Two Faces of Liberalism

Two Faces of Liberalism

By JOHN GRAY
The New Press

Read the Review


Liberal Toleration



The liberal state originated in a search for modus vivendi. Contemporary liberal regimes are late flowerings of a project of toleration that began in Europe in the sixteenth century. The task we inherit is refashioning liberal toleration so that it can guide the pursuit of modus vivendi in a more plural world.

Liberal toleration has contributed immeasurably to human well-being. Nowhere so deep-rooted that it can be taken for granted, it is an achievement that cannot be valued too highly. We cannot do without that early modern ideal; but it cannot be our guide in late modern circumstances. For the ideal of toleration we have inherited embodies two incompatible philosophies. Viewed from one side, liberal toleration is the ideal of a rational consensus on the best way of life. From the other, it is the belief that human beings can flourish in many ways of life.

If liberalism has a future, it is in giving up the search for a rational consensus on the best way of life. As a consequence of mass migration, new technologies of communication and continued cultural experimentation, nearly all societies today contain several ways of life, with many people belonging to more than one. The liberal ideal of toleration which looks to a rational consensus on the best way of life was born in societies divided on the claims of a single way of life. It cannot show us is far from being peculiarly modern. On the contrary, in their diversity of ways of life late modern societies have something in common with the ancient world. What is new in the modern world is not acceptance of diversity in styles of life. It is hostility to hierarchies.

The cultures from which European moral philosophy emerged contained many forms of ethical life. Greek polytheism expressed the belief that the sources of value are irreducibly plural. If it recognized the idea of the best human life, it was one in which many distinct and at times conflicting sources of value were honoured. In their acceptance of many sources of value, the Greeks were at one with other ancient cultures: ancient Judaism imposed few universal obligations; Hinduism recognized different duties in different stations and stages of life.

Ancient societies were more hospitable to differences than ours. This is partly because the idea of human equality was weak or absent. Modernity begins not with the recognition of difference but with a demand for uniformity. There is nothing new in the idea that the good life may vary with different people. To think that it is distinctively modern is a mere prejudice.

Ancient pluralism found few echoes in Greek philosophy. The founders of European ethical theory were monists. Neither Plato nor Aristotle was in any doubt that one way of life was best for humankind. Whether the good for humans was finally one, as Plato imagined, or many, as Aristotle was sometimes ready to admit, the best kind of life was the same for everyone — even though they never doubted that it could be lived fully only by a few leisured Greek males. In this classical view, conflicting judgements about the human good are symptoms of error. For the founders of European ethical theory, as for the Christians who came after them, conflicts of value were signs of imperfection, not a normal part of ethical life.

From its beginnings, moral philosophy has been a struggle to exorcize conflict from ethical life. The same is true of political thought. European political philosophy has been deeply marked by the resistance to conflict that shaped Greek ethics. In the city, as in the soul, harmony has been the ideal. Most liberal thinkers have taken over the Socratic, Christian and Enlightenment faith in the harmony of values. But an ideal of harmony is not the best starting-point for thinking about ethics or government. It is better to begin by understanding why conflict — in the city as in the soul — cannot be avoided.

In the form that we have inherited it, liberal toleration is an ideal of rational consensus. As heirs to that project, we need an ideal based not on a rational consensus on the best way of life, nor on reasonable disagreement about it, but instead on the truth that humans will always have reason to live differently. Modus vivendi is such an ideal. It embodies an older current of liberal thought about toleration, and applies it to our own new circumstances.

Modus vivendi expresses the belief that there are many forms of life in which humans can thrive. Among these there are some whose worth cannot be compared. Where such ways of life are rivals, there is no one of them that is best. People who belong to different ways of life need have no disagreement. They may simply be different.

Whereas our inherited conception of toleration presupposes that one way of life is best for all of humankind, modus vivendi accepts that there are many forms of life, some of them no doubt yet to be contrived, in which humans can flourish. For the predominant ideal of liberal toleration, the best life may be unattainable, but it is the same for all. From a standpoint of modus vivendi, no kind of life can be the best for everyone. The human good is too diverse to be realized in any life. Our inherited ideal of toleration accepts with regret the fact that there are many ways of life. If we adopt modus vivendi as our ideal we will welcome it.

Ethical inquiry does not yield a single way of life or scheme of values for all — not even for a single individual. Instead it shows that people have reason to live in different ways. Different ways of life embody incompatible aspects of the human good. So, in different contexts, may a single human life. Yet no life can reconcile fully the rival values that the human good contains.

The aim of modus vivendi cannot be to still the conflict of values. It is to reconcile individuals and ways of life honouring conflicting values to a life in common. We do not need common values in order to live together in peace. We need common institutions in which many forms of life can coexist.

The span of good lives of which humans are capable cannot be contained in any one community or tradition. The good for humans is too beset by conflict for that to be possible. For the same reason, the good life cannot be contained in any one political regime. A theory of modus vivendi is not the search for an ideal regime, liberal or otherwise. It has no truck with the notion of an ideal regime. It aims to find terms on which different ways of life can live well together.

Modus vivendi is liberal toleration adapted to the historical fact of pluralism. The ethical theory underpinning modus vivendi is value-pluralism. The most fundamental value-pluralist claim is that there are many conflicting kinds of human flourishing, some of which cannot be compared in value. Among the many kinds of good lives that humans can live there are some that are neither better nor worse than one another, nor the same in worth, but incommensurably — that is to say, differently — valuable. Even so, there may be good reasons for preferring some incommensurable goods over others.

Value-pluralism is closer to ethical theories which affirm the possibility of moral knowledge than it is to familiar kinds of ethical scepticism, subjectivism or relativism. It enables us to reject some judgements about the good as being in error. At the same time, it means giving up a traditional notion of truth in ethics. To affirm that the good is plural is to allow that it harbours conflicts for which there is no one solution that is right. It is not that there can be no right solution in such conflicts. Rather, there are many.

The good is independent of our perspectives on it, but it is not the same for all. It is not just that different ways of life honour different goods and virtues. More, what one way of life praises another condemns. Value-pluralism is the claim that both may be right. This claim is paradoxical. It seems to imply a tolerance of contradiction that classical logic prohibits. There is paradox here, but not — or so I shall argue — of the kind that should concern us. It may be that the good cannot contain contradictions; but it shows itself in ways of life that are incompatible.

Conflicts of value need not express any uncertainty, practical or intellectual, about what is good. At their starkest, they exclude any such uncertainty. They are conflicts within the good itself. However variously they may be understood, peace and justice are universal goods; but sometimes they make demands that are incompatible. When peace and justice are rivals, which is worse, war or injustice? Neither has automatic or universal priority. Peace may be more urgent than justice; the claims of justice may override the immediate needs of peace. In conflicts of this kind, people need not differ about the content of the good or the right. Where they differ is on how their rival claims are to be reconciled.

Justice does not speak always with one voice. The communities that are locked in conflict in Israel and Ulster may claim that they invoke the same principles of justice. Yet their judgements of what is just and what unjust in the context of their contemporary conflicts are deeply at odds. In part, this reflects their different interpretations of their shared history. Partly, no doubt, it is also an expression of the fact that their interests are in many ways opposed. When communities contend for power over scarce resources, they are likely to seek to justify their rival interests by arguments of fairness. Where interests are at odds and political power is at stake, shared principles of justice are likely to yield incompatible judgements of what justice demands.

But conflicts over what justice demands do not come only from these familiar facts. Justice itself makes incompatible demands. When justice requires that restitution be made for injustice done to communities in the past, the result may be unjust to present generations. A claim for the return of land that was unjustly expropriated may collide with a no less just claim to the land that is based on generations of working it. Such conflicts do not arise from an imperfect sense of justice. They express the truth that justice itself encompasses conflicting values.

Even if a conception of justice could be formulated that received universal assent, it would make conflicting demands about which reasonable people could differ. Once again, this is not because human reason is imperfect. It is because incompatible solutions of such conflicts can be equally reasonable.

That conflicts between universal values can be settled in incompatible ways is one reason why people belong to different ways of life. The many ways in which humans can live well embody different settlements among discordant universal values. Contrary to the liberal ideal of toleration, the fact of divergent ways of life is not a result of the frailty of reason. It embodies the truth that humans have reason to live differently.

At the same time, some conflicts of value do arise from rival views of the good. They come not from rivalry among values that are universal but from the different goods that are honoured in particular ways of life. Some goods that are central in some ways of life are absent, or else marginal, in others. In late modern societies, personal autonomy and romantic love are highly valued; but these rival goods are far from being valued by everyone. Today, as in the past, there are ways of life that do not celebrate them, or which condemn them.

To be caught between the demands of different ways of life is a common source of moral conflict. Many people face conflicts among values for which there is no single right solution. The fact that ways of life honour different goods and virtues is not a mark of imperfection. It is a sign that humans can live well in different ways.

Yet not all ways of life allow humans to live well. There are universal human goods and evils. Some virtues are needed for any kind of human flourishing. Without courage and prudence no life can go well. Without sympathy for the suffering and happiness of others, the artefact of justice cannot be maintained. Forms of life that are deficient in these virtues are lacking in the conditions of human well-being. Such values are generically human. Because they are universal they can be used to assess any particular way of life.

That some values are incommensurable does not mean that all ways of life have the same value. The bottom line for value-pluralism is the diversity of goods and evils, not of ways of life. Different ways of life can be more or less successful in achieving universal goods, mitigating universal evils and in resolving conflicts among them.

Even so, universal values do not fit together to compose an ideal life — for the species, for particular societies or for individuals. Rather, if universal values can be rivals, there can be no such thing as an ideal life. There may be a best life for any individual; but not one that is without loss. In particular ways of life there may be better or worse solutions to conflicts of value; but none that meets fully every legitimate claim. There are better and worse regimes, and some that are thoroughly illegitimate; but none that fully realizes all universal values, and is thereby a model for all the rest.

The most fundamental differences amongst ways of life arise from the manner in which they deal with conflicts among values that are universal. Universal values enable us to assess particular ways of life; but they do not add up to a universal morality.

In the world as we find it, even the barest requirements of a life worth living cannot all be always met in full. Toppling a tyranny may trigger civil war. Protecting a broad range of liberal freedoms may result in the regime that guarantees them being short-lived. At the same time, supporting a strong state as a bulwark against anarchy may worsen the abuse of power. Wise policy can temper these conflicts. It cannot hope to overcome them.

Conflicts of value go with being human. The reason is not that human beings have rival beliefs about the good life. Nor is it — though this comes closer to the nub of the matter — that the right action sometimes has wrong as its shadow. It is that human needs make conflicting demands. The idea of a human life that is without conflicts of value runs aground on the contradictions of human needs.

It is not only that, because they make incompatible demands on scarce material resources, human needs may be practically at odds. More, they can be met fully only in forms of life that cannot be combined. The lives of a professional soldier and a carer in a leprosarium, of a day trader on the stock market and a contemplative in a monastery, cannot be mixed without loss. Such lives embody virtues that do not easily coexist; and they may express beliefs that are contradictory. Yet each answers to a human need.

The best human lives are very different from one another, and often incompatible. This is not a truth of logic. It is a fact about human nature. As such it is not unalterable. Perhaps, as technologies of genetic engineering advance, human beings will be tempted to alter the biological endowments that have enabled them to live in so many different ways. There is nothing to say such attempts cannot succeed; but if they do they will destroy much that has hitherto been of value in human life.

Conflicts of value come from the competing needs of our common human nature. A kind of moral scarcity is built into the fabric of human life. It is because human needs are contradictory that no human life can be perfect. That does not mean that human life is imperfect. It means that the idea of perfection has no meaning. The idea of conflicting and incommensurable values is far from the Augustinian notion that all things human are imperfect. Augustine contrasted the imperfection of the human world with the perfection of the divine. By contrast, rivalry between incommensurable values destroys the very idea of perfection.

The fact that good harbours conflicts of value does not mean that the human condition must always be tragic. To be sure, tragic choices cannot be eliminated from ethical life. Where universal values make conflicting demands, the right action may contain wrong. When values clash in this way, there may be irreparable loss. Then there is surely tragedy.

But the plurality of values signifies more than simply tragedy. It means that there are many kinds of life in which humans can thrive. Where these lives are so different from one another that their worth cannot be compared, it makes little sense to speak of gain or loss. When such lives cannot be combined, they need not be antagonists; they may be alternatives. If we choose among them, as sometimes we must, the choice need not be tragic. It may simply bespeak the abundance of flourishing lives that is open to us.

If this is true, it has always been so. Value-pluralism is an account of ethical life, not an intepretation of pluralism in late modern societies. If it is true, it is a truth about human nature, not the contemporary condition. Nevertheless, value-pluralism has a special application to late modern societies.

In nearly all contemporary societies the coexistence of many ways of life is an established fact. Though distinct, these ways of life are not independent. They interact continuously — so much so that it may be hard to tell the difference between them. Indeed, since many people belong to more than one, it may be impossible to distinguish them completely. Ways of life are tricky things to get to know. They do not come ready labelled. There is no sure-fire method of enumerating them. And they come in many varieties.

There is the way of life of religious fundamentalists and secular liberals, of countryfolk and `young urban professionals', of Taliban and Quakers, of first-generation immigrants and that of their children, of Homer's warrior-class, the Desert Fathers and twenty-first-century Hasids, and indefinitely many more. It is impossible to specify the necessary and sufficient conditions that must be met for a style of human activity to qualify as a way of life. Nor is it necessary. We can distinguish them by a loose bundle of criteria.

Ways of life must be practised by a number of people, not only one, span the generations, have a sense of themselves and be recognized by others, exclude some people and have some distinctive practices, beliefs and values, and so forth. Often these criteria do not yield a clear result. Two communities may honour many of the same values but be locked in an historical conflict. We might say of them that they have the same way of life but are divided in their allegiances to the regime under which they live. (Think of Ulster.) Or two communities may have distinctive and opposed beliefs about the historical sources of their present conflicts, contrasting attitudes to a number of social issues, and a strong propensity to exclude one another (by avoiding intermarriage, for example). Then we might be inclined to say they have conflicting ways of life. (Think again of Ulster.) What counts as a way of life may not always be decidable.

When the standard types of contemporary liberal thought refer to pluralism they mean the diversity of personal ethical beliefs and ideals. That is not the kind of pluralism that should most concern political philosophy. Late modern societies are notable for the diversity of ways of life they contain. Immigration and the partial erosion of the cohesive national cultures that were constructed earlier in the modern period have increased the number of ethnic and cultural traditions that coexist in the same societies. At the same time, continuing cultural experimentation has produced a number of new styles of life. This fact of pluralism was not foreseen in liberal thought. Even now it has not been fully comprehended.

The conflicts of value that rightly shape the agenda of political thought come not from the divergent ideals of individuals but from the rival claims of ways of life. Recent liberal orthodoxy passes over these conflicts because it takes for granted that one way of life is dominant in society. In contrast, value-pluralism has particular relevance to late modern societies in which, by choice, chance or fate, many ways of life have come to coexist.

Liberal thought needs revision if the ideal of toleration is to be refashioned to suit this circumstance. In standard liberal accounts, pluralism refers to a diversity of personal ideals. Liberal thought rarely addresses the deeper diversity that comes when there are different ways of life in the same society and even in the lives of the same individual. Yet it is this latter sort of pluralism that should set the agenda of thought about ethics and government today.

To think of this condition as a peculiar disability of modern times is mistaken. The pot-pourri sometimes called western civilization has always contained conflicting values. Greek, Roman, Christian and Jewish traditions each contain distinctive goods and virtues that cannot be translated fully into the ethical life of the others. The notion of a `western tradition' in which these irreconcilable elements were once fused cannot withstand philosophical — or historical — scrutiny. There was never a coherent synthesis of these values, nor could there have been. Still, for many centuries, these diverse inheritances were subordinated in European societies to a single ethical ideal. With all its doctrinal variations, and the many prudent allowances it made for the intractability of human nature, the Christian ideal of life succeeded in subjugating or marginalizing others that had been part of the European inheritance for centuries or millennia. Liberalism needs to be rethought to fit a context in which different ideals of life coexist in the same societies — and often the same individuals.

In recent liberal writings, the fact of pluralism refers to a diversity of personal ideals whose place is in the realm of voluntary association. The background idea here is that of the autonomous individual selecting a particular style of life. This type of diversity resembles the diversity of ethnic cuisines that can be found in some cities. Like the choice of an ethnic restaurant, the adoption of a personal ideal occurs in private life. But the fact of pluralism is not the trivial and banal truth that individuals hold to different personal ideals. It is the coexistence of different ways of life. Conventional liberal thought contrives to misunderstand this fact, because it takes for granted a consensus on liberal values.

In reality, though there is unprecedented lip-service to them, most late modern societies contain little consensus on liberal values. Many people belong at once in a liberal form of life and in communities which do not honour liberal values. At the same time, many who stand chiefly in liberal ethical life do not subscribe to some of its traditional values. The liberal ideal of personal autonomy is the idea of being part-author of one's life. For some, the pursuit of autonomy comes into conflict with allegiance to an established community. For others, it is in tension with the freedom to respond to the needs of the present. For all these kinds of people, `traditional', `liberal' and `postmodern', ethical life is inescapably hybrid.

Most late modern societies are far from exhibiting an overlapping consensus on liberal values. Rather, the liberal discourse of rights and personal autonomy is deployed in a continuing conflict to gain and hold power by communities and ways of life having highly diverse values. Where it exists, the hegemony of liberal discourse is often skin-deep.

If it can be found anywhere, an overlapping consensus on liberal values should exist in the United States. And it is true that in the USA there is virtually no group that does not invoke liberal principles. Yet America is no different from the rest of the world in being riven by conflicts between ways of life. The quarter of the American population that espouses creationism, `the right to life' and other fundamentalist causes does not repudiate liberal values explicitly — as people with similar beliefs might do elsewhere in the world. It appropriates them for its own purposes. The strategic deployment of liberal discourse for fundamentalist ends by a large segment of the population is not a consensus on liberal values. It is practically the opposite. Like other late modern societies, the United States is not hegemonically liberal but morally pluralist.

Recent liberal thinkers claim that the appropriate response to the fact of pluralism is a `theory of justice'. The `political liberalism' of John Rawls and his followers claims to advance an account of justice that can be accepted by people who have different conceptions of the good. According to this recent orthodoxy, the liberal state is not just one among a number of regimes that can be legitimate. It is the only mode of political organization that can ever be fully legitimate. In recent liberal thought this claim is conjoined with another — that what makes a liberal state legitimate is its protection of human rights.

For Rawls, as for Ronald Dworkin, F.A. Hayek and Robert Nozick, political philosophy is a branch of the philosophy of law — the branch which concerns justice and fundamental rights. The goal of political philosophy is an ideal constitution, in principle universally applicable, which specifies a fixed framework of basic liberties and human rights. This framework sets the terms — the only terms — on which different ways of life may coexist.

These thinkers claim that conflicts among goods and ways of life can be resolved by reading off the requirements of justice or rights. But the requirements, of justice and rights can and do conflict. It is not just that the demands of one right may clash with those of another. A single right may make incompatible demands. There is no uniquely rational way of resolving these conflicts. This truth has large consequences. It means that there can be no such thing as an ideally liberal regime. Because rights make conflicting demands that can reasonably be resolved in different ways, the very idea of such a regime is a mistake.

To say that there cannot be an ideal political regime is not to mount a defence of imperfection in politics. It is to reject the idea of an ideal regime. Different regimes can rightly resolve conflicts among vital human rights in different ways. Some such settlements are better than others, but there is nothing which says that the best regimes will resolve conflicts among rights in similar ways.

On the contrary, because their circumstances and histories vary so much, the best regimes are very different from one another. (So are the worst.) Politics abounds in tragic choices. Even so, it is not because of the tragedies of politics that the idea of an ideal regime lacks sense. It is because the best regimes come in many varieties.

When we differ deeply as to the content of the good, an appeal to rights will not help us. For in that case we will differ as to which rights we have. Fundamental differences about rights express rival conceptions of the good. When rational inquiry leaves our views of the good deeply at odds, it is vain to appeal to rights. Basic human rights can be justified as giving protection against universal human evils; but even such rights clash with one another, and incompatible settlements of their conflicts can be equally legitimate. When universal evils clash, no theory of rights can tell us what to do.

It is the same with social justice. We cannot avoid judgements of fairness regarding the distribution of goods in society. The notion that fairness in procedures is all that society needs in the way of a shared conception of justice, which Hayek made familiar, has little to be said for it. Yet no contemporary society contains a consensus on fairness that is deep or wide enough to ground a `theory of justice'.

There is no more consensus on what justice means than there is on the character of the good. If anything, there is less. Amongst the virtues, justice is one of the most shaped by convention. For that reason it is among the most changeable.

When many ways of life share the same society, it is natural that the sense of justice should vary. It is therefore hardly surprising that liberal philosophers differ about the most fundamental requirements of justice. Today, most liberal thinkers affirm that justice is the supreme virtue of social institutions; but some declare that it demands equal distribution of social goods, others that it requires respect for the supposed fact that each of us owns his or her natural endowments, yet others that it involves matching resources with basic needs or merits — and still others that it has nothing to do with distribution at all. Such differences are to be expected. They mirror differences in moral outlook in the wider society. What is surprising is that they are not seen as an objection to the enterprise that most contemporary liberal thinkers have in common — the attempt to construct a theory of justice. When recent liberal thinkers claim that liberalism is a strictly political doctrine, they mean that it does not depend on any comprehensive conception of the good. They never tire of telling us that the demands of justice must take priority over any ideal of the good. They appear to have overlooked the fact that different views of the good support different views of justice.

Only this oversight can account for the fact that in `political liberalism' nothing of importance is left to political decision. The basic liberties and the distribution of social goods are matters of justice, and in political liberalism what justice demands is a matter not for political decision but for legal adjudication. The central institution of Rawls's `political liberalism' is not a deliberative assembly such as a parliament. It is a court of law. All fundamental issues are removed from political deliberation in order to be adjudicated by a Supreme Court. The self-description of Rawlsian doctrine as political liberalism is supremely ironic. In fact, Rawls's doctrine is a species of anti-political legalism.

Liberal legalists differ about the rights we have. Egalitarian legalists, such as Rawls and Dworkin, think we have welfare rights to resources, whereas libertarian legalists such as Nozick and Hayek insist that the only human rights are rights against aggression and coercion. These are fundamental differences. They reflect different beliefs about whether human beings can be said to own themselves, how they acquire property rights in natural resources, and what their well-being consists in.

Liberal legalists are at one chiefly in their common illusion that their views on rights do not express rival views of the good. In reality, Rawls and Hayek have opposed conceptions of justice, not because they take different stances in the philosophy of right, but because they hold to antagonistic conceptions of the good life. In their accounts, as in all theories of rights and justice, differing views of rights spring from different views of the good.

What these egalitarian and libertarian variants of liberal legalism have in common is more fundamental than the points at which they differ. Each supposes that principles of justice and rights can be formulated that are at once highly determinate and ideally universal. (That the later Rawls appears to have retreated from the ideal of universality does not affect the present argument.)

Libertarian liberals such as Nozick believe that a universal economic system is required by justice. For them, rights of property and laws of contract are not social and legal conventions, which can reasonably vary in accord with the changing requirements of human well-being. They are direct applications of universal human rights. It is not merely that modern economies cannot prosper without well-functioning market institutions. Rather, the institutions of the market embody timeless dictates of justice. Indeed, on this strange view, only a single type of market economy — the highly singular type of capitalism found intermittently in some English-speaking countries over the past century or so — is fully compatible with the demands of justice.

Thinking of market freedoms in this way, as derivations from fundamental human rights, is a fundamental error. Like other human freedoms, the freedoms embodied in market institutions are justified inasmuch as they meet human needs. Insofar as they fail to do this they can reasonably be altered. This is true not only of the rights that are involved in market institutions. It is true of all human rights.

The institutions of the market advance human well-being to the extent that they enable individuals and communities with different or incompatible goals and interests to trade with one another to mutual advantage. This classical defence of market institutions can be given another formulation. Individuals and communities animated by rival and (in part) incommensurable values can interact in markets without needing to reconcile these rival conceptions of the good. Market institutions assist personal autonomy and social pluralism by enabling such communities to replace destructive conflict by beneficial competition. In short, there is a value-pluralist defence of market institutions; but there is no one best variety of market institutions, either for every society, or for every context in a single society.

Markets are not free-standing. They are highly complex legal and cultural institutions. They do most to promote pluralism and autonomy when they are complemented by other, non-market institutions. Without the `positive' freedoms conferred by enabling welfare institutions, the `negative' liberties of the market are of limited value.

Egalitarian liberals such as Rawls do not claim that only one kind of economic system can be just. They recognize that justice can be realized in a variety of economic systems. Depending on historical circumstances, sometimes socialism may be best, at others some species of capitalism. In Rawlsian theory, justice is silent on the choice of economic systems. Despite this, whatever system is chosen must satisfy Rawls's principles of distribution.

This last requirement presupposes that an overlapping consensus on distributive issues can be reached across the numerous ways of life that exist in late modern societies. But insofar as different ways of life are animated by different ideals of the good, they will think of issues of distribution differently. A strongly individualist way of life will take for granted that the social unit of distribution is the individual. Others will nominate the family or intermediate social institutions for that purpose.

The several ways of life that may be found in most contemporary societies do not share a conception of the primary goods of human life. They are animated by different conceptions of the good life, which may overlap enough to make compromise possible, but which have too little in common to permit the development of a single, overarching conception of justice.

For liberal legalists, when different ways of life clash, all that needs to be done is to ask what justice demands. Once the principles governing an ideally liberal constitution have been stated, they need only to be applied. Applying the law is applying a theory of justice to particular cases; and there are no hard cases that cannot be decided. But when society contains not one but many ways of life, each with its own conception of the good, will there not be as much divergence in views of fairness as there is in understandings of the good? When ways of life differ widely in their view of the good, will they not support different views of justice?

Liberal legalists aim to circumvent conflict about the good life by appealing to ideas of justice and rights. In this they claim a lineage that goes back to Kant, who sought to develop a political philosophy based solely on the right. Whether or not this is a correct interpretation of Kant, a pure philosophy of right is a quixotic enterprise. The right can never be prior to the good. Without the content that can be given it only by a conception of the good, the right is empty.

A strictly political liberalism, which is dependent at no point on any view of the good, is an impossibility. The central catsegories of such a liberalism — `rights', `justice', and the like — have a content only insofar as they express a view of the good. At the same time, insofar as they have any definite content, claims about rights and justice are enmeshed in conflicts of value. If we differ about the good life, we are bound to differ about justice and rights. Political liberalism presupposes that justice can stand aloof from conflicting claims about the good. In truth the enterprise of a theory of justice is undone by these conflicts.

Recent liberal political philosophy ascribes infinite weight to a value that is almost infinitely complex. The requirements of justice are not everywhere the same. Because expectations vary from society to society, what is just in one may be unjust in another. What justice demands is not a matter of subjective preference, but it varies with history and circumstances.