San Diego Law Review

Fall, 1993

Symposium: The Role of Religion in Public Debate in a Liberal Society

*729 CONSTRUCTING AN IDEAL OF PUBLIC REASON [FNa]

Lawrence B. Solum [FNaa]

How should citizens in a modern pluralist democracy debate and discuss public affairs? There is wide agreement that the government should not censor public debate about politics, at least not without very good reason. But when it comes to a related question of political morality - "To what ideal should citizens aspire in political debate?" - the issue is cloudy. For example, some have argued that religious reason should be excluded from public debate; others argue for the exclusion of statements which degrade people on the basis of their religion, race or ethnicity. Still others contend that in public debate, an ideal of political morality should mirror the freedom of expression: all viewpoints should contend in a marketplace of ideas. An ideal of public reason can provide guidance on these issues. Thus, an investigation of the idea of public reason may illuminate the relationship between religion and politics.

This Article undertakes the construction of an ideal of public reason. [FN1] It begins with an investigation of the idea behind the phrase "public reason"- focusing on the work of John Rawls. The idea is further developed by considering the various possibilities for an ideal or normative standard of public reason. As each option is considered, some possible formulations are discarded and additional specifications are added. The penultimate section of the Article restates the ideal that is constructed through this process of elaboration, evaluation, and elimination. Finally, a brief survey of historical uses of the *730 idea of public reason is contained in the Appendix to this Article.

I. The Idea of Public Reason

The phrase "public reason" is ambiguous and might be used to express any number of distinct ideas. As used in this Article, "public reason" refers to the common reason of the public in its capacity as citizens constituting a polity. An ideal of public reason is a normative standard for the use of public reason. The phrase, "public reason," has a number of uses - some distantly and others closely related to the use in this Article. This section explores the use of public reason in contemporary political philosophy. I will focus on Rawls' idea of public reason, along with related uses by others. [FN2] Rawls' notion of public reason has already received considerable attention from legal scholars [FN3] and philosophers [FN4] and will serve as a focus for discussion. The idea of public reason was introduced in several of his essays in the 1980s, [FN5] was extensively developed in his Melden Lectures entitled "The Idea of Free Public Reason" delivered in 1990, [FN6] and published in revised form in Political Liberalism in 1993. [FN7]

*731 In an early formulation, Rawls explained what he has called the "idea of free public reason":

[G]reat values fall under the idea of free public reason, and are expressed in the guidelines for public inquiry and in the steps taken to secure that such inquiry is free and public, as well as informed and reasonable. These values include not only the appropriate use of the fundamental concepts of judgment, inference, and evidence, but also the virtues of reasonableness and fair-mindedness as shown in the adherence to the criteria and procedures of common sense knowledge, and to the methods and conclusion of science when not controversial, as well as respect for the precepts governing reasonable political discussion. [FN8]

Although this discussion contains the core of the Rawls' position, a few additional points deserve separate discussion.

First, Rawls understands public reason as the reason of a political society. A society's reason is its "way of formulating its plans, of putting its ends in an order of priority and of making its decisions accordingly." [FN9] Public reason contrasts with the "nonpublic reasons of churches and of many other associations in civil society." [FN10] Both public and nonpublic reason share features that are essential to reason itself, such as simple rules of inference and evidence. [FN11] Public reasons, however, are limited to premises and modes of reasoning that can appeal to the public at large. Rawls argues that these include "presently accepted general beliefs and forms of reasoning found in common sense, and the methods of science when these are not controversial." [FN12] By contrast, the nonpublic reason of a church might include premises about the authority of sacred texts and modes of reasoning that appeal to the interpretive authority of particular persons.

Second, the limits imposed by Rawls' ideal of public reason do not apply to all actions by the state or even to all coercive uses of state power. Rather, his ideal is limited to what he calls "the constitutional essentials" and "questions of basic justice." [FN13] Thus, the scope of the freedom of speech and qualifications for the franchise would be subject to the Rawlsian ideal, but the details of tax legislation and the regulation of pollution control would not. [FN14]

*732 Third, Rawls' ideal of public reason applies to citizens and public officials when they engage in political advocacy in a public forum; it also governs the decisions that officials make and the votes that citizens cast in elections. The ideal does not apply to personal reflection and deliberation about political questions; by implication it could not apply to such reflection or deliberation about questions that are not political in nature. [FN15]

With these features in mind, we can offer a summary of the Rawlsian ideal of public reason; this ideal has three main features: (1) The ideal of public reason limits the use of reason to (a) the general features of all reason, such as rules of inference and evidence, and (b) generally shared beliefs, common- sense reasoning, and the noncontroversial methods of science. (2) The ideal applies to deliberation and discussion concerning the basic structure and the constitutional essentials. (3) The ideal applies (a) to both citizens and public officials when they engage in public political debate, (b) to citizens when they vote, and (c) to public officials when they engage in official action - so long as the debate, vote or action concerns the subjects specified in (2). With Rawls' view in mind, we proceed to two preliminary subjects: first, the role of the idea of public reason in the regulation of public discourse and, second, the ways in which a particular ideal of public reason might be justified.

II. The Role and Justification of Public Reason

In this part of the Article, I will examine two issues that set the foundation for specifying and justifying a liberal ideal of public reason. The first issue concerns the role of an ideal of public reason; the second issue concerns the standards by which the case for such an ideal should be measured.

A. The Role of an Ideal of Public Reason

Public reason is the common reason of the public at large - in our case, the citizenry of a democratic society. Of course, an understanding of the role of an ideal of public reason requires a prior understanding of the role of public reason itself. Thus, the discussion that follows assumes a notion of that purpose similar to that advanced by Rawls: Public reason is the reason of the public as the citizens of a democratic polity. Public reason is used in political debate in the public sphere and is used by public officials to justify the Constitution, laws, executive actions, and judicial decisions.

An ideal of public reason, or standard of civility, is intended to *733 serve a regulative role. More particularly, an ideal of public reason regulates public reason-giving practices in two ways: (1) as a standard for self-evaluation and, (2) as a standard for political criticism. Each of these aspects of the role deserves comment.

An ideal of public reason serves as a standard for self-evaluation in the sense that it can be used by citizens to guide their own use of reason in the public sphere. One can ask, "What kinds of arguments should I give or refrain from giving in public political debate?" An ideal of public reason answers this question by articulating a set of reasonable standards of civility to one's fellow citizens. The first role for an ideal of public reason assumes voluntary compliance. In this role, the ideal is affirmed voluntarily because it is seen as reasonable by members of a political community.

An ideal of public reason can serve another role - as a standard for the political criticism of argument in the public sphere. One can ask, "When is it proper for me to criticize the argumentation of a fellow citizen on the ground that the reasons offered transgress the limits of civility?" An ideal of public reason answers this question by defining standards for political criticism of reasons given in public. The second role of public reason does not assume coercive enforcement by the state, as such enforcement would violate the political right of freedom of expression. But the second role of public reason does not rule out the use of social pressure to encourage compliance with the ideal. Political criticism can change behavior in two ways: (1) by offering reasons that are accepted and, (2) by communicating disapproval that motivates because of citizens' desires for the approval of their fellows.

B. Justifying an Ideal of Public Reason

In the case of a noncompulsory ideal of public reason, we must formulate the ideal and provide justifications for it that are addressed to those for whom the ideal is intended to have normative force. The ideal must be one which they could reasonably accept, and the justifications must be reasonably persuasive. Why should these limits be observed? Consider two answers.

One answer is instrumental: a noncompulsory ideal of public reason must generate its own support. If the ideal and its justifications cannot reasonably be accepted by significant groups in our polity, then the ideal will not serve the purpose of facilitating and regulating deliberation in the public sphere.

The second answer is based on the idea of respect for one's fellow *734 citizens as free and equal. The force of this reason can be illustrated by examining David Smolin's review [FN16] of Michael Perry's recent book, Love and Power . [FN17] In that book (his position has since changed), [FN18] Perry argued for an ideal of "ecumenical political dialogue" [FN19] that includes basic standards of civility and commitment to two attitudes called "fallibilism and pluralism." [FN20] Fallibilism requires a commitment to the idea of self-critical rationality, and pluralism involves the affirmation that moral pluralism can result in richer moral insight than moral monism. [FN21] Smolin's review objects to these requirements from the perspective of an evangelical [FN22] Christian perspective.

In particular, Smolin argues, "Perry's theory excludes from dialogue those groups unwilling to accept the dialogic virtues of fallibilism and pluralism and the consequent distinction between religious faith and religious belief." [FN23] Smolin argues that Perry's ideal would exclude members of various evangelical, fundamentalist, and pentecostal Protestants and traditional Catholics, Anglicans and Lutherans. [FN24] He then argues, " T hose excluded by Perry's criteria, such as myself, are going to protest that their exclusion is unfair." [FN25] Smolin does not offer a reason for his charge of unfairness at this point, but moves to another line of critique. "More importantly, those excluded, who comprise a culturally significant and politically active portion of the population, are less and less willing to accept the kind of exclusion Perry perpetuates." [FN26] Why unwilling?

Perry . . . makes no serious attempt to persuade those he excludes of the propriety of their exclusion. Why should those who view pluralism and fallibilism as vices accept them as norms of civic virtue? Why should those who consider Perry's distinction between religious faith and religious belief to be heretical be persuaded by a vision of dialogue premised on acceptance of this distinction as the mark of "authentic religion?" [FN27]

Smolin's point is that Perry has not offered reasons for his ideal of *735 ecumenical discourse that could be taken as reasonable by someone who shares Smolin's convictions. Perry offered reasons from within his comprehensive religious view. But from within Smolin's own and quite different religious conception, Perry's starting points are not axiomatic. Indeed, some of Perry's premises are heretical from Smolin's evangelical perspective. In an Article written after Smolin's review, Perry expresses concern about Smolin's objection and retreats somewhat, stating that "the essential criterion [for ecumenical dialogue] is less fallibilism than public accessibility." [FN28]

The exchange between Perry and Smolin has an important lesson for the question at hand: What sort of justification should be given for an ideal of public reason? An ideal of public reason must be justified by arguments that can be accepted as reasonable by members of the public to which the ideal is addressed. By formulating and arguing for his ideal of ecumenical discourse from a liberal Catholic perspective, Perry advanced reasons that could not be accepted as reasonable by Smolin, who rejected many of Perry's religious premises. Smolin asks for arguments that he can view as fair and reasonable from his evangelical perspective.

The point of the exchange between Smolin and Perry can be expressed in the following more abstract (yet illuminating) formulation: An ideal of public reason is reflexive in the sense that it applies to its own public justification. An ideal of public reason must be justified by public reason. Moreover,this requirement of justificatory reflexivity has implications for the content of the ideal. An ideal should include the idea that public reason gives citizens reasons that are public , in the sense that they could reasonably be considered as motivating by those to whom they are addressed. This point is reinforced by the role of public reason. For example, one role of public reason is to give citizens reasons to obey the law of their own free will on the basis that the law is reasonably justified. If the sole justification for a law is a nonpublic reason (such as a deep premise of a comprehensive moral theory or a sectarian religious view), then there will be many citizens who cannot obey the law because on the ground that they see the law as reasonably justified - although they still may obey because of fear of punishment.

The reflexivity of public reason has a practical corollary: An ideal of public reason will emerge in a public political culture as a result of public debate and discussion. As a practical matter, the ideal will *736 change over time and debate over the contours of the ideal will itself be shaped by the prevailing ideal at any given point in time.

The exchange between Perry and Smolin has a second implication for the content of a liberal ideal of public reason. Such an ideal should include the idea that public reason gives citizens reasons that are reasonable in the sense that they are limited by the principle of fairness. Smolin's objection that Perry's ideal of ecumenical dialogue is unfair can be reconstructed as an argument based on the notion of respect for other citizens as free and equal. Such respect does not mean only giving other citizens reasons which they already accept as true. Treating others as free and equal does not mean catering to existing beliefs and desires; indeed, one could argue that catering to existing beliefs and desires is disrespectful because it treats others as unreasonable. The principle of respect for citizens as free and equal does mean that we should give our fellow citizens the sort of reasons that they could reasonably accept.

The modal operator "could" is crucial here. The requirement is not to give reasons that all or most of one's fellow citizens will accept; rather the requirement is to give reasons they reasonably could accept. For example, an ideal of public reason that excluded Smolin from public discourse is not one that he reasonably could accept - as Smolin himself argued. A more inclusive ideal may not be one that he does accept - he might think that atheist or blasphemous discourse ought to be excluded from public discourse, but the inclusive ideal is one that he reasonably could accept. This ideal does not treat him as a second-class citizen.

With a core notion of the role and justification of an ideal of public reason in place, I now turn to the content of the ideal. The following section outlines basic distinctions about the possible structures of an ideal of public reason.

III. The Structure of Ideals of Public Reason

This section explores the possible structures of an ideal of public reason by discussing distinctions about the content of such an ideal. Each distinction marks a dimension in conceptual space: Any ideal of public reason will occupy a particular position with respect to each of the dimensions that is identified. If this point seems too abstract, consider a less formal version. We can imagine many different ideals of public reason. Can we organize and categorize their features to systematize the ways in which they differ?

*737 A. Distinction One: Four Contexts to Which an Ideal of Public Reason

Might Apply

The first distinction relates to the domain of deliberation and discussion to which an ideal of public reason might be applied. At one extreme, an ideal might apply to all reasoning, whether deliberation or discussion, whether public or private, about any topic whatsoever. [FN29] At the opposite extreme, we might imagine an ideal that was limited to the most formal and public uses of reason. For example, only the written opinions of the courts of law and the texts of statutes and executive orders might be subject to the requirement. For practical purposes, we can distinguish four contexts in which an ideal of public reason might be applied.

Call the first context "private deliberation and discussion." The first context is important because it marks the domain to which an ideal of public reason does not apply. Thus, private deliberation about private life is not regulated by the ideal. By private deliberation, I mean the use of reason by an individual without discussion with others. By deliberation about private life, I mean deliberation about one's life plan, one's intimate associations, one's family and so forth. In addition, private discussion about private life is not the subject of an ideal of public reason.

In a sense, all discussion could be said to be public (or exterior) in that discussion involves more than one participant. This is not the sense in which "public" is used in the phrase "public reason." Discussions that are not addressed to the public at large include those between family members or within voluntary associations. These discussions are private for the purposes of an ideal of public reason.

Call the second context "public discussion about ethics and culture." One can imagine the formulation of an ideal of public reason that did apply to this context. It could be argued that any use of reason that is addressed to the public at large should be subject to the ideal of public reason. Moreover, for some groups and individuals there will be instrumental reasons for adhering to the ideal of public reason in public debate about ethics and culture. Persuading as many of one's fellow citizens as possible may require adhering to an ideal of public reason. But this instrumental reason is not one of political morality. It does not violate the requirement of civility to *738 offer one's deepest convictions about ethics or culture in public debate, even though reasons expressing such convictions may not be viewed as reasonable by many groups or individuals.

Call the third context "public discussion about the coercive use of state power." The case for formulating an ideal of public reason to cover this context is stronger still. Reason used in the third context is public in two senses. First, reason-giving in public discussion about the coercive use of state power is public in the sense that reasons are given to the public at large. Of course, in any given public discussion, the whole public is not literally addressed in the sense that it will read or hear the communication. Newspaper editorials, speeches to which the public is invited, and so forth are addressed to the public in the sense that they are intended for any member of the public who chooses to read or listen. Second, reason-giving in public discussion about the coercive use of state power is public in the sense that the subject is public. State power is the power of the public; in a democratic society, the state acts on behalf of the body of public citizens. Coercive state power also is directed at the public; when the requirements of the rule of law are observed, laws and regulations are addressed to the public at large.

As I have defined it, the third context is public discussion about all coercive uses of state power, but reasonable arguments can be made for a narrower formulation. For example, Rawls limits his case for an ideal of public reason to public debate about the constitutional essentials and basic liberties. [FN30] Of course, there is a sense in which the constitutional essentials and basic liberties lie at the core of public reason; the constitutional order is the foundation upon which the public political order is built.

But there are good reasons to extend the ideal of public reason to all coercive uses of state power. First, most citizens encounter the state most directly and concretely through the coercive exercise of power. In a sense, the basic structure and constitutional liberties lie behind the scene. If public reason is to reconcile citizens with the political order, it should do so in those instances in which citizens are likely to ask for justification. For many and perhaps most citizens, those instances are likely to be ordinary applications of coercive power and not the extraordinary cases in which the basic structure or constitutional liberties are called into question.

Second, citizens who ask for justification when they encounter the coercive power of the state are making a reasonable demand. We may reply that we can justify the basic structure and constitutional liberties to them. Suppose they counter that they believe that their dignity as free and equal citizens is offended if they are coerced *739 without a justification that they see as reasonable. How are we, their fellow citizens, to argue that their request is unreasonable? Would we not ask for reasons that we could accept as reasonable, if we were in their place? In sum, the third context is public discussion about the coercive use of state power. Although a narrower formulation is reasonable, the better view is the broader one.

Call the fourth context "deliberation and discussion by officials acting in their official capacity." It would seem that an ideal of public reason must apply to this context if it is to count as an ideal of public reason at all. The use of reason by public officials in their official capacity is public in both senses in which public discussion about the coercive use of state power is public: Public officials address the public at large, and their actions in their official capacities are on behalf of and directed at the public. In addition, when citizens use reason to discuss public matters, they may do so in their capacity as private citizens. However, when officials use reason in their official capacity, they are public persons (they personify the public) in the sense that they occupy the role of public official. Because this context is public in all three senses, we might say that an ideal of public reason is essentially concerned with deliberation and discussion by officials in their official capacity.

B. Distinction Two: Causal Influence Distinguished from Use in Reasoning

The second distinction is between causal influence and use in reasoning. This distinction goes to the kind of constraint that an ideal of public reason places on individual action. It might be argued that citizens and officials should not allow nonpublic reasons, such as their comprehensive moral or religious views, to have any causal influence on their public actions. For example, a legislator may not cast a vote for or against a bill if she suspects that her vote would be caused by her religious views. This way of formulating an ideal of public reason seems implausible. Of course, one might ask citizens to engage in a thought experiment, asking "Would I take this action even if I did not hold my comprehensive view?" But if one believes that one's nonpublic beliefs have a pervasive influence on one's action, compliance with a causal formulation of the ideal may simply be impossible. One can imagine a Catholic saying, "Any action I take is influenced by my religion, because my identity is constituted in part by my Catholicism. I have no idea how I would act if this were not so."

*740 A more plausible formulation of the ideal would focus on the use of nonpublic reasons in deliberation and discussion. A requirement of public reason can readily be applied to one's giving of reasons in debate and discussion. Application to deliberation may be more difficult, but seems possible. One's conscious deliberations are open to introspection, and one can attempt to deliberate in ways that limit the role of nonpublic reasons.

C. Distinction Three: Direct versus Foundational Use of Reason

The third distinction is between direct and foundational use of reason. This distinction addresses the following difficulty. If an ideal of public reason were to require that reasons be public all the way down , then the ideal would be far too stringent. Take the following example of a seemingly public reason: Every citizen deserves respect as a human being. This reason is public in the sense that it can be derived from our public political culture, and it does not rely on any particular moral or religious view. On the other hand, it might turn out that many citizens believe this public reason holds for nonpublic reasons. For example, some citizens may believe that every citizen deserves respect as a human being because Scripture reveals that all humans were created by God in God's image. In this case, the public reason has a nonpublic foundation.

Why would an ideal of public reason that excluded beliefs held on the basis of nonpublic reasons be too stringent? One answer to this question is that such exclusion might be viewed as unprincipled and unfair. Ruling out the foundational use of nonpublic reasons would be inconsistent with the Kantian idea that the only limits of reason should be those imposed by reason itself. [FN31] For example, some citizens may believe that they have a moral obligation to bring their deep moral beliefs to bear in their private deliberations on questions of political morality. Although these citizens might be willing to accept an ideal of public reason that excluded the foundational use of nonpublic reasons as a modus vivendi, as an unprincipled compromise necessary to avoid the greater evil of public disorder, they could not affirm the exclusion of public reasons with private foundations on principled grounds. This objection - that public reason as mere modus vivendi is unprincipled - is important, because the justification for an ideal of public reason should be one that the public at large can affirm as reasonable.

*741 D. Distinction Four: Secular Reasons Distinguished from Public Reasons

The fourth distinction is between two pairs of reason categories. The first pair of categories, frequently discussed in connection with the role of religion in politics, includes religious reasons and secular reasons. The second pair of categories includes public reasons and nonpublic reasons. The point of the fourth distinction is that these pairs do not map the same difference between kinds of reasons. Reasons that are religious may be public, and reasons that are secular may be nonpublic.

For example, the injunction "Thou shalt not kill" is a public reason, even though it is a quotation from a religious text. This injunction expresses a belief that the killing of humans is a wrong that can be derived from our public political culture, of which the text itself is likely a part. Consider by contrast a secular proposition of the hedonistic variant of utilitarian moral theory - that the moral evaluation of an action is determined solely by the pain or pleasure that will result. This proposition is not derived from our public political culture, but is instead a controversial part of a comprehensive moral doctrine. For this reason, this utilitarian proposition is nonpublic, even though it is formulated in secular terms.

If reasons that are in some sense religious can be public, and reasons that are secular can be nonpublic, then it follows that the religious/secular reason distinction does not map directly onto the public/nonpublic reason distinction. This is not to deny that many religious reasons are nonpublic or that many public reasons are also secular. In some contexts, subsets of the two reason categories will map onto each other (roughly, at least); in contemporary American society, for example, if one focuses on sectarian religious reasons on the one hand, while focusing on secular public policy debate, on the other hand, then the religious reasons will almost all be nonpublic and the secular reasons will almost all be public. The point of this section is to warn against overgeneralizing from such special contexts.

E. Three Kinds of Principles: Laissez Faire, Inclusionary, and Exclusionary

The fifth and final distinction is between three different principles by which an ideal of public reason can express the requirement that reason be public. Call the first expression of the requirement "the *742 principle of laissez faire." This principle interprets the idea of public reason as reason which is free of constraint. Call the second expression of the requirement "the principle of exclusion." An exclusionary requirement of public reason is a requirement that the reasons given or relied upon exclude nonpublic ones. Call the third formulation of the requirement of public reason "the principle of inclusion." An inclusionary requirement of public reason is a requirement that the reasons given or relied upon include public ones. This section begins with the principle of laissez faire, then explores a variety of inclusionary and exclusionary principles.

1. The Principle of Laissez Faire

One might argue that the best ideal of public reason would draw upon Kant's notion that public reason ought to be free, limited only by standards internal to reason itself. [FN32] This might be read as implying that all reasons should be allowed so that the truth can be revealed in public debate. If the principle of laissez faire were accepted as the ideal of public reason, then any reason that was sincerely believed to have argumentative force could be advanced in public argument. For example, public officials might advance religious reasons for the coercive use of state power if they believed that these arguments were valid. No reason would be excluded from public discussion or deliberation on the ground that it was not public - in Rawl's sense that public reasons are those based on common sense, science, or the public political culture.

Although the principle of laissez faire is a conceptual possibility, it is not a live option for our political culture. Given the fact of pluralism, we do not expect the rough and tumble of political argument to reveal the truth about such deep matters as our duties to God, the nature of good, or the meaning of life. Even if offered sincerely and with a willingness to engage in wide- ranging discourse until the dispute over their truth is resolved, reasons that directly rely on premises concerning these deep matters will be rejected by many as unreasonable justifications for political action. One can imagine that if the world were arranged differently, this would not be so. If humans were immortal and their reason as perfect as that of angels, then even the deepest questions might be seen as resolvable by conscientious public debate. But our lives must come to an end, and our reason is not perfect. For us, the principle of laissez faire cannot serve to express the ideal of public reason.*743 2. Some Exclusionary Principles

An ideal of public reason could be formulated in terms of a variety of exclusionary principles. Examination of such principles begins with the most stringent requirement - that only reasons accepted by the whole public be allowed as public reasons.

a. Exclusion of Contested Beliefs

The most stringent formulation of an exclusionary principle would be the exclusion of all beliefs and modes of reasoning that are not accepted by the whole public. This principle could be interpreted as literally requiring universal agreement on all premises and inferences. In this form, the principle would constrict public reason to the vanishing point, because there is at least one citizen who will deny the truth of almost every conceivable premise of political argument. Perhaps there is also a philosopher who will contest the validity of every mode of reasoning.

Consider then a first step in relaxing the exclusion-of-contested-beliefs principle: The principle could be relaxed so a belief is counted as contested only if it is not contested by any adult citizens in full possession of their faculties, acting on the basis of a sincere desire to reach the truth. Even this relaxed version of the exclusion-of-all-contested-belief principle would leave public reason little room to operate. In a sense, public reason is most important when there is disagreement. When everyone accepts the premises and inferences that justify a policy, then public justification is least urgent. It is when there is disagreement about government action that reasons need to be given, but the stringent principle, excluding all contested beliefs from any role in public reason, seems to rule out public reason-giving in precisely this situation.

Consider a second relaxation of the principle: A more plausible exclusionary principle might exclude all reliance on a subset of contested beliefs that concern morality. This principle would allow the inclusion of contested factual beliefs and contested rules of inference, but would exclude any moral belief from public reason if the belief was contested by any adult citizens in full possession of their faculties, acting on the basis of a sincere desire to reach the truth. Thus, public reason would include any moral belief that was accepted by all competent citizens; for example, the idea that the killing of an innocent is a prima facie wrong might count as such a moral belief.

This exclusionary principle would allow for more robust public *744 discourse than the more stringent principle, but it, too, would impose a severe limitation. It might well be the case that public reasons could be developed for many policies from uncontested moral premises, but discussion would stop whenever contested moral beliefs became part of the argument. The exclusion of all contested moral beliefs would mean that even an argument that begins with uncontested moral beliefs is disqualified if it builds to conclusions which themselves are moral beliefs derived from the uncontested premises, but nonetheless are not accepted by all competent citizens.

Moreover, the kind of moral belief that gains universal acceptance is likely to be quite weak. For example, even the Pareto principle, that a state is preferred if it makes at least one person better off and none worse off, would not meet this requirement. There are many competent citizens who believe that being made better off is not a good if it results in any inequality.

Consider a third relaxation of the principle - this time by excluding contested moral beliefs only if they are not used to directly support the conclusions of political arguments. This principle would allow the disclosure of foundational beliefs that are contested, but would disallow their use as direct support for conclusions of public reason. For example, one might argue from a theological premise to a respect for human life to the conclusion that the government should provide adequate food and housing for all citizens. The contested theological premise serves a foundational role, but direct support for the proposition of public policy is provided by an uncontested belief - that human life should be respected. Even when limited in this way, the principle of excluding contested beliefs does not seem plausible. It limits the efficacy of public discussion to cases in which disagreement can be resolved on the basis of arguments that include only uncontested beliefs in roles of direct support. Given the wide diversity of moral opinion, public debate that was conducted in conformity with this ideal would likely be shallow and truncated.

b. Exclusion of Religious Beliefs

Another exclusionary principle might prohibit only the use of religious beliefs, allowing other moral beliefs in public debate. This principle is related to that of excluding contested beliefs, because in modern democratic societies, all religious beliefs are contested. [FN33] For example, in the United States, Islam, Hinduism, Shinto, and a variety of other religions coexist with Protestantism, Catholicism, and *745 Judaism. Moreover, the beliefs of all religions are disputed by secular viewpoints.

Contrasted with the principle of excluding all contested beliefs, the principle of excluding religious beliefs seems at least feasible. Assuming that the principle is relaxed to allow the use of nonreligious reasons (e.g. killing is morally wrong) that are affirmed on the basis of religious foundations (e.g. because God has commanded it), this principle would allow for reasonably robust public discussion. Assuming there were good reasons to adopt it, the principle could be implemented.

There are, however, severe problems with this principle. First, the exclusion of religious beliefs from public discourse is unfair to those who are believers. Construed so as to be least fair to believers, this principle would exclude only theist beliefs, but allow atheist beliefs as public reasons. On the one hand, believers could not argue for public aid to religious school on the ground that religious training would be good for children. On the other hand, nonbelievers would be free to argue against such aid on the ground that religious schools promote false and dogmatic belief systems. This would appear to be an unfair double standard for believers.

Of course, this blatant unfairness might be corrected by reformulating the standard to exclude all beliefs about religion, including, for example, atheist beliefs. Taken literally, this formulation would be self- defeating because this proposed ideal of public reason is a political belief about religion, and hence would itself be excluded from public discourse. Let us assume that the ideal could be modified to allow a limited class of public debate about the role of religion in political life, without appeals to the truth or falsity of particular religious doctrines. The question of fairness still remains.

Those with religious beliefs will argue that it is unfair to allow nonbelievers to appeal to secular moral and philosophical doctrines while requiring believers to refrain from voicing the full range of their moral and religious beliefs. Why should it be permissible for utilitarians to appeal to the truth of utilitarian moral theory if it is impermissible for Catholics to appeal to the truth of Catholic doctrine? A full answer to this question is beyond the scope of this Article, although other articles in this Symposium do address that topic. The point that bears emphasis is that both secular and religious views contain deep and controversial doctrines that are not public in the sense that they can reasonably be affirmed by the public at large. Atheists cannot reasonably be asked to affirm the proposition that *746 belief in God is an essential component of the good life; believers cannot reasonably be asked to affirm the proposition that pleasure and pain are the ultimate sources of value. Given that both religious and secular doctrines contain such deep and controversial beliefs, a liberal ideal of public reason ought not to exclude religious beliefs and allow secular beliefs. Some other criteria must be formulated to give content to an ideal of public reason.

c. Exclusion of Nonpublic Reasons

The final proposal is to formulate an ideal of public reason that excludes from public debate all reasons that are nonpublic. Without definition, this principle would be empty. By "nonpublic," I mean those reasons that are not public reasons as defined by Rawls. If public reasons are (1) common-sense beliefs, (2) ideas from our public political culture, and (3) the noncontroversial conclusions of science, then nonpublic reasons are the rest. This includes deep beliefs about the nature of the good that form part of various comprehensive religious and moral doctrines.

Another category of nonpublic reason is not captured by the discussion so far. Statements that deny the freedom and equality of fellow citizens may be considered nonpublic reasons. Of course, this simple formulation would need to be spelled out in some detail. For now, the principle might be explicated in terms of its application to racist speech. For example, a racist might argue against antidiscrimination laws on the ground that a particular group was not fully human and hence was not deserving of the equal respect accorded citizens of other racial groups. This argument would be excluded on the basis of a principle excluding statements that deny the free and equal status of citizens. The principle itself might be justified on the ground that such racist remarks cannot be accepted as reasonable by those of one's fellow citizens who belong to the denigrated group. Hence, racist statements which deny the free and equal status of citizens are not addressed to the public at large; they are not public reasons.

With this important addition in place, let us stipulate that the principle of excluding nonpublic reasons will not exclude reasons that are public except in the sense that they may have a nonpublic foundation. The nonpublic foundation for such public reasons would itself be excluded from public discussion and from a direct role in deliberation, but public reasons would be admitted to public political debate irrespective of their foundations. The proposal is that an ideal of public reason might be given content by a principle excluding any direct role for nonpublic reason in public debate or discussion. Should we accept this proposal?

*747 First, it must be conceded that the principle of excluding nonpublic reasons would be chosen over the principle of laissez faire. Given the fact of pluralism, that principle does not insure that the reasons offered in public political debate can reasonably be accepted by the public at large.

Second, the principle of excluding nonpublic reasons would be chosen over the principle of excluding all contested reasons (or all contested moral reasons). If all contested reasons were excluded, public political debate would be severely restricted. Allowing all public reasons (as defined here) would facilitate a more robust and effective role for public political debate.

Third, the principle of excluding nonpublic reasons would be chosen over the principle of excluding religious reasons. Whereas that principle could not be accepted as fair by believers, the principle of excluding nonpublic reasons could be. Exclusion of all nonpublic reasons puts the deep beliefs of both theists and atheists on equal footing.

Based on these three comparisons, we have a prima facie reason to accept the principle of exclusion of nonpublic reasons as the best interpretation of the ideal of public reason. An important comparison, however, has yet to be made. The next section introduces the possibility of an inclusionary ideal of public reason.

3. Principles of Inclusion

An ideal of public reason might be specified in yet another way - by principles of inclusion. [FN34] A principle of inclusion offers criteria for what must be included in an argument in order for it to comply with an ideal of public reason. It is possible to comply with a principle of inclusion, even if nonpublic reasons are part of an argument. For example, policy A might be justified by public reason P and nonpublic reason Q. So long as P provided a sufficient justification for A, the inclusion of Q would not violate an inclusionary principle satisfied by P. [FN35]

There are several possible principles of inclusion. The most stringent principle of inclusion corresponds to the most stringent principle *748 of exclusion: An ideal of public reason might require that every public policy be supported by a sufficient reason that relies on uncontested beliefs. This possibility does not require extensive consideration. The weaknesses of the principle of excluding contested belief would apply to a principle that required a sufficient reason for including only uncontested beliefs. Such reasons would run out too soon.

Analogously, a principle that required the inclusion of sufficient secular reasons would suffer from the same defect as a principle excluding religious reasons. Believers could not accept such a principle as reasonable, and because our society includes large numbers of believers, an effective ideal of public reason must be reasonably justifiable to them.

The most promising inclusionary principle, therefore, is one that requires the inclusion of public reasons. This principle would allow for robust debate but would not be unfair to those with religious views. Unlike the principle of excluding nonpublic reasons, the principle of including public reasons would allow citizens to advance nonpublic reasons in public debate. Of course, nonpublic reasons would only be allowed if sufficient public reasons were also given. This implies that nonpublic reasons could only be given in two circumstances: (1) if the nonpublic reason were the foundation for a public reason, and (2) if the nonpublic reason were an additional sufficient justification for a policy that would be given an independent and sufficient justification by a public reason. Both the exclusive and inclusive versions will result in the giving of sufficient public reasons, but the inclusive version will allow a specified role for nonpublic reasons as well.

Would the principle of including public reasons be chosen over the principle of excluding nonpublic reasons? Like the principle of excluding nonpublic reasons, the principle of including public reasons would foster civility and the civic virtue of tolerance. It would do this in two ways. First, by requiring citizens to give a public reason, the principle of inclusion assures that no citizen will call for the coercion of another citizen without giving a reason the other views as reasonable. Second, by requiring citizens to exclude nonpublic reasons that are not the grounds for public reason, the principle of inclusion assures that no citizen will call for the coercion of another on the basis of grounds the other would view as wholly unreasonable. That is, no one will be coerced on the basis of a reason that cannot be seen as an alternative deep foundation for a reason that is public and thus can be accepted as reasonable. The principle of including public reasons is not inconsistent with the duty of civility - the duty of citizens to explain how the coercive use of state power can be supported by public reason.

*749 Of course, it must be conceded that there is at least one prima facie reason to believe that the principle of excluding nonpublic reasons might do a better job of respecting the values captured by the duty of civility than would the principle of including nonpublic reasons. The principle of including public reasons allows some nonpublic reasons to be given in public constitutional debate. Even if these nonpublic reasons are limited to a supporting role for public reasons, allowing them at all risks undermining the value of civility. After all, the fact that a religious reason is given at all may be offensive to some.

There are, however, reasons to believe that the principle of including public reasons might do a better job of fostering civility and tolerance than would the exclusionary principle. First, it is possible that the giving of nonpublic reasons (which are not shared) that are the foundations of public reasons (which are shared) will foster a sense of political solidarity and tolerance. If I see that you and I agree about fundamental public values, despite our disagreement about the moral foundations of those values, I may come to see your fundamental views as reasonable - despite my unwillingness to accept them as true. [FN36]

The second reason to prefer the principle of including public reasons is that theprinciple of exclusion has as its corollary a duty of limited intolerance. Adherence to the principle of exclusion implies two duties. The first is a duty of forbearance: I myself must forbear from giving nonpublic reasons. This follows from the first role of an ideal of public reason - as a standard for self-restraint. The second is a duty of limited intolerance: I must disapprove when others fail to forbear by violating the principle of exclusion. This follows from the second role of an ideal of public reason - its role as a standard of public criticism. Of course, my disapproval must itself be civil, expressed in terms that respect the freedom and equality of my fellow citizens who violate the principle of exclusion. But I must disapprove because, on the principle of excluding nonpublic reasons, my fellow citizens have acted wrongly in giving nonpublic reasons. By contrast, the principle of inclusion has as its corollary a duty of tolerance. Correlative to the permission granted by the principle of including public reasons, for citizens to advance their own comprehensive moral or religious doctrines as supporting grounds for public reasons that bear on the constitutional essentials, is the duty to listen with *750 respect and tolerance for these views.

Compliance by citizens with the duty of limited intolerance imposed by the principle of excluding nonpublic reasons may well undermine their development of the civic virtue of tolerance. By contrast, compliance with the duty of tolerance imposed by the principle of including public reasons would naturally seem to foster the civic virtue of tolerance. Thus, comparing the two interpretations of the ideal of public reason, the principle of including public reasons would seem to do a better job of fostering the civic virtue of tolerance and thus indirectly supporting the value of civility. Because this virtue is a very great political good, there is a strong reason (all else being equal) to prefer the principle of including public reasons.

The third reason for preferring the principle of including public reasons is that giving the nonpublic reason expresses the ideal of full respect for the autonomy of fellow citizens. The third argument begins with the premise that there is an ideal of full respect for fellow citizens as free and equal. Call this "the ideal of full respect." This ideal expresses the notion that one ought to treat one's fellow citizens as possessing an equal human reason that grounds their capacity to exercise their freedom. Treating one's fellows as reasonable in this sense requires that one give them reasons when acting in a way that affects them; one such action might be voting on the coercive use of state power. The full requirements of this ideal include: (1) that I give fellow citizens reasons which they could accept as reasonable (i.e., I give them public reasons), (2) that I do not give them reasons that they could not accept as reasonable (i.e., I give them only public reasons), and (3) that I disclose to them all of the reasons that are the basis for my position (i.e., I make a full disclosure).

But given the fact of pluralism, we cannot attain completely the ideal of full respect. The fact of pluralism creates tension between refraining from the giving of nonpublic reasons and giving all the reasons upon which I am actually relying. Given the fact of pluralism, citizens will frequently, even usually, have nonpublic reasons as foundations for the public reasons that support their views with respect to constitutional essentials. The question then becomes, given that the ideal of full respect can be only partially realized, which aspect of the ideal should give way?

It seems that it is the ideal of refraining from the giving of nonpublic reasons that must give way. Why? The principle of excluding nonpublic reasons assumes that one's fellow citizens should not even be allowed to listen and think about the nonpublic reasons, because they might not understand that these nonpublic reasons play only a supporting role. By contrast, the principle of including public reasons assumes that citizens have the ability to listen with tolerance and even learn from others' religious and moral beliefs.

*751 The fourth reason to prefer the inclusion of public reason principle over that of excluding nonpublic reasons is that we ought to favor authenticity in political debate. Allowing the disclosure of nonpublic reasons reinforces authenticity; suppressing nonpublic reasons that really do play a foundational role detracts from authenticity. There are two supporting grounds for this fourth reason. First, one cannot regard something as a political virtue when it amounts to dissembling. But if nonpublic reasons really do play a foundational role, then their suppression may amount to dissembling - at least in those particular political debates where their foundational role becomes relevant to the debate. In those debates, the ordinary canons of fair argument would require disclosure. Second, there are reasons to believe that inauthenticity in political argument can be a very great evil. Given our history, we have good reason to fear political positions that are advanced on the basis of hidden agendas.

The fifth and final reason to prefer the principle of including public reasons over that of excluding nonpublic ones is that there may be times when allowing nonpublic reasons is necessary to prevent a great evil. The use of religious argument by the abolitionists may have been a case of this sort. [FN37] Assume that the abolitionists' case for emancipation of the slaves rested in part on the argument that it was an offense against God to treat as property a being that he created in his image. Assume further that offering this nonpublic reason in political debate was an essential precondition for the events that led to the freeing of the slaves. The nonpublic reason (the argument from the assumption that God created all humans in his image) served as the ground for a public reason (the fundamental equality of persons). The abolition of slavery was a very great good, and I think many Americans share the intuitive sense that the use of religious argument by the abolitionists was not an offense against political morality. That the principle of including public reasons fits this intuitive sense counts in its favor. Moreover, there are other cases of this sort. The contemporary civil rights movement also uses religious appeals, as powerfully illustrated by the speeches of Dr. Martin Luther King, Jr. [FN38]

*752 IV. A Restatement of the Liberal Ideal of Public Reason

This final section summarizes the argument by restating the liberal ideal of public reason in light of the many distinctions, qualifications, and arguments that have been explored so far.

First, the principle of laissez faire for private discussion should govern (1) private discussion, whether of political or private matters and (2) public discussion of moral and cultural matters. Two considerations are central here. Initially, there is no need for a more restrictive principle governing private discussions or public discussions that do not concern politics in a direct way. Civility and tolerance can flourish without extending the political ideal of public reason to discussions in families and communities of voluntary association such as churches. Public debate over ethics and politics may undermine civility and tolerance, but the risk is far less grave than in the case of public political debate over the coercive use of state power. In addition, the freedom of thought and expression would be gravely infringed if nonpublic reasons were viewed by citizens as not legitimately voiced and debated in the private sphere. Although the threat would be less grave if public discussion of ethics and politics were proscribed by an ideal of public reason, the loss would still be great. The discussion of art, literature, morality, and the like would be far less robust if citizens observed an ideal of restraint that removed deep and controversial questions about the good and the meaning of life from such discussions.

Second, public debate by private citizens should be governed by two principles. The first of these is the inclusion of public reasons in public discussion and private deliberation concerning the coercive use of state power. The case for the principle of including public reasons already has been stated at length, but the core of the argument is simple. Public reasons should be included because only public reasons can be viewed as reasonable by the public at large. The giving of reasons to one's fellow citizens is required by the notion that one should respect them as free and equal. An ideal which requires that public reasons be given but allows nonpublic reasons to be given as well will do the best job of fostering the virtues of civility and tolerance, as compared to the principle of laissez-faire or a principle of exclusion.

The second principle governing public debate is the exclusion of intolerance and disrespect for the freedom and equality of fellow citizens in public discussion. The ideal of public reason supports the exclusion of racist, sexist, and homophobic speech from public debate over the use of coercive state power. Allowing the inclusion of this sort of speech would not foster civility and tolerance; rather, it would undermine those values.

*753 Finally, the behavior of public officials in their official capacity should be governed by the principle of excluding nonpublic reasons. Public officials are different from private citizens because they personify the state; the statements of public officials in their official capacity are, in a real sense, the statements of the state and hence of the public at large. For this reason, it would be unfair to allow public officials to express their own deep convictions about the good as the official reasons for state action. Allowing public officials to advance nonpublic reasons would violate the requirement of treating all citizens fairly. Moreover, this requirement does not violate the freedom of conscience or expression of public officials. Public office is entered voluntarily, and public officials retain their full freedoms when they speak in their capacity as private citizens.

Together, these principles constitute a liberal ideal of public reason. Such an ideal is not aimed at limiting the common human capacity for reason. Public reason does limit the expression of some reasons in some contexts, but not for arbitrary or unprincipled reasons. Public reason does not require that we refrain from reflection about the matters of deepest significance; it does focus our reason on the pluralism that characterizes modernity. In the sphere of public political debate, an ideal of public reason must be justified in accordance with public reason. Because of this constraint, the justification of public reason by political liberalism is shallow in the sense that it can only draw on common sense and our public political culture. But even those resources are sufficient for public reason to make the case to each citizen that he or she should find the roots of the political ideal of public reason in his or her own deepest convictions. In this way, public reason transcends itself.

*754 Appendix: Historical Formulations of Public Reason

This Appendix presents a brief survey of the historical uses of the phrase "public reason"; this survey is not intended to provide an intellectual history. Rather, this recounting of the use of "public reason" in the writings of Hobbes, Rousseau, Jefferson, and Kant is intended to illustrate the variety of uses to which the phrase can be put. Some common themes emerge, but we should not assume that all uses of the phrase refer to the same concept. We begin with the earliest use of the phrase to be discussed, that by Thomas Hobbes. [FN39]

A. Hobbes

The phrase "public reason" is found in Thomas Hobbes' Leviathan . [FN40] The section of Leviathan in which this passage appears addresses the question, whose reason should govern the question of whether a purported miracle has occurred?

For in these times, I do not know one man, that ever saw any such wondrous work, done by the charm, or at the word, or prayer of a man, that a man endued but with a mediocrity of reason, would think supernaturall: and the question is no more, whether what we see done, be a Miracle; whether the Miracle we hear, or read of, were a reall work, and not the act of a tongue, or pen; but in plain terms, whether the report be true, or a lye. In which question we are not every one, to make our own private Reason, or Conscience, but the Publique Reason , that is, the reason of God's Supreme Lieutenant, Judge; and indeed we have made him Judge already, if wee have given him a Soveraign power, to doe all that is necessary for our peace and defence. A private man has alwaies the liberty, (because thought is free,) to beleeve, or not beleeve in his heart, those acts that have been given out for miracles, according as he shall see, what benefit can accrew by mens belief, to those that pretend, or countenance them, and thereby conjecture whether they be Miracles, or Lies. But when it comes to confession of that faith, the Private Reason must submit to the Publique ; that is to say, to God's Lieutenant. [FN41]</