1. Rawlsian Deliberative Democracy

There is some disagreement in the literature about the defining features of deliberative democracy. I follow Samuel Freeman in thinking that a Rawlsian deliberative democracy is one in which

• • “political agents or their representatives aim to collectively deliberate and vote their sincere and informed judgments regarding measures conducive to the common good of citizens.”Footnote ⁶

• • “citizens recognize a duty in their public political deliberations to cite public reasons—considerations that all reasonable citizens can accept in their capacity as democratic citizens—and to avoid public argument on the basis of reasons peculiar to their particular moral, religious, and philosophical views and incompatible with public reason.”Footnote ⁷

• • “[p]olitical agents are seen and see one another as democratic citizens who are politically free and equal participants in civic life.”Footnote ⁸

To fill in details, let us start with the word “public” in the second feature. Rawls distinguishes what he calls “the public political forum” from “the background culture.”Footnote ⁹ The public political forum consists of the discourse of judges, public officials, and candidates for office.Footnote ¹⁰ The background culture is the culture of civil society. It includes public media of all kinds; it also includes contributions to political philosophy like Rawls’s own. Discussion and deliberation about politics takes place in both. But the second feature above refers to “public deliberations.” What makes these deliberations public is that they are conducted in the public political forum. What makes them deliberations, rather than, say, declamations?

As Rawls understands it, public deliberation is discussion in the public forum that aims at reaching and justifying what I call “political outcomes”: the election of a candidate, the passage of a law, or the adoption of a policy. I take the first feature of deliberative democracy to require that in such discussion, citizens or their representatives do not just state their preferences, but rather exchange reasons purporting to justify the political outcomes they favor. When they exchange such reasons with the intention of reaching and justifying outcomes, they deliberate. According to the Rawlsian account of public reason, not all the considerations that could be brought forward in public deliberation have justificatory force. Only a subset does: that which the second feature refers to as “considerations that all reasonable citizens can accept in their capacity as democratic citizens.” These are “public reasons.”

Some considerations which do not have justificatory force may be brought forward in disregard of the fact that they do not. Others may be brought forward in good faith, by citizens who believe that they do justify. Even exchanges in the public forum which include reasons brought forward with such disregard might count as public deliberation, but let’s restrict ourselves to the exchange of reasons that is and is mutually known to be in good faith: the exchange of reasons in which participants try in good faith to reach justified political outcomes. This deliberation satisfies an important part of the first feature of deliberative democracy: it is sincere. When I refer to public deliberation from now on, it is sincere public deliberation that I have in mind.

Sometimes the phrase “public reason” is used—as by Rawls, as we shall see—to refer, not just to a consideration with reason-giving force in public deliberation, as in deliberative democracy’s second feature, but to an activity which includes the exchange of such considerations. On those occasions, it is used to mean “public reasoning.” To what activity does the honorific “public reason,” in this sense of “public reasoning,” apply?

To see the possibilities, consider a chain or set of sincere exchanges which results in a political outcome, perhaps because it is adopted after all “vote their sincere and informed judgments”—informed, that is, by the preceding deliberation. Even public deliberations which include reasons that are not themselves public can result in justified political outcomes. Does the whole set of deliberations which results in the justified outcome count as public reasoning? Or does the honorific apply just to that—perhaps artificially abstracted—fragment of deliberations that consists only of the exchange of reasons that are public?Footnote ¹¹

Call the former “the wide view of public reasoning” and the latter “the narrow view.” The difference between them matters because we want to know what is meant by saying that Rawlsian public reasoning does without the concept of truth. It also matters because, according to Rawlsian deliberative democracy, political outcomes should be based on sincere public deliberations. This raises the question of what the appropriate basis of a political outcome is. Is the whole set of exchanges, including non-public reasons, an appropriate basis? Or does the appropriate basis include only that part of public deliberation in which public reasons are exchanged? How the account of deliberative democracy is to be filled in depends upon how we answer these questions. I shall say something later about how I think it should be filled in, but first I will turn to the readings of Rawls I want to examine.

2. The Non-Permissive Reading

There are a number of passages in Political Liberalism in which Rawls abjures the concept of truth. Some commentators have taken these passages to have radical—and ultimately self-defeating—implications. Edmund Santurri cites a passage from Political Liberalism to claim that “political liberalism ‘does without the concept of truth’ altogether.”Footnote ¹² Rawls does say that political liberalism does without the concept of truth, but as we shall see, speaking of what “political liberalism does” is ambiguous. Without disambiguation, it is not clear that Santurri’s addition of the word “altogether” is warranted.

Other readers have drawn out implications of Rawls’s remarks about truth for what we have seen is an important element of Rawlsian deliberative democracy and what is an important element of Rawls’s political liberalism: public reasoning specifically. Thus Linda Zerilli takes Rawls’s “method of avoidance” to apply to what she calls “public reason” and implies that for him, “claims to the truth of one’s beliefs, values, or judgments have no place.”Footnote ¹³ The interpretation of Rawls according to which public reason does without truth is elaborated and defended by David Estlund and Joshua Cohen. Estlund might initially seem to echo Santurri when he says Rawls’s “stated view” is that “political liberalism need[s] to avoid truth claims.”Footnote ¹⁴ In a later essay, Estlund reiterates his commitment to what I call “the non-permissive reading,” remarking that “Rawls says that he hopes political liberalism can proceed without putting forward anything as true, preferring to rely on the property of reasonableness instead.”Footnote ¹⁵ But in the same essay, he says “Rawls proposes that political liberalism should be understood as banning the concept of truth from our proper justificatory practices.”Footnote ¹⁶ On one natural interpretation of the phrase “justificatory practices,” they are part of public reasoning somehow understood.

Cohen also endorses the non-permissive reading of what he calls “public reason,” saying that Rawls suggests “we can have truth or public reason, but not both. He embraces the idea of public reason as common ground under conditions of doctrinal disagreement, and concludes that we should leave truth aside.”Footnote ¹⁷ Cohen says that “Although I find the idea of public reason compelling, I disagree with Rawls’s claim that the concept of truth finds no place in it.”Footnote ¹⁸ He goes on to emphasize that:

Because he thinks Rawls claims that “public reason lacks the concept” of truth,Footnote ²⁰ Cohen dubs the view that his non-permissive reading attributes to Rawls Rawls’s “No Concept view” of public reason.Footnote ²¹

Santurri, Zerilli, Estlund, and Cohen cite a number of passages from Political Liberalism to support their readings. All but Zerilli cite Rawls’s claim that “political liberalism, rather than referring to its political conception of justice as true, refers to it as reasonable instead.”Footnote ²² All also cite the page where he says that “within itself the political conception does without the concept of truth.”Footnote ²³ And all cite Rawls’s discussion of political constructivism, where he says that “political liberalism … need not go beyond its conception of a reasonable judgment and may leave the concept of a true moral judgment to comprehensive doctrines.”Footnote ²⁴ What licences the move from claims about what political liberalism does and does not do, to claims about the relationship between public reason and truth?

Let’s understand political liberalism as an approach to certain fundamental questions in political philosophy the practitioners of which—call them “political liberals”—assume what Rawls calls “the fact of reasonable pluralism.”Footnote ²⁵ Because political liberals assume reasonable pluralism, they attempt to answer those fundamental questions without addressing the moral and metaphysical questions on which many political philosophies have been premised, such as questions about the nature and ends of persons. Instead, political liberals start from ideas found in the public political culture of democratic societies. The results are political conceptions of justice.

Rawls’s political conception includes distributive principles and a principle of legitimacy. It also includes guidelines of public reasons which say what considerations have justificatory force in favor of political outcomes. The guidelines spell out the duty of civility and elucidate the second feature of deliberative democracy by saying what kinds of considerations citizens must appeal to in their attempts to justify political outcomes. They also say what considerations citizens should not appeal to, since appealing to them—or appealing to them without offering other, public, reasons besides—would run afoul of the third feature of deliberative democracy, which implies citizens’ duty to address one another as free equals.

The values and principles furnished by a political conception of justice therefore provide what Rawls calls “a guiding framework for deliberation,”Footnote ²⁶ by which I take him to mean a guiding framework for public deliberation. In saying that political liberalism does without the concept of truth, and makes do with reasonability instead, Rawls might be making one or more of several claims about the exclusion of truth from the reasoning or public deliberation which takes place using that framework. Among the things he might mean are:

1. (1) Public deliberation need not include truth-claims for its political outcomes to be justified.

Another possibility is:

1. (2) Truth-claims are never public reasons.

Since public reasons are claims with reason-giving force in public deliberation, (2) implies that truth-claims never have such force even if the propositional content of which truth is predicated is in fact true.

(1) and (2) concern public reasons understood as considerations which count in public deliberation. But in saying that political liberalism does without the concept of truth, Rawls might be staking a claim about public reasoning. Because, as we saw in the previous section, “public reasoning” can mean one of a couple of things, there are a couple of claims about public reasoning Rawls might be making. One takes public reasoning narrowly and says:

1. (3) Truth-claims have no place in public reasoning, understood as that part of public deliberation that includes only reasons which are public.

Another takes public reasoning widely and says:

1. (4) Truth-claims are never public reasons. Even though they are not, they could still enter into public reasoning but those who enter them thereby violate the duty of civility.

But in saying that political liberalism does without concept of truth, Rawls might not be referring to such reasons at all. He might instead be referring to the defense he—writing as a political liberal—provides for justice as fairness and its “guiding framework.” That defense is offered in the background culture rather than in the public forum. It is founded on basic intuitive ideas of the person and society, and draws out their implications. And so, in saying that political liberalism does without the concept of truth, Rawls might mean

1. (5) The presentation of political liberalism should defend the basic ideas on which it is based, and the principles drawn out on their bases, as reasonable, while not addressing the question of their truth.

How are these five claims related? (1) clearly does not imply (2). For truth-claims could be public, contrary to (2), but unnecessary for the justification of political outcomes, as (1) says, because for any justified outcome, there is always a set of public reasons sufficient to justify it which does not include truth-claims. But if there are some political outcomes that are justified, and justification is always by sufficient public reasons, then (2) does imply (1). Moreover, (2) clearly implies (3). (3) does not, however, imply (2) since one could consistently hold that truth-claims have no place in public reasoning narrowly construed, while holding that there is some other explanation for their having no place in it than that they are not public reasons.

Because (1) does not have any of the allegedly problematic implications for Rawls that I discuss in the next section, I ignore it in what follows. What is most important for my purposes is that (5) is consistent with, but logically independent of, (2), (3), and (4). To state a possibility I elaborate later, Rawls could—without inconsistency—propose as reasonable a guideline of public reasoning according to which some truth-claims count as public reasons, while remaining silent about the truth of that guideline.

We have seen that Estlund reads Rawls as “banning truth from our proper justificatory practices.”Footnote ²⁷ Whether he takes Rawls to endorse (2) depends upon what he means by “proper justificatory practices.” I have said that public deliberation is the exchange of putatively justificatory reasons for political outcomes. When citizens engage in public reasoning so understood, they exchange reasons they take to contribute to the justification of the outcomes they favor. If Estlund took public deliberation to be a justificatory practice because citizens engaged in it try to justify the outcomes they favor, then he would read Rawls as taking the wide view of public reason and banning truth from it, perhaps on pain of violating the duty of civility. In that case, he would read Rawls as endorsing (4). But I do not believe that is what Estlund has in mind. I believe he takes “proper justificatory practice” to refer to the practices of providing considerations that really do justify a political outcome. Taken this way, proper justificatory practice includes the practice of a court’s or a legislature’s displaying reasons that really do justify the outcomes it has reached, and of citizens adducing reasons that really do have justificatory force. This is the set of practices from which Estlund thinks Rawls bans truth. He takes Rawls to ban it because he reads Rawls as endorsing (2).

By contrast, I believe Cohen takes Rawls to endorse (3). It is possible to hold that truth-claims have no place in public reasoning because truth-claims are never public reasons, and so to endorse (3) because one endorses (2). But (3) does not entail (2), and so Cohen is not committed to thinking that Rawls endorses (2) just because he thinks Rawls endorses (3). There is a passage, to which I shall return, in which Cohen discusses someone who believes that human beings have a right to religious liberty on the basis of natural law. That person “is simply not permitted to say everything he or she believes about that proposition—in particular, not to say that it is true and a consequence of more fundamental truths.”Footnote ²⁸ If “not permitted” means “not morally permitted,” and if the duty of civility is the source of the moral prohibition, then Cohen would read Rawls as endorsing (4). But I take Cohen to mean that the prohibition is conceptual rather than moral: he takes Rawls to mean that the conceptual resources of public reason do not permit the person “to say everything he or she believes.” And so the quoted remark is consistent with my claim that Cohen takes Rawls to endorse (3) rather than (4).

One advantage of distinguishing (2), (3), (4), and (5) is that the distinctions make it possible to pry apart versions of the non-permissive reading—those according to which Rawls endorses (2) or (3)—which may initially have seemed quite similar.Footnote ²⁹ They also make it possible to distinguish my reading from others. For on my reading, many of the passages in Political Liberalism and elsewhere that are taken to support the imputation of (2) or (3) to Rawls in fact support the imputation of (5). Because (5) is compatible with the denials of (2), (3), and (4), this leaves open the possibility that Rawlsian public reasoning can allow for at least some truth-claims.

3. Untoward Implications

If some version of the non-permissive reading of Rawlsian public reason is right, then Rawls’s view has implications that seem problematic, some of which have been surveyed by Cohen and Estlund.

4. What do the Texts Support?

It seems unquestionable that Rawls endorses (5) since it seems unquestionable that it describes the way Rawls himself presents political liberalism. That it does makes (5) importantly different from (2)–(4) because they concern reasons or reasoning in the public forum while (5) does not. In this section, I question the textual support for imputing (2)–(4) by arguing that passages which are often taken to support their imputation to Rawls in fact support the imputation of (5).

We saw that Zerilli takes Rawls’s “method of avoidance” to apply to citizens’ public reasoning and to imply that, for him, “claims to the truth of one’s beliefs, values, or judgments have no place.”Footnote ³⁹ If this were what the method of avoidance implied, then Rawls would be committed to (3). But what Rawls says about the method of avoidance supports a claim that is critical to my reading, the claim that many of the passages in which Rawls forswears appeal to truth refer to his own exposition of justice as fairness. He says that the method applies “to the presentation of justice as fairness and … how it is set up.”Footnote ⁴⁰ He never says that it applies to citizens’ public reasoning.

In a passage which is taken to support the imputation of (2) or (3), Rawls does say that “within itself the political conception does without the concept of truth.”Footnote ⁴¹ But the phrase “within itself” and the context are crucial. The passage is part of an extended contrast between rational intuitionism and constructivism. The rational intuitionist, Rawls says, claims that moral principles are arrived at by accessing a prior and independent moral order. When Rawls says that “within itself the political conception does without the concept of truth” in this context, I take him to mean that the political constructivist can show how to arrive at political principles without claiming to access such an order, by constructing them on the basis of conceptions of the person and society. And so, in laying out the political conception, he need not appeal to the concept of truth or make claims about being true to such an order.

In a passage in “Political Not Metaphysical” that might be taken to support the imputation of (2) or (3), Rawls says

The impersonality of Rawls’s language is confusing, since justice as fairness does not present itself. Rawls presents it. I take him to be referring here to how he presents it. He presents or proposes it as a possible basis for consensus, rather than as a conception that is true. But this is compatible with people who accept his proposal reasoning with it as if it is true. It is therefore compatible with the denial of (2) and (3), and so does not lend them support.

In “Reply to Habermas”, Rawls does say:

The remark about the sufficiency of the “idea of the reasonable" might be thought to tell in favor of (2) or (3), but I do not believe that it does.

For one thing, the sufficiency of the concept of reasonability would not imply that truth-claims are not good reasons or have no place in public reasoning; it would only imply their superfluity. For another, here as in the passage from “Political Not Metaphysical,” the impersonality of the phrases “political liberalism says” and “political liberalism views” make the whole passage difficult to interpret. On my reading of this passage as of the earlier one, Rawls is not talking about public reasoning. Rather, when he uses these impersonal phrases, what he really means is what he says and views. When he puts forward political liberalism, he does not put it forward as true, nor does he say that truth is necessary for “discussing questions of constitutional essentials and matters of basic justice.” But as I argue in the next section, if some citizens believe that asserting the tenets of political liberalism are true is necessary, Rawls leaves it open to them to do so, provided that their interventions are also received as reasonable.

My claim that the texts support (5) rather than (2), (3), and (4) might seem to depend on a distinction I drew between the former claim and the latter three: that (5) refers to reasoning in the background while (2), (3), and (4) refer to reasons or reasoning in the public forum. But, it may be objected, that distinction is untenable once what Rawls calls “the full publicity condition” is taken into account. For when that condition is satisfied, “everything that we would say—you and I—when we set up justice as fairness and reflect why we proceed in one way rather than another” is “publicly available” and “present in the public culture.”Footnote ⁴⁴ If what is present in the public culture can be introduced into public reasoning, then my distinction between (5) on the one hand, and (2), (3), and (4) on the other, is effaced.

But this objection can be lodged by those who would impute (2), (3), or (4) to Rawls only if they are willing to saddle him with inconsistency. Those who would impute (2) or (4) to him would have to read him as thinking both that what is present in the public culture can be entered into public reasoning and that truth-claims are never good public reasons. But since the public culture includes truth-claims,Footnote ⁴⁵ Rawls could not consistently hold both. Those who impute (3) would have to read him as holding both that what he says when he sets up justice as fairness can enter into public reasoning and that public reasoning lacks the concept of truth. But since among the claims Rawls makes when he sets up justice as fairness is that “political liberalism does not use the concept of truth,” he cannot consistently hold both of those theses either.

Why does Rawls present political liberalism without putting it forward as true? The phrase “some other idea instead” in the passage just above provides a clue. Political conceptions are put forward in the hope that they will be the objects of an overlapping consensus. Among the comprehensive doctrines whose consent is hoped for, there will be some—like the natural law view to which Cohen refers—according to which principles that bear on constitutional essentials and matters of basic justice can have truth-values. These comprehensive doctrines will apply the concept of truth to practical principles. But there may be other doctrines according to which those principles are not the sorts of things that can be true or false. According to these doctrines, Rawls’s principles are, in one respect at least, like the rules of a game: they are not truth-apt because there is nothing they are or can be true of. And so rather than using the concept of truth, these doctrines predicate “some other idea” of the principles and rules they favor, such as reasonability. Proposing a conception of justice without claiming truth for it leaves open the possibility that such views can be among those that overlap on the conception.Footnote ⁴⁶ I suspect that the Kantian constructivism Rawls endorsed in the Dewey lectures is one such view.Footnote ⁴⁷ If so, then he was intimately familiar with the kind of comprehensive view that required him to refrain from presenting justice as fairness as true once he turned to political liberalism.

5. What of (1)–(4)?

(5) is logically independent of (2), (3), and (4) so, even if the passages which are usually taken to support the attribution of (2), (3), or (4) to Rawls are better read as supporting the attribution of (5), that does not show that Rawls does not endorse one of (2)–(4). Of these possibilities, (4) is the one whose imputation to Rawls can most obviously be dispensed with. It says:

• (4) Truth-claims are never public reasons. Even though they are not, they could still enter into public reasoning but those who enter them thereby violate the duty of civility.

To see how the imputation of (4) can be dispensed with, consider how we might be led to the reading that imputes it.

Paradigm cases of reasons which are not public are reasons drawn from moral and religious doctrines. These are not reasons that can be recognized as such by citizens just as free equals because their recognition as reasons depends upon the acceptance of the doctrines from which they are drawn. But though moral and religious claims are paradigms of non-public reasons, claims which depend upon contested philosophical theses—such as contested theses in the philosophy of mind—are non-public as well. The nature of truth is also philosophically contested and so, it might be thought, Rawls would recognize that assertions of truth are no less problematic than assertions of claims about the nature of persons and minds. Thus truth-claims are relevantly similar to other reasons which are not public, and that is why they are not public either, just as (4) says. If Rawls also believes that the introduction of non-public reasons into public deliberation violates the duty of civility, then he holds that the introduction of truth-claims violates that duty, as (4) also says.

But the last step in this argument is contrary to Rawls’s proviso. According to the proviso:

I take “public reason” here to mean “public reasoning.” And I take the first phrase of the proviso—that “reasonable [comprehensive] doctrines may be introduced into public reason[ing] at any time”—to show that Rawls takes a wide view of public reasoning, according to which the phrase “public reasoning” denotes the whole set of deliberations that leads to a justified political outcome.

This part of the proviso indicates that public reasoning so understood can include reasons which are not public, such as contested claims in morals, religion, or metaphysics. If truth-claims are relevantly similar to those claims, then this part of the proviso also allows for their introduction into public reasoning, as the second sentence of (4) says. But if truth-claims are relevantly similar to other contested claims, then the proviso also seems to imply—contrary to (4)—that their introduction into public reasoning does not violate the duty of civility, since the introduction of those claims does not violate it. So if truth-claims are relevantly similar to other contested claims, Rawls’s proviso implies that he does not endorse (4). And if they are not relevantly similar, then it is not clear what ground would be for imputing (4) to him, since the reasoning I sketched above supporting its imputation depends upon relevant similarity.

Recall that (3) says:

1. (3) Truth-claims have no place in public reasoning, understood as that part of public deliberation that includes only reasons which are public.

Perhaps Rawls would endorse (3). But if he consistently holds the wide view of public reasoning that is at work in his proviso, then he does not actually endorse it. The argument that he would needs to be qualified to make clear that (3) uses “public reasoning” more narrowly than Rawls himself does.

In “The Idea of Public Reason Revisited,” Rawls states the proviso differently, saying that it “allows us to introduce into political discussion at any time our comprehensive doctrine.”Footnote ⁴⁹ Rawls does not explain a shift of language that seems puzzling once the wide and narrow views of public reasoning are distinguished. One possible explanation is that he takes “public political discussion” and “public reasoning” to be co-referential, as the wide view holds, so that the conclusion in the text is correct. Another is that he came to think the two are distinct and that the narrow view of public reasoning is correct. In that case, (3) could be imputed to him in unqualified form. But what seemed problematic about the imputation of (3)—or part of what seemed problematic about its imputation—is that it implies No Truth-Claims Backing the Overlap. But since the proviso would allow the introduction of comprehensive doctrine into public political discussion, it is hard to see what is problematic about that implication or why Rawls’s acceptance of (3) would motivate revision of his view.

I suggested earlier that the difference between the two ways of understanding public reasoning is significant for our understanding of deliberative democracy. Before I elaborate this, let us turn to the claim that Rawls endorses:

1. (2) Truth-claims are never public reasons.

(2) is the first sentence of (4). When I asked whether Rawls endorses (4), I laid out reasons he might have for accepting it. I said that truth-claims are not public because truth is a philosophically contested concept, as are the concepts employed in paradigm cases of non-public reasons. But it seems to me that whether truth-claims are contested depends upon the contestability of the propositional content to which the truth-predicate is attached. Truth-claims about morals and religion, even if admissible into public reasoning understood widely, seem controversial in a way that truth-claims about empirical matters are not. If that is right then, despite what I said above for the sake of argument, truth-claims about empirical matters can be public reasons and (2) cannot be attributed to Rawls without qualification.

There is also some textual basis for thinking it cannot be. In Political Liberalism, Rawls says:

Reasoning about constitutional design is conducted in the public forum since it is conducted at stage 2 of Rawls’s Four-Stage sequence.Footnote ⁵¹ That reasoning depends upon “the scope and limits of political and social institutions and how they can be made to work effectively,” and ultimately upon a society’s history. The justification of a choice of constitution depends upon empirical claims. It seems incontestable that, as Rawls observes, “here the concept of truth applies.” If the empirical claims can help to justify a “design of the constitution” as “most appropriate,” then why not the assertions that those empirical claims are true? These assertions may not be necessary to justify the design of a constitution; perhaps the empirical claims suffice. That would lend some support to the imputation of (1). But if the truth-claims can play any role at all, then they are public reasons, contrary to (2).

Earlier in Political Liberalism, Rawls also seems to allow appeal to truth in the justification of legislation that bears on fundamental questions. He says:

If Rawls thinks the application of principles of justice to a matter of basic justice like income distribution can be justified by claims which are widely accepted as true, it is hard to see why he would think the claim that they are true could not do the same justificatory work. If he thinks they could, then he does not endorse (2) in unqualified form.

In Section 3, I surveyed implications of the non-permissive reading of Rawls that motivated suggestions for revising Rawlsian public reason. The claim that Rawlsian public reason has those implications depended upon the imputation of (2), (3), and (4) to him. I have argued that many of the passages that were taken to support the imputations are best read as supporting the imputation of (5). I have also argued that Rawls does not endorse (4), that (3) can be attributed to him only if “public reasoning” in it is taken in a narrower sense than the sense Rawls attaches to that phrase, and that (2) can be attributed to him only if qualified so as to allow that truth-claims about empirical matters can be public reasons. These arguments undercut the support for No Truth-Claims Backing the Overlap, The Irrelevant Truth of Science, and No Soundness. Qualifying (2) in the way that I have suggested does not undercut the argument for No True Guidelines. But that implication depends upon the supposition that a defense of admissible reasons is itself a necessary part of public reasoning. It is not clear that it is. Proposed revisions of and additions to Rawls’s account of public reason may be useful, but not because that account has the problematic implications that are alleged.

6. Deliberative Democracy Revisited

By undercutting support for at least some of the untoward implications, the arguments offered here do much to block the further implications that the non-permissive reading was said to have for Rawlsian deliberative democracy. They also open the possibility that Rawlsian public reasoning is more congenial to truth-claims than the non-permissive reading allows. This possibility, in turn, affects how we fill out the ideal of deliberative democracy.

As we have seen, the proviso says that “reasonable [comprehensive] doctrines may be introduced in public reason at any time, provided that in due course public reasons, given by a reasonable political conception, are presented sufficient to support whatever the comprehensive doctrines are introduced to support.”Footnote ⁵³ As I read this, natural law adherents may defend claims about religious liberty, for instance, on grounds drawn from their comprehensive doctrine, provided that in due course they defend those claims on the basis of political values as well.

I take Rawls’s proviso to suggest that what he wants to rule out is citizens moving from their belief that a proposition is true on the basis of their comprehensive doctrine to the thought that its truth is sufficient to give that proposition reason-giving force in public deliberation. If a proposition drawn from comprehensive doctrine is to have reason-giving force, it must be reasonable and supportable as such on the basis of political values. Citizens are permitted to introduce claims from their comprehensive doctrines that they believe to be true, provided they also believe those claims to be reasonable and are prepared to show that they are. And so the problem against which Rawls warns is not that of citizens’ introducing what they explicitly describe as truth-claims into public deliberation. The problem is that of citizens’ taking their truth to be sufficient for introducing them as good reasons. Thus, speaking of political conceptions rather than their constituent propositions, Rawls says that “holding a political conception as true, and for that reason alone the one suitable basis of public reason” is a mistake.Footnote ⁵⁴

In societies as we know them, ordinary citizens rarely engage in public deliberation because they rarely enter the public political forum. But “[a]t the heart of the institutionalization of the deliberative procedure,” Cohen writes, “is the existence of arenas in which citizens can propose issues for the political agenda and participate in debate about those issues.”Footnote ⁵⁵ And so if contemporary democracies were deliberative democracies, perhaps ordinary citizens would participate in public deliberation to a much greater extent than they currently do. On my picture of how they would deliberate, those who think their comprehensive doctrines have implications for public policy would draw out those implications in deliberation, and may well articulate the metaphysical backing of their views. In doing so, they must satisfy Rawls’s proviso, and so justify their policy conclusions by appeal to public reasons “in due course.” The question is what “in due course” requires of them.

We have seen that citizens must believe the positions they advocate are reasonable and must be ready to show that they are. But accounts of public reason are regrettably short on the study of conversational pragmatics. My picture of public deliberation depends upon a fact that is not sufficiently taken into account in the public reason literature: that citizens engaged in it operate with knowledge of background and context. Rawls gives some indication of this, saying that

Rawls also says that, in the public justification of a conception of justice, citizens do not look into the content of one another’s comprehensive doctrines but “take into account and give some weight to only the fact—the existence—of the reasonable overlapping consensus itself.”Footnote ⁵⁷ But he does not ask how citizens come to know that others’ comprehensive doctrines overlap on the political conception. Their coming to know this surely involves their gaining some acquaintance with the comprehensive doctrines that are widely adhered to in their society. That acquaintance is part of the background knowledge citizens bring to public deliberation. Having this background knowledge, they may recognize the goodwill of someone who argues for racial equality or religious liberty on the basis of her comprehensive doctrine, and see that she is defending a position that is reasonable and supportable by political values. In that case, they may never press her to provide a defense based on those values. “In due course” never comes.Footnote ⁵⁸

It may be thought that, when there is an overlapping consensus on a political conception of justice, it is mutual knowledge among citizens that all are collectively committed to reasoning within that conception just as the philosopher proposed it, and so to eschewing the concept of truth. It may be thought that, when there is an overlapping consensus, there is overlap on (2), (3), or (4). But nothing in the idea of an overlapping consensus compels this reading. My view is that, when there is an overlapping consensus, part of the object of the consensus is that public deliberation be conducted as the proviso permits, and the proviso permits appeal to a wide range of claims, including appeals to truth.