Rawls on Religion and Public Reason
MAPH THESIS AWARDS 2013 By Gustav Steinhardt
For some theists in pluralistic democracies, political deliberation presents a moral dilemma. Their faith may incline them to understand all political values as ultimately expressions of the divine, and so they will want to make political arguments in terms of their religious views; yet they recognize, if they are reasonable, that their fellow citizens do not share those religious views, and so may find arguments in religious terms to be offensive. These religious citizens, then, could feel that the demands of their faith are fundamentally incompatible with the demands of citizenship in a pluralistic democracy.
John Rawls, in Political Liberalism, proposes a solution to this dilemma that has sparked a vigorous debate among political theorists. Rawls’s proposal is a conception of public reason that allows citizens to base their political views in their religious doctrines, but insists that religious reasons are not sufficient justification for those views under some circumstances. Thus, religion can be the basis of individual political conviction, but it is excluded from the public political forum.
In recent years, however, this proposal has come under fire from critics who see it as symptomatic of liberalism’s excessive discomfort with religion. Critics of Rawls, especially from within the Christian tradition, have worried that public reason imposes onerous and unnecessary constraints on some religious citizens insofar as it forces them to leave behind their religious convictions when they enter the public square. In other words, public reason, as Rawls describes it, may conflict with certain reasonable forms of religious life. Recent scholars, such as Jeffrey Stout, Nicholas Wolterstorff, and Christopher Eberle, have tried to generate new theories of public reason that are more accepting of religious citizens, and have generally concluded that the Rawlsian theory is inadequate to this task. These critics typically understand their project as an effort to break the cycle of mutual reinforcement that currently obtains between an old-fashioned liberalism (one that is deeply suspicious of religion) and a growing religious radicalism (one that is deeply suspicious of liberalism). These two political phenomena feed and encourage one another; as Stout puts it
: “we are about to reap the social consequences of a traditionalist backlash against [Rawlsian] liberalism. The more thoroughly Rawlsian our law schools and ethics centers become, the more radically Hauerwasian the theological schools become.” In order to reconcile liberalism with monotheism, and thus to avoid empowering radical fundamentalism, Stout and others like him find it necessary to seek alternatives to Rawls – alternatives that, hopefully, will placate the demands of both Rawlsians and monotheists.
I think, though, that Rawls anticipated his critics’ arguments in ways that have been overlooked. His theory of public reason is far more inclusive of religious voices than is generally understood, and thus the conflict between Rawlsian liberalism and monotheism may in fact be illusory. If true, this would indicate that Rawls himself provides the most promising lead in the conciliatory effort that Stout, Eberle, and their allies have rightly called for. To see this, of course, it will be necessary to examine in some depth the ambiguities in Rawls’s thought, and the ways in which his thinking changed over the course of his career.
I will argue for a reading of Rawls in which public reason imposes strict discursive requirements on public officials, but leaves citizens qua citizens free to justify their political views in terms of whatever religious or secular ideals they choose. I call this the “institutional view” of public reason. The institutional view consists of two parts: first, it insists that those involved in coercive decisionmaking must be able to justify their decisions in accordance with a Rawlsian conception of public reason; second, it requires citizens to try, in good faith, to reason in this way, but it acknowledges that some morally responsible citizens will be unable to do so in the end.
The first of these points draws on a proposal put forward by Charles Larmore, while the latter comes from the work of Christopher Eberle; unlike Larmore or Eberle, however, I conceive of my proposal as essentially a reading of Rawls rather than an alternative to Rawls. The Rawlsian conception of public reason seems to me to be compatible with the arguments of both Eberle and Larmore. Rawls does make claims that seem to be inconsistent with my reading, and it is these claims that provoke the worries of his critics, but I will argue that they can be overcome on internal, exegetical grounds. With the help of this reading, I will try to show that Rawls has ready answers to the most basic concerns of his religious critics
; under most circumstances, reasonable religious citizens need not worry that their faith and their commitment to public reason will come into conflict.
There is, however, one circumstance where possible conflict lingers. When citizens step into the polling booth, they seem to inhabit a liminal space in between the arena of political discussion (where public reason applies loosely, if at all) and the coercive institutions of the state (where it applies more strictly). A question remains as to whether citizens in the polling booth may vote solely in accordance with their religious views, or whether the constraints of public reason must prohibit them from doing so. My answer to this question is somewhat inconclusive, but I hope at least to indicate the outlines of the problem. I begin, then, with an account of the Rawlsian conception of public reason, including the limitations on its scope. With the basic form of public reason laid out, I will examine the changes Rawls made to it later on in his career.
I. Public Reason
Content of Public Reason
Rawls develops his conception of public reason in response to a problem of legitimacy under conditions of reasonable pluralism. In a pluralistic democracy, reasonable citizens will be irreconcilably divided on moral, philosophical, and of course religious issues, yet all must live under the same basic institutions and abide by the same set of laws. Given the vast diversity of justificatory frameworks this pluralism implies, and assuming that legitimacy requires justifiability to the coerced, how is it possible that one system of laws and institutions can be justifiable to all citizens and thus legitimate? Rawls’s answer is expressed in what he calls the “liberal principle of legitimacy”:
Our exercise of political power is proper and hence justifiable only when it is exercised in accordance with a constitution the essentials of which all citizens may reasonably be expected to endorse in light of principles and ideals acceptable to their common human reason. This is the liberal principle of legitimacy. To this it adds that all questions arising in the legislature that concern or border on constitutional essentials, or basic questions of justice, should also be settled, so far as possible, by principles and ideals that can be similarly endorsed. 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.
Later, in restating the liberal principle of legitimacy, Rawls adds this:
And since the exercise of political power itself must be legitimate, the ideal of citizenship imposes a moral…duty — the duty of civility — to be able to explain to one another on those fundamental questions how the principles and policies they advocate and vote for can be supported by the political values of public reason.
Public reason, then, requires citizens (under certain conditions to be specified later on) to seek and offer properly public justifications for their political positions — a responsibility that emerges from their commitment to the liberal principle of legitimacy. For Rawls, this minimally entails a good-faith effort to employ only principles that we believe our fellow citizens can, at least in principle, accept. We should not appeal to our own religious, moral, or philosophical worldviews (“comprehensive doctrines” in Rawls’s terminology) since in a pluralistic democracy many of our reasonable fellow citizens will not share these views. Justification in terms of any particular comprehensive doctrine therefore shows deficient respect for fellow citizens and constitutes a violation of the duty of civility. In other words, public reason only allows justification based on what we all share as reasonable citizens.
So what do reasonable citizens, who differ widely in matters of religion and philosophy, all share as such? Naturally, what they share is precisely their reasonableness. For Rawls, this means something very specific: reasonable citizens are those who seek, offer, and accept fair terms of cooperation with others regarded as free and equal. They “desire for its own sake a social world in which they, as free and equal, can cooperate on terms all can accept,” and they “recognize the independent validity of the claims of others.” Even if we disagree vehemently about religion, cosmology, morality, metaphysics, and a host of other vitally important topics, Rawls argues that we can at least agree on this much as reasonable citizens.
The shared perspective of reasonableness allows us to develop freestanding political conceptions of justice that can be affirmed from within a variety of comprehensive doctrines, but that nonetheless are derived from and depend for their validity on no one doctrine or subset of doctrines. Thus, the political conception of justice must, in accordance with the terms of reasonableness set out above, include some notion of human equality; citizens’ reasons for affirming that equality, however, will vary. A Quaker citizen might understand human equality as a consequence of the divine spark innate in the human soul, whereas a Kantian citizen might ground that equality in ideas of rationality and autonomy, and so on. Because the citizens in question are all reasonable, they can all affirm a political conception of justice, or at least a family of political conceptions, and thus share a common foundation of public reason. Their reasons for affirming that common foundation will vary along with their comprehensive doctrines; but ideally the foundation itself should remain constant. Rawls calls this kind of agreement an overlapping consensus.
Reasonable citizens, then, can form an overlapping consensus on a freestanding political conception of justice in light of their own comprehensive doctrines. When they justify their political views to one another, they will ground their justifications in that conception, and thus offer reasons that can be acceptable to all their reasonable fellow citizens. To reason in this way is to fulfill the “duty of civility.” But if we are to conclude that all reasonable citizens can fulfill the duty of civility, then the constraints it imposes must be grounded in the idea of reasonableness itself; otherwise, those constraints might turn out to be fundamentally in conflict with reasonable comprehensive doctrines. In the following sections, I will explore possible conflicts of this sort, and try to resolve them by rooting the duty of civility more firmly in the Rawlsian conception of reasonableness.
At this point, we need to distinguish two possible understandings of the duty of civility, as highlighted by Christopher Eberle: civility as pursuit and civility as restraint. The difference comes into view if we find, after a long process of deliberation in good faith, that we are unable to justify our views in terms of public reason. Civility as restraint would require that we change our views in this situation, whereas civility as pursuit would consider the duty discharged by our good-faith effort and would leave us free to fall back on our comprehensive doctrines. The wording of the duty of civility seems to suggest a principle of restraint, but Rawls is not explicit on this question; later on, we will see how Rawls’s statement of the duty of civility is open to either interpretation. In the end, I will argue it may be best to adopt a mixed interpretation in which public officials are saddled with a duty of restraint, whereas citizens qua citizens need only fulfill a duty of pursuit.
Leaving aside this slight ambiguity, it’s reasonably clear what Rawls means by public reason. We know what counts, by his lights, as a suitable justification. But by whom and under what circumstances is such justification to be attempted? It’s crucial on Rawls’s view that not all political conversations among all citizens at all times are to be governed by the norms of public reason.
Scope of Public Reason
The scope of public reason is limited by three general constraints. Justification in terms of a political conception is required only
: with respect to certain topics ; in a certain kind of forum ; and where such justification is possible. In addition to these constraints, Rawls introduces two qualifications – the “inclusive view” and, later, the “wide view” – identifying special conditions under which it may be appropriate and even desirable to introduce nonpublic reasons into political discussion, even when the general constraints are met.
Rawls insists, first of all, that the norms of public reason apply only when “constitutional essentials and matters of basic justice are at stake,” or to what he calls the “special subject of public reason.” Many political discussions – perhaps most – may involve nonpublic reasoning simply because they bear only on ordinary matters of policy, not on anything as grave as the basic institutions of society. The precise delineation of the “special subject of public reason” is not entirely clear in Rawls, but for my immediate purposes this ambiguity is not significant. Since the arguments that concern me do not turn on the meaning of the “special subject,” I will leave this first constraint very loosely outlined.
The second constraint on the scope of public reason concerns discursive setting: citizens must use appropriate reasoning in the public forum, but not in what Rawls calls the “background culture.” The background culture includes civil society and the various institutions to which citizens belong, and here the norms of public reason do not apply. But they do apply “when [citizens] engage in political advocacy in the public forum.” In Political Liberalism, Rawls argues that the public forum extends well beyond “official forums” such as courts and legislatures, and thus that citizens qua citizens are responsible for upholding public reason. In particular, the public forum extends into the polling booth, and so public reason applies for citizens when they think about how to cast their ballots. As we will see in the next section, the line separating the public forum from the background culture shifted, and by the time Rawls came to write Public Reason Revisited, he was delineating the public forum very narrowly. The shift, and the ambiguities it created, will become important when we start to ask how Rawls can answer his religious critics.
Finally, Rawls is careful to acknowledge that the norms of public reason can only apply where they can in fact be satisfied. Recall his statement that within the public forum, matters of basic justice “should be settled, so far as possible, by principles and ideals that can be similarly [i.e. publicly] endorsed.” Rawls implies, then, that there may be times when this is not possible, and leaves open the question of how often this might be the case. He is also not explicit about what we ought to do when a problem cannot be settled by public reason; as I have indicated, there are two possibilities, the “restraint” and “pursuit” interpretations of public reason, both of which are plausible based on his language. At the very least, the phrase “so far as possible” seems to signal Rawls’s willingness to recognize that the ideal of public reason may not be practical or appropriate in all actual contexts. This too will play an important role in the Rawlsian response to religious criticisms.
So the three conditions delimiting the scope of public reason are these: we should refrain from appealing to comprehensive doctrines (a) in deliberation about constitutional essentials and matters of basic justice; (b) when these take place within the public forum; and (c) when it is possible to do so. Even when all these conditions are met, however, Rawls imagines circumstances in which nonpublic reasons may be admissible. In certain cases, he argues, when a society is not well ordered, it may be necessary for citizens to invoke their comprehensive doctrines in order to “strengthen the ideal of public reason itself” and encourage their fellow citizens to honor that ideal – encourage them, in other words, to be more civil. Rawls imagines that nonpublic reasoning can strengthen the ideal of public reason by helping citizens see the ways in which their comprehensive views can support a political conception of justice, an idea he calls the inclusive view of public reason. In the introduction to the paperback edition of Political Liberalism, Rawls abandoned this view in favor of the wide view, which I will discuss in the next section.
Changes to Public Reason
Rawls introduces the wide view of public reason in The Idea of Public Reason Revisited, arguing that reasons derived from comprehensive doctrines may be introduced by anyone at any time and under any circumstances “provided that in due course proper political reasons…are presented sufficient to support whatever the comprehensive doctrines introduced are said to support.” The eventual presentation of adequate public reasons, then, retroactively vindicates a violation of public reason by a kind of supplementation or translation. This is often referred to simply as “the proviso.”
There are, of course, significant ambiguities in Rawls’s formulation of the wide view. Insofar as it implies a duty of translation or supplementation, we might wonder on whom this duty is incumbent – on the person who initially introduced nonpublic reasons? Or on someone else? The phrase “in due course” is similarly unclear. “It is important,” Rawls says, “that it be clear and established that the proviso is to be appropriately satisfied in good faith. Yet the details about how to satisfy this proviso must be worked out in practice and cannot be governed by a clear set of rules given in advance.” I think, though, that if Rawls is to answer the objections of his religious critics, it will be necessary to clarify the proviso and place the duty of translation squarely on the shoulders of public officials, leaving citizens free to engage in nonpublic reasoning as they see fit.
Rawls made one other change in Public Reason Revisited that deserves attention here. In Political Liberalism, the public forum was taken to include citizens in their capacities as voters and activists. In Public Reason Revisited, however, Rawls draws a very different line around the public forum. There, he defines the public forum as including three parts: the reasoning of judges, the reasoning of public officials, and the reasoning of candidates for office. Everything outside of this forum is now defined as part of the background culture, presumably including voters and activists. Rawls does not make it plain whether he is intentionally backing off from his earlier insistence that public reason applies to citizens qua citizens, but this passage in Public Reason Revisited at least suggests that it is within the boundaries of the state that public reason is to apply most stringently. The ambiguity will, hopefully, be cleared up by my mixed interpretation of public reason. Before discussing that view, however, we need to see more clearly the criticisms to which it is addressed.
II. Problems for Public Reason
Rawls’s view of public reason has come under fire in recent years as placing an undue burden on some religious citizens. To limit citizens’ appeal to their comprehensive doctrines implies that religious views must be separable from political views: that justice, for example, can be discussed and defined in a way that does not touch on discussions of theology and general morality. But many citizens find that such separability is inconsistent with their religious convictions – to discuss justice without reference to God is, from this perspective, not to discuss justice at all. Moreover, as we will see a little later, this could turn out to be true of reasonable religious citizens and not simply of fanatics and theocrats.
Among those who take this line, Nicholas Wolterstorff is perhaps the most widely cited. His statement of the point has become definitive:
It belongs to the religious convictions of a good many people in our society that they ought to base their decisions concerning fundamental issues of justice on their religious convictions. They do not view it as an option whether or not to do so…Their religion is not, for them, about something other than their social and political existence; it is also about their social and political existence.
For Wolterstorff, the duty of civility stands in direct contradiction to a Christian duty to pursue wholeness, integration, and unity across the entirety of life. When he finds that his conception of justice is supported only by his Christian faith and not by public reason, he says, it is not possible for him to view this support as insufficient and so not feasible for him to exercise restraint. The duty of civility, then, may come into conflict with his religious views rather than being supported by them.
Although Wolterstorff’s basic point indicates an important concern about Rawls, his articulation of that point is incomplete insofar as it lacks an extended treatment of reasonableness. After all, just because something “belongs to the religious convictions of a good many people” does not imply that it is reasonable; and if it can be demonstrated that Wolterstorff’s understanding of theistic integrity is not reasonable, then Rawls would have no reason to be concerned with citizens who affirm it. So we need to look elsewhere to find out whether Rawls ought to be troubled by arguments like Wolterstorff’s.
Reasonableness and Theistic Integrity
It may indeed seem that there is something patently unreasonable about Wolterstorff’s theistic citizen, in that he seems to be prioritizing his own comprehensive doctrine over public reason. Is this compatible with reasonableness? Rawls derives his principles of public reason directly from the ideal of cooperation among citizens regarded as free and equal; if these citizens feel compelled to reject public reason, might it be because they either reject or fail to understand the implications of reasonableness itself? Rawls seems to argue that the political conception of justice in which public reason is rooted must have priority over comprehensive doctrines, and thus that reasonable citizens will generally not violate public reason. He states clearly that “the political values expressed by its [political liberalism’s] principles and ideals normally have sufficient weight to override all other values that may come in conflict with them,” and that therefore conflicts between the duty of civility and our comprehensive doctrines should ordinarily be settled in favor of the former.
Rawls never fully explains how this could be the case given that our affirmation of any political conception of justice is inevitably grounded in our comprehensive doctrines. It seems that, on his account, all reasonable doctrines must prioritize the duty of civility over other sacred duties (such as the duty to pursue theistic integrity); this order of priorities, in other words, must be entailed by the idea of reasonableness itself. From a Rawlsian perspective, then, all reasonable doctrines must prioritize the duty of civility simply because this is what reasonableness requires. It’s possible that this is true. But there is good reason to suspect that the idea of reasonableness does not imply anything so strong, and thus that reasonable citizens may reasonably reject the duty of civility in favor of a more limited duty – here the work of Christopher Eberle is invaluable.
In Eberle’s view, the desire to cooperate with others on fair terms requires us to show respect by seeking public justifications, but it does not require us to restrain our recourse to comprehensive doctrines when that pursuit fails, nor does it require us to withhold support from policies that we are unable to justify in a suitably public manner. That is, reasonableness entails a “principle of pursuit” but not a “doctrine of restraint.” We may at times be unable to come up with public justifications that we think our reasonable fellow-citizens can all accept, and we should be free to fall back on our own comprehensive doctrines (and to express our views publicly) when this happens.
Eberle acknowledges that some of his critics experience a certain resistance at this point – they still feel intuitively that there is something inherently disrespectful or unreasonable about appealing to one’s religious views in political deliberation. But, Eberle says, this could be because their paradigmatic examples are citizens who are disrespectful according to other criteria. “If an evangelical,” he says, “particularly one who fulminates about abominations to God, constitutes our paradigm of the refusal to exercise restraint, it’s no wonder that we have an allergic reaction to the refusal to exercise restraint.” When one imagines citizens introducing religious views in the public political forum, one might immediately think of Jerry Falwell or Fred Phelps, Ali Khamenei or Danny Danon, or of similarly strident religious hard-liners. Such characters, whom I will simply call “fanatics,” conjure up the specter of theocracy in our political imagination, and the duty of civility is one way of explaining what it is that we find so repugnant in their behavior. But Eberle says it is not the right explanation, and he offers three alternatives, to which I think I can add a fourth.
First, Eberle suggests that what really bothers us about the fanatics is not their use of nonpublic reasoning but the substantive injustice of their views. It is easy to find xenophobia repugnant, and doubly so when it comes sanctimoniously wrapped in scripture. But in the first instance what is vile about the fanatics is their xenophobia itself – and their failure to appreciate the profoundly anti-xenophobic ideals within their own traditions. What we rightly find objectionable is not their insistence on the inseparability of their religious and political views, but their religious and political views themselves.
Second, Eberle says, we may be repelled by the fanatics personally – their manner of speaking, the poverty of their arguments, or their unwillingness to hear other points of view. Their views, already substantively repellent, are often presented in a style that makes them even more so. In Rawlsian terms, their unwillingness to reason with us in a spirit of respect and civic friendship indicates simple unreasonableness, and of course we have a right to criticize them on that basis. Yet their unreasonableness has nothing to do with their refusal to restrain the use of comprehensive doctrines in political justification.
Finally, and most importantly for Eberle, most of those who fail to discharge the duty of restraint also fail to discharge the duty of pursuit. The appeal to their comprehensive doctrines does not come at the end of a long process of thought and consideration – it is, in many cases, their first and only attempt at a justification. So, Eberle says, we can fault them for not even trying to satisfy the duty of civility. But citizens who do try, in good faith, to satisfy the duty of civility cannot be faulted for refusing to exercise restraint when nonpublic justifications are all they have.
There is a fourth concern which Eberle does not take up, but which is obviously important for my purposes, namely that fanatics may find nonpublic reasoning to be perfectly acceptable within coercive institutions such as courts and legislatures. They may argue, as some have, that only theists should be allowed in government, or that judges must be schooled in a particular religious tradition before they can be deemed competent to interpret the law. Beyond not subscribing to the duty of civility, in short, they might not subscribe to the liberal principle of legitimacy itself – that is, they may be willing to coerce nonbelievers without any concern for public justification of that coercion. Thus, one of the confounding variables that may lead to our discomfort with nonpublic reasoning is the fear that such reasoning will make its way through the institutions of power, and that illegitimate and unjustifiable coercions will come about as a consequence.
Eberle does not address this issue; he, like many of Rawls’s religious critics, says that he is only discussing what citizens should do, leaving aside what we ought to expect from legislators, judges, and the like. But I think we need to include both issues, since what we expect from public officials affects what we can accept from our fellow citizens. Before discussing this issue further, however, we need to see in greater detail what a reasonable conception of theistic integrity might look like, and how such a conception might come into conflict with the duty of civility.
The Reasonable Theist: Elijah
Nonpublic reasoning in the public square need not be guilty of any of these errors, and when it is not, Eberle sees no reason to view it as unreasonable or disrespectful. As long as a citizen is not guilty of other offenses such as the ones just mentioned, we should not condemn her for being unable to fulfill her obligations of restraint. To show this, Eberle describes a hypothetical religious citizen, “Elijah,” who falls short of the duty of civility that Rawls describes, yet does not appear to be in any way unreasonable. Eberle’s particular articulation of this thought experiment is stated entirely as a hypothetical, which may provoke suspicion that it models a class of citizens who do not actually exist. As a result, it may be helpful to give the Elijah experiment some more specific, and less hypothetical, content.
My own research at the 57th Street Friends Meeting inclines me to think of Elijah as a Liberal Quaker, and one whose politics are rooted in a conception of the divine nature of human dignity. Thus, he is eminently reasonable: for Elijah, respecting his fellow citizens as free and equal and seeking to cooperate with them on fair terms is not only an expression of political virtue, but indeed a religious obligation. The basic commitments of reasonableness are thoroughly supported by his comprehensive doctrine. Yet Elijah’s theological conception of human personhood heavily influences his political views, which include fairly radical notions of egalitarianism and pacifism, and a firm abolitionist stance on capital punishment. Elijah realizes that his views may be justified in part by public reason, but that they can only be justified in full by reference to some nonpublic principles. He cannot quite justify his views in terms of principles that he expects all of his reasonable fellow citizens can accept. Nonetheless, Elijah finds his Quakerism to be morally binding as a guide to political action – he finds that he must protect the sacredness of the human person by supporting any policies that reduce violence and inequality, even if he is unable to justify those policies publicly. In this case, Eberle reminds us, Elijah must view his situation as “tragic,” and must sincerely wish that he could find public justifications for the policies his Quakerism inclines him to support; if he is unable to do so, then he might feel compelled, in spite of his commitment to reasonableness, to violate the duty of restraint. It is not clear that he can be faulted as unreasonable under these circumstances. (I will shortly explore some possibilities of this sort in greater detail.)
So we have reason to believe that Elijah is reasonable, and his use of nonpublic reasons in the political forum does not prima facie compromise his reasonableness; but why would this reasonable citizen be concerned that the duty of civility will come into conflict with his religious convictions? I will now try to lay out some of the concerns Elijah might have regarding the duty of civility, then show how Rawls’s view of public reason can answer all of them, so long as it is understood in accordance with the institutional view.
1. A Sterile Public Forum
Elijah’s first worry might be something like this: by fulfilling the duty of civility, he will fail to engage in a contestation of politically potent religious symbols, and so will cede vital discursive territory to fanatics and fundamentalists. That is, Elijah may feel that his use of nonpublic reasoning is valuable in that it challenges unreasonable articulations of religious identity and constitutes a strong, visible, public example of reasonable religiosity. In the absence of such examples, the exclusionary voices of religious homogeneity may come to dominate public (as well as internal) understandings of particular religious traditions.
Martha Nussbaum’s account of the Hindu right in India is a case in point. Here, Hindu extremists have formulated a politically potent narrative of religious, national, and ethnic identity that espouses the exclusion of Muslims and other non-Hindu elements – and this narrative has established itself as one of the more conspicuous conceptions of Hindu identity. As a result, “the liberal face of religion is underdeveloped and undervalued,” and the playful, cooperative pluralism innate in the Hindu tradition – as good an example of reasonable religiosity as any – is being neglected. At the same time, Indian liberals in the tradition of Jawaharlal Nehru have often treated religion with suspicion and even hostility. Claude S. Fischer’s essay on “the alienation of religion from the left” describes a related phenomenon in the United States. In short, reasonable religious people may be losing a discursive fight over the meaning over their own traditions, and reasonableness itself is suffering as a consequence. In the clash between reasonable pluralisms and exclusionary fanaticisms, Elijah’s nonpublic reasoning is a powerful resource, and he may be concerned that the duty of civility will prevent him from bringing that resource to bear in the public forum. Elijah may worry that the Rawlsian conception of public reason leads to a form of public life that is suspicious of religious voices, which in turn fuels binary oppositions between secular liberals and strident religious nationalists. Rawls himself, of course, was not at all hostile to religion, and seems to have been sensitive to precisely this worry; in §4 I will show how his conception of public reason accommodates it.
2. Unjust Argumentation
Elijah’s second worry might be that public reason will require him to justify his political views in ways that not only misrepresent those views, but may even constitute acts of injustice. Imagine that Elijah is campaigning against the death penalty. He realizes, of course, that there are many arguments available in public reason that would adequately justify his abolitionist stance on capital punishment – he could argue, for example, that the death penalty is easily abused, ineffective as a deterrent, expensive, irreversible, and cruel. These arguments would be rooted in a political conception of justice and so employed as part of public reason. From Elijah’s perspective, however, there may be a problem with arguing in this way. If he were to employ public arguments of this sort, Elijah would be implying that the question of capital punishment ought to be settled by practical considerations. To say, for example, that the death penalty is unjust because it is easily abused is to say that it would be acceptable were it not so easily abused. Yet this could be inconsistent with Elijah’s Quakerism – the act of stating the argument could be inconsistent with adequate respect for the sacredness of the human person. Elijah might well feel that the only argument he can employ without violating his religious convictions is one that acknowledges those convictions, and that grounds his radical opposition to the death penalty in a suitably transcendent conception of personhood. Thus, only by claiming that the death penalty is unjust because it violates the sacredness of the human person can Elijah avoid violating that sacredness in the utterance of his argument. Does this mean Elijah cannot participate in public reason without contravening his Quaker faith?
3. Institutional Reason
I have hinted that my response to the second worry will be to lift some of the burdens of public reason from citizens and place them on the shoulders of public officials – I will describe this proposal in greater detail in §3. But this naturally provokes a third worry: if Elijah finds himself unable to employ public reason without violating his faith, why should he expect such reasoning from public officials? Moreover, if he does expect public officials to engage in public reasoning, does this bar him from public office?
Finally, imagine that Elijah steps into the polling booth to vote for a proposition that he favors as a result of his radical pacifism, which of course is rooted in his Quakerism. Imagine further that Elijah has no justifications for his support of the proposition other than his Quakerism. It would seem that his reasonableness inclines him to withhold support from this coercive policy in view of his justificatory deficiency
; yet he still conceives of the human person as an innately divine entity, and so to withhold support from the policy is to accept an evil that borders on deicide. Does his Quakerism not have priority over public reason in this case?
III. The Institutional View of Public Reason
I suspect, as I have indicated, that Rawls can answer all of these worries, so long as public reason is understood in accordance with what I call the “institutional view.” This view, drawing on Eberle and Charles Larmore, is a mixed conception of public reason in which different duties attach to actors in different social roles – discursive roles are governed by a duty of pursuit, whereas coercive roles are governed by a more stringent duty of restraint. In both cases, the governing norms are grounded in the demands of reasonableness, and therefore should be acceptable to all reasonable citizens in light of their own (reasonable) comprehensive doctrines. To establish this claim – and moreover to establish that it is fundamentally consistent with Rawls’s own argument – it will be necessary to rethink some of Rawls’s claims in the body text of Political Liberalism. On the whole, however, my proposal does not contradict Rawls’s claims in Public Reason Revisited (or the introduction to the paperback edition of Political Liberalism), which Rawls understood as superseding his earlier work. Thus, my view is consistent with those Rawls was defending by the end of his career. With my proposal described, and its compatibility with Rawls’s argument clarified, I will sketch the answers it gives to the four worries we saw in the previous section.
Discursive Engagement: Eberle
As we have seen, Eberle’s answer to the worries outlined above is to replace the duty of civility with a simpler duty of pursuit. In accordance with this duty, citizens are expected to show respect for one another by seeking public justifications, but are not required to restrain their recourse to comprehensive doctrines when that pursuit fails; nor are they required to withhold support from policies that they are unable to justify in suitably public ways. Eberle calls this a “duty of conscientious engagement.” We may at times be unable to come up with public justifications that we think our reasonable fellow citizens should accept, and when this happens, all citizens should feel free to make up their minds in light of their own comprehensive doctrines, and to share their views through nonpublic reasoning. For Eberle, this is to be distinguished from the Rawlsian duty of civility, which he identifies with a doctrine of restraint.
Although Eberle’s interpretation of the duty of civility (i.e. as including a doctrine of restraint) is different from my own, it is certainly a plausible reading of Rawls. Eberle cites two passages as evidence that there is a doctrine of restraint implicit in Rawls’s duty of civility, both of which say effectively the same thing: “to abandon public reason whenever disagreement occurs in balancing values is in effect to abandon public reason altogether.” Eberle seems here to be interpreting the word “abandon” as synonymous with “violate.” We should not violate public reason when we are unable to find suitable justifications, which implies that we should exercise restraint.
But I think it is possible to interpret the word “abandon” in another way. We might say instead that to abandon public reason is to abandon it as an ideal. Eberle, of course, doesn’t object to the ideal of public reason on this level; he only insists that it must be an overarching regulative ideal, not a strict duty like the duty of restraint. Rawls, always careful to build in phrases like “where possible” and “political liberalism tries,” seems willing to accept this understanding of public reason. Furthermore, Rawls’s project is itself one of ideal theory; his insistence, then, can only be that citizens should aspire to fulfill their duty of civility, not that they must do so under all actual circumstances. Public reason, then, should be interpreted as a goal toward which we strive, thus entailing nothing beyond the duty of pursuit. Eberle’s “duty of conscientious engagement” can be understood as a reading of the duty of civility rather than an alternative to it, and it is this reading that I intend to build on in what follows.
Coercive Decisions: Charles Larmore
Of course, something more than mere pursuit is needed on the part of certain social actors in order to secure the liberal principle of legitimacy. If judges, for example, fail to exercise restraint, then their opinions may lack legitimacy and citizens may be coerced without proper justification. Thus, we need to make a clearer distinction between what is required of citizens qua citizens and what is required of those who make coercive decisions.
Charles Larmore’s article on “Political Liberalism and Public Reason” makes this distinction clearly, and criticizes Rawls for overlooking it. For Larmore, Rawls’s error is that he fails to ground his conception of public reason explicitly in the principle of equal respect for persons. As a result, Rawls loses sight of the important distinction between the liberal principle of legitimacy (which is entailed by the principle of equal respect) and the duty of civility (which may not be). Larmore’s argument runs something like this: equal respect, clearly enough, entails the liberal principle of legitimacy; the liberal principle of legitimacy, in turn, implies that coercive decisionmaking must be guided by public reason; but coercive decisionmaking is distinct from open discussion, and Rawls’s justification for his conception of public reason says nothing about this latter category. Thus, public reason must apply when we shape and justify coercively binding norms, but need not apply when we are simply discussing political matters with our fellow citizens. The duty of civility is only incumbent upon citizens in their capacities as decisionmakers (whatever these may be), not as participants in public justificatory discourse.
I agree with Larmore’s proposal, although I am puzzled by his reading of Rawls. In essence, Larmore’s reading of Rawls differs from mine in that he sees a firm distinction between reasonableness and the principle of equal respect where I find this distinction unclear. For Larmore, there is an important difference between reasonable citizens and citizens who are “not only reasonable but also committed to the norm itself of equal respect for persons.” Thus, in his view, it must be possible for a citizen to be reasonable without being committed to the norm of equal respect. But on my reading, the reasonable citizen must be so committed. As we have seen (§1), reasonableness for Rawls entails regarding one’s fellows as free and equal, and recognizing the independent validity of their claims. If it is possible for a citizen to be reasonable but not committed to the norm of equal respect, then we would have to imagine a citizen who could say: “all of my fellow citizens are free and equal, their moral claims have independent validity, and I wish to cooperate with them on fair terms. But I what I owe them is something less than equal respect.” I confess I am unsure what such a citizen would mean by that claim. As a result, I’m inclined to believe that the principle of equal respect is contained within the Rawlsian idea of reasonableness. If this is true, then Larmore’s proposal can be understood as a reading of Rawls rather than a critique of Rawls.
Moreover, as we also saw in §1, the Rawls of Public Reason Revisited adopts a definition of the “public forum” that seems to be in alignment with Larmore’s proposal. If the public forum includes only the reasoning of judges, public officials, and candidates for office, then it is only decision-makers who are to be constrained by the norms governing that forum, i.e. by public reason. So by the end of his career, Rawls was describing public reason in terms that sound similar to the argument Larmore puts forward. So it seems to me that Larmore’s article, like Eberle’s book, can be accommodated by the Rawlsian conception of public reason.
If I’m right that reasonableness entails the principle of equal respect, then Larmore’s argument is also consistent with my earlier claim that moral demands on religious citizens ought to be grounded in the idea of reasonableness. Thus, by asking what exactly reasonableness entails, we can develop a conception of public reason that serves Rawls’s objectives and that places no undue constraints on Elijah and those like him. Any constraint that is derived from reasonableness, after all, ought to be acceptable to reasonable citizens of whatever religious persuasion. In what follows, I will try to develop an interpretation of Rawls that foregrounds reasonableness and its corollary, the liberal principle of legitimacy.
The “Institutional View”’
I want to argue for a view of public reason that combines Eberle’s “duty of conscientious engagement” with Larmore’s distinction between coercive decisionmaking and open discussion. Briefly stated, the view is this: when citizens are engaged in open discussion, they must fulfill the duty of conscientious engagement, but need not exercise restraint; within the institutional boundaries of the state, however, they should exercise restraint. If we adopt a view of this kind, we will welcome nonpublic reasons from our fellow citizens, but be reluctant to accept them from public officials, paradigmatically from judges writing their opinions. We will expect public officials to justify their decisions in public terms, and to exercise restraint when public justifications are not available.
The immediate objection to this view is, of course, that it is not only judges and other public officials who are involved in coercive decisionmaking. Citizens, in their capacity as voters, act within the institutional boundaries of the state in casting their ballots, and so the norms of public reason must apply here as well. I do not deny this, but I think the case of voting deserves a more extended treatment, which I will try to offer in my discussion of Elijah’s fourth objection.
Since I have framed my proposal as an interpretation of Rawls rather than an alternative to Rawls, it’s important to point out that the institutional view is fully consistent with the wide view of public reason. Indeed, the institutional view may be taken simply as a more specific version of the wide view – one in which “in due course” is taken to mean “before political ideas become coercive realities.” As long as someone (in this case the makers of coercive decisions) provides suitably public reasons, the wide view as Rawls describes it will be satisfied. This version of the wide view also satisfies the purposes of public reason in that it prevents illegitimate coercion; as long as decision-makers are bound by a duty of restraint, every coercive decision will be backed up by public reasons. Nothing further should be needed in order to secure the liberal principle of legitimacy. So the Rawls of Public Reason Revisited would not necessarily object to the institutional view of public reason. With this interpretation of Rawlsian public reason in view, we can try to develop answers to the four objections we saw in §2.
1. A Sterile Public Forum
I think the wide view can accommodate the first worry easily enough. As long as Elijah is not barred from offering religious reasons in open discussion, he is free to contest the meaning of religious symbols and languages, and to combat interpretations that he sees as unreasonable. So Elijah’s nonpublic reasoning is fully permissible in accordance with public reason. Moreover, Rawls’s views enables us to go still further, to see how Elijah’s nonpublic reasoning can be celebrated rather than simply permitted. Rawls allows us to say, in accordance with the inclusive view of public reason, that Elijah is “strengthening the ideal of public reason” by strengthening the dispositions of religious reasonableness that underpin that ideal. Thus, the wide view and the inclusive view taken together amply accommodate Elijah’s first worry.
2. Unjust Argumentation
So long as we adopt the “institutional” version of the wide view, we can accommodate Elijah’s second worry as well. If the responsibility of supplementation is incumbent upon citizens – if Elijah himself is responsible for presenting public reasons “in due course” – then public reason is indeed an onerous burden, since it would require him, in due course, to violate his understanding of Quakerism. But if the duty of supplementation falls on the shoulders of public officials, then Elijah need not worry about being required to produce public justifications if he feels that to offer these is to compromise his integrity as a Quaker. He may still make nonpublic arguments, entrusting others to supplement them appropriately.
3. Institutional Reason
The third worry is settled by Elijah’s commitment, in accordance with reasonableness, to the liberal principle of legitimacy. Since their comprehensive doctrines are reasonable, reasonable religious citizens accept the liberal principle of legitimacy on religious grounds and therefore seek public justifications for their decisions because this is what their religious values ask of them. In other words, so long as the move from reasonableness to the liberal principle of legitimacy is secure, and so long as the liberal principle of legitimacy requires restraint on the part of public officials, then that restraint must be entailed by reasonableness itself. So reasonable theists must be inclined, by the nature of their reasonable theism, to comply with it. This is not to say that public officials must utterly abandon their religious convictions when they take office – which seems impossible – but only that they should not ground their decisions solely in those convictions.
Of course, Elijah may find that his religious convictions prevent him from stepping into roles that require him to employ public reason (see the previous concern); but this does not mean that he necessarily rejects the principle that the requirements of public reason do in fact attach to those roles. That is, he can affirm the liberal principle of legitimacy, conclude from it that public reason applies to judges and other officials, and then decide that certain social roles are not right for him in light of his particular interpretation of Quakerism. He may still affirm that those roles, and the justificatory duties attached to them, are appropriate for others whose comprehensive doctrines are slightly different. Indeed, he must affirm something like this if he gives sufficient credibility to the liberal principle of legitimacy.
Here, as I have suggested, we come to a difficult point. It seems that when Elijah decides whether or not to cast his ballot in accordance with his radical pacifism, he is pulled in opposite directions by two forces that are both expressions of reasonableness. As a reasonable person, he acknowledges the liberal principle of legitimacy, and so wants to exercise restraint when he is in the polling booth; yet his radical stance on violence, which inclines him to support the pacifist measure nonetheless, is also entailed by his conception of what it means to treat other persons as “free and equal” – i.e. it is also a corollary of his reasonableness. Furthermore, Elijah could reasonably feel that his duty in the polling booth, like his duty in open discussion, is to follow his sincere opinion, and so to support the pacifist measure. Since reasonableness itself does not obviously recommend one of these options over the other, the conflict here is not easily resolved in the abstract, and I am not sure that the Rawlsian ideal of public reason can offer guidance in this hypothetical case.
What we can say is that the abstract example provoking the worry may be so rare in practice as to be of little consequence. In most actual cases, I suspect that Elijah’s views would be justifiable in both public and nonpublic terms – anti-war views generally are justifiable in this way, as are Elijah’s stance on capital punishment and his radical egalitarianism. Thus, in most cases Elijah will be free to vote in accordance with his Quakerism without violating the norms of public reason at all. So maybe we can assume that any actual view held by a reasonable theist will be one that public reason could, one way or another, support. So Elijah’s fourth worry is certainly plausible in the abstract, but may perhaps be laid to rest on the basis of a reasonable, if hardly verifiable, assumption.
We might also say that, at least in the ideal case, citizens would not ordinarily vote on constitutional essentials. If the “special subject of public reason” concerns the basic rights and privileges of the social order, then we might rightly place these things beyond the reach of majoritarian trampling. In that case, public reason would never be implicated in the polling booth to begin with, and Elijah’s fourth worry would be moot. This argument, however, depends on clarifying the meaning of “constitutional essentials and matters of basic justice,” a task I have avoided in the interest of brevity. Suffice it so say that this line of argument is plausible, but not one that I can elaborate and defend fully in this paper.
The Rawlsian conception of public reason, understood in accordance with the institutional version of the wide view, welcomes religious voices in the discursive arena of political justification, and at the same time holds firm to the liberal principle of legitimacy. Reasonable citizens may, if faith dictates, ground their political views in their comprehensive doctrines, and may employ religious or otherwise nonpublic reasons in public justification. In special circumstances, when they are making coercive decisions on constitutional essentials and matters of basic justice, citizens will be bound by a duty to restrain their reasoning and support only those policies that can in fact be justified by public reason. So long as this duty applies to decisionmakers, Rawls’s proviso will be satisfied in due course, and illegitimate coercions will be ideally minimized. As citizens, then, we need not be offended when we hear justifications in terms of a comprehensive doctrine; our fellow citizens, if they are reasonable, will show respect by affirming the liberal principle of legitimacy and by trying to find and employ public justifications when they can. When this is not possible, we will not fault our neighbors for their refusal to exercise restraint. We may, of course, object to nonpublic reasoning on the part of unreasonable citizens – i.e. fanatics – but we must do so in response to their unreasonableness, not their nonpublic vocabulary. A reading of Rawls that foregrounds reasonableness, in short, allows us to shift our critical attention away from nonpublic justification (which, as we have seen, may be a constructive force in religiously diverse societies) and onto unreasonable views and actions (which are, almost by definition, harmfully divisive), thus leaving reasonable religious citizens safe from undue moral censure.
When citizens engage each other in the manner I have described, they are still guided by the Rawlsian ideal – an ideal in which citizens cooperate on fair terms, respect the claims of others, and coerce each other only reluctantly and within the guidelines of legitimacy. They still affirm that the exercise of political power ought, ideally, to be exercised only in accordance with principles that all reasonable citizens can accept. Rawls religious critics, on the whole, do not take issue with the Rawlsian ideal as an ideal – they simply wonder about the practical constraints it imposes on religious citizens. We can now see that the Rawlsian ideal might be construed so as to obviate their concerns.
Because the Rawlsian conception of public reason is an ideal rather than a firm set of rules, we need not be surprised by the ambiguity surrounding the example of voting. Citizens in the polling booth, as I have tried to show, may reasonably feel some ambivalence about whether or not their nonpublic reasons are sufficient to justify their voting one way or another. They will accept the liberal principle of legitimacy, and so feel bound by a duty of restraint in their capacity as coercive decisionmakers; yet their faith may have overriding weight nonetheless. Rawls may not be able to give clear guidance in this case. Fortunately, it is plausible to suppose that the dilemma is a purely hypothetical one. For the concern about voting to work, we would have to imagine a citizen, possessed of all the skills and dispositions of reasonableness, voting in favor of a view that both bears on constitutional essentials and cannot be justified in terms of public reason, but that is nonetheless compatible with reasonableness. I have, for my part, been unable to think of a non-artificial example that fits these criteria.
In general, then, religious citizens need not worry that their faith commitments will conflict with the Rawlsian conception of democratic citizenship. Rawls’s arguments do not push us in the direction of a sterile secularity or an anti-religious public ethos, but in fact support a robust atmosphere of public deliberation that celebrates the interconnection of religion and politics – provided only that believers and nonbelievers alike agree on minimal principles of freedom, equality, and mutual respect.
 In speaking “religious citizens,” I am of course describing a certain subset of religious citizens, not religious citizens as such. Nor do I mean by this language to suggest that the objections I will examine apply only to theists and other religious adherents; I suspect that similar objections could be raised on behalf of many secular views. This question, while interesting in its own right, is outside the scope of the current discussion.
 Jeffrey Stout, Democracy and Tradition (Princeton, NJ: Princeton University Press, 2004); Nicholas Wolterstorff, “The Role of Religion in Decision and Discussion of Political Issues,” in Nicholas Wolterstorff and Robert Audi, Religion in the Public Square: the Place of Religious Convictions in Political Debate (Lanham, MD: Rowman & Littlefield, 1997); Christopher Eberle, Religious Conviction in Liberal Politics (Cambridge: Cambridge University Press, 2002)
 Stout 75
 PL 137
 PL 217
 PL 50, 52
 Rawls also argues that reasonable citizens “affirm reasonable comprehensive doctrines,” but the difference between the reasonableness of citizens and the reasonableness of their doctrines need not concern us here. Suffice it to say that reasonableness consists of a certain set of attitudes and dispositions – these are affirmed by reasonable citizens and supported by reasonable doctrines.
 PL 218
 PL 217
 PL liii
 PL 214
 PL 215
 PL 217
 PL 137, emphasis mine
 PL 248
 PL l
 John Rawls, “The Idea of Public Reason Revisited” in Political Liberalism: Expended Edition (New York: Columbia University Press, 1993), 463
 PRR 462
 PRR 443. Rawls actually says that the public forum includes the discourse of judges, etc., but I think the slight change in wording is warranted in light of Rawls’s statement that public reason is about reasoning, not discourse (PL 220).
 Wolterstorff 105, emphasis in original
 Wolterstorff explicitly avoids discussing the meaning of “reasonableness,” and rather unfairly criticizes Rawls as “inarticulate” on this issue (Wolterstorff 170). Wolterstorff treats reasonableness as a vague epistemological category, whereas for Rawls it is a fairly specific set of moral attitudes – as a result, Wolterstorff’s critique does not quite make contact with Rawls.
 PL 138
 PL 218
 Eberle 68
 Eberle 111
 Eberle 110
 Eberle 112
 Eberle 7; see also Wolterstorff 150
 I assume, for the sake of simplicity, that all of these issues bear on constitutional essentials in some sense. Certainly from Elijah’s Quaker perspective they involve the basic scheme of constitutional rights, and so fall under the “special subject of public reason.”
 Eberle 113
 Eberle finds that Elijah seems intuitively to be reasonable, and I share this intuition. But some readers may still feel viscerally disrespected by Elijah’s violation of the duty of restraint. I hope that the discussion in §3 will help satisfy their doubts.
 I do not mean to suggest that all, or even most, Liberal Quakers would reason exactly as Elijah does. But many do, and the content of their views warrants serious consideration regardless of their numerical strength.
 Martha Nussbaum, The Clash Within: Democracy, Religious Violence, and India’s Future (Cambridge: Harvard University Press, 2007), xi
 Claude S. Fischer, “Can Liberals Get a Witness?” Boston Review, Jan/Feb 2013
 PL l
 Eberle 68, cf. §1
 Eberle 104
 PL 241; cf. a nearly identical statement at PL lv. Both cited in Eberle 141
 PL 137
 PL 140, cf. PL 336
 I delay, temporarily, the question of whether citizens in their capacity as voters should count as coercive actors.
 Charles Larmore, “Political Liberalism and Public Reason: A Critique of John Rawls” in Julian Nida-Rümelin and Elif Özmen, eds., Welt der Gründe (Hamburg: Felix Meiner Verlag, 2012), pp. 1249-1262
 That is, to regard one’s fellow citizens as free and equal, and to desire cooperation with them on fair terms, entails an insistence that those fellow citizens should only be coerced in accordance with a constitution whose essentials they can reasonably be expected to endorse.
 Larmore 1254-5
 I am also not sure I agree with Larmore when he says that the principle of equal respect “is rejected by those…who believe that conformity to God’s will, not respect for human reason, ought to be the decisive consideration” in political decisionmaking (Larmore 1255). Since many theists consider equal respect (and, often, the more robust ideas of compassion, fellowship, service, and love) to be the essence of God’s will, I am not convinced that conformity to God’s will is always or even usually in tension with a principle of equal respect. Likewise, on the other side, I am not convinced that “respect for human reason” is necessarily a more reliable foundation for equal respect, since such a foundation might, in some cases, compromise the equality of our respect for persons deficient in reason, i.e. children, the mentally ill, and perhaps even the uneducated.
 The Rawls of Political Liberalism, of course, would raise objections here (see PL 215-218); but those objections are, I think, obviated by the later changes in Rawls’s view.
 Of course, this is only true if we have both the wide view and the inclusive view on hand; so Rawls is perhaps being hasty when he says that the wide view “secures what is needed” and drops the inclusive view altogether (PL l). The wide view may secure what is needed in the sense that it still determines when nonpublic reasoning is permissible; but the wide view is still valuable in showing that nonpublic reasoning can, in some instances, be a positive good.
 In the absence of arguments to the contrary, I will assume that reasonableness does indeed entail the liberal principle of legitimacy, which in turn requires restraint on the part of public officials. Of course, I have not defended these assumptions here, but they seem reasonably secure nonetheless.