erman - 'what is wrong with agonistic pluralism

24
Eva Erman What is wrong with agonistic pluralism? Reflections on conflict in democratic theory Abstract During the last couple of decades, concurrently with an increased awareness of the complexity of ethical conflicts, political theorists have directed attention to how constitutional democracy should cope with a fact of incommensurable doctrines. Poststructuralists such as Chantal Mouffe claim that ethical conflicts are fundamentally irreconcilable, which is indeed a view shared by many liberal theorists. The question of whether ethical conflicts are in principle irreconcilable is an important one since the answer has implications for what democratic institutions are desirable. In light of this question the article investigates the notion of conflict in agonistic pluralism and discourse theory. At first glance, Mouffe’s agonism seems apt to accommodate ethical conflict in democratic governance, since it focuses on conflict as the core of politics, whereas Habermasian deliberative democ- racy seems inappropriate for this task, as it focuses on consensus. However, through an inquiry into the conditions of conflict this article will argue the opposite, namely, that conflict cannot be adequately understood within Mouffe’s agonistic framework. The thesis defended is (1) that discourse theory offers a more accurate account of conflict than agonistic theory because it embraces the idea that deliberation is constitutive of conflict, and (2) that some of Habermas’ assumptions concerning ethical discourse need to be revised in order for his democratic theory to fully accommodate this insight. Key words agonistic pluralism · communicative action · discourse theory · ethical conflict · Jürgen Habermas · Chantal Mouffe PSC PHILOSOPHY & SOCIAL CRITICISM vol 35 no 9 pp. 1039–1062 Copyright © The Author(s), 2009. Reprints and permissions: http://www.sagepub.co.uk/journalsPermissions.nav http://psc.sagepub.com DOI: 10.1177/0191453709343385 at Ateneo de Manila University on March 8, 2015 psc.sagepub.com Downloaded from

Upload: kam-ho-m-wong

Post on 01-Oct-2015

230 views

Category:

Documents


1 download

DESCRIPTION

critique of the agonistic model, plus the necessity of modifying the communicative-deliberative model

TRANSCRIPT

  • Eva Erman

    What is wrong with agonisticpluralism?Reflections on conflict in democratictheory

    Abstract During the last couple of decades, concurrently with anincreased awareness of the complexity of ethical conflicts, political theoristshave directed attention to how constitutional democracy should cope witha fact of incommensurable doctrines. Poststructuralists such as ChantalMouffe claim that ethical conflicts are fundamentally irreconcilable, whichis indeed a view shared by many liberal theorists. The question of whetherethical conflicts are in principle irreconcilable is an important one since theanswer has implications for what democratic institutions are desirable. Inlight of this question the article investigates the notion of conflict in agonisticpluralism and discourse theory. At first glance, Mouffes agonism seems aptto accommodate ethical conflict in democratic governance, since it focuseson conflict as the core of politics, whereas Habermasian deliberative democ-racy seems inappropriate for this task, as it focuses on consensus. However,through an inquiry into the conditions of conflict this article will argue theopposite, namely, that conflict cannot be adequately understood withinMouffes agonistic framework. The thesis defended is (1) that discoursetheory offers a more accurate account of conflict than agonistic theorybecause it embraces the idea that deliberation is constitutive of conflict, and(2) that some of Habermas assumptions concerning ethical discourse needto be revised in order for his democratic theory to fully accommodate thisinsight.

    Key words agonistic pluralism communicative action discourse theory ethical conflict Jrgen Habermas Chantal Mouffe

    PSCPHILOSOPHY & SOCIAL CRITICISM vol 35 no 9 pp. 10391062

    Copyright The Author(s), 2009.Reprints and permissions: http://www.sagepub.co.uk/journalsPermissions.navhttp://psc.sagepub.com DOI: 10.1177/0191453709343385

    at Ateneo de Manila University on March 8, 2015psc.sagepub.comDownloaded from

  • Introduction

    During the last few decades, hand in hand with an increased emphasison ethical pluralism and the incommensurability of moral and religiouscomprehensive doctrines, democratic theorists have devoted more andmore attention to understanding ethical conflict, which previously hasbeen too little theorized. At the same time, this development has resultedin a shift of focus from moral questions of, for example, justification,towards a more pragmatic focus on political agreements or compromises.The most famous example is perhaps John Rawls political liberalism,which suggests that the institutions of a constitutional democracy shouldcope with a fact of incommensurable religious and moral doctrines.1Poststructuralists such as Chantal Mouffe take the position that ethicalconflicts are fundamentally irreconcilable, which is indeed a view sharedby many liberal theorists. The claim made by Mouffe is not that prospectsfor resolving ethical disagreement are unlikely, but that such conflictsare not even in principle reconcilable. This is the reason why Moufferegards social consensus as a dangerously utopian idea.

    As pointed out by Maeve Cooke, the question of whether ethicalconflicts are in principle irreconcilable is an important one, since theanswer has implications for what democratic institutions are desirable.If ethical conflicts can never be eradicated they must be dealt withthrough certain kinds of devised institutional arrangements.2 If they arenot irreconcilable, we should bet on institutions that implement mech-anisms and procedures for promoting cross-cultural dialogue and inter-ethical understanding. Indeed, in light of globalization, pluralization andfragmentization of societies, democratic theory must continue the trendof investigating the conditions of ethical conflict, since any democratictheory applicable to pluralist societies or to supranational arrangementsmust be able to lodge conflict in one way or another. A starting pointof this article is that a more profound knowledge of conflict is neededin order to reach a better understanding of how to handle conflicts, andthat such knowledge cannot be reached either through a presuppositionof incompatible comprehensive moral doctrines or a purification ofpolitics of an explicit normative dimension.

    In order to address the question of whether ethical conflicts are inprinciple irreconcilable this article investigates the notion of conflict indemocratic theory. More specifically, it will focus on Mouffes agonisticdemocracy and Habermasian discourse theory of democracy (delibera-tive democracy). At first glance, Mouffes agonism seems to be able toaccommodate ethical conflict in democratic governance, as it focuses onconflict as the core of politics. Indeed, this is probably one reason whyMouffe is one of the most frequently cited agonist theorists. Discoursetheory, by contrast, seems to be inappropriate for this task, as it focuses

    1040Philosophy & Social Criticism 35 (9)

    at Ateneo de Manila University on March 8, 2015psc.sagepub.comDownloaded from

  • on consensus and is criticized precisely for neglecting the conflictualdimension of the political process. However, through an inquiry into theconditions of conflict this article will argue the opposite, namely, thatconflict cannot be adequately understood within Mouffes agonisticframework. The thesis defended is that discourse theory offers a moreaccurate account of conflict than agonistic theory because it embracesthe idea that deliberation is constitutive of conflict, where deliberation isdefined as speech-acts oriented performatively towards validity-claims. Ifthis is a tenable position then it does not make sense to claim that ethicalconflicts are irreconcilable. However, in order for Habermas democra-tic theory to fully accommodate this constitutive thesis, some of its basicassumptions concerning the view of ethical discourse (including ethicalconflicts) need to be modified.

    The argument proceeds in three steps. First, I provide a brief sketchof Mouffes agonistic pluralism, partly through her criticism of Haber-masian deliberative theory, and a critical examination of her view ofconflict. My argument draws on previous criticism of agonism butdiverges by focusing on conflict rather than on consensus. Agonists havebeen criticized for being dependent on consensus, which is the very notionthey dismiss as impossible. But while Andrew Knops has convincinglyshown that Mouffes notion of agonism is dependent on deliberative pre-suppositions, it is argued here that such presuppositions are presumedby Mouffes notion of antagonism as well.3 My criticism of Mouffe istwofold: one part is directed towards the idea of antagonism, the othertowards the alleged transformation from antagonism to agonism viaMouffes ethico-political principles. Second, I investigate the conditionsof conflict by elaborating the thesis that deliberation is constitutive ofconflict. In this section I also respond to the objection that this consti-tutive idea neglects the fundamental distinction between communicativeand strategic action within discourse theory. The final section addressessome problems that Habermasian democratic theory faces in light of theproposed view of conflict.

    Conflict in Mouffes agonism

    Agonism implies a deep respect for the other and agonists attempt toreorient political theory towards the question of how to deal with irre-concilable differences. In contrast to, for example, liberal, deliberativeand multicultural views, which in different ways emphasize social co-operation, agonists are doubtful about the capacity of politics to over-come these differences through democracy. Contemporary society isbetter understood as contested and deeply divided, which has import-ant normative implications for how we are to understand democracy.

    1041Erman: What is wrong with agonistic pluralism?

    at Ateneo de Manila University on March 8, 2015psc.sagepub.comDownloaded from

  • A democratic system should be designed to provide arenas where citizenscan express their disagreements and where difference can be confronted.4

    In Mouffes view, liberal democratic societies are ill-equipped toconfront the challenges raised by pluralism due to the fact that they donot grasp its nature. The reason for this inability lies in the presumedpolitical theoretical framework which is rationalistic, universalistic andindividualistic in character.5 According to Mouffe, deliberative democ-racy represents the most recent paradigm of liberal democratic theory.Yet, this theory also fails to give a correct account of the political sinceit reformulates the idea of the public sphere by simply moving from aneconomic model to a moral one. Mouffe chooses to examine the Haber-masian version of deliberative democracy as this is in her view the mostsophisticated one.6

    Deliberative legitimacy in Mouffes reading is based on the idea thatdemocratic decisions are legitimate to the extent that they represent animpartial standpoint that is equally in the interest of all.7 The problemwith this consensus approach is not only empirical, since deliberativetheorists agree that consensus will only rarely be reached, but onto-logical. The deliberative universalistic understanding of the political over-looks important ontological aspects; most importantly, that the politicalis constituted through power, that antagonism is inevitable, and thatevery agreement is an expression of hegemonic power and thus unstable.These are the reasons why the deliberative notion of consensus must berejected.8 Deliberative democracy denies the political by defining it as aspace of (deliberative) freedom rather than a space of power and conflict.9

    Habermas assumes that the values articulated through the idealspeech situation are of guidance here, e.g. equality, openness and impar-tiality. Mouffe argues that, through the approximate realization of theconditions of such an ideal discourse, Habermasians presume that discus-sion will produce legitimate outcomes to the extent that it is orientatedtowards the generalizable (impartial) interests of the discourse partici-pants. Through reason-giving from the moral point of view the generalinterest will emerge.10

    It seems as if the core of the problem, in Mouffes view, lies not onlyor even primarily in the notion of consensus but in the deliberative(mis)understanding of conflict. She argues that no deliberation couldever take place without impediments to free and unconstrained publicdeliberation. Therefore, we have to conclude that the very conditionsof possibility of deliberation constitute at the same time the conditionsof impossibility of the ideal speech situation.11 Deliberative democracythus ends up in a paradox since the ideal speech situation demandsconditions from the outset, a shared set of discourse rules, which areontologically and conceptually impossible, preventing the possibility ofdeliberation.12 According to Mouffe, the agonistic model offers a more

    1042Philosophy & Social Criticism 35 (9)

    at Ateneo de Manila University on March 8, 2015psc.sagepub.comDownloaded from

  • viable solution to meet the demands of pluralism due to the fact that itis designed not to optimize the prospects of consensus but to optimizethe prospects for citizens to confront their clashing views.

    To understand the agonistic view we have to go to the bottom of thenotion of antagonism and the central role it plays for agonistic democ-racy. Antagonism on Mouffes account is an ineradicable conflict betweenus and them (friend and enemy). It is constitutive of the politicaland forms the limits of the social.13 Antagonism does not consist of areal opposition (A B), which occurs between two objects that havetheir own positivity independent of their relation to each other. Thus,there is nothing antagonistic in a crash between two cars.14 But neitheris antagonism a contradiction (A not-A), which occurs because Acannot be not-A at the same time. A contradiction too has to do withfull identities (i.e. it is an objective relation in the sense that it is madeintelligible by the fact that the objects already are), although this timebetween concepts rather than physical objects.15 Antagonism is differ-ent: the presence of the Other prevents me from being totally myself.The relation arises not from full totalities, but from the impossibilityof their constitution. Insofar as antagonism exists, I cannot be a fullpresence for myself.16

    Mouffe makes use of Derridas notion of the constitutive outside toexplain the emergence of antagonism. She argues that identity can beestablished only through an us/them distinction. Similarly, the democ-ratic logic always necessitates the drawing of a boundary between thosewho belong to the demos (us/friends) and those who are outside of it(them/enemies).17 Strictly speaking, antagonisms are external to societyas they constitute its limits.18 And since antagonism is ineradicable,pluralism should not be viewed as a mere fact, as liberals such as Rawlsdo, but as an axiological principle. Pluralism is constitutive at the con-ceptual level of the very nature of democracy.19 It implies the permanenceof antagonism and conflict. Conflicts should not be seen as disturbancesthat unfortunately cannot be completely eliminated, or as empiricalimpediments that prevent the realization of consensus. Against Habermasshe argues that the obstacles to consensus are not empirical.20 Rather,pluralism makes consensus a conceptual impossibility.21 In fact, thedeliberative attempt to negate the ineradicable character of antagonismby aiming at rational consensus constitutes the real threat to democ-racy,22 since it does not take into account that the aim of politics is alwaysto create an us (unity) by a determination of a them in a context ofconflict.23

    Starting out from these ontological assumptions about conflict, Mouffedraws the conclusion that democratic theory can never get rid of con-flict since the overcoming of the us/them opposition is an impossibility.Without an us and a them, no identity.24 The aim of democratic politics

    1043Erman: What is wrong with agonistic pluralism?

    at Ateneo de Manila University on March 8, 2015psc.sagepub.comDownloaded from

  • should rather be to transform the them from being perceived as enemiesto be destroyed to being recognized as friendly enemies, which meansto transform the conflictual relation from an antagonistic to an agonisticone. The latter is a relation between adversaries, i.e. between legitimateenemies who subscribe to the ethico-political principles of agonisticdemocracy.25 There is a crucial difference between antagonism and agon-ism: a relation is antagonistic when A and B have no common sym-bolic space, while it is agonistic when A and B share a commonsymbolic space but wish to organize it in different ways.26 Thus, byadhering to these principles participants turn to an adversarial mode ofinteraction and the hostile antagonistic conflict transforms to an exchangeof views where the disputants respect one anothers beliefs, and so on.27Through agonistic contestations participants can reach agreement andcommon political decisions. But these agreements are both temporaryand constituted by hegemonic power, and should always be looked uponas temporary respites in an ongoing confrontation.28 On Mouffesaccount, deliberative democracy neglects the hegemonic nature of conflictand thus the antagonistic dimension of democracy. Politics is a contestfor power, and legitimacy is nothing but successful power.29

    Problems with the agonistic approach

    Criticism has been raised against the agonistic approach for relying onseveral of the ideas embraced by the deliberative theory it attempts todistance itself from. The main target has been the agonistic notion ofconsensus, which is criticized for being implied but not spelled out.Obviously, with the purpose of elaborating a democratic theory, Mouffedoes not wish to accept antagonism between enemies as it is, which isthe reason why she pleads to some ethico-political principles that alldisputants must accept in order to transform this antagonism into demo-cratic agonism. As has been pointed out time and again, Mouffe is vagueon the contents of these normative principles, although she specifiesequality and liberty as important ingredients.30 In addition, in order toadhere to these principles some kind of consensus is needed, althoughMouffe is not clear about what kind of consensus this is supposed to be.The aim of this section is to show that the connection between conflictand deliberation not only concerns agonism, as pointed out by Knops,31but cuts deeper into the agonistic theory, all the way to the notion ofantagonism.

    At the core of Mouffes agonistic pluralism, inspired by poststruc-turalism, lies the acknowledgement and profound respect for differencein terms of the concrete Other. It is in the meeting between the I andthe Other (or between the us and the them) that antagonism arises.

    1044Philosophy & Social Criticism 35 (9)

    at Ateneo de Manila University on March 8, 2015psc.sagepub.comDownloaded from

  • To the extent that Mouffe were to equate antagonism with difference,I cannot see any problems with her argument that identity presupposesdifference (i.e. antagonism) and that difference is the condition of thepossibility of constituting unity and totality.32 Similarly, it would be aconceptual impossibility to imagine similarity or sameness without differ-ence. Within such an interpretive framework, Mouffes criticism of con-sensus would in fact make sense; in so far as antagonism equals differenceone could coherently draw the conclusion that consensus is a conceptualand metaphysical impossibility. Every action requires choice, which neces-sitates discrimination and thus exclusion. In this sense no consensus willever be a fully inclusive we.33

    But Mouffe clearly does not want to make a semantic and concep-tual point. Without doubt antagonism is supposed to do some norma-tive work. For Mouffe, antagonism is not difference per se, but concretedifference between us and them or me and the Other as real persons, asenemies. And this has political implications.34 On the one hand, Mouffedoes not want to relativize difference by reducing it to a purely descrip-tive notion.35 On the other, although antagonistic relations are not simplythere but construed in the meeting between subjects, she insists thatantagonism is solely about ontology, i.e. before the ought.36 Conse-quently, antagonism is ascribed a dual role. The problem is that thisduality hides the fact that, while difference is descriptive, antagonismis normative.37 One reason why this is difficult to detect in Mouffeswriting is that she often slides from using difference in the abstract anddescriptive sense, to a normative use of difference as conflict/antagon-ism. On the very same pages where she argues that unity presupposesdifference and that a complete reabsorption of alterity into oneness isimpossible, she draws the conclusion that the aim of democracy shouldbe to bring the excluded to the fore so that they can enter the terrainof contestation.38 This dual role thereby confuses particularity (whichis implied by difference in general) with individuality (a difference suchas antagonism) and neglects Hegels insight that the latter requiresspecific attitudes among the subjects involved.39

    It is argued here that conflict cannot be adequately understood withinan agonistic framework. Mouffes notion of antagonism fails because itdoes not embrace the idea that deliberation is constitutive of conflict.This has two implications: on the one hand, the notion of antagonismbecomes untenable, and on the other, it becomes impossible to explainhow antagonism can transform into agonism. Let us start with the firstproblem. While Knops argues that agonists take no notice of the fact thatdiscussants (adversaries) become aware of difference in discourse,40 I wishto take the argument one step further and claim that becoming awarepresupposes that a conflict is there to be aware of. But which conflictcould this be? An immanent feature of conflict is that it presupposes an

    1045Erman: What is wrong with agonistic pluralism?

    at Ateneo de Manila University on March 8, 2015psc.sagepub.comDownloaded from

  • element of X against Y, e.g. someone against someone else if the conflictis interpersonal. That is, although concrete differences are supposed tobe expressed and confronted on Mouffes political arena, two adversariescannot conceive of their differences as incommensurable or ineradicableand still compare (and thus confront) them. Neither could they drawthe conclusion that they are ineradicable without comparing them. Anycomparison must start out from common presumptions. Some kind ofconsensus is thus needed to even understand this against. For as soonas the two adversaries interact, they immediately presuppose a naverealism and thus take for granted that they speak about one and the sameworld.41 As emphasized by Habermas, the moment they stop taking thisfor granted their communication would lead to complete breakdown.42

    Indeed, Mouffe might reply that, even if this is true regarding themeeting between adversaries, it is not true of antagonists. But thisargument does not hold. To begin with, agonism and antagonism arenot two kinds of conflicts. Rather, in Mouffes own words, agonism isa different mode of manifestation of antagonism.43 This makes sense,since a manifestation always presupposes something behind, so to speak,something to become manifested. Mouffe makes a point of emphasizingthat, while adversaries are not enemies to be destroyed, they are stillenemies. So while antagonism can take two different forms, what Mouffecalls antagonism proper is its original form. Antagonists can only be-come adversaries under certain conditions (i.e. by accepting some ethico-political principles). If we leave agonistic conflict aside for a moment,how is it possible for antagonism proper to be a conflict between us andthem (or me and the Other) without any common symbolic space, touse Mouffes words? A shared symbolic space presupposes a commonunderstanding and would not make sense outside of an intersubjectivelinguistic context. In fact, the description of antagonists as not sharingany symbolic space is quite surprising against the background of Mouffesearlier work, where she goes against the Foucaultian distinction betweenthe discursive and the non-discursive and argues that every subject posi-tion is located within a discursive structure.44 So perhaps Mouffe usesthe expression no common symbolic space to portray the differencebetween antagonism and agonism within a thinner yet implicitly pre-supposed discursive (intersubjective) space. However, even on this morecharitable reading the notion of antagonism is flawed. The distinctionbetween subject and object (and between particular and general) canonly be meaningfully understood within an available conceptual andthus symbolic space, which presupposes a common understanding. Thismeans that the actors involved can only identify an antagonistic conflictas such through some common presumptions about each other as subjects.Unfortunately, following Schmitt, Mouffe would never attribute to antag-onism such a dimension of common understanding, since the point with

    1046Philosophy & Social Criticism 35 (9)

    at Ateneo de Manila University on March 8, 2015psc.sagepub.comDownloaded from

  • the distinction between friend and enemy is precisely that there is nocommon understanding at all.

    What is suggested here is thus that there could be no conflict withoutdeliberation, i.e. without speech-acts oriented performatively towardsvalidity-claims. Conflict is dependent on some shared idea of what is atstake. A possible objection to this view of conflict is that the presupposedcommon understanding of what is at stake at the most contributes aminimally shared empirical background knowledge of the situation, nota minimally shared normative consensus. But here we must ask: howcould the parties involved conceive of the situation as conflictual solelyon the basis of empirical knowledge about the outside world? Indeed,they must also value several aspects of it in specific ways for the conflictto emerge. In fact, as persuasively argued by Quine, settling ones mean-ings is a process that cannot be separated from settling ones beliefs.45So even if people are likely to disagree (and in a conflict they mostcertainly will) on which value is the most important and which reasonshould be decisive for who is right and who is wrong, by sharing anormative understanding of a conflictual situation as conflictual, theyhave already entered into the Sellarsian space of reasons.46 This is anormative space which cannot be reduced to an empirical descriptionof causal relations. Agents are not solely caused to act by their strongestdesire in a Humean sense. Rather, they take up a desire (or some otherpro-attitude) as a reason for action. And this endorsement of motiva-tion requires a capacity for (self-)reflection through which their desiresare coupled together with some conception they have of themselves.Moreover, the space of reasons is a social space, since this capacity forreflectivity takes place in a social setting reasons are essentially public.As emphasized by Kenneth Baynes, the normativity of reasons is depen-dent on the attitudes of mutual expectation that agents who ascribe oneanother the capacity to take a yes/no stand on validity-claims raise intheir speech-acts.47 To claim that A acts for a reason is to claim that Ais responsible or accountable for the action under (at least) that descrip-tion. And to claim that A is accountable is to claim that she is held tobe accountable by others.48 To act for a reason is to act under a norm,and norms are created in intersubjectivity, between subjects.49 AgainstMouffe, it is argued that the reason why antagonists must share a sym-bolic space in order to become antagonists is that both the empiricaland normative parts of the conflict emerge in it.

    There is another route to take in order to try to save the antagonisticnotion of conflict. One could claim that antagonists share no symbolicspace because antagonism is pure violence. In fact, it is sometimes hardto discern if Mouffe means to say that antagonism is violence, or if antag-onism is the possibility of violence.50 I think the latter reading is whatMouffe intends, because if antagonism would mean merely violence it

    1047Erman: What is wrong with agonistic pluralism?

    at Ateneo de Manila University on March 8, 2015psc.sagepub.comDownloaded from

  • would both be superfluous and hard to discern what normative workit would do for a democratic theory (of any kind). By antagonism Mouffewants to accentuate the possibility of violence in the meeting betweenthe I and the Other in every temporary agreement or consensus. To viewthe Other as an enemy to be destroyed does not mean that she actuallyis destroyed. So while violence is not a necessary condition for antagon-ism, the possibility of violence is. At the same time, to claim that everymeeting between two subjects involves a possibility of violence seems un-controversial. However, this goes for all relations, not only antagonisticones. Unless it is further qualified, a possibility always implies otherpossibilities as well. To the extent that Mouffe wishes to distinguishantagonistic relations from, say, agonistic relations, this would take usback to the problem of how agents could have access to this antagon-ism if they do not share any symbolic space. How could they know thatthe Other is an enemy to be destroyed rather than a friend?

    To argue that deliberation is constitutive of conflict might seemprovocative when considering horrifying events, such as genocide, whichare the outcomes of conflicts where there is seemingly no common under-standing or consensus whatsoever. But when persons are treated exclu-sively as objects, this would more appropriately go under the headingof violence. Of course, this view would not entail that there could be nosuch thing as political or social violence. However, when we distinguishpolitical violence from other kinds of violence through the epithet poli-tical, we do not usually refer to some essentialist element in the violentacts as such, but to the political causes of them. There are simply differ-ent causes for violent acts. Moreover, if we took a closer look at an em-pirical case of genocide, we would probably discern a conflict betweenpolitical leaders and at the same time find it more appropriate to assignto the actual genocide of civilians the label of violence.

    Now let us take a closer look at the second problem for agonistictheory, namely, that it cannot explain how antagonism can transforminto agonism.51 As was stated before, according to Mouffe agonism is acertain mode of manifestation of antagonism.52 However, even if antag-onism and agonism were to constitute two kinds of antagonisms (ratherthan two forms of one and the same), we would not know when, letalone how, we had succeeded in transforming from one (where we donot share any symbolic space) to the other (where we do share a sym-bolic space) without some common understanding, i.e. a shared symbolicspace. Since the subjects involved do not share any symbolic space untilthey have accepted the ethico-political principles and become adversaries,how can they accept some common principles before this moment? Thus,it seems as if we have to share a symbolic space not only as adversaries,as Mouffe claims, but as antagonists as well not only in order to identifyantagonism as such, as argued before, but also to be able to become

    1048Philosophy & Social Criticism 35 (9)

    at Ateneo de Manila University on March 8, 2015psc.sagepub.comDownloaded from

  • adversaries (i.e. legitimate enemies) and to know what it means to complywith some ethico-political principles.

    What Mouffe seems to suggest then is that the transformation fromantagonism to agonism is a moral choice that can be neither explainednor grounded. Let us follow this line of thought for a moment and assumethat some ethico-political principles simply exist and that we somehowhave access to them through introspection. Since antagonists make thismoral choice without sharing any intersubjective space, it must be akind of solipsistic exercise. In fact, the only way to make sense of theidea that antagonists are able to accept some common ethico-politicalprinciples is that each and every subject subscribes to them within hisor her own subjective world. But this leads to a subjectivism whichseems to go against Mouffes anti-subjectivist pretensions.53 Moreover, itpresumes that conflicts are interpersonal (or inter-group) while the intra-personal structure is already there, so to speak, ready to use to makemoral choices. Although, admittedly, Mouffe advocates a dynamic viewof antagonism where subjects are situated at a range of subject positions,making antagonism emerge, transform and disappear, the question ishow this dynamics is possible within an agonistic theoretical framework.For Mouffe, antagonism is always a conflictual relationship between theI and the concrete Other (or between us and them), i.e. antagonisticrelations are inter-dimensional. Her irreducibility thesis in fact prohibitsany other inference since antagonistic conflicts are there as soon as theI meets the Other and could at the most be manifested in a different way(by transforming to agonistic conflicts). To the extent that inescapableantagonistic conflict is an ontological starting-point of the analysis placing the interpersonal (or intergroup) dimension in the limelight anddefining pluralism as an axiological principle Mouffe has to presumethat the identity of the subjects involved is a premise of their agency.That is, she has to presume that an intrapersonal structure of some sortis already at the subjects disposal to make the exercise of agencypossible and thus the coming to the moral decision to transform fromantagonism to agonism.54

    Once again similar to liberals, Mouffes moral choice echoes aKantian idea of autonomy, i.e. that one is bound only by norms/rules(in this case the ethico-political principles) one has laid down for oneself.The question arises: If whatever I acknowledge as correct is correct, inwhat sense is it meaningful to speak about what I did as binding forme? As Robert Brandom argues, while the authority of the self-bindergoverns the force that attaches to a certain rule, that authority cannotextend also to the content of this rule. Because if it did, she has not byher endorsement really bound herself by a rule at all.55 The determinacyof the content of what you have committed yourself to is dependent onthe attitudes of others, which means that for me to be committed to a

    1049Erman: What is wrong with agonistic pluralism?

    at Ateneo de Manila University on March 8, 2015psc.sagepub.comDownloaded from

  • rule, I have to have acknowledged a commitment that others attributeto me.56

    Winding up, if we find it plausible that human interaction involvesboth an interpersonal and intrapersonal dimension, identity (and an intra-personal structure) cannot be a premise of agency which is the unfor-tunate consequence of Mouffes ontological statements but must be aproduct of it. It is something that must be achieved through deliberation.57Agency is not only an exercise of (interpersonal) self-determination, but atthe same time a cognitive exercise of (intrapersonal) self-interpretation.58We cannot exercise one without the other since they are interdependent.Consequently, conflicts do not only occur between persons (or groups)but also within them, and neither can emerge before or without deliber-ation. Instead of viewing antagonism as an experience of the limit ofthe social within the social itself and the Other as preventing me frombeing totally myself,59 I find it more plausible that the way to becometotally oneself to the extent this is possible is through the recogni-tion by others, not through their absence. This would lead to the conclu-sion that antagonism is not the experience of the limit of the social butrather an experience of the social, or, more simply put, a social experi-ence. When antagonists motivate themselves to make the moral choiceof becoming adversaries they are not solely caused to act. Rather, theydecide to act by reflecting on whether this choice fits with their concep-tion of themselves. And the assessment that it would be morally rightto accept the ethico-principles depends neither on the predictable conse-quences of the action nor on any property in the action itself, but onthe fact that such action could be justified to others in a moral discourse.Since reasons are inherently public, as acknowledged by Baynes, actingfor a reason requires a conception of ourselves as standing under theexpectation that we can justify the reasons for our action to others.60So even if we accepted Mouffes idea that the transformation from antag-onism to agonism involves a moral choice that is impossible to explainor ground, it would be hard to see what this choice would consist of atall. While something similar to good reasons for acting could be iden-tified on the basis of insight in the sense that we can be convinced thatour considerations are good ones, they gain their status as good reasonsin the possible agreement of others.61

    Some further notes on a discourse theoretical view of conflict

    In contrast to agonistic theory, a discourse theoretical view sustains thatwe cannot ontologically presume that certain conflicts are ineradicable,because we would not know which conflicts these are. Conflict originatesfrom the Latin word conflictus, where the prefix con- means together

    1050Philosophy & Social Criticism 35 (9)

    at Ateneo de Manila University on March 8, 2015psc.sagepub.comDownloaded from

  • or jointly. Thus, conflict not only presupposes difference but alsosomething in common, a common understanding. When deliberation isat work, through an interplay between an interpersonal and intra-personal dimension, conflicts (within and between people) both emergeand transform. The meeting with the other affects the search for coher-ence among my own conflicting values (intrapersonal dimension) at thesame time as the (potentially contested) relationship between me and theother (interpersonal dimension) is shaped by this inner undertaking.62

    To fully grasp these discourse theoretical assumptions about conflictwe must take a look at what kind of rationality is built into delibera-tion. But before we do so, a clarification concerning the notion of rationalconsensus needs to be made. Mouffe is wrong to presume that rationalagreements of democratic procedures for Habermas must reflect ageneralizable interest and be justified by impartial moral reasons. Thisonly concerns moral norms within his moral theory (discourse ethics),which should not be conflated with his democratic theory. Although boththeories draw on the same language-philosophical basis, there are crucialdifferences. Habermas discourse principle (D-principle) states that actionnorms are valid to the extent that all possibly affected persons couldagree as participants in rational discourses. Rational discourse includesany attempt to reach an understanding over problematic validity-claimsunder communicative conditions that enable all topics and contributionsto be freely processed.63 Concerning discourse ethics and the justificationof moral norms, however, the D-principle takes the form of a universal-ization principle (U-principle).64 It is this principle which privileges thegeneralizable interest, to use Mouffes words, and differentiates impar-tial moral reasons (i.e. agent-neutral reasons) from agent-relative ones.Moral norms are agent-neutral (impartial) all the way down to the reasonsthat justify them and moral reasons are reasons that cast off their agent-relative meaning.65

    If we move to the domain of democratic theory, Habermas formu-lates a notion of legitimacy and establishes a procedure of legitimatedecision-making (e.g. law) through a principle of democracy, which isa specification of the D-principle. What the U-principle and the demo-cratic principle have in common is that they express the underlyingmoral idea of the D-principle to give the interests of each person equalconsideration in discourse, e.g. by including as many voices as possible,by giving one another a respectful hearing, and by requiring everyoneto justify their views upon request (and, as we have seen, equality andtolerance are ingredients built into Mouffes ethico-political principles aswell). This moral idea reflects the mutual relations of recognition builtinto communicatively structured life-forms in general and is by Habermasreferred to as the moral point of view. But the moral point of viewshould not be conflated with the point of view from which moral norms

    1051Erman: What is wrong with agonistic pluralism?

    at Ateneo de Manila University on March 8, 2015psc.sagepub.comDownloaded from

  • are justified.66 What is overlooked by Mouffe is that the democratic prin-ciple allows room for the validation of all kinds of norms (not only moralones) which supply outcomes (decisions/law) with their legitimacy.67

    Mouffes own way of avoiding the notion of consensus is to speakabout a thin so-called conflictual consensus.68 In order to distinguishthis thin conflictual consensus from Habermas thicker one, Mouffe callsthe latter rational consensus and, as we have seen, connects it to thejustification of moral norms and to impartial generalizable interests. Butthis is not the basic meaning given to rational on the Habermasianaccount. Habermas does not view communication as the telos of lan-guage, but rather follows Humboldt and emphasizes understanding asits core. Speech-acts oriented towards understanding are indispensablefor the hermeneutic access to the world. Habermas argues, first, thatwe cannot understand an utterance without being acquainted with thereasons given for its validity, and, second, that we cannot understandsuch reasons without at least implicitly evaluating their validity. Itis in this sense that communicative action demands interpretations thatare rational.69 Thus, rationality is ascribed a pragmatic meaning.

    It is important to elucidate this pragmatic meaning in order torespond to one seemingly obvious objection against the defended thesis,namely, that the idea that deliberation is constitutive of conflict neglectsthe fundamental distinction between communicative and strategic actionwithin discourse theory. For not any kind of social interaction countsas communicative. And, the argument goes, the only requirement thatis needed for antagonism to appear is to form the idea that anotherpersons interest is an obstacle, and to decide to act in order to removeor reduce that obstacle. That is, the only thing that is required is Asinstrumental understanding of Bs interests as being in conflict with As.If this is true, it would imply that deliberative presuppositions are onlynecessary for conflict as agonism, where the respect for others beliefsimplies a shared moral framework, a point already made by Knops. Inorder to defend my thesis against this criticism, and show how antagon-ism too relies on such presuppositions, I will elaborate the relationshipbetween communicative and strategic action in relation to meaning,validity and contestability.

    One of the basic presumptions of Habermas social theory is thatlinguistic interactions such as strategic, figurative or symbolic action areparasitic on communicative action. Strategic actions alone cannot forma stable system of social interaction but would lead to pathological mis-understandings and instantaneous breakdown. To show that a conflictcannot consist of strategic action alone, but has to include the practiceof reason-giving (communicative action), we have to analyze this para-sitic feature on two levels: in connection to understanding and in con-nection to contestability. It is only if both levels are fulfilled that we candiscern a conflict.

    1052Philosophy & Social Criticism 35 (9)

    at Ateneo de Manila University on March 8, 2015psc.sagepub.comDownloaded from

  • The distinction between communicative and strategic action eluci-dates a difference between illocutions and perlocutions. Communicativeaction involves the illocutionary aim of reaching understanding, and,following Wittgenstein, Habermas sees the telos of reaching understand-ing as inherent in the linguistic medium itself. A communicative speech-act is not only supposed to be understood by the hearer, but also, as faras possible, to be accepted by her or him. The illocutionary success thusdepends on whether the act is sufficiently comprehensible and accept-able, i.e. it is proportionate to the intersubjective recognition accordedto the validity claim raised with it. By contrast, perlocutionary effectsof speech-acts are effects that can be brought about causally by non-linguistic actions. For even if strategic actions are social actions and assuch are linguistically mediated, language is not used communicativelyin the sense elucidated, but with an orientation toward consequences,i.e. perlocutions.70 While the illocutionary aims still dominate the perlocu-tionary effects in actions that are teleologically oriented toward success what Habermas calls weak communicative action, which is intertwinedwith purposive (teleological) rationality and whose validity-claims aresupported by agent-relative reasons this dominance disappears instrategic action. This, however, does not mean that strategic actions cando without illocutionary force. Rather, they derive their nourishmentfrom communicative action since perlocutions too require successfulillocutionary acts as their vehicle, i.e. they ride on the backs of illocu-tionary acts and hence require indirect mutual understanding.71

    If we were to apply this to the example above, A could only recog-nize Bs interests through speech-acts oriented performatively towardsvalidity-claims. Thus, the recognition on a pragmatic account must beunderstood as a communicative achievement. This means that there couldbe no purely instrumental understanding of someones interests. Strat-egic action is only possible if the participants understand one another,i.e. if they feed parasitically on a common linguistic knowledge and ona communicative competence. The difference is that in strategic actionthis competence is used indirectly in order to make each other under-stand what they want or believe.72 To this it might be responded thatantagonism on Mouffes account is not about a specific form of actionat all, but about a way of seeing the other. But on the pragmatic accountdefended here this would not make sense, since there could be no suchseeing without linguistic mediation of some sort between the partiesinvolved.

    It is thus argued that meaning and validity are internally connectedin the sense that a person understands a speech-act only when he or sheknows the conditions under which it may be accepted as valid. So theorientation towards the possible validity of utterances is not only partof the pragmatic conditions of reaching understanding but of linguisticunderstanding itself.73 Still, while the presumption that strategic action

    1053Erman: What is wrong with agonistic pluralism?

    at Ateneo de Manila University on March 8, 2015psc.sagepub.comDownloaded from

  • cannot occur without action oriented toward understanding says some-thing about the interconnection between social action and communica-tive action broadly speaking, it does not (yet) say anything about conflictspecifically. Indeed, we might act strategically toward others withoutthere being a conflict at all. For example, the slave-owner might treather slaves strategically in a non-conflictual way, i.e. coordinating heractions by exerting influence over them. In this case we might speak ofoppression but not (necessarily) of an antagonistic relation, i.e. of twoparties regarding each other as enemies to be destroyed. For such aconflict to emerge something must be contested. And, according to theview defended here, contestability is best understood as inseparable fromvalidity.

    As was stated before, strategic actions are social actions even if com-municative rationality is not directly at work. However, there are linguis-tically structured actions that are non-social (i.e. non-communicative);for example, for the epistemic purposes of pure representation of knowl-edge or for purposes of mentally planning an action. Such actions aremonological in the sense that their meaning content is independent ofthe illocutionary acts in which they might be embedded (monologicalhere meaning that they can be expressed intelligibly without referenceto a second person). Even if language in every case has to be acquiredcommunicatively, illocutionary acts do not play a fundamental role inthe epistemic use of language or in the calculation of action effects.74However, as soon as a speaker wishes to be taken seriously, or if whatshe or he says is questioned, an illocutionary force is immediately atwork. For then the speaker is required to justify to others what he orshe before considered monologically (by the raising of validity-claims inan argumentative practice). As soon as values are transformed into claimsof validity, social interaction has a built-in orientation toward intersub-jective recognition. In fact, in these non-communicative situations, theillocutionary force was never really absent, but merely abstracted awayby a temporary suspension of the reference to a second person.75

    The main point made here is this: for something to be contestable,thus for a conflict to emerge, it must involve an illocutionary force, foronly illocutionary acts that can be valid or invalid may be contested. Forexample, in a conflictual situation imperatives or announcements areoften expressed, such as we will not give up this fight, you are wrong,this is our land, and leave my property now. Of course, neither imper-atives nor announcements aim at agreement in the strong sense whatHabermas calls strong communicative action, in which validity-claimsare supported by agent-independent reasons but they move within thehorizon of a mutual understanding based on validity-claims and thusstill within the domain of communicative rationality.76

    1054Philosophy & Social Criticism 35 (9)

    at Ateneo de Manila University on March 8, 2015psc.sagepub.comDownloaded from

  • Some challenges for Habermasian democratic theory

    The aim so far has been to elaborate the first part of the thesis, thatdiscourse theory offers a more accurate account of conflict than Mouffesagonistic theory because it harbours the idea that deliberation is consti-tutive of conflict. Against Mouffe, I have tried to show that it does notmake sense to presume that ethical conflicts are irreconcilable in prin-ciple. It is important to keep in mind, however, that this is not a prog-nosis of the possibility of reconciling ethical conflicts. Indeed, mostprobably ethical conflicts will continue to be a part of any modernpluralist democracy. But the reason why ethical discussions fail, I wouldargue, is not due to the ontological essence of ethical conflict, but couldinstead be explained by analyzing the impediments that prevent theirgetting started, most commonly unequal power relations. Apart frombeing wrongly construed, it is in fact hard to see the attractiveness ofMouffes notion of conflict. Participants in intersubjective communica-tion have the tools to recognize these impediments and perhaps try todo something about them in order to improve the possibility of reachingunderstanding or a reason-based substantive compromise of some sort.If, however, we exclusively attribute the explanatory force of ethicalconflicts to the ontological fact of irreconcilability, this might have theeffect of preserving unequal power relations rather than trying to identifyand remedy them. It thus seems as if Mouffe draws too hasty conclu-sions from her premises. Just because we know for a fact that everyactual agreement is impregnated with power, it does not follow (at leastnot without further argumentation on Mouffes part) that all agreementsare equally impregnated with power, i.e. are equally bad or unjustified.77In fact, in order even to be able to draw the latter conclusion it seemslikely that participants must engage in the game of giving and askingfor reasons in an open-ended critical public deliberation.

    We will now move to the second part of the thesis, namely, that somemodifications are needed in order for Habermas democratic theory tofully accommodate the idea that deliberation is constitutive of conflict.More specifically, one important issue is in need of further elaboration:Habermas definition of ethical discourse.78 Although I agree with muchof the criticism raised against Habermas sharp distinctions betweenmoral, ethical and pragmatic discourses, my focus here is primarilydirected to the narrower question of how ethical discourse as such isconstrued and the kind of validity claim Habermas connects to it. ForHabermas, discussions of ethical matters go under the heading of ethical(or ethical-existential/ethical-political) discourses. In contrast to moraldiscourses, which are universal and concern the equal respect of allhuman beings and what is equally good for all, ethical discourses areparticularistic and concern what is best for me or for us from our point

    1055Erman: What is wrong with agonistic pluralism?

    at Ateneo de Manila University on March 8, 2015psc.sagepub.comDownloaded from

  • of view.79 Thus, ethical deliberation is foremost a hermeneutic explica-tion of individual or collective self-understandings or value-orientations.Interlocutors of ethical discourses are participants in processes of self-clarification and self-realization, guided by the validity-claim of authen-ticity.80

    As we have seen, the constitutive thesis defended here makes thepresumption that human interaction embraces an interpersonal and anintrapersonal dimension. Since agency is simultaneously an exercise ofself-determination (inter-dimensional) and a cognitive exercise of self-interpretation (intra-dimensional), conflicts both emerge and transformbetween and within persons through deliberation. This dynamic cannotbe accounted for by Habermas narrow definition of ethical discourse.In his view, ethical discourse is guided by the validity-claim of authen-ticity and thus concerns an intra-dimension, i.e. about who I or wewant to be, while moral discourse is exclusively connected to the inter-dimension, i.e. about what I or we ought to do. If the constitutive thesisis plausible, however, questions about who I (or we) want to be andquestions about what we ought to do are interdependent and cannot beseparated. Even if Habermas acknowledges that all three validity-claimscan be criticized in moral discourses, the opposite is not the case.81 Andit is exactly the exclusion of an ought to do dimension that makesHabermas characterization of ethical discourses too narrow. As empha-sized by Cooke, Habermas does not sufficiently acknowledge the dimen-sions of struggle against and challenge to established (often officiallylegitimated) perceptions of goals and self-understandings.82 So ratherthan dismissing ethical matters to the private sphere where they canbe protected against the critical scrutiny of public deliberation ethicaldiscourses are better viewed as a kind of political discourse. Like anyother validity-claims, ethical claims have a cognitive and thus universal-ist dimension,83 they rest on the presupposition that, if valid, everyonewould have to accept their validity. Therefore, a privatization of ethicalcommitments restricts the recognition of their validity to those who sharethe relevant substantive ideas of the good life and frustrates a vitalaspect of the subjectification process, namely, the strive towards univer-sal recognition of the validity of deeply held ethical beliefs.84 Consider,for example, discourses in which particular group needs or collectiveself-understandings are articulated and critically examined. It is doubtfulwhether such political discourses on ethical matters are guided evenprimarily by the norm of authenticity.85 So while Habermas, like Mouffe,recognizes that any consensus is ethically patterned, he draws thewrong inference from it, ending up mistrusting the notion of rationalconsensus on ethical matters.

    1056Philosophy & Social Criticism 35 (9)

    at Ateneo de Manila University on March 8, 2015psc.sagepub.comDownloaded from

  • Conclusion

    The aim of this article has been to examine the notion of conflict in demo-cratic theory against the background of the larger question of whetherethical conflicts are irreconcilable in principle. The last couple of years,concurrently with an increased awareness of conflict and pluralism havingto be taken into account more seriously when theorizing about the condi-tions of democracy, Mouffes agonistic pluralism has gained support fromdemocratic theorists and attracted a broad spectrum of social scientistsand activists. Although the agonist idea to place conflict at the heartof politics is appealing, this article has argued that conflict cannot beadequately understood within Mouffes agonistic framework. An inquiryinto the conditions of conflict exposes that, due to the fact that Mouffedoes not embrace the idea that deliberation is constitutive of conflict,the notion of antagonism is flawed. I have tried to show that it is impos-sible both to make sense of the notion of antagonism and to demonstratehow antagonism on Mouffes account can transform into agonism. Thearticle has defended the thesis that discourse theory offers a more plaus-ible ontological understanding of conflict than agonism because it canaccommodate precisely this insight about the relationship between con-flict and speech-acts oriented performatively towards validity-claims.

    It would be unfortunate to follow agonists and define antagonisticconflicts as irreducible, because which conflicts would we have in mind?In fact, I see a possible danger in the tendency within contemporary poli-tical theory to start out from the ideas of incommensurable conflict andfact of pluralism. If these ideas are not carefully elaborated, they mightprohibit a deeper understanding of conflicts, both concerning how theyemerge and what they consist of. This is related to another tendency,namely that of mystifying the other. Carried to an extreme, the otheris treated like an extraterrestrial being who is always violated by beingmisunderstood, suggesting that understanding comes from within thesubject itself rather than from the understanding between the I and theother. Against this I have argued that understanding and mutual under-standing must be seen as inextricably connected and, in the words ofBrandom, that Nothing is absolutely other.86

    Stockholm University, Sweden

    Notes

    I am much indebted to Arash Abizadeh, Stefan Rummens, Andrew Knops, SofiaNsstrm, Ulf Mrkenstam and Niklas Mller for their constructive comments

    1057Erman: What is wrong with agonistic pluralism?

    PSC

    at Ateneo de Manila University on March 8, 2015psc.sagepub.comDownloaded from

  • on earlier drafts of this paper. Thanks also go to the participants of the NOPSAconference in Troms, Norway (August 58, 2008), in particular, Johan Hyrn,as well as to the Swedish Research Council for financing my research projecton human rights, deliberative democracy and conflict resolution.

    1 John Rawls, Political Liberalism (New York: Columbia University Press,1993). Rawls assumes that it is a human condition that people disagree withone another about matters of deep moral and religious concern. However,in contrast to Mouffes scepticism, he remains open to the question of truthconcerning moral matters. See John Rawls, The Law of Peoples (Cambridge,MA: Harvard University Press, 1999).

    2 Maeve Cooke, Are Ethical Conflicts Irreconcilable?, Philosophy & SocialCriticism 23(2) (2007): 119.

    3 Andrew Knops, Debate: Agonism as Deliberation On Mouffes Theoryof Democracy, The Journal of Political Philosophy 15(1) (2007): 11526.

    4 See, for example, the works by Chantal Mouffe, Ernesto Laclau, WilliamConnolly, and Bonnie Honig.

    5 Chantal Mouffe, Deliberative Democracy or Agonistic Pluralism?, SocialResearch 66 (3) (1999): p. 745.

    6 ibid., p. 746.7 ibid., p. 747. In fact, this reading is common of theorists sympathetic to

    agonism. Andrew Schaap, for example, states that citizens in a deliberativesetting are not just supposed to assert their own particular interests butshould be able to articulate them in terms of general moral principles thatall can accept. Andrew Schaap, Agonism in Divided Societies, Philosophy& Social Criticism 32(2) (2006): p. 257.

    8 Chantal Mouffe, The Democratic Paradox (London: Verso, 2000),pp. 99100.

    9 Mouffe, On the Political (New York: Routledge, 2005), p. 9.10 Mouffe, Deliberative Democracy, p. 748.11 ibid., pp. 7512.12 ibid., p. 752.13 In most writings, Mouffe describes antagonism as being inherent in all

    human relations, emerging in the meeting between the I and the Other.Thus, the opposition between us and them is necessarily an antagonisticone. Antagonism is ineradicable and constitutive of the political. However,it is possible to discern a different view, where the us/them is seen as a plaindifference that could be transformed into an antagonistic friend/enemy. Forexample, Mouffe writes that identities are established on the mode of anus/them perceived as simple difference, which under certain circumstancesare transformed into antagonistic relations. From that moment it becomespolitical. This is why antagonism constitutes an ever-present possibility inpolitics (The Democratic Paradox, p. 13).

    14 Ernesto Laclau and Chantal Mouffe, Hegemony and Socialist Strategy:Towards a Radical Democratic Politics, 2nd edition (London: Verso, 2001),p. 122.

    15 ibid., p. 124.16 ibid., p. 125.17 Mouffe, The Democratic Paradox, p. 4.

    1058Philosophy & Social Criticism 35 (9)

    at Ateneo de Manila University on March 8, 2015psc.sagepub.comDownloaded from

  • 18 It is not clear what Mouffe means by external to society. She argues thatantagonism is the experience of the limit of the social, but not a limit asa boundary separating two territories, since this would imply somethingon the other side, so to speak. Rather, [t]he limit of the social must begiven within the social itself as something subverting it (Laclau andMouffe, Hegemony and Socialist Strategy, p. 127). Mouffe writes that theimpossibility of closure is the impossibility of society (ibid., p. 122). But atthe same time she argues that [i]f society is not totally possible, neither isit totally impossible (ibid., p. 129). It is not easy to understand Mouffesdistinction between external and internally external and between impossi-ble and totally impossible.

    19 Mouffe, The Democratic Paradox, p. 19.20 ibid., p. 88.21 ibid., p. 33.22 ibid., p. 22.23 ibid., p. 101.24 ibid., p. 12.25 ibid., pp. 1012.26 ibid., p. 13.27 ibid., p. 102.28 Mouffe, Deliberative democracy, p. 755.29 Mouffe, The Democratic Paradox, pp. 1003.30 ibid.31 Knops, Debate: Agonism as Deliberation, pp. 11526.32 Mouffe, The Democratic Paradox, p. 33.33 Laclau and Mouffe, Hegemony and Socialist Strategy, p. xvii.34 For a clarifying account of Mouffes misuse of Derridas ideas of the consti-

    tutive outside and difference for understanding collective identity, see ArashAbizadeh, Does Collective Identity Presuppose an Other? On the AllegedIncoherence of Global Solidarity, American Political Science Review 99(1) (2005): 4560.

    35 Chantal Mouffe, Hegemony and New Political Subjects: Toward a NewConcept of Democracy, in Marxism and the Interpretation of Culture,ed. C. Nelson and L. Grossberg (Chicago, IL: University of Illinois Press,1988), p. 98.

    36 For the sake of clarity, I do not raise any general objections to the drawingout of normative implications of a described or established ontologicalfeature of human existence. This is being done by political philosophers allthe time. But in order to succeed in such an endeavour, Mouffe would haveto draw on empirical evidence to make plausible that human beings startout by viewing each other as enemies to be destroyed, e.g., similar to howphilosophers use psychological data to make the case that humans strivefor recognition. However, as Mouffe is well aware, if she were to do so,she would soon discover that this is not how human beings predominantlyperceive of each other.

    37 Of course, the concept of difference can be used normatively in a specificcontext. But it is primarily a descriptive concept standing by itself, similarto water and blue. Antagonism, by contrast, is an evaluative concept,similar to happy and courage. It is important to notice that such an

    1059Erman: What is wrong with agonistic pluralism?

    at Ateneo de Manila University on March 8, 2015psc.sagepub.comDownloaded from

  • analytical descriptive/normative distinction is separate from the largerphilosophical question of whether all concepts are normative. The latterposition which I find very plausible concerns a different or at least athinner notion of normativity. It says that semantic content or meaning isnormative in the sense that it must be related to a practical ability (a knowhow), which invokes rules and procedures that could be applied in the right(or the wrong) way. See, for example, Robert Brandom, Making it Explicit(Cambridge: Cambridge University Press, 1994); Saul Kripke, Wittgenstein,Rules and Private Language (Cambridge, MA: Harvard University Press,1982); John McDowell, Wittgenstein on following a rule, in Meaning andReference, ed. A. W. Moore (Oxford: Oxford University Press, 1993); PhilipPettit, The Reality of Rule-Following, Mind 99(393) (1990).

    38 Mouffe, The Democratic Paradox, pp. 334.39 G.W. F. Hegel, Hegels Science of Logic, trans. A. Miller (New York: Human-

    ities Press International, [1831] 1969).40 Knops, Debate: Agonism as Deliberation, p. 125.41 Eva Erman, Conflict and Universal Moral Theory: From Reasonableness

    to Reason-Giving, Political Theory 35 (5) (2007): p. 607.42 Jrgen Habermas, Truth and Justification, trans. B. Fultner (Cambridge:

    Polity Press, 2003).43 Mouffe, The Democratic Paradox, p. 13.44 The relationship between antagonism and the social is thus unclear. On the

    one hand, antagonists share no symbolic space, on the other, when speakingabout subjects, Mouffe refers to subject positions and argues that everysubject position is a discursive position (Laclau and Mouffe, Hegemonyand Socialist Strategy, p. 115). At any rate, to the extent that conflict(antagonism) is understood as a social phenomenon, it is argued here thatit is social in the wrong way.

    45 Indeed, it is a common presumption of Quine and Hegel that institutingconceptual norms and applying them are two aspects of the same process.See, for example, Robert Brandom, Some Pragmatist Themes in HegelsIdealism, European Journal of Philosophy 7(2) (1999): p. 186, note 29.

    46 Wilfrid Sellars, Empiricism and the Philosophy of Mind (Cambridge,MA: Harvard University Press, [1956] 1997); In the Space of Reasons,ed. K. Scharp and R. Brandom (Cambridge, MA: Harvard University Press,2007).

    47 Kenneth Baynes, Practical Reason, the Space of Reasons, and PublicReason, in Pluralism and the Pragmatic Turn: The Transformation ofCritical Theory, ed. J. Bohman and W. Rehg (Cambridge, MA: MIT Press,2001), pp. 5386.

    48 ibid., p. 64.49 See Robert Brandom, Review: Knowledge and the Social Articulation of the

    Space of Reasons, Philosophy & Phenomenological Research 55(4) (1995):895908; Making it Explicit: Reasoning, Representing, and DiscursiveCommitment (Cambridge MA: Harvard University Press, 1994), chapter 1.

    50 Jacques Derrida directs a similar critique towards Schmitt one of Mouffesmajor sources of inspiration. He argues that Schmitt implicitly slides frompossibility to eventuality, and from eventuality to effectivity, cited inAbizadeh, Does Collective Identity Presuppose an Other?, p. 53.

    1060Philosophy & Social Criticism 35 (9)

    at Ateneo de Manila University on March 8, 2015psc.sagepub.comDownloaded from

  • 51 One could reasonably argue that it is not necessary to address this problemonce one has dismissed Mouffes notion of antagonism. However, I wishto show that, by not embracing the thesis that human interaction is consti-tutive of conflict, Mouffe faces some further challenges.

    52 Mouffe, The Democratic Paradox, p. 13.53 One of the main criticisms agonists raise against the liberal tradition is its

    presupposition of an individualistic subjectivist framework. Even thoughMouffe thinks that Habermas deliberative theory does a better job thanmost traditional liberal theories since it attempts to reformulate classicalnotions of democratic theory in communicative terms she claims that heis still caught up in the same individualism in his move from an economicview to a moral one (Mouffe, Deliberative Democracy, pp. 7456).

    54 A similar argument against liberal impartiality theory is discussed in Erman,Conflict and Universal Moral Theory, pp. 6057.

    55 Brandom, Some Pragmatist Themes, pp. 1701.56 ibid., p. 172.57 Susan Hurley, Natural Reasons (Oxford: Oxford University Press, 1989),

    pp. 31617, 2623.58 In fact, as Robert Goodin has shown, the intrapersonal dimension is often

    overlooked by deliberative democratic theorists as well, for whom the inter-personal dimension always takes precedence. According to Goodin, theappropriate deliberative answer to the complexity of modern mass societieswhere face-to-face interaction becomes increasingly problematic, is that wefocus less on external-collective deliberation (interpersonal) and more oninternal-reflective deliberation (intrapersonal). On the view taken here,however, this is a problematic suggestion, since intrapersonal and interper-sonal deliberation have different normative status. While something similarto good reasons could be reached through introspection, they gain theirstatus as good reasons interpersonally by being accepted by others. SeeGoodin, Reflective Democracy (Oxford: Oxford University Press, 2003),pp. 16994 (chapter 9).

    59 Laclau and Mouffe, Hegemony and Socialist Strategy, pp. 125 and 127.60 Baynes, Practical Reason, p. 61.61 ibid., p. 66.62 See Hurley, Natural Reasons.63 Jrgen Habermas, Between Facts and Norms: Contributions to a Discourse

    Theory of Law and Democracy, trans. W. Rehg (Cambridge, MA: MITPress, 1996), pp. 1078.

    64 ibid., p. 109.65 Jrgen Habermas, The Inclusion of the Other (Cambridge, MA: MIT Press,

    1998), p. 42.66 The moral point of view expresses the intuition that valid norms require

    wide agreement. But Mouffe does not take into account that this point ofview is elucidated through the D-principle, not justified by it. Rather, thejustification of the D-principle weighs heavily upon a theory of argumen-tation (Habermas, Between Facts and Norms, pp. 1089).

    67 ibid., p. 110. Unless, of course, the question concerns constitutional matters,that is, the establishing of the equivalence to Mouffes ethico-political prin-ciples, in which case moral norms are privileged.

    1061Erman: What is wrong with agonistic pluralism?

    at Ateneo de Manila University on March 8, 2015psc.sagepub.comDownloaded from

  • 68 Mouffe, Deliberative Democracy, p. 756.69 Jrgen Habermas, On the Pragmatics of Communication, ed. M. Cooke

    (Cambridge: Polity Press, 1998), pp. 198200.70 ibid., pp. 3269.71 ibid., p. 330.72 ibid., p. 332.73 ibid., p. 339.74 ibid., p. 318.75 ibid., p. 319.76 ibid., p. 323.77 Also Habermas acknowledges this fact. Even when democratic policies and

    laws satisfy the principle of legitimacy, they reflect some ethical beliefswhile excluding others and are thus ethically biased. He refers to this asethical patterning and sees it as unavoidable. Indeed, one of the centraltasks of a democracy is precisely to constantly challenge these biases. SeeJrgen Habermas, Struggles for Recognition in Constitutional States,European Journal of Philosophy 1 (2) (1993): 12855.

    78 I cannot go into this here but another challenge facing Habermas, whichmight turn out to be of relevance for an analysis of conflict, concerns theideal speech situation. As we have seen, Mouffe (along with numerousothers) interprets the role of the ideal speech situation as an ideal that issupposed to be approximately realized. But this is a misunderstanding. Theideal speech situation is a methodological fiction, a thought experiment.Through a hypothetical speech situation Habermas aim is to reconstructthe necessary conditions for enabling communicative action. It is supposedto elucidate what makes communicative sociation possible. However, it isprecisely as a regulative idea it has been challenged by Honig and Derrida,who instead defend the normative force of a permanent possibility of crit-ique. Still, it is doubtful whether the ideal speech situation is a regulativeidea in the Kantian sense, as they suggest, since it is both regulative andconstitutive on Habermas account and as such does not fit the classicalopposition between the two. I thank Maeve Cooke for pointing this out tome. See Bonnie Honig, Dead Rights, Live Futures: A Reply to HabermassConstitutional Democracy, Political Theory 29 (6) (2001): 792805;Jacques Derrida, Force of Law: The Mystical Foundation of Authority,Cardozo Law Review 11 (1989/1990): 9191045.

    79 Jrgen Habermas, Reply to Symposium Participants, Cardozo Law Review17 (1996): p. 1490.

    80 Jrgen Habermas, Justification and Application: Remarks on DiscourseEthics, trans. C. Cronin (Cambridge, MA: MIT Press, 1993), p. 12.

    81 Habermas, On the Pragmatics of Communication, p. 326.82 Cooke, Are Ethical Conflicts Irreconcilable?, p. 11.83 Of course, in contrast to moral claims, ethical claims are universal in a

    singular sense, claiming to be valid for specific individuals in particular situ-ations, rather than for everyone. See Maeve Cooke, Realizing the Post-Conventional Self, Philosophy & Social Criticism 20 (12) (1994): 87101.

    84 Cooke, Are Ethical Conflicts Irreconcilable?, p. 4.85 ibid., pp. 1112.86 Brandom, Some Pragmatist Themes, p. 185, note 22.

    1062Philosophy & Social Criticism 35 (9)

    at Ateneo de Manila University on March 8, 2015psc.sagepub.comDownloaded from