Jürgen Habermas is perhaps the leading contemporary defender of the ‘unfinished project’ of modernity and Enlightenment reason. His work is often contrasted with post-structuralist thinkers who are typically cast as intellectual enemies of the Enlightenment. Habermas’s defence of the political and philosophical legacy of the Enlightenment has long been a benchmark against which political theorists have measured their own positions. For example, Jean-François Lyotard claimed that Habermas’s discourse-theoretical approach, with its ‘search for universal consensus’, is outdated because it denies the incommensurate and ‘heteromorphous’ nature of the postmodern social bond (Lyotard, 1984, pp. 65, 66). In response to such critiques, Habermas labelled various postmodernists in the 1980s as ‘young conservatives’, because in his view they rejected the progressive side of Enlightenment reason, and he accused them of falling into a ‘performative contradiction’ (Habermas, 1987, p. 13). Habermas presented the contributions of thinkers such as Jacques Derrida as inconsistent because they tend to deny the epistemological and normative presuppositions they use to make their arguments. More recently, a number of prominent political theorists in the Anglo-American academy have differentiated their ‘agonistic’ conceptions of democracy from Habermas’s project of discourse ethics. Perhaps the most critical is Chantal Mouffe, who maintains that Habermas puts liberal democracy at risk because his theory of discourse ethics supposedly overburdens contemporary citizens with a necessity to strive towards rational consensus. She says that the ‘very condition of consensus’ and ‘the process of argumentation’ of Habermas’s theory is ‘fatal for democracy’, because it eliminates ‘pluralism from the public sphere’ (Mouffe, 1999, pp. 47, 52).

Various commentators have examined the relationship between Habermas and Mouffe, or between deliberative and agonistic forms of democracy, and have argued that their positions are incommensurate.Footnote 1 In some ways this is understandable, given that Habermas and Mouffe draw on different philosophical traditions. Mouffe invokes Claude Lefort’s conception of the ‘democratic revolution’ to explain political modernity and Carl Schmitt’s distinction between friend and enemy, which for her represents the ‘specificity of the political’ (Mouffe, 1993, p. 2; 2000, p. 101). For (Mouffe, 2005, p. 11), the political is ultimately the realm of decision and conflict, not discussion and consensus. She also draws upon Antonio Gramsci’s notion of hegemony, a form of power that entails coercion and consent (Mouffe, 2000). Habermas (1989, p. 137) is critical of ‘decisionism’ because it reduces democracy to elite rule. Drawing upon Kant’s and Hegel’s theories of modernity, Habermas’s theory of communicative action emphasises instead the capacity of citizens to transcend their particular self-interest through the communicative generation of norms of conduct that are generalisable.

Every political theory invokes a ‘set of fundamental’ assumptions about ‘the relation of humans to themselves, to others, and to the world’, and these assumptions inevitably generate political differences (Connolly, 1995, p. 9). However, the agonistic critiques of discourse ethics have focused so exclusively on the philosophical assumptions of Habermas’s approach that the political similarities between Habermas and Mouffe have been largely overlooked. In order to bring out these similarities, I adopt a hermeneutic strategy that partially suspends the philosophical differences between Habermas and Mouffe, and focuses instead on what they have to say about democratic politics. This is not to deny the importance of the philosophical differences between them. Indeed, Habermas’s transcendentalism, his communicative understanding of consensus and his account of modernity as a ‘project’ that is realised dialectically are clearly at odds with Mouffe’s radical anti-foundationalism, her understanding of consensus as a hegemonic construct and her presentation of modern democracy as inherently paradoxical. Nevertheless, on closer inspection, we will see that these ontological differences do not actually generate radically different conceptions of democratic practice. In fact, I show that despite these philosophical differences, Habermas and Mouffe share a similar political imaginary, one that bears distinct marks of the republican tradition.

The aim of this article is therefore to push beyond the agonistic-deliberative polarity by showing how Habermas and Mouffe are both committed to a broadly critical republican vision of collective self-determination. Mouffe’s agonism invokes the work of republican thinkers such as Michael OakeshottFootnote 2 and Quentin Skinner, and I will show that on closer examination, there are similar republican aspects underpinning Habermas’s discourse ethics. By bringing their work into the framework of critical republicanism, we shall see that they provide valuable resources for theorising the normative conditions of the modern democratic political community, and also how best to address the challenges of pluralism. Indeed, both Habermas and Mouffe advocate the extension of the principles of liberty and equality to greater areas of social life, and they both argue that constructive disagreement stemming from cultural pluralism is central to the flourishing of the democratic public sphere and vital to creating a more inclusive and participatory polity.

The article commences with a brief explanation of three key features of critical republican theory: the importance of civic accord, the value of conflict and the value of collective self-government. Throughout the article, I demonstrate that these features are also central to Mouffe’s and Habermas’s political framework. In the next section, I show that Mouffe’s dismissal of discourse ethics as a ‘moral’ rather than ‘political’ theory does not hold up to scrutiny, because for Habermas morality represents a form of civic solidarity that provides a space where groups and individuals can pursue their particular conceptions of the good. This is actually similar to Mouffe’s use of Oakeshott’s notion of societas. In the following section, I argue that, contrary to Mouffe’s claim, Habermas recognises the inevitability of conflict, and his emphasis on consensus needs to be understood in terms of the communicative regulation of strategic action rather than as an overly idealised substitute for strategic forms of action. After this, I examine Mouffe’s criticism that Habermas prioritises democratic self-rule before individual rights, and I show that this line of critique does not make sense, because the same priority is true of her own agonistic conception of liberal democracy. Indeed, both Habermas and Mouffe reject the liberal conception of the foundational status of individual rights, and both share the republican ideal of collective self-government as a condition of individual rights. I conclude that Habermas and Mouffe both offer valuable insights for a theory of critical republicanism and have much to offer in contemporary discussions about how to critically engage the challenges of value pluralism.

Critical Republicanism

Republicanism in the broadest sense is a tradition of political thought that originates in Roman antiquity and resurfaces in the context of the Renaissance and the American War of Independence. The tradition includes thinkers such as Cicero, Livy, Niccolò Machiavelli, James Harrington, James Madison, Thomas Jefferson, Hannah Arendt and John Pocock. The term republicanism is a combination of two Latin words, res (matter, thing) and publica (public), so together referring to the ‘public matter’ or the ‘public thing’. Historically, republican theory has encompassed a variety of features including an emphasis on popular sovereignty, the rule of law, a mixed government, a deliberative senate, a popular assembly and an elected executive (Sellars, 1998, p. 8). Nevertheless, classical republicanism is often criticised for its imperialism, and its emphasis on masculine virtues of warlike glory. Indeed, despite the fact that all laws and elections of magistrates were dependent on the popular vote, it is also true that the Roman republic was nonetheless an oligarchy run by a narrow ruling class. However, contemporary republican thinkers rework the tradition to develop a more egalitarian theory appropriate to present conditions. For example, Quentin Skinner’s interpretation of Renaissance republicanism (1998, 2002), Philip Pettit’s democratic republicanism (1997) and Cécile Laborde’s critical republicanism (2008) all advocate a non-hierarchical and inclusive conception of economic, political and social life (see also Laborde and Mayor, 2008). Indeed, (Laborde, 2008, p. 4) draws on the Frankfurt School in order to argue that critical republican theory begins with an immanent critique of existing institutions that identifies forms of domination and oppressive practices in order to transform them.

Like classical republicanism, ‘critical republicanism’ seeks to construct the common good of the people and to prevent self-interest, corruption and tyranny from prevailing. However, critical republicans champion the ‘promotion of universal democracy’ and reject the elitist, masculine and imperial assumptions of the classical tradition (Laborde and Mayor, 2008, pp. 14, 17, 19). They acknowledge and accept moral individualism, ethical pluralism and individual rights as constitutive of modern societies. Critical republicans also challenge the predominant liberal accounts of politics because they argue that domination is still prevalent in late modern societies under a system of formal rights (Laborde and Mayor, 2008, p. 16). In a liberal state, one might be free to pursue one’s interests without interference yet still remain subject to forms of domination or dependent on arbitrary forms of power and authority. For example, a bearer of legal rights can still be hindered in her ability to act freely if she is in a subordinate position in any professional or personal relationship. Indeed, critical republicans stress independence from domination and arbitrary forms of power as the basic aspiration of freedom (Pettit, 1997; Skinner, 1998). Thinkers like Skinner point to the need to challenge not only political domination, but also economic, domestic and social forms of domination (see Pettit, 1997, p. 343). The republican language of liberty as non-domination has great scope and emancipatory potential, and freedom as non-domination is never guaranteed by the state, but needs to be maintained through active citizenship in the public realm.

Three key features of this new critical republicanism also align with the core of Habermas’s discourse ethics and Mouffe’s agonistic democracy. First, critical republicanism acknowledges the value of religious and cultural pluralism, and understands the dangers of fragmentation, which, if ignored or repressed, have the potential to undermine the basis of the polity (Skinner, 2002, p. 25; Laborde, 2008, p. 2). In contrast to liberal individualism, critical republicans place strong emphasis on addressing civic discord – which as we will see is not to be confused with disagreement or contestation – by ensuring that an identification with a civic bond, or a common ethos of engagement, takes precedence over other self-interested considerations (Skinner, 2002, p. 24). Skinner (2002, p. 25) shows that the Roman orator Cicero and pre-humanist writers in the Italian city states understood the importance of preserving concord as a condition of the public good. This is not to be understood as an agreement on substantive values, but rather adherence to a mode of civic conduct that conditions one’s interactions with others and the pursuit of individual ends. In her discussion of the hijab controversy in France, for example, Laborde emphasises not just the importance of good laws and institutions but also the importance of the quality of relations between citizens. She stresses the role of social norms and civic attitudes as a precondition for addressing inequality and injustice (Laborde, 2008, p. 12). Indeed, critical republicanism requires ‘that people willingly share in practices of cooperation (such as wealth distribution), to be able to make compromises for the sake of the common good’ (Laborde and Mayor, 2008, p. 14).

Second, like both Habermas and Mouffe, critical republicans acknowledge the value of non-violent conflict in the public realm. Indeed, they regard contestation and disagreement as a productive force by which to strengthen a free state rather than as something that needs to be repressed, overcome or pacified. Skinner (2002, pp. 153–154) shows that Machiavelli saw the lack of ‘serenity’ and the domestic unrest and tumults of Rome as indispensable to the freedom of the republic. Machiavelli shows that conflicts between different classes were necessary for the maintenance of liberty, because they prevented the wealthy from dominating the whole society. We will see below that the same idea remains relevant today as a response to the challenges of pluralism.

Finally, as we have said, critical republicanism understands liberty as ‘non-dependency’ or ‘non-domination’, which is different from the liberal notion of freedom as ‘non-interference’ (Pettit, 1997, p. 40; Skinner, 1998). With this formulation of liberty as independence, critical republican thinkers challenge the dichotomy between positive and negative liberty. Negative liberty is associated with the liberal tradition and – following Hobbes – liberty is understood as the absence of external constraint. Positive liberty is associated with the work of Aristotle and is associated with self-mastery and the pursuit of a substantive notion of the good. Skinner interprets Machiavelli’s neo-Roman conception of liberty as offering a third conception of liberty that differs from each of these options. Skinner does not subscribe to an Aristotelian or perfectionist conception of liberty. Instead, on his account, liberty is simply a means by which to challenge arbitrary forms of power, and thereby to continually ensure that one remains in a position of freedom as ‘non-dependence’ (Pettit, 1997; Skinner, 1998, 2002). However, (Skinner 2002, p. 9) also shows that the neo-Roman idea of freedom contrasts with the Hobbesian conception of negative liberty, where individuals hand over the right to govern themselves to a sovereign body tasked with the protection of their individual rights. Indeed, Machiavelli and other Renaissance republican writers emphasise the need for the citizens to remain ‘the sole power of making laws’ so that ‘all individual members of the body politic – rulers and citizens alike – remain equally subject to whatever laws they choose to impose upon themselves’ (Skinner, 1998, p. 74). In other words, there is an intrinsic link between liberty as non-domination and the need for collective self-government. We will now turn to Habermas and Mouffe, and we will see that these three key components of critical republican politics are also at the core of their respective political theories.

Morality as Civic Solidarity

For (Mouffe, 1993, p. 66; 2000, p. 69), one key challenge of contemporary pluralist societies is the need to construct a form of political association that does not postulate a substantive notion of the good, and yet is able to create a civic bond between citizens of diverse backgrounds and beliefs. She is critical of both liberal and communitarian accounts of modern politics. (Mouffe 1993, pp. 61, 64) argues that liberal theories – such as those of John Rawls and Ronald Dworkin – are inadequate because they are based on a mythical notion of a morally ‘unencumbered’ self, whereas communitarian accounts are grounded in a substantive notion of the good that is equally untenable in heterogeneous societies. By way of alternative, Mouffe invokes Michael Oakeshott’s concept of societas in order to theorise the basis of the democratic political community. Societas ‘designates a formal relationship in terms of rules’ between citizens, rather than a ‘teleological form of association’ based on the satisfaction of a ‘substantive common action’ or purpose, which Oakeshott calls universitas (Oakeshott, 1975, p. 205; Mouffe, 1993, pp. 66, 67). Following Oakeshott, Mouffe says societas binds people from different forms of life into an identifiable association, but one that retains the ‘multiplicity of voices that contemporary societies encompass’ (Mouffe, 1993, pp. 66, 67; 2000, pp. 97, 100). In other words, individuals from different religious, cultural and ethical communities identify with a civic identity that does not privilege a particular individual or group’s interest over any of the others. However, this is more than the constitutional framework of a minimal liberal state that regulates differences by relegating them to the private sphere, because individuals not only recognise the authority of conditions of acting in pursuing their own ends, they also recognise that they have common public concerns.

According to Mouffe, citizens in contemporary Western societies share a set of principles specific to the liberal democratic tradition. Following Oakeshott, Mouffe calls this the respublica, which is a ‘common or “public” concern’ (Oakeshott, 1975, p. 201; Mouffe, 1993, pp. 66, 67). Respublica is a practice that consists of a ‘complex set of rule like prescriptions which do not prescribe satisfactions to be sought or actions performed but moral considerations specifying conditions to be subscribed to in choosing performances’ (Oakeshott, 1975, p. 183; Mouffe, 1993, p. 67). Invoking Rawls’s terminology, she says that in a societas there exists a certain priority of the right over the good. However, for (Mouffe, 1993, p. 68), the ‘principles that specify the right, the respublica, are not conceived in a Kantian manner as in Rawls, but [in] a Hegelian way’. By this she means – again following Oakeshott – ‘to be associated in terms of the recognition of the respublica is to enjoy a sittlich relation’ (Oakeshott, 1975, p. 259; Mouffe, 1993, p. 68). A sittlich relation suggests something more than simply being related to another person via a formal legal category or abstract relation such as universal rights. It is a ‘general mode of behaviour’ that appears as ‘custom’ or norms of conduct (Hegel, 1991, p. 195). Citizens identify with these norms as part of a collective ethos or sense of community. The idea of Sittlichkeit is central to republicanism and resonates strongly with the notion of respublica. It refers to the obligations an individual has to a specific socio-historical community, and is not an abstract individual obligation to a transcendental principle like Kant’s categorical imperative.

Mouffe’s conception of societas and respublica provide crucial insights for addressing questions of social cohesion in a pluralistic society. In effect, what she offers is a third way between liberalism and communitarianism, which would allow for the rebuilding of public life without subsuming value differences under a communitarian conception of a common good. This might enable citizens with very different value systems – traditional Christians or Muslims or secular humanists – to identify sufficiently with one another as citizens to enable them to adopt relations of civility while retaining their substantive differences.

Mouffe (1993, p. 113; 2005, p. 5) maintains that the ‘ethics of the political … should be distinguished from morality [and moral philosophy], which concerns [itself with] individual action’. She contrasts her republican-based understanding of the political with what she calls Habermas’s moral approach to politics. According to (Mouffe, 1993, p. 10; 2000, p. 66), Habermas’s Kantian theory of discourse ethics seeks to find a viewpoint standing above politics. In her view, Habermas’s principle of universalisation is a moral agreement rather than a political relationship, one that provides no room for rational disagreement (Mouffe, 2000, pp. 66, 89; 2005, pp. 121–122). In addition, she argues that the procedures of discourse ethics cannot work properly, because, in marked contrast to the republican tradition, they lack a commitment to a civic bond or some ‘specific form of ethos’ (Mouffe, 2000, p. 69). In other words, she maintains that Habermas’s project of discourse ethics fails to offer a Sittlichkeit because he isolates politics from any embedded ethical norms or practices. Mouffe (1992, p. 277) concludes that this approach inevitably leads to a conception of the citizens as a ‘mere legal status, indicating the rights an individual holds against the state’ rather than a celebration of an active and participatory form of citizenship.

Mouffe touches on some important differences between her perspective and Habermas’s. Nevertheless, she overlooks the republican elements in his work and therefore fails to see the similarity in their respective approaches. Mouffe is correct to point out that Habermas’s discourse ethics is a moral theory. He explicitly acknowledges this (Habermas, 1990, pp. 120–121, 195–199). However, she is mistaken in accusing Habermas of reducing politics to a Kantian conception of personal morality (Mouffe, 2005, pp. 121–122). Elsewhere I have made the case that Habermas does not use the term ‘morality’ like the Kantian conception of ‘good will’, which refers to the motives and intentions of individual actors (Khan, 2012a, 2012b). The moral dimension of discourse ethics is not a command addressed to individual agents. It constitutes a civic bond between different individuals and groups, whereby the validity/legitimacy of the norms that ought to regulate their mutual interactions is established under conditions of reciprocity (Habermas, 1998, p. 217). Like Mouffe, politics for Habermas is not the pursuit of a common purpose or a vision of the good life (what Oakeshott calls universitas). Instead, like Oakeshott’s notion of societas, discourse ethics establishes a civic space where diverse individuals and groups pursue their own particular notions of the good life, while adhering to and constructing shared norms of conduct.

Habermas is well aware that modern societies are characterised by an irreducible ‘pluralism of value orientations’ (Habermas, 1984, p. 180; 1990, p. 76; 1996, pp. 64, 473; Mouffe, 1993). He says each form of life has its particular ‘self-description of itself, its persons and groups and thus their identities and life projects’, and he calls this a diversity of ethical forms of life, or ‘ethics’ (Habermas, 1993, p. 59). Like Mouffe, Habermas recognises that for plurality to flourish limits must be placed on legitimate forms of conduct. In order to prevent a single culture from dominating the public sphere, discourse ethics seeks to establish a universal principle by which the validity of contested norms can be adjudicated (Habermas, 1996, pp. 64, 473). For Habermas, the limits are established by the discourse principle that states that ‘only those norms can claim validity that could meet with the acceptance of all concerned in practical discourse’ (Habermas, 1998, p. 41). This principle ensures that a ‘norm is valid [only] when the foreseeable consequences and side effects of its general observance for the interests and value-orientations of each individual could be jointly accepted by all concerned without coercion’ (Habermas, 1998, p. 42). In other words, diverse ethico-political practices within a given society are legitimate only if they do not violate the moral principle (Habermas, 1998, p. 227).

In contrast to Mouffe’s theory, the discourse principle establishes an impartial ‘universalistic point of view’ from which to adjudicate contested norms (Habermas, 1990, p. 122; 1996, pp. 108–109). Habermas (1990, pp. 121, 196–197) also describes discourse ethics as de-ontological and formalistic. The structure of discourse ethics privileges the right or the ‘just’ over any particular conception of the good life (Habermas, 1990, pp. 120, 210; 1993, p. 2). For Habermas, a moral procedure is a form of association that is ‘not based on a shared conception of the good, but on a more abstract form of recognition contained in the idea of free and equal consociates under law’ (Baynes, 1995, p. 221). Nevertheless, the Habermasian discourse principle does not regulate individual will in the manner of the categorical imperative. Instead, the principle of universalisation performs a similar function to Mouffe and Oakeshott’s notion of respublica in that it provides rules that prescribe certain ‘norms of conduct’ to be followed in the pursuit of one’s individual chosen ends (Mouffe, 1993, p. 67). Indeed, Habermas does not understand these norms as a set of a priori rules, as Mouffe implies, but rather as a ‘practical discourse’ or an embedded ‘public affair, practiced inter-subjectively by all involved’ (Habermas, 1990, p. 198).

Mouffe (1993, p. 68) acknowledges that Oakeshott uses the term ‘moral’ to describe societas as a non-instrumental form of association, and she says that he uses the term to refer to an essentially ‘ethico-political’ relation. Yet the same can be said about Habermas’s use of the term ‘moral’. It describes an embedded civic relationship rather than an abstract formula for directing personal moral duty.Footnote 3 This is an important similarity between Habermas and Mouffe, which is missed by the majority of commentators, and revealed in the proximity of their respective positions on Oakeshott.Footnote 4 Habermas’s references to universality may appear to conflict with this emphasis on the situated ethos or the ‘common ethical–political dimension’ of a distinct political community. This is because the term ‘universal’ might be taken to suggest something that is valid for all people at all times, whereas the notion of societas clearly points towards the specific forms of civil association characteristic of distinct communities. However, (Habermas, 1998, p. 217) qualifies his use of the term universal and says that a universal moral norm is valid only for those affected by it. Indeed, as (McCarthy, 1990, p. xiii, n. 8) points out, the universal for Habermas is distinct from the transcendental and ‘need not be based on a priori reasoning or pretend to infallibility’. It is better understood as something like Hegel’s ‘concrete universality’, which is embedded in a particular community. For (Habermas, 1990, p. 89), relations of reciprocity (the presuppositions of argumentation) are context-transcending in the sense that they are said to be implicit in the fact of human communication. This clearly differs from the epistemological groundlessness of Mouffe’s anti-foundationalism. We will come back to this point below (Habermas, 1984, p. 95; 1990, pp. 82, 89). However, for Habermas the context-transcending rules of argumentation always need to be embedded in a particular historical community.

Discourse ethics is primarily concerned with the democratic determination of civic – or ‘moral’ – norms that are agreed to by all concerned within a contextually specific community. This emphasis is particularly evident in Habermas’s critique of the cosmopolitanism of theorists such as David Held, who advocates a form of global democratic governance (Habermas, 2001b, p. 107). Unlike Mouffe, (Habermas, 1996, pp. 514, 515) does defend the idea of cosmopolitan principles of human rights, and he embraces the ideals of ‘world citizenship’ and a ‘world public sphere’. However, like Mouffe, he nonetheless argues that an ‘inclusive community of world citizens’ is not a viable project, because it would lack the sort of ‘civic solidarity’ that emerges within the nation-state or, on his (and Mouffe’s) account, from supranational states such as a federal Europe.Footnote 5 According to Habermas, a global democracy would be unable to construct ‘an ethico-political self-understanding’ from the multiplicity of worldwide traditions and value orientations, and it would consequently be unable to ‘generate any normative cohesion’ (Habermas, 2001b, pp. 107, 108). He says that any such democracy would be one only in name, because it would be limited to a mere legal category rather than being a civic identification (Habermas, 2001b, pp. 107, 108). Indeed, in Habermas’s account a democratic political community needs to be based on something more than a constitutional document or a set of legal rules. It also requires a ‘common political culture’ that binds people together through common moral norms, and these are at the core of his notion of ‘constitutional patriotism’ (Habermas, 2001b, p. 745). In other words, like Mouffe, Habermas seeks to find a republican midway point between the liberal emphasis on abstract rights and the communitarian emphasis on a fully contextualised conception of the good. In a position that is very similar to Mouffe’s, his emphasis is on the need to combine mutual identification with an acknowledgement of differences, in order to bring the citizens of contemporary pluralistic societies together in an acknowledgement of legitimate forms of authority and commonality.

The Positive Value of Conflict

The idea that conflict is an inevitable – and in some forms a valuable – aspect of modern democratic politics is a prominent theme in Mouffe’s agonistic theory of democracy. Mouffe claims that ‘antagonism’ is the ‘essence of the political’, and this is clearly at odds with Habermas’s argument that reciprocity and mutual understanding are implicit in human communication. She concludes from this fundamental difference that Habermas’s theory cannot account for conflict and antagonism (Mouffe, 2000, p. 83; 2005, p. 89). However, in this section, I show that Habermas’s project of discourse ethics also recognises the inevitability and value of conflict, and that Mouffe is wrong to suggest that his work is exclusively concerned with the politics of consensus. On the contrary, one can see in his work an aspiration towards the communicative regulation of conflict and strategic action. Again, the similarities between the two thinkers are clear and pronounced, and they both share the republican insight into the positive value of conflict.

Mouffe accuses Habermas of failing to grasp the nature of ‘the political’ and therefore for failing to adequately address the question of ‘division and antagonism’ (Mouffe, 1993, pp. 1, 69). Following Schmitt, she equates ‘the political’ with the inherently conflictual dimension of human relations. In her view, Habermas’s ‘moral’ approach fails to see that ‘violence is ineradicable’, and therefore it can only ever be sublimated rather than eliminated or overcome (Mouffe, 2005, p. 89). Mouffe understands agonistic democracy as a set of practices that consist in challenging hostility and in ‘trying to defuse the potential antagonism that exists in human relations’ (Mouffe, 1993, p. 3; 2000, p. 101). Agonism refers to constructive and non-violent relations between adversaries, which are contrasted to antagonism, or potentially violent conflict between enemies. The achievement of agonism rather than antagonism depends on societas, a relationship in which citizens agree to abide by the democratic rules of the game. The temporary consensus established through a collective identification with societas facilitates the agon between diverse subject positions. Adversaries are ‘friends because they share the common symbolic space, but they are also enemies because they want to organise’ the common symbolic space in a different way from one another, although through non-violent means (Mouffe, 2000, p. 13).

In Between Facts and Norms, Habermas starts from premises similar to Mouffe’s by stating that in modern secular society normative orders cannot be guaranteed by appeal to divine or metaphysical authority (Habermas, 1996, p. 26; Mouffe, 2000, p. 18). Habermas says that, in the absence of a transcendental power, individuals are likely to evaluate events ‘in light of their own preferences’ (Habermas, 1996, p. 27). Violence and conflict are therefore a real possibility, given the plurality of beliefs and preferences characteristic of modern societies. According to (Habermas, 1996, p. 26), modern subjects either try to resolve their differences through communicative action, in which they act together to reach an agreement on norms of interaction (not on substantive ends), or they shift to strategic action, in which they act according to their particular ends. In his recent work, Habermas indicates that he considers the pure ideal of communicative action to be unrealisable in practice. For example, he emphasises the importance of bargaining, where participants agree to a particular norm for different reasons through a negotiated consensus (Habermas, 1996, p. 166). Indeed, (Habermas, 1996, p. 25) says that ‘compromises make up the bulk of political processes’. The discourse principle remains relevant, but it is brought to bear indirectly through setting the conditions of the bargaining process.

As Habermas sees it, the only way to prevent a situation of escalating conflict – what Mouffe calls relations of antagonism – is for inter-subjective actors to ‘come to some understanding about the normative regulations of strategic actions’ (Habermas, 1996, pp. 26–27). In other words, Habermas’s ‘inter-subjective’ agreements ‘consensually establish suitable socially integrating constraints on strategic action’ by binding citizens to certain norms of conduct in their interactions with each other. Indeed, ‘a socially integrating force’ must be developed that ‘(imposes) obligations on the addressees’, which is ‘possible only on the basis of inter-subjectively recognised normative validity claims’ (Habermas, 1996, p. 27). However, these norms do not seek to transcend or eliminate strategic action altogether, as it is often assumed (see, for example, Bickford, 1996; Mouffe, 2000). According to Habermas, generalised norms must be presented in such a way that the strategic actor is able to identify with the norm, and to shape his/her strategic behaviour in line with the norm, and he stresses that those norms are only in play as long as they are continually re-enacted. In other words, Habermas argues that in the ‘cooperative regulation of their common life’, individuals renounce violence and also ‘concede one another the right to remain strangers’ (Habermas, 1996, p. 308). In discourse ethics, subjects create a civic community through deliberation by coming to an uncoerced agreement regarding the norms that ought to govern the political community, and this is designed to facilitate a more constructive form of pluralism. Indeed, we could go so far as to say that the basic objective of Habermas’s politics is similar to Mouffe’s: he stresses the need to defuse ‘antagonism’ by transforming it into something ‘agonism’ (or normatively regulated strategic action).

Moreover, this is not only a late adjunct to Habermas’s theory. Indeed, at no point in the development of his work can Habermas be said to have underestimated the role of conflict in modern politics. This is also evident in his earlier writings, where he insisted that strategic action cannot be excluded from the political. In fact, in this regard, Mouffe’s criticism of Habermas parallels Habermas’s own criticism of Hannah Arendt. He argues that Arendt’s overly idealised concept of the republic ‘screens all strategic elements, as force, out of politics … [and] removes politics from its relations to the economic and social environment in which it is embedded’ (Habermas, 1977, p. 16). Discourse ethics does not seek to resolve conflict between different interpretations of the good life. In fact, Habermas’s theory of the ‘public sphere and the maintenance of democracy’ requires political contestation (see Markell, 1997, p. 394; Brady, 2004, p. 334). Unlike Rawls, Habermas insists that participants in discourse bring their particular values from ‘the lifeworld’ into the process of argumentation by presenting them as validity claims to be challenged (Habermas, 1996, p. 22). The objective of democratic politics for Habermas is not to agree on a substantive notion of the good, as is often supposed, but to arrive at some sort of temporary agreement about the norms that ought to coordinate the strategic action of groups and individuals in the pursuit of their particular conception of the good life.

Mouffe also recognises the importance of establishing consensus in the formation of the political community. A pluralist democracy requires a consensus between citizens on the ethico-political values of the community (res publica). However, in her view, every consensus necessarily involves exclusions and is therefore always only a temporary hegemonic stabilisation of power, entailing the mobilisation of passions and rhetoric (Mouffe, 1993, p. 4; 2000, p. 104).Footnote 6 Following Gramsci, Mouffe argues that every consensus is a provisional hegemony, a combination of ‘coercion plus consent’, whereas, on her view, the ‘procedural constraints’ of Habermas’s ‘ideal speech situation’ and his ideal of a rational consensus preclude ‘the possibility of contestation’ (Mouffe, 2000, p. 92). Mouffe implies that Habermas understands the democratic process as oriented towards a fully inclusive consensus, and some kind of permanent end point. Commentators sympathetic to discourse ethics have also levelled a similar critique at Habermas’s work (Bohman, 1994, p. 925; Rehg, 1994, pp. 75–76). However, Habermas’s notion of consensus is more complex than these critics acknowledge. At the constitutional level, Habermas, like Mouffe, seeks to establish a consensus on the values of civility that citizens ought to identify within their interaction with each other as members of a political community, and in the pursuit of their self-chosen ends. For (Habermas, 1990, p. 207; 1996, p. 107), a ‘rational consensus’ on moral values and norms is also always a temporary achievement that is subject to contestation, revision and refinement through practical discourse. Indeed, (Habermas, 1996, p. 9) recognises that any consensus on epistemological and moral claims is intrinsically fallible, and therefore it is always subject to the possibility of revision and change. The goal of political theory for Habermas, as well as for Mouffe, is to create ‘unity [on contested norms] in a context of conflict and diversity’ (Habermas, 1992, pp. 115–148; Mouffe, 2000, p. 101).Footnote 7 His acknowledgement that ‘there can be no inclusion without exclusion’ also steers away from the idea of a fully inclusive procedure, and resonates with Mouffe’s claim that ‘collective identities can only be established on the mode of an us/them distinction’ (Habermas, 2004, p. 7; Mouffe, 2005, p. 13). Indeed, (Habermas, 2001b, p. 107) says that any ‘political community that wants to understand itself as a democracy must at least distinguish between members and non-members’. Contrary to Mouffe’s claims, Habermas’s ‘regulative ideal’ does not aim to resolve the problem of antagonism or violence (Mouffe, 2000, p. 48). For (Habermas, 1996, p. 116), violence or conflict is only ever kept at bay as long as actors share norms and values of engagement, which create space for competing and diverse ethical forms of life.Footnote 8 In other words, Mouffe’s and Habermas’s alternative reflections on the nature of consensus do not produce radically different conceptions of democratic practice. Both Habermas’s and Mouffe’s theories of conflict resonate with the critical republican emphasis on overcoming discord – or violent forms of conflict – through civic participation and the cultivation of a common concern, and this is understood in turn as a condition of legitimate forms of conflict and pluralism.

Liberty before Liberalism

Berlin (1969) distinguished negative liberty (understood as the ‘absence of constraint’) from positive liberty (understood as self-mastery or the freedom to ‘lead one prescribed form of life’). He claimed that these are incompatible ways of understanding freedom, and he defended negative liberty, arguing that positive liberty could easily lead to forms of political authoritarianism (Berlin, 1969, pp. 118–172). He says that the pursuit of positive liberty can have disastrous implications, because the quest for self-realisation or the attainment of some ideal can lead groups of motivated citizens to coerce and manipulate others. In other words, Berlin provides a paradigmatic account of the liberal conception of the political, which understands liberty exclusively in terms of individual rights that ensure a space that is free from interference. However, as I have said above, critical republican thinkers challenge this dichotomy and put forward an alternative conception of liberty as ‘non-domination’, which is maintained through active citizenship (Pettit, 1997; Skinner, 1998, 2002). In this final section, I show that both Habermas and Mouffe emphasise the importance of the republican conception of freedom, understood as form of independence, which is ensured through self-government. They both regard civic participation as central to guaranteeing private rights and the maintenance of democratic legitimacy.

Drawing upon C.B. Macpherson, Mouffe affirms the importance of both liberty and democratic equality as two fundamental principles in modern liberal democratic politics. Moreover, she stresses their paradoxical relationship, because the principle of individual liberty potential conflicts with the majoritarian principle and the idea of popular sovereignty (Mouffe, 2000, p. 2). In contrast to liberal thinkers with their emphasis on the priority of individual rights, Mouffe draws upon Skinner’s work, in order to argue that citizens need to guarantee their individual liberty through active participation. Indeed, she suggests that democratic citizenship – expressed through the cultivation of civic virtues and devotion to the common good – is a prior condition of individual liberty (Mouffe, 1993, pp. 63, 69–70). Democratic participation brings diverse individuals and groups together and reinforces the legitimacy of liberal democratic institutions. The idea of active participation as a necessary condition of the maintenance of individual liberty is central to Mouffe’s conception of agonistic democracy.

Moreover, Habermas also argues that the legitimacy of democratic institutions rests on active citizenship and expressions of popular sovereignty. For him, popular sovereignty and individual liberty are ‘co-original’, meaning they ‘require each other’ (Habermas, 2001a, p. 767). Habermas says that individuals can only protect their private autonomy if they exercise their public autonomy and vice versa. Indeed, the idea of the importance of participation has been a central theme running throughout Habermas’s writings. In The Legitimation Crisis, Habermas argued that the modern state is potentially subject to systemic crises of legitimacy under late capitalism, because it cannot rely on traditional forms of legitimation and therefore needs ‘genuine participation of citizens’ (Habermas, 1988, pp. 36–37). However, he argued that in the increasingly ‘depolitized public sphere’ legitimation has been reduced to ‘civic privatism’ that entails ‘political abstinence’ and a concern with one’s ‘career, leisure and consumption’, and this tendency is reinforced by ‘democratic elite theories’ that effectively support the idea of the naturalness of the ‘capitalist economic society’ (Habermas, 1988, p. 37). To counter these developments, (Habermas, 1996, p. 467) insists that legitimacy needs to be maintained through the realisation of the active consent of the citizens, in their participation in political institutions in civil society. As (Habermas, 1996, p. 318) sees it, if citizens do not actively participate in the continuous recreation of the political community then there is a likelihood that ‘paternalistic monopolies’ will be consolidated that seriously undermine the process of democracy. In (Habermas’s, 1996, pp. 79, 454) view, paternalistic authorities are anti-democratic because they deny the addressees of the law the opportunity to be its authors, and they also make citizens ‘passive’ by conferring rights upon them. In republican terms, elite rule is illegitimate because it renders the citizen body dependent on the good will of their rulers.

In her evaluation of Habermas, (Mouffe, 2000, pp. 92–93) invokes Charles Larmore’s liberal critique of Habermas’s co-originality thesis. According to Larmore, Habermas’s claim that rights and self-rule are co-original is misleading because he ‘makes democratic self-rule a principle prior in status to that of individual rights’, and he therefore regards popular sovereignty as the ‘ultimate basis’ of the organisation of political life (Larmore, 1999, pp. 621, 613). It is curious that Mouffe chooses to deploy Larmore’s critique of Habermas’s co-originality thesis, because this liberal anxiety could equally be levelled at her own republican conception of democracy. For (Mouffe, 1999, p. 42), too, the enjoyment of individual liberty is grounded in active democratic citizenship. In other words, the key point of contention here is not between Habermas and Mouffe, both of whom stress the ultimate priority of democratic participation as a condition of individual rights, but rather between the republican stress on self-government and Larmore’s liberal conception of the absolute priority of individual rights. In my view, both Habermas’s and Mouffe’s approaches are superior to Larmore’s liberal foundationalism because they both avoid the trap of naturalising individual rights. Their republican idea of participation as necessary to guaranteeing private rights is an attractive feature of their respective conceptions of the political community, and Mouffe misses this point of connection between them. Indeed, Habermas is clear that rights have to be politically constructed through active participation; otherwise the danger is that citizens treat paternalistically imposed rights as ‘pre-given moral facts’ (Habermas, 1996, p. 454).

This does not mean, however, that either Habermas or Mouffe does not embrace popular sovereignty uncritically. The defence of certain civil liberties acts as a safeguard against expressions of popular sovereignty that might contravene the rights of minorities. In Habermas’s work, these liberties are reflected in the presuppositions of argumentation, which are the immanent rules of regulation expressed in communicative action. By way of contrast, (Mouffe, 2000, p. 5) defends individual rights as part of the inheritance of the liberal democratic tradition, which is defined in terms of a constant articulation of the paradoxical principles of liberty and equality, and in her view there can be no appeal to a meta-theoretical grounding external to this tradition. Indeed, for (Mouffe, 2000, p. 5) the relationship between liberalism and democracy is a contingent historical phenomenon that evolved in the eighteenth and nineteenth centuries. This is in contrast to contemporary liberals like Ronald Dworkin, for whom individual rights are foundational and thus always necessarily trump popular sovereignty. Mouffe’s position is also distinct from so-called radical democrats like Slajov Žižek, who advocate the suspension of liberal rights in the interests of revolutionary social change. Like the critical republicans, both Habermas and Mouffe emphasise that the exercise of democratic citizenship is a means to an end, that civic participation enables citizens to protect their rights under formal law and this legitimates the social order. Active participation is necessary to counter political apathy and the crisis of legitimacy characteristic of actually existing democracies. The duty to participate in the public life of the community is a precondition of the enjoyment of personal liberty. Individuals can only protect their private autonomy (liberal rights) if they exercise their public autonomy (popular sovereignty) as citizens and vice versa (Habermas, 2001a, p. 767).

In other words, the real differences between Habermas and Mouffe on the issue of democracy and individual rights are philosophical, but not, it seems, political. Indeed, Mouffe’s criticisms are primarily targeted at Habermas’s understanding of modernity as a dialectical learning process. Habermas’s understanding of modernity is premised on his rationalisation of the lifeworlds thesis. He claims that as traditional societies evolve, cultures undergo a rationalisation process of social and cultural modernisation. For Habermas, modernity is marked by a reflexive form of reasoning where modern subjects increasingly acquire a capacity to evaluate the claims of their particular tradition, custom and habit in the ‘light of good reason’ (Habermas, 2001b, p. 133). In modern societies, morality (which is conventionally bound to tribal and religious authorities) is uncoupled from conventions and replaced by general principles of legitimacy such as those embodied in liberal democratic constitutional states (Habermas, 2001b, p. 152). In Habermas's view, this move from traditional to modern societies is a progressive and developmental learning process in which individuals and groups from different cultures can learn increasingly how to solve the problems of social coordination. Habermas is optimistic that modern citizens will progressively realise the principles of equality and liberty, or democracy and the rule of law, that is, if we conceive of constitutional government in the long run as a fallible ‘self-correcting learning process’ (Habermas, 2001a, p. 774).

By way of contrast, Mouffe draws attention to the inherently contingent nature of liberal democracy, and rules out the eventual reconciliation of the principles of individual liberty and popular sovereignty. For Mouffe, liberty and equality always remain in paradoxical tension with each other, and Habermas’s desire to eliminate this tension, she insists, is detrimental to the future of liberal democracy and puts democracy at risk (Habermas, 2001a, p. 768; Mouffe, 2005, p. 84; see also Honig, (2007). In her view, liberal democracy is not to be understood as part of a collective learning process supposedly embedded in the ‘project of modernity’; rather, liberal democracy is a historical achievement, and its provisional hegemony needs to be defended and enhanced by actively fostering identification with democratic values (Mouffe, 2000, p. 96).

Habermas’s dialectical understanding of modernity is clearly distinct from Mouffe’s conception of the liberal democratic paradox. However, Mouffe’s claim that Habermas’s dialectical approach threatens democracy does not hold up to scrutiny, because in terms of his political prescriptions, Habermas’s argument is again very close to Mouffe’s. He acknowledges that individuals cannot accept liberal democracies as their ‘fortunate inheritance from the past’ but rather they should regard them as something that must be carried forward through active participation (Habermas, 1996, p. 471; 2001a, p. 774). Indeed, his conception of the dialectic of modernity is not as deterministic as may first appear, because Habermas recognises that the unfolding of modernity is not a linear process, one that is steered by Geist or the ‘cunning of reason’. In much the same way as Mouffe, (Habermas, 1990, p. 208) stresses that the further expansion of liberal democracy cannot be left to the unfolding of history, but needs to be actualised and carried forward by active socio-political movements. In this respect, both Habermas and Mouffe insist on the agency of the people in ensuring the maintenance and the future flourishing of liberal democratic practices and institutions. Their positions on specific challenges, such as the difficulties associated with religious fundamentalism, are also very similar (see Olson, 2009). Habermas’s theory entails a critique of fundamentalism, because the fundamentalist claims ‘exclusiveness for a privileged way of life’ and ‘leave[s] no room’ for ‘reasonable disagreement’ or ‘reflection with the other world views with which [s/he] share[s] the same universe of discourse’ (Habermas, 1998, p. 224). Similarly, (Mouffe, 2005, p. 122) says that the liberal democratic framework can only recognise those individuals and groups that accept the constitutional and legal framework of a democratic political association. In other words, those individuals and groups who do not accept ‘agonistic (legitimate) forms of expression’ should not be tolerated, and thus are legally and symbolically excluded from the societas (Mouffe, 2005, p. 82). In other words, once again, Mouffe and Habermas arrive at broadly equivalent strategies for addressing intractable forms of conflict or antagonism despite their philosophical differences.

Conclusion

In this article, I have shown that Mouffe exaggerates the differences between her political prescriptions and Habermas’s approach. For both Mouffe and Habermas, pluralism is a vital and irreducible element of the modern public sphere. Both thinkers also theorise consensus in republican terms, as a set of rules, norms or practices that preside over the civic association. Contrary to Mouffe’s criticism, Habermas’s notion of consensus as a regulative ideal does not seek to eliminate conflict but rather to distinguish between legitimate and illegitimate expressions of conflict, similar to her distinction between ‘agonism’ and ‘antagonism’. Indeed, they both recognise the positive value of certain forms of conflict to the ongoing health and vitality of democratic systems. Moreover, they also reject the liberal priority of individual rights over participation, and they share instead the republican emphasis on the need for participation as a prior condition of individual liberty. In other words, despite their philosophical differences, Habermas and Mouffe provide remarkably similar viewpoints about how existing liberal democracies might be further developed in the direction of critical form of republicanism. The different elements of the critical republican approach that are explored in this article offer a pertinent response to the challenges associated with value pluralism and increasing inequality. Although there is no doubt much to be done in developing this perspective, and applying it to the challenges we face today, the acknowledgment of how Habermas and Mouffe can both contribute to this broad outlook, and especially in contrast to the liberal viewpoint, provides an important step in this direction.