The Political Classics: Green to Dworkin. Contributors: Murray Forsyth - Editor, Maurice Keens-Soper - Editor. Publisher: Oxford University Press. Place of publication: New York. Publication year: 1996.

Chapter 10 John Rawls: A Theory of Justice

JOHN DAY

IN the 1950s and 1960s political philosophers in the English- speaking world debated whether their academic discipline was dead. The ground for such pessimism was partly that increasing specialization in political philosophy seemed to have made traditional theorizing in the grand manner both superfluous and overambitious. Furthermore, linguistic analysis, then the dominant mode of philosophical investigation, argued that the task of moral and political philosophy was merely to clarify the language used in moral and political discourse, not, as had been done in the past, to offer moral and political prescriptions. The philosopher was said to have no special qualifications that entitled him to tell people what they should do. He could clear away some of the confusions and misunderstandings in political arguments, but should then leave moral and political decisions, which depended ultimately on each person's subjective values, to the individual. The classics of political philosophy that made strong political recommendations, like Plato Republic and Hobbes Leviathan, were thought to be phenomena of a past that had muddled true, that is, analytical, philosophy with moral advice.

Then, in 1971. John Rawls published A Theory of Justice,1 which vigorously challenged his contemporaries' conception of what was possible and desirable in moral and political philosophy. Although employing the analytical skills honed by the linguistic analysts, Rawls did not accept the limitations that they had tried to impose on political philosophy. Not content to illuminate merely corners of the subject, he elaborated a sophisticated argument for an extensive and complex theory of social justice, which has made his book a modern classic. While

____________________

1 John Rawls, A Theory of Justice ( Cambridge, Mass., 1971; Oxford, 1972). References in the text are to this edition.

219

recognizing that people have different notions of what constitutes their own good, Rawls nevertheless tried to demonstrate, in the tradition of Plato and Kant, that rational men could by the use of their reason reach agreed, objective conclusions about what constituted social justice. The argument, while highly abstract in form, was intended to provide standards of social justice that could and should be applied to actual societies. Politicians, as well as philosophers, have in practice taken Rawls seriously. In the USA and Britain some liberal politicians to the left of centre have used ideas from his Theory of Justice in formulating the theoretical justifications of their practical policies.

For over twenty years evaluation and criticism of Rawls Theory of Justice have dominated discussion in moral and political philosophy. Just as Isaiah Berlin Two Concepts of Liberty2 tends to be the starting-point of modern discussions of liberty, so philosophers examining the nature of justice cannot ignore Rawls Theory of Justice. Some critics have concluded that Rawls's idiosyncratic method and his original conclusions are fundamentally misconceived, but most of his opponents believe that his arguments have to be taken seriously. One of the most radical attacks upon Rawls's position was made by his Harvard colleague, Robert Nozick, in Anarchy, State, and Utopia.3 Yet he wrote in appreciation of Rawls:

A Theory of Justice is a powerful, deep, subtle, wide-ranging, systematic work in political and moral philosophy which has not seen its like since the writings of John Stuart Mill, if then. It is a fountain of illuminating ideas, integrated together into a lovely whole. Political philosophers now must either work within Rawls' theory or explain why not.4

____________________

2 Isaiah Berlin, Two Concepts of Liberty ( Oxford, 1958), which is discussed above in ch. 6.

3 Robert Nozick, Anarchy, State, and Utopia ( New York, 1974). This book is discussed in ch. 11.

4Ibid. 183; cf. Michael Walzer's acknowledgement in his Spheres of Justice ( New York, 1983), p. xviii. 'No one writing about justice these days can fail to recognize and admire the achievement of John Rawls. In the text, I have mostly disagreed with A Theory of Justice . . . [but my enterprise] might not have taken shape at all-- without his work.'

-220-

The Subject of the Book

The rest of the chapter discusses the text of A Theory of Justice. In his Preface Rawls, recognizing that the book is long, tells the reader which sections, amounting to about a third of the book, provide most of the essentials of the theory. In addition, he recommends certain other sections without which the theory might be misunderstood (pp. viii-ix). The exposition in this chapter concentrates on these two sets of passages. The modifications that Pawls has made to his theory since the publication of A Theory of Justice are not discussed here.

Rawls is precise about the kind of justice that he is concerned with: his aim is not to construct a general theory that covers all kinds of justice, but to formulate the principles of social justice. Social justice, as he understands it, is concerned with the distribution of rights, duties, and advantages within a society. However, he is not absolutely clear about what constitutes a society. He assumes rather vaguely that a society is 'a more or less self-sufficient association of persons' (p. 4) that has a 'political constitution' (p. 7), which perhaps suggests that he has states in mind, although he does not positively disqualify communities without sovereigns. Hopefully we shall not seriously misunderstand him if we imagine him to be thinking primarily of societies like Britain or the USA.

His concentration on justice within societies means that he does not consider justice within small groups like the family, nor, except incidentally, justice between states. He leaves it an open question how justice in these other cases might differ from social justice (pp. 7-8). Another form of Justice which Rawls does not discuss in the book is that which deals with people who break the rules of society (pp. 8-9). Social justice, as he defines it, is not concerned with that, but with the distribution of what people value in society, like wealth, and of what they regard as burdens, like taxes.

Rawls points out that, 'although society is a cooperative venture for mutual advantage', there are likely to be conflicts of interest between its members about the distribution of the benefits produced by social co-operation (p. 4). Invariably there will be insufficient amounts of these benefits, like wealth and power, to satisfy those who seek them. Hence competition for

-221-

larger shares of what is available is inevitable. The principles of social Justice lay down the right ways for the political, economic, and social institutions of a society to make distributions of the resources for which people compete. In doing so, they assign rights and duties to members of the society (pp. 4, 6-7).

The Original Position

One of the most original and most controversial parts of Rawls's theory is the method by which he works out the principles of social justice, although, with excessive modesty, he disclaims any originality for his method (p. viii). Rawls first outlines his method and the principles of justice that result from it (pp. 11- 17). Later in the book, the main explanation of the principles of justice (in ch. II, pp. 54-117) precedes, somewhat curiously, the main explanation of the method which generates them (in ch. III, pp. 118-92). The exposition in this chapter explains first the method and then how the method produces the principles of justice.

Rawls imagines people coming together in certain very specific conditions, which he calls the 'original position' (pp. 11-12, 17-19, 118-50), in order to discover what these principles should be. These persons would be free, in that they could make decisions without any constraint. Also, they would be rational in the narrow sense of seeking the most effective means to their ends (pp. 14, 142-5). Each in the special circumstances of the original position would be 'mutually disinterested' (pp. 13, 127-8), considering only his own interests in the process of deciding what should be the principles of justice that would regulate societies.

The particularly distinctive, and at first sight, strange feature of the original position is that people in it choose the principles of justice behind a 'veil of ignorance' (pp. 12, 136-42). No one knows the nature of his own society, whether it is slave- owning, or capitalist, or socialist. Nor does he know his own position within a society, whether king or serf, factory owner or factory worker, master baker or apprentice candlestick maker. Those behind the veil of ignorance do not even know what their own abilities are, whether intellectual or manual, great or small. Furthermore, they are ignorant of their own psychologies,

-222-

not knowing whether they are brave or cowardly, with or without common sense, extrovert or introvert. Finally, persons in the original position do not know their own conceptions of the good: they are ignorant whether, outside the original position, they prefer, for example, wealth or power, whether they are ascetic or hedonist, religious or atheist.

However, although they know nothing about their own individual characteristics or their places in society, they do have some general knowledge (pp. 137-8). This is necessary if they are to make sensible choices about the principles of justice. They do, therefore, know general facts about the workings of human minds, societies, and governments.

These people in the original position, knowledgeable about human beings in general, but ignorant of their own individual characteristics and social positions, are imagined choosing principles of justice that should regulate all societies. The purpose of the original position is to ensure that the choice of these principles is made with impartiality, which is the essence of justice. None of the persons behind the veil of ignorance could know whether he would benefit or not from any principles that might be chosen to shape societies. From the equality of ignorance of those in the original position emerges automatically what Rawls calls 'justice as fairness'.

Rawls claims that his method of determining social justice is a form of the traditional argument from an original contract (p. 11). The original position resembles a convention of individuals in a state of nature deciding the conditions on which they are prepared to form a society. However, in the original position the persons are not setting up a particular society, but working out the principles that should govern any society. Both in the traditional model of an original contract and in Rawls's original position, the individuals wish to secure the benefits of cooperation that a society brings, but also to safeguard themselves against the dangers that certain kinds of society might impose on them. One of Rawls's achievements is to show how the idea of social contract can be developed so as to produce a theory of social justice at a time when many thinkers believed that contractual theory was of merely historical interest in political philosophy.

Rawls removes the ambiguities that existed in some theories

-223-

of social contract about whether a social contract had existed in the past or might exist in the future. The original position is purely hypothetical, a device that Rawls invents to argue for a particular theory of justice. He is inviting his readers to take part in a thought experiment. He is not suggesting that the original position might have existed at some point in history, nor is he recommending that some people set it up at some stage in the future.

When Rawls says that the people in the original position do not know their own personalities or their own positions in society, he is speaking as if they are historical persons abstracted from concrete historical social situations, rather than unknown people existing before the creation of society, as in traditional contractual theories. It is as if a number of people with varying characters and abilities, from different types of society, have been lifted from their historical settings and inflicted with selective amnesia. Then, in this condition they are asked to choose principles of justice that should regulate all societies, which they do in ignorance of how these principles would affect them in their situations back home. Afterwards they might be supposed to be returned by the philosophical sorcerer to their full personalities in their actual societies, where they would feel obliged to apply the principles of justice that they had impartially agreed to in the original position.

The decision on the principles of justice in the original position is in one significant respect unlike most actual, historical agreements. The people in the original position do not have different views about what constitutes social justice. Their agreement consequently does not follow bargaining from different positions and does not represent a compromise between competing ideas (p. 139). This is because each person behind the veil of ignorance is identical to every other. He has the same general knowledge and the same ignorance of who he is and where he has come from. Each person in the original position acts rationally and in pursuit of his own self-interest. Hence, the arguments for certain principles of social justice appeal equally to all those in the original position. In one sense, therefore, all that Rawls needs in the original position is one hypothetical person. However, the way in which he sets up the original position is valuable, because it both acknowledges

-224-

that people have an infinite variety of personal characteristics and social positions and shows how the impartiality that is intrinsic to justice insists that these differences are ignored.

In working out the principles of justice Rawls does not wish to rely entirely on arguments from the original position, although these arguments form the core of his theory. He proposes checking the conclusions that are derived from the original position against our intuitively held moral beliefs, which he calls 'considered convictions' (pp. 19-20, 47-8). If the original position leads to principles at variance with these convictions, there are two alternatives: either to modify the convictions, or to change the conditions of the original position so that it Yields principles that no longer conflict with these convictions. Rawls expects to do a little of each, holding on to strongly felt convictions when they seem incontrovertible, but changing them if their correctness seems open to question and the original position leads to a firm principle. Rawls claims, for example, that the considered conviction that religious intolerance is wrong is beyond question. Therefore, it would be necessary to change any model of the original position that led to a different conclusion. However, considered convictions are less likely to provide certainty about the correct distribution of wealth and here the original position provides positive guidance (pp. 19-20). Rawls recommends going back and forth between considered convictions and conclusions from the original position, making adjustments in each where necessary, until the discrepancies are ironed out. This state he calls 'reflective equilibrium' (pp. 20, 48-50). As he acknowledges, he does not work through this process in the book, but he seems to say that what he does present in the book is the result of such a process (p. 21).

This section of Rawls's argument may be said to introduce a kind of moral common sense into an otherwise rigorously rational argument. However, it risks opening the floodgates to subjective moral judgements and undermining the strong argument from the original position, since different people are likely to have different 'considered convictions'. Rawls is content with the reflective equilibrium that he has reached, but are others, on checking the conclusions from the original position against their own considered convictions, likely to come up with different reflective equilibria? Clearly, Rawls

-225-

TABLE 1. The maximin rule

Decisions Circumstances

C1 C2 C3

d1 -7 8 12

d2 -8 7 14

d3 5 6 8

wants to avoid presenting an abstract argument for certain principles of justice that contradicts intuitively held convictions about justice, but his advocacy of working towards reflective equilibrium seems to throw some doubt on the validity of the argument from the original position. It is difficult not to feel uncomfortable with the idea that Rawls recommends of tampering with the conditions of the original position if that is necessary to ensure that its conclusions do not violate considered convictions. However, it should be emphasized that Rawls is prepared to modify the conditions of the original position only if its conclusions fly in the face of convictions that have been well considered, that clearly do not result from prejudice, and that are held with the greatest confidence.

The Choice of Principles from the Original Position

Rawls's main concern in A Theory of Justice is, of course, to derive principles of justice from the original position (pp. 150- 92). A crucial stage of the argument occurs when he demonstrates that the persons rationally considering their own interests behind the veil of ignorance would choose principles of social justice by adopting the 'maximin rule' (pp. 152-8). This 'tells us to rank alternatives by their worst possible outcome: we are to adopt the alternative the worst outcome of which is superior to the worst outcomes of the others' (pp. 152-3). The significance of this rule can be illustrated with the help of Table 1, which Rawls uses (p. 153).

A person has the choice of three alternative decisions, and he knows that for each decision there are three possible

-226-

outcomes, depending on which set of possible circumstances actually occurs. Perhaps he is making an investment and has the opportunity of buying shares in one of three companies. An unusually prescient financial adviser is able to predict how each company will do in each of three sets of economic circumstances, and also has, miraculously, reliable inside knowledge that one of these three sets of circumstances will actually occur. If the investor makes decision d1, he may lose £7,000 (if the units in the table are thousands of pounds) or gain either £8,000 or £12,000; if he makes decision d2, he may lose £8,000, or gain either £7,000 or £14,000; if he makes decision d3, he may gain either £5,000 or £6,000 or £8,000.

The attraction of decision d2 is that, if he is lucky, he will gain more than is possible from either of the other decisions, although he also risks losing more. Decision d1 has its own advantages, because it offers the chance of gaining only £2,000 less than d2 if the most favourable circumstances come up, while the loss, if the worst circumstances happen, is £1,000 less. Decision d3 offers distinctly less reward if the best circumstances crop up, but its appeal lies in the fact that even in the worst possible circumstances there is, unlike the other two cases, no loss, but a small gain, £5,000. If the investor takes decision d3, he is acting in conformity with the maximin rule.

Which is the best decision for an individual in this sort of case may depend on his personality--for example, whether he is adventurous or cautious--or upon his personal circumstances--for example, whether he can afford to lose money, or whether a member of his family is desperately in need of the money that he might win. However, there are very special features of the original position that make following the maximin rule the only rational choice. Behind the veil of ignorance no one knows whether he will be, once he is out of the original position and in a concrete historical situation, at the top, or at the bottom, or in the middle of the social heap. He could be a slave on a sugar plantation, or a Russian serf tied to the land, or a factory worker in nineteenth-century England working twelve hours a day, or a subsistence farmer in contemporary Africa on the verge of starvation. Consequently, it is sensible to take no risks, but to ensure the best possible outcome for the worst-off in society. That entails establishing principles of

-227-

justice that guarantee the improvement of whoever is most disadvantaged.

At first sight it might appear attractive to gamble and to work out principles of justice that would protect your power and wealth if you are lucky enough in real life to find that you are a millionaire monarch. However, it is not rational to count on being a slave owner when you might turn out to be a slave, or to set up principles of justice that would advantage you if you find that you are Queen Elizabeth I or Blackadder, but push you further into the mire if you end up as a pauper or Baldrick. In the original position it is not rational to accept, for example, Nozick's entitlement theory of justice,5 which entitles a person to property if he has legitimately acquired it, as this does nothing for those who, for whatever reason, have no property.

Rawls shows the power of his argument by comparing the appeal to those in the original position of the maximin rule and of the principle of average utility, which he believes is the strongest rival to his own theory of justice (pp. 161-6, 167- 75). At first sight, it might appear that those in the original position would choose the principle of average utility as the basis of social justice. The society would, according to this principle, aim to increase the average amount per capita of whatever people enjoyed. Justice would include, for example, increasing the wealth of the country so that, on average, everyone was better off. However, there would be no guarantee that any particular individual would get an average share of the increased wealth: he might get more than the average, but he might get less. He might even lose some wealth that he previously had, even though the average wealth per head increased. Yet, someone arguing against Rawls might say that it would be rational for people in the original position to prefer the principle of average utility. They would gamble on receiving the average increase or better, although risking getting less than the average.

Rawls, however, argues that it would not be rational in the original position to choose the principle of average utility to regulate distribution of what is valued in society. His argument

____________________

5 Anarchy, State, and Utopia, 150-3.

-228-

is that those in the original position would not be sensible to take the risks that accepting the principle of average utility entails. Those behind the veil of ignorance have no basis for knowing their propensity to take risks (p. 172), so social justice cannot be formulated on the assumption that temperamentally they are gamblers. Nevertheless, there are some situations where non-gamblers are sensible to take a chance, because there is no certainty which course of action will be the most advantageous. In that case it is rational to weigh probabilities. For example, if there is a danger of aerial bombing, the evidence may point to people having a greater chance of survival if they shelter in a cellar, although there is a risk of their being buried alive there. In order to calculate probabilities, however, you need evidence, and this, according to Rawls, is what those behind the veil of ignorance lack (pp. 172-3). In those particular conditions they have no idea of the probabilities, if they accept the principle of average utility, of their ending up as a winner or a loser (p. 168). In these circumstances of ignorance it is not rational to risk becoming a loser, when the alternative strategy of acting on the maximin principle safeguards your position if you turn out to be a disadvantaged or unlucky member of society.

The clinching argument against people in the original position choosing the principle of average utility is that it gives no guarantee that individuals will not be used for others' ends (pp. 177-83). There might be circumstances where the greatest average utility could be achieved only by the institution of slavery. So those in the original position would not choose the principle of average utility in case they were to lose their liberty in order to benefit others in the society. In general, people would not want to risk any form of long-term social, economic, or political subordination for the sake of improving the average utility, from which they did not personally benefit.

The Two Principles of Justice

The purpose of constructing the original position is to explain which principles of social justice would be chosen by anyone who did not know his own position in society and who, therefore, would be absolutely impartial. Rawls argues that

-229-

such a person would want to ensure that the position of the worst-off in any society would always be improved, irrespective of desert. Acting according to the maximin rule, those in the original position behind the veil of ignorance would, according to Rawls, choose two principles of justice to ensure the fair distribution by societies of the 'primary goods' that those societies help to produce.

Primary goods are the things that people specially value, that they want more of, and about whose distribution they are consequently particularly concerned. The primary goods that social justice is concerned with are those over whose distribution society has some control. Wealth and income are clearly important among these socially controllable primary goods, but by no means the only ones. People also want rights and liberties, opportunities for self-development and self-advancement, and power (pp. 62, 92). The most important primary good, Rawls argues, is self-respect, which 'includes a person's sense of his own value, his secure conviction that his good, his plan of life, is worth carrying out' (p. 440). Self-respect depends, paradoxically, partly on appreciation by others.

Some goods, such as vigour and imagination, Rawls says, are 'natural' in the sense of not being under the control of society (p. 62). Consequently, their distribution is not the concern of social justice. Among the things that Rawls regards as influenced by society, but not under its control, is health. However, although diseases are biological and therefore 'natural', society can, of course, do much to limit their effects. Counting health as a natural rather than a social primary good leads Rawls to ignore questions of social justice that arise in discussing the distribution of health care.

One further point needs to be made about the primary goods, the distribution of which social justice seeks to regulate. Rawls does not assume that all people in fact will be interested in all these primary goods in the same ways and to the same extent (p. 93). He is aware that there are some who despise wealth and others who are uninterested in power. In the original position the principles of justice provide for the fair distribution of the social primary goods that most people in most societies want, without assuming that everyone in all societies always wishes to take up his fair share of them all.

-230-

The first of the two principles that emerges from the original position states that each person is to have an equal right to as much liberty as possible provided that it does not prevent others having the same liberty (p. 60). The similarity with John Stuart Mill's argument in On Liberty that the only justification for restricting anyone's liberty is to prevent harm to others is immediately obvious. It is also interesting and significant that Rawls speaks of rights to liberty in his crucial first principle, since he does not in his theory of justice discuss at length, or emphasize the importance of, the concept of rights in the sense of moral entitlements, as many other philosophers of justice do, especially those in the Lockean tradition. Nozick, for example, opens his book with the declaration: 'Individuals have rights, and there are things no person or group may do to them (without violating their rights). So strong and far- reaching are these rights that they raise the question of what, if anything, the state and its officials may do.'6 It is important to notice that Rawls, although not exploring the concept of rights,7 as Nozick does, nevertheless has this concept at the centre of his theory of justice.

In his first, provisional formulation of the first principle Rawls speaks of the right to liberty, but later changes 'liberty' into the more cumbersome but more precise 'total system of equal basic liberties' (p. 302). He is here acknowledging that it makes more sense to talk of liberties than liberty8 and recognizing that one liberty may come into conflict with another. For example, in some historical circumstances, political liberty, where all are free to participate in political affairs, may produce governments that infringe civil liberties such as freedom of thought (p. 201). So, the liberties to which people are entitled have to be balanced against each other in a complex system of liberties.

The right to equal liberties, which the first principle prescribes, would be chosen in the original position because no

____________________

6 Anarchy, State, and Utopia, ix; cf. Ronald Dworkin, Taking Rights Seriously (2nd impression, London, 1978).

7As distinct, of course, from the right, which is of central concern to Rawls, particularly where he contrasts it with the good (e.g. pp. 446-452). For a full study of Rawls on rights, see Rex Martin, Rawls and Rights ( Lawrence, Kan., 1985).

8Cf the argument against the existence of a right to liberty in Ronald Dworkin, Taking Rights Seriously, 266-72.

-231-

one behind the veil of ignorance would want to risk finding himself in a society where he was deprived of liberty and where the principles of social justice said nothing about the necessity to give him liberty. Rawls believes that those in the original position fear above all else discovering when they reenter the real world that they are slaves denied all basic liberties. Consequently they would regard it as their first priority to ensure that slavery was pronounced unjust. People need liberties because without them they cannot seek whatever ends they have chosen for themselves. Liberty is the necessary condition of self-fulfilment.

The heart of the first principle of justice is the protection of liberty, but equality is also prominent, both in the origins and the substance of the principle. Those in the original position are equal in their ignorance of their own societies and their positions within them. It follows that they are equally concerned to safeguard the least advantaged in society whom they may turn out to be. They consequently support a principle of justice that prescribes reform of any society in which the greater liberty of the rich and the powerful is enjoyed at the cost of the liberty of the poor and the weak. Equality of liberty is vital to a just society. So all must have equal rights to equal systems of liberties.

The second of the two principles of justice, which has two subprinciples, regulates social and economic inequalities. These are to be arranged so that they are, first, attached to positions and offices equally open to all, and, secondly, to the greatest benefit of the least advantaged (pp. 60-1, 83, 302). The second principle, then, does not demand that societies should move towards absolute social and economic equality. There is not, therefore, an exact parallel with the first principle, where equality of people's ignorance in the original position about their social positions leads to the prescription of equal liberty within a society.9

What is immediately obvious about the second principle is the effect of arguing according to the maximin rule. Just as it leads in the first principle to the provision of safeguards for those who are worst off in terms of liberty, so in the second

____________________

9Why this is so is explained below, pp. 236-7.

-232-

principle it results in special treatment for those who are worst off in terms of wealth and power.

Before closer scrutiny of the second principle, it is important to understand the priority that those in the original position would give the first principle over the second. A central part of Rawls's theory consists in his placing the principles of justice in what he calls a 'lexical order' (pp. 34-45, esp. pp. 42-3). This arranges the principles in order of moral importance, ranking the first principle that establishes equal liberty before the second principle that regulates social and economic inequalities. It follows from this lexical ordering that the first principle must be satisfied before the second principle. In practical terms, governments must take whatever measures are necessary to move towards equal liberty before they introduce policies to modify social and economic inequalities.

Those in the original position would choose to give priority to liberty because they would not, in historical situations, want to sacrifice liberty in order to obtain more wealth or power (pp. 151-2, 542-8). Since liberty is a necessary condition of people obtaining their ends, and thus fulfilling their human potentialities, they are not prepared to trade it in for any other benefits. In the original condition they would not accept principles of justice that endorsed sacrificing some liberties in order to secure a more equitable distribution of wealth.

Rawls recognizes only one exception to the priority of liberty: at a very early stage of social development, when people are very poor, they would accept a temporary denial of equal liberty, 'if it is necessary to enhance the quality of civilization so that in due course the equal freedoms can be enjoyed by all' (p. 542; see also p. 152). Presumably Rawls means that people in abject poverty cannot enjoy liberties, so it is necessary to build up wealth in a society to a point where they can. Then they can function as full human beings.

Although it is never right for a society to sacrifice liberty to achieve some other end (except in circumstances of dire poverty), it is sometimes necessary to restrict one liberty in order to protect another. 'The precedence of liberty means that liberty can be restricted only for the sake of liberty itself' (p. 244, see also pp. 243-50). For example, in some historical circumstances it would be best to restrict the franchise, and hence

-233-

restrict political freedom, in order to prevent a popularly elected government suppressing the more fundamental civil liberties, such as liberty of conscience and freedom of the person (pp. 228-34, especially pp. 229-30, and 246-7). Another situation where liberty may have to be limited to defend liberty is where an intolerant group endangers the liberties of others (pp. 216-21). Lexical ordering is necessary in order to decide priorities between competing liberties, but the aim remains to increase the overall enjoyment of liberty.

The significance of those in the original position choosing a lexical order of principles can be seen very clearly when this method of arbitrating between independent and potentially conflicting moral principles is contrasted with alternative approaches. One way of dealing with conflicting principles of justice is that recommended by Mill in Utilitarianism: to use an overriding criterion--in his case, utility--to decide between them. In trying to decide whether it is just to reward skill more than effort in a factory worker, Mill claims that the only way to reconcile the competing arguments is by calculating social utility.10 A totally contrasting way of dealing with conflicts between moral principles forms the conclusion of Isaiah Berlin Two Concepts of Liberty. There Berlin argues that the essence of individual liberty is being able to choose which moral value should have priority in which circumstances: 'human goals are many . . . and in perpetual rivalry with one another. To assume that all values can be graded on one scale, so that it is a mere matter of inspection to determine the highest, is to falsify our knowledge of men as free agents.'11 So, there is no fixed, rationally agreed hierarchy of moral values, as there is in Rawls's theory of justice. Rawls, in arguing for lexical ordering, is rejecting both the moral relativism, which Berlin positively praises, and the belief in a superior moral criterion to be used as an arbitrator between competing lower level moral principles, in which Mill has faith.

Rawls proposes a lexical order not only between the first and second principles of justice, but also between the two parts of the second principle. Just as some forms of liberty have

____________________

10 J. S. Mill, Utilitarianism ( Everyman edition, London, 1991), 60.

11 wo Concepts of Liberty ( Oxford, 1958), 56.

-234-

priority over others, so, in arranging social and economic inequalities, Rawls's particular version of equality of opportunity has priority over increasing the benefits of the least advantaged. Those in the original position would regard the establishment of the fair distribution of positions of power in a society as having moral priority over the redistribution of wealth.12

Rawls believes that those in the original position would insist on some form of equality of opportunity as an integral part of social justice. However, he argues that they would not be content with a simple 'career open to talents' (p. 72). This would merely ensure that the limited number of positions and offices available in a society would go to whoever performed best in open competitions. However, what Rawls calls 'fair equality of opportunity' (pp. 73-4) demands more than allotting positions of power on grounds of ability and not according to personal contacts. In addition, no one should be disadvantaged in the competition for places by a defective education resulting from an inferior social position. In some way, those with such a handicap should receive compensation to put them on an equal footing in the competition with those who have no such handicap. In effect, fair equality of opportunity embraces positive discrimination. Significantly, those in the original position are, once again, looking after the interests of whoever is at the bottom of the pile--in this case, those people whose education has suffered because of their social position.

The other part of the second principle of justice, dealing with the distribution of social and economic benefits, has probably aroused more interest and controversy than any other element in Rawls's theory except the veil of ignorance. One method by which he reaches what he calls the 'difference principle' is to consider an alternative way of distributing social and economic benefits that at first sight might seem attractive to those in the original position. This approach is similar to the one used in the search for the general principle that those in the original position would adopt in working out the principles

____________________

12Curiously, Rawls calls his version of equality of opportunity the second part of the second principle (e.g. p. 83), although in lexical order it has priority over the other part about increasing the benefits of the worst off (e.g. pp. 302-3).

-235-

of justice, where the superficially persuasive theory of average utility is rejected in favour of the maximin rule.

Rawls considers whether those in the original position would prefer the principle of efficiency, known in economic theory as Pareto optimality, as the just way to distribute benefits such as wealth within societies (pp. 65-72). This would recommend as social policy that everyone should be made better off, provided that no one is consequently made worse off. However, this proves unacceptable as a principle of justice, since it provides the worst off with no guarantees of escaping from permanent poverty. For example, the principle of efficiency would justify one person holding all the wealth in a community, since there is no rearrangement that would make some better off, but none worse off. The owner of all the wealth is bound to be a loser in any rearrangement. Similarly, in certain circumstances, serfdom would be justified, when there was no way of improving the lot of the serfs without worsening the position of their lord.

In place of the efficiency principle Rawls proposes as a basis for the fair distribution of social and economic benefits his original and controversial 'difference principle' (pp. 75-80). This states that 'the higher expectations of those better situated are just if and only if they work as part of a scheme which improves the expectations of the least advantaged members of society' (p. 75). This part of the second principle of justice requires, then, that, once equal liberty and fair equality of opportunity have been secured, as the first principle and the lexically prior part of the second principle demand, rearrangements in the distribution of social goods such as wealth should normally benefit the worst off. This would often entail some form of taking from the rich and giving to the poor.

The important and interesting exception is when improving the lot of the best off also improves the lot of the worst off. If, for example, paying a senior member of the government a larger salary enables or encourages him to evolve an economic plan that increases the dole of the unemployed or provides them with better job prospects, this is justified by the difference principle. What is distinctive about this principle is that, although it tends to greater social and economic equality, it positively endorses inequality if that is the best method of

-236-

improving the conditions of the most deprived. Consistently, Rawls's form of social justice has as its priority the advancement of the least advantaged, which the maximin rule that was adopted in the original position requires.

Self-Interest and Justice

Rawls's theory of social justice contains a central paradox which may took like an inherent and fatal contradiction. In order to deduce principles of justice the essence of which is impartiality, he initially supposes that people in the original position argue from self-interest. It may seem at first sight that Rawls sets himself the impossible task of deriving justice from self- interest.

To some thinkers, of course, there is no problem in moving from self-interest to justice. Acting justly to others, on the understanding that they act justly to you, is what self-interest dictates as the most efficient way of dealing with other self- interested people. Glaucon in Plato Republic and Hobbes in Leviathan both take this line of argument. However, since Rawls does not regard justice as a branch of self-interest in this way, the problem remains of how in his theory justice can emerge out of self-interest.

The problem would never arise for Rawls, it might be thought, if he did not start by making the false, Hobbesian assumption that human beings can only have self-interested motives. However, the answer to this criticism is that, in fact, he is not denying that people can be altruistic. If people were always unavoidably self-interested, there would be no chance that his principles of justice would be implemented, except when they happened to coincide with powerful people's self- interest; but he does not make this pessimistic assumption. It is only when he is constructing the purely hypothetical original position that he posits the existence of persons arguing solely from their own self-interest. He expects, or at least hopes, that in real life, by contrast, people will try to implement the principles of justice because they are right, whether or not they personally stand to gain from their implementation (pp. 147-8). Rawls's theory of justice, far from claiming that

-237-

human beings are irrevocably self-interested, assumes that they have the moral potentiality to act justly.

The question then arises as to why he sets up the original position in such a way that people within it consider only their own interests. The explanation starts from the fact that the function of social justice is to arbitrate between conflicting interests: 'A society . . . is typically marked by a conflict as well as by an identity of interests' (p. 4). If there were no conflicts of interest between individuals, there would be no need for social justice. Individuals are sometimes benevolent towards each other, but in those circumstances the issue of social justice does not arise. Social justice comes into play on those occasions when people do have conflicts of personal interest. The original position is hence established to deal with those conflicts of interest over which social justice presides.

Rawls is keen to emphasize the important role of justice in protecting the separate interests of individuals. His original position is set up with individuals seeking principles of justice that safeguard their own interests in order to rule out any principle, like average utility, that would justify sacrificing one individual's good for the sake of another's. Those in the original position find a safe form of justice in the two Rawlsian principles, which in effect elaborate Kant's categorical imperative to 'act in such a way that you always treat humanity . . . never simply as a means, but always at the same time as an end'.13 This Kantian principle insists on respect for the individual and, therefore, reinforces the individual's respect for himself, which Rawls regards as his chief good.

In the original position each individual knows that in a real society he has a distinct and particular self-interest, although he does not, behind the veil of ignorance, know what it is. In the original position the individual self-interestedly chooses principles of social justice that will improve his conditions in a real society if he happens to be one of the most disadvantaged in that society. Thus, he chooses the principles of equal liberty, fair equality of opportunity, and the difference principle. These principles, emanating from the self-interest of people in the hypothetical original position, should be adopted in real societies

____________________

13As formulated by Nozick in Anarchy, State, and Utopia, 32. See Theory of Justice, 178-83. Interestingly, this principle is central to both Rawls's and Nozick's theories, in spite of their fundamental differences about how it should be interpreted.

-238-

because they are just. The veil of ignorance in the original position is intended to produce principles of justice that are not the product of any particular social group arguing on its own behalf. In their impartiality lies their fairness. So it is a paradox and not a contradiction that self-interested behaviour in the original condition leads to principles of justice in real societies.

Conclusion

A Theory of Justice contains and engenders a host of paradoxes. It derives practical altruistic principles from arguments about the choices of hypothetical egoistic persons. Rawls produced a classic of egalitarianism that denies the value of absolute equality. Some of his critics believe that Rawls's arguments for the priority of liberty merely rationalize the prejudices of late- twentieth-century American liberals, but ostensibly his book is a work of pure philosophy seeking through reason to discover eternal moral truths. While Rawls himself, with characteristically liberal open-mindedness, has moved on and modified the ideas of his Theory of Justice, this work still commands attention as a modern classic, provocative in both its methods and its conclusions.

Bibliographical Note

The work discussed in this chapter is John Rawls, A Theory of Justice ( Cambridge, Mass., 1971; Oxford, 1972). Since the publication of this book, on which Rawls's extensive reputation was built, he has developed and modified his ideas. His most recent thoughts on political philosophy are contained in John Rawls, Political Liberalism ( New York, 1993).

No contemporary political philosopher has stimulated more philosophical discussion than Rawls. The extensive debate in philosophical journals soon after the publication of his Theory of Justice is reflected in a comprehensive collection of journal articles: Norman Daniels (ed.), Reading Rawls: Critical Studies of 'A Theory of Justice' ( Oxford, 1975). A few years later two other editors invited scholars to write essays for a book on various aspects of Rawls's theory: H. G. Blocker and E. H. Smith (eds.), John Rawls' Theory of Social Justice ( Athens, Oh., 1980).

Three books that are devoted wholly to analysis of Rawls's theory of justice regard his book as a major work, but disagree with many of

-239-

his arguments: Brian Barry, The Liberal Theory of Justice: A Critical Explanation of the Principal Doctrines in 'A Theory of Justice' by John Rawls ( Oxford, 1973); David Schaefer, Justice or Tyranny? A Critique of John Rawls's 'A Theory of Justice' ( Port Washington, NY, 1979); Robert P. Wolff , Understanding Rawls ( Princeton, 1977). A critical but more sympathetic exegesis of Rawls's theory is contained in Rex Martin, Rawls and Rights ( Lawrence, Kan., 1985).

For those seeking an easily understood and relatively short book on Rawls and his critics, Chandran Kukathas and Philip Pettit, Rawls: 'A Theory of Justice' and its Critics ( Cambridge, 1990), may be recommended. Good chapter-length introductions can be found in the chapter by Samuel Gorovitz, 'John Rawls, "A Theory of Justice"', in Anthony de Crespigny and Kenneth Minogue (eds.), Contemporary Political Philosophers ( London, 1976), and in Bhikhu Parekh, Contemporary Political Thinkers ( Oxford, 1982).

Three years after the publication of A Theory of Justice, a colleague of Rawls at Harvard University produced a classic refutation of Rawls's arguments about justice: Robert Nozick, Anarchy, State, and Utopia ( Oxford, 1974), especially chapter 7, section 2. Since Nozick's book appeared, discussions of justice often treat the theories of Rawls and Nozick together as of major importance (for example, Brown and Kymlicka, whose books are mentioned below). One book contains articles about both Rawls and Nozick: J. Angelo Corlett (ed.), Equality and Liberty: Analyzing Rawls and Nozick ( London, 1991).

Special attention should be paid to two other major philosophical works, besides Nozick's, that incorporate discussions of Rawls's ideas on justice: Ronald Dworkin, Taking Rights Seriously (second corrected impression with appendix, London, 1978), ch. 6; Alasdair MacIntyre, After Virtue: A Study in Moral Theory (second (corrected) edn. with postscript, London, 1985), ch. 17.

Interesting sections on Rawls's theory of justice may be found in some books on the general theme of justice: Brian Barry, Theories of Justice ( London, 1989); Tom D. Campbell, Justice (Basingstoke, 1988); David Miller, Social Justice ( Oxford, 1976), ch. 1, sect. 4; Philip Pettit, Judging Justice ( London, 1980), pt. 5; Michael J. Sandel, Liberalism and the Limits of Justice ( Cambridge, 1982).

There are also useful discussions of Rawls in works on political philosophy that regard justice as the central theme: Alan Brown, Modern Political Philosophy: Theories of the Just State (Harmondsworth, 1986), ch. 3; Will Kymlicka, Contemporary Political Philosophy: An Introduction ( Oxford, 1990), ch. 4.

For a full bibliography, see J. H. Wellbank, Denis Snook, and David T. Mason (eds.), John Rawls and his Critics: An Annotated Bibliography ( New York, 1982).

-240-

II

Robert Nozick: Anarchy, State, and Utopia

MICHAEL LESSNOFF

ROBERT NOZICK'S Anarchy, State, and Utopia was published in 1974,1 three years after his colleague at Harvard University, John Rawls, published A Theory of Justice, and the two books have been bracketed together ever since. Politically they are sharply opposed. While Rawls's 'liberal' theory calls on governments to undertake extensive economic transfers to the poor, the 'libertarian' Nozick defends the 'minimal state', a state 'limited to the narrow functions of protection against force, theft, fraud, enforcement of contracts and so on' (p. ix) --a state forbidden to engage in any economic redistribution whatever. Such an extreme position shocked many; but all have been forced to pay careful attention to Nozick's arguments.

These arguments can be considered in four stages, as follows. (1) Nozick's fundamental moral philosophy; (2) his justification of the minimal state; (3) his theory of economic justice, which forbids any more-than-minimal, or redistributive state; (4) his argument that the minimal state provides a framework for utopia.

Nozick's Fundamental Moral Philosophy

Nozick's moral philosophy is a particular version of the priority of the right over the good (a very particular version--for Rawls also asserts this priority). As Nozick himself puts it, moral goals are subordinate to moral 'side constraints' (p. 29), which forbid absolutely the violation of individual rights. This

____________________

1 Robert, Nozick, Anarchy, State, and Utopia ( Oxford, 1974). References in the text are to this edition.

-241-

side-constraints morality, Nozick stresses, is not an injunction to minimize the total amount of rights violation (which he calls a 'utilitarianism of rights'). There is, for example, a well- known objection to the utilitarian theory of punishment that if, in certain circumstances, framing and punishing an innocent person would effectively deter crime, utilitarianism would then prescribe punishing the innocent person. In Nozick's terms this would be violating one person's rights in order to minimize rights violations overall. Such actions are absolutely ruled out by his side-constraints morality.

As Nozick explains, his side-constraints morality rests on a fundamental moral principle--the 'Kantian imperative', that is, 'the Kantian principle that individuals are ends and not merely means' (pp. 30-1). Individuals, therefore, 'may not be sacrificed or used for the achieving of other ends without their consent'. ('Other ends' means the ends of other persons.) Side constraints, Nozick says, reflect the separateness and inviolability of individual persons: harm to one individual cannot be justified by greater gain to another or others, for there is no overall 'social good' advanced thereby, only the separate individuals enjoying goods or suffering evils. 'There is no justified sacrifice of some of us for others' (p. 33).

Since states and governments, just like any other agent, are bound by Kantian moral side constraints, their power of coercion must not be used to force some persons to serve the ends of others. This 'libertarian side constraint' is fundamental to Nozick's political philosophy.

Why must persons be treated in accordance with the 'Kantian' side constraints? Because, according to Nozick, a person is a being with the ability to regulate and guide its life in accordance with some overall conception it chooses to accept. And why is this so morally crucial? Nozick suggests that the answer has to do with an 'elusive and difficult notion: the meaning of life' (p. 50). A person gives meaning to his or her life by shaping it in the way indicated: only so can a life be meaningful. Here we arrive at the very bedrock of Nozick's moral and political philosophy, the presumed premiss of all his conclusions. Whether they are really derivable from it is an important question.

Let us return to politics. Given Nozick's Kantian imperative,

-242-

one might wonder whether the existence of the state--an intrinsically coercive institution--can be justified. Nozick's answer is yes--but only the minimal state.

Nozick's justification of the Minimal State

In seeking to justify even the minimal state Nozick is, he says, taking seriously the anarchist view that the state, of necessity, 'must violate individuals' rights and hence is intrinsically immoral' (p. xi). He therefore begins his argument by considering an anarchic, or stateless, situation--the traditional 'state of nature'. Nozick does not claim to know what this anarchic state of nature would be like, but it is, he maintains, possible to frame a theoretically appropriate description. This should be neither absurdly optimistic, nor excessively pessimistic (like Hobbes's version). Given that the aim of the argument is to convince the anarchist, the appropriate starting-point is a state of nature corresponding to 'the best anarchic situation one reasonably could hope for' (p. 5), namely, one in which people generally, though not always, act as they should. Anarchism would be refuted, says Nozick, 'if one could show that the state would be superior even to this most favoured situation of anarchy . . . or would arise [from it] by a process involving no morally impermissible step'. Since Nozick in fact adopts the second strategy, the (relevant) state of nature must be subject to moral constraints. A state of nature subject to moral constraints, and in which people generally, though not always, act as they should, is, approximately, Locke's state of nature. A Lockean state of nature, therefore, is the starting- point of the argument.

Nozick claims that he takes seriously the position of the anarchist, but it would be more accurate to say that he responds to one particular kind of anarchism, individualist anarchism (as Nozick sometimes puts it) or (as it is often called) anarcho-capitalism.2 This, I believe, is a further reason why

____________________

2Some anarcho-capitalist works referred to by Nozick are M. W. Rothbard, Power and Market (Menlo Park, 1970); D. Friedman, The Machinery of Freedom ( New York, 1973); J. Hospers, Libertarianism ( Los Angeles, 1971); and J. J. Martin, Men against the State: The Expositors of Individualist Anarchism in America, 1827-1908 ( Colorado Springs, Colo., 1970). See Anarchy, State, and Utopia, ch. 2 n. 4.

-243-

Nozick's argument begins from a Lockean state of nature. The anarcho-capitalist objection to the state needs to be precisely defined. It does not rest on any general objection to coercion. On the contrary, like Locke, the anarcho-capitalists take for granted a 'natural' right, not only to self-defence, but of self- protection by punishing those who have violated rights. Their objection to the state is that it (seeks to) monopolize (or at least to regulate) the right to punish, and that it forcibly extracts property from its citizens by taxation. According to the anarcho- capitalists, protection against invasions of person, property, and so on should be, like other goods and services, supplied and purchased voluntarily through the free market. As we shall shortly see, Nozick's argument for the state is, to a considerable extent, a response to such arguments.

Nozick's justification of the minimal state starts from the Lockean state of nature, which 'has a law of nature to govern it' (because human beings were not 'made for one another's uses').3 In Nozickian terms, the law of nature imposes moral side constraints protecting individual rights. But the law is not always obeyed. Therefore, every person also has a right of self- defence against transgressors, a right to punish them to a degree appropriate to the transgression, and a right, if injured, to exact compensation. Such a situation, Locke says, is subject to serious inconveniences. For example, many people will lack the power to enforce their rights effectively. Also, the right of men to be judges in their own cases leads to biased judgment, and to unjustified or excessive punishment (or punishment so considered by its victims). This in turn leads to violent feuding ('war', as Locke calls it) between the parties involved.

According to Nozick, Locke's conclusion that 'civil government' is the appropriate remedy for these 'inconveniences' is too quick. We need first to explore what resources are available to deal with them in the state of nature itself--namely, 'all those voluntary arrangements and agreements persons might reach' (p. 11). First of all, Nozick says, individuals could (or would?) strengthen their power to enforce their rights by

____________________

3 J. Locke, The Second Treatise of Government and A Letter Concerning Toleration, ed. J. W. Gough ( Oxford, 1957), 5.

-244-

establishing mutual-protection associations, in which 'all will answer the call of any member' for defence against and punishment of rights violators. But such simple associations would not be satisfactory. For one thing, Nozick says, they would need to set up procedures of adjudication to handle disputes between members (where one member accuses another of violating his rights). Also, it would be highly inconvenient for all members to be constantly on call to carry out I protective functions' for all other members. To deal with this problem, they would have recourse to 'division of labor and exchange': 'Some people Will be hired to perform protective functions, and some entrepreneurs will go into the business of selling protective services' (p. 13). In other words, protective functions will now be performed by commercial firms ('protective agencies') selling their services on the market. This is, of course, the solution favoured by the anarcho-capitalists.

According to Nozick, however, it is not yet a satisfactory solution. 'Initially, several different protective associations or companies will offer their services in the same geographical area. What will occur when there is a conflict between clients of different agencies [and] they reach different decisions as to the merits of the case?' (p. 15). There is a conflict, as one agency tries to protect its client while the other attempts to punish him. The problem of 'war' between individuals in the state of nature seems to have been replaced by one of war between protective agencies.

However, the situation is not, according to Nozick, as bad as it seems. The forces of the opposed protective agencies may well 'do battle' (p. 16). But the outcome will be the establishment, in each geographical area, of a single 'dominant protective association' either because one association always wins its battles in the area, and thus puts its rivals out of business there; or else (if this is not so), to avoid the wasteful costs of war, the rival agencies

agree to resolve peacefully those cases about which they reach differing judgments. They agree to set up, and abide by the decisions of, some third judge or court [in these cases] . . . Thus emerges a system of appeals courts and agreed upon rules about jurisdiction . . . Though different agencies operate, there is one federal judicial system of which they are all components. (p. 16)

-245-

Nozick now asks, 'Is the dominant protective association a state?' (p. 22). His answer is that it appears not to be, but 'appearances are deceptive' (p. 25). (This is rather misleading. It would be more accurate to say that the dominant protective agency is not yet quite a state, but is poised to become one.) The association appears not to be a state because, although dominant in its area, it does not include everyone therein. A small number of 'independents' coexist with it, but are not protected by it, because they have chosen not to buy its protective services but instead to enforce their own rights. To turn the dominant protective association into an indubitable state, the independents must be incorporated into it. Nozick argues that this will happen, that it will be accomplished by force, and that this exercise of force is morally justified. Why so?

Nozick's argument is long and complex, but in essentials is as follows. The independents who enforce their own rights, unlike the dominant protective association, do not have a relatively reliable and impartial Judicial procedure for determining guilt and appropriate punishment. Given individuals' tendency to partiality in their own favour when they are judges in their own case, right-enforcement by independents creates a serious risk of unjust punishment of the association's clients. Therefore, the association, on behalf of its clients, is entitled to prohibit right-enforcements by independents against them. Furthermore, it has the power to enforce this prohibition, and will do so (chs. 4, 5).

At this stage, the dominant protective agency has become what Nozick calls an 'ultraminimal state' (pp. 52, 113). It is a state because it monopolizes right-enforcement, but not yet a 'minimal state', because it is not willing to protect the rights of all--only of those who buy its protection services. Although Nozick, rather strangely, argues that the move to the ultraminimal state is morally justified in itself, he also (and surely correctly) states that it is morally obligatory to move from the ultraminimal to the minimal state. If this were not done, the former independents would be left without any means of protecting their rights. There is, however, a further issue--the redistributive (or apparently redistributive) nature of (even) the minimal state. For (even) the minimal state is obliged to protect

-246-

the rights, not only of those in its territory who can and do pay the 'economic' price of such protection, but also of those who cannot (p. 141). As Nozick asks (alluding to the transition from dominant protective association to minimal state): 'How can a protection agency, a business, charge some to provide its product to others?' (p. 25). Presumably, it cannot, but a (minimal) state can, and justly so, Nozick argues. Prohibition of 'independent' right-enforcement is justified only if substitute protective services are offered by the prohibitors, i.e. the clients of the dominant protective association. If necessary, they must subsidize those ex-independents who would otherwise be disadvantaged by their forcibly altered status--in particular, 'those of scanter resources' (p. 112). So, the upshot is that the Nozickian minimal state may and must provide protective services, uniformly and monopolistically for everyone in its territory, whether or not they wish it, and whether or not they can afford to pay for the services. Those who cannot afford them, but receive them none the less, may appear to be beneficiaries of compulsory redistribution of income from other citizens. According to Nozick, however, the apparent redistribution is not really redistribution at all, but compensation for coercive removal of the right of self-protection. It is part of the obligatory compensation owed to the independents for loss of rights.

Assessment of Nozick's justification of the Minimal State

Many commentators have been puzzled by Nozick's justification of the minimal state. No state, minimal or other, has ever been established via the sequence of events described in Nozick's story, as he must know. What, then, is the significance of the fact that this sequence can be imagined as a bare possibility? But Nozick's claim is more than this. It is that, if men began in the most favourable state of anarchy that can reasonably be expected, they would, through the promptings of self-interest but without violating any moral side constraint, eventually move into a state in the way described (pp. xi, 7, 114, 118). The argument is framed thus because it is aimed at anarchists, a fact which is particularly clear in the move from the dominant protection association to the minimal state: in effect, this

-247-

says to anarchists (independents) that the (minimal) state is entitled to impose its authority on them without their consent.

In these terms, does Nozick's argument succeed? I believe it is deeply flawed, because Nozick grossly underrates the differences between the state and commercial firms selling protective services. It is not possible to derive the former as an extension or modification of the latter. The fault-line between the two shows up in many ways in Nozick's argument. For example, Nozick has to admit that he can give no reason why the dominant protective association would be entitled to force independents to accept its jurisdiction in relation to disputes with other independents (p. 109). Thus, the Nozickian 'state' has not actually achieved a monopoly over right-enforcement and (despite Nozick's protestations) is only dubiously a state. Again, Nozick notes that independents not wishing to pay the price charged by the dominant protective agency would be entitled to refuse to do so, at the cost of forgoing protective services (p. 113). Peal states, however, would prosecute such people for tax evasion, but would nevertheless otherwise protect their rights on the same basis as anyone else's. Yet again, Nozick notes that private protection agencies, including a 'dominant' one, would (in line with normal commercial practice) offer clients a variety of protective packages, at different prices, to cater to demand for more or less extensive or elaborate protection (p. 13): states, however, are supposed to provide the same protection of rights to all citizens.

Nozick's dominant protection association is, we see, much less like a state than he claims. In fact, it is in some respects a very strange animal, in so far as it is (or may be) the result of an amalgamation between a number of different commercial protection agencies competing with one another in the same geographical area. Nozick calls this outcome 'federal' but it is not, like a federal state, a federation of geographical units, but rather one of (competing) protection agencies, which continue to operate (and presumably to compete) in the same territory. The so-called federation is only a system of 'higher' courts set up to avoid conflict between protective agencies in cases 'about which they reach [or might reach] differing judgments' (p. 16). Such a system is nothing like a federal state or any kind of state. Yet when Nozick describes the dominant protection

-248-

association turning into the minimal state, it has mysteriously become a unified organization with a single body of clients. This sleight-of-hand is necessary to bridge the huge gap between a commercial protection agency and a state. In sum, Nozick's argument as to how a minimal state would evolve out of a Lockean state of nature is unpersuasive: thus, on his own terms, he has not justified the minimal state.

There is another kind of problem besetting Nozick's attempted justification of the minimal state. To see this, we must look more carefully and critically at the idea of a commercial protective agency. What services, exactly, do such agencies sell? According to Nozick, they perform the functions of detection, apprehension, judicial determination of guilt, punishment, and exaction of compensation (p. 13). Presumably they have at their disposal detectives, police forces, judges, and prisons. A client buys their services by paying, say, an annual premium, and calls on his agency when he considers his rights to have been violated. It is rather difficult to envisage these agencies working in a satisfactory way. For one thing, one characteristic police function--the routine patrolling of neighbourhoods, in order to discourage and detect crime--would not be offered by them, since this is not a service that can be sold to individuals (it is what economists call a 'public good'). Nor is It at all clear that a system of commercial protective agencies would be an effective deterrent to crime (or rights violation). Providing the full panoply of services described by Nozick would probably be extremely expensive, and might well be beyond the means of many, perhaps most, private citizens. If so, there would be no deterrent to violating the rights of the relatively poor (perhaps the majority) but only of the rich. Nozick, therefore, has no right to assume that his 'dominant protection association' is faced with only a small number of 'independents'. The association would in fact confront two categories of non-members: genuine independents, or anarchists, who have chosen not to join, but to enforce their own rights; and reluctant independents, who are simply too poor to buy the association's protective services, but would if they could. Incorporation-plus-subsidization, as prescribed by Nozick, is thus not coercion for the latter, but a bargain they would freely accept; hence the subsidization is not, as

-249-

Nozick argues, compensation for forced incorporation, but genuine redistribution. This makes it all the more difficult for him to justify the 'minimal' state, but not the more-than- minimal state.

Nozick's attempted justification of the minimal state underrates both the difference between the state and a commercial protective agency, and also the inadequacies of a regime of the latter type. In this sense, he takes the arguments of the anarcho- capitalists too seriously, and concedes too much to them. What he should have argued is, not that the state would evolve legitimately out of a Lockean state of nature, but rather that what would evolve, short of the state, would be so unsatisfactory that it would be justified to establish the state, if necessary imposing it by force on anarchists, so long as it provides equal protection of the rights of all. This conclusion, incidentally, is unaffected by the number of anarchists so coerced.

Nozick's Theory of Economic Justice

Nozick's theory of economic justice--'justice in holdings'--is the most discussed, and the most iconoclastic, part of his political philosophy. It is called by him the 'entitlement theory', and is contrasted with two types of (allegedly) erroneous theories. One such type consists of 'end-state' principles, which 'hold that the justice of a distribution [of wealth] is determined by how things are distributed . . . as judged by some structural principle(s)' (p. 153). From this point of view, two distributions are to be judged equally just as if they have the same structure, or 'profile', but different persons occupy different positions in the structure. Thus, 'my having ten and your having five, and my having five and your having ten are structurally identical distributions' (p. 154). Nozick's objection to end-state principles (such as Rawls's difference principle, or a principle limiting inequality) is that they are non-historical, i.e. they treat as irrelevant to the justice of a distribution how it came about, how individuals acquired the holdings they have. Nozick's entitlement theory consists of historical principles of justice.

Nozick's second class of erroneous theories is made up of what he calls 'patterned' principles of justice. Such a principle 'specifies that a distribution is to vary along with some natural

-250-

dimension(s)'. It has the form, to each according to his--', where the blank is filled in by 'moral merit, or needs, or marginal product', or whatever (p. 156). Patterned principles, though they may or may not be historical, are rejected because they are incompatible with any freedom whatever for individuals to use their wealth as they choose. If at a given moment a patterned distribution were established, it would of necessity be destroyed if persons could freely (and hence un predictably) transfer some of their holdings in purchases, gifts, loans, and so on. As Nozick sums up: 'Liberty upsets patterns, (p. 160). Nozick's entitlement theory prescribes no pattern.

What does it prescribe? First, a principle of Justice in acquisition, which specifies how unowned things may be appropriated (for the first time) by persons. Second, a principle of justice in transfer, which specifies how things already owned may become (legitimately) the property of other persons. According to the entitlement theory, a person is entitled to what he has if and only if he acquired it in accordance with the principle of Justice in acquisition, or has acquired it in accordance with the principle of justice in transfer from someone entitled to it. (The complication created by the issue of rectification of injustice is ignored for the time being, though it will later be seen to be important.)

What exactly are Nozick's principles of justice in acquisition and transfer? To the bewilderment of many commentators, Nozick declines to specify them in any detail (p. 153). Despite this, it is possible to give some account of them. Thus, it seems clear that, according to Nozick's principle of Justice in transfer, a person becomes the (new) legitimate owner of something already legitimately owned by another owner, if and only if the latter freely conveys it to that person. This transfer may take the form of a gift (charitable or otherwise), a bequest, a market exchange, etc. It follows, of course, that taxation by the state to provide welfare services, etc.--a forced transfer--is unjust. According to Nozick, such forced redistribution violates the Kantian imperative: it uses some persons for the benefit of others. It is morally on a par with theft.

It is harder to specify Nozick's principle of Justice in acquisition, but not impossible. Notoriously, and to the perplexity of many, Nozick discusses Locke's labour theory of acquisition

-251-

at length, only to reject it, without putting anything definite in its place. However, there is no doubt that for Nozick, as a general principle, labour creates entitlement. When new goods are produced, entitlement to them depends on how they were made. 'Whoever makes something, having bought or contracted for all other held resources used in the process . . . is entitled to it' (p. 160). If he uses no held (already owned) resources, he must then be entitled to what he makes out of unheld resources.

A similar conclusion is implied by Nozick's famous argument that 'taxation of earnings from labor is on a par with forced labor' (p. 169). Such taxation is 'seizing the results of someone's labor [which] is equivalent to seizing hours from him'; it forces him to devote some of his hours of labour to ends not his own. Those who so force him, says Nozick, treat him as if (during these hours) they owned his person and abilities. They flout 'the classical liberal notion of self-ownership' and embrace 'a notion of (partial) property rights in other people' (p. 172). Taxation of earnings from labour is not only like theft; it is like (partial) enslavement.

But taxation of earnings from labour is not the only way of seizing the results of someone's labor'. The result of a person's labour may be something he has made out of unheld resources. If that thing did not become his property, other persons would be entitled to seize it, and, in effect, become part-owner (by Nozick's argument) of its, maker. His position would be like that of a (partial) slave. Presumably, therefore, the maker must in such cases become the rightful owner.

This principle of just acquisition is also implied by an argument of Nozick against Pawls. Whereas Rawls conceives of justice as the right distribution of the fruits of social cooperation, Nozick asks us to imagine 'ten Robinson Crusoes, each working alone for two years on separate islands' (p. 185), a scenario neglected by Rawls, he thinks, because it constitutes 'a clear case of application of the correct theory of justice: the entitlement theory' (p. 186). Nozick's point is that each Robinson Crusoe is the rightful owner of what, working alone, he has produced (none has any claim on the product of another). AN in all, it seems almost certain that Nozick believes that a person who makes something out of unheld resources

-252-

is entitled to appropriate it. Presumably, this implies a right to appropriate unheld natural resources in order to make something. And in that case, it would seem illogical not to admit also a right to appropriate unheld natural resources simply in order to use them. Nozick's principle(s) of just acquisition do not, after all, seem to be so very different from Locke's.

It need hardly be said that Nozick's assertion of the property rights created by labour has nothing in common with socialism, or with any critique of capitalism. For one thing, he would undoubtedly include within 'labour' all forms of economic activity, including management. For another, it is obvious from his principle of just transfer that the 'earnings' to which labour creates an inviolable title are whatever others are prepared to pay for it, in the market. Furthermore, Nozick thinks it feasible and desirable to extend his argument against taxation of earnings from labour to 'interest, entrepreneurial profits, and so on' (p. 170)--that is, to the specifically capitalist sources of income. Although he does not give this extended argument, he nevertheless assumes it when he objects to principles of justice that 'give each citizen an enforceable claim to some portion of the total social product' (p. 171). No matter the kind of income or wealth taxed for this purpose, the redistribution involved amounts to partial ownership of some people by others.

As is already clear, Nozick denies that there exists a right to life, in the sense of a right to what is necessary in order to live. Such a 'right' would be a right to resources that may be justly held by others, according to the entitlement theory, and of which they cannot justly be deprived. Nevertheless, Nozick's entitlement theory is not--quite--a licence for unlimited appropriation without regard to the interests of others. The right of appropriation is limited by what Nozick calls the 'Lockean proviso' (p. 179). Locke imposed, as a condition on appropriation, that 'enough and as good' be left for others. This Nozick interprets to mean that appropriation must not worsen the situation of others. He then poses the question: 'Is the situation of persons who are unable to appropriate . . . worsened by a system allowing appropriation and permanent property?', and answers it by pointing to the numerous ways in which private property 'Increases the social product' and also provides 'sources

-253-

of employment' (p. 177). He thus concludes that in a free- market system the proviso '(almost?) never will come into effect' (p. 179). Not quite never, however. No one is allowed to appropriate the world's total supply of drinkable water, for example, presumably because it is a natural resource necessary to human life. Monopolistic ownership of such a resource would confer on its owner a degree of power which, if misused, could bring about what Nozick calls a 'catastrophe'. It is only in such cases of catastrophe, he says, 'that the question of the Lockean proviso being violated arises' (p. 181). And this is so because, in judging the effects of a private property system, the decisive point is that 'the baseline for comparison is so low as compared to the productiveness of a society with private appropriation'. If private property had not arisen, Nozick thinks, men would be so wretchedly poor that their situation would be little better than catastrophic. But Nozick does not think that, in practice, private property in accordance with the entitlement theory is likely to produce catastrophes. He concludes that 'the free operations of a market system will not actually run afoul of the Lockean proviso' (p. 182).

Assessment of Nozick's Theory of Economic Justice

One difficulty in appraising Nozick's theory of economic justice is that one may find it unacceptable simply because one's ultimate value-judgements differ from his. About such matters not much can be said, except to register disagreement. There are, however, many points at which Nozick is open to criticism on purely logical grounds. But before criticizing Nozick I shall mention one point on which he should, I believe, be defended against his critics.

This is his argument against 'patterned' principles of distributive justice. Nozick's argument shows conclusively, in my opinion, that all such principles must be rejected because they would destroy all freedom to dispose of one's wealth as one chooses. Some critics of Nozick have denied this. G. A. Cohen, for example, argues against Nozick that 'patterns preserve liberty'.4 However, his argument does not show this. The gist of

____________________

4 G. A. Cohen, Robert Nozick and Wilt Chamberlain, "'How Patterns Preserve Liberty'", in J. Arthur and W. H. Shaw (eds.) Justice and Economic Distribution ( Englewood Cliffs, NJ,. 1978), 246-62.

-254-

it is that untrammelled market exchanges in a capitalist economy lead to inequalities of wealth and therefore of power between social classes so great that subsequent market transactions cannot be described as free (the workers, who need access to the means of production in order to live, are forced to work for the capitalists, who own and control access to the means of production). This, if true, is an argument against untrammelled free-market capitalism, but not an argument for any patterned principle of justice.

It is essential, in this connection, to keep firmly in mind the difference between patterned and end-state principles of justice ( Nozick himself is extremely careless in this regard).5 A principle that outlaws extreme inequalities, or one that calls for any ,social minimum' (for example, Rawls's difference principle, or one that requires provision for the basic needs of all) is an end-state principle, not a patterned principle. These end-state principles require state redistribution, of course, but they do not destroy all freedom to dispose of one's wealth as one chooses. The success of Nozick's argument against patterned principles does not mean that his argument against all state welfare transfers is likewise successful. That remains to be considered.

One peculiarity of Nozick's argument is that, most of the time, it seems to presuppose a world consisting entirely of adults. The rights of children are not separately considered. Nozick's philosophy might be summed up as follows: no one has any (enforceable) obligation to help anyone else, no matter their need, unless they voluntarily consent to do so. Can Nozick apply this to children? Do parents not have an (enforceable) obligation, if they can, to look after their children? (The act of begetting children is not consenting to look after them.) If the parents are unable to look after their children, or are dead, do the children have no rights to be looked after? Does no one--their other relations, if any, or the wider community-- have any (enforceable) obligation to ensure that orphans are looked after? Nozick has to say that there is no such obligation. If he were to make an exception for children or orphans,

____________________

5See Anarchy, State, and Utopia, 156, where Nozick 'extend(s) the use of "pattern" to include the overall designs put forth by combinations of end-state principles'; and p. 209, where Rawls's difference principle, incomprehensibly, is called a 'patterned end-state principle'.

-255-

he would be hard-pressed to resist extending his concession to adults who are equally unable to look after themselves--for example, the chronic sick and disabled. But he is, we know, quite unwilling to have the state use its taxation powers to help such people.

These are matters of value-judgements. Let us turn to logical problems related to Nozick's argument against (welfarist) redistribution. These begin right at the beginning. Nozick, we recall, bases his entire moral and political philosophy on his notion of a 'meaningful life': a person's life is meaningful if and only if he or she is able to shape it in accordance with his or her own choices. Is it necessary to point out that extreme poverty inhibits and may even destroy the ability to lead a meaningful life, thus defined? Despite the high value Nozick places on the living of a meaningful life, he does not believe there is any right to its material preconditions, or even to the preconditions of life itself Yet meaningful life is certainly impossible without life. Does it then make sense to give private property rights absolute priority over life itself? It would be quite absurd to argue that a wealthy person, taxed as part of a redistributive scheme to help the poor, is thereby prevented from living a meaningful life. Although the value of living a meaningful life is supposed to ground Nozick's side- constraints morality, it cannot do so. That morality, therefore, is left ungrounded.

Another problem, noted by many of Nozick's critics, is the implications of the so-called Lockean proviso. In Nozick's view, as we saw, it has, in practice, virtually no implications at an, because the 'baseline' against which the entitlement system of private property has to be compared is a situation of such miserable poverty. What assumptions is Nozick making here? What (hypothetical?) situation provides the appropriate baseline, and why? Nozick, as often, admits he cannot give a fully adequate answer ('[The] question of fixing the baseline needs more detailed investigation than we are able to give it here' (p. 177)), but it looks as if his conception of the relevant baseline is one where no property rights whatever are recognized, where, in other words, everyone is considered free to use everything. No doubt such a state of affairs would be, as Hume argued, little short of catastrophic: but it is not clear why this is the

-256-

appropriate baseline for comparison. It appears that Nozick thinks so, because of the historical nature of his entitlement theory. He conceives of the world's resources as being initially unowned by human beings, then becoming owned in accordance with his entitlement theory, and then asks if this makes things worse. But surely this story is unhistorical. It is not very likely that the world's resources were first appropriated by individuals, working alone. The human race, in its earliest days, was not organized on capitalist or even individualist principles. If, as seems not unlikely, acts of original acquisition were carried out by groups--clans, tribes, or whatever-- recognizing a right of all members of the group to some share of the product, then--so long as no coercion was involved-- Nozick would have no reason to object to such an arrangement. And, if individual private property was preceded by some such system, should not the latter be the baseline for comparison? Notwithstanding the admittedly far greater productivity of a private property economy, the Lockean proviso might not then appear so nugatory as it does to Nozick. Despite its great productivity, the private property, free-market economy almost never produces a situation of full employment, and occasionally generates mass unemployment. Under Nozick's entitlement theory, the unemployed would (charity apart) have no income. They could starve to death. Is this not a 'catastrophe'? Can one be sure that persons starving to death in a wealthy Nozickian society are not worse off than they would be in a poor society differently organized? Of course, one can be sure of nothing where such counter-factuals are involved. Still, if one wants to be reasonably sure of satisfying the Lockean proviso, one should, as several commentators have argued, provide, out of taxation, at least a basic social minimum.

Discussion of the Lockean proviso illustrates what is perhaps the most fundamental problem with Nozick's theory of economic justice. It is very hard to see how, in practice, to apply it. According to Nozick, a distribution is just if and only if everyone is entitled to what they own, and this is the case now if the entire history of appropriation has conformed to Nozick's principles of justice in acquisition and transfer. As everyone (including Nozick) knows, this has not happened. Apart from theft, fraud, and coercive redistribution by governments, current

-257-

ownership has been much affected by past acts of military violence, warfare, invasions, and so on. According to some commentators, the establishment of capitalist private property itself depended, to no small degree, on the violent or fraudulent expropriation of previous, often collective, owners. To cope with such problems, Nozick adds to his principles of just acquisition and just transfer a third principle--the principle of rectification of injustice. This principle is at once important, and problematic in the extreme. To apply it, Nozick says, one must use 'historical information about previous situations and injustices done in them' (p. 152) to estimate the difference between the existing distribution of holdings and the (probable) distribution that would now exist if no injustice had taken place. If there is a difference, the best estimate of the latter distribution 'must be realized' (p. 153). As Nozick probably realizes, this is a hopeless enterprise. He is thus led to a quite surprising conclusion: 'One cannot use the analysis and theory presented here to condemn any particular scheme of transfer payments, unless it is clear that no consideration of rectification of injustice could apply to Justify it' (p. 231).

But it can never be clear what should be done now to rectify the accumulated effects of centuries of past injustices. It thus turns out that Nozick's theory of economic Justice is, indeed, purely a theory, without any practical application. If we agree with it, we learn from it what the human race, throughout its history, ought to have done, but not what we should do now.

The Minimal State and Utopia

In view of the conclusion of the previous section, it is perhaps not surprising that the last part of Nozick's book portrays the minimal state as a utopia (or, more accurately, a framework for utopia). However, this is not an admission of its unrealizability or lack of practical relevance. Rather, Nozick wishes to show that 'the minimal state is inspiring as well as right' (p. ix).

Nozick writes that the subject of this last section of his book is 'the best of all possible worlds' (p. 298). He notes, rightly, that not all goods--not all desires--can be simultaneously realized. 'The best of all possible worlds for me will not be that for you.' Utopia, therefore, must be, in some sense, 'best for

-258-

all of us; the best world imaginable, for each of us'. What sort of society does this imply? Accoding to Nozick, one which has 6 a wide and diverse range of communities which people can enter if they are admitted, leave if they wish to, shape according to their wishes; a society in which utopian experimentation can be tried, different styles of life can be lived, and alternative visions of the good can be individually or jointly pursued' (p. 307). The more diverse the range of communities available, the more people will be able to live (nearly) as they wish to live. As Nozick puts it, 'utopia will consist of utopias' (p. 312)--'no one can impose his own utopian vision on others'. The minimal state, Nozick claims, provides a framework for this utopian society.

However, I cannot see why the minimal state is a better framework for Nozick's utopia than a state which imposes taxes to provide a basic social minimum. Formal freedom, perhaps, is greater in the minimal state: but the capacity to live as one chooses, or in the type of community one would wish to live in, depends also on economic resources. If, as Nozick says, the aim is to create a world which is 'best for all of us' (emphasis added), this creates a strong argument against the minimal state, and for extensive redistribution, perhaps along the lines of Rawls's difference principle.

Nozick, however, makes a striking (and perhaps surprising) claim for his minimal state utopia. It 'has many of the virtues, and few of the defects [that] people find in the libertarian vision': 'For though there is great liberty to choose among communities, many particular communities internally may have many restrictions . . . which libertarians would condemn if they were enforced by a central state apparatus' (p. 320). Thus 'communities' (though not the state) may ban capitalism, or enforce economic redistribution among their members, refusing to allow individual members of the community to opt out of their arrangements. Why? Nozick cannot, and does not, argue that the relevant difference between communities and states is that people have joined communities voluntarily, since it is obviously possible to be born into a community, as into a state. Nor does he argue that one is free to leave a community --perhaps because he would then have to allow redistributive states so long as they too leave this freedom to their citizens.

-259-

So what is (are) the relevant difference(s)? After wrestling with this problem for about a page, Nozick concludes, with great candour: 'I do not see my way clearly through these issues' (p. 323). Nevertheless, he does not withdraw his previous claims about communities.

Conclusion

One of the great virtues of Anarchy, State, and Utopia has just been illustrated--Nozick's disarming honesty. Few political philosophers--probably few philosophers, or intellectuals of any kind--are as explicit and open as Nozick is in acknowledging the gaps and weaknesses in his arguments. This makes the job of a critic that much easier, and I have not scrupled to take advantage. Very early in his book, Nozick frankly tells the reader that it will not present 'a finished, complete and elegant whole', with all details thought through (p. xii). He adds: 'I believe there is also a place for a less complete work containing unfinished presentations, conjectures, open questions and problems . . .'. One can only agree. Anarchy, State, and Utopia is a book written in an unusual style, remarkable not only for its intellectual verve, but for the very high proportion of sentences which take the form of questions--questions which are often not answered, but are (almost always) good questions. It also includes many interesting discussions not mentioned in this survey, because peripheral to the main argument (the most famous and brilliant of these is probably on the 'experience machine' (pp. 42-5)).

But what, finally, is the value of Nozick's political philosophy? In my opinion the most valuable part is also the least acceptable, namely the theory of economic justice set out in part 11. The reasons why it is not (in my opinion) acceptable have been stated above. But its value is to present a strong and impressive case for private property rights. One need not agree with Nozick that (in an ideal world) these would take precedence over every other consideration; none the less, one may be convinced by Nozick's arguments that there have to be very good reasons for overriding them. Vague and implausible goals like 'equality' are not enough; nor is the realization of some 'patterned' principle of distribution. Nevertheless, good reasons

-260-

--such as the relief of suffering, and provision for the needy--- do exist.

Nozick's iconoclasm is also useful. He forces us to face up to such questions as, for example, what the (moral) difference is between theft and state redistribution. Why is state redistribution not theft by means of a powerful third party? Or would theft sometimes be morally Justified? Arguably it would, in cases where it is necessary to prevent starvation or great suffering and does relatively little harm to the person stolen from. Can state redistribution be justified in cases where theft cannot? If so, why? Anarchy, State, and Utopia provokes us into asking, and trying to answer, questions such as these.

Bibliographical Note

Anarchy, State, and Utopia has attracted at least one book-length commentary, Jonathan Wolff Robert Nozick ( London, 1991). This is an excellent discussion of the whole of Nozick's political philosophy, sympathetic but ultimately sceptical. Other commentaries are briefer-- articles, chapters, and reviews--and (except for the reviews) usually concentrate on only one aspect of Nozick's theory, either the justification of the minimal state, or (much more commonly) his libertarian theory of economic justice. However, twenty of these briefer discussions have been usefully collected between two covers by Jeffrey Paul in Reading Nozick: F-ssays on Anarchy, State, and Utopia (Totowa, NJ, 1981), to provide a general overview.

Most of the essays in Reading Nozick are respectfully critical and most are on the theory of economic justice. Perhaps the best is Samuel Scheffler , "'Natural Rights, Equality, and the Minimal State'", which presents a careful argument deriving universal rights (and duties) from Nozick's premiss of meaning life. Thomas Nagel, in "'Liberalism without Foundations'", gives reasons why a rational person would welcome legislation to compel him to fulfil such duties. Cheyney Pyan, in "'Yours, Mine and Ours: Property Rights and Individual Liberty'", presents an interesting but, I think, unsuccessful counter-argument to Nozick's attack on patterned principles of distribution. A careful consideration of the circumstances in which invasion of property rights is and 15 not justified is in Judith Jarvis Thomson "'Some Ruminations on Rights'", while Onora O'Neill, in "'Nozick's Entitlements'", argues convincingly that Lockean premisses do not Yield Nozickian conclusions on absolute

-261-

property rights. David Lyons, in "'The New Indian Claims and Original Rights to Land'", provides a rare discussion of Nozick's principle of rectification of injustice. An attempt to provide a grounding for Nozickian property rights is Eric Mack, "'How to Derive Libertarian Rights'".

Quite a number of books on political philosophy devote a chapter to Anarchy, State, and Utopia. For example, Chapter 10 of Tony Honore Making Law Bind ( Oxford, 1987) takes Nozick to task for the parochiality of his 'liberal Western' conception of private property rights. Gordon Graham , in chapter 3 of Contemporary Social Philosophy ( Oxford, 1988), is generally sympathetic to Nozick's theory of economic justice, but doubts whether market transactions between extremely unequal parties are genuinely free. Less sympathetic to Nozick is Jeremy Waldron, chapter 7 of whose The Right to Private Property ( Oxford, 1988) contains a penetrating critique of Nozick's principle of just acquisition of unowned resources--namely, that it involves loss of rights by others consequent on unilateral action of the appropriator, on terms they would not accept.

Some more articles critical of Nozick may be mentioned. J. Exdell, in "'Distributive Justice: Nozick on Property Rights'" ( Ethics, 87 ( 1976-7), 142-9, challenges Nozick's assumption that prior to their private appropriation resources are (morally) unowned--perhaps individuals generally have a human right to benefit from them, which private property rights should respect. J. R. Kearl ( "'Do Entitlements Imply that Taxation is Theft?'", Philosophy and Public Affairs, 7 ( Fall 1977), 74-81, suggests that the advantages to property owners from the state's protection of property rights entitles it to tax them for redistributive purposes; while M. H. Lessnoff castigates Nozick for his confusion of patterned and end-state principles in "'Capitalism, Socialism and Justice'" (in J. Arthur and W. H. Shaw (eds.), Justice and Economic Distribution, Englewood Cliffs, NJ, 1978). Finally, Charles Sayward and Wayne Wasserman, in "'Has Nozick Justified the State?'" (in J. A. Corlett (ed.), Equality and Liberty: Analyzing Rawls and Nozick (Basingstoke, 1991), 261-7), conclude that he has not. Corlett's collection also contains another five articles on Nozick.

-262-

12

Ronald Dworkin: Taking Rights Seriously

PAUL KELLY

JOHN RAWLS, Robert Nozick, and Ronald Dworkin together form the trinity of American liberal political philosophers which has dominated much Anglo-American political thought since the 1970s. Dworkin made his reputation with his book Taking Rights Seriously, which was published in 1977. Unlike Rawls A Theory of Justice or Nozick Anarchy, State, and Utopia, it was not written as a book. Although the arguments of each chapter are intricately connected, most of the chapters were first published elsewhere. In this sense Dworkin's book is much closer in style to Oakeshott Rationalism in Politics and Other Essays or Berlin Four Essays on Liberty than it is to the works of his American colleagues. As with Berlin and Oakeshott, Dworkin's adoption of the essay style allows him to develop an argument that is more than merely the sum of the individual chapters.

The philosophical position developed in Dworkin's book has provided the base for a host of subsequent books and articles which have further enhanced his reputation as one of the foremost theorists of a right-based liberal political order, and, in jurisprudence, as one of the most influential critics of legal positivism. Nevertheless, Taking Rights Seriously remains Dworkin's most important book. This is in part due to its style, in part due to its direct engagement with practical issues. Although Dworkin holds prestigious chairs in jurisprudence at Oxford University and New York University, and although five chapters of the book first appeared in law journals, he was not writing for a purely academic audience. Instead his concerns were often unambiguously political, showing how certain theses about law or constitutional interpretation or the theory of rights have a direct bearing on our understanding of the values inherent in a liberal democratic society. An important element of his rejection of legal positivism is precisely that

-263-

it separates legal philosophy from other aspects of political theory. For Dworkin, being a political philosopher or a philosopher of law involves engagement with practical life and, given the centrality of law to modern liberal democratic societies, it is, in his view, quite proper for the philospher of law to contribute to debate in the public realm of political activity. It is this willingness to connect philosophical questions with matters of public concern that makes his book important as a conception of what a political philosophy can do. In it Dworkin outlines a public philosophy for a society that takes rights seriously--that is, he outlines a genuinely liberal theory of law.

A Liberal Theory of Law in Place of the Ruling Theory

Dworkin brings together the varied themes and arguments of each chapter of Taking Rights Seriously under the heading of 'a liberal theory of law'. In his introduction he claims that his liberal theory of law was developed at a time when liberalism as a tradition of political thought was facing considerable hostility. The liberal theory of law that Dworkin develops forms part of the wider dissatisfaction with the received liberal tradition of utilitarian ethics and legal positivism. His own position is 'sharply critical' of legal positivism, which is conventionally seen as the paradigmatic liberal theory of law. Legal positivism has its roots in the philosophy of the eighteenth-century English legal philosopher Jeremy Bentham ( 1748-1832), and has had such a pervasive influence on modern legal theory and liberal political ideas that Dworkin dubs it the 'ruling theory of law'. As he goes on to argue throughout the rest of this book, much of the dissatisfaction that liberalism has engendered has been a result of the theoretical inadequacy of the 'ruling theory'.

The 'ruling theory' has two parts: first, a theory about the necessary and sufficient conditions for the truth of propositions about law--that is, a theory of what the law is; secondly, a theory of what the law ought to be. As Bentham argued in A Fragment on Government ( 1776), jurisprudence can be divided into the expository and the censorial. Expository jurisprudence is concerned with identifying what the law is, which is a straightforward scientific enterprise that does not appeal to moral

-264-

standards. For Bentham and subsequent legal positivist philosophers, it is possible to identify true propositions about English law by appealing to certain institutional facts about the authoritative source of legal commands. The second task of jurisprudence, for Bentham, is that of the censor, whose job it is to argue for the reform or transformation of the law, and whose tool for identifying what the law ought to be is the principle of utility. These two tasks are logically distinct and, according to Bentham, the confusion of these two goals has dangerous consequences.

Dworkin's theory challenges both parts of this 'ruling theory', not only in its Benthamite guise, but also in its later manifestations such as H. L. A. Hart's legal positivism and modern utilitarian theories of political morality. He contests the logical separation of legal, political, and moral theory which underlies Bentham's distinction. Whilst propositions about the law are not the same as propositions about morality, there are nevertheless significant connections which are derived from the role of rights in both legal and political or moral argument. If we take the role of rights seriously, and reject Bentham's view of. them as merely 'nonsense upon stilts', then, according to Dworkin, we will see that an adequate theory of law is an extension of political and moral theory rather than logically distinct from it. Given the centrality of rights to Dworkin's argument, he also rejects the utilitarian view that the general welfare can override individuals' basic interests. Instead he is concerned with restoring the concept of rights within liberal political thought, by defending a view of rights as 'political trumps' which individuals hold against the pursuit of collective goals which impose loss or injury upon them.

The success of Dworkin's enterprise, and his claims for the continuity of legal, political, and moral philosophy, depends on his being able to defend a conception of rights which bridges the divide between the logical and normative components of law. Traditionally, arguments which have attempted to bridge this gap have appealed to pre-political moral or natural rights, precisely the sorts of rights that Bentham, Burke, Marx, and a host of other philosophers dismissed as metaphysical nonsense. According to positivists, the problem with such natural-rights arguments is that one can engage in meaningful

-265-

Rules, Principles and Policies: The Critique of Positivism

Dworkin's critique of legal positivism is subtle. Unlike the standard natural-law view, Dworkin's argument does not merely construct a moral conception of law, neither does it deny the positivist's claim that a satisfactory account of law must make appeals to authoritative institutional sources--that is, statutes, precedents, and legal decisions--nor, most importantly, does it

-266-

deny that one can decide legal questions in clear cases by appeals to such valid rules. The rules that positivists place so much weight on do, indeed, play an essential role in any account of the law. However, he denies the positivist's claim that the idea of law as an authoritative system of rules is a sufficient account of law.

There are three basic features of legal positivism which Dworkin challenges. There is, first, the idea that law is a special set of rules for the use of public power, and that these rules can be distinguished from other social rules, such as moral rules or customs, by their 'pedigree' rather than their content. Secondly, and most importantly, there is the idea that this set of authoritative rules is exhaustive of the law, and that, if a case cannot be settled by applying any of these rules, then its settlement is an example not of applying the law, but rather of official discretion. As Hart argues in The Concept of Law, there are cases where the rules run out, and in such cases a judicial decision becomes a case not of applying the law but of making new law by exercising discretion. Thirdly, there is the view that in such cases it does not make sense to argue that the judge is enforcing a legal right. In other words, in cases that do not fall under a valid legal rule, neither plaintiff nor defendant has a right to a particular decision and legal rights and obligations exist only in uncontroversial cases. Thus, when a judge is trying a 'hard case', his decision will, according to the positivists, involve making new laws and extending new rights that did not previously exist.

The first part of Dworkin's assault on positivism takes the form of a philosophical account of judicial adjudication. Whereas positivists take legislation to be paradigmatic of the law, Dworkin takes adjudication, or what judges do in deciding cases, and by focusing on adjudication he argues that we can see that the model of rules provides an inadequate account of what law is like.

When judges are confronted by 'hard cases' that require adjudication, they do their utmost to appeal to rules, but they will also often appeal to standards that are not rules, standards that Dworkin calls principles and policies. The distinction between principles and policies is complex and controversial, and Dworkin gives a number of different accounts of the

-267-

distinction. Basically a principle is a standard observed as a requirement of justice or morality. Later on he gives an account of principles which is connected with rights. Principles are individuated claims on behalf of groups or individual persons that can be made against the pursuit of policies or other social goals. Principles concern rights. Policies, on the other hand, concern social goals that the community wishes to achieve. Wealth maximization or the maximization of utility would constitute a policy.1 An example that Dworkin gives of a principle is that used to decide the case of Riggs v. Palmer in 1889 in the state of New York. The case was brought to decide whether the heir to a will could inherit even though he had murdered his grandfather in order to claim the inheritance. Although the statutes regulating the making, proof, and effect of wills did not make a distinction to cover such cases, the judge decided against the plaintiff, by appealing to the principle that 'no man may profit from his own wrong' (p. 26).2 Principles and policies are both standards used in adjudication. The difference between them is a logical one, rather than a matter of content. Principles are not special cases of rules. A principle like that given above is different in kind from a legal rule such as 'A will is invalid unless signed by three witnesses'. A rule like the latter applies in an all-or-nothing fashion. If three persons have witnessed a will, and the rule that three persons must witness a will is valid, then the answer supplied by the rule must be accepted. If the rule is not valid, then it plays no role in a decision at all. In the case of such valid rules, the judge is left with little room to exercise his judgement about the outcome. But in the case of principles, things are different. A principle has to have a weight attached to it; in adjudicating a case, therefore, it does not apply in an all-or- nothing fashion. In some cases, such as Riggs v. Palmer, the weight of the principle will outweigh adherence to the rules pertaining to the proof and effect of wills. Because rules do not have the dimension of weight or importance, it is not possible to see principles as a subset of rules. If we think of the analogy

____________________

1See R. Posner, The Economics of Justice ( Cambridge, Mass., 1981), for the view that judges ought to maximize wealth through their decisions.

2 R. Dworkin, Taking Rights Seriously ( London, 1977; 2nd corrected edn., 1978). References in the text are to the second edition.

-268-

of a game, an umpire can determine whether the rules have been broken or when specified exceptions occur, but he is not generally empowered to judge whether the rule applies or not. However, in the case of principles, the judge does have to decide the weight of a principle in a given case, and on a case- by-case basis. Consequently, the positivist's argument for law as a body of rules is inadequate.

The second part of Dworkin's critique is an assault on the idea that a non-normative account of the law can be given by appeal to the source or pedigree of law. Hart, in The Concept of Law, had argued that the law can be divided into a system of primary and secondary rules, the primary rules being the basic rules of law prescribing actions and conferring obligations, which are identified by appeal to a secondary rule of recognition.3 A secondary rule is identified when the 'practice conditions' of the rule are met--that is, when individuals behave in a certain way. The rule exists not because judges or functionaries believe that they have a moral obligation to recognize only certain sources of rules as sources of law, for this would thereby involve normative political questions of what sources can legitimately give rise to law. Rather Hart assumes that the mere fact that judges act in a certain way constitutes the secondary rule. This secondary rule is a 'social' rule, because it is not conferred from some external authority, but are instead constituted by a certain social practice, and it is from this alone that it derives its authority. Thus a judge behaving in a certain way exhibits through his actions that he is under an obligation to recognize only certain sources of law as authoritative. For positivists the existence of this secondary rule of recognition is simply a matter of fact exhibited by conformity to the practice conditions of the rule--that is, behaving in a certain way.

In challenging the account of the secondary rule of recognition as a 'social' rule, Dworkin draws a distinction between 'social' rules that are constituted by a practice and 'social' rules that are justified by the practice. He argues that Hart's theory works only if the content of this 'social' secondary rule of recognition is the same as the practice which gives rise to these rules. That is, the 'social' rule of recognition which specifies

____________________

3 H. L. A. Hart, The Concept of Law ( Oxford, 1961), 97.

-269-

that only certain authoritative sources give rise to law will only be a genuine source of obligation to recognize only those sources, if it is indeed the case that only those sources can give rise to law. But, according to Dworkin, this begs the question at issue. For this 'social' secondary rule of recognition win only be obligation-creating if it does indeed constitute the practice of law. Why should we recognize that what judges and functionaries do is both all that law is and all that it can be? For the social-rule theory to work, Hart must assume that all that law can be--that is, the full practice of law--is captured in his account of these two sets of primary and secondary rules. Hart has not provided an argument as to why the social-rule account of the secondary rule of recognition can be the only criterion of the source of law. Contrary to Hart, Dworkin argues that what legal theories do is provided contestable characterizations of what the practice of law is, each of which is justified by the practice of law. The practice cannot be fully comprehended by Hart's mechanistic social-rule account, for this would rule out the sort of argument that takes place within the practice of law over what are authoritative sources of law-- that is, precisely what rule of recognition is justified by the practice. On this question Hart answers that only certain institutional sources give rise to law, whereas Dworkin argues that the resources of a community's morality are also necessary for identifying law. Both authors claim that their theory is justified by the practice of law.

The other key point to Dworkin's argument is that, if the social-rule theory is true and that genuine obligations can be inferred only from regularities of behaviour constituting social practices, then the only obligations we can have are conventional obligations constituted by such practices. The problem with this account is that it ignores an important sense in which we use the concept of obligation and duty in moral language to distinguish our moral duties from what convention tells us to do. It is often the case that appeals are made to duties and obligations in moral and political arguments for the express purpose of criticizing conventional moral rules. If Hart's social thesis is correct, then this would considerably impoverish our moral language and conduct, for his account of the authority of law could equally apply to the practice of morality.

-270-

The third component of the critique of positivism concerns the nature of judicial discretion. Positivists assert that in cases where the rules run out the judge can often be required to exercise his discretion, and, given that this is appropriate only in cases where the rules have been exhausted, it is properly understood not as applying the law, but rather as making new law and creating new legal rights.

Dworkin's response is to distinguish three senses of discretion which are confused by the positivists. In the first case a person has discretion if his 'duty is defined by standards that reasonable men can interpret in different ways' (p. 69). A person has discretion in a second sense if no higher authority may review or overturn his decision. Finally, a person can have discretion where a set of standards which impose duties on him do not impose a duty to make a particular decision. According to Dworkin a judge might regard himself as having discretion in the first and second cases whilst still considering his discretion as raising a question about what his duty as a judge is. It is only if we acknowledge that duties and obligations must be derived from 'social' rules that the judge would have a strong discretion of the sort positivists claim. But, as we have already seen, there is no good reason to believe that obligations and duties can be derived only from 'social' rules. The positivist's conception of law distorts his account of what judges actually do in adjudicating 'hard cases'. If we can have duties which are not wholly derived from 'social' rules, then there is no difficulty in claiming that, although the rules might have been exhausted, in 'hard cases' judges are nevertheless able to decide on the basis of law, and reason about legal obligations, rather than make purely arbitrary decisions.

This last point is crucial not only because it challenges the positivist account of what judges actually do in deciding such cases, but also because it has significant political consequences which the positivist ignores. If judges are to be understood as making purely arbitrary non-legal decisions in 'hard cases', by appealing simply to their own preferences, feelings, or instincts, then a question immediately arises. Why should we allow judges to make such privileged decisions rather than leave them to accountable political institutions? This is an issue for any liberal democratic community, but is perhaps most

-271-

vital in the USA, where a non-elected and non-accountable Supreme Court is empowered to make very significant decisions about individual rights, the interpretation of the Constitution, and the bounds of legislation. If we consider the landmark case of Roe v. Wade, which established a woman's right to choose to have an abortion, and if, following the positivists, we interpret this as Judicial legislation or the creation of a new right, then we can legitimately ask what gave those judges the authority to legislate on such a controversial moral question? Why are their opinions or moral beliefs a superior criterion for settling the issue than the political beliefs of the majority? The positivist thesis concerning judicial legislation raises the question of why unaccountable judges should be allowed to legislate, and also why individuals should regard judges' opinions as legally binding when they have based their decisions not on a matter of law but on some other standard such as their moral preferences. All the positivist can say is that it is merely a fact of our community that we do allow courts to legislate. But this will certainly not persuade those who regard abortion as a fundamental issue of morality, and who might not think convention is particularly weighty in matters of life and death.

To answer such questions we need an account of adjudication which connects with a wider moral and political theory, and is able both to ground legal rights, and justify certain decisions being confined to the courts. This leads us to Dworkin's right-based account of adjudication and political morality.

A Herculean Task: Taking Rights Seriously

Dworkin needs to provide an account of adjudication which connects with a wider moral and political theory, but in constructing such a theory he is constrained. Unlike standard natural-law theory he cannot construct and defend an external theory of moral rights which underpins the law, for he has to show that judges should appeal to right-based arguments in adjudicating hard cases, and that such a theory is the best account of how judges actually adjudicate. The theory of legal rights will be an account of the practice of law, and this is

-272-

crucial because his purpose is to show that positivism mistakes what judges do. He is constrained still further by the important additional requirement of showing that arguments for legal rights are 'matters of principle' rather than 'matters of policy', for an important rival school of law--the law-and-economics movement which has grown up around the work of the American jurist Richard Posner and the British economist Ronald Coase--has attempted to show that judicial decisions in key landmark cases can be explained in terms of the policy of maximizing wealth. If the law-and-economics movement is right, then the best account of our current legal practice win be based on utilitarian policies of wealth maximization; hence their adoption of an economic model of rationality for explaining legal practice and their advocacy of judges applying a version of the utilitarian calculus in adjudicating 'hard cases'.

Alongside Dworkin's critique of positivism there are two further themes which underlie the remainder of his book, and which dominate all his subsequent writings; first, an account of how judges actually decide 'hard cases' by appeal to principle; and, second, a critique of utilitarianism as a rival account of the 'community morality' underlying the practice of law. The latter question will be addressed in the next section.

How does he show that judges appeal to principle in adjudicating hard cases? The argument is in three parts. First, Dworkin repeats his crucial distinction between principle and policy. An argument of principle 'shows that some political decision advances or protects some individual or group right', whereas an argument of policy 'advances some collective goal of the community' (p. 82). He then argues that, although the distinction is controversial, it is the 'distributional character' of arguments of principle which distinguishes them from arguments of policy. The second part of the argument focuses on principle and democracy. If judges are given strong discretion to legislate, and they do so by appeal to policy, then the following question naturally arises in a democracy: what gives unelected public officials such as judges the right to decide political questions? Unlike other political representatives, judges are neither elected, nor ordinarily politically accountable for their decisions. Decisions of policy involve the weighing and balancing of different interests in identifying the common good.

-273-

The political institutions of a democratic state provide a forum in which public discussion and debate can take place to identify these interests. Matters of principle, on the other hand, are not decided by balancing such interests, and, if the question at issue is one of determining rights, then it is not clear that democratic institutions with their reliance on compromise, bargains, and majority decisions are the best place for settling such issues. The very point of appealing to rights is that they have what Dworkin calls a 'threshold weight' against collective goals or majority decisions. If one has a right to free speech, then that right must have a 'threshold weight' sufficient to trump any consideration of collective good which might be suggested to limit free speech. It would not make sense of our use of rights to argue both that an individual has a right to freedom of speech and that the government can restrict that right whenever its exercise becomes inconvenient. This fact provides a good reason for protecting those institutions which protect and enforce rights from the vicissitudes of majoritarianism. The third part of the argument attacks a two-stage view of judicial decision-making. The standard model claims that judges enforce legal rights until the rules run out, and then appeal to their own beliefs and values. In contrast, Dworkin argues that judges are bound by a conception of political responsibility: 'This doctrine states, in its most general form, that political officials must make only such political decisions as they can justify within a political theory that also justifies the other decisions they propose to make' (p. 89). This conception underlies the practice of appealing to precedent where judges have to show how their decisions are consistent with an interpretation of other decisions already taken.

Dworkin also needs to identify the rights which the judge should appeal to in deciding such 'hard cases' as legal rights. This he does by distinguishing between 'background rights' and institutional rights, and then between institutional rights and specifically legal rights. Each has the same normative character, but they have a different specification. Background rights are general rights that justify a political decision by a society in the abstract, but do not specify who should enforce them and how. Institutional rights, on the other hand, provide a justification for a decision by some person or institution. Background

-274-

rights are abstract, whereas institutional rights are concrete-- that is, they are rights which are clearly defined so that they can be shown to 'trump' particular political alms and decisions. Institutions, in their turn, can be either fully or partly autonomous. The game of chess is autonomous in the sense that one cannot appeal to morality to argue for an institutional right to a decision in chess. Institutional rights in such a case are given by reference to the rules which are constitutive of the game and their authoritative interpretation. However, unlike chess or baseball ( Dworkin's favourite examples), law is only partly autonomous, for the rules of legislation are not generally considered sufficient to decide questions of right--especially not in 'hard cases'. An individual can appeal to wider morality to claim that a certain right should be enacted. So law is in part continuous with political morality and not fully autonomous. However, it is important to recognize that institutional legal rights are nevertheless rights, and should be enforced even where they diverge from a political theory's background rights. The claim that some institutions are semi-autonomous is important, as positivists claim that all institutions are autonomous, with the corollary that questions of law are conceptually distinct from questions of morality. How we distinguish between autonomous and semi-autonomous institutions is not, as positivists claim, a matter of conceptual analysis, but a matter of critical reflection on the character of the practice. This involves a philosophical interpretation of the character of the practice, which is not only a logical construction, but also an appropriate phenomenology of adjudication recognizable to lawyers. Although law is only a semi-autonomous institution, it is not the case that judicial argument can arbitrarily introduce moral and political claims, so that the question of what law is and what law should be collapses. The judge does not appeal to external moral principles in the way that some traditional natural-law theories suggest. Dworkin does not want to collapse the question of what rights should be enacted in law, into the distinct but equally normative question of what rights an individual has in a particular 'hard case'. The first case is a matter of legislation, and consequently the appeal to moral principle is different from that in adjudication. The judge is concerned with what rights an individual has. Thus, when

-275-

Dworkin's fictional super-judge Hercules is deciding a 'hard case', he is not appealing to external moral principles which would change the law, but looking for a principled decision which is consistent with other key decisions, precedents, and rules in showing what the law is. When making his decision, Hercules can incorporate an appeal to background moral and political rights, but the appeal is one that applies these in giving the best account of the practice. The virtue that guides the judge in adjudicating 'hard cases' is 'integrity'. He must respect the integrity of the practice and provide the best possible account of it when making an appeal to moral principles to account for specific institutional rights. What is wrong with traditional natural-law theory is that the appeal to moral principle is not rooted in the best account of the practice. Such principles are not shown to provide a more coherent account of the practice; instead they are used to show the weaknesses of the practice. What makes Hercules' task a model of adjudication is that he is trying to show the connection between background rights and their institutional implications--that is, how they give rise to legal rights. The Herculean task of the judge is to show the continuity between the best account of law and background moral and political rights--that is, to show how they are connected and how this connection provides an account of key decisions and principles. This is a complex procedure that involves an appeal to legal history, precedent, convention, and an account of the role and function of law within a given political community. At its most complete, as practised only by the super-judge Hercules (who has both the time and the expertise to make these extremely complex decisions), the process Will involve giving a full account of what law is, how the decision he makes is consistent with that general account, and how it is also consistent with the normative aspirations of that community.

In doing this the judge is doing more than exercising discretion or legislating. He is responding to an incompleteness in a particular conception of law which has to be resolved. Where the positivist claims that, when the rules run out, law ceases and discretion takes over, Dworkin argues instead that it is not law that has run out, but merely a conception of law. The positivist's rule-based conception of law is merely one contestable

-276-

account of the law. The task of adjudication in 'hard cases' is to provide an alternative conception of the concept of law that will take the practice on further.4 The implication of this view-- namely, that legal argument in 'hard cases' consists of articulating alternative conceptions of law--explains why Dworkin rejects 'strict constructionist' interpretations of the law or the US Constitution. Some politicians (most notably President Richard Nixon) criticized various Supreme Court judgments for incorporating political arguments into judicial decisions and relying on 'political' philosophies of law, rather than the law itself. Dworkin's response to such arguments is that the law consists of a complex institutional structure of rules, principles, practices, and precedents, and that any coherent account of that complex practice can only be a 'conception' of law. So the 'strict constructionist' is providing a 'conception' of that practice just as much as the 'liberal'. The important difference is that Dworkin's 'liberal' is explicit about the connection between political theory and judicial decision-making in his account of law, whereas the 'strict constructionist' disguises his conservative political agenda through trying to discover the intention of the 'founders' or historical legislators. The view that we should be bound by the historical intentions of the founding fathers of the US Constitution is just as much a 'conception' of law as any other, which calls for justification as the best interpretation of the practices of law and its connection with the community's wider political and moral aspirations.

Both the positivist's theory of judicial discretion and the conservative's hostility to the liberal's use of political theory masquerading as law rely on an inadequate understanding of the function of moral reasons in arguments from principle. Their shared assumption that moral beliefs are subjective preferences gives rise to the dilemma of leaving controversial cases to courts: when it comes to abortion or desegregation, why should the moral preferences of a group of liberal Supreme Court justices have greater authority than those of the 'moral' majority? If in appealing to moral beliefs the judge were doing

____________________

4This distinction between a concept and its conceptions is derived from both W. D. Gallie, "'Essentially Contested Concepts'", Proceedings of the Aristotelian Society, 56 ( 1956), pp. 167-98, and J. Pawls, A Theory of Justice, Oxford, Oxford University Press, 1972, pp. 3-6.

-277-

no more than saying, 'the law has run out, what do I think?', there would be some justice to the claim. But this is not how judges appeal to moral principles. The judge's appeal is not to personal preferences but to the key moral intuitions which underlie the law and society's conception of the function of law. Both sets of critics ignore the way in which moral principles function in such arguments: unlike bald assertions, they are arguments which make truth claims. Dworkin challenges the belief that moral and political assertions cannot be objects of knowledge because they are merely expressions of preference. His denial that moral assertions are expressions of preference rather than objects of knowledge connects his jurisprudence with his political philosophy and in particular with his critique of the other key component of the 'ruling theory of law'--utilitarianism.

Rights, justice, and Utilitarianism

Utilitarianism is an essential component of the 'ruling theory of law' because it follows from the belief that all moral assertions are merely expressions of subjective preferences: the product of will not reason. Modern utilitarianism is but the most sophisticated and successful version of arguments that have their roots in David Hume's rejection of the possibility of morality being derived from reason. Dworkin's cognitivist account of judicial reasoning depends upon the view that moral reasons can be given greater weight than is provided by the emotivism or subjectivism of modern utilitarianism. However, utilitarianism poses a further and more recent problem for Dworkin, for a rival school of law, the law-and-economics movement, has argued that the best account of what judges do in deciding 'hard cases' is given in terms of policy-based arguments for wealth maximization. If Dworkin's account of right- based adjudication is to stand, he needs to show that policy-based utilitarian arguments concerning wealth maximization fall to capture the underlying public morality of law and of the wider community.

Dworkin's characterization of the right-based morality which underlies our use of the notion of rights in political and legal discourse is most clearly articulated in his extended discussion

-278-

of John Pawls's conception of the social contract in his A neory of Justice. Dworkin is to a large extent in agreement with Pawls, but his extended discussion of Pawls's philosophical method allows him to develop his own account of morality as right based and show how such a view can be developed and defended against critics such as utilitarians. He begins with a famous critique of Rawls's conception of the 'original position' in which individuals, deprived of important knowledge of themselves and their circumstances, agree to two principles of justice which then structure a fair society. Like many others, Dworkin points out that the 'original position' cannot provide reasons for accepting the two principles of justice, for an agreement made in such a condition of ignorance could have no binding force whatever: 'A hypothetical contract is not simply a pale form of contract; it is no contract at all' (p. 151). He then goes on to probe Rawls's argument more deeply, and it is here that he identifies the resources for his own defence of a right-based morality. He turns in particular to Rawls's method of reflective equilibrium. This method is similar to the constructivist method that Dworkin uses in the person of Hercules to justify particular decisions in 'hard cases'.

The method of reflective equilibrium starts from the common intuitions and sense of justice that individuals share. Although individuals share a sense of justice, they differ considerably in their accounts of what justice is and what it requires. Dworkin suggests that we share a belief that slavery is unjust, but we might differ significantly in our accounts of why this is so. Moral and political argument starts from these shared common intuitions. When we engage in arguments about justice we appeal to such intuitions, but we do so in complex ways. These intuitive commitments are in part constitutive of a community, and it could only abandon its core intuitions about justice by becoming significantly different. Moral argument, at least in its philosophical guise, involves the construction of theories which account for these intuitive convictions. Although these theories are developed independently of the beliefs and tested against them, in the process of articulating and developing them they undergo modification and refinement; that is why the method is described as reflective. Equilibrium is achieved when there is a match between

-279-

the structure and outcome of our theory of justice and our considered intuitions about justice, and when the theory is shown to provide a better reflective account of our intuitive commitments than any rival theory does. This constructivist method of reflective equilibrium is how we justify moral and political principles. So, once again, Dworkin is rejecting a method of moral argument which presents theory as external to a particular community. Moral argument has to be grounded somewhere, and it starts not in the realm of philosophical abstraction but with these common intuitions. That said, the methodology allows for considerable philosophical abstraction and subtlety in the critique of rival theories. Dworkin then goes on to characterize Rawls's device of the 'original position' as a constructive model for public justification of moral or political principles in circumstances where individuals differ in their convictions. What underlies this device is the intuition that individuals deserve equal concern and respect, and what the 'original position' does is model a situation of choice characterized by equality of concern and respect. In such a situation equal individuals choose the principles which regulate their association. This leads Dworkin to characterize Rawls's theory as right based. He distinguishes three broad categories of moral theory: right based, duty based, and goal based. A right-based morality takes some basic right--in this case the right of individuals to equal concern and respect--as morally fundamental. Dworkin makes a good case for characterizing Rawls's argument as right based in this way, because of the priority he attaches to the individual over the maximization of welfare. However, he still has to show that our wider public morality is also right based, and show how this basic right to equality of concern and respect relates to individual rights claims.

Satisfying the first of these two requirements is easy in one sense, because all Dworkin needs to do is show that we do appeal to rights in moral and political argument. Rights certainly have an important role in public moral argument, particularly in the USA, which has a Bill of Rights and a conception of fundamental rights written into certain key constitutional amendments. However, he also needs to show that this appeal to rights is fundamental and not merely a 'short hand' for more

-280-

fundamental moral principles such as utility. To overcome this challenge he needs to give a more detailed account of how institutional and basic rights used in moral, political, and constitutional argument are grounded on the fundamental right to be treated with equality of concern and respect. This he does in a discussion of 'Civil Disobedience' occasioned by demonstrations against the draft and the war in Vietnam and in the chapter entitled "'Taking Rights Seriously'". Here he identifies the point of moral and political rights as protections of certain basic human interests against the coercive power of the state. To have a right, he argues, is to enjoy protection against the coercive power of the state or community, even when the exercise of that right offends the public's sense of what is the right thing to do. Not all legally enacted rights are rights in this strong moral sense. It was not the case in the USA that individuals had an enacted legal right to dodge the draft. However, Dworkin points out that this does not entail that an individual does wrong in dodging the draft. This question depends on whether the legal enactment of the draft is constitutional--that is, whether it can be shown to be consistent with the best account of law and its function within the community. Similarly, it is not the case that an individual does wrong in protesting against the draft law, even if this runs against the grain of majority opinion, if it is the case that the legal enactment violates constitutional rights.

Dworkin's arguments about 'Civil Disobedience' are complex and controversial and deserve a much fuller treatment than space allows. What is important here is the distinction between 'rights' and 'the night thing to do'. This distinction is crucial to understanding the function of rights in public moral discourse. To have a right only when it is consistent with majority opinion about the right thing to do would undermine the idea of rights as special protections of crucial individual interests. What sort of protection would a right to freedom of expression be if it could be restricted every time the public interest required it? If we accept that rights function in this way as protections against majority tyranny and government power, then we have good grounds for rejecting any utilitarian or goal-based foundation of those rights. Any such goal- based theory would always make the enforcement of rights

-281-

conditional on the achievement of the goal rather than on the principled protection of the individual irrespective of outcomes. The claim that our public morality is right based and that what underlies this right is the requirement of equal concern and respect provides a better account of the weight we attach to rights arguments than can be provided by utilitarianism. The discourse of rights which we employ in public moral and political arguments in a liberal democratic society is not merely a 'short hand' for utility, and to assume that it is falls to take seriously the function and weight of rights.

A further crucial feature of claiming that our public moral discourse is best characterized as right based is that it helps explain the complex relationship between the foundational 'background' right to equality of concern and respect and the institutional rights that are grounded on it. Dworkin's rights thesis does not give rise to a simple list of basic rights such as life, liberty, and property, such as we find in Nozick Anarchy, State, and Utopia. Instead the foundational right to equality of concern and respect gives rise to a complex of rights to particular liberties, rather than to liberty as such, and to other rights which comprise equal concern and respect rather than to a right to sameness of treatment. The rejection of a crude right to sameness of treatment allows him to justify policies of reverse discrimination which on the surface at least seem to contradict a commitment to equality. The public moral argument over rights to equal treatment, reverse discrimination, and distributive justice are all, according to Dworkin, best explained in terms of articulating the full implications of this commitment to equality of concern and respect. Utilitarian arguments cannot provide an adequate account of this basic egalitarian intuition despite a commitment to 'Bentham's Dictum' of counting everyone as one and no one as more than one. The supposed egalitarianism of utilitarianism falls because it does not treat everyone as one, but actually discriminates against minorities as less than one, through its counting of external preferences. Even though preference utilitarianism seeks to satisfy as many preferences as possible, according to Dworkin it fails to satisfy strict impartiality because it includes both personal and external preferences. Personal preferences are for an individual's own enjoyment of some good or service, whereas

-282-

external preferences concern the assignment of goods or benefits to third parties. If a preference utilitarian calculation is to satisfy the preferences of a group, but these are external preferences for restriction of law-school places to whites only because blacks are somehow inferior, then the simple egalitarianism of utilitarianism collapses in favour of the institutionalization of discrimination which falls to accord all individuals equality of concern and respect. Utilitarianism can be egalitarian only by excluding such external preferences, and it can only do that by ceasing to be utilitarian and instead incorporating some right-based principle which justifies this exclusion. If, however, utilitarianism seeks to abandon egalitarianism altogether as some strict act utilitarians do, then the justification of the principle is challenged by its strong incompatibility with our intuitions about rights and equality of concern and respect. And if it ignores such intuitions in its justification, there is the much more complex philosophical problem of what does ground the principle of utility and what makes it a source of moral and political obligations. Because it is wholly abstracted from even the best reflective account of our core moral and political intuitions, the argument of the radical utilitarian, like that of the traditional natural-law theorist, neither challenges nor assists us in public moral and political debate.

Conclusion

Superficially, Dworkin's arguments appear to be addressed to a political culture such as that of the USA, where rights play a significant role in public argument, and the and the natural implication is that his theory has little to say to a legal and political order which is not rights based. This is a farmiliar charge levelled against much American political philosophy--namely, that it universalizes its own political culture. Such a response falls to take seriously Dworkin's important distinction between rights- based conceptions of law and political community and right- based conceptions. In Britain and many other countries which have no written constitution, there is no basic set of rights to which all subscribe. Rights are not foundational. However, Dworkin is clearly aware of this; what he is concerned with is a conception of the moral significance of the individual as a

-283-

free and equal subject. That, he rightly argues, is a view which is deeply held across very different legal and political systems. To argue that this is a peculiarly American idea would be to misunderstand the way in which this idea has its roots in modern European culture. This sort of commitment to the individual may well have its origins in historically specific communities, but it would be a mistake to see it as the peculiar property of any one community or culture. Whether we subscribe to a given set of rights as politically foundational or not, it is clear that Dworkin's right-based theory has a much wider significance and application than the USA.

In challenging the 'ruling theory of law', Dworkin set himself against some of the most cherished beliefs of modern moral and political philosophy, which denied the continuity of law, morality, and politics. His thesis is controversial enough in itself, and modern positivists have been reluctant to abandon their position when faced with his arguments. The debate between Dworkin and the 'ruling theory' goes on. However, one implication of his position has been particularly troubling, even for those who are sympathetic to his assault on positivism and utilitarianism. This is what has come to be known as the 'right-answer thesis'. If Dworkin is right in his claim that in adjudication the judge is finding law and not making it, then it follows that there is in principle a right answer to each controversial 'hard case'. Similarly, in moral and political arguments there must be a right answer to questions of justice and right. Dworkin's reasons for this thesis are complex. First, he argues that, in the case of legal and moral reasoning, the reasons never just run out. In making a moral decision about the requirements of distributive justice or law, a judge does not only argue as far as he can and then simply make arbitrary decisions like tossing a coin. This fails to capture adequately the practice of reason-giving which underlies moral and political argument. For Dworkin the practice of law is in part the institutionalization of that practice of public argument. The second reason has to do with the logic of such reasoning. The 'right-answer thesis' excludes the possibility of any middle ground between two controversial claims in law or morality. If one is true, then the other cannot also be true at the same time and must therefore be false. Such a view challenges the

-284-

incommensurability which many find an irreducible feature of modern pluralist society. Traditionally, natural-rights arguments and their legacy for modern procedural theories of justice and conceptions of law have presented a picture of humans as inherently conflictual. This conflict cannot be eradicated but can be channelled into competition rather than conflict by law. Even modern procedural conceptions of the political community such as Rawls's and Dworkin's theories of justice appear to be ways of accommodating fundamental, and, at least in Rawls's case, intractable disagreements. If at the same time Dworkin wants to advance a cognitivist theory of law and adjudication, then he has the problem of reconciling this commitment with a liberal theory of justice as neutrality. After all, neutrality is usually presented as a response to an intractable disagreement over which 'right answers' are not possible. I do not wish to suggest that Dworkin cannot reconcile the cognitivism of his legal theory with his theory of justice as neutrality, but the two strands appear to pull in opposite directions, and reconciling this tension is not an easy task. In favour of Dworkin's position it can be argued that he takes seriously the practice of reason-giving which underlies legal, moral, and political decision-making. If we reject the Dworkinian view, we are left with a profound challenge: why should we trust unaccountable individuals to make such fundamentally important decisions? If their views are arbitrary rather than principled, then what place do they have in a liberal democratic society? What can legitimize a liberal democratic order is its commitment to the open public justification of political decisions, rather than the conventional acceptance of arbitrary majority will. Despite the ultimately incomplete character of his argument in Taking Rights Seriously, the importance of the book lies in its addressing some of the most fundamental and urgent questions of political philosophy today.

Bibliographical Note

Taking Rights Seriously was first published in 1977. In 1978 a second corrected edition appeared with the addition of a long appendix, 'Reply to Critics', originally published in an issue of the Georgia Law Review

-285-

devoted to Taking Rights Seriously. The edition I have used is the second corrected edition. Taking Rights Seriously, although still undoubtedly Dworkin's most influential and widely read work, is very much a work in progress, an outline of a position that has been developed and expanded in a number of books published since. A further volume of essays in law and political theory, A Matter of Principle, was published by Oxford University Press in 1985, and in 1986 Dworkin published a complete monograph statement of his jurisprudence, Law's Empire ( Cambridge, Mass.). Law's Empire develops and expands many of the arguments in Taking Rights Seriously and has attracted considerable attention in its own right.

Following the publication of Law's Empire, much of Dworkin's attention has been devoted to political philosophy rather than the philosophy of law more narrowly conceived. In particular he has been working on an egalitarian theory of Justice and the foundations of liberal political theory. On the subject of Justice and equality, see 'What is Equality? Part 1: Equality of Welfare', and 'What is Equality? Part 2: Equality of Resources', Philosophy and Public Affairs, 10 ( 1981), 185-246, 283-345, 'What is Equality? Part 3: The Place of Liberty', Iowa Law Review, 73 ( 1987), 1-54, and 'What is Equality? Part 4: Political Equality', University of San Francisco Law, Review, 77 ( 1989), 1-30, 'In Defense of Equality', Social Philosophy and Policy, 1 ( 1983), 24-40, and 'Liberal Community', California Law Review, 77 ( 1989), 479-504. On the foundations of egalitarian liberalism, see 'Foundations of Liberal Equality', The Tanner Lectures on Human Values, xi, ed. Grethe B. Peterson ( Salt Late City, 1990), 3-119. Dworkin is currently reworking the essays on equality for a book on egalitarian justice. His most recent book, which is a departure from these more abstract concerns, is Life's Dominion: An Argument about Abortion and Euthanasia ( London, 1993). Although this book is a significant departure from the abstractions of contemporary political philosophy, it continues to testify to Dworkin's concern to apply his political and legal philosophy to matters of current controversy, and his view of political philosophy as contributing to the public philosophy of a modern liberal democracy.

Although there are few books on Dworkin's thought, the secondary literature is vast. The only complete guide to Dworkin's ideas so far is S. Guest, Ronald Dworkin ( Edinburgh, 1992). This is useful but written largely from a lawyer's point of view for law students, and is thus less helpful on Dworkin's political philosophy or the wider context of philosophy of law. Other useful collections of essays are: M. Cohen (ed.), Ronald Dworkin and Contemporary jurisprudence ( London, 1994), and A. Hunt (ed.), Reading Dworkin Critically ( Oxford, 1992). The latter is rather polemical. Most good studies of the philosophy of H. L. A. Hart will have a significant discussion of Dworkin's arguments. Although not

-286-

devoted to Dworkin, both W. Kymlicka, Contemporary Political Philosophy ( Oxford, 1990), and J. Waldron, The Right to Private Property ( Oxford, 1988), are particularly useful. Both are philosophically close to Dworkin and develop some of his arguments concerning rights and equality. Dworkin's contributions on constitutional law, politics, and political philosophy appear regularly in the pages of the New York Review of Books.

-287-