Platform1Democracy 2001

Democracy between Autonomy and Heteronomy


Democracy between Autonomy and Heteronomy

Ernesto Laclau

In this presentation, we will address the question of the relationship between democracy and autonomy. "Autonomy" is an elusive category: its conceptual specification can involve the operation of very different social logics. The first and most radical way of defining it would be in terms of self-determination. An entity is autonomous as far as it does not have to go outside itself in order to be determined in its being. It is in these terms that self-determination, freedom, and infinitude formed, for Hegel, an indissociable whole: true infinitude, as different from a spurious one, involves finding, within itself, the principle of its own determination. And this is the very definition of freedom conceived as autonomy. But this triple equation — freedom, self-determination, and autonomy — involves also the notion that the truly autonomous subject can only be a universal one. As a result, freedom and necessity become, for this universal subject, strictly synonymous. In the classical formulation, freedom can only be the consciousness of necessity. It is only as far as my true self is the universal that nothing is external to myself and that I am really autonomous; anything less than this universal self will be limited by something essentially alien which will be the source of an irreducible heteronomy.

The question that we want to raise concerning democracy is to what extent this ultimate heteronomy is incompatible with freedom — in other terms: to what extent self-determination, conceived as pure universality, is the condition of freedom. We want to suggest that, on the contrary, it is the very failure of a total freedom that makes possible the freedoms (in the plural) of contingent and finite subjects. If it could be shown that this is indeed the case, a fully self-determined subject would certainly be impossible; but one would have also to conclude that heteronomy cannot be entirely excluded from the workings of freedom. Freedom would involve an undecidable tension between autonomy and heteronomy and would thus become the name of that very undecidability. The referent of that name, however, would have been displaced: it would no longer be the closure of the gap between universality and particularity, but the very impossibility of that closure. So we will have to explore the different uses to which heteronomy can be put in relation to democracy.

Let us start with the hypothesis that the gap between the universality of the community conceived as a totality and the particularism of a plurality of demands or groups of demands is a constitutive one. If this gap were radically unbridgeable, we would have to conclude that there is no constituency corresponding to the "universal" and that democracy would be impossible. There would be no way of constituting "the people" (peuple) and we would just have a civil society conceived as a locus of dispersed demands (the "system of needs," as Hegel called it). In that case, the moment of universality would be transferred to a separate sphere (the State), bureaucracy would become the "universal class," and no democracy (let alone a radical one) would be conceivable. So the very possibility of democracy seems to depend on being able to construct a bridge between the particularism of the demands and the community conceived as a whole — that is, it turns on the possibility of constructing "the people" as a historical agent. There are, however, various ways of conceiving this bridging process. If it is seen as a necessary transition from the particularism of the demands to the universality of the community, the emancipation involved in radical democracy would be seen as the movement from the kingdom of necessity to the kingdom of freedom — the latter being conceived as self-determination in the strong sense of radical autonomy. But this is the possibility that we have excluded ex hypothesi.

The other alternative is that the bridging of the gap does take place, but that the bridge is not a priori inscribed in the nature of the particular demands, for it depends on a contingent process of political construction. In that case, an element of exteriority becomes constitutive of the identity of "the people." But if this exteriority is truly constitutive, this means that the democratic/emancipatory agent cannot be entirely self-determined and that, consequently, its identity will necessarily include an element of heteronomy. We are confronted with the paradoxical situation that autonomy requires its exact opposite — heteronomy — as the condition of its constitution — or, rather, that the couple autonomy/ heteronomy is the name of a tension, of a continuum where the conditions for the elimination of one or the other pole of the dichotomy never arise. Democracy, as it were, consists in the negotiation between these two contradictor), logics, not in the elimination of heteronomy in the name of a fully fledged selfdetermination.

The traces of this heteronomy can be found in two central aspects of a democratic polity: the double inscription of the identity of the democratic subjects and the emptiness of the place of inscription. Let us start with identity. As we have asserted, there is no concrete demand which finds inscribed, within itself, the conditions of its own universalization. Demands are many — housing, education, freedom of the press, political and economic rights, etc. — and they can be articulated between themselves in a variety of ways. Some of these ways will not be democratic at all — there is no reason why some popular or democratic demands cannot be articulated to authoritarian projects; and, within the field of democratic politics, a plurality of articulating projects will necessarily coexist. (This coexistence is essential if we are going to speak of democratic politics.) What is crucial to democratic politics, in terms of what we have called double inscription, is: (1) that the universalization of a demand takes place in terms of contingent chains of equivalence1 with other demands; (2) that, while this universalization will bring about a certain closure, the latter is never total — otherwise a certain articulating content would cease to be contingent and will become necessary, and in that case all possibility of dissent will be excluded and democracy will be at an end.

Let us suppose that a society is experiencing a situation lived as radical injustice: justice, in that case, will be seen as what is present through its very absence; justice will not have a positive content of its own, but certain particular contents will be lived as the embodiment of a radical political redress capable of bringing about justice. The overthrow, for instance, of bureaucratic rule in Eastern Europe in the 1980s is a good example. We see here the operation of what we have called double inscription: on the one hand, certain concrete demands are "universalized" when they enter into equivalential chains with other demands; on the other hand, these equivalential chains are the embodiment of something constitutively exceeding them: the empty signifiers "order," "justice," "truth," "revolution," etc. The important point is that the investment of these empty symbols into concrete contents embodying them is a radical investment: there is no possibility of moving from one level to the other in terms of a merely logical transition.

The matter could be formulated in terms of Lacanian theory: because the subject is, in a radical sense, the subject of the lack, its relation to a concrete content can only be a relation of identification. The universalization achieved through the inscription of a certain demand into a chain of equivalences confers a certain power to the subject thus constituted, and in this way endows him with a certain autonomy, but it is an autonomy achieved only at the price of an identification. This involves the impossibility of a pure self-determination and is the source of a heteronomy which is always at the heart of autonomy.

If we consider our second aspect concerning the relationship autonomy/heteronomy, similar considerations can be made. The contingent articulation between empty universal identity and particular content incarnating the latter is to be found again in the relationship between the empty place of power and the actual force occupying it. This is the very pertinent distinction introduced a long time ago by Claude Lefort. On the difference with the hierarchical societies of the past, where the place of power has a "natural" occupier — where, in our terminology, this is an overlapping between universality and differential particularity — the main change brought about by the "democratic invention" is that this overlapping is unmade: the place of power becomes empty, while its links with its successive occupiers become essentially contingent. I entirely subscribe to this vision of the changes resulting from the democratic revolution. I would only add that the condition of emptying the place of power in the way described by Lefort is that a similar process takes place at the level of the subject: it is because the subject is now split — in the way we described — between particularity and empty universality, that the place of power, in modern democracy, can become empty.

If this conclusion is correct, democracy, as the space of negotiation between universality and particularity, necessarily blurs the boundaries between State and civil society. For Hegel, these boundaries were strict: civil society was the realm of pure particularity, while the universal class was exclusively located in the political sphere. Marx eliminates the latter and transfers the attributes of the "universal class" from bureaucracy to the proletariat, but in this transfer the moment of particularity vanishes: the simplification of class struggle under capitalism leads to a universal subject whose self-determination will not be tainted by any heteronomy. It is only with Gramsci that we see the emergence of a political logic which cuts across the distinction State/civil society and, in that way, makes the interaction between autonomy and heteronomy an integral part of the democratic negotiation.2 We can add that the undecidability of the game universality/particularity permeates many contemporary debates concerning agency. We have, on the one hand, positions asserting an issue-oriented politics based in cultural diversity, multiculturalism, affirmation of difference, etc. On the other, the insistence that the broader problems linked to emancipatory struggles are abandoned through this emphasis on difference.3 It should be clear why, in our view, both emphases are unilateral and limited. It is not a question, for us, of denying the radical democratic potential that differential struggles present, but it is not a question either of limiting oneself to the punctual character of those struggles, leaving aside broader strategic considerations. The construction of contingent chains of equivalence is, in our view, the terrain in which the link between universality and particularity has to be established. The tension between these two polar alternatives — the building up of a universal emancipatory subject, the enlargement of the democratic revolution through the expansion of the equalitarian principles to increasingly larger sections of the population — runs through the whole history of modern democracy.

We have to stress a last point concerning the dialectic between heteronomy and autonomy and its reproduction at all levels of political argument. It concerns the ways of addressing the relationship inclusion/exclusion. As we have seen, an autonomy conceived as strict self-determination cannot leave anything outside itself, it has to reduce anything apparently alien to an internal moment of its self-development. Self-determination has to be all-inclusive. Now, a notion of inclusion conceived this way is not incompatible with some kind of exclusion: one which makes the otherness of the other the condition of constitution of the self. The otherness of the other becomes, in that way, an internal moment of a wider totality under which "self" and "other" are subsumed. Their incompatibility is reduced by referring them to a universality transcending both poles of the initial opposition. Hegel was able, in that way, to present "world history" as a purely internal development unified by a cunning of reason which dialectically supercedes and interiorizes all apparent externality. This could not however be, even for Hegel, the totality of the story. For the main line of historical development, as described by him, had to come to terms with the presence of a contingent excess escaping its mastery. Thus, he could not avoid having to speak of "peoples without history." And here we have an exclusion of an entirely different kind: it is an exclusion radically refractory to self-determination, one that the latter cannot retrieve. It involves the emergence of a heterogeneity which threatens the purely internal character of "world history."

Hegel probably thought that the purely marginal character of this social excess was no real threat to his story. And Marx argued along similar lines: the proletariat was part of a world history unified by the category of productive labor.4 But even in Marx's account, the irreducible remainder was present: it was given by the notion of lumpenproletariat as a social excess, without history, living in the interstices of all social formations. The violently dismissive way in which both Marx and Engels referred to the lumpenproletariat is well known. Some of their contemporaries, like Bakunin, took, however, a different view: it was for them the radical exteriority of the lumpenproletariat vis-à-vis the existent system that ensured its revolutionary potential. And some later writers take a similar line. Thus, Fanon writes: "The lumpenproletariat, once it is constituted, brings all its forces to endanger the 'security' of the town, and it is the sign of the irrevocable decay, the gangrene ever present at the heart of colonial domination. So the pimps, the hooligans, the unemployed, and the petty criminals ... throw themselves into the struggle like stout working men. These classless idlers will by militant and decisive action discover the path that leads to nationhood ... The prostitutes too, and the maids who are paid two pounds a month, all who turn in circles between suicide and madness, will recover their balance, once more go forward, and march in the great procession of the awakened nation."5

We see thus emerging, from within the very logic of construction of the emancipatory subject, the same tension between autonomy and heteronomy that we have pointed out since the beginning of our reflection. Total autonomy would presuppose that only inclusion is ultimately thinkable and that any exclusion is purely transitory or appariential and destined to be finally mastered by the inclusive identitary logic. But it follows that the subject of this logic cannot be any actually existing subject but only a transcendental one (the latter being, of course, the Absolute Spirit, which is not even achievable in the sociopolitical sphere). The second kind of exclusion is, on the contrary, constitutive: the alien opposing it is irreducible to any type of interiorization. Does this mean that we have pure heteronomy and that the possibility of any autonomy has to be discarded? Not at all, for in the very process of opposing something alien one is able to construct the conditions of one's own efficacy within a certain area; one does become self-determined in a weak and partial sense. Now, the autonomy thus achieved depends on heteronomy for its emergence in a double sense: firstly, because the autonomy thus conceived is indistinguishable from power, and the latter presupposes something alien over which it is exercised (a power whose object is oneself would not be power at all), and secondly, because it is only through its opposition to a power external to itself that the identity of the relatively autonomous entity is constituted. (Going back to Fanon's example: it is only their opposition to colonialism that makes possible the union of the marginals in a new historical actor.)

On the difference with the absolute subject of a total self-determination, the partially autonomous/partially heteronomous identity of the subjects emerging from the second relation of exclusion corresponds exactly with the identity of actually existing social actors. And it is important to realize that these are the only subjects of any possible democracy, the latter involving the presence of irreducible dissent and of a constitutive gap between the particularism of the social agents and the universality of the communitarian space. But this gap — without which, as we have seen, there would be no democracy — translates the undecidability which is its necessary corollary to the main categories which have historically structured the discourse of democratic theory. We will discuss its operation in two of these categories, sovereignty and representation, and we will later conclude this presentation by stating the centrality that the notion of hegemony has for a radical approach to democracy.

Sovereignty. It is important to see why sovereignty is a specifically modern category. The notion of an ultimate source of power was certainly not absent from the Ancient World or the Middle Ages — notions such as summa potestas or plenitude potestas moved certainly in that direction — but on the whole, the limitations of the royal power by natural law, by custom, and by a feudal organization that reduced the king to the role of primus inter pares conspired, before the 14th century, against the formulation of a fully fledged theory of sovereignty. (To this, one should add the presence of the two universal powers, the Empire and the Church, which equally limited the sovereignty of the nation states.) The important point for our discussion is that, at the moment of its mature formulation in modern times — for instance, in the Hobbesian principle according to which auctoritas, non veritas fizcit legem — the notion of sovereignty is linked to the absence of limitations in the exercise of power. This means that we are dealing with autonomy in the strong sense of the term discussed above. The State centralization brought about by the absolute monarchies worked in that direction, and democratic theory, at the moment in which it attempted to replace the sovereignty of the king by that of the people, had the whole terrain prepared by centuries of bureaucratic unification.

But a democratic sovereignty has problems of its own, which are not so visible in an authoritarian one. In the case of Rousseau, for instance, we have the whole ensemble of paradoxes linked to the notion of volonté générale. The whole question of making possible a will which becomes the locus of a true universality could only be approached by him in terms of homogeneous social actors in small communities, communities that — he was the first to recognize — were every day less compatible with the conditions of the modern world. That is the reason why, from the very beginning, the theory of sovereignty was challenged by opposite approaches which turned constitutionalism into a weapon to prevent the total concentration of power in a single point of the political organism. The American Constitution, for instance, as justified in the Federalist Papers, advocates a loose federation of states which prevents the formation of a strongly centralized juridical structure.

It is easy to see that this opposition overlaps with the two senses of autonomy that we have discussed before. Either we have autonomization of a particular sector against the community as a whole — in which case no subject can truly be a sovereign (and democracy, consequently, identifies itself with a limitation of sovereignty), or we have the total autonomy of a fully sovereign power, but in that case such a regime can only be democratic if it overlaps with the homogeneous will of a volonté générate. The important point to stress is that these two logics are ultimately incompatible: there is no square circle that could bring them together into a logically coherent intellectual mode. The fact that democracy exists in the space of this irreducible tension does not mean, however, that it is impossible, but only that the language game that we call "democracy" consists in negotiating between these two incompatible poles. A society which leans too much toward extreme particularism would not be able to build up any form of global collective representation and collective will — and would be easily manipulated by an administrative power which is not submitted to any political check or challenge. But a society which is exclusively universalistic in terms of the collective identities that it is capable of creating, would have to suppress dissent and would fall into the worst forms of authoritarian unification (the Class, the Race, the Fatherland, etc.). Now, the terrain thus drawn for democratic negotiation is no other than that of an autonomy contaminated by heteronomy, for it will be the negotiation of social agents who are less than the "universal class."

Representation. We have asserted that a certain universalization — which stops short of full self-determination — is a condition of democracy. How is that universalization, however, achievable? We have already given some elements to start answering that question. Two central dimensions should be taken into account. The first is what we have called equivalential relations: a social identity universalizes itself when it enters into an equivalential relation with other identities. And identities are structured around demands. A demand which remains closed within its own particularity without establishing equivalences with other identities can never become political. Politics supposes negotiating connections between demands and constituting wider social identities as a result of those connections.

But there is a second and equally important connection. Wider popular identities are organized around equivalential chains. But it is not enough to enumerate the links of the chain in a purely additive exercise: it is also necessary to name the chain, to signify it as a whole. The symbolic unity of "the people" is crucial in any process of political construction. The means of representation of these collective ensembles, however, are only the particular links constituting the chain. This requires that some of these links become split from their own particularities and that, without ceasing to be particular, they become the signifiers of a certain overflow of meaning. "Socialization of the means of production," for instance, is a technical way of running the economy, but in the socialist discourses of the beginning of the 20th century it signified a much wider project of human emancipation, equivalentially embracing demands coming from a variety of fields. Without this bringing together of demands through naming those wider ensembles, there would be no possibility of universalization.

This has, however, an important consequence: there is only universalization through representation. The idea of a purely self-transparent autonomy should be resolutely discarded. Why? Firstly, because no autonomy could exist except by making the subject more universal — which is only achievable through the expansion of the equivalential chains. A demand closed with its own particularity, far from being autonomous, would be reduced to a hopeless heteronomy: it could not constitute any power and it would be unable to establish any war of position in the negotiation of the tension autonomy/heteronomy. Secondly, as we have seen, the condition of the equivalential chain becoming a wider collective ensemble is the representation of that ensemble by a particular signifier whose identity is split between its own particularity and its wider representative function. This is the operation that Gramsci called "hegemony." But if the wider (more universal) identity requires being represented by a universal equivalent — what in my terminology I have called an empty signifier6 — this means that there are no social identities except through a process of representation. Representation is not a subsidiary relation, one that completes an identity constituted outside and previous to the process of representation, but is inscribed in the originary construction of social identities. The process of representation and the process of formation of social identities is actually one and the same process. And here we find again the dialectic between autonomy and heteronomy: no autonomy can exist except through the equivalential universalization of demands and the power relations that the latter involves (this is the autonomous moment); but no universalization can take place except through representation (this is the heteronomous moment). Representation involves a double movement between representative and those who are represented, in which both sides contribute something to the representative process without being possible to assign any ultimate priority to either side.7

Hegemony. Let us recapitulate. We have argued in this paper that some of the stark oppositions that have dominated social and political theory for a long while are simply the result of making a choice for one extreme of opposition and presenting the other as its strict antithesis. We have maintained, on the contrary, that in most cases the two extreme opposites, far from rejecting each other, contaminate each other, so that it is only by focusing on their processes of mutual subversion that new language games can be designed which take into account the historical possibilities for democratic theory and practice that those apparent blind alleys actually open.

The center of our analysis was the relation between autonomy and heteronomy. We tried to show that the strict identification between autonomy, freedom, and self-determination leads to a situation in which: (a) such identification does not correspond to any possible subject; and (b) that it is only by articulating autonomy and heteronomy in their tense relation that it is possible to develop more complex strategic games that open the way to actual democratic interactions. The same can be said of other couples of traditionally antagonistic concepts — some of which we have referred to briefly — such as universality and particularity, power and emancipation, inclusion and exclusion, etc.

Central to our concern has been the category of "power" which, in our view, points to the terrain in which the negotiation between most of our dichotomies takes place. To go back to autonomy/heteronomy: for a subject who is less than the Absolute Spirit to be autonomous involves the construction of a power relation, something which makes possible autonomy as a result of a relation of forces which, however, presupposes the moment of heteronomy. The same can be said of all other oppositions. And this leads us to the question of hegemony, which is for me the basic category of political analysis. I have defined "hegemony" as the process by which a certain particularity assumes the representation of a universality which is incommensurable with it. We have here an undecidable terrain between universality and particularity. We are not simply in the terrain of pure particularity, which would involve straight domination; but we do not have pure, uncontaminated universality either. What we have is a mutual subversion between universality and particularity which creates the field of a tension which cannot be overcome. As in the cases of power/emancipation, inclusion/exclusion, and autonomy/heteronomy (but perhaps they are not different cases), it is this logic of undecidability which is at the root of the political productivity of the notion of hegemony. As Kant writes in the Critique of Pure Reason: "Much is already gained if we can bring a number of investigations under the formula of a single problem."8

Ernesto Laclau. Democracy between Autonomy and Heteronomy. In: Okwui Enwezor, Carlos Basualdo, Ute Meta Bauer, Susanne Ghez, Sarat Maharaj, Mark Nash, Octavio Zaya (ed.): Democracy Unrealized. Documenta11_Platform1. Hatje Cantz Verlag, Ostfildern-Ruit. 2002, pp. 378-386.



On the notions of equivalence and difference, see Ernesto Laclau and Chantal Mouffe, Hegemony and Socialist Strategy: Towards a Radical Democratic Politics (London: Verso, 1985), ch. 3.


Antonio Gramsci, Selections from the Prison Notebooks, ed. and trans. Quintin Hoare and Geoffrey Nowell Smith (London: Lawrence & Wishart, 1971), pp. 206-277.


This is the position maintained by Slavoj Žižek, among others, over recent years.


See Peter Stallybrass, "Marx and Heterogeneity: Thinking the Lumpenproletariat," Representations 31, no. 1 (Summer 1990), pp. 69-95.


Frantz Fanon, quoted in ibid., p. 89.


See my essay "Why Do Empty Signifiers Matter to Politics?," in Emancipation(s) (London: Verso, 1996).


See my essay "Power and Representation," in ibid.


Immanuel Kant's Critique of Pure Reason, trans. Norman Kemp Smith (London: Macmillan, 1985), B 19, p. 55.

About the Platforms

In the course of our research and preparation for Documenta11 in 2002 the curatorial team discussed the possibility of organising a sixth platform. It would feature the voices of artists, curators, critics and intellectuals formed by the experience of Documenta11 and its various platforms. The spirit of the event should be that of “reculer pour mieux sauter”, looking back to look forward, using the event to reformulate the issues most urgent to our practices just as Documenta11 itself enabled us to rethink our political, cultural and aesthetic engagements.

Documenta11 team members Ute Meta Bauer, Angelika Nollert and myself visited Okwui in Munich where he was confined by illness but where he continued to work on his Haus der Kunst and critical and curatorial projects.

Introduction by Mark Nash, Ute Meta Bauer and Angelika Nollert

Read more