Postsecular Democracy? Habermas and Rawls on Religion

Postsecular Democracy? Habermas and Rawls on Religion

Postsecular Democracy? Habermas and Rawls on Religion

Tom Bailey

(Draft, not for circulation. Comments welcome: )

In this paper, I argue that neglected elements of Habermas’s and Rawls’s treatments of the ‘postsecular’ show their secular political frameworks in an underappreciated light. In particular, rather than considering their much-discussed attempts to accommodate religions within these frameworks, I analyse and compare Habermas’s account of ‘translating’, or ‘learning’, from religions and Rawls’s account of religious contributions to public reasoning under the ‘proviso’ and by ‘conjecture’. Both accounts concern the peculiar task of transforming religious into shareable terms of political justification. And in both cases, I argue, Habermas and Rawls imply that the secular grounds to which they otherwise appeal – deliberative rationality and mutual respect, respectively – have no ultimate authority over religion and, indeed, like religions presuppose an unjustifiable ‘faith’ in their own possibility. This postsecular ‘modesty’ conceives the secular framework concerned as merely the internal elaboration of a contingent secular commitment, with no ultimate authority over alternative, religious forms of social coordination. I also argue that Rawls pursues this modesty further than Habermas. For Rawls conceives of secular grounds not as a substantial philosophical constraint dividing religious from secular politics, as Habermas does, but as a contingent and dynamic achievement of citizens in elaborating their particular moral resources, including religious ones. By contrasting this vision of politics with the formal and overarching secular frameworks that Rawls and Habermas otherwise endorse, then, I show that, at least in these postsecular moments, Rawls not only expresses a postsecular modesty, also indicates a novel postsecular sense of liberal politics.

To introduce my analysis, I begin with a brief summary of the more well-known elements of Habermas’s and Rawls’s secular state frameworks and their attempts to accommodate religions within them. I then consider Habermas’s account of ‘translating’, or ‘learning’, from religions, bringing out its ‘postsecular’ implications. Finally, I consider Rawls’s presentation of his ‘proviso’ and ‘conjecture’, and show how Rawls develops their postsecular implications further than Habermas.

Habermas’s and Rawls’s secular frameworks

I take it that in their later work, both Habermas and Rawls no longer claim that modern societies are, will be or should be secular, and so attempt to accommodate religions in politics. But both nonetheless continue to endorse ‘secular’ frameworks for state politics. That is, for both, political justification must be based ultimately on grounds independent of religion, since these ultimate grounds of political justification must be equally acceptable – ‘available’, or ‘shareable’ – for citizens of different moral persuasions. If religions are to be accommodated in politics, then, both Habermas and Rawls insist that they must be accommodated only within such an overarching secular framework.

Thus, in response to sociological literature on the topic, Habermas has replaced his earlier account of social development towards secularization with an emphasis on the ‘post-secular’, understood as a ‘return’ of religions at least to the public spheres of modern societies. Yet he nonetheless continues to treat the justification of state policies as ultimately based on the shareable grounds provided by deliberative rationality, as exercised in the public sphere and parliament. For he treats political justification in terms of communicative reason, understood as a practice of coordinating social actions through reason-giving that pursues an ideal of objective rational justification, thus superseding ‘pre-modern’, conventional ways of organizing social life like religions. Admittedly, he insists on religious freedoms and – particularly now that he acknowledges the ‘postsecular’ character of modern society – on an informal public sphere that is open to all arguments, including religious ones. In political contexts he also relaxes the rational justification required to what is universally acceptable, rather than a universal consensus; to procedures rather than direct participation; and to admit ethical (including religious), pragmatic and fairly bargained arguments as well as universally ‘moral’ ones. But nonetheless, at the utmost level of political justification, parliamentary deliberations, he insists that religious arguments must be filtered out, excluded, because the ultimate reasons for solving political problems must be equally accessible to all citizens, if the communicative ideal of objective rational justification is to be achieved. At this level, he considers religious arguments to resist the rational criticism necessary for political deliberation, and therefore limits religions’ possible contributions to cultivating citizens’ non-rational motivation to participate in such deliberations, by developing their sense of collective self-determination, their awareness of the need to give universally accessible arguments, and even their sensitivity to injustices.

The later Rawls too recognizes the persistence of religions in modern societies like the U.S.– indeed, this was an important motivation for him to replace his earlier exclusion of comprehensive doctrines from political justification with a more accommodating account. But he too insists that religions must ultimately be excluded from political justifications. Rather than communicative reason, he treats political justification in terms of mutual respect, understood as a fundamental value implicit in modern democratic societies and one distinct from the different moral worldviews of religious and other citizens. For him, such respect should structure our social cooperation in the form of a recognized constitution. Admittedly, like Habermas, Rawls makes certain accommodations to religions in his later work: he allows that religious arguments may be introduced into political deliberations over all but basic constitutional issues; that religious citizens may elaborate their own religious accounts of the mutual respect expressed in a constitution, so that the consensus over it is ‘overlapping’; and that they may even introduce religious ideas into deliberations over the constitutional requirements of respect, on the condition that ‘in due course’ they can find supporting reasons acceptable to others – his ‘proviso’. But at the utmost level of political justification – for him, deliberations over basic constitutional issues by political officials and voters – Rawls too insists that religious arguments must be excluded, in order for the ultimate justification of political coercion to be equally acceptable to all citizens. For, he claims, providing each other with reasons that each other could accept is the only way for citizens to respect each other in the absence of a shared moral worldview. He consequently also echoes Habermas in relegating religions to a supplementary role: at best, he sees them as valuable resources for cultivating consensus over a mutually respectful constitution, insofar as they offer alternative moral terms for expressing and affirming this respect.

Despite coming to appreciate the post- or non-secular character of modern societies, then, both Habermas and Rawls continue to endorse secular frameworks for politics, insisting that political justification must be ultimately based on grounds independent of religions, such that these grounds are equally acceptable to citizens of different moral persuasions. In other words, both respond to the moral pluralism, or moral conflict, which religions and other moral worldviews represent by appealing to further, overarching and ‘liberal’, grounds of ideal consensus. In Habermas’s case these grounds are to be provided by deliberative rationality, as exercised in the public sphere and parliament; in Rawls’s case they are those of a mutually respectful consensus, as expressed in a constitutition, government decision-making, and citizens’ voting.

Now, there are lively debates under way over how inclusive and acceptable Habermas’s and Rawls’s secular frameworks are: some commentators welcome their attempts to accommodate religious argument within such frameworks, some propose alternative secular frameworks, while others object to such frameworks simply because, however accommodating they are, they are ultimately secular. Here, though, my concern is not with these debates directly, but with how some neglected claims which Habermas and Rawls make about the transformation of religious into equally acceptable terms of political justification show their secular frameworks in a different light, and, at least in Rawls’s case, have unexpected implications for the treatment of religions in liberal politics. In particular, I will focus on Habermas’s discussions of ‘translating’, or ‘learning’, from religions and Rawls’s ‘proviso’ about religious contributions to public reasoning and his associated notion of ‘conjecture’. My aim is to show that these notions are distinctively ‘postsecular’ in implying that the secular grounds to which Habermas and Rawls appeal – deliberative rationality and mutual respect, respectively – have no ultimate authority over religion and, indeed, like religions presuppose an unjustifiable ‘faith’ in their own possibility. These implications thus stand in significant tension with ‘orthodox’ understandings of Habermas’s and Rawls’s secular frameworks, and even, I will suggest, with some of the conclusions about the right treatment of religions in liberal politics that Habermas and Rawls themselves draw from these frameworks.


To begin with Habermas, it should first be noted that much of what Habermas says about ‘translating’ or ‘learning’ from religions presents religions merely as sources to be mined for otherwise rational ‘contents’. Like his frequent suggestion that religions might provide non-rational motivations to engage in political deliberations, then, these remarks do not upset his secular framework, but merely supplement it. For he is generally very careful to distinguish between the universally accessible ‘content’ that might be translated or learned from religion, and the dogmatic ‘form’ in which religions present that content when they appeal to a revelatory justificatory source – in other words, he carefully distinguishes ‘knowledge’ from ‘faith’. As he puts it in his essay, ‘Religion in the Public Sphere’, he considers the Western philosophical tradition to have ‘learned’ many of its most fundamental moral concepts from religion by ‘freeing cognitive contents from their dogmatic encapsulation in the crucible of rational discourse’, but he insists that nonetheless, ‘[a]t best, philosophy circumscribes the opaque core of religious experience … [which] remains as profoundly alien to discursive thought as the hermetic core of aesthetic experience’ (2008b: 142, 143). As examples of such learning, he refers to the ideas of equality, respect, autonomy, and community (2008b, 2010 interview, 2011), and he appears to think that more might be fruitfully learned in this way – think of his own appeal to the Christian notion of a given ‘nature’, or ‘creation’, in warning of the dangers of genetic engineering (2003, 2008c s10). But the neat distinction between rationally and universally accessible ‘content’ and inaccessible justificatory ‘form’ – between ‘knowledge’ and ‘faith’ – means that whatever can be learned or translated from religions is just what can be justified on rational, and not only on religious, grounds. In other words, what is learned from religions is just what could already be learned from rational enquiry.

However, some of what Habermas says in elaborating on this translating or learning from religions suggests that it involves much more than such mining of religions for otherwise rational ‘content’. In particular, he claims that it reveals the contingency of his own sense of political justification itself, as well as a certain ‘faith’ that such political justification, like religions, must have in its own possibility.

He makes the first claim, about the contingency of his sense of political justification, most clearly in the final section of ‘Religion in the Public Sphere’, when considering the prospects for successful learning among religious and nonreligious citizens. He begins by remarking that the development of citizens’ ‘mentalities’ in this respect is so ‘unpredictable’ it can hardly be considered ‘a cognitively steered process at all, one that may be described as a learning process’, and that what he considers ‘learning’ in this context is relative to his own sense of rational justificationand the secular political framework he bases on it: ‘these changes in mentality count as complementary “learning processes” only from the perspective of a specific normative self-understanding of modernity’. This raises the question of the validity of this sense of political justification. If this validity is assumed, then the problem of learning is simply one of the practical limits of theory, or what he calls the ‘self-limitation of political theory’ – that is, it is simply the problem of the degree to which citizens actually, in practice endorse, or can be brought to endorse, the theoretically ‘correct’ sense of political justification. But here Habermas does not insist that his is the ‘correct’ sense. Rather, he admits that the validity of his sense of rational justification, and hence also of his secular political framework, is contingent:

[T]his discourse concerning the correct understanding, and the correctness tout court, of a liberal constitution and a democratic civic ethos extends into a terrain where normative arguments do not go far enough. The controversy also extends to the epistemological question of the relationship between faith and knowledge, which itself impinges upon key elements of the background understanding of modernity. Interestingly enough, both the philosophical and the theological efforts to define the relationship between faith and knowledge in a self-reflexive manner throw up far-reaching questions concerning the genealogy of modernity.

What is crucial here is that Habermas does not claim or suggest that his theory answers these questions, that it establishes objectively the distinction and relationship between ‘faith’ and ‘knowledge’. In other words, Habermas admits that his own sense of rational justification – and thus of what counts as accessible ‘content’, as ‘knowledge’ rather than ‘faith’ – is a contingent one, that the secular political framework he bases on it is therefore normatively contingent too, and, generally, that the grounds of modern political justification are open to alternative interpretations of rational justification, and its distinction from ‘faith’. Indeed, in the final sentence of ‘Religion and the Public Sphere’, he concludes that the acceptance of his secular framework depends on ‘whether secular and religious citizens, from their respective points of view, are prepared to accept an interpretation of the relationship between faith and knowledge that first makes it possible to treat one another in a self-reflexive manner in the political arena.’

Also crucially, Habermas presents this insight into the contingency of rational justification as ‘learnt’ not by rationality alone, but only through an engagement with religion. In this, it is an insight into more than the otherwise rational ‘content’ of religions to which Habermas often limits himself. In particular, he claims that it derives from engagements with such religious perspectives as ‘radical orthodoxy’, and that ‘[c]ontroversies with such opponents must be conducted within the proper disciplinary terrain. This means that theological claims can only be met with theological counterarguments, historical and epistemological claims only with historical and epistemological counterarguments.’ He proceeds to indicate that in a possible ‘genealogy of modernity’s understanding of itself’, rationality will not be reducible to scientific rationality, progressively liberated from religions, since scientific rationality itself is rather a product of religions – it is ‘the outcome of a history of reason of which the world religions are an integral part’.

Habermas makes the second claim, that regarding political justification’s ‘faith’ in its own possibility, in elaborating on how his understanding of translation or learning is exemplified by Kant’s philosophy of religion. He does this particularly extensively in another important essay, ‘The Boundary Between Faith and Knowledge’ (2008c; also 2010). Notably, there he explicitly dismisses Kant’s tendency to reduce religion to a mere source of non-rational motivations or otherwise rational ‘content’ – something that, as I have mentioned, Habermas himself tends to do elsewhere. Instead, he focuses on reading Kant’s accounts of the ‘highest good’ and the ‘ethical community’ as expressing a necessary ‘faith’ that individuals’ right actions will combine to improve the social world. He calls this a ‘faith’, rather than a moral ‘duty’ or a political ‘law’, because such a successful combination of actions is too complex to be pursued by individuals or enforced by a state. Yet Habermas insists that it is nonetheless necessary if we are not to ‘despair’ of the possibility of coordinating our social actions rationally. Indeed, he presents this ‘faith’ precisely as the orientation to pursue objective rational justification, by moving from the particular terms employed in our existing communities towards ever more universal ones. That this orientation cannot be established rationally, and yet is necessary for rational coordination, is crucial. For it follows not only that this orientation exceeds the rational ‘content’ of religion – as Habermas emphasizes, Kant can reveal it only through Christian doctrine – but also that rational justification in general, and political justification in particular, rest on a non-rational orientation towards coordinating actions in this way. In other words, for Habermas ‘knowledge’ rests on a ‘faith’ in its own possibility.

These claims about the contingency and the faith of Habermas’s secular sense of political justification clearly shed a new light on that sense. In particular, they imply that, for Habermas, both rationality and religion promise shareable terms with which political action can be coordinated, and thus in this regard have the same political structure or function, for those oriented towards them; but that neither rational nor religious terms possesses ultimate authority, such that either rationality or religion could claim priority over the other. In other words, if our social actions are to be coordinated in shareable terms, then both rationality and religion promise such terms, but – normatively, at least – the ‘choice’ between them is simply a leap of faith. Habermas’s secular framework of political justification, like his theory of rational justification generally, thus appears as merely the internal elaboration of an unjustifiable faith in the possibility of such rational justification, proposed merely as an alternative to other, equally unjustifiable faiths in the possibility of coordinated action, such as religions.

Also worth emphasizing is that, as Habermas presents them, these insights into the contingency and non-rational ‘faith’ of rational justification cannot be ‘learnt’ by rationality alone, but only through an engagement with religion. In this, these insights contravene Habermas’s otherwise strict distinction between rationality and faith, by implying that religion can provide truths, or knowledge, not accessible to rationality alone, as well as the orientation to truth, or knowledge, that is presupposed but cannot be provided by rationality alone. Thus, on Habermas’ account, not only do rational justification in general and political justification in particular rest on a contingent ‘faith’ in their own possibility, but this cannot be understood or provided without engaging with non-rational religion. In other words, for Habermas, rationality, and with it political justification, cannot be fully understood or realized without religion.