Journal of Political Ideologies (October 2005), 10(3), 289–309

Antagonism, hegemony and ideology after heterogeneity LASSE THOMASSEN Department of Politics and Public Administration, University of Limerick

ABSTRACT This article examines the implications of the introduction of the category of ‘heterogeneity’ in Ernesto Laclau’s most recent work. Laclau’s theory of hegemony and discourse theoretical approach to ideology is often associated with the category of ‘antagonism’. I argue that heterogeneity should be the central category of hegemony and discourse analysis, and that antagonism can be seen as a strategy of ideological closure. In addition, heterogeneity— understood as the simultaneous condition of possibility and impossibility of hegemonic articulation—renders the theory of hegemony closer to Derridean deconstruction. Hegemony analysis and deconstruction are often presented as different and complementary theoretical moves. I argue that this is not the case, and that they can instead be seen as dealing with the same issues of the conditions of possibility and impossibility of the discursive constitution of ideology and identity.

Introduction One of the most influential approaches to the study of ideology is Ernesto Laclau’s theory of hegemony and the so-called ‘Essex School’ employing Laclau’s work. They treat ideology as a special kind of discourse, that is, a structured, meaningful totality,1 and ideology refers to a particular kind of discourse that attempts to conceal the always-already dislocated character of any meaningful totality.2 The hegemonic approach to discourse and ideology draws upon different strands of thought, among them Wittgenstein, Heidegger, Lacan and post-Saussurean linguistics.3 Another source of inspiration is Derridean deconstruction which has played a central role for Laclau and those inspired by his work.4

Corresponding Address: L. Thomassen, Department of Politics and Public Administration, University of Limerick, Limerick, Ireland. Email: [email protected] ISSN 1356-9317 print; ISSN 1469-9613 online/05/030289–21 q 2005 Taylor & Francis DOI: 10.1080/13569310500244313

lasse thomassen The appropriation of deconstruction for the theory of hegemony is seemingly unproblematic: deconstruction and the theory of hegemony are often presented as the two sides of the same coin, with deconstruction showing the contingency of structures and identities and the theory of hegemony explaining the constitution of structures, identities and ideology.5 Recently, Aletta Norval has put this complementary relationship between deconstruction and the theory of hegemony into question,6 and here I wish to continue this problematisation of the relationship between deconstruction and hegemony. This meta-theoretical question of the relationship between hegemony analysis and deconstruction is linked to a more specific issue arising from Laclau’s conceptual apparatus. In his most recent work, Laclau has introduced the notion of ‘heterogeneity’ to refer to an excess escaping the attempt to discursively objectify the boundaries of identities.7 One example of an heterogeneous entity is the lumpenproletariat, which, in Marx’s work, is a discursive excess escaping the creation of a conceptual frontier between bourgeoisie and proletariat. I shall return to the notion of heterogeneity and to this and other examples in more detail below. The introduction of the notion of heterogeneity necessitates a reconsideration of the notion of antagonism, which has held a central place in Laclau’s work. This, in turn, forces us to consider what is the end of the analysis of ideology: antagonism or heterogeneity; and whether ideology analysis comprises simply the dual strategy of deconstruction and hegemony analysis. My claims in this article are threefold. First, with Norval, I will argue that deconstruction is not a negative preparation for hegemony analysis. This is linked to the notions of antagonism and heterogeneity. The central category, if there is one, of hegemony and discourse analysis is heterogeneity, I will argue, not antagonism. As a result, hegemony analysis, like deconstruction, is also concerned with showing the contingency and precariousness of discourses and social identities. Here, deconstruction and hegemony are shown to be not simply different and complementary discourse analytical strategies. Second, and following from the previous point, my argument implies that social identities are not necessarily constituted around antagonistic frontiers, and that there are only degrees of antagonism, never ‘pure’ antagonisms. Third, I will argue that it is possible to see antagonism as an ideological type of discourse because it is one way of attempting to achieve discursive closure. Although the argument is mainly theoretical and conceptual, I shall use a number of concrete examples. In the first section, I examine the existing literature on the relationship between deconstruction and hegemony and discourse analysis. In subsequent sections, I examine the key parameters of Laclau’s theory of hegemony—empty signifiers, logics of equivalence and difference, and so on— and show the implications of Laclau’s recent reformulations of these for the notion of antagonism. I argue that the notion of antagonism must be qualified, and this has implications for the use of the theory of hegemony for discourse and ideology analysis. This conclusion is further emphasised with the introduction of the notion of heterogeneity, the consequences of which I discuss in the last section of the article. 290

antagonism, hegemony and ideology after heterogeneity Deconstruction, hegemony and discourse analysis Derridean deconstruction has been part of the theory of hegemony from its inception in Laclau and Mouffe’s Hegemony and Socialist Strategy. In this section, I will first consider Laclau’s appropriation of Derrida’s work and then the views of several of Laclau’s commentators. Since Aletta Norval has already dealt with Laclau’s appropriation of Derrida’s deconstruction, I will only make some brief comments on this. In Hegemony and Socialist Strategy, Laclau and Mouffe make reference to Derrida’s deconstruction of structuralism, showing that ultimately no structure or system can be held together by a transcendental signified. This enabled Laclau and Mouffe to question the emphasis on structure, determinism and necessity in Marxist thought.8 Deconstruction provides an argument for contingency and, hence, for the centrality of hegemony understood as the articulation of contingently linked differential elements into a more or less stable whole. In short, no hegemony without contingency and the deconstruction of structure. In the 1990s, Laclau reformulated this insight in terms of the Derridean notion of undecidability. Deconstruction, he argued, shows the undecidability of structures and identities, and the theory of hegemony provides a theory of the decision in an undecidable terrain.9 As Norval has shown,10 this rests on a misunderstanding of undecidability. Not only is undecidability not a general infrastructure, but a specific notion introduced by Derrida in specific contexts; in addition, the decision does not dissolve undecidability. This goes against the temporal dimension of Laclau’s use of undecidability/decision: first undecidability, then decision. Likewise, deconstruction is not a negative and preparatory move that one needs to make and can subsequently leave behind before embarking on the analysis of the decision or hegemony. Deconstruction and hegemony, undecidability and decision, can neither be temporally separated, nor conceived as different and complementary analytical moves. It can, of course, be argued that Laclau merely appropriates and, in the process, rearticulates deconstruction for his own purposes and that it is therefore wrong to accuse him of having misinterpreted Derrida. Yet, the problem I want to emphasise here is that his particular interpretation of deconstruction implies the aforementioned temporal division of deconstruction and hegemony. As I shall argue below, this is of utmost importance for the way one does hegemony analysis because it means that it cannot simply consist in showing how a hegemonic project was successful. It must also consider the limits of any hegemonic constellation, that is, incorporate the purportedly specifically deconstructive move into the hegemony analysis. Laclau’s appropriation of deconstruction and undecidability as different from and complementary to the theory of hegemony is reflected in the work of some of his commentators. Thus, three introductions to Laclau’s work and to discourse analysis all argue that deconstruction and the theory of hegemony are different and/but complementary discourse analytical strategies.11 For instance, Jacob Torfing writes: 291

lasse thomassen Deconstruction in a certain sense implies a theory of hegemony and the theory of hegemony implies deconstruction. . . . Whereas hegemony brings us from undecidability to decidability, deconstruction shows the contingent and constitutive character of decidable hegemonic articulations by revealing the undecidability of the decision.12

While I agree that there is a mutual implication between hegemony and deconstruction, this should not be understood as if the two ‘constitute two complementary and reciprocal movements’.13 Deconstruction is not a merely preparatory analytical move, nor is it a purely negative undertaking. As Derrida has argued, deconstruction is affirmative. Deconstruction does not merely take texts and discourses apart, showing the contingency of structures, identities and binary oppositions, and it does not leave us with a terrain of indeterminacy. As most forcefully argued by Rodolph Gasche´, deconstruction aims to ‘account’ for the conditions of impossibility as well as the conditions of possibility of identities, distinctions, and so on.14 This is the case, for instance, in relation to undecidability and decision: the former at once makes the latter possible and impossible. There is no decision without undecidability, but, importantly, because of undecidability, no decision is ever complete or final. Hence, it is a mistake to argue that deconstruction and hegemony analysis are different and, as such, complementary discourse analytical strategies. Deconstruction already involves what the theory of hegemony is thought to add to it, namely an account of the possibility of identities, distinctions, and, more generally, the stabilisation of meaning. Conversely, as I will try to show in the following, the theory of hegemony should not be seen as exclusively concerned with an account of how decisions and hegemonic totalities come about. Hegemony, empty signifiers and antagonism In a recent work, Laclau uses the following model to clarify the way hegemonic articulation works (Figure 115). In the model, D1, D2, D3, D4, and so on, represent particular signifiers (or demands: D), which are articulated into a chain of equivalence (D1 ; D2 ; D3 ; D4. . .). One of the signifiers (D1) has been able to empty itself of its particular content. As a result, it can stand in for and represent the chain as a whole, thereby establishing the equivalence among the different signifiers. This creates an antagonistic frontier (F) vis-a`-vis an antagonistic force (T), in relation to which the particular signifiers of the chain of equivalence stand in the same relation insofar as they take part in the chain of equivalence as represented by the empty signifier. This links together the creation of a chain of equivalence, the empty signifier and antagonism. In the original model, T refers to Tsarism and the chain of equivalence is formed by different demands united through their opposition to the Tsarist regime. A more recent example of the same kind of hegemonic construction is the so-called ‘War on Terror’ in the aftermath of the 9 –11 attacks. The US 292

antagonism, hegemony and ideology after heterogeneity

Figure 1.

government was able to put together a coalition united through their opposition to ‘terrorism’, especially the international terrorism associated with the names of Osama bin Laden, al-Qaeda and the Taliban. There was an antagonistic frontier dividing ‘us’ from ‘them’: ‘you are either with us or you are with the terrorists’. This hegemonic formation made certain things possible that had not previously been possible (the coalition of hitherto mutually opposed states, for instance), but also excluded certain possibilities from the hegemonic political space (putting into question the particular way the ‘War on Terror’ was carried out, for instance). Tendentially empty signifiers and internal divisions It is not always clear whether Laclau’s theory of hegemony refers to the articulation of identity and meaning as such or to a particular way of articulating meaning and identity, for instance populist discourses emphasising antagonism, equivalence and emptiness.16 This also applies to the model presented above. The question here is whether or not, and to what extent, (contingent) hegemonic articulation is subsumed to antagonism as the necessary end of hegemony. Part of the aim of my analysis is to show that antagonism is not a necessary outcome of hegemony, and that, even in the case of, for instance, populist discourses, it is necessary to pay attention to tendential emptiness, the relativisation of antagonism, and heterogeneity. Just as the world has turned out to be more complex than the simplistic discourse of George W. Bush would allow, so Laclau has complexified his model. The signifiers in the chain of equivalence are equivalent and not identical, that is, they retain some of their mutual differences (hence ; and not ¼ ). Like the other signifiers, the empty signifier is split between its equivalential content and its differential content, thus making it only tendentially empty. Laclau speaks 293

lasse thomassen here of a differential remainder17 or ‘a remainder of particularity’.18 The tendentially empty signifier is not a transparent medium; rather, the signifier has an ineliminable materiality.19 This is of utmost importance for the model and has implications for the analysis of hegemony and ideology. The tendentially empty signifier simultaneously represents the whole chain and is one part of it among others; this is why it appears twice in the model. As a result, the tendentially empty signifier is unable to represent the whole as a whole. In Emancipation(s), Laclau links the notion of the empty signifier to the subversion of the play of differences: insofar as the empty signifier does not have any particular differential content, it is able to represent the totality of the relations of differences without being merely one more difference in an infinite field of differences. Yet, insofar as it is only tendentially empty, the (tendentially) empty signifier is not able to fulfil this role. It never ceases to be also one particular signifier among others; that is, it is never just any signifier but always also this rather than that signifier. Similarly, the equivalence never completely dissolves the relations of difference: ‘just as the logic of difference never manages to constitute a fully sutured space, neither does the logic of equivalence ever achieve this’.20 Since Laclau conceives of meaning and identity in post-Saussurean terms as constituted through relations of difference, equivalence supposedly halts or subverts the play of difference. Again, insofar as the equivalence is only tendential, this subversion is only tendential, and the potential divisions within the chain cannot be dissolved in the antagonistic frontier (for instance, in the case of the ‘War on Terror’). As a result, the signifiers of the chain of equivalence are partly floating signifiers. Their interiority or exteriority to the chain cannot be clearly and unambiguously established, and they can therefore, be dis- and re-articulated. This is an essential part of hegemonic struggles; if there were no floating signifiers, it would be the end of any future hegemonic struggle. The hegemonic operation of a particular signifier taking up the signification of the chain of equivalence is essentially a relation of representation. Importantly it is not the representation of something already present. What is missing is precisely the equivalence, and this is what the relation of representation brings into existence. What is represented does not pre-exist the relation of representation; rather, the latter constitutes the former in a performative fashion.21 Yet, as we know from Derrida’s deconstruction of performativity, the performative is never pure, but always made possible and contaminated by a constative. The performative representation of the chain of equivalence involves an irreducible element of citation: one of the signifiers of the (as yet not fully constituted) chain is partly emptied of content. We thereby have an operation involving both continuity and discontinuity, both citation and performative institution. The relation of representation is only possible insofar as the particular signifier is gradually emptied, yet this process of emptying is never complete. Thinking of hegemony as a relation of representation in this way means that there is no pure (performative) origin of naming; but nor does the process of representation come to an end and establish the full identity of the chain, where 294

antagonism, hegemony and ideology after heterogeneity there would be a simple relation of repetition across the different parts of the chain. This suggests that we can think of the hegemonic relation not only in terms of decision/undecidability and representation but also of the Derridean notion of iterability.22 One could think about identification along the same lines: it involves citation of an already existing object, which is signified in a particular way. It is not a pure citation, though, because there is an element of naming and rearticulation when the object of identification becomes the central organising principle of my identity. The latter process is never complete because there is a tension between citation and naming, with the result that we have continuous identifications rather than a fully established identity. It matters which particular signifier is cited, that is, which signifier takes up the task of representing the whole. While this is ultimately contingent, it is not arbitrary. The particular signifiers are not equally able or likely to take up this task because it takes place in an already partly sedimented terrain permeated by relations of power. Hence, the particular signifier taking up the task of signifying the totality must not only be available, but must also compete with other particular signifiers. This takes place, not on a level playing field, but in a terrain that is itself the result of prior hegemonic articulation.23 Moreover, since the empty signifier is only tendentially empty, it matters which particular signifier takes up this role. The emptying of the signifier opens up a space within which other signifiers can be included and represented, but this opening up is only possible via a simultaneous closing off, because it is the relative emptying of a particular signifier. It is one of the tasks of hegemony analysis to examine why some signifiers come to represent the whole and why others do not. The analysis of hegemony cannot stop at the identification of a successful hegemony, but must also examine which alternatives have been excluded for the present hegemony to be possible. For instance, in the case of the ‘War on Terror’, one must ask what would have been possible, and what would have been excluded, if it had not been a ‘War’, but a police or law-enforcement operation. And, one might ask why ‘freedom’ and not ‘solidarity’ became the organising signifier for the Bush administration. These are not just empty terms that everybody can agree with, rendering it irrelevant which signifier comes to dominate; instead Bush was (also) continuing certain existing discourses in American society, and this had implications for what became possible and what did not. Glenn Bowman’s study of the way in which Palestinians in exile have imagined their ‘lost’ nation provides a further example. Bowman shows how the empty signifier (the Palestinian nation) is necessarily split between its emptiness and a differential remainder. Since the nation is imagined through synecdoche, the lost whole (the Palestinian nation) depends on which part of it one puts in its place, that is, which particular signifier takes up the task of signifying the whole.24 In short, synecdoche also relies on the citation of already existing signifier. The means of signification at our disposal will depend on our embeddedness in a particular context, which is always partially sedimented and permeated by relations of 295

lasse thomassen power. Not every means of signification is equally available in every context, and a signifier will have different meanings in different contexts. As a consequence, different significations of the ‘same’ ‘empty’ thing (for instance, the ‘lost’ Palestinian nation) may not be compatible, and some of them may eventually need to be suppressed in order for a coherent and unitary identity to emerge. Insofar as the split in the signifier is constitutive, it is always possible to imagine things otherwise, for instance, for the nation to be a different nation. And there will always be persons and groups whose points of views cannot be represented within the space of representation opened up—but simultaneously closed—by the tendentially empty signifier. In the case of Palestinians in exile, Bowman shows how these different images of the nation emerge, how they depend on their different contexts of enunciation, and how they mutually conflict.25 One note of caution must be raised at this point, though. Bowman’s explanation of the inherent split in the signifier is ambiguous. At times, he suggests that this is so for the intrinsic reasons explained above. At other times, however, he explains it with reference to the de facto geographical dispersion of Palestinians and their physical, political and social separation from their land.26 Likewise, Bowman sometimes presents the dispersion of Palestinians from one another and from their land as the result of some external force (namely, Israeli occupation), rather than as an inherent part of identity that cannot be overcome.27 This would imply that if only they were not geographically dispersed, and if only the Israelis did not occupy their land and discriminate against them, the Palestinians would regain their lost fullness. Yet, given that there are only tendentially empty signifiers, and that there is always some internal division, this is impossible. Antagonism: The limit of objectivity? If the empty signifier is only tendentially empty, and if the equivalence is unable to completely subvert the relations of difference, then the frontier vis-a`-vis the antagonistic outside will not be a clear and stable frontier (it should be represented in Fig. 1 as a dotted line). Antagonism only exists as a discursive effect and only as one end of a spectrum that is never reached. If anything, there are tendential antagonisms, that is, frontiers and identities that are constructed as more or less antagonistic. As argued above, this means that the frontier is open to dis- and rearticulation. For instance, in the ‘War on Terror’ it was not difficult to create a coalition for freedom and democracy and against bin Laden and the Taliban. Once the ‘War’ went on to Iraq, however, the cracks that were already to some extent present in the coalition started to open up. With Iraq, it was no longer possible to represent the enemy as an absolute threat (or ‘evil’). The ‘freedom’ that was supposed to hold the coalition together appeared to be the freedom of a particular agent, and, as a result, the coalition was unable to stay intact. Before turning to some of the implications of the relativisation of antagonism for 296

antagonism, hegemony and ideology after heterogeneity discourse analysis, it is necessary to examine Laclau’s different formulations of antagonism. In Hegemony and Socialist Strategy, antagonism is introduced as the limit of objectivity and as a threat to my identity: ‘[I]n the case of antagonism . . . the presence of the “Other” prevents me from being totally myself. The relation arises not from full totalities, but from the impossibility of their constitution. . . . Insofar as there is antagonism, I cannot be a full presence for myself’.28 And: ‘Antagonism as the negation of a given order is, quite simply, the limit of that order, and not the moment of a broader totality in relation to which the two poles of the antagonism would constitute differential—i.e. objective—partial instances’.29 Antagonism is conceived as the limit of objectivity because it refers to the subversion of the discursive constitution of meaning and objectivity. More specifically, antagonism refers to the moment when the system of differences is constituted into a chain of equivalence opposing an external threat. The equivalence is established through an empty signifier signifying a set of differences as a totality. Supposedly, antagonism both makes meaning possible (because it provides the condition of possibility for the differences to coalesce into a totality) and impossible (because it denotes a point where the relations of difference, which are constitutive of meaning, are subverted by equivalence). Yet, the antagonistic relation does not actually threaten the identity established through the chain of equivalence. Instead, antagonism is the flipside of equivalence: it is constituted by and constitutes equivalence, because the equivalent signifiers are equivalent insofar as they are all opposed in the same way to the antagonistic Other. Hence, the emptiness of the empty signifier and the antagonistic relation go hand in hand: the emptiness of the empty signifier signifies the fullness of an identity (for instance, of a communal identity by fixing the meaning of its differential elements in relation to that fullness), and the antagonistic Other is supposed to threaten this fullness.30 However, although this type of antagonistic relation is indeed not a positive, differential relation, as Laclau and Mouffe rightly argue, this is not what precludes the possibility of its representation. ‘Antagonism’ precisely refers to and presupposes ‘a broader totality in relation to which the two poles of the antagonism would constitute’, if not differential, then at least objective, ‘partial instances’. Since antagonism presupposes the fullness of the community with a clear division between inside and outside, antagonism presupposes the space of signification within which both the community and its antagonistic other are constituted. With an antagonistic frontier, you have a clear inside and a clear outside. Hence, Laclau and Mouffe are both right and wrong when they write in Hegemony and Socialist Strategy that the ‘“experience” of the limit of all objectivity [has] a form of precise discursive presence, and this is antagonism’.31 Antagonism has a ‘precise discursive presence’: it is a discursive representation, but, as such, it is not the limit of objectivity or signification. Similarly, the empty signifier may be the limit of signification insofar as the latter is constituted through relations of difference. Yet, the empty signifier makes the signification of a space of fullness with clear boundaries possible. Both pure equivalence (or the pure emptiness of the empty signifier) and pure difference would constitute a set of 297

lasse thomassen fixed relations. The limit of objectivity, I suggest, is instead reflected in the mutual contamination of equivalence and difference. Slavoj Zˇizˇek has argued that antagonism is a way to externalise the ineliminable split (or lack) in the subject and, in this way, to discursively master the alwaysalready dislocated character of any identity.32 A good example is racist and xenophobic ideologies: if only we could get rid of the foreigners, then crime and unemployment would disappear . . . If the identity of the community is constituted through a hegemonic articulation, and if hegemony is essentially a relation of representation, then the identity of the community will be marked by an inherent split (or lack), which it can only erase by projecting it onto something represented as external to and negating the identity. We do not start with a pure inside; the inside is always already dislocated, and it is only the ‘negation’ of this dislocation—its externalisation—that creates the purity of the inside. As a consequence, Zˇizˇek argues, we should distinguish between dislocation and any discursive response to it, including antagonism. Since New Reflections (1990), Laclau has thought of identity and discourse in this Lacanian sense, as constituted around a lack and, as such, inherently dislocated, with antagonism being one way of discursively mastering dislocation, even if this always eventually fails.33 As Nathan Widder has rightly argued, ‘the turn to dislocation would be redundant if it necessarily manifests itself in antagonistic relations’.34 Laclau concludes: In Hegemony and Socialist Strategy the notion of limit is more or less synonymous with antagonistic frontier. Objectivity is only constituted through a radical exclusion. Later on I came to realize that this assimilation presented two flaws. The first, that antagonism is already a form of discursive inscription—i.e. of mastery—of something more primary which, from New Reflections on the Revolution of our Time onwards, I started calling ‘dislocation’. Not all dislocation needs to be constructed in an antagonistic way. The second flaw is that antagonism is not equivalent to radical exclusion. What it does is to dichotomize the social space, but both sides of the antagonistic relation are necessary in order to create a single space of representation.35

Bowman’s study of Palestinian nationalism provides a good example of this point. He shows how the Palestinian nation is defined through its antagonistic relationship with the state of Israel. At the same time, he argues, Palestinian identity is internally divided and only the result of contingent articulations.36 Whatever alternative antagonisms there may be within the Palestinian community occur within and are subsumed to the antagonistic frontier vis-a`-vis the state of Israel. At present at least, these ‘secondary’ antagonisms do not put the defining antagonism with Israel into question.37 However, the antagonistic frontier vis-a`vis Israel and the resultant fullness of the lost Palestinian identity masks, not an essence, but the essential lack of an essence, that is, the inherently dislocated character of identity. It is the task of hegemony analysis to examine whether, and how, antagonism is the response to dislocation, and how the construction of one antagonism may rely on the suppression of alternative antagonisms (even if the analysis cannot stop at this).38 As I shall argue in the conclusion, antagonism can be seen as ideological insofar as it conceals the dislocatory character of identity by 298

antagonism, hegemony and ideology after heterogeneity externalising the dislocation onto an external antagonistic force, thereby also establishing the unity of the discourse. Hegemony analysis and the relativisation of antagonism The relativisation of antagonism—antagonism is one possible discursive response to dislocation among several others, and there are only tendential antagonisms39— has implications for the strategy one pursues in analysing ideology. The hegemony analysis of ideology can neither assume antagonism to be always in existence nor can it consist only in looking for antagonisms to emerge. Hegemony analysis must look both ‘behind’ and ‘beyond’ antagonism. It must look ‘behind’ antagonism in order to see whether and why an antagonism was constructed, something only possible through a careful analysis of the historical context. It must also look ‘beyond’ antagonism in order to examine how the antagonism is never fully constituted and may subsequently be transformed. That is, hegemony analysis must take a dynamic perspective that takes antagonism as one possible outcome among others and not as the teleological aim of any identity formation. Significantly, this outcome is not the end of the matter, but instead requires further analysis of the possibilities of its transformation. Hegemony analysis does not necessarily aim at the identification of antagonisms, and even if an antagonism is identified, this cannot be the last word on the matter. A good example of the kind of hegemony analysis that takes antagonism as its endpoint is Sebastia´n Barros and Gustavo Castagnola’s attempt to explain the shape of Argentine politics after World War II. They argue that Peronism was not only a particular hegemonic project, but became the imaginary horizon of other hegemonic projects, thus setting the terms of Argentine politics long after its fall from power in 1955. Peronism shaped Argentina in two ways: it divided Argentine politics and society into two antagonistic camps, and these camps were Peronism and anti-Peronism—you were either for or against Peronism. Barros and Castagnola write that ‘Peronist populism introduced the representational resources which functioned as a “negative” imaginary precluding the stability of Argentine politics’. This ‘prevented the formation of a common imaginary sustaining a stable political order. Social differences were immediately read in terms of political exclusion. The political frontiers, thus framed by the constitution of political identities, prevented the emergence of a stable hegemonic articulation’.40 The problem with this analysis is the link between antagonism and instability. If the imaginary horizon—including the central antagonism—of Argentine politics was the same for almost half a century, then it seems wrong to talk about instability. There was stability with regard to the terms of political and social struggles, and there was a common imaginary shared by the whole political order, including the antagonistic forces. Barros and Castagnola write: ‘[t]his strict split of the political space into two fields overdetermined by an equivalential division prevented the constitution of two conditions for a stable hegemonic practice: the presence of a plurality of antagonistic forces and the instability of the frontiers separating them’.41 This is 299

lasse thomassen the crux of the matter, however. It is correct that hegemony is only possible insofar as frontiers and identities are ultimately unstable. In this sense, dislocation and heterogeneity, not antagonism, are the conditions of possibility (and impossibility) of hegemony. Since antagonisms divide the social field into two (for instance, Peronism and anti-Peronism), to have a plurality of antagonisms would require their relativisation. You cannot have both pure antagonism and instability of frontiers and identities, and the possibility of hegemony is linked to the relativisation of antagonism. Moreover, the constitution of the imaginary horizon (through the emptying of a signifier and the creation of equivalence) relies on the exclusion of those who are barred from representation within the imaginary horizon because the latter is not infinite but limited by the differential remainder. One must therefore, not stop at the identification of antagonistic frontiers and ideology, as Barros and Castagnola do; discourse analysis must go one step further and identify the discursive heterogeneity—the remainders of particularity— resulting from and making possible these antagonistic frontiers and identities. Heterogeneity, antagonism and discourse analysis In his most recent work, Laclau has introduced the notion of heterogeneity to refer to a discursive excess escaping categorisation and conceptual mastery. Heterogeneity stands ‘in an undecidable tension between internality and externality’ vis-a`-vis the boundaries of the discourse.42 As examples of heterogeneity, Laclau gives the lumpenproletariat in Marx’s work, the peoples without history in Hegel, and the subaltern.43 The point of this section is to develop a more systematic account of the category of heterogeneity in the context of the hegemonic approach to ideology and discourse analysis. I will, first, discuss Peter Stallybrass’s analysis of Marx’s notion of the lumpenproletariat in order to show how heterogeneity relates to antagonism. I will then further develop the notion of heterogeneity through a discussion of Georges Bataille’s notions of homogeneity and heterogeneity. The figure of the lumpenproletariat in Marx’s discourse is an example of a heterogeneous entity. The lumpenproletariat is a discursive excess escaping the conceptual categories of Marx’s analysis of capitalism, in particular the determination of the antagonistic relation between proletariat and the capitalist class. Yet, as Peter Stallybrass has argued,44 the lumpenproletariat not only shows the limit of the objectification of the relation between proletariat and bourgeoisie. The exclusion of the lumpenproletariat from the other categories—it belongs neither to the proletariat, nor to what is antagonistically opposed to the proletariat, namely the bourgeoisie—makes it possible to theorise the relation between proletariat and bourgeoisie as an antagonistic relation. The lumpenproletariat is precisely the ‘irreducible remainder’45—Laclau’s characterisation of heterogeneity—from the constitution of the identity of the proletariat, which is constituted through the antagonistic relation vis-a`-vis the bourgeoisie. The exclusion of heterogeneity from the chain of equivalence supports the unity of the chain and of the identity in question. However, this heterogeneity is not 300

antagonism, hegemony and ideology after heterogeneity excluded in an antagonistic fashion as opposed to the identity. Heterogeneity is excessive and undecidable; in the case of Marx’s lumpenproletariat, it escapes the attempt to conceptualise social relations in antagonistic terms. Yet, the ‘exclusion’ of the heterogeneous from the antagonism also makes the antagonism possible. In Laclauian terms, the condition of possibility of antagonism is the exclusion of the heterogeneous differential remainder. The heterogeneous is not excluded from the discourse—for instance, from Marx’s texts. It is not something external to the discourse, which would presuppose a closed discourse with clearly demarcated limits. That is why we can refer to it as discursive heterogeneity, and why it also undermines the antagonism. The heterogeneous does not simply disappear from the discourse. The existence of these heterogeneous elements shows the ultimate contingency of the constitution of an identity or a discourse, including antagonistic identities and discourses. Heterogeneity therefore, provides a privileged point of entry for ideology analysis. One must locate the heterogeneous elements in a discourse, examine what this heterogeneity is the trace of, and how it is dealt with in the discourse. Stallybrass’s analysis of Marx’s texts is exemplary in this regard. The lumpenproletariat provides a point of entry for the analysis of Marx’s texts tracing the former as an effect of Marx’s discursive decisions and examining how Marx deals with the resulting heterogeneous excess. We should not be led to think that, normatively, there is anything inherently progressive about heterogeneity. For instance, although Marx finds some revolutionary potential in the spontaneity of the lumpenproletariat, he also identifies the lumpenproletariat as a regressive force and as the foundation for the conservative discourse of Bonapartism. Georges Bataille’s analysis of ‘The Psychological Structure of Fascism’ is also telling in this regard. Bataille refers to homogeneity as ‘the commensurability of elements and the awareness of this commensurability’. ‘Production’, according to Bataille, ‘is the basis of social homogeneity’,46 and money is the equivalent through which each person exists: As agents of production, the workers fall within the framework of the social organization, but the homogeneous reduction as a rule only affects their wage-earning activity; they are integrated into the psychological homogeneity in terms of their behaviour on the job, but not generally as men [sic]. Outside of the factory, and even beyond its technical operations, a labourer is, with regard to a homogeneous person (boss, bureaucrat, etc.), a stranger, a man [sic] of another nature, of a non-reduced, non-subjugated nature.47

Science can only take homogeneity as its object of knowledge, thus excluding the possibility of a science of heterogeneity. Therefore, ‘it is necessary to posit the limits of science’s inherent tendencies, and to constitute a knowledge of the nonexplainable difference, which supposes the immediate access of the intellect to a body of material, prior to any intellectual reduction. Tentatively, it is enough to present the facts according to their nature’.48 Heterogeneity does not lend itself to ‘any intellectual reduction’ within a coherent and closed scientific system. This suggests that the study of the heterogeneous can only proceed through categories 301

lasse thomassen and examples—including the category of heterogeneity and the example of the lumpenproletariat—that gesture towards, but never appropriate, the heterogeneous. ‘[W]ith a view to defining the term heterogeneous’,49 Bataille, echoing Marx’s description of the lumpenproletariat, gives the following examples of heterogeneous ‘waste’ and ‘unproductive expenditure’: ‘the numerous elements or social forms that homogeneous society is powerless to assimilate: mobs, the warrior, aristocratic and impoverished classes, different types of violent individual or at least those who refuse the rule (madmen, leaders, poets, etc.)’.50 Pointing to these things makes Bataille able to talk about the heterogeneous as a positive experience without falling back upon the homogeneous to understand the heterogeneous and without viewing heterogeneity merely as the failure or negation of homogeneity.51 Here one must avoid two opposite pitfalls. On the one hand, as pointed out by Bataille, heterogeneity is not simply the negation of homogeneity, and one should avoid thinking of heterogeneity and homogeneity in a dualistic fashion. Heterogeneity is, rather, the simultaneous condition of possibility and impossibility of homogeneity. On the other hand, one must also avoid Bataille’s more or less implicit references to an immediate and vitalistic access to the world of heterogeneity. Instead, heterogeneity is inherently linked to representation. A related question concerns the relationship between the category of heterogeneity and the examples of it. One must avoid the temptation to reduce heterogeneity to Stallybrass’s, Bataille’s or Laclau’s examples or to any other example. Likewise, we must not think that these examples express an underlying essence. The notion of heterogeneity, as I have used it here, is what Derrida calls a ‘non-synonymous substitute’52 to name different discursive aporia, which could also be referred to in Laclauian terms as, for instance, the differential remainder or the tension between equivalence and difference. ‘Heterogeneity’, then, is neither a pure concept nor simply a name. It is rearticulated each time it is applied in concrete analyses. The explanatory force of the category of heterogeneity can, of course, only be shown through its use in concrete analyses; however, it can at least be used to pose a set of questions about the ideological nature of particular discourses.53 Social homogeneity, according to Bataille, may become dislocated by economic contradictions, causing elements to split off from the homogeneous sectors of society. These elements may then be articulated with already heterogeneous elements and form a new social formation. This is how Bataille accounts for the emergence of Fascism in the 1920 s and 1930 s.54 Despite Bataille’s tendency to reduce things to the economy in the last instance, from the perspective of hegemony, it is interesting to note that the existence of heterogeneity is the condition of possibility of hegemonic (re)articulation. What orthodox Marxism cannot explain, namely the emergence of the non-class ideology of Fascism, can be accounted for in this way. Fascism thrives upon the heterogeneity that cannot be accommodated within the relation between worker and capitalist. Likewise, in the case of Louis Bonaparte’s successful articulation of the lumpenproletariat, 302

antagonism, hegemony and ideology after heterogeneity the phenomenon of Louis Bonaparte is not reducible to the relation between the proletariat and the capitalist class.55 In both cases, the heterogeneous elements are articulated into an anti-system antagonism, even if that articulation is never completely successful. Significantly, the heterogeneous is not something wholly other as the example of the lumpenproletariat in Marx’s discourse might suggest. It is discursive heterogeneity, part of the discourse, which can, then, be rearticulated. To sum up, I suggest that the notion of heterogeneity is a non-synonymous substitute for different discursive aporia and for different ways in which we can refer to the limit of discursive objectivity. It stands in for other terms, yet it does not refer to some underlying principle or essence, which the other terms merely reflect. One can use heterogeneity to refer to the differential remainder, to the inherent split in the signifier and to the undecidable relationship between equivalence and difference.56 Similarly, we can say that there is something heterogeneous in the relationship between performative and citation. In these cases, we are dealing with the limit of objectivity. Yet, what is thus the limit to hegemonic articulation is simultaneously the condition of possibility of hegemonic articulation. Without the unstable relationship between equivalence and difference, for instance, no hegemonic articulation would be possible. In the terms of Hegemony and Socialist Strategy, heterogeneity not only appears in the undecidable relationship between equivalence and difference, but also in the fact that ‘[t]he transition from the “elements” [i.e. ‘any difference that is not discursively articulated’] ‘to the “moments” [i.e. ‘differential positions . . . articulated within a discourse’] is never entirely fulfilled’.57 We are dealing here with something that is both the condition of possibility and limit of hegemonic articulation. Laclau and Mouffe use the term ‘field of discursivity’ to refer to the inherent inability of any discourse to close itself as a totality: a [discursive] system only exists as a partial limitation of a ‘surplus of meaning’ which subverts it. Being inherent in any discursive formation, this ‘surplus’ is the necessary terrain for the constitution of every social practice. We will call it the field of discursivity . . . it determines at the same time the necessarily discursive character of any object, and the impossibility of any given discourse to implement a final suture.58

The field of discursivity refers to the discursive constitution of objects, and to the ultimate unfixity of moments within a discourse. Like heterogeneity, the field of discursivity is not something external to discourse in the sense of a region lying beyond the borders of the discourse. If the subversive character of the field of discursivity was due to its location in a region beyond the limits of the discourse, it would presuppose what it was supposed to subvert, namely the discourse as an inside with an outside. Rather, the field of discursivity is an inherent characteristic of any discourse, an internal limit to discourse. The field of discursivity is closely linked with what Laclau and Mouffe refers to as the ‘discursive exterior’ which ‘is constituted by other discourses’.59 The field of discursivity refers to the ultimate unfixity of the moments within the discourse, whereas the discursive exterior refers to the competing discourses potentially able to rearticulate the discursive 303

lasse thomassen moments. The field of discursivity and the discursive exterior are, thus, two sides of the same coin: without one, the other would hardly matter. In both cases we are dealing with a heterogeneity that cannot be discursively mastered—it is the heterogeneity that, in Bataille’s analysis, makes the fascist articulation of the workers possible (but also ultimately limits it). Indeed, we are dealing here with something undermining the very possibility of establishing a discursive inside with clearly demarcated boundaries. Hegemony analysis then involves not only the identification of emerging and persisting discourses, but also—as a way of examining the historical character of these discourses—the identification of the field of discursivity and the discursive outside. As with the lumpenproletariat in Marx’s texts, heterogeneity does not disappear from a more or less stable and hegemonic discourse. If there is no discourse or hegemony without heterogeneity, there is no hegemony analysis without attention to heterogeneity, that is, without consideration of what is simultaneously the condition of possibility and impossibility of hegemony. The identification of contingency and conditions of impossibility cannot simply be referred to as a specifically deconstructive move; rather, it is an inherent part of hegemony analysis. As a result, deconstruction and hegemony analysis cannot be distinguished in a dualistic fashion. Both are concerned with the conditions of possibility and impossibility of texts or discourses, and the conditions of possibility cannot be clearly distinguished from the conditions of impossibility. Similarly, whatever heterogeneity there may be in a discourse, it is not simply a heterogeneity preceding its supersession with the establishment of a hegemonic discourse and, for instance, an antagonistic frontier. Heterogeneity persists. Conclusion: Deconstruction and hegemony, heterogeneity and antagonism— and ideology The introduction of the category of heterogeneity into the hegemonic approach to ideology and discourse analysis has implications for the relationship between hegemony analysis and deconstruction, for the category of antagonism, and for ideology analysis. Heterogeneity refers to the simultaneous condition of possibility and impossibility of hegemonic articulations, including antagonism. Accordingly, hegemony analysis and deconstruction cannot be distinguished according to dualisms of possibility/impossibility, closure/contingency, or decision/undecidability. Both hegemony analysis and deconstruction address both sides of these dualisms, and they are not opposed and complementary discourse analytical strategies. The supposedly ‘deconstructive’ move is an inherent part of hegemony analysis (and vice versa). It is at least possible to conceive of hegemony analysis in this way, something that is more likely with the introduction of the category of heterogeneity, even if the theoretical tools—for instance the ‘field of discursivity’—were already there in previous formulations of the theory of hegemony. Indeed it could be argued that much of the argument made here about 304

antagonism, hegemony and ideology after heterogeneity heterogeneity could also have been made starting from the formulation of the theory of hegemony in Hegemony and Socialist Strategy, albeit with some rearticulation of it. Although I do not think it necessary to go outside the theory of hegemony in order to make the argument of this article, a similar, but not identical, argument may have been made starting from alternative theoretical sources. One possible source is Wittgenstein’s later philosophy;60 another source are theories of abundance inspired by Nietzsche and Deleuze, for instance the work of William Connolly.61 As for heterogeneity and antagonism, the conclusion is that, just as there are only tendentially empty signifiers, so there are only tendentially antagonistic frontiers. Pure antagonism is impossible, and we should rather speak of different degrees of antagonism. As a consequence, hegemony analysis can neither take antagonism as the necessary outcome of ideology and identity formation, nor, should it be the case, as the end of the process. Hegemony analysis must examine if and how discursive heterogeneity is articulated into antagonism, but also the heterogeneity created through the articulation of antagonism. This suggests that hegemony is not necessarily linked to antagonism, even in its tendential forms. Not only is Fig. 1 in need of qualification, it may not be the whole story either. It may well be the case that (relative) antagonism is restricted to particular ideologies, such as populism, which, in Laclau’s work, is associated with equivalence, empty signifiers and antagonism.62 However, the theory of hegemony may be able to cover non-antagonistic cases by also focusing on relations of difference as a way of constructing hegemonic discourses.63 The implications for the analysis of ideology have already been hinted at above. If discursive closure is ultimately impossible, if there is always something heterogeneous, then the misrecognition of this and the attempt to conceal it can be said to be ideological. Here, in the words of Laclau, ‘[t]he ideological would not consist of the misrecognition of a positive essence, but exactly the opposite: it would consist of the non-recognition . . . of the impossibility of ultimate closure’.64 In this context, heterogeneity provides a point of entry for the analysis of the ideological character of particular discourses. If ideology is identified as the attempt to conceal or exclude the heterogeneous, then ideology analysis taking this perspective should seek to identify the heterogeneous elements that resist attempts at closure as well as the various ways which agents deal with these elements. For instance, one might ask what are the strategies pursued by agents in the United States today in order to externalise and suppress what is termed ‘unpatriotic’ and ‘un-American’. Here the notion of antagonism is central, although in a different way than originally conceived by Laclau and Mouffe. Antagonism can be seen as one way of suppressing and externalising heterogeneity. It can be seen as a way of establishing coherence and closure.65 As such, antagonism is ideological: it is a strategy to achieve closure and to suppress its ultimate impossibility. One could argue that this is the case in the case of the ‘War on Terror’ and in the discourses analysed by Barros and Castagnola (Peronism) and Bowman (Palestinian nationalism). This is especially so when the antagonism is naturalised, and the 305

lasse thomassen antagonistic frontier and the opposing camps are endowed with allegedly noncontingent characteristics. Here antagonism is part of a naturalisation of what are properly speaking contingent phenomena, a naturalisation that makes these phenomena appear to be beyond politics and hegemonic dis- and rearticulation. What we are dealing with here is what Michael Freeden, coming from a different perspective, has called decontestation: the identities of the antagonistic camps as well as the antagonistic frontier itself are decontested and, thereby, depoliticised.66 Acknowledgements The author would like to thank Jason Glynos, Leonard Williams, Aletta Norval, Lars Tønder, Ernesto Laclau, Michael Freeden and two reviewers for the Journal of Political Ideologies for their comments on earlier versions of the argument.

Notes and references 1. E. Laclau and C. Mouffe, Hegemony and Socialist Strategy: Towards a Radical Democratic Politics, 2nd edition (London: Verso, 2002), p. 105. 2. E. Laclau, New Reflections on the Revolution of Our Time (London: Verso, 1990), p. 92; and E. Laclau, ‘The death and resurrection of the theory of ideology’, Journal of Political Ideologies, 1 (1996), pp. 201 –220. 3. For Laclau’s work, see Laclau and Mouffe, op. cit., Ref. 1; Laclau, New Reflections, op. cit., Ref. 2; E. Laclau, Emancipation(s) (London: Verso, 1996); and J. Butler, E. Laclau and S. Zˇizˇek, Contingency, Hegemony, Universality: Contemporary Dialogues on the Left (London: Verso, 2000). On the intellectual influences of Laclau and the ‘Essex School’, see J. Townshend, ‘Laclau and Mouffe’s hegemonic project: The story so far’, Political Studies, 52 (2004), pp. 269– 288; A. Norval, ‘Theorising hegemony: between deconstruction and psychoanalysis’, in L. Tønder and L. Thomassen (eds), Radical Democracy: Politics between Abundance and Lack (Manchester: Manchester University Press, forthcoming 2005); and Y. Stavrakakis, Lacan and the Political (London: Routledge, 1999). For discourse analytical studies using Laclau’s theory of hegemony, see the contributions in E. Laclau (ed.), The Making of Political Identities (London: Verso, 1994); D. Howarth, A. J. Norval and Y. Stavrakakis (eds), Discourse Theory and Political Analysis: Identities, Hegemonies and Social Change (Manchester: Manchester University Press, 2000); D. Howarth and J. Torfing (eds), Discourse Theory in European Politics (Basingstoke: Palgrave, 2004); and F. Panizza (ed.), Populism and the Mirror of Democracy (London: Verso, 2005). 4. For instance, A. Norval, Deconstructing Apartheid Discourse (London: Verso, 1996); and D. Howarth, ‘Complexities of identity/difference: Black consciousness ideology in South Africa’, Journal of Political Ideologies, 2 (1987), pp. 51 –78. 5. For this view, see E. Laclau, ‘Deconstruction, Pragmatism, Hegemony’, in C. Mouffe, Deconstruction and Pragmatism (London: Verso, 1996), pp. 47– 60; E. Laclau, ‘Discourse’, in R. E. Goodin and P. Pettit (eds), A Companion to Contemporary Political Philosophy (Oxford: Blackwell, 1993), p. 435; Laclau, Emancipation(s), op. cit., Ref. 3, pp. 78f, 90; Laclau and Mouffe, op. cit., Ref. 1, p. xi; J. Torfing, New ˚ kerstrøm Theories of Discourse: Laclau, Mouffe and Zˇizˇek (Oxford: Blackwell, 1999), pp. 102f; N.A Andersen, Discursive Analytical Strategies: Understanding Foucault, Koselleck, Laclau, Luhmann (Bristol: ˚ kerstrøm Andersen, ‘Political Administration’, in Howarth and The Policy Press, 2003), pp. 56 –62; N.A Torfing, op. cit., Ref. 3, pp. 142–145; and L. Phillips and M. Winther Jørgensen, Discourse Analysis as Theory and Method (London: Sage, 2002), pp. 48f. For a good use of a combination of deconstruction and Laclauian discourse theory, see Vicki Squire, ‘“Integration with diversity in modern Britain”: New Labour on nationality, immigration and asylum’, Journal of Political Ideologies, 10 (2005), pp. 51–74. 6. A. Norval, ‘Hegemony after deconstruction: the consequences of undecidability’, Journal of Political Ideologies, 9 (2004), pp. 139 –157. 7. E. Laclau, ‘Paul de Man and the politics of rhetoric’, Pretexts 7:2 (1998), pp. 154 and 156; E. Laclau, ‘Democracy between autonomy and heteronomy’, in O. Enwezor et al. (eds), Democracy Unrealized: Documenta11_Platform1 (Ostfieldern-Ruit: Hatje Cantz, 2002), p. 381; C. Pessoa et al., ‘Theory, democracy, and the Left: an interview with Ernesto Laclau’, Umbr@, (2001), pp. 9f; and E. Laclau,

306

antagonism, hegemony and ideology after heterogeneity

8.

9.

10. 11.

12. 13. 14. 15. 16. 17. 18.

19. 20. 21. 22. 23. 24. 25.

26. 27. 28. 29. 30.

On Populist Reason (London: Verso, 2005), Chapter 5. ‘Heterogeneity’ is used in a similar fashion in J. Derrida, Positions, 2nd edition (London: Continuum, 2002). Laclau and Mouffe, op. cit., Ref. 1, pp. 111f. In New Reflections (Laclau, op. cit., Ref. 2, pp. 17f, 172f), Laclau introduced the deconstructive notion of ‘constitutive outside’ from H. Staten, Wittgenstein and Derrida (Lincoln, NE: University of Nebraska Press, 1984), pp. 16 –18, 24. On Staten’s and Laclau’s uses of this notion, see L. Thomassen, ‘In/exclusions: towards a radical democratic approach to exclusion’, in Tønder and Thomassen, op. cit., Ref. 3. Laclau, ‘Deconstruction, Pragmatism, Hegemony’, op. cit., Ref. 5, pp. 47–60; Laclau, Emancipation(s), op. cit., Ref. 3, pp. 78f, 90; and Laclau and Mouffe, op. cit., Ref. 1, p. xi. Yet, Laclau also rightly acknowledges that Derrida himself theorises the decision, cf. Laclau, ‘Deconstruction, Pragmatism, Hegemony’, op. cit., Ref. 5, p. 48. A. J. Norval, ‘Hegemony after deconstruction: the consequences of undecidability’, Journal of Political Ideologies, 9 (2004), pp. 139–157. ˚ kerstrøm Andersen, Discursive Analytical Strategies, op. cit., Ref. 5, Torfing, op. cit., Ref. 5, pp. 102f, 300;A pp. 56–62; and Phillips and Jørgensen, op. cit., Ref. 5, pp. 48f. When distinguishing deconstruction from hegemony analysis, they all make reference to Laclau’s distinctions between deconstruction and hegemony and between undecidability and the decision. For a similar view of the complementarity of deconstruction and hegemony in relation to politics, see S. Critchley, ‘On Derrida’s Specters of Marx’, Philosophy & Social Criticism, 21:3 (1995), p. 21. Torfing, op. cit., Ref. 5, p. 103. Ibid. Here I am assuming that the distinction between hegemony and deconstruction is conceptual and not just methodological or heuristic. R. Gasche´, The Tain of the Mirror: Derrida and the Philosophy of Reflection (Cambridge, MA Harvard University Press, 1986). E. Laclau, ‘Constructing Universality’, in Butler, Laclau and Zˇizˇek, op. cit., Ref. 3, p. 303. In Laclau’s original model, the equivalence is represented by ¼ , but it is clear from the text that it should be ;, that is equivalence and not identity. For a good example of the use of Laclau’s theory for the analysis of populist ideology and a discussion of this point, see Y. Stavrakakis, ‘Antinomies of formalism: Laclau’s theory of populism and the lessons from religious populism in Greece’, Journal of Political Ideologies, 9 (2004), pp. 253–267. E. Laclau, ‘On the names of God’, in S. Golding (ed.), The Eight Technologies of Otherness (London: Routledge, 1997), p. 262. Laclau, op. cit., Ref. 15, p. 305. See also E. Laclau, ‘Glimpsing the future’, in S. Critchley and O. Marchart (eds), Laclau: A Critical Reader (London: Routledge, 2004), p. 281; and Laclau, ‘The death and resurrection of the theory of ideology’, op. cit., Ref. 2, p. 219. In the model, this is represented by the division (split) of the elements in the chain. E. Laclau, ‘Identity and Hegemony: The Role of Universality in the Constitution of Political Logics’, in Butler, Laclau and Zˇizˇek, op. cit., Ref. 3, p. 70. Laclau and Mouffe, op. cit., Ref. 1, p. 129. Laclau refers to the empty signifier as a name as opposed to a concept or description, cf. E. Laclau, ‘The uses of equality’, in Critchley and Marchart, op. cit., Ref. 18, pp. 342 –344. J. Derrida, Margins of Philosophy, trans. A. Bass (London: Harvester Wheatsheaf, 1982), pp. 307– 330. Laclau, Emancipation(s), op. cit., Ref. 3, pp. 41– 43; and E. Laclau, ‘Structure, History and the Political’, in Butler, Laclau and Zˇizˇek, op. cit., Ref. 3, p. 208. G. Bowman, ‘“A Country of Words”: Conceiving the Palestinian Nation from the Position of Exile’, in Laclau, The Making of Political Identities, op. cit., Ref. 3, p. 142. For good examples of similar analyses, see A. J. Norval, ‘Social Ambiguity and the Crisis of Apartheid’, in Laclau, The Making of Political Identities, op. cit., Ref. 3, Chapter 5; D. Howarth, ‘The difficult emergence of a democratic imaginary: Black Consciousness and non-racial democracy in South Africa’, in Howarth, Norval and Stavrakakis, op. cit., Ref. 3, pp. 175–177; and N. B. C ¸ elik, ‘The constitution and dissolution of the Kemalist imaginary’, in Howarth, Norval and Stavrakakis, op. cit., Ref. 3, pp. 200f. See, for instance, Bowman, op. cit., Ref. 24, p. 139. For an analogous problem in an analysis of post-war Italian anti-Fascism, see J. Martin, ‘Ideology and antagonism in modern Italy: poststructuralist reflections’, Critical Review of International Social and Political Philosophy, 8:2 (2005), p. 154. See, for instance, Bowman, op. cit., Ref. 24, p. 164. Laclau and Mouffe, op. cit., Ref. 1, p. 125. Laclau and Mouffe, Ibid, p. 126. My account of Laclau’s theory of hegemony enjoys the comfort of hindsight as the notion of antagonism is already present in Hegemony and Socialist Strategy (1985), whereas the notion of the empty signifier is

307

lasse thomassen

31. 32. 33.

34.

35. 36. 37. 38.

39. 40. 41. 42.

43. 44. 45. 46. 47. 48. 49. 50. 51. 52. 53. 54.

introduced in Emancipation(s) (1996). This is not to suggest that one can understand Laclau’s work as the teleological unfolding of his present position. As is evident, Laclau’s texts are themselves heterogeneous. On this, see L. Thomassen, ‘Reading radical democracy: a reply to Clive Barnett’, Political Geography, 24 (2005), pp. 631–639. Laclau and Mouffe, op. cit., Ref. 1, p. 122 (also p. 146). S. Zˇizˇek, ‘Beyond Discourse-Analysis’, in Laclau, New Reflections, op. cit., Ref. 2, pp. 251 –254. Laclau, New Reflections, op. cit., Ref. 2, pp. 39–41, 44f, 65; Laclau, ‘The death and resurrection of the theory of ideology’, op. cit., Ref. 2, pp. 203ff; Laclau in L. Worsham and G. A. Olson, ‘Hegemony and the Future of Democracy: Ernesto Laclau’s Political Philosophy’, in L. Worsham and G. Olson (eds), Race, Rhetoric and the Postcolonial (Albany: SUNY Press, 1998), p. 137; and Pessoa et al., op. cit., Ref. 7, p. 15. At times, Laclau and his commentators continue to treat antagonism as the limit of objectivity, for instance, Laclau, op. cit., Ref. 19, pp. 72 and 77. Dislocation may also be said to be a discursive construction because it depends on the discursive construction of what is dislocated and of what dislocates, cf. Laclau, op. cit., Ref. 18, p. 319; and T. B. Dyrberg, The Circular Structure of Power: Politics, Identity, Community (London: Verso, 1997), pp. 146– 148. N. Widder, ‘What’s lacking in the lack: a comment on the virtual’, Angelaki, 5 (3) (2000), p. 133 (note 23). For the argument that we need to distinguish dislocation and antagonism and understand the latter as merely one way among others to construct identities, see also A. J. Norval, ‘Frontiers in question’, Filozofski vestnik, 18:2 (1997), pp. 51–75; and U. Sta¨heli, ‘Competing figures of the limit: dispersion, transgression, antagonism, and indifference’, in Critchley and Marchart, op. cit., Ref. 18, pp. 234– 239. Laclau, op. cit., Ref. 18, pp. 318f. Bowman, op. cit., Ref. 24, p. 156. Bowman, Ibid, pp. 143, 155. In ‘Social Ambiguity and the Crisis of Apartheid’ (Norval, op. cit., Ref. 25, pp. 121f, 127, 131f), Aletta Norval does to the South African apartheid regime what Bowman does to the imagined Palestinian nation. Norval shows how, despite a seemingly clear antagonistic frontier dividing South African society into two camps, there are, in fact, criss-crossing frontiers. These competing frontiers not only undermine the stability and clarity of any particular frontier. When suppressed, they return as discursive heterogeneity or, in Norval’s terms, ‘ambiguity’ and ‘indeterminate elements’. Similarly, in the context of post-war Italy, James Martin (op. cit., Ref. 26, pp. 154, 156– 158) shows how the antagonism between fascism and anti-Fascism cannot suppress all potential divisions within the anti-Fascist camp because the meanings of ‘fascism’ and ‘anti-Fascism’ cannot be completely fixed. See Martin, op. cit., Ref. 26, for a similar position. S. Barros and G. Castagnola, ‘The political frontiers of the social: Argentine politics after Peronist populism (1955–1973)’, in Howarth, Norval, and Stavrakakis, op. cit., Ref. 3, p. 35. Barros and Castagnola, Ibid. For Laclau and Mouffe’s formulation of the same, see Laclau and Mouffe, op. cit., Ref. 1, p. 136. Laclau, ‘Paul de Man and the politics of rhetoric’, op. cit., Ref. 7, p. 156 (also p. 154); E. Laclau, ‘Politics, polemics and academics: an interview by Paul Bowman’, Parallax, 5:2 (1999), p. 103; and E. Laclau, ‘Can immanence explain social struggles?’, Diacritics, 31:4 (2001), p. 5. For an ambiguous characterisation of heterogeneity, see Laclau, ‘Democracy between autonomy and heteronomy’, op. cit., Ref. 7, p. 382. Pessoa et al., op. cit., Ref. 7, pp. 9f; Laclau, ‘Democracy between autonomy and heteronomy’, op. cit., Ref. 7, pp. 381f; and Laclau, On Populist Reason, op. cit., Ref. 7, Chapter 5. P. Stallybrass, ‘Marx and heterogeneity: thinking the Lumpenproletariat’, Representations, 31 (1991), pp. 69– 95. Laclau, ‘Democracy between Autonomy and Heteronomy’, op. cit., Ref. 7, p. 381. G. Bataille, ‘The Psychological Structure of Fascism’, trans. F. Botting and S. Wilson, in F. Botting and S. Wilson (eds), The Bataille Reader (Oxford: Blackwell, 1997), p. 122. It is ‘tendential homogeneity’, though. For Laclau’s use of Bataille, see Laclau, On Populist Reason, op. cit., Ref. 7, Chapter 5. Bataille, op. cit., Ref. 46, p. 123. Bataille, Ibid, p. 126 (also pp. 125 and 146 [note 3]). Bataille, Ibid, p. 126. Bataille, Ibid, p. 127. In addition to these things, Bataille refers to the unconscious, the sacred, and affect (pp. 126–128). Bataille, Ibid., pp. 126f. Derrida, op. cit., Ref. 22, pp. 12, 25f. On the difficulties of a post-structuralist approach to explanation, see the forthcoming work of Jason Glynos and David Howarth. Bataille, Ibid., pp. 125, 140, 142.

308

antagonism, hegemony and ideology after heterogeneity 55. Stallybrass, op. cit., Ref. 44, pp. 79ff. 56. Heterogeneity can also express the irreducible gap between what is represented (the absent fullness of the community, an absent universality) and the means of representation (a particular signifier). It refers simultaneously to a lack (the particular signifier is never quite up to the task) and an excess (the empty signifier is only tendentially empty because it contains too much difference). 57. Laclau and Mouffe, op. cit., Ref. 1, pp. 105, 110. 58. Laclau and Mouffe, Ibid, p. 111. 59. Laclau and Mouffe, Ibid, pp. 111 and 146 (note 20). 60. M. Freeden, Ideologies and Political Theory: A Conceptual Approach (Oxford: Oxford University Press, 1996), pp. 60–91; and Norval, op. cit., Ref. 34. 61. W. E. Connolly, ‘Review essay: twilight of the idols’, Philosophy and Social Criticism, 21:3 (1995), pp. 130 –136; Widder, op. cit., Ref. 34. 62. See Stavrakakis, op. cit., Ref. 16. 63. Norval, op. cit., Ref. 34. 64. Laclau, New Reflections, op. cit., Ref. 2, p. 92. See also Laclau, ‘The death and resurrection of the theory of ideology’, op. cit., Ref. 2, pp. 203 –207. 65. Cf. also Martin, op. cit., Ref. 26, pp. 152 and 154. 66. Freeden, op. cit., Ref. 60, p. 76. This is why I have suggested that a radical democratic approach to inclusion and exclusion needs to go beyond the notion of antagonism, cf. Thomassen, op. cit., Ref. 8.

309