‘Spheres of Justice’: An Exchange

To the Editors:

Ronald Dworkin’s review of my Spheres of Justice [NYR, April 14] is also, and quite legitimately, a defense of his own approach to moral and political philosophy. It raises hard questions about how that enterprise ought to be carried on, but it provides, I’m afraid, easy answers. It avoids the difficulties of morality and politics. Indeed, that avoidance is, if I understand Dworkin correctly, the greatest advantage that he claims for his approach. I would like to argue (again) that the difficulties are unavoidable.

My own claim is that we cannot distribute goods to men and women until we understand what the goods mean, what parts they play, how they are created, and how they are valued, among those same men and women. Distributions flow out of and are relative to social meanings. But this argument fails, according to Dworkin, because social meanings are not in fact shared. They are “endlessly contested and debated.” Our political arguments, he says, begin with disagreements about how this or that good functions in our social life and about what principles are appropriate to its distribution. Such disagreements can’t be resolved unless we move outside our own traditions and understandings and appeal to “general” principles. We have to turn to an “inclusive formula that can be used to measure justice in any society” (my emphasis). I expect that Dworkin plans to propose such a formula, and I don’t doubt that his proposal will be technically interesting and philosophically skillful (see his provisional statement, which is already very clever and very elaborate, in Philosophy and Public Affairs, Summer and Fall, 1981). But it probably won’t surprise him if people find his formula contestable and if the debates go on “endlessly.” The suggestion that someone is going to end the contest with a single knock-down argument is a piece of philosophical impetuosity.

What is the best way of carrying on the argument? Curiously, Dworkin has himself provided an excellent model in his account of how “hard cases” ought to be decided in a legal system like our own (Taking Rights Seriously [Harvard University Press, 1977], pp. 81-130). Hard cases are contestable cases, and in an important sense the contest is endless: the judge’s decision is merely one moment in an ongoing argument. But in principle, Dworkin insists, the judge can make the right decision. Exactly how he does this is not entirely clear; Dworkin’s account is at once persuasive and ambiguous, and I shall pursue only one strand, which seems to me the most powerful strand, of his reasoning. The judge reaches the right decision not by appealing to principles external to the legal system, but by exploring the internal principles of the system itself—and of the legal and political culture in which it is embedded. He searches for “the political morality presupposed by the laws and institutions of the community” (p. 126). That is exactly the procedure that I too would recommend, and for all my “relativism,” I share Dworkin’s sense that in a particular case, in a particular culture, there is, in principle, a right decision. To be sure, Dworkin stipulates a miraculously intelligent judge, a stand-in for the author, who is named Hercules. But Hercules is not privy to some universal theory of justice; he is merely superhumanly learned in his own tradition, patient and skillful in studying its history, its underlying philosophy, and its institutional details. He teases out the deepest understanding of the “legal community.”

I can think of no better way of discovering the appropriate distributive principles for medical care, political power, bureaucratic office, education, punishment, and so on. The moral world, of course, is more loosely structured than the legal world: there is no basic text like the constitution and no authoritative decisions (precedents), but there is a history; there are institutions and practices and underlying ideas. We must interpret these as best we can in order to get some grasp on the character of the goods we distribute to one another (they have no abstract or general character, or none that will help in determining who, among ourselves, should get how much of what). There will be different interpretations and, absent Hercules, no final and definitive interpretation. But that is not to say that we can’t mark off better from worse arguments, deep and inclusive accounts of our social life from shallow and partisan accounts.

Dworkin seems to think that such a procedure will have no critical bite. But he recognizes readily enough that Hercules can be a legal critic, and he ought to recognize that I can be (I often am) a social critic. Social critics commonly don’t, and certainly needn’t, invent the principles they apply; they don’t have to step outside the world they ordinarily inhabit. They appeal to internal principles, already known, comprehensible to, somehow remembered by, the people they hope to convince. Most often, as I have tried to show, they claim that such and such a good is not being distributed in accordance with its own meaning and the principles that flow from that meaning, but has been usurped and tyrannically controlled by men and women who hold some other good. They complain that religious communion is available only on terms set by the powerful, or that offices are bought and sold by the wealthy, or that pubishment falls upon the poor rather than upon the guilty, or that the best commodities and services are reserved for members of the Party, and so on.

But this sort of complaint won’t make for determinate criticism, according to Dworkin, unless one assumes (as he claims I do) that the different spheres of distribution, within which this or that principle is appropriate, are “fixed and preordained,” established “on an all-or-nothing basis.” I don’t in fact make any such assumption; indeed, it is contrary to the method and intention of my book. Social goods and distributive spheres have first to be found through a process of empirical investigation, and then they have to be understood through a process of interpretation. They have the forms they take in a particular society; there are no preordained forms. It is entirely possible, on my view, that for some goods we will have complex rather than unitary distributive principles; and it is possible too that the boundaries we draw around a particular good, our understanding of what counts as a commodity, an office, or a punishment, will change over time.

“We cannot just rule out in advance,” Dworkin writes, “the possibility that though justice requires the state to intervene in the market for medicine in order to ensure that the poor have some care, it does not require that the poor be provided the same medical care the rich are able to buy. Walzer takes the opposite view that justice demands a full national service.” Not quite right: I don’t rule out in advance, but consider and reject the possibility that Dworkin raises; and I argue that “justice demands” a more egalitarian distribution in our society because of what medical care means to us, the value we collectively assign to it, and the decision, already made, to provide it out of communal resources for some but not all the members of the community (Spheres of Justice, pp. 86-91). The argument is historical, sociological, contingent. Dworkin wants an entirely different kind of argument, so that one might say at the end, flatly, that a rich society that leaves medical care to the market “would not be a just society.” I am in fact disinclined to say that just like that, for it may be the case that the wealth of some particular society ought to be spent on the cure of souls, not of bodies, or on defense, or drama, or education. I don’t see how these priorities can be philosophically determined. But that is not to rule out radical criticism, for the actual distribution of salvation, security, and culture is likely to be distorted, has historically been distorted, by wealthy and powerful elites, and it is one of the tasks of moral philosophy (and of social theory too) to explain and condemn the distortions.

The case is the same with place and office. Here I criticize an argument that Dworkin made some years ago about the Bakke case (NYR, November 10, 1977), and he responds to my criticism, though without quite saying what is going on. “Walzer…believes that a certain conception of talent is automatically assigned to certain university places and professional offices, no matter how thoroughly the community is divided about this.” Not quite right: I argue at considerable length (Spheres of Justice, Chapter 5) against the “automatic assignment” that meritocrats commonly make, and I try to account for the leeway that search and selection committees have and ought to have. But such committees don’t have infinite leeway; powerholders are constrained in the way they give out offices by what offices are in the United States today. Dworkin thinks of offices, or at least of a very large number of offices, as a kind of currency, a social resource, to be distributed on essentially utilitarian grounds. But offices, like all other social goods, are not just things lying about to be used in any way we please; they are the products of a particular history; they have a special meaning in our culture—a meaning in this case hinted at though by no means fully revealed by, the revolutionary slogan about careers open to talents. I try to explore that meaning, inadequately, I suppose, since I am not Hercules. But without such an exploration, no distributive claim can ever be anything but arbitrary, no enforcement of a claim anything but tyrannical.

Let me come back to the problem of disagreement. Dworkin has some fun with the notion that people disagree about the meanings they share: if they disagree, he says, there are no shared meanings. In fact, two different kinds of disagreement are possible, which I have probably not sufficiently distinguished. First, people can disagree within a cultural tradition. They interpret meanings in somewhat different ways, or they take different positions on boundary disputes and on overlapping or entangled goods. This is the best way to understand current arguments about quotas and affirmative action. Dworkin’s position on these questions has not figured in the actual debates; perhaps it is genuinely original, a philosophical feat. The standard arguments are very different from his, and they all pay tribute (there is honesty and hypocrisy on both sides) to the common understanding of office—which connects its distribution, though not very strictly, to talent and performance. This kind of disagreement displays rather than denies the existence of shared meanings.

Second, people disagree because they come out of radically different cultural traditions, as in many third world states today. I was thinking of cases of this second sort when I suggested that divided societies might have to provide “alternative distributions.” Does this mean, Dworkin asks, “medical care for the poor in some cities, but not in others”? It might mean that, if the different cities were inhabited by people with, say, radically different understandings of medical care. If the populations were mixed, as they most often are, then it might be (morally) necessary to work out a political accommodation. Politics must sometimes substitute for justice, providing a neutral frame within which a common life slowly develops. (Even this is a critical idea, allowing us to deny the justice of imposed, rather than negotiated, distributions.)

Dworkin’s deepest worry is that I am a relativist. (He is especially concerned about my views on the Indian caste system, of which he provides a very partial account: but that is an issue best left to another occasion.) I see his point, though most relativists would think me just as tiresomely judgmental as he is. Indeed, we share a desire to reach moral conclusions. We only disagree about the force or, better, the scope of our conclusions. I don’t hope to make arguments that are conclusive for all human beings in all societies that exist or will exist or have ever existed. I don’t subscribe to the idea—it seems to me distinctly odd—that the principles of justice appropriate to Americans must be appropriate as well to ancient Babylonians. Not that such an idea makes it impossibly hard to arrive at principles of justice; it makes it too easy, for the principles need not apply to anyone in particular. The hard task is to find principles latent in the lives of the people Dworkin and I live with, principles that they can recognize and adopt.

Michael Walzer

Princeton, New Jersey

Ronald Dworkin replies:

I appreciate Professor Walzer’s thoughtful letter, but I do not think it meets my argument. I said that his theory, which makes justice depend on shared conventions, cannot be useful for us because our society is divided rather than united over which principles of distributional justice to apply to different types of resources. Many Americans disagree about how far medical care, for example, should be distributed in accordance with need rather than with ability to pay.

Walzer now replies that there are two kinds of disagreement about justice: disagreement within a cultural tradition when people, “interpret” their own conventions and practices in somewhat different ways, and disagreement between very different traditions. He believes that our own political differences—about medicine and affirmative action, for example—are disputes of the first sort, and that his conventionalist account of justice permits him to take up critical positions in these debates, because it permits him to argue about what our traditions, properly understood and interpreted, really require.

It is unfortunate that Walzer did not develop this idea in his book, for he would then have told us what, in his view, could make one interpretation of a moral tradition better than another. Our record in providing medical care is a mixed one. The state sponsors some medical research, and we have some federal and state programs of free medical care for those most in need. But the rich have always been able to buy much better medical care for themselves. How shall we interpret this record? Shall we say that our traditions assign medicine to the market, with some inconsistent exceptions that should now be abandoned? Or that they assign medicine to the sphere of need, but with inconsistent backsliding in favor of wealth and privilege? Or that they express the more complex principle that justice requires leaving medicine to the market but insists on just the qualifications and exceptions that we have made? What could make one of these interpretations superior to the others?

Walzer says that he can think of no better account of the process he has in mind than my own description of the way judges decide hard cases at law. The ideal judge I imagine—Hercules—interprets the immense variety of past judicial decisions in his jurisdiction so as to find the scheme of principle that “underlies” them, and then decides new cases that come before him by applying that scheme. The set of principles he deploys as the best interpretation may not fit all the past decisions; it may show some of these to have been “mistakes” which should now be eliminated to achieve integrity in the law as a whole. Walzer suggests that we take my account of interpretation of past judicial decisions as a model for interpreting and reforming the moral culture of the community as a whole.

But this partial statement of my views about adjudication simply raises the same puzzle about interpretation in a new form. What does it mean to say that a particular principle—that medical care must be treated as a matter of need, for example—“underlies” the political accommodations we have reached? How is that metaphor to be unpacked? Walzer seems unaware of how I myself answered that question. I said that a principle underlies a body of rules if it provides the best available justification for those rules, and I emphasized that this is not simply a mechanical test of counting how many past decisions “fit” the proposed principle. I said that if different schemes of interpretation each fit a great many rules but are inconsistent with others then Hercules must choose, as the “correct” interpretation, that which in his view comes closest to what abstract justice would require.

But of course this makes adjudication in hard cases turn on judgments of political morality that different judges will make differently. (Many critics reject my argument for that reason.) So Walzer cannot adopt my account of judicial interpretation without undermining his entire case. He would then have to say that the best interpretation of our arrangements about medicine, of those I just described, is the interpretation which is to be preferred on grounds of abstract justice. This would concede what he mainly wishes to deny, that justice is at bottom independent of the conventional arguments of any particular society. For if he continued to claim that justice is only a matter of what these conventions, properly interpreted, provide, his argument that one interpretation comes closer to abstract justice would be wholly circular and wholly ineffectual.¹

So if Walzer is to protect his main position he cannot use my analysis after all. What other options does he have? I suggested one in my review. He might think that the various “spheres” of justice he describes are each preordained and distinct, so that an interpreter could ask whether our practices about medicine, taken together, are closer to those required by the pure, preordained sphere of need than they are to the pure, preordained sphere of market transactions, and then insist that they be reformed so as to be entirely like those of the sphere they most resemble. Of course these judgments of resemblance would be impressionistic—two interpreters might disagree—but the process does not seem to require any assumptions about nonconventional requirements of abstract justice. So Walzer could accept this story without abandoning his general position. That is why I suggested, in my review, that the idea of fixed, preordained spheres is the “hidden” and “tacit” premise of his argument. I meant that I thought he was relying on that idea whether he understood this or not.

I am not surprised that he disowns the idea once it has been made explicit. But how else can he explain how one interpretation of our practices can be better than another when it is controversial which is better? Why is the argument defending the status quo in medical care—that the balance between market and need that politics has achieved itself provides the best interpretation of its moral traditions with respect to medicine—any worse than the argument for reform in either direction? If Walzer says that the compromise is a poorer interpretation because it is unprincipled, because it does not express a coherent and defensible vision of justice, then he is appealing to the idea of abstract justice he rejects; if he says it is illegitimate because it does not consistently enforce either the market or the need model then he is appealing to preordained spheres. Perhaps he can provide a different account of interpretation that will escape both these problems, but he has not yet indicated what this might be.

Walzer says that the remarks I quoted in my review, about what justice requires when “people disagree about the meanings of social goods,” were intended to apply to disagreement of the second sort, between rival moral traditions. He says that in such cases politics must substitute for justice. This is the heart of our disagreement. The idea that the world is divided into distinct moral cultures, and that it should be the goal of politics to foster the value of “community” by respecting the differences, has for a long time been associated with political conservatism and moral relativism. It is once again fashionable in political theory, but its proponents have paid insufficient attention to their central concepts. Moral traditions are not clubs into which the peoples of the world are distributed so that everyone carries a membership card in one but only one. On the contrary, these traditions can be defined at different levels of abstraction, and people who belong to a common tradition at one level of abstraction will divide at another, more concrete, level.

American liberals and members of the “Moral Majority,” for example, belong at one level to a common moral tradition, because they share an idea that the state must act justly toward all, and share many opinions about what justice is. (So do two people from what Walzer calls radically different cultural traditions, like citizens of different “third world” countries.) But American liberals and members of the “Moral Majority” do not share a common moral tradition at another, more concrete, level because they disagree sharply about the role that justice permits or requires the state to play in the moral lives of its citizens. We can, if we wish, say that one side or the other better “understands” the more abstract principles about which they are agreed; that liberals have a better theory of what justice really requires. But this is not a neutral, anthropological judgment we can defend simply by studying the practices of American society. It is a moral judgment taking sides in the dispute, the kind of judgment that would make no sense if justice were simply a matter of convention. So the idea of a shared moral tradition cannot do the work Walzer wants; if society is divided on some issue, the tradition runs out where the dispute begins. If we accept his dictum, that politics must then replace justice, justice all but disappears for us. We are left with the politics of selfishness.

Walzer makes some harsh remarks about my own views about justice, though he does not say which of my views he has in mind. He says that my arguments provide only “easy” answers, that because they try to ground particular claims about particular societies in more general critical principles they can apply to no one in particular. He does not elaborate, and since he offers no examples I find it difficult to respond or even to see what he means. He mentions only my views about affirmative action, and what he calls my “elaborate” arguments about distributional justice, but he says these are both contentious and offers no reason to doubt my claim that anyone who does accept them is committed to particular positions in the practical debates of American politics.

He seems to assume that an argument can have no practical bite if it begins in principles of some generality. The truth, as I have tried to show, is quite the contrary. Of course an argument of political morality cannot reach particular conclusions without taking into account relevant features of the community under discussion. Of course “we cannot distribute goods to men and women until we understand what these goods mean…among those same men and women.” There is indeed no reason to expect that the concrete conclusions we reach about what justice requires for us will apply as well to the ancient Babylonians. Certainly an abstract theory of justice is useless politically if it cannot be advanced in practical politics. But the principles of justice we use to decide which features of a community are relevant to a just distribution of its goods and opportunities—and therefore which principles we should struggle to promote politically—must be principles we accept because they seem right rather than because they have been captured in some conventional practice. Otherwise political theory will be only a mirror, uselessly reflecting a community’s consensus and division back upon itself.²