucial question remains that of the ‘pattern’ theory of justice, and the potential conflict between that and liberty. A liberal could admit that liberty upsets patterns, and respond that this simply means that we have two values: justice and liberty – and that they are in potential conflict. More serious is the conservative response that there are not two values here, but one, and that is justice. If we tie justice to patterns we authorize injustice in the individual case. But since we derive our understanding of the concept from the individual case this is an argument for refusing to attach the concept to a pattern. (Wittgensteinian arguments about language-learning are relevant here.)

(There are deep questions here abut the motivating force of justice. Aristotle’s view of justice, as a virtue – is this an attempt to give a theory of the motivation that makes justice into a real issue for all of us?)

Desert, right, and need.

Much of the conservative argument about justice concerns the supposed illegitimacy of taking the concept of justice out of the context that provides its sense. What is this context? The establishment of relations of trust between individuals, the upholding of promises and contracts, and the punishment of wrongdoing. Egalitarians have a rooted tendency to apply the concept in a context where it is no longer clearly guided by its own internal principles. By adopting the idea of social justice they convert charitable gifts into rights of those who receive them, and sets up patterns of distribution that seem to have little connection to the deserts of the recipients. We have an intuitive understanding of right, desert, and related concepts from our ordinary transactions with each other; but when these concepts are transferred to the context of society as a whole, in order to denote obligations established by no specific relationships, and in order to justify claims based on need alone, they float free of their logical base. If, for example, welfare benefits are really the expression of a claim of justice, then they are a right in the recipient, and a non-defeasible claim against the common fund. Many conservatives argue that, if we see them in this way we do not merely undermine the motive to provide charity, we also produce relations of dependency that become institutionalised, to the detriment both of those who receive and of those who provide.

There are arguments on many levels here, and it is important to distinguish those which are arguments about justice from those of a more utilitarian kind. We might criticize the ‘dependency culture’ on utilitarian grounds, while nevertheless maintaining that need is sufficient (given a few assumptions about society) to establish a claim of justice and therefore of right. Or we might try to argue that the dependency culture is itself the direct result of transferring concepts of right to a context where the individual relations do not exist that would sustain its application. (This takes us back to the discussion of human rights last week.)

What emerges clearly from this is the need, for the cogency of the conservative position, for a theory of desert, and for a theory as to why a claim of justice provides a reason for action. These are tough demands. But the conservative response is to say that the demands on the liberal are even tougher: to give a theory of desert that will cover all members of society, regardless of individual histories, and also to show why a claim of social justice provides a reason for action. Against whom is the claim to be made? And whence comes the obligation to comply with it? Not from a contract that I never explicitly made, surely?

Return to the Topics
Read Culture