Post-foundational Political Ontologies

Oliver Marchart defines post-foundationalism as the realisation of the fact ‘that there is no Archimedean point, no ethical or rational principle from which it is possible to found society’ (2022: 74). While metaphysics has traditionally been ‘the quest for such ground, which can assume different names: logos, idea, cause, substance, objectivity, subjectivity, will or, in more theological terms, the supreme being or God’ (2007: 23), post-foundational thought, in contrast, develops ‘a constant interrogation of metaphysical figures of foundation, such as totality, universality, essence, and ground’ (2007: 2). Although Marchart does not make this distinction, it is possible to draw on his theoretical coordinates to identify two different varieties of foundationalism. On the one hand, we would have a transcendent type of foundationalism, in which the foundation that gives meaning to the social is effectively located outside its own sphere. Don Herzog calls ‘foundationalist [ . . . ] any political justification [that] must be grounded on principles that are (1) undeniable and immune to revision and (2) located outside society and politics’ (1985: 6). The second clause of his definition refers to this type of transcendent foundationalism, the most recognisable example of which is the theological one. However, one could also define another type of immanent foundationalism, one in which the foundation remains ‘undeniable and immune to revision’ even though it is no longer located ‘outside society and politics’. One case of this immanent foundationalism would be found, for example, in the Noûs of the pre-Socratic Anaxagoras, who, according to Onfray, ‘approaches things as a philosopher for whom the cosmos reveals itself as a godless order that must be read in an immanent and never transcendent way [ . . . ] He considered Spirit, Intellect, Reason, Energy – the Noûs in Greek – to be the cause of everything, the first principle of the universe’ (2022: 20).

Marchart presents Laclau as a post-foundational thinker par excellence, alongside Jean-Luc Nancy, Claude Lefort and Alain Badiou. In chapter III of Hegemony and Socialist Strategy Laclau and Mouffe build a post-foundational theoretical framework by arguing that ‘the social itself has no essence’ (1985: 96) and that ‘we must consider the openness of the social as the constitutive ground, the “negative essence” of the existent’ (1985: 95). On the basis of this ontological framework, both authors develop a theory that thinks of political articulations in a hegemonic and discursive fashion. The affirmation that the social lacks essence comes along with the deconstruction of the traditional role of universals when it comes to thinking about society and politics. According to Laclau,

if one can talk about universality, it is only in the sense of the relative centralities constructed hegemonically and pragmatically [ . . . ] This means, of course, that there is no a priori centrality determined at the level of structure, simply because there is no rational foundation of History. The only ‘rationality’ that history might possess is the relative rationality given to it by the struggles and the concrete pragmatic-hegemonic constructions. (1989: 78)

Does it also make sense to present Castoriadis as a post-foundational thinker? In principle, this hermeneutical hypothesis would face two interrelated obstacles. The first, more superficial, has to do with what Marchart (2007) calls ‘left Heideggerianism’, which would represent the philosophical affiliation shared by ‘his’ four post-foundational authors (Nancy, Lefort, Badiou and Laclau). There is not an ounce of Heideggerianism in Castoriadis’ thought. To give just a couple of enlightening examples: already in the second part of The Imaginary Institution of Society he was very critical of ‘philosophers like Heidegger, who have endeavoured to ignore psychoanalysis (and sexuality, as well as society, power, politics)’ (1987a: 330). It was in ‘The “end of philosophy”’ where Castoriadis engaged in a more critical and detailed way with Heidegger and his philosophy. Although he conceded that ‘it would be silly to deny that Heidegger was one of the important philosophers of the twentieth century’ (1989: 9), he stressed again that

Heidegger's congenital blindness before the political/critical activity of humans [is] at the root of his adherence to Nazism and the Führerprinzip [leader principle]. This blindness is fittingly complemented by a seemingly equally congenital blindness concerning sexuality and, more generally, the psyche. Here we have the bizarre spectacle of a philosopher talking interminably about the Greeks, and whose thought draws a blank in the place of polis, eros, and psyche (1989: 4).

In spite of this blatant anti-Heideggerianism, Marchart himself mentions Castoriadis as one of those post-foundational thinkers who were left out of his book but who could perfectly well have been included, along with others such as Jacques Rancière, Julien Freund or Zygmunt Bauman (2007: 155), a list into which female thinkers such as Chantal Mouffe, Judith Butler or Wendy Brown could have also been incorporated. The second obstacle, of greater philosophical significance, has to do with the question of postmodernity. Castoriadis’ critical allusions to postmodernism are numerous. He disqualifies postmodernism for its eclecticism and nihilism (2007: 84; 1997b: 347). In a similar manner, he also accuses ‘deconstruction’, linked to the figure of Jacques Derrida, of being ‘one more manifestation of the sterility of our times’ (Morgin et al. 1995: 110). In his view, postmodernism ‘ends up glorifying eclecticism, covering up sterility and providing a generalised version of the “anything goes” principle’ (2001: 25). Since it is relatively common to use the terms ‘post-foundationalism’, ‘post-structuralism’ and ‘postmodernism’ interchangeably, as if all three were equivalent and synonymous, one might conclude that to present Castoriadis as a post-foundational thinker is perhaps hermeneutically illegitimate and unjustifiable. But let us take a closer, more nuanced look.

The project of ‘radical democracy’ that Laclau and Mouffe proposed in Hegemony and Socialist Strategy would be simultaneously modern and postmodern (Mouffe 1989). It would be modern insofar as it aims – in Lefortian terms – to deepen the ‘democratic revolution’ and develop its political project of ‘equality and freedom for all’ (1989: 34). It would be postmodern, however, insofar as it requires ‘abandoning the abstract universalism of the Enlightenment’, epistemological foundationalism and ‘the essentialist conception of social totality’ (1989: 44). Laclau describes the transition from modernity to postmodernity in the following terms:

If something has characterised the discourses of modernity, it is their pretension to intellectually dominate the foundation of the social, to give a rational context to the notion of the totality of history, and to base on the latter the project of a global human emancipation. As such, they have been discourses on essences and fully present identities, based in one way or another upon the myth of a transparent society. Postmodernity, on the contrary, begins when this fully present identity is threatened by an ungraspable exterior, that introduces a dimension of opacity and pragmatism into the pretended immediacy and transparency of its categories. (1989: 71-72)

Laclau and Mouffe's understanding of ‘the postmodern’ explicitly removes the ghosts of eclecticism, nihilism and sterility that frightened Castoriadis so much. Mouffe rejects the assimilation of the dichotomy modernity versus postmodernity with the dichotomy objectivism versus relativism and discredits the identification of her post-foundational theory with any kind of nihilistic approach (1989: 37). Laclau positions himself unequivocally along the same lines: ‘abandonment of the myth of foundations does not lead to nihilism [ . . . ] Humankind, having always bowed to external forces – God, Nature, the necessary laws of History – can now, at the threshold of postmodernity, consider itself for the first time the creator and constructor of its own history’ (1989: 79-80). To reflect on a conceptual difference between post-foundational positions such as those of Laclau and Mouffe and postmodern nihilisms such as those denounced by Castoriadis, Marchart coined a distinction between two varieties of non-foundational theories: post-foundationalism and anti-foundationalism. According to his definition, ‘post-foundationalism does not stop after having assumed the absence of a final ground, and so it does not turn into anti-foundational nihilism, existentialism or pluralism’ (2007: 14). Post-foundationalism ‘does not assume the absence of any ground; what it assumes is the absence of an ultimate ground, since it is only on the basis of such absence that grounds, in the plural, are possible’ (2007: 14). Marchart concludes that Laclau ‘does not fall into the trap of an extreme anti-foundationalism and equally attacks those Baudrillardian versions of postmodernism which assume, on the basis of a critique of foundationalism, the implosion of all meaning’ (2007: 151).

Having explained the distinction between post-foundationalism and anti-foundationalism, the way is clear to answer the question of whether it is possible to read Castoriadis as a post-foundational thinker. In view of the above considerations, it can be assumed that the answer will be positive. To test this hypothesis it will suffice, for the moment, to refer to some fragments that testify to the post-foundational character of Castoriadis’ political ontology. Already in one of his early texts, Castoriadis argued that the possibility of an ‘autonomous society excludes radically the idea of a transcendental source of the institution (of a law given by God or gods, by nature or even by reason, if by reason we mean a totality of exhaustive, categorical, atemporal determinations, anything but the very movement of human thought)’ (1980: 98). Later, in one of his best-known texts, he reaffirmed the same unambiguously post-foundational reasoning, in the same terms as Marchart, Mouffe and Laclau:

If the law is God-given, or if there is a philosophical or scientific ‘grounding’ of substantive political truths (with Nature, Reason, or History as ultimate ‘principle’), then there exists an extra-social standard for society. There is a norm of the norm, a law of the law, a criterion on the basis of which the question of whether a particular law (or state of affairs) is just or unjust, proper or improper, can be discussed and decided. This criterion is given once and for all and ex hypothesi does not depend upon human action. Once it is recognized that no such basis exists [ . . . ] and once it is also recognized that there is no ‘science’, no episteme or téchne, of political matters, the question of what a just law is, what justice is – what ‘the proper’ institution of society is – opens up as a genuine, that is, interminable, question. (Castoriadis 1997d: 281)

Finally, there is an important idea that is often repeated by Castoriadis and which structures his entire ontology: ‘being is chaos, abyss or the groundless’ (2009: 64). The Greek–French thinker understands this chaotic nature of reality as implying that

there is not, as a treasure that has been hidden or that is to be found, any ‘signification’ in being, the world, history, or our own lives. In other words, it entails acceptance of the fact that we create signification on the basis of the baseless, the groundless, that we too give form to the chaos through our thought, our action, our labour, our works, and that therefore this signification has no ‘guarantee’ external to itself (1997b: 344).

Ontologies of Signification

The starting point of post-foundationalism is the thesis that the social has neither a transcendent essence nor an immanent essence. The denial of the ultimate foundation of society has – taking up Herzog's definition – two consequences. First, the meaning of social practices does not derive from metaphysical principles whose establishment and development is independent of the thought and action of human collectivities. Castoriadis defines this situation as an abandonment or an overcoming of ‘heteronomy’, the form of institution of society that has governed most human societies and has been historically sanctioned by ‘inherited thought’. Secondly, the meaning of social practices depends on what Butler calls ‘contingent foundations’ (1992), whose political character determines that they cannot be considered as principles ‘undeniable and immune to revision’. The political ontologies of Castoriadis and Laclau and Mouffe are built upon the realisation that the meaning of social practices comes neither ‘from the will of God, nor from the nature of things, nor from the Reason of history’ (Castoriadis1976: 72). In addition to being post-foundational – or precisely because they are so – they are ontologies that assume the discursive character of social reality (according to Laclau and Mouffe's formulation in Hegemony and Socialist Strategy) or the imaginary dimension of social meanings (according to Castoriadis’ formulation in The Imaginary Institution of Society).²

Laclau defines his own ontology, as opposed to Badiou's mathematical ontology, as a rhetorical one (Laclau 2008: 6). That his ontology is rhetorical, symbolic or discursive does not mean that it is idealist or anti-materialist: ‘the fact that every object is constituted as an object of discourse has nothing to do with the question of whether there is a world outside thought, nor with the alternative realism or idealism’ (Laclau and Mouffe 1985: 108). Laclau and Mouffe reject the distinction between discursive and extra-discursive practices, conceiving the social as ‘the field of discursivity’, while pointing to ‘the material character of every discursive structure’ (1985: 108). In their conception, discourse cannot be reduced exclusively to language, as every discursive structure also articulates non-linguistic elements. Both authors assert in an oft-repeated example that ‘an earthquake or a falling of a brick is an event that certainly exists, in the sense that it occurs here and now, independently of my will. But whether their specificity as objects is constructed in terms of “natural phenomena” or “expression of the wrath of God”, depends upon the structuring of a discursive field’ (1985: 108).

In Castoriadis’ theory, what gives meaning to a society are its ‘imaginary meanings’ (God, the polis, the citizen, the nation, spirits, norms, laws, sin, virtues, human rights, etc.), which even in a certain sense are ‘more “real” than the “real”’ (1987a: 140). In his view, ‘reality’ for each society, can only be described within the network of institutionally legitimated meanings that the network itself defines’ (1992: 89). Castoriadis identifies in ‘inherited thought’ an illegitimate separation between the field of ‘the historical’ and the field of ‘the social’, which according to him share the same logics and must be thought together. The analysis of ‘the historical–social’ begins with the premise that ‘the institution of society is the institution of a world of meanings’ (1987a: 235) and ‘everything that is, in one way or another, grasped or perceived by society, has to signify something, has to be endowed with signification’ (1987a: 235); in short, ‘it always has a meaning’ (1987a: 235). We would thus find ourselves in front of an ontology of signification or an ontology of meaning: ‘the sole “problem” that the institution of society has to resolve everywhere and always [ . . . ] is the “problem” of meaning: creating a world [ . . . ] invested with signification’ (1997c: 383). Like Laclau and Mouffe, Castoriadis also does not reduce imaginary significations to exclusively linguistic phenomena, since ‘the full significance of a word is everything that can be socially stated, thought, represented or done on the basis of this word’ (1987a: 243). If Laclau and Mouffe resorted to the case of the earthquake to illustrate their discursive conception of the social, Castoriadis performs the same operation by alluding to an alternative example:

[The] transformation of the natural fact of being-male and being-female into the imaginary social signification of being-man and being-woman [ . . . ] refers to the magma of all the imaginary significations of the society considered. Neither this transformation itself, nor the specific tenor of the signification in question can be deduced, produced or derived on the basis of the natural fact, which is always and everywhere the same. (1987a: 229)

The Institution of the Social and the Political

Starting from a post-foundational and discursive political ontology, our three authors analyse the shaping and transformation of social processes on the basis of the interaction between their two dimensions: the instituting and the instituted. Castoriadis defines the ‘social–historical’ as ‘the union and the tension of instituting society and of instituted society, of history made and of history in the making’ (1987a:108). On the one hand there is ‘the social as instituted’, the set of imaginary social meanings in force at a given moment that structure and provide meaning to the practices that shape the collective life of a society; however, the social as instituted ‘always presupposes the social as instituting. “In ordinary times”, the social is manifested in the institution, but this manifestation is at once true and, in a sense, fallacious, – as in those moments in which the social as instituting bursts onto the stage and pulls up its sleeves to get to work’ (1987a: 112). Castoriadis conceives the action of the social as instituting – the process of the imaginary institution of society – as an exercise of true ‘ontological creation’, whereby existing imaginary social meanings are transformed or new ones appear to replace the old ones. In ‘archaic’ or ‘cold’ societies, the social as instituted predominates; in them ‘change would be marginal or simply non-existent, the essence of their life unfolding as stability and repetition’ (1987a: 185). At the other end of the scale, we find the moments of revolutionary transformation of society, those in which the social as instituting predominates at the peak of its intensity.

In Laclau and Mouffe's work, the Husserlian distinction between sedimentation and reactivation of social practices is in a sense analogous to Castoriadis’ distinction between the social as instituting and the social as instituted. According to Laclau, ‘there is in all society a normative order governing institutional arrangements, contacts between groups, the circulation of goods, and so on. This is what we have called “the realm of the sedimented social practices”’ (2014: 121). The social as instituted is a sedimented order, although such sedimentation is never absolute and definitive, for that would mean the death of politics. In Laclau's terms: ‘if the sedimentation of social relations were to prevail to the point of rendering invisible the contingency of the acts of original institution, no hegemonic displacement would be possible’ (Pessoa et al. 2001: 8). Since Laclau identifies politics with hegemony, the halt of hegemonic displacements would be tantamount to the death of politics. A social practice being sedimented means that its meaning becomes naturalised and the traces ‘of the contingent instance of its original institution’ are erased (Pessoa et al. 2001: 8). Reactivation would consist, precisely, in the return to that political instance in which the meaning of every social practice is perpetually reopened. The reactivation and transformation of sedimented meanings is the activity of the social as instituting – or, to be more faithful here to Laclau and Mouffe's terminology, of the political as instituting. Returning to Castoriadis’ earlier examples, we would say that cold or archaic societies are those in which generalised social sedimentation predominates, while revolutions are moments in which the political reactivation of a multiplicity of previously sedimented social practices is intensified to the maximum. Between the extremes of a frozen society and a revolutionary eruption, there would be room for a whole range of intermediate cases, the nature of which would depend on which dimension (instituting or instituted) is more prevalent, and which social spheres and practices are affected by each of them. Continuing with this series of geological metaphors, Castoriadis defines the social as a ‘magma’ (or a ‘magma of magmas’) in which ‘there are flows that are denser, nodal points, clearer or darker areas, bits of rock caught in the whole. But the magma never ceases to move, to swell and to subside, to liquify what was solid and to solidify what was almost inexistent’ (1987a: 243-244).

Castoriadis, Laclau and Mouffe share the idea that the institution of the social is eminently political, in the sense that it is the exclusive product of the thought and action of human collectivities. For this reason, both Castoriadis and Mouffe have developed a conceptual distinction that seeks to overcome the limitations derived from the meaning we give to the term ‘politics’ in its common, everyday use, where we identify it with the activity that takes place within the state, that is carried out by political parties or that occurs during electoral periods. Thus, both authors distinguish between ‘politics’ and ‘the political’. However, the meaning that Castoriadis and Mouffe give to each of the two terms – and therefore to the distinction as a whole – is far from equivalent.

Castoriadis encompasses under the term ‘the political’ all institutions of ‘explicit power’, that is, ‘instances capable of formulating explicitly sanctionable injunctions’ (1991: 156), which throughout history have been present in any human society. On the contrary, he defines ‘politics’ as ‘the lucid and deliberate activity whose object is the explicit institution of society’ (1991: 174). Understood in this sense, ‘politics’ would have been created by the ancient Greeks when they simultaneously invented democracy and philosophy (1991: 159). Castoriadis understands politics as ‘the questioning of the law in and through the actual activity of the community’ (1991: 164). Mouffe, on the other hand, makes a total inversion of the terms. For her ‘the political’ is ‘the dimension of antagonism [ . . . ] constitutive of human societies’ that refers to ‘the very form in which society is instituted’, while ‘politics’ would refer to ‘the set of practices and institutions through which an order is created that organizes human coexistence in the context of conflictuality provided by the political’ (2005: 9). Mouffe affirms that ‘to make this distinction suggests a difference between two types of approach: political science which deals with the empirical field of “politics”, and political theory which is the domain of philosophers who enquire not about facts of ‘politics’ but about the essence of “the political”’ (2005: 8). In Heideggerian terms, ‘politics’ would be situated at an ‘ontic’ level and ‘the political’ at an ‘ontological’ level.

It follows from the above definitions that ‘politics’ in Mouffe and ‘the political’ in Castoriadis refer to the more conventional sense of the term (the one that has to do with institutions and explicit power), while ‘the political’ in Mouffe and ‘politics’ in Castoriadis refer to the instituting dimension of the social. With respect to these last two categories, it is relevant to highlight the centrality that Mouffe's definition of ‘the political’ gives to antagonism, hostility and conflict, an element that is not so central to Castoriadis’ definition of ‘politics’.

Emancipation: Autonomy and Radical Democracy

At the beginning of this article, it was pointed out that both the first part of The Imaginary Institution of Society and the first part of Hegemony and Socialist Strategy are devoted to a deconstruction of certain fundamental theoretical postulates of the Marxist tradition, a tradition within which Castoriadis, Laclau and Mouffe began their respective intellectual trajectories.³ However, none of the three advocated at any point an abandonment of socialism as an emancipatory political horizon; rather, they set out to reformulate socialism, respectively, as a project of individual and collective autonomy (Castoriadis) and as a project of radical democracy (Laclau and Mouffe), thus freeing it from theoretically essentialist and politically reductionist remnants linked to Marxism.

From a historical point of view, there is a crucial difference between Castoriadis’ project of autonomy and Laclau and Mouffe's project of radical democracy, although both do indeed acquire their proper meaning in the contemporary political and social context. For Castoriadis, most of the history of human societies is characterised by a situation of heteronomy or ‘closure of meaning’, except for two major exceptional periods corresponding to ancient Athenian democracy and modern Western democracies, in which ‘an extraordinary historical creation’ – autonomy – ‘then occurred’ (2010: 57). The creation of autonomy emerges and becomes visible in ‘Greek democracy, and in a much larger, but also in some respects more problematic, form, [in] the modern democratic and revolutionary movement’ (1997a: 113). Castoriadis writes that ‘an autonomous society, a truly democratic society, is a society that puts all pregiven meanings into question; it is a society in which, for this very reason the creation of new significations is liberated’ (Morgin et al. 1995: 98). It is at this point where the link between post-foundationalism and democracy becomes most explicit. Although ‘we have been unable to speak of a society that would be autonomous in the full sense of the term’ (Castoriadis 1993: 106), autonomy ‘is that which has defined our history since ancient Greece and to which Europe has given new dimensions: we make our laws and our institutions, we want our individual and collective autonomy, and only we can and must limit this autonomy’ (Castoriadis 1990: 382-383). Castoriadis does not glorify, however, the ‘regimes of liberal oligarchy’ in which we currently live, which are the result of a compromise ‘our societies have reached between capitalism properly speaking and the emancipatory struggles that have attempted to transform or liberalize capitalism. This compromise guarantees, it cannot be denied, not only liberties, but also certain possibilities for certain members of the dominated categories of the population’ (1997a: 115). It is important to note another analytical and normative difference, whose political consequences we cannot develop here: while Castoriadis defines contemporary Western societies as ‘liberal oligarchies’, Laclau and Mouffe prefer to speak of liberal democracies hegemonised by neoliberalism.

Laclau and Mouffe's project of radical democracy has an exclusively modern genealogy, linked to what Tocqueville originally called the ‘democratic revolution’. They argue that ‘the key moment in the beginnings of the democratic revolution can be found in the French Revolution, since [ . . . ] its affirmation of the absolute power of its people introduced something truly new at the level of the social imaginary’ (1985: 155). The principle of popular sovereignty has since then been, together with the promotion of social equality, the main distinguishing feature of the democratic tradition (Mouffe 2000: 2). Laclau and Mouffe conceive the historical emergence of ‘democratic discourse’ as ‘a decisive mutation in the political imaginary of Western societies [that] took place 200 years ago [and] initiated what Claude Lefort has shown to be a new mode of institution of the social’ (1985: 155).

The fact that Laclau and Mouffe's theorising is confined to the modern world and does not introduce any reference – unlike Castoriadis’ – to ancient Athenian democracy is directly linked to the importance they attach to the principle of democratic representation. At this point, what is probably the most explicit and radical disagreement between the theories of the three authors comes to the fore. Here, ‘two positions confront each other: one sees representative democracy as an oxymoron and argues that a “real” democracy needs to be a direct or even a “presentist” one; another claims that far from contradicting democracy, representation is one of its very conditions’ (Mouffe 2013: 123). For Laclau, representation is not only ‘the core of the political’ (2012: 393) but also ‘the primary terrain of constitution of social objectivity’ (2005: 163). Castoriadis considers that where there is representation there is no democracy, and that the only true democracy is a direct and participatory one, being a necessary requirement that the people hold and exercise power. Although it is true that Castoriadis envisages two historical stages in which the project of autonomy emerges (ancient Greece from the seventh century BC onwards and the modern world born of the democratic revolutions), in the end he tends to privilege in his analysis ancient Athenian democracy, which he defines if not as a model – for ‘we would be crazy to copy the political organisation of 30,000 citizens to organise 35 or 150 million citizens’ (1997c: 407) – then at least as a ‘germ’ (1997a: 114). Even when he shifts the focus of his analysis to the modern world, Castoriadis states that ‘it is not by chance that every time there were great revolutionary or reforming movements in society, in the true sense of the term, they began almost without exception with an impulse of restoration or establishment of direct democracy’ (2012: 106). He gives as examples the town hall meetings in New England in the 1770s and 1780s, the Parisian sections during the French Revolution, the Paris Commune or the soviets’ and workers’ councils, including those of Hungary in 1956 (2012: 106). As a negation of these germs of direct democracy, Castoriadis sees

a metaphysics of political representation [that] determines everything without it ever being voiced or made explicit. What is this theological mystery, this alchemical operation that makes your sovereignty, one Sunday every few years, a fluid that spreads throughout the country, enters the ballot box and comes out again that same evening on the television screen, on the faces of the ‘representatives of the people’ or on the face of the representative of the people, the Monarch with the title of President? This operation is clearly of a supernatural character, and no one has ever attempted to provide a foundation for it or even to explain it. (1993: 108)

We are not going to discuss here to what extent this disdain for the fiction of representation, enunciated from naturalistic parameters (as if anything that is not strictly physical or material immediately becomes metaphysical or theological), is coherent with the centrality that Castoriadis gives in his work to imaginary significations as ontological constructors of the social–historical world. As Laclau and Mouffe remind us, it is not because representation is a fiction that it ceases to constitute a principle that structures actual social relations (1985: 119).

The authors of Hegemony and Socialist Strategy believe that there can be no democracy without representation. In their book they mention the classic size argument, according to which direct democracy ‘only works in reduced social spaces’ (1985: 185). Mouffe goes on to argue that in reality ‘the difference between ancient democracy and modern democracy is not one of size but of nature. The crucial difference resides in the acceptance of pluralism. By “pluralism” I mean the end of the substantive idea of the good life, what Claude Lefort calls the “dissolution of the markers of certainty”’ (2000: 18). Laclau and Mouffe point out that ‘the original forms of democratic thought were linked to a positive and unified conception of human nature and tended to constitute a unique space in which that nature was to manifest the effects of its radical freedom and equality’ (1985: 181). Laclau also indicates that in conditions of social dispersion and dislocation, democratic representation can contribute positively to the articulation and consolidation of collective political identities (1997: 51).

Despite their antagonistic approaches to the question of democratic representation, the works of the three authors still present important affinities. One of the most relevant ones has to do with the fact that both Castoriadis’ project of autonomy and Laclau and Mouffe's project of radical democracy include an important socialist dimension, but neither of the two projects is reduced to it. For Laclau and Mouffe, ‘every project for radical democracy presupposes a socialist dimension, as it is necessary to put an end to capitalist relations of production, which are at the root of numerous relations of subordination; but socialism is one of the components of a project for radical democracy and not vice versa’ (1985: 178). Within both projects there is room for all those struggles linked to what Castoriadis calls ‘movements of the sixties’ (1987b) and Laclau and Mouffe ‘new social movements’ (1985). ‘What did these movements, be it the working class, women's, youth, or ecology movement, express?’ asks Castoriadis. ‘I think they all come under the same heading, with the same signification: movements toward and for autonomy’ (2010: 66). They are all struggles against different kinds of injustices or oppressions that are not strictly economic in nature, which does not give them an inferior ontological relevance. Laclau and Mouffe also argue for ‘redefining the socialist project in terms of a radicalisation of democracy; that is, as an articulation of struggles against different forms of subordination – of class, gender, race, as well as those opposed by ecological, anti-nuclear and anti-institutional movements’ (1987: 6). It is here that the convergent contributions of Castoriadis’ project of autonomy and Laclau and Mouffe's project of radical democracy to a post-foundational emancipatory horizon become most clearly manifested.

Conclusion

It was already noted at the beginning of this article that a parallel, systematic and detailed reading of the works of Castoriadis and Laclau and Mouffe is a task that has yet to be accomplished. Here we have simply outlined some general ideas that can serve as a starting point. We have worked on four main hypotheses, which have been outlined respectively in each of the preceding subsections: 1) in Castoriadis we find, malgré lui, a post-foundational political ontology, which due to its depth and philosophical rigour can make an outstanding contribution to enriching and broadening the debates taking place in the field of post-structuralism and which anticipates some central ideas of Laclau and Mouffe's political ontology; 2) the main common element shared by the political ontologies of the three authors is undoubtedly their status as ‘ontologies of signification’, insofar as they conceive the creation and allocation of meaning as the central process in the structuring of the social and political world; 3) the institution of the social (imaginary for Castoriadis; discursive for Laclau and Mouffe) is an eminently political process, which ultimately always depends – and it could not be otherwise – on the wills, ideas and agencies of individuals and human collectivities; 4) from a political point of view, we find in the work of all three authors a theoretical and practical commitment to emancipatory ideals, as well as a will to renew them from a conceptual point of view, which in the case of Castoriadis manifests itself through his project of autonomy and in the case of Laclau and Mouffe takes the form of a project of radical democracy. Within this last section, we are obliged to recall for the last time that the representative principle is the great theoretical object of discord between their conceptions of democracy. Although in this debate our theoretical sympathies lie closer to Laclau and Mouffe, we believe that Castoriadis’ radical critique should be taken into account as it can be useful to give more solidity and substance to positions in favour of democratic representation.