Between autonomy and representation: toward a post-foundational discourse analytic framework for the study of horizontality and verticality

ABSTRACT This paper sets out to think the relationship between horizontality and verticality from the perspective of post-foundational discourse theory, taking as a starting point the diachronic development from Laclau’s and Mouffe’s joint work on radical democracy to Laclau’s theory of populism. The argument here is that the shift in conceptual terrain from the autonomy of ‘democratic struggles’ to the representative function of ‘empty’ popular signifiers points to deeper shifts and slippages – especially around the category of antagonism – as well as a more general tension between a horizontal politics of autonomy and a vertical politics of representation, for which radical democracy and populism respectively take on a paradigmatic character. While horizontality is predicated on the autonomous multiplicity and ‘multiplication of antagonisms’, verticality entails the simplification and concentration of antagonism around central representative signifiers. The question thus becomes how antagonism is organized, or – drawing on Nonhoff’s concept of contrariety – to what extent the contrarieties defining the constituent parts of an equivalential chain are more multiple or more concentrated. This is followed by empirical considerations on how horizontality and verticality are organizationally mediated within current political projects, including a distinction between movement parties and Volksparteien neuen Typs (‘people’s parties of a new type’).


Introduction
The manifold new forms of contentious politics of the past decade have provoked lively debate in academic and activist circles alike on the relationship between horizontality and verticality. Public intellectuals on the left have made prominent interventions in recent years in an attempt to rethink the horizontality-verticality relation as a strategic question for radical politics today (e.g. Hardt & Negri, 2017;Mouffe, 2018;Nunes, 2021;Žižek, 2015).
More than ever, the so-called movements of the squares such as 15-M, Aganaktismenoi, and Occupy Wall Street as well as their aftermathsincluding the electoral rise (and fall) of new left-wing parties such as Podemos and Syrizaappear to have brought the architectonics of horizontal/vertical to the very foreground of political thought, often in conjunction with the likes of left/right and bottom/up (a case in point being Mouffe's (2018) plea for a 'left populism' of the 'underdog' against 'those in power'). As Dyrberg (2003) has argued, spatial orientational metaphors such as left/right, up/down, front/back, and in/ out play a constitutive role in symbolically structuring political order and lend themselves to application for empirical analysis following the discourse theory of Mouffe (1985/2001). De Cleen and Stavrakakis (2017) have made a similar argument with their influential 'discursive architectonics' approach, proposing a distinction between nationalism and populism in terms of the 'horizontal' in/out architectonics of national vs. foreign in nationalism and the 'vertical' bottom/up architectonics of underdog vs. power in populism. Against the background of these recent advances within post-foundational discourse theory, this paper sets out to consider horizontality and verticality not only as forms of antagonistic frontier-building, but also as types of structuring relations internal to political organizations: put simply, what does it mean for a political party or movement to be 'horizontalist' or 'verticalist'? Here, the diachronic development of post-foundational discourse theory, with the shift in focus from Laclau's and Mouffe's early joint work on radical democracy to Laclau's later theory of populism, offers a particularly fruitful perspective by providing both a common conceptual vocabulary for the general study of discourse and the possibility of a focused reconstruction of shifts and slippages that point, in turn, to an underlying tension between horizontal and vertical politics. In short, the argument to be advanced in the following is that horizontality is predicated on the multiplicity and 'multiplication of antagonisms', whereas verticality entails the simplification and concentration of antagonism around a central representative struggle and the empty signifiers associated with it. In Laclau's and Mouffe's work, radical democracy and populism can be understood to take on a paradigmatic character for horizontal and vertical politics, respectively: a radical-democratic project entails the autonomy of 'democratic struggles', each of which constructs an antagonistic terrain against forms of domination specific to it (e.g. 'women's rights' against 'patriarchy', 'anti-racism' against 'racism', 'LGBT rights' against 'homophobia', etc.) prior to being intersectionally linked around a common struggle for 'democracy', whereas populism foregrounds the antagonistic frontier between a popular subject ('the people') and a power bloc ('the elite') as one taking on a representative function for other struggles. The distinction between a horizontal politics of autonomy and a vertical politics of representation is especially relevant in the present political conjuncture and, as will be argued, prototypically maps onto contrasting forms of party organization that have come to the fore in the past decade: 'movement parties' such as the CUP in Catalonia correspond to an emphasis on the horizontal integration of autonomously organized movement actors, whereas 'people's parties of a new type' (Volksparteien neuen Typs) such as Podemos and France Insoumise are characterized by vertical plebiscitary links between a leadership and a largely undifferentiated 'people' as the representative subject interpellated within party structures.
By keying in on Laclau's and Mouffe's workincluding its diachronic shift in focus from radical democracy to populismthis paper locates horizontality and verticality as two dimensions internal to a broadly post-foundational or post-Marxist understanding of hegemony, in contrast to numerous existing research contributions that, drawing on the likes of Hardt's and Negri's biopolitics or Rancièrian politics-as-equality, situate horizontality in a relationship of exteriority to hegemony or representation as a kind of post-hegemonic or non-representational form of politics (e.g. Arditi, 2007;Eklundh, 2017;Kioupkiolis, 2016;Kioupkiolis & Katsambekis, 2014;Prentoulis & Thomassen, 2013). This is not to deny the relevance or productiveness of these other approaches, but rather to tap into the advantages that a narrower focus can provide: in particular, the diachronic movement from radical democracy to populism in Laclau's and Mouffe's work within the overarching framework of post-foundational discourse theory points to deeper conceptual shifts and slippages around the central category of antagonismwhich, in turn, offer a useful springboard for developing a more systematic framework for the analysis of horizontality and verticality. In doing so, I draw on Nonhoff's (2006) concept of contrariety understood here as something like antagonism in miniaturewhich has found application in the German-speaking literature in particular (Kim, 2022;Marchart, 2013Marchart, , 2017Marttila, 2015) and proves to be useful for bridging the shifting understandings of antagonism within Laclau's and Mouffe's work. The argument in the following is first developed by way of a reconstruction of the theories of radical democracy and populism followed by a more conceptually focused discussion in relation to the category of antagonism and the implications for distinguishing between horizontality and verticality, before the final section circles back toward empirical considerations on horizontal and vertical politics in the present conjuncture.
Radical democracy and populism in Laclau and Mouffe: from the politics of autonomy to the politics of representation With a few notable exceptions (such as the concept of 'empty signifier'), the conceptual building blocks necessary for our discussion are already in place in the early joint work of Mouffe (1985/2001), namely Hegemony and Socialist Strategy (hereafter HSS). Here, the authors develop a 'post-foundational' (Marchart, 2007) theory of discourse that posits the contingent nature of discursively mediated social relations following the logics of difference and equivalence: if the construction of meaning and identity rests on the production of difference (A is A because it is distinct from B, C, D, …), the logic of equivalence holds out the possibility of breaking up and re-instituting existing fields of difference by drawing an antagonistic frontier that re-defines the coordinates of the social around new oppositions (A is distinct from B, C, and D, but united with them in the chain of equivalences A ≡ B ≡ C ≡ D against E). The ultimate 'unfixity', or contingency, of differential identities means that the social is constantly subject to the re-instituting intervention of the political (Laclau & Mouffe, 1985/2001: the attempt to fix the identities of signifiers such as A, B, C … is constantly susceptible to dislocations (Laclau, 1990), or interruptions in the structured production of meaning (e.g. the claim that B does not, in fact, belong with A, C, and D against E), which can give rise to ever newer reconfigurations of meaning via the re-drawing of antagonistic frontiers (e.g. B is now incorporated in the chain of equivalences B ≡ D ≡ E against A ≡ C). The political as antagonism thus takes on the ontological function of a precondition for all politicsa 'limit of all objectivity' (Laclau & Mouffe, 1985/2001) that makes visible the constitutive incompleteness and contestability of every social order. In this vein, Laclau's and Mouffe's discourse theory can be situated within a wider lineage of post-foundational ontologies of contingency that argue that social being is predicated not on stable transcendental grounds (whether this might be divine right, modernization, or the cunning of reason), but rather on the possibility of conflict, i.e. antagonism, as the limit to every attempt to invest the impossible object 'society' with a positive ground (Marchart, 2007(Marchart, , 2018. Antagonism in this ontological sense, in turn, manifests itself empirically in the form of specific antagonisms as effects of discursive articulations (e.g. the frontier separating A ≡ C from B ≡ D ≡ E in the previous example; see also Nonhoff, 2017); as will be seen, the question of how these antagonisms are organized will be crucial for understanding the differences between horizontal and vertical politics.
In HSS, Laclau and Mouffe couple this theoretical framework with a strategic plea for a radical-democratic politics as the most promising path for a hegemony project of the left in the context of Thatcherism and the New Social Movements. As I have argued elsewhere (Kim, 2020(Kim, , 2021, radical democracy following Laclau and Mouffe can be understood to consist of at least two dimensions: (1) an ontology of contingency, i.e. the recognition that discursively mediated social relations are contingent and, therefore, that politics as a never-ending struggle over hegemony cannot be predicated on an a priori privileged subject such as 'the working class'; and (2) a politics of autonomy that seeks to build a chain of equivalences of ever more 'democratic struggles' for liberty and equality rights ('feminism, anti-racism, the gay movement, etc.') as loci of antagonisms in their own right while linking them around the nodal point 'democracy' (Laclau & Mouffe, 1985/ 2001. Autonomy here means recognizing each and every 'democratic struggle' as constituting its own antagonistic terrain vis-à-vis forms of domination specific to it, such that 'the antagonisms within each of these relatively autonomized spaces divide them into two camps […] against objects constituted within their own space' (Laclau & Mouffe, 1985/2001: for instance, 'feminism' against 'patriarchy', 'anti-racism' against 'racism', 'LGBT rights' against 'heteronormativity', and so on. A politics of autonomy entails a 'principle of democratic equivalence', more than just a formal logic of equivalence (Laclau & Mouffe, 1985/2001: 'democratic struggles' such as feminism, antiracism, and LGBT rights are linked equivalentially not only in common opposition to all these forms of domination, but also by virtue of a mutual recognition of the autonomous particularity of these antagonisms. Radical democracy, in this sense, can be understood as a politics of intersectionality and horizontal solidarity: indeed, the strategic thrust of HSS is not least to refute the notion that the so-called New Social Movements 'are secondary struggles and that the struggle for the "seizure of power" in the classical sense is the only truly radical one' (Laclau & Mouffe, 1985/2001. Here, again, the defining link between contingency and autonomy can be seen: in recognizing the 'irreducible plurality of the social' (contingency), a project of radicalizing democracy must also embrace the irreducible multiplicity of struggles for democratic rights and aim at 'the multiplication of antagonisms and the construction of a plurality of spaces within which they can affirm themselves and develop' (autonomy) (Laclau & Mouffe, 1985/2001.
Seen this way, radical democracy following Laclau and Mouffe points in prototypical fashion to a horizontal politics of autonomywhich, at the same time, stands in a productive tension to a vertical logic of representation. The basic structure of horizontality entails a multiplicity of overlapping antagonisms that each define the constituent parts of the equivalential chain and are recognized as such by the others (e.g. 'feminism', 'anti-racism', and 'LGBT rights' are each autonomously constituted by, and then linked in common opposition to, the likes of 'patriarchy', 'racism', and 'homophobia'). The process of equivalential unification of these struggles, however, entails a hegemonic relation whereby a nodal pointin this case, 'democracy'takes on the function of a representative signifier for all these struggles. Representation (or 'hegemonic representation' in the preface to the 2001 edition) is a process whereby 'political practice constructs the interests it represents' (Laclau & Mouffe, 1985/2001: 'feminism', 'anti-racism', and ' LGBT rights' take on their identities as 'democratic struggles' by being articulated in terms of the nodal point 'democracy' in common opposition to the various forms of domination, in addition to their particular claims to liberty and equality rights. Here, Mouffe (1985/2001, p. 155) invoke the Lefortian notion of the 'democratic revolution' as a 'new mode of institution of the social', whereby claims to rights can be articulated in terms of their universalizing appeali.e. following a logic of equivalential expansion into ever more domains in the name of more 'democracy'. The tension between autonomy and representation is thus an irreducible, but also productive one insofar as the unifying appeal to 'democracy' helps to sustain the commitment to a radical-democratic politics of autonomy in turn: Mouffe (1985/2001, pp. 140-141) emphasize here that 'autonomy itself can only be defended and expanded in terms of a wider hegemonic struggle' that maintains a commitment to the autonomy of 'democratic struggles' in the first place. Indeed, it is precisely because of the radical contingency of the social that the mediating intervention of radical-democratic nodal points is needed so that claims to liberty and equality are articulated as part of an emancipatory project rather than, say, being differentially incorporated into the logic of neo-liberalism (e.g. Fraser's (2016) critique of 'progressive neoliberalism' or 'liberalized feminism') or equivalentially co-opted by far-right discourses claiming to defend women's or LGBT rights against the supposed intolerance of 'Islam' (with examples abounding from Pim Fortuyn to Marine Le Pen; see section 'Horizontality and Verticality Today').
In HSS, in short, Laclau and Mouffe stipulate a horizontal politics of autonomy as a constitutive dimension of radical democracy, while also situating it as the locus of a fundamental tension vis-à-vis a vertical politics of representation. In Laclau's On Populist Reason (hereafter OPR), published 20 years after HSS, a pronounced shift in the autonomy/representation balance can be observed in favor of the latter. This shift arguably begins with the more systematic use of the category of 'demand' as 'the elementary unit of politics' (Marchart, 2018, p. 111): the idea here is that the most basic political relation is produced by the articulation of a demand directed at a locus of power that is called upon to either fulfill or reject it. Central to Laclau's theoretical buildup in OPR is the distinction between 'democratic' and 'popular' demands: a 'democratic demand' is articulated in isolation, i.e. following the logic of difference, without generating an antagonistic terrain vis-à-vis the locus of power that it addresses; an antagonistic break only emerges when unfulfilled 'democratic demands' become 'popular demands' by forming an equivalential chain (i.e. following the logic of equivalence) in common opposition against a locus of power to 'constitute a broader social subjectivity' around an antagonistic frontier dividing the social into two camps (Laclau, 2005a, p. 74). Laclau (2005a, p. 162, emphases in original) takes this argument further to maintain that a 'tendentially empty signifier' (a concept first introduced in Laclau, 1996Laclau, /2007)in the case of populism, the name of a 'people' or, in the last instance, the name of a leadertakes on the function of 'a point of identification' that 'represents an equivalential chain' and, in so doing, 'constitutes that totality' as such via the act of naming. This understanding of representation as an operation that retroactively enacts the subject that is supposed to be representeda point arguably prefiguring the more recent 'constructivist turn' in representation studies (e.g. Disch, 2015;Saward, 2006) was already seen in HSS; in OPR, this argument is now made with a greater emphasis on a psychoanalytically informed notion of the performative effect of naming, whereby 'popular demands' take on their collective identity by virtue of being articulated in the name of 'the people' (and/or the name of a leader) as an empty signifier. 1 A shift in accent can be seen here between the theories of radical democracy and populism: whereas the autonomy of 'democratic struggles' constituted a defining dimension of radical democracy, it is now the logic of representation that takes center stage not only as the sine qua non for populist politics, but also 'the primary terrain of constitution of social objectivity'for which the construction of a 'people' in populism, in turn, constitutes a 'paradigmatic case' (Laclau, 2005a, p. 163; emphasis in original).

The horizontal and vertical politics of antagonism: toward an analytical framework
How is this shift to be made sense of? The contention of this paper is that, beyond strategic considerations in how Laclau and Mouffe set out to think the political 'within the conjuncture' in the 1980s and two decades laterwith radical democracy pointing to an opportunity to re-invent left-wing politics in the context of Thatcherism and the New Social Movements, while populism takes on a similar function in the era of financialized neo-liberalism and later the movements of the squares (see also Errejón & Mouffe, 2015;Mouffe, 2018)the passage from radical democracy to populism points to deeper shifts and slippages, especially in relation to the category of antagonism, and, as such, offers a useful springboard for developing a systematic framework for the analysis of horizontal and vertical politics more generally. To begin with, the conceptual displacement from 'democratic struggles' in HSS to 'democratic demands' in OPR as the basic unit of the chain of equivalences has important implications: whereas each 'democratic struggle' within the radical-democratic equivalential chain autonomously constitutes an antagonistic terrain in its own rightindeed, 'the autonomy of social movements […] is a requirement for the antagonism as such to emerge' (Laclau & Mouffe, 1985/2001) -'democratic demands', by contrast, cannot generate an antagonistic break on their own prior to being incorporated into the equivalential chain of 'popular demands'. Indeed, a kind of reversal can be seen whereby the representative agency of the empty signifier is ascribed a certain autonomy when Laclau (2005a, p. 93) refers to 'the construction of the "people" as a crystallization of a chain of equivalences in which the crystallizing instance has, in its autonomy, as much weight as the infrastructural chain of demands which made its emergence possible'. What is at stake here is no less than the question of what it takes for the political qua antagonism to emerge: in HSS, a radical-democratic bloc can only emerge on the basis of autonomously constituted 'democratic struggles', whereas in OPR, the antagonistic frontier of 'people' vs. 'power' can only emerge through the equivalential unification of 'popular demands', ultimately via 'the people' as an empty signifier. There is a slippage here in how the concept of antagonism itself is understood: in HSS, Mouffe (1985/2001, pp. 131-132) speak of the possibility that individual discursive elements can become 'points of antagonism' prior to being incorporated into a chain of equivalences, such that 'the antagonisms within each of these relatively autonomized spaces divide them into two camps'the examples here being the 'political space of the feminist struggle' or that of 'the anti-racist struggle'. In this vein, Mouffe (1985/2001, p. 133) even draw a distinction between 'democratic' and 'popular' struggles, referring to the latter as an exceptional situation in which individual antagonisms crystallize into a single overarching one between 'the dominant groups and the rest of the community'. In other words, what is held to be the standard version of the political in OPRthe crystallization of an antagonistic frontier dividing the social into two camps around chains of equivalences tied to empty signifiers, for which populism takes on a paradigmatic characteris considered the exception rather than the rule in HSS. In the latter, therefore, Laclau and Mouffe hold out the possibility that antagonism emerges within autonomous spaces prior to their incorporation into the 'popular' subjectivity that is at the center of attention in OPR.
It is here that the concept of contrariety, as introduced by Nonhoff (2006) and subsequently taken up especially in the German-speaking post-foundational discourse-analytic literature (Kim, 2022;Marchart, 2013Marchart, , 2017Marttila, 2015), is particularly useful for capturing this slippage in the understanding of antagonism. Contrariety refers to a relation of mutual exclusion between individual discursive elements such that A is opposed to B and vice versa; it is distinct from a relation of difference whereby A is merely not B. Nonhoff (2006) speaks here of a 'transfer' (Überführung) of difference into contrariety: relations of contrariety are predicated on difference but constitute a distinct logic beyond itnamely, that of a mutual incommensurability between two differential elements. 2 Understood thus, contrariety corresponds to something like antagonism in miniature and, indeed, what Laclau and Mouffe have in mind when they refer to 'democratic struggles' in HSS as loci of antagonisms in their own right: 'women's rights' against 'patriarchy', 'anti-racism' against 'racism', and 'LGBT rights' against 'heteronormativity' are all constituted via relations of contrariety and then linked to each other in relations of equivalence through the overlapping construction of these contrarieties. 3 Seen this way, contrariety precedes equivalence and, indeed, constitutes a requirement for the emergence of antagonistic frontiers; the equivalence A ≡ B presupposes common relations of contrariety between A and B to a third element, such as A ↔ Z and B ↔ Z. Nonhoff (2017) argues in this vein that an antagonistic frontier is always the result of the construction of two opposing chains of equivalences as well as the contrarieties between the individual elements constituting them (see Figure 1 for illustration, with the equivalential chain A ≡ B ≡ C ≡ D facing W ≡ X ≡ Y ≡ Z ). In this sense, the concept of contrariety is arguably presupposed, but not explicitly introduced, in Laclau's and Mouffe's theory of radical democracy; to return to their earlier terminology, 'democratic struggles' are autonomous precisely in the sense that they are each defined by their own contrarieties that then overlap to form a chain of equivalences.
In OPR, by contrast, Laclau tends to present the logic of equivalence in a simplified manner such that an equivalential chain of 'popular demands' faces a largely undifferentiated power bloc on the opposing side of the antagonistic frontier. Drawing on the example of an equivalential unification of demands against an 'oppressive regime' such as Tsarism, Laclau (2005a, p. 130) presents a diagram showing a chain of equivalences with D 1 as an empty signifier linking the demands D 2 , D 3 , and D 4 around it against the Tsarist camp T S . Notably, Laclau (2005a, pp. 131-140) uses this diagram as a starting point for a discussion of social heterogeneity and the contingent nature of identities, such that the 'popular demands' in a chain of equivalences can either be differentially incorporated by the system (e.g. when the tsar promises partial land reform) or equivalentially re-incorporated as floating signifiers onto the Tsarist side of the antagonistic frontier (e.g. when the tsar claims to be on the side of the peasants against Germany as a common national enemy). What is foreclosed in the discussion, however, is the possibility that individual demands can be autonomous without being merely differential, i.e. the possibility that they autonomously maintain contrarieties of their own prior to their incorporation into a chain of equivalences. 4 If, however, the same conceptual framework is applied whereby contrariety constitutes a precondition for equivalence, the equivalential chain D 1 ≡ D 2 ≡ D 3 ≡ D 4 is predicated on each of these demands being individually articulated in contrariety to the Tsarist regime (i.e. the demand for bread is blocked by the regime, as are the demands for peace, land, electrification, etc.). Figure 2 presents a modified version of Laclau's diagram with the relations of contrariety now made visible.
The contrast here to Figure 1 is telling: there is now a concentration of contrarieties around the opposing element T S , such that the popular demands on the protagonist side of the antagonistic frontier are not defined in terms of autonomous contrarieties of their own prior to their equivalential unification in common contrariety against the Tsarist regime. It is here that a distinctly vertical logic of representation can be seen: namely, the simplification of antagonism around empty signifiers ('people' vs. 'power') in this example, the Tsarist bloc T S as an empty signifier that (from the perspective of the 'popular' discourse) concentrates all the opposition to the 'popular' side around itself and thus comes to represent the opposing side in its entirety (following here Nonhoff's (2006Nonhoff's ( , 2017 operationalization of the empty signifier as the element that has relations of contrariety with all elements of the opposing equivalential chain). In empirical reality, to be sure, things are more complex and differentiated: the Tsarist bloc T S is itself constructed in equivalential terms by the 'popular' discoursewith distinct elements such as 'nobility', 'clergy', 'military', etc. forming an equivalential chain around a unifying Figure 2. Schematic illustration of a vertical equivalential discourse. designation such as 'Tsarist power'but the defining character of every populist discourse consists in the concentration of contrarieties around the empty signifiers standing for each equivalential bloc (the name of 'the people' and the name of power, e.g. 'Tsarist power' as a recurring signifier that comes to stand for the entirety of the power bloc). 5 We now have a basis for operationalizing the distinction between horizontal and vertical politics in discourse-analytic terms: horizontality is predicated on a multiplicity of contrarietiesor what Laclau and Mouffe referred to as a 'multiplication of antagonisms' that constitute autonomous loci of strugglewhereas verticality entails the simplification and concentration of contrarieties around a central representative signifier. Horizontality and verticality can be understood as two ends of a spectrum, for which autonomy and representation are the constituent principles and radical democracy and populism the paradigmatic forms, respectively. The key question for the autonomy/representation balance is not least that of how antagonism is organized: namely, to what extent the contrarieties within the equivalential chain are more multiple or more concentrated. At the same time, it is worth emphasizing that both horizontality and verticality, within the outlined framework, permeate all politics as a question of degree, not either/or: on a theoretical level, there cannot be total autonomy or total representation insofar as the identity of every signifier within a chain of equivalences is constitutively split between its differential or contrarietal particularity on the one hand and its attachment to the equivalential collectivity of which it is part on the other (see also Laclau & Mouffe, 1985/2001. The ultimately irreducible but potentially productive nature of this tension becomes particularly clear in the relationship between radical democracy and populism (see also Kim, 2020Kim, , 2021: it is easily conceivable that a radical-democratic politics of autonomy might be supplemented by references to popular representative signifiers (e.g. what Gerbaudo (2017) has analyzed as 'citizenism' in the movements of the squares) or that populist political projects might additionally feature autonomous local alliances rooted in social movements (e.g. Barcelona En Comú or the former Ahora Madrid in the case of Podemos). These examples, in turn, point to certain patterns of interaction or cross-fertilization between horizontal and vertical politics in the present conjuncturea topic to which the final section now turns.

Horizontality and verticality today: movement parties and Volksparteien neuen Typs
It has been pointed out that Laclau's and Mouffe's workincluding its diachronic shift in focus from radical democracy to populismhas taken on heightened relevance in a context of increasing attempts to construct 'vertical structures (party) rooted in horizontalist practices (protests, social movements)' (Thomassen, 2016, p. 173). The decade of the 2010s saw the global emergence of the so-called movements of the squares, characterized by their conspicuous lack of centralized leadership coupled with calls for 'real democracy now' in the name of the 'citizenry', the 'people', or the '99%', as well as the rise of a diverse array of political parties that sought to channel these protest discourses by means of novel forms of political organization. As Gerbaudo (2017, p. 7) has argued with his concept of 'citizenism' in analyzing the likes of 15-M, Occupy Wall Street, and Nuit debout, these movements themselves saw a productive synthesis between 'the neoanarchist method of horizontality and the populist demand for sovereignty': on the one hand, a radical-democratic politics of autonomous, decentralized assemblies; on the other hand, a populist invocation of the sovereignty of a popular underdog excluded by the elitewhat Gerbaudo (2017, p. 97) refers to as a passage from 'intersectionality' to 'trans-sectionality', which 'aims to transcend rather than simply ally diversity' through the appeal to a singular sovereign subject. Here, already, a fundamental tension between the politics of horizontality and verticality, autonomy and representation can be seenone that is magnified still further in the emergence of various party formations that have sought to performatively integrate these protest discourses into new forms of political organization. As I have argued elsewhere (Kim, 2021), a basic distinction can be drawn here between movement parties and people's parties of a new type (Volksparteien neuen Typs) 6 : movement parties tend to emphasize the horizontal integration of autonomously organized social movement actors, whereas the Volksparteien neuen Typs are characterized by vertical plebiscitary links between a leadership and a largely undifferentiated 'people' as the representative subject interpellated by party structures.
In contrast to 'movement parties' understood in an expansive sense to encompass all parties that engage in a combination of institutional and protest activity (e.g. Caiani & Císař, 2019;Della Porta et al., 2017), movement parties within the framework outlined here can be understood to refer more specifically to parties that practice a horizontal politics of autonomy that interpellates social movement actors, in their autonomy, as internal decision-making agentssuch as by giving them organized platforms within the party. The paradigmatic example of a movement party in this narrower sense is arguably the Popular Unity Candidacy (CUP) in Catalonia, which not only elevates democratic struggles such as 'anti-racism', 'ecologism' and 'feminism' to the status of programmatic pillars, but also consists of a decentralized network of autonomous 'local assemblies' as the basic organizing unit of the party that are then represented at the level of the 'national assembly'. Another, much more marginal and less visible, example can be seen in the case of the Ukrainian left-wing party project 'Social Movement', which was conceived as an attempt to bring together social movement actors (e.g. independent trade unionists) and, in turn, represent them with their own platforms within the party's organizational committee (e.g. 'the Defense of the Working Class Platform') (Kim, 2020). As these examples illustrate, the relationship between the principles of autonomy and representation is neither one of radical exteriority nor linear trade-off, insofar as a horizontal politics of autonomy requires representative structures in order to be instituted as such in the first place; as Prentoulis and Thomassen (2013, p. 181) have argued with reference to 15-M and Aganaktismenoi, '[t]here is no horizontality without verticality' insofar as horizontality is not a natural state but needs to be instituted through organizational structures. At the same time, there is very much a trade-off involved in the extent to which 'democratic struggles' can be constructed as autonomous within a logic of being represented via internal structures: instituting a 'women's platform', an 'LGBT platform', or indeed a 'trade union platform' within a party organization entails reducing the complex reality (and, indeed, the autonomy) of women's struggles, LGBT rights struggles, or labor struggles so as to make them representable in the form of a handful of activists operating in their name within the context of a political party or alliance. Here, the potentially productive but ultimately irreducible nature of the tension between autonomy and representation can be seen: the two can lead a mutually beneficial coexistence by generating multiplier effects for both the movement and party actors involvedwhat Mouffe (2018) refers to as 'synergies' in her discussion of left populism and radical democracybut there cannot be an indefinite expansion of the one that is not at the expense of the other (much like Laclau's (2005b, p. 261) argument that 'there is no square circle' for reconciling the populist, pluralist, and institutional dimensions of radicalizing democracy).
If movement parties interpellate autonomously organized movements as decisionmaking agents and representative subjects within the party, the Volksparteien neuen Typs interpellate a largely undifferentiated 'people' as the representative subject by reducing traditional barriers to participation and creating vertical plebiscitary links to the leadership, e.g. via internal referenda initiated by the latter and open to anyone registered onlinesimilarly in this vein to Gerbaudo's (2018) conceptualization of 'digital' or 'platform parties'. The paradigmatic examples here are Podemos, France Insoumise, and the Five Star Movement, with Podemos constituting a distinctive case of 'reflexive populism' (Kioupkiolis, 2016) that explicitly set out to build a party not around organized movements, but (in its initial phase) around the representative identifiability of a 'mediatic leadership' around Pablo Iglesias, known as the 'Podemos hypothesis' (see also Iglesias, 2015). It is worth noting that, at Podemos' launch event in January 2014, Iglesias went so far as to claim that the purpose of the new party was precisely not to 'represent' 15-M or any other social movement because 'the movements are unrepresentable' (Público, 2014); yet a vertical logic of representation can be seen in a more subtle way, such as in the January 2015 'March of Change', when Iglesias first thanked the 'thousands of youths on that 15th of May' for 'being here' again and then claimed that 'we haven't filled the Puerta del Sol in order to dream, but to make our dreams reality in 2015 […] we are going to win the elections over the People's Party' (La Marea, 2015; see also Kim, 2020). Here, the promise of 'winning' functions as a nodal point that incorporates a wide range of subject positions, including 15-M, against a common adversary without these subject positions being articulated in terms of autonomous contrarieties specific to themthus pointing in prototypical fashion to a vertical logic of representation. 7 The centrality of this promise of 'winning' and its verticalizing implications in justifying the leading role of the 'mediatic leadership' can be seen in other examplessuch as when Carolina Bescansa claimed in January 2015 that Iglesias' leadership deserves support because it stands for a 'Podemos for winning' as opposed to a 'Podemos for protesting' (InfoLibre, 2015). Even with Podemos' entry as a junior partner in a center-left coalition government in 2019 and its electoral decline from its initial peak of over 20%, the central representative signifier 'winning' has arguably remained a consistent feature in the party's discourse (see, for instance, an interview with the new party executive following its election in summer 2021: HuffPost.es, 2021).
Another area in which the foregoing discussion on horizontality and verticality takes on heightened relevance is in relation to the increasingly visible phenomenon of far-right appropriation of demands for women's and/or LGBT rights against the supposed threat of intolerant 'Islam' as a common enemy. Here, the discursive structure is distinctly inimical to autonomy: 'women's rights' or 'LGBT rights' are constructed as belonging to the side of 'the people' or 'the nation' solely by virtue of common opposition to 'Islam' and not in terms of autonomous contrarieties specific to women's or LGBT struggles, such as against 'patriarchy' or 'heteronormativity'. Indeed, the implication is that problems of discrimination faced by women or LGBT people are external to the national community and purely a result of imported 'foreign' cultures such as Islamas epitomized by Marine Le Pen's claim that it is unsafe to be 'a woman, homosexual, Jewish, or even French or white' in parts of France as long as there is a 'Muslim occupation' of the country (Le Parisien, 2016). In this context, therefore, a pronounced verticalism can be seen in the very logic of how 'democratic demands' are equivalentially articulated in contrariety to a single unifying Otherin direct contrast to the 'principle of democratic equivalence' in the form of autonomous contrarieties as invoked by Laclau and Mouffe in HSSeven prior to, in this case, questions of organizational mediation. This kind of articulatory verticalism is not unique to the far right, however: indeed, much of the debate around 'identity politics' within left-wing circles in numerous countries ultimately comes down to the question to what extent the particularity of 'democratic struggles' ought to be subordinated to a single unifying struggle for 'material' or 'class' interests, however these are then conceived (this being the position represented, for instance, with increasing polemicism by Sahra Wagenknecht within the German context).

Conclusion
The aim of this paper was to contribute toward a systematic framework for the analysis of horizontality and verticality following the post-foundational discourse theory of Laclau and Mouffe. In doing so, the paper takes as a springboard the diachronic movement from Laclau's and Mouffe's joint work on radical democracy to Laclau's theory of populism as well as deeper conceptual shifts and slippagesespecially in relation to the category of antagonismthat point to a fundamental tension between autonomy and representation that, in turn, characterizes the relationship between horizontality and verticality. It is here that a specifically post-foundational discourse-theoretical framework that situates both horizontality and verticality as dimensions internal to a politics of hegemonyas opposed to immanentist or biopolitically informed approaches that understand horizontality as an outside of hegemony or representationcan yield insights into the constitutive interplay between horizontal and vertical politics in the discourse-analytically informed terms of autonomy and representation, including the manifold possibilities of pursuing a politics of autonomy within projects of constructing hegemony. It is here also that the concept of contrariety as introduced by Nonhoff proves to be crucial for bridging the shifting understandings of antagonism in Laclau's and Mouffe's early joint work on radical democracy and Laclau's later theory of populism. While horizontality is predicated on a multiplicity of contrarieties defining each element within a chain of equivalences, verticality entails the simplification of antagonism around a central contrariety as represented by empty signifiers ('people' vs. 'power') on either side of the antagonistic frontier. The horizontality-verticality relation thus becomes a question of how antagonism is organizednamely, to what extent the contrarieties defining the constituent parts of an equivalential chain are more multiple or more concentratedas well as a question of organizational practices, which can in turn be understood as performative enactments of discursive articulations following the logics of contrariety, difference, and equivalence. This ultimately irreducible, but potentially productive, tension can be seen not least in the manifold attempts to combine a politics of autonomy and a politics of representation in the post-squares conjuncture, where the question is not least one of how horizontality and verticality are organizationally mediated in political practice and to what extent there is a relative preponderance of horizontal autonomy or vertical representation. In this vein, the distinction between movement parties and people's parties of a new type (Volksparteien neuen Typs) offers a classificatory scheme for understanding horizontal and vertical politics at work in empirical practice in relation to new and emerging forms of party organization. In sum, these reflections open up productive avenues for future research by offering not only a focused reading of post-foundational discourse theory in its diachronic development, but also rich potential for the further development of empirical typologies based on the dimensionality horizontal/vertical as well as systematic discourse analyses drawing on an expanded conceptual toolkit for the study of political organization. For practitioners of radical politics, the considerations on the dynamic tension between autonomy and representation can contribute toward a much-needed reflexive understanding of both the manifold possibilities and limits of combining horizontal and vertical politics today.
Notes signifiers (i.e. 'the people'-as-underdog as an empty signifier) and (2) the dichotomizing character of the people vs. power discourse (i.e. following the logic of equivalence). Seen this way, populism can be understood as conceptually distinct from (but empirically combinable with) the likes of nationalism, in which popular signifiers are either less central and/or constructed following a distinct architectonics of national vs. foreign rather than underdog vs. power (De Cleen & Stavrakakis, 2017). 6. The term is meant as a conscious amalgamation of the Germanophone notion of Volkspartei and the Leninist concept of the 'party of a new type', which has been re-fashioned in mainstream political science literatures to describe new parties ranging from the Greens (Müller-Rommel, 1982) to the far right (Gebhardt, 2013). Given this terminological background, the German rendition is used preferentially throughout the rest of the paper. 7. In other words, the Sawardian 'representative claim' of Podemos vis-à-vis 15-M is a very much implicit, yet (from the perspective outlined here) distinctly vertical one: 'we represent you by giving all of us (including 15-M) the possibility of winning against the PP'not 'we represent you by building our party organization specifically around 15-M activists' (this would have been the horizontal 'movement party' path that Iglesias & co. explicitly rejected).