1

Freedom of speech is a basic right in virtually all liberal constitutions and an ‘institutional condition’ for a polity to qualify as democratic (PL 309).Footnote ¹ It is likewise entrenched in international charters such as the Universal Declaration of Human Rights and the International Covenant on Civil and Political Rights.Footnote ² In the works of Kant and Rawls, we find an acute sensibility for the pre-eminent importance of freedom of speech, indeed both authors can be placed among its great philosophical champions. Both are aware that claims to free speech draw on a variety of normative sources, and that they can serve individual as well as political goals. For both authors, free speech is a characteristic and necessary liberal as well as democratic institution, but their conceptions differ markedly when applied to non-liberal and non-democratic societies.

This basic difference between Kant and Rawls comes to the fore in Rawls’s treatment of non-liberal, hierarchical societies. The difference is that freedom of speech, for Kant, is a universal claim that can serve as a test of legitimacy (Rechtmäßigkeit) of all legal orders, while for Rawls, some legal orders are owed full recognition even if they do not in principle guarantee freedom of speech. I want to discuss this difference in four steps. In the first two steps, I sketch some main features of free personal and political speech in Kant (section 2) and Rawls (section 3). I argue that a much-overlooked shift in Rawls’s development ties his account of free speech to issues concerning justice. In the next step, I discuss Rawls’s perspective on some non-democratic regimes in his Law of Peoples, regimes that he understands as well-ordered but which do not guarantee freedom of speech. I criticize Rawls’s account from Kant’s perspective (section 4). In a fourth step, I introduce a ‘module’ from Kant’s pre-republican thought into Rawls’s conception, aiming to secure a core area of justice-related speech, and respond to some initial objections (section 5). In contrast to earlier critiques of Rawls, which lambasted his lack of concern for freedom of speech in non-Western societies as insufficiently liberal, my remarks are offered in a constructive spirit. I suggest that under Kant’s premises for autocratic legitimacy, an important extension of speech rights is called for even in non-liberal, non-democratic states, and that a Rawlsian account should and can adopt it.

2

Most conceptions of free speech fall neatly into one of two categories, depending on whether they foreground individual entitlement or political functionality. Roughly speaking, liberal-individualist positions interpret free speech as protecting a realm of sovereign idiosyncrasy against social and political intervention, while republican positions interpret free speech as serving citizens’ informed participation in political will-formation. This immediately entails differences of scope and of extension between the two types of conceptions.Footnote ³ While republican positions only protect politically salient speech, liberal-individualist views protect all kinds of expression in order to guarantee, as put in a beautiful formulation by Rudolf Smend, ‘a sphere of morally necessary breathing space for the individual’.Footnote ⁴ While republican views award speech rights to citizens, liberal-individualist conceptions extend them to denizens and citizens alike, conceptualizing free speech as a human right. At least at first glance, the republican perspective seems applicable exclusively to democratic societies, since the ‘principle of the unqualified freedom of public speech … has no political justification where men are governed without their consent’ (Meiklejohn Reference Meiklejohn1948: 84). The individualist perspective on free speech is in principle available to all types of polities, but less favoured outside liberal societies.

Kant’s conception of freedom of speech contains both liberal-individualist and republican elements. The first element is an ‘innate’ right to free speech as a personal liberty, valid in all contexts (1). A second element is a claim to free speech in circumstances of political subjection, empowering subjects to articulate their judgements on the justice or otherwise of laws and decrees (2). Kant introduces this second claim in two variants: one, as a necessary feature of republican constitutions, and two, as a reactive claim against illegitimate oppression in all polities. In other words, Kant argues for free political speech in a republican and a non-republican variant, applying his argument to democratic as well as autocratic political systems, but in both cases geared towards politically salient speech. A third element, covering political and non-political speech, is developed in Kant’s account of personal and political enlightenment (3). For reasons set out below, I shall not draw on this third element in the present context.

(1) Kant introduces free speech as a personal liberty in his introduction to the Doctrine of Right.Footnote ⁵ A person is entitled to ‘communicating [his or her] thoughts to [others], telling or promising them something, whether what he says is true and sincere or untrue and insincere’ (DR, 6: 238). We know that it is a private liberty since Kant introduces it as a natural legal position in the sense that persons can hypothetically draw on it in a ‘state of nature’, i.e. in the absence of all government, as well as under any existing legal order. Kant refers to this entitlement as part of ‘innate right’, i.e. the right to such freedom as can coexist with all others’ freedom under law, because innate right contains the authorization ‘to do to others anything that does not in itself diminish what is theirs, so long as they do not want to accept it’ (DR, 6: 237). Since others are always free to reject one’s statements or offers, one is free to make those statements and offers. By ‘what is theirs’, Kant means other people’s innate and – as he goes on to explain in the section on Private Right – acquired rights, including but not limited to their rightful possessions. In other words, even if a communicative act negatively affects the rights of its hearers, such communications should be free if the harm effected by them results from persuasion.Footnote ⁶ The underlying idea is that acts of communication freely accepted by the hearer shift the responsibility for any consequences onto them. Hearers make use of their capacity to decide what to believe and what statements or promises to accept, ‘for when someone merely says what he thinks, another always remains free to take it as he pleases’ (DR, 6: 238n.).

In the remainder of this article, I will return to the idea that individuals have rights; however, I will not put any weight on the concrete scope of Kant’s ‘innate right’. This, for three reasons: first, I do not want to run the risk of importing substantive content from the natural law tradition into this discussion. Second, I want to bypass the debate between those who read ’innate right’ as providing us with something like a catalogue of individual rights to be ‘extracted’ from Kant’s formulation (Unruh Reference Unruh2016: 200; cf. Ripstein Reference Ripstein2007) and others who read it as a way of expressing our ‘equal legal accountability’ (Flikschuh Reference Flikschuh2017: 85). Finally, although the innate right to free speech is a central tenet of Kant’s legal theory, the extent of its protection is fully contingent on its congruence with other rights and laws, whether basic or not. The examples Kant introduces, sometimes no doubt inadvertently, show that he takes his vindication of innate free speech to be consistent with narrow restrictions, such as a right to have one’s reputation protected after death (DR, 6: 295). Nor will such freedom extend to begging or disorderly speech (DR, 6: 325). Innate right covers a right to say what ‘is true and sincere or untrue and insincere’, but does not cover fraudulent lies (DR, 6: 238 and n.). Kant is prepared to defend the expression of contingent falsehoods, but will not countenance freedom of communication regarding convictions that are demonstrably false: He is comfortable with admitting the assertion that ‘there is no God’, but will not countenance the public assertion that ‘“it is impossible that there is a God” … On the contrary, the state is authorized to ban such extravagant assertions of a paradoxical mind.’Footnote ⁷ Violations of honour and violations of dignity restrict the natural right to free communication; one central case is the right to be beyond reproach (DR, 6: 237). This entails that although the claim to free personal speech is universal, its scope is limited by the open-ended nature of potential clashes with conflicting claims. In societies protective of personal, professional, caste or gender honour (VN, 23: 363–70), this may amount to severe limitations. I conclude that ‘innate’ right, however universal in its extension, should not be thought to carve out a definitive scope for free speech.

(2) Kant’s derivation of a right to political speech is very different. While the Introduction to the Doctrine of Right equips people with an ‘innate right’ to free speech in private law, its sections on Public Right add an entitlement under public law. The republican constitution envisaged in this work is to guarantee full political as well as personal freedom, based on personal and political equality (DR, 6: 313–14; cf. TPP, 8: 350n.). Kant’s conception of popular sovereignty, introduced in the section on Public Right, conceives of active citizenship as equal political freedom, that is, as the entitlement ‘to discuss the state (den Staat zu behandeln), organize it, or contribute to legislation as active members’ (DR, 6: 315, tr. PN). In the same way, since citizens in a republic will be deciding on war and peace, directly or via their representatives (TPP, 8: 350; DR, 6: 341), the necessity or otherwise of wars will become a matter for public debate. It is reasonable to assume that, as a condition of all-round active contributions to legislation, ‘discussing the state’ cannot be restricted in content, and free political speech will in principle extend to all areas of public regulation.

Since Kant provides parallel arguments for free political speech elsewhere, with varying scope and limits and based on different assumptions, it is important to trace the particular genealogy of this view in the development of Kant’s political republicanism.Footnote ⁸ Two years before Kant’s fully fledged republican turn in Toward Perpetual Peace and four years before the republican Doctrine of Right, he presented a preparatory step in the argument for comprehensive political freedom in his essay on ‘Theory and Practice’, a work still very much torn between old autocratic and new republican allegiances. In its ‘autocratic’ part (TP, 8: 297–305), Kant takes issue with Hobbes’s comprehensive denial of citizens’ rights, while criticizing others for encouraging violent resistance in cases of their infringement. His position is that citizens do have rights even though they may not be recognized everywhere, such that wherever one finds oneself subjected to public law, an entitlement to the public expression of peaceful dissent follows from this fact. This argument for freedom of ‘the pen’, the ‘sole palladium of the people’s rights’, is based on two strands, one reflexively triggered by subordination under law, the other building on this predicament in an epistemic perspective. Both strands of argument presuppose that in autocratic political systems in which personal rights may be imperilled, ‘[a] nonrecalcitrant subject must be able to assume that his ruler does not want to do him any wrong’ (TP, 8: 304). From this, it follows that:

The first argument (a) is meant to safeguard subjects’ ‘claims to right’, but the entitlement in question is not a ‘claim to right’ in the sense of ‘innate’ or historically acquired private right claims in the Doctrine of Right. A rights-bearing citizen is ‘authorized to judge for himself’ an arrangement that he takes to be unjust. This is a new and reflexive entitlement, which can only arise once public law is in place: a second order right, a right to defend whatever one conceives of as one’s own and others’ ‘inalienable’ first order rights. Freedom of speech under the description of ‘freedom of the pen’ is thus a higher order claim that refers to the claims citizens believe they have, but which they are deprived of in the current political situation. The higher order claim is triggered by the supposition that, if people have rights, it must be possible for them to reconstruct government activity as non-hostile, and that this puts government under a duty to let people voice their dissent. The second argument (b) in the above quotation points out that free speech is necessary for constructive policy-making in the discovery and undoing of perceived wrongs. Like the first one, it is not just a wager on the epistemic benefits of free speech, but a condition of legitimate rule. All those subjected to a legal order have a claim to view government as acting under the supposition that knowledge of facts and of consequences matters for legislation: that wrongs would be rectified and abuses stopped, if only access to such knowledge would be provided. It is enough to argue that if people have non-contingent claims at all, then, under all governments, articulations of perceived violations of such claims must be free, as reactive responses to presumed wrongs and as constructive contributions to righting them.

‘Theory and Practice’ addresses all governments, including autocratic ones,Footnote ¹⁰ and the justification of free speech it delivers is therefore limited in scope. While Kant’s pre-republican argument stringently protects speech addressing presumptive rights violations, political speech directed at other targets may fall prey to limitations. Kant distinguishes questions of right from questions of policy and asserts that, under autocratic government, subjects ‘are not entitled to appraise’ the latter (TP, 8: 297n.). While, for example, a people in an autocracy must be free to criticize a war tax that levies unequal burdens on it ‘since it cannot take this distribution to be just’, citizens cannot complain that such a tax is dispensable, or that ‘in their opinion the war may be unnecessary’ (TP, 8: 297n.). They cannot let go of their rights and their interpretation of what is just,Footnote ¹¹ but are not thereby entitled to weigh in on administrative controversies, nor on decisions over collective well-being. Nor does Kant extend pre-republican free speech to challenging ‘the limits of esteem and love for the constitution (Verfassung) within which one lives’ (TP, 8: 304), such that pre-republican speech can be held to conformity with the basic principles of the established order.Footnote ¹² We have seen that this restriction is to be lifted, and all topics for regulation to be dealt with freely under a republican constitution, since all legislative authority then lies in the hand of the people. It is therefore obvious that Kant’s basic two-level argument will generate different scopes for autocracies and republics.

(3) There are further arguments for freedom of speech in Kant, most forcefully stated in his 1783 essay ‘An Answer to the Question: What is Enlightenment’. Kant’s arguments from shedding one’s ‘self-incurred immaturity’ in morality, religion and politics are designed to apply to autocratic and republican societies alike. However, they are strongly dependent on values such as enlightened maturity and emancipation from traditional loyalties and authorities. They are committed to progress and improvement, based on a perfectionist understanding of autonomy.Footnote ¹³ I avoid these arguments here for methodological reasons, as they may exceed what Rawls is prepared to countenance as an uncontroversial basis for free speech even in liberal and democratic societies. Although Rawls famously picks up on the notion of a ‘public use of reason’, which in Kant stands in the service of increasing personal and collective autonomy, he steers clear of Kant’s enlightenment justification for free speech (PL 210). Rawls does interpret John Stuart Mill’s arguments for free speech as perfectionist, and we can see that their reliance on a ‘permanent interest in the conditions of individuality’, on the one hand, and in the ‘progress of civilisation’, on the other, may make them less than neutrally and universally acceptable under conditions of reasonable pluralism (LPP 304). Although more parsimonious than Mill’s position, Kant’s enlightenment arguments for free speech run a parallel risk of introducing controversial normative orientations, and I do not draw on them here.

3

Rawls acknowledges that all liberal political conceptions include ‘the right to free political speech of all kinds’ as well as ‘the right of free non-political speech’ (LPP 12). In his first book, A Theory of Justice, he includes both elements. When introducing freedom of speech, in enumerating the set of basic liberties in the first principle of justice, Rawls mentions it as a basic liberty that belongs with and complements political freedom, and lists ‘freedom of thought’ as a separate basic liberty:

Note the semicolons grouping batches of basic liberties together, and separating freedom of speech from freedom of thought. This means that the list of basic liberties ‘contains two distinct items, each of which might reasonably be included in in the category of free speech’ (Haworth Reference Haworth1998: 175). Samuel Freeman (Reference Freeman2007: 47) explains that freedom of thought, for Rawls, is ‘similar to freedom of conscience but extends more broadly to include freedom of belief and the expression of belief on all subjects, be they political, literary, artistic, scientific, or philosophical. This basic liberty protects freedom of inquiry and discussion, as well as the freedom to communicate and express one’s views on all subjects.’ In using ‘freedom of thought’ to entail freedom of personal speech, Rawls is not only echoing the second chapter of Mill’s On Liberty, but also Kant’s ‘What does it Mean to Orient oneself in Thinking?’, which claims that without public articulation, free thought would lack all claims to meaningfulness or correctness (WOT, 8: 144). ‘Freedom of political speech and discussion’, on the other hand, Rawls includes under the ‘equal political liberties’ which also contain the liberty to assemble, associate, and take part in elections as equals (Freeman Reference Freeman2007: 46). While freedom of thought is to be understood, in Benjamin Constant’s influential distinction, as a personal or ‘modern’ liberty, freedom of speech, in relating to the political liberties, represents an aspect of the freedom ‘of the ancients’ (JF 2, cf. 143). Freedom of thought is thus to be kept separate from the political rights and freedoms, and Rawls insists that ‘freedom of thought and liberty of conscience, freedom of the person and the civil liberties, ought not to be sacrificed to political liberty, to the freedom to participate equally in political affairs’ (TJ 1999: 177) In its irreducible dualism, Rawls’s conception of free speech parallels Kant’s.

What commentators seem to have missed, as far as I see, are the changes over time that Rawls’s conception undergoes in the identification and derivation of types of expressive liberties.Footnote ¹⁴ When first introducing free speech, in enumerating the basic liberties in the original version of A Theory of Justice, Rawls refers to it as a basic liberty that belongs with and complements political freedom, and contrasts it with freedom of thought and liberty of conscience. In Political Liberalism, he locates free speech differently. In the chapter ‘The Basic Liberties and their Priority’, first published in 1982, he starts out with an informal enumeration of the basic liberties contained in the first principle of justice. He includes freedom of thought and conscience, and ‘the political liberties’, separated off by semicolons, but does not mention freedom of speech. Likewise, in the Preface to the Revised Edition of TJ, dated 1990, in listing the basic liberties, he omits freedom of speech (TJ 1999: p. xii). Rawls does not thereby instil any doubts as to its continuing status as a basic liberty, but no longer affiliates it explicitly with a type or kind. Later in ‘The Basic Liberties’, when taking up some controversial issues in First Amendment jurisprudence, he does address freedom of speech, but surprisingly groups it with a different class of liberties, introducing it as ‘freedom of political speech and press which falls under the basic liberty of freedom of thought and the first fundamental case’ (PL 340). I will take up the meaning of the phrase ‘the first fundamental case’ in the next paragraph. Here, I want to highlight the change in Rawls’s account. While Rawls had kept the political liberties, alongside free speech, separate from freedom of thought in his original conception, and had discussed freedom of thought and its expressive entitlements predominantly as a personal liberty, in his new account he groups freedom of political speech ‘under’ the basic liberty of freedom of thought. What is new to Political Liberalism is thus not that Rawls should sway in the importance of political speech with regard to the basic liberties. What is new is the connection between free political speech and freedom of thought, in a chapter that sets out the justification for, and the limits of, the basic liberties.

I suggest that Rawls’s change of mind results from the new justification for the basic liberties that had prompted him to revise TJ and pursue a more complex derivation of them in a number of interim articles and in PL. According to Rawls’s own reconstruction, in the earlier work he had derived the basic liberties and their priority from an insufficiently narrow conception of the person – one consonant with the idea of self-interested parties choosing basic liberties as primary goods in the original position (TJ 1999: p. xiii). In the later work, he understood that the basic liberties depend on a more complex conception of the person (which was of course already present in the design of the original position in TJ, but less prominently so). This more complex conception sees persons pursuing two highest-order interests in realizing their two moral powers in ‘two fundamental cases’ (TJ 1999: p. xii), first, in the development and application of a sense of justice and, second, in the development and pursuit of a conception of the good. The two moral powers stand for the reasonable and the rational elements in our souls, for our capacities for a sense of right and justice, on the one hand, and for the pursuit of a conception of the good, on the other. They correspond to persons’ two highest-order interests, the interests in securing fair terms of social cooperation and in realizing self-chosen aims in the pursuit of a good life. When, in his new account, Rawls asserts that freedom of political speech ‘falls under the first fundamental case’, he means that the justification for freedom of political speech refers exclusively to the capacity to develop a sense of justice and the highest-order interest in securing fair terms of social cooperation, not the capacity and interest in pursuing a good life. The justification does not invoke what he calls ‘the second fundamental case’, namely that basic liberties are necessary in order to ‘decide upon and revise, and rationally to pursue, a conception of the good’ (CP 366). The interest referred to by Rawls in his derivation of free political speech as the ‘first fundamental case’ is our interest in judging our social world in terms of right and justice, and in developing and exercising this capacity. Bypassing the complex ‘original position’ representation of the first moral power, what is ‘basic’ about both freedom of thought and freedom of speech is that they are orientated toward issues of right and justice.

Rawls’s change of mind has at least two consequences for his view of free political speech. The first is that his new justification may narrow down our conventional understanding of the value of political speech and, by association, our conception of freedom of thought as far as it relates to political speech. Note that the new justification has freedom of thought and liberty of conscience coming apart. While ‘[t]he equal political liberties and freedom of thought are to secure the free and informed application of the principles of justice’, liberty of conscience is now justified under ‘the second fundamental case’, the interest in ‘forming, revising, and rationally pursuing a conception of the good over a complete life’ (PL 334–5). Unlike what may continue to be true for personal speech, therefore, the foundational importance of free political speech is not motivated by whatever function it may serve in the pursuit of the good life.Footnote ¹⁵ This sends a strong message to interest-based conceptions of democracy, such as preference-aggregative accounts. Viewed as an element of the basic liberty of freedom of thought, there is no value to individuals using political speech in getting what they want, or in becoming clear about what they want, but rather in deliberating about what they should want. Rawls gives no justification for using political speech in shaping identities, or for promoting, criticizing and revising affiliations with the goals that people have adopted in their pursuit of the good life. The latter is the domain of the liberty of conscience, from which free political speech is now separate. This entails that free political speech is now a rather ‘moralized’ notion, in that it focuses more clearly on questions of justice, and at least not immediately on the conditions of rationally pursuing the good life, for example in the struggles for political power in a democracy. In focusing on issues of justice, it is less tightly bound to ensuring the functionality of democratic systems, and can perhaps be prized free from an exclusively democratic interpretation. The second consequence of the shift of Rawls’s doctrine after TJ has the opposite effect, in that the justification for the basic liberties is now more contextually specified to democratic regimes, at least more explicitly so, and thus more limited in its appeal. The new justification of the basic liberties lies in a conception of the person allegedly derived from ‘our’ view of persons as free and equal citizens in democratic political systems. Rawls finds the idea of citizens as equipped with two moral powers in ‘the tradition of democratic thought’ (PL 12–13), and he accordingly situates all justification for the basic liberties in this tradition. This may seem to entail that the lessons learned from the new justification for free political speech cannot apply outside the realm of liberal and democratic societies. I will suggest that the opposite is true, that tying political speech to freedom of thought and to ‘the first fundamental case’ liberates it from a narrow functionality for democratic preference-expression and will-formation and makes a version of it applicable to non-liberal societies. The new argument supports people’s power to judge – and their interest in judging – the justice of the basic structure of their societies.

4

Freedom of speech is notoriously not among the human rights delineated by Rawls in The Law of Peoples.Footnote ¹⁶ Moreover, while he singles out for special attention the difficult case of religious toleration, he does not explicitly reflect on the absence of free speech claims from his rights catalogue. Absent from Rawls’s list of human rights are free political speech as well as free personal speech, and likewise full and equal liberty of conscience, though he does say that the remaining rights, or what remain of those rights, are to be guaranteed to a degree ‘sufficient to ensure freedom of thought’ (LoP 65). This is puzzling, since we saw that freedom of thought in the early Rawls entailed freedom of personal speech; moreover, according to the later Rawls freedom of thought entails freedom of political speech. Rawls now backtracks on this view. Take the imaginary society of Kazanistan, introduced by Rawls as a hypothetical and ideal-theoretic example for a human rights-respecting ‘decent hierarchical society’, where people have unequal religious and civil freedoms, and therefore differential liberty of conscience, and no principled claim to free personal or political speech (LoP 75–8). If free personal speech, or free political speech, are necessary elements of freedom of thought, safeguarding the Rawlsian list of human rights will not be sufficient to ensure it. The remainder of the article therefore has a dual task. In the first step, I will show that if human rights are to be ‘sufficient to ensure freedom of thought’ then Rawls’s list will have to be amended. In the second step (section 5), I will show that such amendment becomes possible through the integration of a Kantian ‘module’ into Rawls’s account.

Rawls gives two justifications for the items he places on and off the list of human rights. The first, political justification is that human rights should not be conceived of as a parochial instrument rooted exclusively in Western traditions (LoP 65; Williams Reference Williams2011: 167). Therefore, patently liberal rights have been left off the list. This is important not least since Rawls considers respecting human rights a sufficient condition for a society to be immune from external military intervention (LoP 79–80). However, it is wrong to ascribe to Rawls the converse thesis – that a violation of human rights is a sufficient ground for intervention.Footnote ¹⁷ Rawls is clear that only under extremely demanding conditions, including ‘egregious’ violations of multiple rights (his examples are slavery and human sacrifice), should military interventions even be contemplated (LoP 93–4n.). The presumption that free speech is left off the list in order to avoid military intervention in cases of non-compliance is therefore misguided. The second, more philosophical explanation is that human rights are to be independent of controversial moral and political ideals of the person that may not be shared everywhere. The conception of the person as free and equal citizen, with two moral powers and their corresponding highest-order interests, which served to derive the basic liberties in Rawls’s liberal conception of justice, cannot therefore form a basis of the Law of Peoples. The assumption that persons are equal, reasonable and rational cannot straightforwardly be globalized, since it would discriminate against alternative conceptions of the person, or challenge the different weighting of group claims as against the claims of individuals in non-liberal societies (LoP 82).

The standard, and in my view correct, interpretation of Rawls therefore is that he cannot justify a right to free speech in autocratic states. Now a first, quick objection could just deny that its absence from the list of human rights entails that freedom of speech will be unprotected in non-liberal societies. It could be argued that Rawls’s list of human rights is not meant to be exhaustive or exclusive. But this objection goes nowhere, since Rawls plainly admits that, even under a full realization of their human rights, ‘individuals do not have the right of free speech as in a liberal society’ (CP 546). A more plausible objection is that, although free speech is not included among human rights, the individual and social good of exchanging conflicting views may still be heeded under the Law of Peoples. Indeed, Rawls (still in ideal theory) provides what he views as a functional equivalent of the participatory and expressive freedoms in his favoured form of non-liberal ‘decent hierarchical society’. ‘Decent hierarchical societies’ fulfil Kant’s condition that citizens must be able to see their state as non-hostile, as not willingly or systematically violating their rights.Footnote ¹⁸ A ‘consultation hierarchy’ is to enable minority group representatives to present their concerns to government officials and ensure a ‘right of dissent’ that is to guarantee dissenters an official reply to their communicated grievances (LoP 61). Below, I will discuss what exactly this ‘right of dissent’ entails (section 4(2)), but if it yielded a non-negligible scope for free speech, it would go some way toward explaining why Rawls can hope that ‘freedom of thought’ can be fully redeemed in non-liberal societies.

The claim that decent hierarchical societies are to honour a ‘right of dissent’ has encouraged interpreters, most prominently David Reidy, to endorse an extensive interpretation of speech entitlements under the Law of Peoples. Reidy claims that ‘citizens or subjects’ will be within their rights when they ‘insist on the content of their basic human rights’. In a decent hierarchical society, they ‘can criticize or dissent from violations or failures to deliver that content. They can demand a public justification’ (Reidy Reference Reidy2008: 177). I agree that if Rawls’s Law of Peoples granted such claims, a core area of free speech, and with it an important dimension of freedom of thought, would be protected. However, in the remainder of this section, I shall argue that The Law of Peoples provides no support for such claims. Neither is the content and interpretation of human rights, in Rawls, a concern for disagreement, nor is challenging them a matter for individual citizens. And while it is true that, within a consultation hierarchy, public justifications can be demanded, they cannot publicly be demanded.

The provision of a ‘consultation hierarchy’ is what distinguishes human rights-respecting and ‘decent’ hierarchical societies from (generally, though not fully) human rights-respecting societies Rawls classifies as ‘benevolent absolutisms’ (LoP 63). While the former are well-ordered, the latter are not. There exists a small industry working away on the question of which contemporary societies, if any, fulfil the conditions for decent hierarchical societies,Footnote ¹⁹ but here I want to start from the one historical example Rawls gives. The difference between ‘decent hierarchical societies’ and ‘benevolent absolutisms’ broadly parallels the distinction between what makes for a legitimate state in Hegel’s Philosophy of Right (LoP 72–3, CP 553) and what may be characteristic of the brand of benevolent despotism envisaged as a legitimate autocracy in Kant’s pre-republican writings. On Hegel’s side, we have a corporatist system that represents society’s groups or ‘estates’, on Kant’s side, a despotic political system that is publicly committed to the rule of law and to individual rights but may nonetheless sometimes infringe them. As we saw in Kant’s ‘Theory and Practice’, a powerful argument can be given for respecting free political speech under such benevolent absolutism. I will now argue, in four steps, that the Hegelian institutional set-up not only offers no functional substitute to guaranteeing freedom of speech, but that it falls short of the Kantian pre-republican version of benevolent absolutism in this regard, and should be amended by it.

(1) The first point is that estates in a ‘consultation hierarchy’, according to Rawls, are to represent ‘the rational interests of their members’ (LoP 73). This should not be construed narrowly, as if corporations were exclusively concerned with realizing their members’ preferences, since as we saw an individual’s ‘rational interests’ include her identity concerns, for example, her religious affiliation and other basic interests in leading a meaningful and good life. In parallel, estates (corporations, associations, groups) claim a communal entitlement to the exercise of religion (although not necessarily to its free or equal exercise, LoP 74). It is thus fair to say that estates will be able to guarantee, to an important extent, the liberty of conscience of their members, in a coherent overall picture that aligns liberty of conscience with rational interest. Recall that liberty of conscience was characterized, by the later Rawls, as fallout from ‘the second fundamental case’, such that, although not based on a liberal individualist conception of the person, the estate model appears as guarantor of people’s capacities for, and interest in, the development and pursuit of a conception of the good. Liberty of conscience had, however, no immediate justificatory connections to ‘the first fundamental case’, the capacity and interest in developing and articulating a sense of justice. In decent hierarchical societies, corporations channel concerns for the good and meaningful lives of their members; they do not dispute about their rights or claims to justice. Moreover, since decent hierarchical societies are defined by their adherence to human rights, it is not obviously the case that a robust denial that they do so will have a place in the exchanges of a consultation hierarchy. It is hard to see how there could be a principled case for such human rights claims from minority groups, given that at least outsiders are barred from voicing criticism of decent hierarchical societies for failing to ‘recognize and protect’ human rights (LoP 61).

(2) By way of introducing the second point, suppose Rawls granted the first. Assume that his narrow view of the function of estates or corporations can easily be repaired: the representative function of estates could be amended to include the discussion of general moral and right-based claims, addressing the conception of justice that is realized in the decent society.Footnote ²⁰ My second point is that such repair, although it clearly puts a strain on decent societies, will not go far enough since it will not unequivocally protect public dissent. Rawls seems to assert the opposite when he says that, in the procedure of consultation, members of corporations have a right to ‘express political dissent’ and that such dissent ‘expresses a form of public protest and is permissible provided it stays within the basic framework of the common good idea of justice’ of a decent society (LoP 72). In contrast, I want to argue that Rawls’s idea of a consultation hierarchy falls behind the Kantian model of legitimate autocracy in an important respect in that it mirrors the grievance (gravamina) model from pre-Kantian natural law. One influential version of this model is propagated in Achenwall and Pütter’s Elementa Iuris Naturae, which Kant used throughout for his lectures on legal philosophy and criticized in ‘Theory and Practice’ as well as elsewhere (TP, 8: 301; DR, 6: 306). Achenwall and Pütter state that citizens who feel they have been unjustly treated can present their complaints to the summus imperans, and that they deserve an answer. However, the communication of such gravamina does not presuppose the existence of a public forum. It can be brought forward in private communication with the ruler, indeed there is ‘no right to bring it to anybody’s attention’ (Achenwall and Pütter Reference Achenwall and Johann Stephan1750: §798, 234–5, tr. PN). Kant, in contrast, demands that people have the right to make presumptive wrongs public by making ‘general and public judgements’ (TP, 8: 305). Although he retains the gravamina rhetoric of older natural law until his late writings (DR, 6: 319), he insists that denouncing such grievances, if it is to be effective, would have to be take place coram publico: ‘[I]f an entire people wants to present its complaint (gravamen), this cannot take place in any way but publicly, … even concerning the least of its demand, its natural right’ (CF, 7: 89, tr. PN). Rawls takes an interim position between Achenwall/Pütter and Kant. One of his examples for consultation hierarchies is the Caliphat in Islamic political thought, where the ruler is required to listen to his subjects’ grievances (LoP 72). In such a confrontation, the process of rectifying wrongs need not involve public consultation and contestation, or their subsequent public condemnation, or even public awareness of them, except ex post in the narrow context of the courts. ‘[G]overnment and judicial officials are required to give a respectful reply’ if confronted with dissent (LoP 61), and, when presented with dissent in court, their answer must be given ‘publicly’ (LoP 67). Outside of formalized contexts like court proceedings, however, it would not make sense to connect the articulation of grievances to the entitlement to receive an answer. Where people voice dissent in a political public sphere, there is no one determinate procedure for answering them, nor will those expressing their dissent be prepared to end their protest, as the gravamina model demands, when presented with a justification. Thus the existence of formal proceedings in consultation, trials or litigation does not entail the existence of a public forum in which grievances are to be articulated. I conclude that when Rawls asserts that ‘dissent expresses a form of public protest’ (LoP 72), he cannot mean by ‘public’ a protest that must be out in the open, accessible to all, but must mean a protest expressing public, not private interests.

(3) A third point is that Rawls is clear that claims to representation in a consultation hierarchy and the ‘right to dissent’ reflect a status that may be conferred on their bearers not as individuals, but mediated by their group membership. In well-ordered hierarchical societies, human rights are typically protected as ‘rights to persons as members of estates and corporations and not as citizens’ (CP 553). Indeed, it must be acknowledged that a group-based, representative exchange of grievances may go some way towards redressing injustices and human rights violations. But whether consultation hierarchies can effectively and comprehensively address violations of rights will depend on how comprehensively groups within society have achieved representation. When Rawls argues that historically oppressed groups such as women need to be represented in a consultation hierarchy (LoP 75), this entails that before such representation succeeds, and in the absence of a right to voice dissent outside the corporatist context, the ‘consultation hierarchy’ equivalent to individual speech rights cannot be fulfilled. From Rawls’s precept that, in a consultation hierarchy, ‘[i]t is necessary and important that different voices be heard’ (LoP 72) we cannot infer that it is necessary or desirable that all voices be heard. Only if we grant that all relevant groups have already and in perpetuity been identified and included in representation in a decent hierarchical regime, can we rule out that there will be further legitimate claims to dissent, such that none will go unheard.

(4) Finally, consider freedom of thought. Though I have rejected Reidy’s interpretation, I have conceded that, to an important extent, freedom of thought would be provided by a regime that gave people a public right of dissent over the contents of their human rights. This is because it values, supports and makes possible their independent judgement about the justice or otherwise of the basic features of its institutions. But it follows from the three above points that the people in decent hierarchical societies do not have that right. They do not have even a modicum of authority over the content of their rights. They have no moral or epistemic authority over the interpretation of their own, and others’, rights claims. What liberty of conscience they have extends to their religious and other affiliations, and is directed to their ‘rational’ interest, not their interest in justice or human rights. They have no share in the public authority over the interpretation of their own, and other peoples’, rights, since established processes of contestation and redress need not be transparent. Finally, they have no individual or other extra-corporate authority over the content of their rights claims. In contrast to what Reidy claims, dissent cannot be brought forward by individuals as ‘citizens or subjects’, but only as members of disadvantaged groups. This generates a problem for members of as-yet-unrecognized groups. Even in the absence of any well-worked out conception of freedom of thought, it seems implausible to assume that if people have no individual moral or epistemic authority over what their rights claims are, and if they cannot claim to have a share in their public interpretation, they have the crucial resources for freedom of thought at their disposal.

5

In the preceding section, I have argued that it is necessary to amend Rawls’s conception, even according to his own requirement, to secure important aspects of freedom of thought. In this section I need to show that it is possible to do so. The ‘module’ the Law of Peoples conception of human rights needs to adopt is one that secures individuals’ free public, justice-orientated dissent. The idea is very simple: if somebody who is subject to a given legal order believes that their or others’ rights have been violated, this yields a right to articulate and contest such a presumptive violation. In Kant’s conception, such a module can be derived from the combination of the claim that there are basic (‘innate’) rights that a person could not give up even if they wanted to, and the claim that people are entitled to publicly judge as to their fulfilment or otherwise. Rawls’s account seems to provide for a parallel dual structure. People ‘have’ human rights, and in principle, dissent over their interpretation could be recognized through the institutions of a consultation hierarchy, complemented by an individual right to public political dissent. In introducing the Kantian ‘module’, dissent and to a certain degree freedom of thought would be protected, albeit not in the full sense as under liberal democracies but rather restricted to topics of rights and justice. Just as in Kant’s pre-republican thought, free speech would not necessarily extend to all policy issues and to non-justice related personal and social issues (some of which would still be addressed in the consultation hierarchy). In contrast to introducing a right to personal and political speech with universal scope, which could plausibly be rejected as a ‘liberal’ import under Rawls’s premises, I propose that Kant’s idea leads to a modest and congenial amendment of Rawls’s conception that could supersede the difficulties addressed in the preceding section. In order to clarify and defend this suggestion, I take up a set of initial objections. I concentrate on objections that deny the coherence with Rawls’s overall framework and leave feasibility issues aside. I am sure further objections can be raised, but hope to present them with a more definitive target in this section.

(1) In the architecture of Kant’s argument, public political speech rights are second-order entitlements to defend a set of pre-existing rights, including, as we saw, a first-order right to free personal speech. Does this entail that the argument only makes sense if we assume that people should be able to claim a comprehensive set of individual rights as in the liberal tradition, thereby going far beyond human rights minimalism? The answer is no. Although the argument may fail to go through if the set of first order rights is empty, since there would then be nothing to defend, it does not presuppose that there is a substantive set of rights to be defended. On the basis of the premise that ‘every human being … has his inalienable rights’ (TP, 8: 304), an individual right to public dissent could articulate two types of claims. It could address presumptive violations of already recognized rights that have gone unnoticed, or it could assert the existence of further dimensions or categories of rights that are as-yet-unrecognized.

(2) A second objection could insist on the distinction between conceptions of natural rights and political conceptions of human rights. Rawls correctly, in my view, rejects the idea that human rights are natural rights as too controversial to serve as a basis for a global conception of international law. But the natural rights rhetoric is an important element of Kant’s dualist strategy. If an individual has ‘his inalienable rights, which he can never give up even if he wanted to’, this seems to give him some moral or epistemic authority over the content of those rights. Likewise, if rights are an inherited dowry from the state of nature, it makes sense to see every individual as ‘authorized to judge for himself’ whether his rights have been violated (TP, 8: 304). Yet it is unclear whether any presumption of moral or epistemic authority on the part of the rights-holder extends to other conceptions of rights. I will address this problem in two parts, in this and the next paragraph. It is especially pressing given the ‘political’ quality of Rawls’s alternative account. ‘Political’ conceptions of human rights look ‘first to the treatment of human rights within the already existing discourse and practice of human rights or what has been called the “human rights regime”’ (Baynes Reference Baynes2009: 374). It seems plausible that political conceptions of human rights should leave it to experts (diplomats, international lawyers, judges, philosophers) to determine their content, and that rights-bearing individuals have no special authority in this regard. In response, we need to distinguish between a claim to authority in specifying the content of a political conception of human rights, and the claim to authority in challenging a given interpretation of its content. Dissenters will not aim to present a comprehensive alternative view of human rights, but they will voice moral outrage if they believe themselves or others to be affected by a violation. Secondly, it is not a requirement of moral or epistemic authority, as Kant’s version may make it seem, that the individual bearer of the presumptive entitlement have an exclusive claim to expertise over her rights. People may have a largely wrongheaded understanding of their rights, and no unilateral final authority over their scope. All we need for the argument for rights-based dissent to go through is a presumption of inalienable partial authority, resulting from the inalienability of the claims in question. What we need to assume is that human rights conceptions, including ‘political’ conceptions, are open to revision and change, and that the necessity for such change may justifiably be articulated by individuals, through their public expression of dissent over presumptive human rights violations.

(3) In a third objection, one might refer to the observation that Rawls is not prepared, in the Law of Peoples, to draw on Political Liberalism’s conception of the person as equipped with two moral powers. We saw that it is the first moral power, a capacity for a sense of justice and the highest-order interest in working out fair terms of cooperation, that linked public political speech and freedom of thought in democratic societies. If we cannot draw on the capacity and interest to offer fair terms of cooperation for the basic structure of society to derive basic liberties, where can the entitlement to criticize existing understandings of right and justice find support? I have already conceded that it is not full but partial inalienable authority over one’s rights that is crucial in order to derive an entitlement to individual and public rights-orientated criticism. But from Rawls’s perspective, it may still be insisted that there is no direct inference from the thought that ‘I have rights’ to ‘I am the moral authority of last resort in interpreting the scope of my rights’. Rawls explains that in a decent hierarchical society, all citizens have authority over ‘their moral duties and obligations’; however, he does not include authority over their claims (LoP 71). All are seen as capable of passive ‘moral learning’, but not of moral intervention or innovation. They are seen as ‘decent and rational’, but not necessarily as reasonable, let alone equally reasonable (LoP 71). They have human rights as members of their groups, within the bounds of an objectivist political conception of human rights, and embedded in a ‘common good idea of justice’, so that it is not obvious where the recognition of all individuals’ capacities and competences to shape their society’s understanding of rights and justice should come from. Decent societies could therefore reject the idea that individuals have the moral authority, as argued by Kant in ‘Theory and Practice’, to claim permanent interpretative expertise over their own rights claims. They could bite the bullet and deny, against Rawls’s assumption, that their members are owed freedom of thought.

In response, recall that Kant based his two-level argument for free political speech on two strands of argument, one resting on interpretative entitlement (a), one on the epistemic virtues of the articulation of rights violations (b). Even if the first strand were to be rejected as a valid inference in decent hierarchical societies, an individual’s public reflection of ‘the wrong that in his opinion is done to him’ (TP, 8: 304) will remain an indispensable outlet for otherwise potentially invisible violations. If society is non-hostile, wrongs cannot but be taken to result ‘from the supreme power’s error or ignorance of certain consequences of his laws’ (TP, 8: 304). Again, individuals may not be final experts on the essence or scope of the wrong, but they can claim to be the first and sometimes only monitors of presumptive violations. Kant’s second strand of argument was to reveal defects in the official understanding of the law, and thereby constitute a necessary condition to overcoming them. In this sense, free political speech is the ‘sole palladium of the people’s rights’ (TP, 8: 304, emphasis PN), in the sense that it is a necessary condition for safeguarding them. In section 4 above, I argued that the epistemic function of public dissent cannot be completely absorbed by an ordered representative procedure. Where people feel moral outrage over violations, it is therefore not only their voices that go unheard, but injustices and wrongs will go undetected if they have no individual claim to free political speech.Footnote ²¹ I conclude that even if decent societies were to reject individuals’ moral authority to partake in public contestations of what a given set of human rights entailed for them, they cannot reject their epistemic authority.

(4) A fourth objection concerns the scope of political speech in decent autocracies. In line with Kant’s suggestions, autocratic societies would have to freely allow dissent over human rights, but not necessarily on all policy issues. A more difficult question is whether dissent should be kept within the bounds of the existing constitution. We saw that, for Kant, free speech in autocracies can be bound to loyalty (and even affection) for the Verfassung in which one finds oneself. If we take Verfassung to refer to the rules and ideas shaping the basic structure of society, the analogy to Rawls’s picture of decent hierarchical societies is again striking. Decent hierarchical societies are characterized by a ‘common good idea of justice’ over which there exists consensus, or at least no legitimate public controversy. According to Rawls, dissent in such societies ‘is permissible provided it stays within the basic framework of the common good idea of justice’ (LoP 72). But consistent with the argument for the module implanted here, the right to dissent will obviously question and inflect the list of human rights itself, thereby allowing for the possibility that rights not yet recognized, for example, a right to political self-government, should be on that list. Binding free political speech to substantive conformity with an existing constitution (autocratic or otherwise) would therefore be incompatible with the view advanced here. On the contrary, it entails that people are within their rights to question existing ‘common good ideas of justice’ through heterodoxical articulations of their human rights. Perhaps, then, the criterion of loyalty to the existing constitution should not be taken to entail the uncritical acceptance of its precepts, or of the rules and ideas of its basic structure, but rather signal loyalty to its reproduction, as opposed to the threat of breakdown into a ‘state of nature’ that would no longer secure any list of rights, let alone their potential amendment by second order right. What must be central for both Rawls and Kant is that the institutional system of decent hierarchical societies is capable of reproducing itself and thereby guaranteeing the first order rights and the second order entitlement based on them, not that it remain unchallenged.

(5) Finally, it could be asked whether, in implanting a non-ideal module into an ideal-theoretical account of legitimacy under the Law of Peoples, I am not conflating ideal and non-ideal theory. Since I have argued that Kant’s account of free political speech comes in two versions, a republican and a ‘pre-republican’ one, my terminology already reveals that, for Kant, autocracy is not a stable or attractive end-state of political constitutions. It is therefore fair to say that Kant’s treatment of autocracies is part of what Rawls would term ‘non-ideal theory’, since Kant takes them to be characterized by large-scale non-compliance with universalistic republican demands, and accordingly develops a political theory of transition from autocratic to republican government.Footnote ²² Rawls, in contrast, intends no such transition; indeed the point of recognizing decent hierarchical societies is that they should not be put under any pressure of democratic transformation. All ‘jawboning’ (Nickel Reference Nickel2008) about human rights from outside decent hierarchical societies is to avoid the suspicion that it is to propel a transparently transformative agenda. This caveat can be projected onto Rawls’s own views on the tasks of non-ideal theory, where he allows for two variants. First, in non-ideal circumstances, we may need to do what we can to transition to an ideal end-state. Second, in non-ideal circumstances, we may need to do as well as currently possible from the point of view of justice (Simmons Reference Simmons2010). These orientations are to be distinguished, although they may not always proscribe distinct practices. Kant’s views on domestic and international law and rights, given his republican teleology, tend towards the former orientation, and therefore run the risk of formulating demands in a non-neutral way. Subverting autocracy, for Kant, is a welcome by-product of free public dissent. In contrast, we can read Rawls’s Law of Peoples as saying that such transformational arguments should be discounted in delineating free speech entitlements. But this should not keep us from delineating those rights in a way that is to improve the overall justice of autocratic regimes.

6

By way of conclusion, I want to reiterate that I have framed my argument in a way that is capable of challenging and improving the Rawlsian position from within. In trying to remain within the premises of Rawls’s account, I have not thereby accepted those premises. I tried to show that, even under a Rawlsian description, hierarchical societies, were they to remain ‘decent’, would have to guarantee an important entitlement besides the agreed-upon list of human rights. This is the entitlement publicly to dissent from established understandings of justice and human rights. Where people in hierarchical regimes protest the dominant ‘common good idea of justice’ by way of articulating a dissenting interpretation of their rights, they remain within those rights. If they are kept from voicing their protest, this detracts from the legitimacy of the hierarchical regime. To reach that conclusion, I needed to show that Rawls’s attempt to specify functional equivalents for free speech under decent hierarchy is insufficient, given his own understanding of what they are supposed to achieve (freedom of thought), and that Kant’s argument for freedom of speech in autocracies is convincing, even disregarding his reliance on natural rights. In short, Kant’s argument for admitting free political speech, circumscribed as speech relating to questions of rights and justice, is to be recognized under an influential idea of ‘decent’ autocratic societies, if they are to claim full international legitimacy.