Political Liberalism (1993) is both the title of John Rawls’s second book and a rallying cry for philosophers, like Martha Nussbaum, who believe that “more than any other modern work of political philosophy [it] carries forward one’s hope for humanity in an era of religious and ideological turmoil.” ^{Footnote 1} Rawls’s book argues that we can establish a just and stable order for ourselves even though we differ profoundly about our conceptions of the good. But it insists that we cannot do so until we are ready to leave aside the “comprehensive doctrines” that we ordinarily rely on to justify our beliefs and take up what Rawls calls a “political conception of justice.”

Students of politics, you might think, should welcome Rawls’s so-called “political turn.” ^{Footnote 2} After all, a common complaint about his landmark Theory of Justice (1971) was that its success in defending liberal principles came at the expense of an almost complete abstraction from political experience. If the world’s most famous philosopher shifts gears and urges us to develop “political, not metaphysical” conceptions of justice, if he draws inspiration from “ideas implicit in the culture of a democratic society” rather than assumptions about rationality, ^{Footnote 3} if he focuses on public reasoning among democratic citizens rather than the choices made by individuals whose identities are hidden by a metaphorical “veil of ignorance,” does that not greatly enhance the importance of the study of politics? Perhaps. But in this case, we should be wary of philosophers bearing gifts. For Rawls overcomes the abstractness of his original defense of liberal principles by writing the philosopher’s interest in justifying arguments into his conception of liberal politics. Seeking to correct an overly philosophical understanding of liberal theory, he comes up with an overly philosophical understanding of liberal practice. And that overly philosophic picture of liberal politics has become the norm for many students of liberalism these days, especially among those who stand at the forefront of the efforts to convince us that public deliberation is the heart and soul of democracy.

Since so many scholars are inclined to treat Rawls and liberalism as near synonyms, few seem to have noticed how much his conception of political liberalism changes familiar understandings of liberal politics. One good way of bringing that change to light, along with a much more genuinely political understanding of liberal politics, is to compare Rawls’s conception of liberalism with that of his longtime Harvard friend and colleague, Judith Shklar. Shklar was also interested in developing a more distinctly political conception of liberalism, as Rawls himself acknowledges. ^{Footnote 4} But while Shklar and Rawls were working on parallel tracks, they were moving in opposite directions. For Shklar’s political liberalism calls for a return rather than a departure. Her “liberalism of fear” points liberal theorists back to what she views as liberalism’s original and abiding purpose—taming the disruptive impact of political power on our lives—rather than forward to a new way of justifying liberal principles and practices. Shklar worried that an increasing focus on the identification and justification of rights was leading academic liberals to lose touch with the problems that inspired their interest in rights in the first place. ^{Footnote 5} Viewed from this perspective, Rawls’s political liberalism, with its even greater emphasis on how we justify our arguments to each other in public life, actually accelerates and deepens a problem it sets out to correct.

Scholars have begun to pay a lot more attention to Shklar’s understanding of liberalism in recent years. But analysis of her work remains something of a cottage industry compared to the vast output of Rawls scholarship. While Thom Brooks and Martha Nussbaum’s Rawls’s Political Liberalism, a collection of essays celebrating its twentieth anniversary, is only the latest in a whole library of books about Rawls, ^{Footnote 6} Andreas Hess’s intellectual biography, The Political Theory of Judith N. Shklar, published 22 years after her death, is the first book wholly devoted to her political thought. ^{Footnote 7} The appearance of these two very useful books provides an appropriate occasion for opening a dialogue between Shklar and Rawls on their competing conceptions of liberal politics.

It is a little difficult, however, to get this dialogue started, given the very different ways in which the two friends wrote and argued. Shklar, one-time president of the American Political Science Association, insisted that she was a political theorist, not a philosopher, by which she meant that she was interested primarily in the political impact of arguments rather than their intrinsic strengths and weaknesses. ^{Footnote 8} Rawls, one-time president of the American Philosophical Association, was, by contrast, the philosopher’s philosopher, carefully weighing each step in an argument, testing and adjusting his premises and conclusions in response to decades of exchanges with other philosophers. Rawls, you might say, was the greatest hedgehog in philosophic history, to invoke Isaiah Berlin’s famous metaphor. He spent 50 years nurturing, refining, and recasting one big idea until it became the foundation of a whole new way of doing moral and political philosophy. Compared to Rawls, Shklar was the foxiest of foxes, to continue with Berlin’s metaphor. She was so impatient to move on to the next subject or insight that attracted her interest that she devoted little attention to explaining why she approached things in the ways that she did. ^{Footnote 9} Getting these two thinkers to engage with each other therefore requires some redirection and extension of their explicit arguments. But it can be done, especially if one draws out some of the implications of Shklar’s books on law and justice, Legalism and The Faces of Injustice. ^{Footnote 10}

The best place to begin this dialogue is probably Rawls’s account of political legitimacy. ^{Footnote 11} “Our exercise of political power,” Rawls writes, “is proper only when we sincerely believe that the reasons we offer for our political action may be reasonably accepted as a justification of these actions.” Such, he insists, is the “liberal political of legitimacy.” ^{Footnote 12} This principle provides Rawls with a guide to how we should address one another as citizens, not just a standard against which to measure the legitimacy of our basic political institutions. “For only a political conception of justice that all citizens might be reasonably expected to endorse can serve as a basis of public reason and justification.” ^{Footnote 13} What renders Rawls’s liberalism “political” is the restraint that he expects liberals to impose on themselves in order to make arguments that could be endorsed by their fellow citizens.

This emphasis on public justification has come to be seen as the “moral lodestar” of liberal political thought by most of the philosophers who currently dominate discussions of liberal theory. ^{Footnote 14} “The foundation of liberal thought,” Jeremy Waldron tells us in an influential 1987 article, “is based on a demand for the justification of the social world.” What makes a position “fundamentally liberal is this:” the insistence that “a social and political institution is illegitimate unless it is rooted in the consent of all those who have to live under it.” ^{Footnote 15} Even self-styled liberal realists—that is, people who ordinarily challenge Rawlsian approaches to political philosophy—describe the demand for public justification as the foundation of “the liberal vision of the political.” ^{Footnote 16}

As Shklar would be quick to remind us, however, identifying “the liberal vision of the political” with contemporary notions about public justification leaves us a big problem: This vision of liberal politics is quite foreign to almost every major liberal thinker between Kant and Rawls, and quite a few who preceded Kant, such as Montesquieu, Smith, and Hume. If you asked most people to name a liberal principle of political legitimacy in the years before this new paradigm of liberal politics became popular, they would most likely have pointed to notions like Mill’s liberty principle or Montesquieu’s separation of powers, principles that constrain the way we exercise power rather than how we justify it. In other words, they would have described practices much closer to Shklar’s conception of liberal politics than Rawls’s.

The new emphasis on public justification moves the focus of liberal politics from something negative to something positive, from trying to locate the points where the exercise of political authority should be deemed questionable or illegitimate to trying to identify the conditions that would make it fully acceptable or legitimate. That might sound like two sides of the same coin, but it is not. For, as Shklar shows in The Faces of Injustice, we should not assume that virtues and vices mirror each other, that negative concepts, like injustice and illegitimacy, are negations of their positive counterparts. In this case, the positive formulation searches for an ideal of political legitimacy that can turn a negative into a positive, the exercise of power over others into a kind of autonomy or self-legislation. The negative formulation, in contrast, takes it for granted that states will exercise power over individuals and seeks means of moderating, regularizing, and rendering more accountable our use of that power. It focuses our attention on the problems associated with the disruptive effects of coercive authority rather than the problems associated with making it justifiable to free people.

Because academic philosophers, like almost everyone else, tend to interpret the activities of others in terms of those with which they are most familiar, it is hardly surprising that they are inclined to identify liberal politics with the activity to which they devote the greater part of their professional lives: the justification of arguments. Nevertheless, their identification of liberal politics with justification itself needs to be justified. ^{Footnote 17} We need to ask what is gained by shifting the meaning of liberal politics in this way and whether it makes up for what may be lost by doing so.

The answer to the first of these questions seems, at least at first, to be quite straightforward: justifying liberal regimes in ways that would be acceptable to people with conflicting beliefs about how to live “carries forward one’s hope for humanity in an era of religious and ideological turmoil.” ^{Footnote 18}

The establishment of a reasonably just and stable order among the world’s warring tribes is a very great good indeed, something most of us would sacrifice quite a bit to achieve. But, on closer examination, the good actually offered us by Rawls’s political liberalism is considerably less impressive. For Rawls’s political conception of justice, we learn, will be acceptable only to holders of “reasonable” comprehensive doctrines about morality and the good life; and “reasonable” comprehensive doctrines exclude those that seek to force themselves on others or reject democratic constitutionalism. ^{Footnote 19} In other words, in order to be “reasonable,” a comprehensive doctrine must “endorse a liberal political conception.” ^{Footnote 20} So what Rawls actually offers is a path to justice and stability for people who already share a commitment to liberal democratic principles. So much for the hope of curtailing communal violence—unless you locate its source in the kind of disagreements that set liberal Protestants against Reform Jews or liberal Kantians against rule-utilitarians. If you are looking to ease the modern world’s “religious and ideological turmoil,” then you will need to look elsewhere for help, to people who are trying to find ways of encouraging those who profess “unreasonable” doctrines to act more “reasonably.”

Rawls’s theory offers liberal democrats the possibility of a common framework within which to adjudicate their arguments about basic social and political questions, a framework that should improve the quality and civility of the way in which they challenge each other’s positions in public life. These are not inconsiderable goods. But they hardly match the promise of establishing a just and stable peace among the world’s warring tribes. As a result, most of us would be willing to sacrifice far less to achieve them. Rawls does not seem bothered by this, since he seems to assume that there is little cost for liberals in pursuing a shared political conception of justice and the constraints on public reason that follow from it. Shklar, however, would strongly disagree. Rawls’s political liberalism, she would argue, makes liberalism both less political and less liberal.

Rawls’s conception of political liberalism de-politicizes liberal practice by trying to squeeze it into the confines of what Shklar calls a “legalist” model of morality and politics. Legalism, for Shklar, is a broad moral ideology that encourages us to measure each other’s behavior against a preexisting set of general rules and principles. As such, legalism models moral and political reasoning on adjudication. It makes the discovery of the relevant general rule and its impartial application to particular cases the key steps in any practical judgment. ^{Footnote 21} Shklar shares with Rawls a strong commitment to legalist norms when it comes to defining and constraining the exercise of political authority. But she parts from him in his application of legalist norms to political argument.

Rawls’s insistence that a just and stable society requires its citizens to share “the very same principles of justice” ^{Footnote 22} reflects his legalist approach to political argument. For it voices his belief that if liberal societies want to avoid the cynicism, violence, and raw exercise of power that he deplores, political argument must become more like legal argument: It must identify a shared set of principles against which to adjudicate the claims that we make against each other, not just a shared set of institutions to manage our disagreements. Shklar rejects that conclusion for two reasons. First of all, she rightly believes that there is a large body of relatively stable and decent liberal practice—the whole history of liberal societies, in fact—that lies between Rawls’s ideal of a well-ordered society and the specter of violence and amorality that he conjures up as its only alternative. Second, she believes that consensus about our conceptions of justice is no more likely than consensus about our visions of the good, and so liberals inevitably end up imposing beliefs on people when they seek common principles of justice within which to adjudicate our arguments against one another.

In A Theory of Justice, Rawls invokes a legalist virtue, fairness or impartiality, in order to help us identify the principles of justice that free and rational individuals should endorse. But in order to generate his famous two principles there, he has to rely on a substantive and very controversial interpretation of rationality, the so-called maximin principle that prescribes the rationality of maximizing minimum returns. In Political Liberalism, he is more inclined to derive these principles of justice from the “fundamental ideas implicit in the culture of a democratic society.” ^{Footnote 23} But, as Shklar complains in another context, these convenient “intimations of shared meaning … are never checked against actual opinions, least of all those of the most disadvantaged and frightened people.” ^{Footnote 24} Were these conclusions about the principles embedded in our political culture investigated in this way, it would quickly become clear that the facts of cultural life are no less complex and controversial than the facts of nature or reason. Shklar therefore urges liberals like Rawls to give up their quest for a consensus on basic principles of justice and listen more carefully to the discordant voices that exist within their communities. “Conflict among ‘us,’” she declares, is “both ineluctable and tolerable, and entirely necessary for any degree of freedom.” Far from searching for some shared principles to adjudicate our disputes, Shklar strains “to make ‘us’ even more aware of our incompatibilities and their consequences.” ^{Footnote 25}

Legalist ethics assume our access to a body of shared rules and principles that can be applied impartially to particular cases. But political argument is largely about the making of general rules and principles, rather than their application. And it is advanced by interested parties to other interested parties in the name of a common or shared interest, not by interested parties to third parties to whom we look for fair and impartial applications of preexisting rules and principles to our cases. Political argument is therefore ill-suited to the legalist constraints that Rawls seeks to impose on it. That becomes clear when we look at the models of public reasoning that he proposes to the members of liberal democratic societies.

Rawls suggests that if we want a quick test of whether our arguments measure up to the demands made by his conception of public reason, we should ask ourselves “how would our argument strike us presented in the form of a Supreme Court opinion.” For the judiciary is the “branch of government that serves as the exemplar of public reason.” ^{Footnote 26} But Supreme Court opinions are products of institutional constraints that have no place in the life of ordinary democratic citizens. They justify authoritative rulings, declarations of what counts as law, rather than opinions exchanged by citizens in the course of public deliberation. (In fact, judicial deliberation is conducted behind closed doors, and from what we have learned about it, does not seem constrained by the niceties of Rawlsian public reason.) These rulings answer questions about the meaning and compatibility of pre-existing rules and principles, rather than questions about the best way for us to proceed. Even when they really do involve judgments about the best way to proceed, those judgments are justified in terms of the best interpretation of preexisting rules; for Supreme Court judges are authorized to make rulings on the basis of their expertise on rules and procedures, rather than their ability to persuade us of the best policies or principles to pursue. And these rulings come down from figures who are raised above us by their role as supposedly impartial third parties to our disputes, rather than from fellow citizens who share with us an interest in the direction of public policy.

Since the constraints on the arguments made in Supreme Court opinions follow from the distinctive role that the court plays in our constitutional structure, an ordinary citizen would appear ridiculous were he or she to model her public speech on Supreme Court opinions. And a political leader who argued as if he or she were impartially applying shared preexisting principles, rather than making a judgment about what serves the common good, would appear irresponsible and, most likely, dishonest. ^{Footnote 27} Rawls cites the example of judicial reasoning partly to demonstrate that there is nothing unprecedented or bizarre about self-imposed constraints on political argument. But his failure to note or comment on the influence of different institutional settings on the way in which we reason in public is troubling. It makes one worry that he simply assumes the value of legalist models of public reasoning without considering the political context within which they are to be practiced.

Shklar would also draw our attention to the way in which Rawls’s emphasis on shared standards of public reason can undermine crucial liberal values. If, as she insists, it is the disruptive impact on our lives of the exercise of coercive authority that should most concern liberals, then liberals need to be especially attentive to what people who feel victimized by such exercises of authority are trying to tell us: “The voices of victims must always be heard first, not only to find out whether officially recognized social expectations have been denied, but also to attend to their interpretations of the situation.” ^{Footnote 28} Demanding that harassed, underrepresented, and often socially isolated citizens show respect to their fellow citizens by presenting their complaints in a “reasonable” manner—that is, without reference to beliefs or passions that their fellow citizens cannot share with them—adds insult to injury. For it suggests an ethical lapse on the part of people who, more often than not, are compelled to shout very loudly in order to be heard at all.

Shklar’s insistence on giving priority to the victim’s voice is part of a larger argument about the limits of what she calls our “normal models of justice,” the bodies of basic rules and principles that we devise to govern our social interactions. Such models, she argues, are useful means of regularizing and moderating the exercise of coercive authority; but they can never provide us with complete and accurate guides to the injustices that they are designed to regulate. For some aspects of injustice become apparent only to those who suffer from it. Hence, Shklar’s emphasis on the importance of paying heed to victims. They have something to teach us, not just a claim on our sympathy. ^{Footnote 29} But whatever one thinks about Shklar’s broader argument here, she gives liberals reason to worry about the extra burden that Rawls’s standards of public reason impose on victims struggling to be heard. She thereby shows that there may be a heavy price to pay for switching the focus of liberal politics from taming power to generating consensus about our conceptions of justice.

Finally, we need to consider the damage that Rawls’s way of talking about political legitimacy might do to settled liberal practices were it to be taken up by modern democrats. Imagine how the following statement would sound if uttered by a prominent public figure in a modern liberal democracy: “When, on a constitutional essential or matter of basic justice, all appropriate government officials act from and follow public reason, and when all reasonable citizens think of themselves ideally as if they were legislators following public reason, the legal enactment expressing the majority is legitimate law.” ^{Footnote 30} Rawls sees nothing controversial in such a statement, since it expresses the norm of justifiability that he identifies with the liberal principle of political legitimacy. But that is not the norm that most citizens today associate with the concept of political legitimacy. Hearing a statement like this, they would probably assume that its speaker was questioning the legitimacy of all the laws that were passed by majorities without the benefit of all this “reasonable” behavior on the part of citizens and legislators. For in ordinary public reasoning today, legitimacy is usually treated as an either-or rather than a more-or-less concept. It describes a threshold condition that exercises of public authority must meet in order to avoid illegitimacy, not a continuum of acceptability. Declaring that “the legal enactment expressing the majority is legitimate law” only when we justify ourselves to each other according to Rawls’s standards of public reason, standards that we rarely live up to, would therefore suggest the failure of most of our laws to meet the minimal conditions of legitimacy.

Of course, that is not at all what Rawls means to say in the passage I have quoted. He is invoking legitimacy as a positive ideal, something that we approach the more completely we meet certain standards, rather than as a minimum condition that every acceptable exercise of public authority must meet. The problem with doing so, however, is that many of the most settled and effective liberal practices, including lots of those that we rely on to tame and regularize the use of power, rely on the older, threshold concept of political legitimacy. Indeed, it is one of liberalism’s greatest achievements to have accustomed so many of us to looking for and rejecting as illegitimate exercises of authority that fail to cross that threshold to legitimacy. The introduction of an alternative understanding of political legitimacy, one that treats legitimacy as an ideal that we approach by degrees, would confuse matters and possibly undo much of that achievement. Rawls’s failure to recognize and account for the way in which his vision of liberal politics departs from his predecessors thus has practical consequences. For current liberal practices, practices that Rawls takes for granted, are often grounded in the conceptions of liberal politics that he leaves behind.

One could, I suppose, complain that I have given Shklar an unfair advantage over Rawls, since almost all of the issues discussed in this essay have played to her strengths rather than his: political conflict, moral psychology, injustice, victimhood, and so on. Had I focused the dialogue between them more on the logic, consistency, and clarity of their arguments about justice, rather than their practical consequences, the result would no doubt have been different. But Rawls chooses the more practical turf himself when he insists on pursuing a distinctly political understanding of liberalism. He is proposing standards to guide the way in which we engage other citizens in democratic forums, not philosophers in seminar rooms. If he does not see why Supreme Court opinions are completely inappropriate models for democratic public reasoning or how foreign most citizens would find his way of talking about political legitimacy, then that creates fundamental problems for his conclusions. A political conception of liberalism needs to investigate the beliefs that are actually expressed by the democratic culture that it invokes. It needs to consider the way in which different institutions structure different political roles before telling us what public reason demands of us. And it needs to consider the political costs, especially to the least advantaged, of pushing liberal societies to share “the same principles of justice,” not just the benefits. Otherwise, it is bound to inspire concerns that it merely replaces an overly philosophical understanding of liberal theory with an overly philosophical understanding of liberal politics.