Political liberalism,free speech and public reason

Introduction

In this paper, I examine an issue which has been surprisingly overlooked in the literature on John Rawls' Political Liberalism ¹ and his conception of public reason. This is Rawls' repeated emphasis on the fact that the constraints of public reason are only moral and should not be enforced by law. According to Rawls, public reason is ‘the reason of equal citizens who, as a collective body, exercise final and coercive power over one another in enacting laws and in amending their constitution’. ² Citizens have a duty (what Rawls calls the ‘duty of civility’ ³ ) to appeal only to political values (rather than their comprehensive conceptions of the good) when making decisions about ‘“constitutional essentials” and questions of basic justice’, ⁴ for example, fundamental issues concerning ‘who has the right to vote, or what religions are to be tolerated, or who is to be assured fair equality of opportunity, or to hold property’. ⁵ The duty of civility, Rawls argues, is ‘a moral, not a legal, duty’. ⁶ More specifically, he claims, ‘it is not a legal duty, for in that case it would be incompatible with freedom of speech’. ⁷

Rawls' claim has been endorsed and reinforced by other authors. In response to those who accuse public reason of unduly restricting freedom of speech, ⁸ for example, Stephen Macedo argues that

[a]dvocates of public reasonableness do not advocate restrictions on political speech … No one is suggesting that the contours of liberal public reason should be used to define the limits of constitutional rights to free speech. Public reason helps define a moral ideal, not a legal requirement’. ⁹

This response, however, begs the question. Macedo, like Rawls, simply fails to provide a proper argument for the claim that the protection of free speech always ought to override any attempt to legally enforce the duty of civility. Neither has any other author (Rawlsian or non-Rawlsian) endeavoured to analyse this problem. ¹⁰

In order to address this significant gap in the literature, my analysis in this paper will proceed as follows. In the first part of the paper, I will assess whether Rawls' claim may simply be due to the practical difficulties raised by any attempt to enforce the duty of civility by law. In response to this claim, I will argue both that the implementation of legislation does not entail the adoption of maximally intrusive and complex measures and that the various provisos gradually introduced by Rawls in his conception of public reason would render the implementation of the duty of civility less impractical than one might expect. If impracticality is not an issue, I will claim, Rawls' unwillingness to legally enforce the duty of civility must be driven by deeper normative reasons grounded in a conception of free speech.

In the second part of the paper, I will therefore assess which arguments for free speech may be consistent with Rawls' political liberalism. Neither in A Theory of Justice ¹¹ nor in Political Liberalism does Rawls offer a systematic philosophical defence of free speech. He simply includes freedom of speech among the ‘equal basic rights and liberties of citizenship’ ¹² and briefly discusses free political speech in connection with the issue of seditious libel. ¹³ Providing a Rawlsian theory of free speech is therefore interesting and useful in its own right but my main goal will be to show how such a theory may contribute in explaining Rawls' rejection of a legally enforced duty of civility.

I will first focus on those arguments that justify freedom of speech based on the ideas of truth and self-fulfilment. Both arguments, I will argue, rely on comprehensive doctrines and therefore cannot provide a freestanding political justification for free speech. Freedom of speech, I will claim, can be justified instead in political terms on the basis of Rawls' political conception of the person and of its two moral powers. However, I will argue, Rawls' wide view of public reason already allows scope for the kind of free speech necessary for the exercise of the two moral powers and therefore cannot explain Rawls' opposition to the legal enforcement of the duty of civility. Such opposition, I will claim, can only be explained on the basis of a defence of free speech grounded in the ideas of democracy and political legitimacy, which is consistent with Rawls' own account of free speech.

For Rawls, therefore, both freedom of speech and the duty of civility are aimed at guaranteeing political legitimacy. This creates a tension within political liberalism. However, while public reason and the duty of civility are clearly central to political liberalism, it is not clear whether and why this should also be the case for unconstrained freedom of speech. I will therefore conclude that, in spite of what Rawls argues in defence of unconstrained freedom of speech, political liberalism is in principle consistent with some restrictions of free speech, including those which would result from the legal enforcement of the duty of civility.

Before proceeding with my analysis, one important point should be stressed. It is not my aim, in this paper, to either defend or reject Rawls' idea of public reason and the duty of civility. Instead, I intend to show that Rawls (and those committed to political liberalism) could endorse the legal enforcement of the duty of civility, given the importance that the latter plays within political liberalism and given that Rawls' wide idea of public reason already allows significant scope for freedom of speech. This would, of course, prevent Rawls and Rawlsians from stressing the purely moral nature of the duty of civility when challenged by those who consider that duty unduly restrictive. Yet this may just be a price that Rawls and political liberals should be prepared to pay. If they are truly committed to the duty of civility, they should be ready to defend it on the basis of its rationale, rather than by stating that the duty is moral and not legal.

Enforcing the duty of civility through law: An impractical task?

First of all, it is worth assessing whether Rawls' rejection of a legally enforced duty of civility may be chiefly driven by practical concerns. One might first think that any attempt by a state to implement a hypothetical ‘Public Reason Act’ would inevitably lead to an excessive level of state censorship and interference with the lives of its citizens. Police control of public debate, for example, would become more intense; CCTV cameras (especially those equipped with audio surveillance microphones) would become even more widespread than they are now; newspapers, television channels and internet users would see their freedom of expression seriously curtailed.

Yet this image is excessively bleak. The approval of a law or policy does not entail that it should be implemented through maximally intrusive measures. For example, even though in liberal democratic polities there are laws against spousal abuse or child abuse governments have not installed cameras in everyone's homes in order to constantly control people's behaviour. This is not only because most people would in fact oppose this kind of measure but also especially because governments (and people more generally) normally weigh the detection of crime (and the security and protection of individual interests resulting from it) against other values which are also important to people generally (e.g. privacy).

Furthermore, the picture becomes even less bleak once one begins to consider the qualifications that Rawls gradually added to his conception of public reason throughout his later work. First of all, it should be remembered that the duty of civility, for Rawls, does not apply to all levels of society. The constraints of public reason, that is

do not apply to our personal deliberations and reflections about political questions, or to the reasoning about them by members of associations such as churches and universities, all of which is a vital part of the background culture. Plainly, religious, philosophical, and moral considerations of many kinds may here properly play a role. ¹⁴

This rules out entirely what is perhaps one of the most feared dangers of state censorship, that is, its interference with the private lives of citizens as members of families and various kinds of groups and associations within civil society. Clearly, state surveillance at this level (e.g. in the form of wiretapping of private houses, police presence at church meetings, censorship of the newsletters produced by sport or music clubs, etc.) would be ruled out by Rawls' account.

However, Rawls also claims that ‘the ideal of public reason does hold for citizens when they engage in political advocacy in the public forum, and thus for members of political parties and for candidates in their campaigns and for other groups who support them’. ¹⁵ Rawls also highlights that, among public officials, public reason applies ‘to legislators when they speak on the floor of parliament, and to the executive in its public acts and pronouncements’. ¹⁶ In ‘The Idea of Public Reason Revisited’, ¹⁷ Rawls remarks again that public reason applies to political deliberation in the ‘public political forum … [which includes] … the discourse of judges … ; the discourse of government officials … ; and finally, the discourse of candidates for public office and their campaign managers, especially in their public oratory, party platforms, and political statements’. ¹⁸

Rawls' claims provide some clear indications of where the state should intervene in order to enforce the constraints of public reason, should the latter be made legally binding. His emphasis on judges and members of the legislature and the executive, that is, gives us clear guidelines regarding where the state should concentrate its action. It would not be difficult, in principle, to enforce the duty of civility within parliament or cabinet meetings, or during the deliberations of constitutional courts. For example, as Patrick Neal notes, the ‘“Lemon test” for determining whether legislation violates the Establishment clause of the First Amendment’ ¹⁹ represents somehow a way of legally enforcing the duty of civility. The Lemon test demands that legislation about religion must have a ‘legitimate secular purpose’. ²⁰ Even more clearly, Jürgen Habermas suggests that ‘[i]n parliament, for example, the standing rules of procedure of the house must empower the house leader to have religious statements or justifications expunged from the minutes’. ²¹ It is somehow surprising that no other author, apart from Habermas, has put forward similar practical suggestions on how to legally enforce conversational constraints analogous to the Rawlsian duty of civility in an institutional context. ²² What these two examples suggest, however, is that the practical implementation of the duty of civility may often be less problematic than expected.

Similar considerations can be made regarding electoral campaigns and party manifestoes, the other area highlighted by Rawls. It would not be excessively difficult to monitor campaign speeches, both in public spaces and on television, in order to check whether candidates and other party affiliates make references to comprehensive doctrines in endorsing coercive legislation about constitutional essentials and matters of basic justice. Those who do could be fined and/or prevented from speaking again in public for a specified length of time. It would be even easier to monitor party manifestoes and programmes and require parties to expunge from them any references to comprehensive doctrines. For example, the present Constitution of Portugal establishes that

[w]ithout prejudice to the philosophy or ideology that underlies their manifestoes, political parties shall not employ names that contain expressions which are directly related to any religion or church, or emblems that can be confused with national or religious symbols. ²³

Even though this constitutional provision only has a symbolic significance, as it expressly does not concern parties’ ideology and philosophy, it could potentially be extended to the latter. Parties, that is, could be required by law not to employ arguments grounded in religious (and non-religious) comprehensive doctrines in order to support programmatic measures about fundamental matters in their manifestoes. This therefore provides us with an idea of what a constitutional or legal application of the duty of civility to party manifestoes and programmes would look like.

More problematic and especially intrusive, however, would be the implementation of Rawls' view that public reason should also regulate ‘how citizens are to cast their vote on these [i.e. fundamental] questions’. ²⁴ Yet Rawls' demand is not as restrictive as it seems. On the one hand, if citizens are involved in campaigning or other forms of ‘political advocacy in the public forum’, ²⁵ they are clearly subject to the constraints of public reason as they no longer differ from, say, candidates or party affiliates. On the other hand, if citizens are not engaged in public deliberation, those constraints only require them to vote on fundamental matters based on public reasons. This means that within their conscience they ought to find public reasons (alongside reasons based on comprehensive doctrines) justifying their choice to vote for a certain policy, candidate or party.

However, when citizens' deliberation regarding how to vote is only conducted within their conscience, this cannot have any implications for freedom of speech as our thoughts are beyond governmental reach. As Frederick Schauer points out

[w]e can think silently. It is not necessary to speak or write in order to think, and when we think silently, our thoughts are beyond the reach of government sanction. Obviously thoughts can be influenced by government. Propaganda is an example, and so is a system of explicit or implicit rewards. But a silent thought qua thought is immune from punishment, and to that extent is discretely different from outward expression or communication. ²⁶

When deciding how to enforce the duty of civility in connection with citizens' voting decisions, therefore, issues of practical implementation become irrelevant. ²⁷

Practical issues, however, might arise from Rawls' view that the constraints of public reason only apply to constitutional essentials and matters of basic justice but not to ordinary legislative issues such as ‘much tax legislation and many laws regulating property; statutes protecting the environment and controlling pollution; establishing national parks and preserving wilderness areas and animal and plant species and laying aside funds for museums and the arts’. ²⁸ I set aside, in the present analysis, the question of whether Rawls' distinction between constitutional and non-constitutional matters is in fact justified. ²⁹ From a practical point of view, it is sufficient to say that once we have drawn a line between constitutional and non-constitutional matters (regardless of whether we actually endorse Rawls' own distinction), then it becomes easy to identify references to constitutional essentials in, say, constitutional court deliberations, parliamentary debates, campaign speeches or party manifestoes, and implement the relevant free speech restrictions based on the duty of civility.

A further aspect, however, deserves greater attention. In ‘The Idea of Public Reason Revisited’, ³⁰ Rawls significantly relaxed his view of public reason by introducing a ‘wide’ conception. This establishes that

reasonable comprehensive doctrines, religious or non-religious, may be introduced in public political discussion at any time, provided that in due course proper political reasons – and not reasons given solely by comprehensive doctrines – are presented that are sufficient to support whatever the comprehensive doctrines introduced are said to support. ³¹

Rawls does not specify what ‘in due course’ means precisely, for example, whether the proviso should be satisfied ‘[o]n the same day or some later day’. ³² He argues instead that

the details about how to satisfy the proviso must be worked out in practice and cannot feasibly be governed by a clear family of rules given in advance. How they work out is determined by the nature of the public political culture and calls for good sense and understanding. ³³

How would the proviso, then, affect the legal enforcement of the duty of civility? Certainly, thanks to Rawls' elusiveness, it would give legislators enough time and room to decide what kind of timescale should regulate citizens' duty to provide public reasons in order to justify, say, laws they have endorsed qua MPs in the parliament, decisions they have made qua judges in the constitutional court or policies they have advocated qua party candidates during a campaign speech or television debate. In the first and second cases, for example, MPs and judges could be asked to provide within a set period of time (e.g. one week, one month, etc.) public reasons to be added to the relevant parliamentary minutes or the text of legislation or constitutional court rulings. This would be a much less restrictive measure than Habermas's abovementioned suggestion that appeals to religious arguments should be removed straight away from parliamentary minutes.

Candidates during election campaigns who have breached the duty of civility, instead, could be required to make a public statement by a certain deadline (which should be sufficiently earlier than the day of the election), providing public reasons (in addition to the non-public reasons they have previously invoked) in support of their advocated measure(s). Similarly, party manifestoes and programmes could be subject to the checks of a ‘Public Reason Committee’ before becoming public. Parties could be asked by the Committee to remove by a certain deadline any references to comprehensive doctrines in support of measures concerning constitutional essentials and matters of basic justice.

Failure to comply with the duty of civility ‘in due course’ would still result, of course, in some form of punishment. For example, guilty MPs or candidates (or their party) could be fined or compelled to remove the ‘illegal’ references to comprehensive doctrines (and the measures they are supposed to support) from the relevant official documents (e.g. draft legislation, party manifesto, etc.). In extreme cases, for example, when a party candidate makes continuous references to a religious doctrine in his/her speeches and fails more than once to provide public reasons ‘in due course’, he/she could be required to withdraw his/her candidature. Similarly, a party that systematically grounds its manifesto in a religious or other comprehensive doctrine and fails to make the necessary changes to it could be banned from the upcoming election and, if recidivist, from future elections too.

It should also be noted that for Rawls ‘the proviso is to be appropriately satisfied in good faith’. ³⁴ People's commitment to public reason, and their compliance with the duty of civility, should therefore be sincere. ³⁵ While some have contested the need for sincerity in public reason deliberation, ³⁶ Rawls clearly endorses this requirement. Here I assume without argument that, as Micah Schwartzman ³⁷ has cogently shown, there are very good reasons for claiming that ‘sincerity is an important condition of public deliberation, and that deliberation is necessary to evaluate, criticise and improve the quality of public justifications’. ³⁸ Political legitimacy, in other words, requires sincere commitment to public reason.

The sincerity requirement may raise serious practical concerns. It is a matter of fact that citizens (especially public officials and politicians) often defend their policy positions on the basis of public reasons that they do not sincerely believe to be sufficient for public justification. This is often due to political expediency, for example, to the aim of gaining popular support or building coalitions with political rivals. ³⁹ As well as contravening Rawls' sincerity requirement, and thus undermining public justification and political legitimacy, this also shows that it may be very difficult, perhaps impossible, to legally enforce and monitor people's compliance with the duty of civility. In the end, there is no way to check whether people are truly motivated by the public reasons that they may expediently invoke in public deliberation.

Yet believing sincerely that certain public reasons are sufficient to justify public policies does not imply that one should be motivated by them. ⁴⁰ In other words, and as it is well captured by Rawls' idea of an overlapping consensus, ‘[o]ne can be moved by a religious claim and believe that political action should follow from it only if it can be substantiated by sufficient public reason’, ⁴¹ that is, without actually being motivated by the latter. The main scope of the wide view of public reason is indeed to enable citizens to work out by themselves how their comprehensive views are related to political values. It does not matter, therefore, if we cannot monitor people's motivations because motivations are actually irrelevant to the sincerity of public reason.

Nevertheless, we are still left with the problem of establishing whether citizens who are often mainly motivated by non-public (e.g. religious, philosophical and ethical) reasons sincerely believe that the public reasons they invoke in public deliberation are sufficient for public justification. How can the state, for example, establish whether religious citizens who appeal to public reasons (whatever these might be) in order to reject abortion or the legal recognition of homosexual relationships sincerely believe that those reasons (i.e. rather than the religious reasons that deeply motivate them) are sufficient for public justification? It seems that we are facing here an insurmountable practical obstacle to the legal enforcement of the duty of civility.

However, as Eric MacGilvray ⁴² has persuasively shown, there is no reason to think that sincerity may not be the object of public scrutiny. More specifically, MacGilvray argues, we can ‘show either that the empirical premises upon which a given appeal rests are false or that those who are making the appeal are not consistent in accepting its broader implications’. ⁴³ In the former case, for example, we can claim that appealing to the welfare of children as a public reason for rejecting the legal recognition of homosexual relationships may not be legitimate if we could show ‘that children of homosexual couples were on average just as “well adjusted” as the children of heterosexual couples’. ⁴⁴ Empirical evidence, that is, can gradually limit the range of public reasons available to insincere people and thus unmask their insincerity. In the latter case, for example, people who reject affirmative action because ‘there should be no “special privileges” for any group’, ⁴⁵ that is, a potentially public reason, may reveal their insincere commitment to that reason if they then justify the unfair outcomes of the lack of affirmative action by appealing either to intrinsic ability differences between ethnic groups (a racist and therefore unreasonable claim) or to differences in training and education (which cannot be justified through public reasons). ⁴⁶

Rawls' wide view of public reason allows plenty of scope for the kind of public deliberation that can contribute in scrutinising and unveiling insincere beliefs in this way. Within this framework, there would be plenty of scope for public institutions aimed at performing this important function. As MacGilvray suggests, for example, we could devise a ‘federal model of public reason’ ⁴⁷ in which local deliberative bodies could ‘help to ensure that the claims made in the name of public reason are constrained within certain bounds: that people cannot get away with making unfounded claims about the dangers of same-sex marriage, or hypocritical claims about the injustice of affirmative action, for example’. ⁴⁸ This kind of deliberation is already possible, of course, in national fora such as parliaments or cabinet meetings, where the insincere beliefs of politicians and public officials are often unmasked during deliberations. Yet a more decentralised system would allow a much more pervasive scrutiny of citizens' (as opposed to just public officials’) appeal to public reasons, and would be in line with the widespread endorsement of forms of local-level deliberation among deliberative democrats. ⁴⁹

The analysis conducted in this section therefore suggests that issues of practical implementation are not insurmountable when discussing whether the duty of civility should be legally enforced. Rawls' wide view of public reason would leave legislators with much room for deciding how this enforcement should be carried out and for adopting tailored measures in order to avoid restrictions of free speech in those areas which for Rawls should fall outside the scope of public reason. Therefore, it seems that we should look elsewhere for the true reasons why Rawls almost axiomatically rejects any legal enforcement of the duty of civility. Rawls' unwillingness to legally enforce the duty of civility, that is, must be driven by deeper normative reasons. As Patrick Neal observes

Rawls' commitment to a robust notion of free speech simply means that he is acknowledging that that freedom will sometimes be abused but that it is not desirable to seek to legally prevent that abuse. Presumably, the good of free speech outweighs whatever benefits would be gained from (trying to) legally enforce the duty of civility’. ⁵⁰

Unfortunately, though, Rawls never provides a systematic theory of free speech in his works. Even though he includes freedom of speech among the basic liberties that ought to be protected and entrenched in the constitution of a ‘well-ordered’ society, ⁵¹ his references to freedom of speech are scattered both in A Theory of Justice and in Political Liberalism and his most detailed discussion of free speech, to which I will return later, concerns the issue of seditious libel. ⁵² My aim in the remainder of this paper is therefore to assess what arguments Rawls (or Rawlsians) could invoke in order to justify free speech and, based on this analysis, to explain why Rawls opposes the legal enforcement of the duty of civility.

Political liberalism, free speech and the argument from truth

The first theory of free speech that I wish to examine is the argument from truth, according to which freedom of speech allows the free exchange of opinions which is crucial for searching and discovering the truth. The most famous version of this argument is the one presented by John Stuart Mill in his essay On Liberty. ⁵³ Mill's arguments are well known and will not be discussed at length here. In summary, Mill argues that we should not silence an opinion because it may be true and silencing it may preserve error. This, Mill claims, would amount to a presumption of infallibility and would constitute a serious obstacle to intellectual development and to the pursuit of truth. Alternatively, if the received opinion (even if obviously true) is not confronted or challenged by other opinions, its meaning may be lost and it may become a dead dogma. Furthermore, the opinion one wants to silence may be partly true and only freedom of speech could allow the true part to emerge. Finally, Mill also claims that an unchallenged opinion is not able to promote action. ⁵⁴

I set aside, for the purpose of this paper, a discussion of the intrinsic merits and flaws of Mill's arguments, which have already been extensively analysed in the literature. ⁵⁵ What I would like to point out, instead, is that Mill's defence of freedom of speech is consequentialist in a twofold sense. First, free speech is functional to the searching and discovering of the truth. Second, and most importantly, searching and discovering the truth through the free exchange of opinions is for Mill not valuable in itself but because it contributes ‘to the mental well-being of mankind (on which all their other well-being depends)’. ⁵⁶ ‘The truth of an opinion’ Mill argues, ‘is part of its utility’ ⁵⁷ and for Mill utility should be ‘the ultimate appeal on all ethical questions; but it must be utility in the largest sense, grounded on the permanent interests of man as a progressive being’. ⁵⁸

It is clear from these remarks that Mill's account of freedom of speech is not unrelated to his overall defence of liberty broadly intended ⁵⁹ but is instead deeply rooted in it. This has important implications in relation to Rawls' political liberalism. If the argument from truth, that is, relies on a comprehensive doctrine such as Mill's liberalism, it could not be invoked by Rawls to justify freedom of speech as a constitutional essential. Indeed Rawls himself often remarks that his political liberalism significantly differs from Mill's comprehensive liberalism and does not endorse, for example, Mill's comprehensive views regarding individuality and autonomy. ⁶⁰ Truth-seeking cannot therefore be invoked to provide a public justification for freedom of speech as a constitutional essential in polities characterised by the fact of reasonable pluralism. Doing so would violate the ‘liberal principle of legitimacy’ ⁶¹ which establishes that political power is legitimate and justified ‘only when it is exercised in accordance with a constitution the essentials of which all citizens as free and equal may reasonably be expected to endorse in the light of principles and ideals acceptable to their common human reason’. ⁶² Furthermore, Rawls himself seems to deny that a freestanding political conception of truth is possible. According to him

political constructivism specifies an idea of the reasonable … [but] … does not … use (or deny) the concept of truth; nor does it question that concept, nor could it say that the concept of truth and its idea of the reasonable are the same. Rather, within itself the political conception does without the concept of truth. ⁶³

This conclusion may not be sufficient, however, for dismissing the argument from truth as a justification for free speech in Rawls' political liberalism. If we untie the searching and discovery of the truth from Mill's emphasis on individual well-being and progress (i.e. ‘utility in the largest sense’), we may be left with a freestanding political justification for freedom of speech. As Frederick Schauer highlights, ‘the advantages of truth are almost universally accepted’. ⁶⁴ They are, to put it in Rawls' terms, ‘presently accepted general beliefs and forms of reasoning found in common sense’, ⁶⁵ or ‘plain truths now widely accepted, or available, to citizens generally’. ⁶⁶ They can therefore be included among those political values which solely, in Rawls' view of public reason, can provide a public justification for the constitutional essentials, including freedom of speech.

It is true, of course, that many people base their lives on ‘false beliefs, say, that sacrificing animals to the Water Gods will prevent hurricanes, or that Creation Science is as adequate theory as Darwinism, or that the Holocaust never occurred’. ⁶⁷ One might also argue that for these people life can only be valuable if the beliefs they hold remain unchallenged. Yet it is plausible to argue that for these people such beliefs are true, even if for (most of) us they are not. This suggests that there is a reasonable overlapping consensus concerning the importance of holding true beliefs, and that very few people (if any) would expressly desire to hold false beliefs.

Joshua Cohen, ⁶⁸ in this connection, provides a persuasive account of how a conception of truth may indeed be political and therefore consistent with (rather than inimical to) the idea of public reason. According to Cohen

a political conception of truth … [involves] … a set of claims about truth – for example, that truth is distinct from warrant, and that it is important – that is suited for the purposes of political reflection and argument in a pluralistic democracy, characterised by doctrinal disagreements. ⁶⁹

Public reasoning, Cohen argues, does not require (and indeed demands renouncing) ‘concepts of salvation, or self-realisation, soul, personal autonomy, purity, courage or honor’ ⁷⁰ but does require a ‘concept of truth’, ⁷¹ in the same way in which it requires ‘the concept of an object, cause, thought, reason or inference or evidence’. ⁷² ‘Truth’, Cohen claims, ‘is so closely connected with intuitive notions of thinking, asserting, believing, judging and reasoning that it is difficult to understand what leaving it behind amounts to’. ⁷³

Two specific aspects of Cohen's account of the political conception of truth are especially relevant to my present analysis. First, Cohen argues, ‘[b]elieving (asserting, judging) is believing (asserting, judging) true, where this slogan is understood to mean that truth is the norm governing beliefs, assertions and judgments’. ⁷⁴ Second, he claims

[t]rue beliefs present things as they are (they “say of what is that it is and of what is not that it is not”), and in that uncontroversial sense correspond to how things are, although it [i.e. the political conception of truth] will not add (or deny) that such beliefs present things as they really are in themselves, determinately and mind-independently. ⁷⁵

These two points suggest that even those people who hold beliefs that are widely considered false can subscribe to this political conception of truth.

Two important considerations follow from Cohen's account. First, and as the passages just quoted show, in order to be the object of a reasonable overlapping consensus a political conception of truth must necessarily be ‘thin’. It must be, to use Cohen's words, a ‘concept of truth, not a theory of truth’, ⁷⁶ as the latter would raise ‘philosophical controversy’, ⁷⁷ the very thing political liberalism aims to eschew. ⁷⁸ Second, and as a consequence, this reasonable overlapping consensus on a thin concept of truth can (and normally is) be accompanied by reasonable disagreement regarding what the necessary conditions for holding true beliefs actually are. In other words, we can agree on a concept of truth (such as Cohen's) while disagreeing on theories of truth, including (most importantly for the present analysis) theories that consider freedom of speech a necessary condition for holding true beliefs.

One might still point out that those who hold certain beliefs (e.g. that the Holocaust never happened), despite clear and opposing evidence, can hardly be considered reasonable. Maintaining such beliefs in the face of contrary empirical and historical evidence seems irrational and it may be difficult to understand how irrational people could be reasonable in the Rawlsian sense. Indeed, even though Rawls makes a clear distinction between the ‘rational’ and the ‘reasonable’, ⁷⁹ Gerald Gaus points out that ‘the reasonable is much more closely bound to the rational than Rawls and other political liberals would have us believe … [and] …  the rational is far more basic to political justification than political liberals suppose’. ⁸⁰ There seem to be good grounds, therefore, for arguing that political liberalism demands standards for belief or ‘epistemic rationality’ ⁸¹ on the basis of which certain irrational beliefs could be considered unreasonable. Yet, as Gaus himself argues, ‘we need not commit ourselves to a notion of truth in order to explicate epistemic rationality’. ⁸² While epistemic rationality may require thicker standards for belief than those entailed by Cohen's political concept of truth, it still leaves scope for reasonable disagreement regarding theories of truth, including theories that consider free speech a necessary condition for holding true beliefs (and, indeed, Gaus himself does not make any references to free speech in his account).

The upshot of this analysis is therefore that while political accounts of truth and of epistemic rationality (such as those defended by Cohen and Gaus) may be consistent with political liberalism, they are irrelevant to a normative defence of free speech. For such a defence to be grounded in the idea of truth, a theory of truth (i.e. rather than a mere concept of truth, or mere standards of epistemic rationality) is required, but we have seen that any such theory would be the object of reasonable disagreement. Cohen's and Gaus's accounts justify at most some of the rules of inquiry that should guide the use of free speech in public reasoning, but do not explain how free speech can be justified on Rawlsian grounds in the first instance, and what the connection between free speech and public reason is in Rawls' account. We therefore need to look elsewhere for a justification for freedom of speech as a constitutional essential in political liberalism.

Political liberalism, free speech and the argument from self-fulfilment

The second argument I wish to examine grounds the protection of free speech in the idea of individual self-fulfilment. The claim, in this case, is that ‘[r]estrictions on what we are allowed to say and write, or (on some formulations of the theory) to hear and read, inhibit our personality and its growth. A right to express beliefs and political attitudes instantiates or reflects what it is to be human’. ⁸³ The argument from self-fulfilment, at least in some if its versions, relies on an Aristotelian conception of human nature and on the view that

[t]he right to freedom of expression is justified first of all as the right of an individual purely in his capacity as an individual. It derives from the widely accepted premise of Western thought that the proper end of man is the realisation of his character and potentialities as a human being. ⁸⁴

Such potentialities include the ability to express one's views, which is ‘an integral part of the development of ideas, of mental exploration and of the affirmation of self’. ⁸⁵ It is clear that, as in the case of Millian liberalism, a view for which human beings should constantly seek progress and intellectual growth through the free expression of beliefs and opinions is inevitably a comprehensive (and therefore controversial) view. It is not shared by those who ‘are likely to be satisfied with other freedoms, or prefer the security or intellectual anaesthesia that accompanies rigid controls on expression’ ⁸⁶ and therefore cannot be invoked as a freestanding public reason to justify constitutional essentials such as freedom of speech.

In a more refined version of the argument from self-fulfilment, Thomas M. Scanlon claims that ‘the powers of a state are limited to those that citizens could recognise while still regarding themselves as equal, autonomous, rational agents’. ⁸⁷ For Scanlon, ‘[a]n autonomous person cannot accept without independent consideration the judgment of others as to what he should believe or what he should do’. ⁸⁸ Freedom of speech is therefore essential to guaranteeing that individual autonomous choices are fully informed and well thought-through. This implies that speech should not be restricted based on the fact that some acts of expression produce certain kinds of harms, for example

(a) harm to certain individuals which consist in their coming to have false beliefs as a result of those acts of expression and (b) harmful consequences of acts performed as a result of those acts of expression, where … the act of expression led the agents to believe (or increased their tendency to believe) these acts to be worth performing. ⁸⁹

Like other versions of the self-fulfilment theory of free speech, Scanlon's argument relies on a comprehensive conception of individual autonomy which cannot be invoked in order to justify the protection of freedom of speech within Rawls' political liberalism. Indeed, Scanlon himself identifies his theory with the ‘Millian principle’, ⁹⁰ as it relies on a Millian understanding of individual autonomy.

Political liberalism, free speech and the two moral powers

There is, however, a slightly different formulation of Scanlon's theory of freedom of speech which seems to provide a more suitable rationale for a Rawlsian account of free speech. In his essay on ‘Rawls' Theory of Justice’, ⁹¹ Scanlon argues that freedom of speech is justified by the fact that the parties in the Original Position would not grant governments ‘the authority to decide matters of moral, religious or philosophic doctrine (or of scientific truth) and … to restrict certain activities on the grounds that they promulgate false or corrupting doctrines’. ⁹² Scanlon calls this ‘the Principle of Limited Authority’. ⁹³ Similarly, in a more recent essay (which, however, does not directly discuss Rawls' work), Scanlon argues that ‘freedom of expression … seeks to protect … our interest in deciding for ourselves how to conduct our private lives’. ⁹⁴

Scanlon's claim bears a clear link with Rawls' political liberalism. More specifically, Rawls ascribes ‘two moral powers’ ⁹⁵ to persons, that is, ‘a capacity for a sense of justice and for a conception of the good’. ⁹⁶ The latter ‘is the capacity to form, to revise and rationally to pursue a conception of one's rational advantage or good’. ⁹⁷ Freedom of expression is clearly necessary for citizens to be able to cultivate this capacity. While the state, in political liberalism, should not embrace any conception of the good, it should provide basic rights and liberties (including freedom of speech) in order to allow citizens to pursue their conceptions of the good (e.g. religious, philosophical, ethical, etc.) in their private lives. Not doing so would encroach upon their liberty of conscience which, for Rawls, ‘is among the social conditions necessary for the development and exercise of this power’. ⁹⁸ If this line of argument justifies the protection of free speech in Rawls' political liberalism, it may also provide the key for understanding why Rawls opposes the legal enforcement of the duty of civility.

Yet this conclusion may be too hurried. For Rawls, we have seen, the constraints of public reason only apply to the ‘public political forum’ ⁹⁹ and not to the ‘background culture’. ¹⁰⁰ Moreover, in ‘The Idea of Public Reason Revisited’ ¹⁰¹ Rawls introduces the ‘non-public political culture [which] mediates between the public political culture and the background culture. This comprises media … of all kinds: newspapers, reviews and magazines, television and radio, and much else’. ¹⁰² The constraints of public reason do not apply to this realm either. It seems therefore that a legal enforcement of the duty of civility in the public political realm would not significantly prevent citizens from cultivating their capacity for a conception of the good. The free flow of information that citizens could access in the background culture and the non-public political culture would be sufficient for them ‘to form, to revise and rationally to pursue a conception of one's rational advantage or good’. ¹⁰³ While providing a plausible justification for a Rawlsian defence of free speech, therefore, this argument still fails to explain why Rawls so strongly opposes the legal enforcement of the duty of civility.

However, a better argument can be found if we turn to the other moral power Rawls ascribes to individuals, that is, the capacity to develop and exercise their sense of justice. This moral power, Rawls argues, is ‘the capacity to understand, to apply and to act from the public conception of justice which characterises the fair terms of social co-operation’. ¹⁰⁴ Cultivating this capacity requires people to work out for themselves whether some argument is properly public or political. This involves not only personal reflection but also deliberation with others and, most importantly, deliberation conducted in the public political realm, rather than solely in the background culture and the non-public political realm. The legal implementation of the duty of civility would seriously undermine this process. Censoring freedom of speech in the public political realm (the only realm, it should be remembered, where that duty applies) would prevent individuals from employing their judgment in order to assess what arguments are properly public and political and therefore would hinder the development and exercise of their capacity for a sense of justice.

This is hardly problematic, however, for Rawls' account. Indeed the ‘wide’ conception of public reason just aims to guarantee this kind of deliberation. By allowing individuals to appeal to comprehensive doctrines, it enables them to participate in this deliberation without having to abide by the duty of civility. Even if the latter was enforced by law, therefore, this deliberation would still be exempted from censorship. This especially reflects Rawls' endorsement of a ‘flexible’ conception of public reason and his view that the latter should not be fixed ‘once and for all in the form of one favoured political conception of justice’. ¹⁰⁵ This is crucial, Rawls argues, if we want to avoid that ‘the claims of groups or interests arising from social change might be repressed and fail to gain their appropriate political voice’. ¹⁰⁶ Take, for example, environmentalist concepts such as ‘sustainability’ and ‘carbon footprint’. ¹⁰⁷ Fifty or sixty years ago, such concepts would have probably been considered non-political and expressions of a comprehensive doctrine. However, thanks to decades of deliberation, during which individuals and groups have often appealed to their comprehensive doctrines (e.g. religious, ethical, etc.), people in western liberal democracies have increasingly acknowledged that these concepts are political, i.e. they are public reasons. They have become, that is, ‘presently accepted general beliefs’ ¹⁰⁸ and ‘plain truths now widely accepted, or available, to citizens generally’. ¹⁰⁹

However, a further problem may arise at this point. Rawls' flexible conception of public reason, we have seen, implies that there is scope for some arguments previously (or currently) considered non-public and non-political (e.g. the abovementioned environmental ideas) to be increasingly acknowledged as public and political after a (sometimes lengthy) process of deliberation. Moreover, the opposite is also true, i.e. some reasons that were previously considered public may gradually become more controversial. The idea, mentioned earlier, that the legal recognition of homosexual relationships can reasonably be rejected in order to protect the welfare of children belongs to this category and so do many other allegedly public reasons whose empirical foundations have gradually been shown to be unsound. This, one might argue, implies that violations of the duty of civility will often be a matter of degree, and will often be contested. Consequently, this will create another ‘in principle’ barrier to the legal enforcement of the duty of civility, analogous to those that I examined in the first section.

Yet this need not be the case. The fact that public reason is flexible and changes over time, and that violations of the duty of civility will often be contested, does not imply that we cannot establish with a good degree of confidence that some (indeed, many) arguments and concepts (e.g. salvation, reincarnation, Kantian or Millian conceptions of personal autonomy, etc.) are definitely non-public and non-political reasons, given the fact of reasonable pluralism. ¹¹⁰ While these arguments and concepts can, of course, be invoked during the unconstrained public deliberation that wide public reasons allows, they should still be relinquished when, in due course, public reasons should be invoked instead in order to support legislation concerning fundamental matters. The fact that the content of public reason changes over time, therefore, only implies that some of the items in the ‘list’ of non-public reasons that should not be invoked in public deliberation will also constantly change, due to new additions and subtractions. It does not imply, however, that the bulk of that list will not remain stable and uncontested over time. ¹¹¹

Rawls' account of public reason therefore already allows the kind of deliberation which enables individuals to develop and exercise their two moral powers. On the one hand, it allows unconstrained debate in the background culture and the non-public political sphere, thanks to which people can cultivate their capacity for a conception of the good. On the other hand, through the wide and flexible conception of public reason, it allows the kind of deliberation that helps citizens to exercise and develop their capacity for a sense of justice. The legal enforcement of the duty of civility would not involve any kind of censorship at these levels and would therefore not hinder individuals’ exercise of their moral powers. Yet, as we have seen, it would indeed entail censorship when, in due course, citizens would be expected by law to relinquish whatever comprehensive arguments or doctrines are deemed non-political and non-public at that point in their society. It still needs to be explained, therefore, why Rawls opposes the legal enforcement of the duty of civility at that stage.

A potential response, still based on Rawls' political conception of the person and, especially, of the first moral power (the capacity for a sense of justice) could be the following. The exercise and development of that capacity requires that citizens comply with the duty of civility sincerely and ‘in good faith’. ¹¹² We have already seen that sincerity is indeed important in public reasoning. This implies that the legal enforcement of the duty of civility might seriously undermine the sincere fulfilment of that duty. If citizens, that is, were under a strong pressure to comply with the duty of civility in due course when advocating legislation about fundamental matters in the public political realm, due to fear of legal sanctions, they might often be inclined to misrepresent or lie about their true beliefs, in order to avoid those sanctions. ¹¹³ For example, if an MP had one week to provide public reasons to support a policy she proposed in the parliament, in order to avoid a fine or suspension, she would probably be more concerned with finding any suitable public reasons to justify that policy (regardless of whether she sincerely believed them) than with developing a sincere commitment to them. The fear of incurring legal penalties, as well as the pressure of a specific deadline, might therefore impose upon her a strong pressure to comply with the duty of civility for merely pragmatic (rather than moral) reasons. This might, as a consequence, prevent her and those in a similar situation from developing their capacity for a sense of justice, that is, the ability to work out for themselves whether certain arguments and claims are suitably public and to articulate their demands about fundamental legislation in terms of public reason.

Yet this conclusion may be challenged. The conversational constraints resulting from the legal enforcement of the duty of civility might in fact support people's sincere commitment to public reason. For example, ‘[i]f a racist argues for a piece of legislation but is forced to defend his claim in terms of public reason, he may over time come to believe the arguments he offers in public’. ¹¹⁴ As Jon Elster points out, ‘this civilising force of hypocrisy is a desirable effect of publicity’. ¹¹⁵ This, one might add, could also be a welcome effect of the public deliberative fora which, as we have already seen, could be devised in order to expose insincere appeals to public reasons. Moreover, even if censorship did not always reinforce people's sincere commitment to public reason, it is plausible to argue that it would at least not preclude that commitment, in the same way in which the existence of laws against murder and theft does not prevent people from sincerely condemning and refusing to commit murder and theft on moral grounds. It still remains to be shown, therefore, why Rawls rejects the legal enforcement of the duty of civility. This is what the next section will aim to accomplish.

Free speech, public reason and political legitimacy

Another major theory of free speech, alongside the arguments from truth and from self-fulfilment, is the argument from democracy. The theory, the most famous version of which was formulated by Alexander Meiklejohn, ¹¹⁶ states that freedom of speech is conducive to citizens' understanding of political matters and that this is essential for their participation in democratic politics and for their self-government. According to Meiklejohn

[t]he First Amendment does not protect a “freedom to speak”. It protects the freedom of those activities of thought and communication by which we “govern”. It is concerned, not with a private right, but with a public power, a governmental responsibility. ¹¹⁷

‘Self-government’, Meiklejohn argues, ‘can exist only insofar as the voters acquire the intelligence, integrity, sensitivity and generous devotion to the general welfare that, in theory, casting a ballot is assumed to express’. ¹¹⁸ Freedom of speech is crucial for this process to be effective.

There are several reasons why the argument from democracy is especially suitable for justifying the protection of free speech in political liberalism. First, the argument has a narrower scope than the arguments from truth and self-fulfilment. This is because it ‘requires for its deployment the a priori acceptance of democratic principles as proper guidelines for the organisation and governance of the state’. ¹¹⁹ In this sense, it is in line with Rawls' political liberalism which expressly relies on ‘certain fundamental ideas seen as implicit in the public political culture of a democratic society’ ¹²⁰ and does not aim to apply to non-democratic polities. Second, unlike the argument from truth and most versions of the argument from self-fulfilment, the argument from democracy does not rely on a comprehensive conception of the good or of human nature. Third, the argument from democracy displays a special focus on political matters and mainly concerns ‘speech relating to public affairs, and even more … criticism of governmental officials and policies’. ¹²¹ This is in line with Rawls' own (brief) account of free speech which mainly concerns ‘free political speech’ ¹²² rather than ‘free speech’ broadly intended.

What does Rawls have to say, then, regarding free political speech? His central argument is that ‘to allow the crime of seditious libel would undermine the wider possibilities of self-government and the several liberties required for its protection’. ¹²³ In a democratic polity, he claims, ‘there is no such thing as the crime of seditious libel; there are no prior restraints on freedom of the press, except for special cases; and the advocacy of revolutionary and subversive doctrines is fully protected’. ¹²⁴ Rawls also recognises and endorses the fact that ‘[w]ithin our [i.e. the United States’] tradition there has been a consensus that the discussion of general political, religious and philosophical doctrines can never be censored’. ¹²⁵

Rawls also claims that ‘subversive advocacy is always part of a more comprehensive political view’ ¹²⁶ and that ‘[a]s Kalven observes, revolutionaries don't simply shout: “Revolt! Revolt!” They give reasons. To repress subversive advocacy is to suppress the discussion of these reasons, and to do this is to restrict the free and informed public use of our reason in judging the justice of the basic structure and its social policies. And thus the basic liberty of freedom of thought is violated’. ¹²⁷ Even more clearly, Rawls argues that

a wise political leadership [in a moderately well-governed democratic society] takes this advocacy [i.e. subversive advocacy] as a warning that fundamental changes may be necessary; and what changes are required is known in part from the more comprehensive political view used to explain and justify the advocacy of resistance and revolution. ¹²⁸

Rawls argues that the only kind of political speech which should remain unprotected is ‘subversive advocacy when it is both directed to inciting imminent and unlawful use of force and likely to achieve this result’. ¹²⁹

Rawls' account shows that he is clearly committed to a justification for free speech grounded in the idea of democratic self-government. Moreover, we have seen, Rawls explicitly states that appeals to comprehensive doctrines in order to advocate changes in the constitutional essentials should be protected by freedom of speech in a democratic polity. This suggests that the legal enforcement of the duty of civility would strongly conflict with the democratic rationale for free political speech and therefore seriously undermine citizens' self-government. Public reason, we have seen, does apply to the public political forum and it is in the latter that subversive advocacy and seditious libel are often voiced, by appealing to comprehensive doctrines and in order to invoke changes in the constitutional essentials and the basic structure of a polity.

The argument from democracy therefore provides a further rationale for freedom of speech in Rawls' account, in addition to the argument from the two moral powers illustrated in the previous section. Yet, unlike that argument, it also explains why Rawls opposes the legal enforcement of the duty of civility. To understand this more clearly, we should observe that the legal freedom of political speech and the moral duty of civility are in fact two sides of the same coin. Freedom of speech, indeed, contributes not only to citizens' democratic participation and self-government but also, most importantly, to the legitimacy of political decision-making and of the legal measures that result from it. It is, in other words, ‘a sine qua non for political legitimacy in a democracy, not just for the quality of democratic engagement’. ¹³⁰ This implies that the purpose of both the right to free speech and the duty of civility is the same, that is, to ensure the legitimacy of coercive legislation (in the case of public reason, only legislation concerning fundamental matters). In order to guarantee this legitimacy, for Rawls we ought to have the right (legal right) to appeal to our non-public reasons in political deliberation but we ought to refrain (moral duty) from using that right when (‘in due course’) we endorse legislation concerning constitutional essentials and matters of basic justice. In other words, we should have ‘a right to do wrong’. ¹³¹ Legally enforcing the duty of civility would paradoxically undermine its purpose, that is, granting legitimacy to legislation concerning constitutional essentials and matters of basic justice.

Yet this conclusion may be open to two serious challenges. First, one might point out that even in the absence of legal constraints, free speech might be hindered by social censorship ¹³² and this might in fact undermine democratic legitimacy. John Stuart Mill, for example, famously highlights how

[s]ociety … practices a social tyranny more formidable than many kinds of political oppression, since, though not usually upheld by such extreme penalties, it leaves fewer means of escape, penetrating much more deeply into the details of life, and enslaving the soul itself. ¹³³

According to Mill

social stigma … is really effective, and so effective is it, that the profession of opinions which are under the ban of society is much less common in England [i.e. during the Victorian era], than is, in many other countries, the avowal of those which incur risk of judicial punishment. ¹³⁴

Social pressure and stigma can significantly restrain citizens' freedom of speech, even when they are not accompanied by legal penalties. The pervasive compliance with the duty of civility that Rawls advocates as a normative ideal for a liberal democratic polity might lead to a widespread moral censorship against those who are in breach of that duty. Rawls himself, for example, argues that

[w]hen firm and widespread, the disposition of citizens to view themselves as ideal legislators, and to repudiate government officials and candidates for public office who violate public reason, is one of the political and social roots of democracy, and is vital to its enduring strength and vigor. ¹³⁵

This suggests that for Rawls, in a similar way, citizens ought to develop a pervasive disposition to morally condemn all those fellow citizens (i.e. not only government officials and candidates) who contravene the duty of civility. This might therefore produce a social climate in which those citizens who wish to contravene the moral duty of civility would formally maintain their freedom of speech while in fact gradually becoming subject to the moral disapproval and social control of the ‘duty-of-civility-abiding’ majority. ¹³⁶ As Candace Cummins Gauthier points out, instances of social control ‘are essential ways in which communities may affect personal change in their members toward behaviour that is more in concert with communal values and ends’. ¹³⁷

Due to the strong constraints that social censorship may impose upon freedom of speech, Rawls' endeavour to endorse both the latter and the moral duty of civility appears problematic. On the one hand, the widespread compliance with the duty of civility that Rawls invokes may lead to a pervasive social censorship which, even in the absence of any legal constraints, may hinder freedom of speech and, consequently, the democratic legitimacy of political decision-making concerning constitutional essentials and matters of basic justice. On the other hand, renouncing the duty of civility would undermine the crucial role the latter plays in guaranteeing the legitimacy of coercive political legislation concerning those fundamental matters in political liberalism.

Yet this conclusion may be too gloomy. It is far from clear that the widespread compliance with the duty of civility, and readiness to condemn those who infringe it, would result in a kind of social censorship as oppressive as legal censorship. The wide and flexible view of public reason, we have seen, would allow a significant degree of freedom of speech even in the presence of a legally enforced duty of civility. The duty of civility, we should remember, only applies ‘in due course’, that is, after a deliberative process during which citizens have had the opportunity to articulate their comprehensive doctrines in the public political realm. For the same reason, therefore, the social enforcement of the duty of civility (i.e. even in the absence of legal constraints) would not affect that deliberation. Rawls, that is, does not expect citizens to morally condemn those who invoke comprehensive reasons during that deliberative process.

Furthermore, the indeterminacy of Rawls' ‘proviso’ implies that the social expectation of compliance with the duty of civility would only (or, at least, especially) be pressing for government officials and candidates for public office. ¹³⁸ This is due to a number of reasons. First, these individuals are more likely than ordinary citizens to be exposed to public scrutiny and, even in the absence of legal rules, they may be expected by most people to comply with the duty of civility. Second, and as a consequence, government officials and candidates for public office might have more pressing practical reasons (i.e. than ordinary citizens) for complying with certain social expectations (including expectations regarding the form and content of their speech), especially if they seek (re-)election. ¹³⁹ Finally, it is difficult to understand what kinds of expectations people might have about each other qua ordinary citizens, regarding compliance with the duty of civility, especially if they are not engaged in formal deliberation within specific institutional bodies (e.g. parliament, constitutional court, etc.) or are not directly involved in political campaigns. The only real requirement, for most of them, would be to justify through public reasons how they vote on fundamental issues. However, as long as voting is secret, it is not clear how social censorship could impose a burden upon them. It is therefore doubtful that the pervasive compliance with the duty of civility would result in any significant kind of social censorship.

Even if that was the case (at least for government officials and candidates), however, that would not have significant implications for political legitimacy. The latter concerns the relationship between citizens and the state, not among citizens themselves. According to Ronald Dworkin, for example

it is illegitimate for governments to impose a collective or official decision on dissenting individuals, using the coercive powers of the state, unless that decision has been taken in a manner that respects each individual's status as a free and equal member of the community. ¹⁴⁰

That respect, Dworkin argues, involves the right ‘to raise a voice in protest or argument or objection before the decision is taken’. ¹⁴¹ It is the state that, in this account, has a duty to allow everyone to express their views (including, for Dworkin, racist and sexist views) in order to claim the legitimacy of its laws and demand citizens' political obligation. The fact that society, or part of it, may exercise some form of social censorship towards some citizens may be unfortunate (from this perspective) but is irrelevant to the issue of political legitimacy.

In fact social condemnation of other people's speech, especially of the speech of those who are government officials or occupy other important institutional roles, is in itself an important aspect of freedom of speech. Take, for example, the way racist or fascist political speech often encounters social disapproval in liberal democracies, even in the absence of legislation that forbids it. This kind of condemnation constitutes an essential aspect of freedom of speech. One could, of course, invoke the legal censorship of certain kinds of hate speech (e.g. fascist, racist, sexist, etc.), and even accept some loss of democratic legitimacy, in order to protect important values such as public order or the dignity of certain citizens (e.g. members of ethnic and religious minorities). ¹⁴² However, it would be incoherent for someone to defend the freedom of expression of fascists or racists and then be concerned about the hostile reaction this may cause within society. This reaction, and the effects it may have on racists and fascists, is in fact what defenders of an unconstrained freedom of speech normally want to encourage.

Applied to Rawls' account, this analysis implies that we should not be too worried about the social enforcement of the duty of civility. It is far from clear that, in spite of Rawls' recommendations, social censorship of those who breach the duty of civility would be pervasive. Even if it was, it would not undermine political legitimacy, which only concerns the relationship between individuals and the state. Furthermore, it would just be part of what freedom of speech inherently involves for those who (like Rawls) refuse to impose any constraints upon it, that is, the clash of ideas and the reciprocal contestation among citizens within the democratic political realm. As Stephen Macedo rightly points out, ‘[i]f some people … feel “silenced” or “marginalized” by the fact that some of us believe that it is wrong to seek to shape basic liberties on the basis of religious or metaphysical claims, I can only say “grow up!”’. ¹⁴³

The second challenge to Rawls' defence of both (legal) unconstrained freedom of speech and (moral) duty of civility is, however, more serious. One might point out, that is, that if political liberals prioritise freedom of speech over the legal enforcement of the duty of civility (as Rawls' own account suggests), that will inevitably lead to a loss of political legitimacy. It is a widely endorsed view, for example, that if public officials adopt a law based on false pretences, that will tend to delegitimise that law in the eyes of the public, even though few people think that those public officials should be punished through legal (as opposed to electoral) means. Similarly, if public officials appeal to non-public reasons to justify a policy (because freedom of speech allows them to do so), that will reduce the legitimacy of that policy. This is a serious concern. It seems that, in the end, Rawls' defence of unconstrained free speech is self-defeating. Given the importance that public reason and the duty of civility play in political liberalism, Rawls' commitment to unconstrained freedom of speech risks undermining the very thing that political liberalism aims to guarantee, that is, political legitimacy.

The question, therefore, is whether even though Rawls does endorse unconstrained freedom of speech, he actually has to do so. While public reason and its constraints, that is, are clearly central to political liberalism, it is unclear whether and why unconstrained free speech should also be. We have seen that, in fact, political liberalism already allows significant scope for freedom of speech, especially on the basis of the wide view of public reason, and that this is sufficient for citizens to develop and cultivate their two moral powers. It is not clear why this freedom should be extended further in order to legally allow citizens to violate the duty of civility when, in due course, they have the moral duty to relinquish non-public reasons and only appeal to public ones. We have also seen that there are not insurmountable practical problems to the legal enforcement of the duty of civility and that even sincere commitment to that duty would not be undermined by legal enforcement. It is therefore unclear why Rawls needs to endorse unconstrained freedom of speech, given his commitment to public reason.

It can therefore be concluded that Rawls and Rawlsians could remain faithful to political liberalism without having to commit to unconstrained freedom of speech. Indeed, as Jeremy Waldron ¹⁴⁴ has recently highlighted, some aspects of Rawls' political liberalism may actually be invoked in order to justify restrictions of free speech (more specifically, of hate speech). Waldron points out that the protection of ‘the assurance of a general commitment to the fundamentals of justice and dignity that [for Rawls] a well-ordered society is supposed to furnish to its citizens' ¹⁴⁵ would justify restricting ‘public and semi-permanent manifestations of racial and ethnic hatred as visible aspects of the civic environment’. ¹⁴⁶ Yet Waldron also acknowledges that his conclusion, while grounded in Rawls' ideas, would probably be rejected by Rawls himself, and that Rawls would in all likelihood endorse ‘First Amendment orthodoxy’, ¹⁴⁷ that is, unconstrained freedom of speech.

Rather than landing a fatal blow to Rawls' political liberalism, however, Waldron's claim, like the analysis conducted in the present paper, just shows that political liberalism simply does not require commitment to unconstrained freedom of speech. There are other values and goals, such as the ‘assurance’ highlighted by Waldron, and the political legitimacy which public reason and the duty of civility aim to ensure, which can be given priority over unconstrained free speech consistently with political liberalism, and in spite of what Rawls actually says about free political speech. It is sometimes possible, perhaps necessary, to be more Rawlsian than Rawls if one is fully committed to political liberalism.

Conclusion

Rawls and Rawlsians often respond to those who claim that the duty of civility is unduly restrictive and exclusionary by stressing that the duty is only moral, not legal. Neither Rawls nor Rawlsians, however, provide substantial arguments to justify why this is the case, apart from the uninformative (and somewhat tautological) claim that a legally enforced duty of civility would undermine freedom of speech. In this paper, I have shown that Rawls' account does indeed contain arguments for rejecting the legal enforcement of the duty of civility, even if Rawls seldom explicitly discusses them. Furthermore, I have claimed, these arguments rely on a Rawlsian account of free speech and especially on the idea that free speech contributes to political legitimacy, the very goal that the duty of civility also aims to achieve. Yet I have concluded that while Rawls does endorse unconstrained freedom of speech, the latter is not essential to political liberalism or, at least, not as essential to it as public reason and the duty of civility are. Rawls and Rawlsians could therefore renounce unconstrained freedom of speech while remaining committed to political liberalism. This would, of course, prevent them from stressing the purely moral nature of the duty of civility when challenged by those who consider that duty unduly restrictive. Yet, if they are truly committed to the duty of civility, Rawls and political liberals should be ready to defend it on the basis of its rationale, rather than by unconvincingly stating that the duty is moral and not legal.