Michael J. Sandel is the Anne T. and Robert M. Bass Professor of Government at Harvard University. His most recent book, Justice: What's the Right Thing to Do, is the distillation of a popular undergraduate course he has taught at Harvard for the past thirty years. The book is aimed at the general educated public, not at scholars.

One can see why the course is popular. The book is filled with fascinating examples, most of them taken from real life, of complicated situations in which it's hard to know what's the right thing to do. Here's an example that occurs early in the book. The U.S. military has traditionally awarded the Purple Heart to soldiers who were killed or wounded by enemy action. The Iraq war has resulted in an unusually large number of veterans who were not physically wounded but suffer from depression or severe post-traumatic stress syndrome. Some have argued that these veterans should be candidates for the Purple Heart on the ground that their "wounds" are often just as debilitating as the physical wounds that have traditionally qualified a veteran for the Purple Heart. Others have insisted that this would be a debasement of the Purple Heart. Who's right? It's not clear.

I received my copy of the book direct from the publisher. It was accompanied by several pages of ecstatic praise from a number of prominent scholars and intellectuals. "A terrific example of moral philosophy at work," said one. These words of praise all focus on the subtitle of the book; they praise Sandel, and rightly so, for the illuminating things he has to say about the right thing to do in various perplexing situations. But the title of the book is Justice. Sandel says that his project is to explore "the strengths and weaknesses of … three ways of thinking about justice." And the structure of the book consists of Sandel discussing in succession those three ways of thinking. The perplexing cases are introduced not for their own sake but to illuminate the strengths and weaknesses of these three ways of thinking. So the question Sandel invites the reviewer to ask is whether readers will emerge with a deepened understanding of justice. I think not. Readers will emerge knowing what's wrong with some theories of justice, but not with a deeper understanding of justice.

In the introductory chapter Sandel declares, "To ask whether a society is just is to ask how it distributes the things we prize—income and wealth, duties and rights, powers and opportunities, offices and honors. A just society distributes those goods in the right way; it gives each person his or her due." My guess is that Sandel's phrase "things we prize" was inadvertent; it gives a relativist twist to his explanation of justice. We need to be able to say that a society prizes things that are in fact bad, with the consequence, sometimes, that the society is unjust.

Apart from that little phrase, this is a pretty good description of a just legal structure; a just legal structure is one that distributes benefits and burdens in accord with what is due people. But having a just legal structure is not enough to make a society just, since there may be widespread violation of the laws. I think it's better to say that a society is just insofar as each individual and social entity is treated by the members and institutions of that society in a way that is due him or her.

If that's correct, then the central challenge facing any account of justice is to illuminate this idea of something being due a person. The main source of my disappointment with Sandel's book is that nowhere does he do this; nowhere does he even attempt to do this. If something is due you, then you have a right to it; to be deprived of it is to be wronged. Nowhere does Sandel give an account of what it is to have a right to something; nowhere does he give an account of what it is to be wronged. The concept of being wronged scarcely makes an appearance in his discussion.

Let's have before us Sandel's own description of the three theories of justice that he distinguishes and discusses. "One says justice means maximizing utility or welfare—the greatest happiness for the greatest number. The second says justice means respecting freedom of choice—either the actual choices people make in a free market (the libertarian view) or the hypothetical choices people would make in a original position of equality (the liberal egalitarian view). The third says justice involves cultivating virtue and reasoning about the common good." Sandel attributes the third approach to Aristotle. This is the approach that he himself favors and argues for.

Many readers will recognize that the first of these three approaches is utilitarianism. To my mind, it's a good question whether utilitarianism should be thought of as a theory of justice; the classical utilitarians never thought of it that way. But let that pass. Sandel quickly disposes of utilitarianism, with arguments that are now standard in the literature and which, to my mind, are decisive. The second theory he engages throughout most of the book, especially in its "liberal egalitarian" version as represented by Immanuel Kant and the late Harvard philosopher John Rawls. He employs Rawls as the main foil for the presentation of his own position.

Sandel locates the root of the disagreement between himself and Rawls in competing conceptions of the person. He ascribes to Rawls what he calls "the voluntarist conception of the person." The voluntarist, Sandel explains, understands human beings "as free and independent selves, unbound by moral ties we haven't chosen." He quotes Rawls' comment that "the self is prior to the ends which are affirmed by it; even a dominant end must be chosen from among numerous possibilities." The voluntarist "ideal," says Sandel, is that "of the freely choosing, unencumbered self."

Sandel calls his own view "the narrative conception" of the self. "We live our lives as narrative quests." He quotes Alasdair MacIntyre as saying, "I can only answer the question, 'What am I to do?' if I can answer the prior question 'Of what story or stories do I find myself a part?' " These stories tell of our solidarities with parents, with children, with friends, with colleagues, with fellow citizens.

Sandel holds that these conflicting anthropologies yield views of justice that conflict on two main points. First, the voluntarist holds that "we can reason about justice by abstracting from our aims and attachments" and from "competing conceptions of the good life." The narrativist denies this: "Many of the most hotly contested issues of justice and rights can't be debated without taking up controversial moral and religious questions. In deciding how to define the rights and duties of citizens, it's not always possible to set aside competing conceptions of the good life." Second, the Rawlsian holds that our duties or obligations are confined to obligations to particular human beings that we take on voluntarily and natural duties that we have to human beings in general. The narrativist argues, to the contrary, that in addition to these we have "obligations of solidarity or membership … for reasons unrelated to a choice—reasons bound up with the narratives by which we interpret our lives and the communities we inhabit."

Before I take up these two points at which Sandel sees disagreement about justice, let me say something about his contrast between two understandings of the self. Sandel's claim that Rawls holds a voluntarist view of the self as unencumbered, in contrast to his own narrativist view of the self as encumbered, goes back to his first book, an unrelenting attack on Rawls titled Liberalism and the Limits of Justice. I hold, as do a good many others, that the contrast Sandel professes to see is simply not there.

Above I quoted Sandel's explanation of what he means by a voluntarist: a voluntarist is one who holds that human beings are "unbound by moral ties we haven't chosen." Rawls does not hold this view; in other passages in Justice Sandel acknowledges as much. Rawls holds, to say it again, that we have natural duties. He explains natural duties as follows: "they apply to us without regard to our voluntary acts. Moreover, they have no necessary connections with institutions or social practices …. Thus we have a natural duty not to be cruel, and a duty to help another, whether or not we have committed ourselves to these actions" (Theory of Justice, §19). From this it's obvious that Rawls does not hold a voluntarist view of the self.

But what, then, are we to make of Rawls' statement, quoted by Sandel, that "the self is prior to the ends which are affirmed by it"? Doesn't this deny that we are "encumbered" selves in Sandel's sense, that is, deny that we have attachments, loyalties, sympathies, commitments and the like by virtue of the various communities and associations in which we have been reared and within which we live? It does not. How could anyone in his right mind deny that we are encumbered selves in this sense? The Rawlsian self is as much encumbered as the Sandelian self.

If one reads the sentence quoted from Rawls in context, one sees that what Rawls is saying is not that all our ends (goals) have been deliberately chosen by us; he's saying that, whatever our ends and however we acquired them, we should subject them to moral judgment. We should ask whether they are morally permissible. And if we judge that they are not, we should reject them. And we can reject them. We are not mere slaves to the goals and ends that emerge in us in the course of living.

Does Sandel disagree? Does he hold that we cannot stand in judgment on, and reject, the ends that our narrative identities have produced in us? I doubt it. But if not, then his distinction between the narrative encumbered self and the voluntarist unencumbered self does not locate a disagreement between himself and Rawls.

Now to the two main disagreements about justice that Sandel sees between himself and Rawls—disagreements that he thinks are the result of their (supposed) disagreement on the nature of the self. The first is this: Sandel attributes to Rawls the view that issues of justice can be separated from competing conceptions of the good life. His own view is that they cannot. Sandel defends his view with two sorts of examples.

He cites our debates over abortion and stem cell research, arguing that at the bottom of these disputes is disagreement over whether aborting a fetus or disposing of a fertilized egg amounts to killing a human being. I think he's right in this analysis of these disputes; this is the basic issue. But if so, then these are not Aristotelian-type disputes over different understandings of the good life; they are disputes over when a human life is being taken.

Second, Sandel presents us with a number of his perplexing cases. Since they are all of the same type, let me arbitrarily choose the Casey Martin case. Casey Martin was a professional golfer who found it difficult to walk a golf course because of a circulatory disorder; he asked the Professional Golfers' Association to be allowed to use a golf cart during tournaments. The PGA refused. Martin took his case to court, arguing that the Americans with Disabilities Act (1990) required reasonable accommodations for people with disabilities, provided the accommodation does not fundamentally alter the nature of the activity. The PGA argued that using a golf cart does fundamentally alter the nature of the activity; becoming tired from walking the course is an essential part of golf. The case went up to the U.S. Supreme Court, with the Court finding in favor of Martin. Sandel observes that the "case raised a question of justice in classic Aristotelian form …. To resolve the questions about rights, the court had to determine the telos, or essential nature, of the game."

I fail to see how this is an example of appealing to a disputed understanding of the good life in order to make a decision about justice. The court did indeed find that to resolve the question about rights it had to decide whether riding a golf cart is compatible with the telos of golf. But determining the telos of golf is not "thinking about the best way to live." Golfers assume that the telos of golf is a good of a certain sort. The justices did not undertake to determine whether golfers are right about that, nor did they attempt to determine how golf fits into the good life. They simply asked whether riding a cart is compatible with the telos of golf.

Though none of Sandel's examples establishes his thesis that disputes about justice sometimes originate in disputes about goods, my own view is that the thesis is correct. My reason is this. I hold that a right is a legitimate claim to the good of being treated a certain way by one fellows; it's a way of being treated that is due one. But if that's correct, then we must expect that disagreements over whether I have a right to being treated a certain way will rather often be rooted in disagreements over whether that way of being treated really would be a good in my life.

The Casey Martin example can be used to make another point about Sandel's "Aristotelian" theory of justice. Does it follow, from the judgment that riding a cart is compatible with the telos of golf, that Martin has a right to be allowed to use a cart? Clearly not. Suppose we add that being allowed to participate in golf tournaments would be a good in Martin's life; does it then follow that he has a right to participate in PGA tournaments? It does not. Here's why: for each of us, there are lots of ways of being treated by our fellows that would be good for us but to which we do not have a right. I think it would be a good in my life if I were awarded the PuIitzer Prize some year. But I don't have a right to be awarded the prize. Martin has a right to participate in PGA tournaments because riding a cart is compatible with the telos of golf and because of the Americans with Disabilities Act. The point is of general importance. To get justice into the picture, we have to introduce something in addition to the idea of life-goods. We have to introduce the idea of its being due one to be treated a certain way by one's fellows. Aristotelian-type reasoning about the good life does not yield a theory of justice.

Let's move on to the second main disagreement about justice that Sandel sees between himself and Rawls. The issue is this: do we have natural duties, and natural rights corresponding to those duties, on account of our particular solidarities and not only on account of our shared humanity? For example, do children under normal circumstances have a natural right to be taken care of by their parents? Rawls quite clearly assumed that we do not have such rights and duties—though nowhere, that I know of, did he discuss the issue. Sandel employs some of his perplexing cases to argue that we do have such rights and duties. I think Sandel is right on this issue and that Rawls was mistaken.

I find Sandel's discussion of the matter seriously inadequate, however. As we have seen several times, Sandel advertises his book as the exposition of an Aristotelian-type theory of justice and a defense of that theory against alternative theories. Surely what is required of a theory of justice at this point is some account of natural duties and natural rights—some explanation of how it is that there are such duties and rights. No such explanation is forthcoming from Sandel, not even an attempt at such an explanation.

Such an explanation would have to fit not only the particular duties and rights on which Sandel has his eye but also the human duties and rights that Kant and Rawls emphasized. And it's my view that, in its treatment of human rights, it would have to account for the fact that even those human beings who are not capable of functioning as moral agents have natural human rights—those in a permanent coma, those who suffer from the last stages of Alzheimer's disease, those who are severely impaired from infancy. Kant and Rawls have no account to offer of the rights of such human beings. Neither does Sandel.

Let me briefly make one final point. Rawls holds that the fundamental virtue of a political order is that it treat all citizens fairly; "justice as fairness" is what he sometimes calls his theory. Rawls took it to be a salient feature of modern democratic societies that citizens have profoundly different views on God, the good, and the right; accordingly, he spent the latter part of his career asking insistently how a political order could treat all of us fairly whatever our views on God, the good, and the right. Sandel quotes Aristotle to the effect that it is the task of government to make people good, and he indicates that he agrees with Aristotle on this point. He presents this as part of Aristotle's theory of justice, and he treats the disagreement between Rawls and Aristotle on this point as a disagreement about the nature of justice.

I think it is misleading to characterize the disagreement between Rawls and Aristotle on this point as a disagreement about the nature of justice. Aristotle never suggests that justice is the fundamental virtue of the political order; justice for him is no more than one virtue among many. He never asked how the political order could be fair to all of us whatever our views about God, the good, and the right.

Rawls used his apothegm, "the right has priority over the good," to mean a rather wide variety of different things. Unfortunately so. But one thing he meant is that, for us in the modern world, the fundamental political question to be addressed is not how to embed in law some particular view about God, the good, and the right, but how to establish a political order that's fair to everyone whatever his or her view on such matters. I think he was right about this; it would be a profound mistake to revert to Aristotelian politics. I think Rawls was seriously misguided in his wan hope that it's possible to arrive at a consensus of some sort on what constitutes a just political order; I am not a Rawlsian. A fair political order is not and cannot be a morally neutral political order. But I think he was right in his insistence that justice is the prime virtue of the political order. I am reminded of what St. Paul says in his letter to the Christians in Rome: it is the God-assigned task of government to secure justice. Not to make people good, but to secure justice.

Nicholas Wolterstorff is Noah Porter Professor Emeritus of Philosophical Theology, Yale University, and Senior Fellow, Institute for Advanced Studies in Culture, University of Virginia. He is the author of Justice: Rights and Wrongs (Princeton Univ. Press).

Copyright © 2010 by the author or Christianity Today/Books & Culture magazine.
Click here for reprint information on Books & Culture.