A Theory of Justice
I think that this book is the most substantial and interesting contribution to moral philosophy since the war, at least if one thinks only of works written in English. It is a very persuasive book, being very well argued and carefully composed, with possible objections and counterarguments fairly weighed and considered: at the same time it conveys a moral vision and a ruling idea, and a strongly marked personal attitude to experience. Although the book is firmly within the traditions of analytical philosophy, and has the virtue of this kind, there is no pretense of a degree of precision that the subject matter does not admit; and this has probably been one cause of the dullness of much analytical philosophy in this field.
Professor Rawls often remarks that any moral theory that is enlightening and not trivial will be an approximation to the truth at some points, and will not fit perfectly all possible cases and situations. In this field, as in so many others, one obtains instructive generalities only at the cost of some looseness of fit in peripheral cases. If the moral theory is a good one, it does clearly distinguish the central cases of justice from the peripheral ones, and it does bring the central cases into a fairly precise and intelligible relation to each other, when before they had seemed a heterogeneous collection without any center at all.
The point of a moral theory, and so of philosophical ethics, is to find some very general guiding principles that explain the apparently unconnected moral beliefs that constitute a prevailing morality. Rather as a linguist and philologist may look for the general principles that determine word order and the structure of sentences in English, so the philosophical moralist looks for the general principles, or the single principle, that explain the apparently diverse arrangements that we would consider unjust and therefore wrong. If we do succeed in finding such principles, which fit the facts of our ordinary moral beliefs fairly well except in a few marginal cases, then we can use the principles as a guide in doubtful cases; just as we would use general principles of grammar as a guide in doubtful cases when our intuitions fail us or are uncertain.
But there is an even more important gain: if our moral beliefs on many subjects, and in many very different situations, are shown to be instances of a few general principles at work, then we have an assurance that our moral beliefs have a rational foundation. At least they are not just a chaos and a jumble: there is a reason why we hold the various beliefs that we do. There is a unitary policy at work, and, in spite of appearances, we are not in our moral judgments just zigzagging from one disconnected prejudice to another.
At most periods in the history of Western thought the fear that moral beliefs may have no rational foundation has been expressed, and expressed with strong feeling. Moral skepticism, familiar from Plato’s dialogues and in Thucydides, is perpetual and natural: in some periods and places it becomes an orthodoxy; usually it is a heterodoxy, but it is always there. Sometimes philosophers, for example Hume, and thinkers of lesser claims, have either overcome or have domesticated the fear by embracing the conclusion, and then trying to rejoice in it. The skeptic may proclaim that in practical matters and the conduct of life we are governed by the heart and not by the head, and that it is better so: our moral sentiments, our immediate repugnances and sympathies, secure our adaptation to an overcrowded environment and have been developed over a long history of socialization.
To look for an underlying consistency across the range of a man’s moral attitudes and judgments is, in the view of such skeptics, to mistake the sense and purpose of moral attitudes. Moral attitudes have the same kind of consistency that a man’s aesthetic and artistic preferences have: they express a particular temperament, and a particular range of feeling, which are both characteristic of their time and place and which also have features that are common to civilized men at most places and times. But there is no rational structure behind the judgments that a man makes: the explanation of the judgments is to be found not in some general principles of rational policy but rather in the largely unknown deep psychology of the sentiments.
After a long history, skepticism became an orthodoxy once again among academic philosophers influenced by logical positivism before the last war, and immediately after it. Wittgenstein had suggested that a man’s expressions of his moral opinions (as I would call them) should be interpreted as expressions of emotion; others suggested that the use of language in moral exhortation should be thought of as the emotive use of language. The word “emotive” had originally been given currency by Professor I.A. Richards as part of a theory of poetic meaning. The habit of dismissing moral arguments, and arguments about the proper ends of action, as delusive and futile was a lingering habit of social scientists also, who were strengthened by the authority of philosophers.
Professor Rawls has been one of the few analytical philosophers who, by the example of their inquiries, have effectively undermined this skeptical orthodoxy. Twenty years ago, in the heyday of what came to be called Oxford philosophy, and while Wittgenstein’s later philosophy was becoming widely known, ethics was still considered an intellectually barren subject, in which no major discoveries were to be expected and no new lines of inquiry could be opened. Professor Rawls’s prolonged inquiries into the nature of justice, and into utilitarian theories of justice, first became well known through articles in philosophical journals about ten or fifteen years ago; they represented a return to the classical, nonskeptical tradition, and yet the method of argument could not be criticized as lacking logical rigor and precaution.
If no account of the virtue of justice can be given that is consistent with the principles of a utilitarian philosophy, no account that systematizes our reflective beliefs about what is just and unjust, then the utilitarian philosophy must be rejected: that was his argument, and it presupposes an underlying consistency as a requirement of rational opinion in ethics. Professor Rawls’s writings on justice are part of a recent movement of thought among philosophers away from the skepticism about rationality in ethics, and therefore in politics, a skepticism best expressed in Professor Charles Stevenson’s Ethics and Language.
The revival of philosophical jurisprudence, associated principally with Professor Herbert Hart in Oxford, together with the pressure of political events in America, has made the search for rational structure in ethics both more widespread and more urgent. Under what conditions is a war a just war? How far may the state justly require a citizen to play his part in a war which he considers unjust? What are the degrees of moral outrage by a government which justify resistance by violent, and also by illegal, means?
These questions have all been rationally debated in the last few years, and must therefore be rationally debatable. The disputants did not in fact fall back on saying: “It is all a matter of how you feel, and there is no point in discussing our different attitudes, once we have clarified the facts.” On the contrary they thought it appropriate, and even necessary, to look for guiding principles very much as lawyers would in arguing, and finally settling, their difficult cases. And this pursuit of rational structure is not only a social necessity, holding together the disputants in a liberal society by common allegiance to methods of argument, even if the methods are ineffective. It is also felt to be a necessity by individuals for their own sake. They feel the need of knowing precisely what stand they are taking, and what general principles they are invoking, particularly when the course of their lives will be disrupted by their moral convictions. They need to make clear to themselves what their action is intended to be, to fix accurately the description, and the banner, under which they are acting.
Rawls’s book is intended to show the kind and form of principle that is involved when a moral stand is a reasonable one; for he who writes about the nature of justice, and attempts some kind of definition of it, is taken to be offering a scheme of rationality for morality in general, because he is proferring such a scheme for the first of the moral virtues. So it was in Plato’s Republic and so it is again in Professor Rawls’s.
In the manner of a traditional theorist of social contract, Rawls considers the choice of a social order that would be made by a sensible and reasonable man who was ignorant of the particular endowments and opportunities, advantages and disadvantages, that he himself would possess in that social order. This imagined ignorance of his own particular situation serves to ensure that the imagined choice will have a general and representative character, as being the choice of an indeterminately characterized Everyman; and yet no special altruism, or disregard of self, need be attributed to this ideal, abstract person who chooses. Lacking any bias derived from a particular known position in the social order, he will, Rawls argues, reasonably choose a method of distribution of advantages and disadvantages that will be fair to all, and that can be agreed to be fair by all, whatever their particular position may be.
By eliminating other possibilities, Rawls shows that fairness can be realized only if two principles are observed: that each person has an equal right to the most extensive equal liberties compatible with similar liberty for all, and secondly, that all inequalities in liberty be justified only as being an advantage to the least advantaged. An enactment of a less extensive liberty must; if it is to be permissible, strengthen the total system of liberty shared by all; and an unequal liberty must be rationally acceptable to those citizens with the lesser liberty. The same double principle of distribution applies to other primary goods, such as opportunity and wealth, as well as to the first of the primary goods, liberty.
The root idea of Rawls’s theory is that an injustice is an arbitrary inequality in the distribution of good things, and primarily in the distribution of liberty. Inequalities of rights and liberty, and inequalities in the distribution of good things generally, are arbitrary from a moral point of view, Rawls argues, if they cannot be shown to produce a benefit in which all the least fortunate have a preponderant share. Aristocratic and plutocratic societies are unjust because their social orders incorporate that arbitrariness in the allocation of liberty and of other benefits which is to be expected in nature.
The moral point of view, for Rawls, precisely consists in the principled rectification of nature’s casualness in distribution. If there is an inequality that we are to accept as reasonable, we must be able to accept it on quite impersonal, objective grounds, discounting altogether our own distinctive needs and interests. For example, restrictions upon freedom of choice in education or medical services may be justified by the net increase in access to these services for the least fortunate. A rational social order, founded upon principles of justice, is to be contrasted with a natural social order, which arises from the blind and natural forces of social competition and heredity.
Professor Rawls is developing, in the lucid language of the modern theory of rational choice, a philosophical position that was common to Rousseau and Kant. In Rousseau a man has a sense of himself as a potential member of a community of free and equal rational beings, while he also finds himself actually constrained by social arrangements that he had no part in making or approving, and that are the outcome of social competition. The basis for a social contract, which moralizes the social order and makes it acceptable to free and reasonable men, is the requirement that each citizen should think of himself only as a citizen when communal decisions are to be made. He should abstract himself altogether from his own contingent interests and preferences, and his own status and opportunities as an individual.
As citizens of a single community, men who would be in conflict with each other in pursuit of their own interests as individuals can reach agreement on the principles and policies that serve their common interests as citizens. These agreed upon common principles and policies must be precisely those that strengthen the community as a community rather than merely increasing the aggregate welfare or satisfaction of individuals. Then self-respect is preserved, because men are no longer at the mercy of the whims of others.
Kant, following Rousseau, similarly argued that every man thinks of himself as belonging to an ideal community of rational beings, as well as having interests that are distinctive and gifts and abilities that are distinctive. The king in his palace and the peasant in his hovel are equal when they meet each other as rational men, quite apart from every contingent attribute that they possess. Whatever their different abilities and interests and opportunities, there is a common ground on which they can meet as rational beings who are ready to listen to argument on matters of principle in an objective, impersonal spirit. They can address each other as equals if and only if they each respect the capacity to follow general principles that exists in themselves and in others.
Hegel reasonably insisted that this conception of morality as the recognition of a common rationality, beneath the arbitrary trappings of natural and social differences, marked a modern spirit, the spirit of the Enlightenment, with its Utopian ideals of the kingdom of reason on earth. Though the idea is traceable to the Stoics, it is principally a Christian idea, which survives in the Enlightenment and in Kant without its dogmatic religious supports. It seems to be a secularized, philosophical version of the notion of the individual soul having its own history, and its own realm, independent of the worldly history of a person: a conception that was soon to be transcended, Hegel believed, in the coming age of political planning, when the individual would learn to define himself by his place in the historical process and in the structure of the state.
Rawls acknowledges that his conception of justice represents a Kantian social ideal that has certainly not prevailed in the minds of men at all times and in all climates. His assimilation of the philosophy of politics to the philosophy of language implies that one can arrive by reflection at the most general principles of the particular social morality within which one lives, and which is one morality among others, as one’s native language is one language among others. We have reliable intuitions about the misuses and solecisms that are intolerable in speech, and some insight into why they are intolerable: so also, Rawls’s method implies, for rights, duties, and obligations in the moral system which we have internalized, as we have internalized the syntactical forms that determine propriety of speech in our language.
In both cases it is still an open question whether there are some common, or even universal, deep-level principles which all moralities, or all languages, in some way exemplify. No one knows. But at least different moralities play a recognizably similar set of parts in the social lives that are governed by their superficially diverse rules. One does not need to be suspicious of the genuineness of the claims that our own morality makes upon us merely because our own morality is in fact only one among many. This is the relativist’s error. Shared loyalty to some one morality is a necessary, though not sufficient, condition of human dignity and self-respect, as the shared discipline of some one language is a necessary though not a sufficient condition of rationality.
But there is a point at which the analogy breaks down, because a morality is a less strictly organized and clearly identifiable entity than a language, and also because a morality is something that one accepts and believes (or believes in), while a language is something that one learns and uses. Also one can, and in rich countries one often does, make up one’s morality for oneself: but not one’s language. Lastly, commitment to a particular morality is of overriding importance, while commitment to a particular language, though important, is not overridingly so.
The absoluteness that we attribute to moral claims, which override all others, resides as much in the fact that they are moral claims as in their specific content. Two men whose lives are formed by conflicting moralities can at least recognize the legitimate absoluteness of claims which they would from their own standpoint reject because of the content or subject matter of the claims. They can acknowledge that their political antagonist is at least being governed by genuinely moral claims exactly as they are themselves.
Professor Rawls does not argue at length against moral skepticism. We do, he thinks, possess and use a concept of justice as fairness. “We” stands for the heirs of a European tradition of law and morality, and it is a tradition that has penetrated into many other places; nor is the concept of justice confined to this one tradition and unrecognizable outside it. The utilitarians, it is true, have given defective accounts of the concept. If you claim, as they do, that all the normative concepts of ethics must be analyzable in relation to measurable, positive goods such as happiness or pleasure, your account of justice is bound to be mysterious and resistant to analysis. For utilitarians, justice is in any case secondary, at best a means of maximizing welfare and at worst not independently identifiable at all. But for nonutilitarians justice is the first of the social virtues, and for Professor Rawls the notion of fairness in the distribution of advantages represents not only a social ideal but the first requirement of a society that is rationally acceptable to all its members.
The social order, in his view, will correct the premoral contingencies of the natural order, in which pestilence, famine, and war are distributed senselessly, as are opportunities for excellence. Inequalities, even gross inequalities, and restrictions upon freedom will persist in a rational social order. It may be the case that some doctors, for example, have more liberty or wealth than some of their patients; but the doctors’ advantage will be serving a legible purpose which everyone who discounts his own special interests will accept as being aimed at a better ultimate distribution: a distribution less disadvantageous to the least fortunate.
My first difficulty with Rawls’s argument begins here: the notion of fairness, and hence Rawls’s notion of justice, seems to involve attaching a sense to the notion of deserving; and this seems to me an utterly obscure notion, until some special explanation is attached to it. Consider this passage, in which the author’s strong ethical culture declares itself:
Perhaps some will think that the person with greater natural endowments deserves those assets and the superior character that made their development possible. Because he is more worthy in this sense, he deserves the greater advantages that he could achieve with them. This view, however, is surely incorrect. It seems to be one of the fixed points of our considered judgments that no one deserves his place in the distribution of native endowments, any more than one deserves one’s initial starting place in society…. Character depends in large part upon fortunate family and social circumstances for which he can claim no credit.
But one may ask: Is there anything whatever that, strictly speaking, a man can claim credit for, or he can properly be said to deserve, with the implication that it can be attributed to him, the ultimate subject, as contrasted with the natural forces that formed him? In the last analysis, are not all advantages and disadvantages distributed by natural causes, even when they are the effects of human agency? And, if we are not strict theists, we will surely not suppose that there is cosmic justice in these distributions?
There seems in Professor Rawls to be an illegitimate (and Kantian) contrast between the essential rational man and his accidental trappings, which include his tastes and talents, and also his desires and interests. Professor Rawls insists that he does not rely on the notion of deserving in defining the original position from which a rational man, ignorant of his own social position and personal characteristics, would rationally choose a just distribution of liberties. This is true; for the fairness that results from the choice is achieved by making the choosers ignorant of all distinguishing facts about themselves, which might provide motives for choosing a social order that rewards some natural characteristics and social placings as against others. The fairness aimed at is the negation not only of aristocracy but also of meritocracy.
A principle of indifference, which counts each man as one merely by virtue of his openness to reason and the idea of justice, is a strong principle of equality: one that I happen to accept, but also one that has only a comparatively recent history, perhaps of two hundred years, of being accepted as normal and not eccentric. This principle of equality becomes a principle of fairness if one thinks of naturally acquired advantages as unearned and undeserved. But I think it would be better to think of all advantages, whether naturally acquired or conferred by men, as unearned and undeserved, and indeed to reject altogether the notion of deserving as having a place in rational and systematic ethics. After genetic roulette and the roulette of childhood environment, a man emerges, so equipped, into the poker game of social competition, within a social system determined by largely unknown historical forces.
Some actual resulting inequalities are justifiable, because they lead to benefits which the least benefited people in society enjoy; they are not deserved, but because of their effects they are fair. Rawls’s principle of justice allows, for instance, that some men should earn wealth and privilege by labors and risks which would not have been performed and incurred in the absence of these incentives—if the poor benefit from these labors and risks no less than those who have earned wealth and privilege. It can be just, in his view, that the artist, inventor, or entrepreneur who has risked failure, unlike the humdrum administrator, should be rewarded for success disproportionately. No doubt exceptional power and willingness to labor and to take risks are finally attributable to accidents of heredity and environment; and in that sense (if it is a sense) they do not deserve the exceptional rewards that they earn. But it is fair that they should get them if others benefit even more.
The objection can be made, and will be made, that a blind chooser might, not irrationally, just choose to gamble, and prefer a social order where the winner takes all, or virtually all: he may prefer to take the risk of being a slave if he has a chance of being a master. The rationality of choosing fairness in social arrangements is not like the rationality of choosing consistency in beliefs; there are tenable alternatives to fairness, but not to mere consistency. But Rawls would reply that such alternatives are not interesting because they are not usually held by those who think about morality and who attach value to justice as a virtue. Suppose a rational man, ignorant of his own social placing and natural gifts, chose to gamble, and consequently chose a social order that leaves nature’s inequitable distribution of liberty and goods uncorrected: he might believe that there was some ultimate justice in another world, or even that there was some cosmic justice in nature’s distributions; but he would scarcely have a use for the concept of social justice.
Professor Rawls arrives, from the premise of rationality in blind choice, at the principle of justice which social democrats have always groped for: when they have presented socialism as aimed at social justice, they have had in mind the principled redistribution of advantages which are randomly distributed in the natural order, since they depend on accidents of heredity and geographical environment and inheritance. Rawls adds an extra liberal scruple, in the spirit of J.S. Mill, in stating the principle of justice with reference to a supreme good, liberty, rather than by referring to the positive goods, such as welfare or happiness.
For him the disadvantaged are all those who, in one way or another, lack the liberty to do what they in fact want to do, whatever that may be; this is where they principally differ from the fortunate, who have inherited or acquired the freedom to act as they please or as they think right. The social order must be an engine geared to correct the differentials in liberty unless they have a function in increasing liberties available to the disadvantaged: the whole social order must perpetually redistribute unequal liberties rather as a genuinely progressive tax system perpetually corrects the otherwise widening differentials in wealth.
It is a noble, coherent, highly abstract picture of the fair society, as social democrats see it. In England, books about the Labour Party’s aims—for example, those written by Douglas Jay and Anthony Crosland since the war—needed just such a theory of justice as this, stated in its full philosophical generality. This is certainly the model of social justice that has governed the advocacy of R.H. Tawney and Richard Titmuss and that holds the Labour Party together. Society must repair the cruelties of nature, and it exists not only to preserve law and order but also to correct the natural differentials between the strong and the weak, and to give institutional support to self-respect, which is for Rawls a primary value.
This is not the moral philosophy of communism, any more than it is of high capitalism. “Socialist competition” in the Soviet Union has in fact allowed immense inequalities to exist provided that they are marginally efficient in adding to the total common pool of economic advantages; this is the kind of defense that was offered, in the time of unapologetic capitalism, for child labor and the rejection of the Factory Acts.
It is important that a “Theory of Justice” should be understood as one conception of justice among others in the tradition of speculations on justice. Each rival conception is still associated with the same concept and has an overlapping history. The classical conception of social justice, continuous from Plato and Aristotle, is of a harmony of pre-existing and recognized spheres of interest, which allots different roles to different classes and which can be modified only by agreed and rational procedures. The two notions, 1) of what is fitting and due to a particular status and role, and 2) of the recognized necessity of following the prescribed procedure when charges are to be made, constitute the classical notion, before the modifications of liberal thought.
The picture of the social order and the natural order that underlies this classical conception is not that the social order properly corrects the arbitrary, and rationally unintelligible, distribution of natural advantages: but rather that the social order properly reflects the hallowed distinctions of degree, status, and role that are perceptible in the natural order, which itself has a harmonious structure pleasing to the Creator. The social order is sick and liable to violence, total disorder, and death when it no longer reflects, as in a mirror, the due subordinations and separations of degree that nature everywhere requires. Examples would be the subordination of female to male, of children to parents, of subject to monarch, of the layman to the priest, and so on. These subordinations, and many like them, represent natural law, in the classical conception. It is evident in the light of reason, for example, that it is just and fitting for parents to care for their young children and for subjects to be loyal to their sovereign. The moral laws are an extension of principles that are innately intelligible to us, and they are not invented, but rather discovered, by us.
The liberal conception of justice, so clearly articulated by Professor Rawls, opposes the designed and invented moral order to the blind causality of the natural order, the moralized and socialized citizen to the natural man. Moral custom and rule are not second nature, but human artifice. In nature the strong prevail and are served by the weak, and form an aristocracy or tyranny: for nature is the domain of mere causality, of a kind that does not conform to the intelligible purposes of a Creator, or to a principle of Sufficient Reason, which might be the last remnant of a cosmic justice and harmony. Equal liberty, and the requirement that each man counts for one in reckoning the common good, are in practice recent inventions, principles of 1789. To their Tory enemies, they are abstract principles of moral geometry, unnatural and unhistorical and therefore delusive: for custom, second nature, prescribes naturally intelligible duties and obligations filling different ranks and roles, both inherited and properly acquired.
The liberal idea of a just society is presumably at home in the United States, and not at war there with notions of legitimacy. The Burkean belief that a just social order, with its inherited degrees and subordinations, reflects a deeper natural order can scarcely have much hold in a freely competitive and socially mobile capitalism. On the contrary, the conservative philosophy of a society which has not been preoccupied with legitimacy and which has had no aristocracy, as in the United States, will rather stress that inequality is the natural result of competition, and will try to justify the inequalities by a strictly utilitarian argument: that is, that the social order ought to be an efficient instrument for maximizing welfare and happiness; and justice and fairness in distribution ought to be no more than useful means to this end. Thus it is argued by conservatives in the US that over-all economic growth has been worth its costs in inequality and injustices.
Professor Rawls’s theory of justice arises from a repudiation of this utilitarian ethics as inadequate to the moral point of view: fairness is for him an irreducible requirement of the moral point of view, and he argues that the efforts of Hume and Mill and Sidgwick to show that justice is a dependent value, dependent on utility, all fail to meet our intuitions.
The enemy on the other side from utilitarianism is intuitionist ethics, which Professor Rawls also rejects: for the intuitionist does not believe that there is a rationally chosen supreme governing principle, either simple of complex, which explains the variety of moral injunctions that we respect. He does not believe that the duties that fall under the heading of being just and fair can be seen as dependent upon a social contract, in which a rational man recognizes a social commitment that other rational men will equally acknowledge. Rather he may believe that the duties and obligations that constitute being just are a specific part of the right way of life which a reflective man imposes on himself as a personal morality and which are binding on him and are independent of any social contract. I write “may believe” because I am choosing what seems to me the most plausible form of intuitionism. The duties and obligations to be just arise not only from his position as a rational member of society subject to a contract but also as a personal morality, representing an opinion about how respect-worthy human beings should behave toward each other, even outside a social order.
The form of intuitionist philosophy that Professor Rawls rejects is that of the Oxford philosopher, the late Sir David Ross. Such a moral philosophy is indeed unacceptable, in so far as it implies that men have a special power of reasoning, or source of knowledge, called “intuition,” which explains our moral judgments. Divorced from such claims, however, a moral philosophy not unlike Ross’s intuitionism seems to me nearer to adequacy than Professor Rawls’s social contract theory. For it seems to me an error of emphasis to explain the virtue of justice, and even more the other essential virtues, as rational consequences of planned co-operation in a rational social setting.
This is one aspect of the virtue of justice and of its function among the virtues; but one may think that it is a derivative, and not the central one. To adopt the moral point of view, one may argue, is to think what kind of character and aims men should have, or try to have, and what kind of life they should lead. This entails having a picture of the wholly admirable man, and of an entirely desirable and admirable way of life. This interpretation of “the moral point of view” Professor Rawls calls “perfectionism,” and he allows that some moral thought is of this kind; but he claims that it is not as fundamental as thought about the necessary bases of rational cooperation, at least where justice and right conduct are to be explained.
It is not obvious how the issue can be settled between Rawls’s social conception of morality and the perfectionist who, like the intuitionist, stresses a moral ideal independent of a social contract. Against Professor Rawls I would cite two kinds of evidence: first, evidence from the history of reflective moral opinions, not all of which can be interpreted in his terms, and all of which can be interpreted in perfectionist terms. Secondly, the psychology of the moral sentiments, which suggests that guilt or shame about injustice and unfairness, and natural respect for their opposites, extends to more primitive relations between people than are imagined by his hypothesis of a rational choice of an unbiased social order.
For many reflective moralities, e.g., Spinoza’s, the institutions that determine the social order are merely the machinery that makes a desirable, natural, and admirable way of life possible: the approved social practices are not the center and starting point of moral concern. There have, moreover, been prevailing moralities that have not taken the distribution of goods within society but rather the relation between persons and the law as the center of virtue: a morality derived from the Old Testament would be of this kind, and will stress the necessity of justice.
The second argument, from the psychology of moral sentiment, associates the virtue of justice mainly with the notions of guilt and innocence, of law and due procedures of law, of reparation, of impartiality in judgment, and only secondarily with the rational distribution of goods in society. This family of moral ideas has primitive origins in personal history, and the trace of these origins is never lost. Evidence for this view can be found in studies of the concept of justice in Greek thought and literature (e.g., by Atkins), in social anthropology, and in psychoanalytical speculations and case studies. And the literature of jurisprudence, centered on the notion of justice, plays only a very small part in Professor Rawls’s book: too small a part, I think.
Moral philosophy, which is the inquiry into reasonable foundations for morality, does not achieve strict proofs, but only the description of possible foundations for moral judgment; and among possible foundations a choice must be made. Rawls’s suggestion that rational choice of an unbiased social order, where individual differences are unknown and discounted, is the foundation for ordinary moral judgments is a persuasive and powerful suggestion, and the fact that this reviewer cannot accept his foundation for his moral judgments, because another possibility—the “perfectionism” that Rawls discards—better fits his intuitions, reflects the fact that moral opinions may differ; for the philosophical difference probably follows a difference of emphasis in moral attitudes.
It is to me surprising, for example, that rationality should enter into the definition of “good” as a predicate, and also into the formal account given of the good, and of the good for X (pp. 399-400): surprising that the good for a man is elucidated by a rational plan for his life. “Rational” seems to me intelligible as an essential predicate of the choice of means toward ends, while the connection between rationality and some of the more important virtues and human aims seems to me, in all other respects, contingent and variable. This difference in theory probably reflects a difference in direct moral opinions: the dispositions and patterns of activity and enjoyment that Rawls and I most admire, and would like to possess and to follow, probably diverge at crucial points, if only slightly.
Again throughout most of his book justice appears as a feature of a well-ordered society rather than strictly as a virtue, that is, as a feature of a man’s character and of his dispositions as manifested in his behavior. The last chapters try to correct, or at least to supplement, this defect in the account of morality; not, I think, altogether successfully, because Rawls’s social contract theory makes right conduct prior to, and independent of, the goodness of a person.
Professor Rawls will also be criticized because he sometimes seems to suggest that the choice of a just social order is necessarily the choice of a rational man, and that any other choice would be irrational: for example, that the gambler, who chooses a social order that allows natural advantages to have their unmitigated effects, is irrational. He may also be attacked, though unfairly, because his model of justice in a well-ordered society is an abstract model, and he pays no attention to the dominant powers of corporations, unions, the military machine, the secret police, the mass media, which make American democracy less than a liberal democracy; and no provision is made for the citizen’s need for fuller participation in government, a critic may say. The answer is that Professor Rawls is providing a theoretical reconstruction of the notion of justice, and not of all the virtues of a good society and of a good life: and he is not concerned with the practicalities of political science or with the theory of democracy.
His argument is at most points wonderfully clear, and the book is a permanent refutation of the reproach that analytical philosophy cannot contribute to substantial moral and political thought. The substance of a critical and liberal political philosophy is here argued with an assurance and breadth of mind that put the book in the tradition of Adam Smith and Mill and Sidgwick: in the best tradition of British moralists, revived now at Harvard.