Arguments have become an appealing research subject for students of global order and normative change (Bjola and Kornprobst Reference Bjola and Kornprobst2011), both because of their causal power and normative content. On the empirical side, the dynamism of argument allows constructivists to transcend the overly static logic of appropriateness and examine how new normative agreements are collectively constructed (Risse Reference Risse2000). On the normative side, the power of reasons promises to lead the way to the conditions of legitimate change in world politics (Payne Reference Payne2001). Persuasion through the ‘better argument’ has been identified as a source of global democratization (Risse and Sikkink Reference Risse and Sikkink1999), decolonization (Crawford Reference Crawford2002), and legalization (Deitelhoff Reference Deitelhoff2006, Reference Deitelhoff2009). Particularly in a global polity where institutions lack coercive power and collective norms are based on persuasion rather than enforcement (Keohane Reference Keohane2001), it seems natural to ask when arguments are effective (Ulbert and Risse Reference Ulbert and Risse2005; Crawford Reference Crawford2011).Footnote 1
This concern is at the heart of recent research on the ‘logic of arguing’ in international politics. On the basis of Jürgen Habermas’ claim of the ‘constraint-free force of the better argument’, the logic of arguing promises to provide an explanatory mechanism for idea-driven change of international norms that usefully complements the otherwise static repertoire of agency logics in International Relations (IR; Risse Reference Risse2000; Müller Reference Müller2004; Deitelhoff Reference Deitelhoff2009). Yet, as I will argue in this article, this promise is based on a treacherous symbiosis of the normative and the causal force of arguments. Studying good reasons as empirical causes undermines both the normative–critical and the explanatory goal of persuasion research. Genuine persuasion is a tenuous, if not unknowable, object of inquiry that can only be ‘observed’ if scholars already know the content of the better argument. Hence, scholars need to reify ‘better arguments’ instead of endogenizing them. This dilemma points to the intricate methodological difficulty of how to causally explain such a normatively laden subject as moral discourse,Footnote 2 a challenge that is regularly acknowledged, but rarely addressed, by IR constructivists.
The Habermasian approach to persuasion, which has its origins in the German IR debate (cf. Risse Reference Risse2000), provides the most explicit justification for studying argumentative change via persuasion.Footnote 3 Thus, it also lays bare the methodological traps inherent in this undertaking and points to the analytical dilemmas faced by students of moral persuasion. Drawing on Jürgen Habermas’ agency-theoretical interpretation of moral discourse, the ‘New German Idealism’ (Steffek Reference Steffek2010) attempts at bridging, rather than ignoring, the normative–empirical gap. Habermas himself started from the concern that argumentation research can hardly be a morally neutral enterprise and attempted to reconcile normative and empirical theory with the help of his concept of a rationally binding argumentative reason–communicative action.
However, as it becomes obvious in Habermasian persuasion research in IR, the Habermasian conclusions from this concern have self-defeating implications for empirical norms research. The affirmative notion of reason and the better argument on which ‘Theory of Communicative Action’ (TCA) is based have led to a reification of moral standpoints as the better argument rather than their problematization. Where persuasion is viewed as an approximation of truth, the supposedly objective (but empty) criterion of universality can be used to ‘observe’ rational discourse. The paradoxical consequence of an empirical–analytical application of Habermasian discourse theory in IR is that discourse is no longer conceptualized as an open-ended process of contestation and normative change, but turns into a compliance mechanism ensuring the socialization of international actors into pre-defined norms.
Hence, the article provides an internal critique of the Habermasian approach to international discourse and exposes its self-defeating outcomes. My argument draws on debates in philosophy and social theory in order to show that Habermas’ strong notion of rationality is an ambivalent solution to the normative–empirical dilemma. I will argue that communicative rationality is a treacherous bridge between an empirical–reconstructive and a normative discourse theory, a bridge that is based on Habermas’ suspicion of ‘metaphysical’ claims rather than an empirical agency theory. This ambivalence not only concerns the much-debated status of Habermasian discourse ethics in moral philosophy, but has profound implications for IR arguing research; it provides the methodologically unstable foundation for turning the ‘power of the better argument’ into a causal mechanism and thereby jeopardizes both the analytical and critical leverage of the ‘logic of arguing’. However, renouncing the Habermasian solution comes at the price of considerably cutting back the promise of the arguing concept. Either scholars give up positivist claims at empirical explanation and take a critical stance instead or they maintain a causal notion of persuasion and acknowledge that this mechanism may be identified where norms are fixed, but hardly where norms are in flux and contested.
The paper is divided into three sections. The first section unpacks the agency-theoretical interpretation of argumentation as authentic persuasion in IR. I will argue that ‘true belief change’ is not necessary for causally effective argumentation, but rather reflects a moral necessity in a truth-based argumentation theory such as TCA. The second section shows that Habermasian IR can only demonstrate the ‘force of the better argument’ by reifying particular standards as universal. The need to fix the meaning of the better argument has led to a recent turn of several arguing researchers toward compliance instead of normative change. The third section traces how scholars can respond to this normative–empirical dilemma: they may engage in open reification and fix the content of valid norms – the socialization move – or adopt a normative–critical stance based on the commitment to a discursive utopia – the critical move. Since both compromises remain analytically dependent on the assumption of valid norms and true persuasion, the conclusion points to the pragmatist sociology of norms as a means to accommodate the normative pluralism and the myriad of non-rational justificatory practices pervading international politics.
Whence the primacy of authenticity?
In the 1990s, IR constructivists started to call for a turn to language in order to fully grasp the sources of international conflict and cooperation. In this context, Habermas’ TCA was brought into play as a way to theorize the normative structure of the international system. As constructivists assume that the precondition of communicative action, a ‘common lifeworld’ of shared meanings and norms, exists not only at the domestic but increasingly also at the international level, Habermasian discourse theory should also apply to international politics (Risse Reference Risse2000, 14–16). World politics could thus be understood and explained as consisting of communicative practices of agents who draw on overlapping normative resources and meanings in order to reach mutual understanding and cooperation (Müller Reference Müller1994).
Habermas’ approach is part of the tradition of critical theory and premised on the assumption that empirical and normative theorizing are closely intertwined, if not mutually constitutive (Benhabib Reference Benhabib1986, 1–15). Accordingly, his communication theory combines a normative discourse ethics with a sociological (empirical–analytical) discourse theory on which an explanatory research program may be based (Müller Reference Müller1994, Reference Müller2004; Risse Reference Risse2000, 2, 7; Mitzen Reference Mitzen2005; Deitelhoff Reference Deitelhoff2009). In particular, German Habermasians have strived to explain the effects of arguments in agency-theoretical terms, as grounded in a distinct logic of social action: the logic of ‘arguing’ or ‘truth-seeking’. Truth-seeking actors are prepared to give reasons for their empirical and moral claims and they are prepared to be persuaded by the ‘better argument’ even if it counteracts their egoistic interests. Thus, the logic of arguing is not only a moral ideal, but the inevitable causal micro-foundation on which effective argument is based (Risse Reference Risse2000; Müller Reference Müller2004; Deitelhoff Reference Deitelhoff2006, Reference Deitelhoff2009). This strand of norms research thus shares the persuasion view of legitimacy prevalent in IR constructivism (see Hurd Reference Hurd1999) and assumes that effective argument can only work through genuine persuasion and motivational change. The following section challenges this causal claim (see section ‘Ontological primacy?’) and unpacks the moral presuppositions that ultimately back up the ‘arguing’ mechanism (see section ‘Arguing as a metaphysical principle’).
Ontological primacy?
From a Habermasian viewpoint, argument is closely tied to motivational change. Arguments have to be intrinsically persuasive, and this requires that they are put forward and received by truth-seeking agents following the ‘arguing’ mode of social action. Truth-seeking speakers differ from instrumentally oriented actors in that they are open to be convinced by the better argument instead of merely pursuing selfish interests with rhetorical means (Habermas Reference Habermas1984, 285–87; Risse Reference Risse2000). Their authentic orientation toward consensus and intersubjective truth is deemed a necessary condition for arguments to be effective at all.Footnote 4
Consequently, purely instrumental modes of argumentation such as the ‘rhetorical action’ model (Schimmelfennig Reference Schimmelfennig2001; Goddard Reference Goddard2006; Krebs and Jackson Reference Krebs and Jackson2007) are viewed as ontologically dependent on persuasion (Habermas Reference Habermas1984, 286–95). If the social force of arguments rests on subjective beliefs about their rightness, manipulative rhetorical maneuvers can only work by feeding on some participants’ honest persuasions: ‘Obviously, the rhetorical action model only works when there is at least one other actor listening who is prepared to change her understanding of the situation in light of the arguments’ (Grobe Reference Grobe2010, 11; drawing on Risse Reference Risse1999, 533). Otherwise, why should anyone pay lip service to a social norm that no one sincerely believes in – would the collectively manipulated norms not quickly lose their ‘legitimating force’ (Deitelhoff Reference Deitelhoff2006, 106, translation T.H.)? Legitimacy thus has to be anchored inside actors’ minds (see also Hurd Reference Hurd2008, 30f).
This viewpoint centrally builds on attitudinal change as the basis of normative change. By delineating communicative from strategic modes of action, TCA highlights the importance of individual action orientations for the way argumentation works (Habermas Reference Habermas1984, 273–337). Evidently, this agency-theoretical approach entails the methodological difficulty of how to access and study the ‘true’ motives of political actors (Holzinger Reference Holzinger2001, 259). A German research group headed by Harald Müller and Thomas Risse attempted to identify instances of true persuasion in a variety of multilateral negotiations covering issue-areas such as security, humanitarian and environmental questions, human rights, and economics. Yet their empirical ambitions could not be fulfiled, as the researchers themselves have recognized (Deitelhoff and Müller Reference Deitelhoff and Müller2005, 150; Ulbert and Risse Reference Ulbert and Risse2005). It proved impossible to distinguish empirically between strategic and truth-seeking actor orientations (Deitelhoff Reference Deitelhoff2006, 150f; Risse Reference Risse2007, 73). Similar difficulties have been encountered by non-German students of argumentative change, who cautioned that instances of genuine persuasion are hard to observe, given that actual discourses are most of the time ‘distorted’ by more coercive dynamics (Payne Reference Payne2001, 52; Crawford Reference Crawford2002, 123–26, Reference Crawford2011).
Yet also from an ontological perspective the recourse to true motives underlying argumentative change is far from evident. What appears to Habermasians like a methodological constraint – that scholars have no immediate insight into the political actor's brain and true motives – can also be turned into an ontological argument: why should actors base their judgment about the social validity of a norm on speculations about others’ motivations, which are ultimately inaccessible not only to researchers but also to the agents ‘out there’? The irreducible intersubjectivity and performativity of the social world (Guzzini Reference Guzzini2000, 155; Herborth Reference Herborth2007, 166) forestalls such immediacy. Moreover, the distinction between thinking (true beliefs) and acting (arguing) that underlies the communicative action concept makes extremely demanding assumptions about the rationality of social actors. Yet it is a debatable assumption that ‘true’ motives are ontological before their utterance or ‘staging’ (Goffman), especially in situations of normative uncertainty and change that are at the center of argumentation research (see Krebs and Jackson Reference Krebs and Jackson2007).Footnote 5
These methodological and ontological problems notwithstanding,Footnote 6 the foundational role of true persuasion is widely accepted among Habermasians who aim to subsume the rhetorical approach under their more substantial notion of communicative action (Risse Reference Risse2000, 8–9).Footnote 7 On closer inspection, however, the theoretical foundation of this primacy of persuasion turns out to be highly fragile.
Arguing as a metaphysical principle
The logic of arguing (communicative rationality) is Habermas’ analytical bridge between normative and empirical statements about discourse. Thus, he centrally addresses a problem that most IR norm researchers, even Habermasians, prefer to bracket: whether empirical analysis of social norms can ever be value-free (Risse Reference Risse2000, 7, fn. 22; Steffek Reference Steffek2003, 253f). Habermas’ solution to the normativity problem represents a peculiar, and radical, fusion of critical and sociological thinking that is based on a strong notion of rationality and truth. On the sociological side, the theory builds on the conviction that value-free investigation of norms is neither desirable nor feasible. Habermas alerts students of moral discourse and of social action more generally that an impartial understanding of argumentation is impossible. Purely empirical observations that certain arguments are factually accepted fall victim to an ‘empiricist truncation’ (1984, 27) of what argumentation actually (i.e. normatively) means and therefore cannot satisfactorily explain the acceptance of arguments. For Habermas, a satisfactory explanation needs to be an outright evaluation of social norms and is subject to a strong, although contextual, truth standard.Footnote 8 Whoever interprets the ‘rationale’ of an argumentative exchange has to pass judgment on the rationality of the speakers, as ‘reasons can be understood only insofar as they are taken seriously as reasons and evaluated’ (Habermas Reference Habermas1991, 30). A more nuanced and relative position regarding norm interpretation would sacrifice Habermas’ goal of enabling rationally binding judgment about empirical or critical claims (Humrich Reference Humrich2008, chap. 2).
The grounds on which these rationally binding interpretations are made are no ultimate foundations, though. This is the second pillar of Habermas’ solution to the normativity problem. In modernity, moral principles are not metaphysical, but have to be found in worldly communicative practices. This is the goal of the so-called ‘universal pragmatics’, a theory of speech that specifies the conditions under which mutual understanding is both possible and reasonable (Humrich Reference Humrich2008, 41–42). Communicative rationality is the analytical core of Habermas’ impressive attempt at formulating a post-metaphysical moral theory that is not based on ultimate metaphysical foundations, but on an explication of laws inherent to speech (Habermas Reference Habermas1988, 88, Reference Habermas1992, 34). This approach is based on the ‘intuition that the telos of [mutual, unconstrained] understanding is internal to language’ (Habermas Reference Habermas1988, 75, translation T.H.). By means of a ‘rational reconstruction’ of the conditions of argumentation, Habermas explicates the preconditions of successful argument, preconditions we do not usually reflect upon, but that we all enact as competent speakers (Habermas Reference Habermas1984, 2; cf. McCarthy Reference McCarthy1985, 276–79).
In his reconstruction, Habermas builds on, but decisively departs from, speech act theory. In order to communicate successfully, Habermas maintains, actors do not only have to master the linguistic rules explicated by speech act theorists (Searle Reference Searle1969, 12–17; Austin Reference Austin1971, 26). Moreover, speakers need to observe practical, moral rules such as truthfulness and respect (Habermas Reference Habermas1995, 307–11). According to this conceptualization, morally bad, threatening, or discriminating speech acts (Butler Reference Butler1997) can only be interpreted as deviations from ‘normal language use’ (Habermas Reference Habermas1988, 72, translation T.H.). Moreover, such deviations from the logic of arguing are still parasitically dependent on true arguing: ‘[TCA argues that] the use of language with an orientation to reaching understanding is the original mode of language use, upon which indirect understanding, giving something to understand or letting something be understood, and the instrumental use of language in general, are parasitic’ (Habermas Reference Habermas1984, 288). Hence, its cooperative thrust is demonstrated to constitute the deep structure of language, and since we all rely on language for thinking and acting, we cannot rationally disagree with this telos. To the contrary, its motivational force is present in every competent speaker.
What makes TCA attractive for social scientists is that Habermas does not classify it as pure moral philosophy, but as (critical) social theory. Our communicative competence is not (only?) a duty, but a social fact, without which cooperation would be impossible (Müller Reference Müller1994, 29–30). If the telos of language – reaching mutually acceptable consensus – is realized, this reasoned consensus benefits from the ‘binding force’ (Habermas Reference Habermas1988, 69, translation T.H.) of language and enables social coordination. This is the progressive work done by ‘the peculiarly constraint-free force of the better argument’ (1984, 28). In this sense, Habermas’ discourse theory is a theory about the possibility of social order, and yet one that can account for and ‘validate its own critical standards’ (Habermas Reference Habermas1984, xxxix). Empirical and critical analysis go hand in hand (Risse Reference Risse2000, 17).
However, this solution to the normativity problem is debatable on both normative and empirical grounds. From the viewpoint of moral philosophy, the status of empirically reconstructed laws of speech is highly ambiguous: are these empirical laws – which are somehow causally determined – or moral imperatives – which actors could also violate (Böhler Reference Böhler1982, 84)? Habermas claims that even skeptics have to recognize that their lifeworld is marked by communicative reason and that they cannot exit the modern reality of rational discourse (Habermas Reference Habermas1983, 112).Footnote 9 Yet how can such empirical determination be the basis of moral criticism (Schnädelbach Reference Schnädelbach2002, 34)? Would this not amount to a naturalist fallacy (Apel Reference Apel1989, 26–28)?
It is beyond the scope of this paper to disentangle these moral philosophical problems. More important for our purposes is to better understand the nature of Habermas’ ‘empirical’ reconstruction of rational speech. For, apart from principled reservations vis-à-vis a post-metaphysical moral theory (Larmore Reference Larmore2001), Habermas’ ‘empirical’ reconstruction of the essence of speech is rather ambiguous. The rules of rational discourse simply are not as indispensable for meaningful speech as TCA suggests. The philosopher Karl-Otto Apel, whose discourse theory has decisively influenced Habermas’ approach (Habermas Reference Habermas1983, 54), has most pointedly formulated the critique that the primacy of arguing cannot be demonstrated on purely reconstructive, that is, empirical grounds (Apel Reference Apel1989, Reference Apel1999; and see Benhabib Reference Benhabib1986, 330–31).
Apel stresses that the conditions of successful speech explicated by speech act theorists do not include an emphatic principle of understanding as mutual consent. Meaning can also be created successfully without resulting in a consensus.Footnote 10 Only if successful communication is initially defined as mutual persuasion, and thus normatively laden, does successful communication require an arguing attitude. This, however, amounts to a petitio principii (Apel Reference Apel1999, 276–77). Apel therefore maintains that the primacy of arguing can only be justified from a philosophical standpoint. From the standpoint of transcendental (metaphysical!) pragmatics, it can be argued that certain norms of argumentation are logically compelling – that is, no one can deny them without ending up in a performative self-contradiction. Yet, this is a philosophical argument (Apel Reference Apel1989, 52), whose validity cannot be tested against empirical evidence (Johnson Reference Johnson1993, 76). That ‘real-life’ speakers share this philosophical attitude while communicating is a tenuous assumption (Apel Reference Apel1989, 54–59; similarly Rorty Reference Rorty2000, 3, 14), an assumption that introduces a moralist bias in favor of persuasion into agency analysis.
Habermas (Reference Habermas1983, 106) has therefore emphasized that his reconstruction can only provide a weak justification for discourse ethics. This cautionary note notwithstanding, the translation of his theory into an empirical research program in IR has exacerbated rather than attenuated the moralist bias of Habermas’ rational reconstruction, which actually jeopardizes the explanatory ambitions of the arguing approach.
From reasoned consensus to compliance
Translating the discourse ethical standard of communicative rationality into an agency mechanism has turned out to be rather daring (Müller Reference Müller2007, 201). Since pure rational discourse is hardly ever encountered in reality, arguing researchers have abandoned their search for truth-seeking behavior and reoriented their efforts toward supposedly more manifest phenomena, namely the ideational and institutional conditions of effective persuasion. Overlapping normative commitments and a discursive setting that allows for the free and open exchange of ideas have been proposed as favorable conditions for effective persuasion (Ulbert and Risse Reference Ulbert and Risse2005; Deitelhoff Reference Deitelhoff2009).
This strategy raises another analytical difficulty, though: how do you know ‘effective argumentation’ when you see it – especially when the result is not a universal consensus, but a contested normative shift? As this section demonstrates, the resulting attempt to observe the success of the ‘better argument’ counteracts the very purpose of Habermasian discourse theory. Where there is no factual observable consensus, researchers have to rely on the theoretical idealizations with which Habermas defends the possibility of a binding reasoned consensus (see section ‘Spotting ideal consensus within actual dissent’). Hence, Habermas’ universalization imperative turns into a treacherous indicator of the better argument (see section ‘The treacherous certainty of universalism’).
Spotting ideal consensus within actual dissent
In order to determine whether arguing has been effective in international debates and negotiations,Footnote 11 Habermasians need to look at the normative outcome: the resulting agreement and its degree of rationality. The problem is, however, that the content of a ‘reasoned consensus’ is itself fixed with recourse to the mechanism of arguing. Just consider how Risse praises the critical potential of research on the logic of arguing: ‘An alleged reasoned consensus might often be an illusion, and the analyst's task is then to criticize it. Interestingly enough, we can only “know” such a false agreement if measured against a reasoned consensus resulting from truth-seeking argumentation’ (Risse Reference Risse2000, 17).
This ‘measure’ is hardly an empirical measure. For establishing what would be a reasoned consensus in a certain international dispute, observation does not take us far. International negotiations will rarely lead to a harmonic situation where everybody agrees on the same principles and their interpretation. Even if disputes are decided with non-coercive, argumentative means, parties can agree not to agree and settle on a compromise (Steffek Reference Steffek2005) whose interpretation may also greatly diverge across parties (Wiener Reference Wiener2008). Telling a reasoned consensus, or elements of it, from somehow distorted collective outcomes has turned out to be close to impossible on purely empirical grounds (Deitelhoff and Müller Reference Deitelhoff and Müller2005, 177). Accordingly, demanding theoretical operations are required, which help to establish a counterfactual measure of the ‘reasoned consensus’.
As Risse (Reference Risse2000) and Nicole Deitelhoff (Reference Deitelhoff2009) have pointed out, this measure can be found in Habermas’ procedural standard of a universalizing rationality. Universalizing reason serves as the principle of inference on which the binding force of reasoned agreements is to be based and functions as the rational presupposition that speakers need to enact whenever they enter into moral discourse (Habermas Reference Habermas1983, 97). This rationale undergirds the idea of the ‘ideal speech situation’, implying equality, openness, and a truth-seeking attitude by all participants to a discourse (Habermas Reference Habermas1983, 98–102). Habermas insists that speakers engaging in rational discourse necessarily argue with recourse to this counterfactual idealization that implies both abstraction (from messy circumstances) and universalization (across all possible contexts; see Risse Reference Risse2000, 10; Deitelhoff Reference Deitelhoff2009, 45, 48). Otherwise, there would be no rational criterion for valid agreements (Habermas Reference Habermas1983, 72–73).Footnote 12
This concept has often been criticized for being unrealistic, since power-free, authentic discourse will probably never see the day in international politics (Schimmelfennig Reference Schimmelfennig1997). However, Habermas’ argument does not rely on a factual realization of the ideal speech situation. Rather, it has to be judged according to its analytical value as a counterfactual operation. The basic claim is that the presupposition of an ideal speech situation is not a representation of empirical discourse, but still consequential for actual argumentation.Footnote 13 All truth-seeking actors counterfactually presuppose ideal discursive conditions and argue as if they were addressing a universal audience, since the quality of their arguments depends on their acceptability in all imaginable contexts (Habermas Reference Habermas1999, 259; Risse Reference Risse2000, 10; Risse, Jetschke and Schmitz Reference Risse, Jetschke and Schmitz2002, 23; Deitelhoff Reference Deitelhoff2006, 92). Thus, the analytical value of this counterfactual idealization does not depend on its explanatory power, but on its capacity to enable rational decisions about the better argument – for both speakers and researchers. And here lies a central problem of Habermasian constructivism:
The treacherous certainty of universalism
Seemingly, Habermas’ neo-Kantian principle of moral inference offers a solution for IR Habermasians seeking to identify the better argument in a given controversy: researchers may determine persuasive arguments with the help of a universality test. As an empirical–analytical tool, this criterion of universalization shall allow researchers to identify a reasoned consensus or traces thereof. Even in the absence of a fully theorized consensus, we may observe that positions have shifted in the direction of the more universal position, and hence that persuasion has occurred.
Deitelhoff (Reference Deitelhoff2009) pursues this strategy in her analysis of the negotiation process leading up to the 1998 Rome Statute of the International Criminal Court (ICC). She identifies two classes of arguments in the debate over the legal design of the Court: ‘Conservative’ arguments framed in terms of ‘political reality’, stability and prudence were brought forward by the opponents of a strong ICC. In particular, great powers maintained that only a court adhered to by all states would be effective; hence, a realistic compromise would have to respect sovereignty and security concerns (2009, 52). Over time, however, a ‘public interest’ frame gained ground that was not power- and sovereignty-preserving, but based on the principles of law and morality. An independent and impartial court was supported by arguments ‘based on the potentially universal principles of equality and impartiality’ (2009, 53). The negotiation shift toward a rather strong ICC against the interests of major powers such as the US is therefore judged to be a product of moral persuasion.Footnote 14
The problem is that this way of ‘observing’ the impact of the better argument crucially depends on the meaningfulness of the universalization criterion. And here we are confronted with the moral philosophical problem that universality is a purely formal quality of norms and therefore indeterminate (Kratochwil Reference Kratochwil1989, 132–38). It is, for example, of little help for deciding whether sovereignty or human rights shall be accepted as the basic norm of the international system (Wiener Reference Wiener2007). Both candidates for a basic global norm can in principle be universalized. Habermasians seem to be barely irritated by this lack of substance and indeterminacy of the ‘better argument’ (Hawkins Reference Hawkins2004, 784). Rather, human rights (Risse, Jetschke and Schmitz Reference Risse, Jetschke and Schmitz2002) or the universal prosecution of human rights violations by an ICC are regarded as unproblematic examples of valid norms. More ‘conservative’, realist positions emphasizing peace and stability, and political prudence, however, are considered to represent invalid moral arguments (Deitelhoff Reference Deitelhoff2009, 51–52). How such normative decisions can be inferred from universality (the seemingly objective indicator) alone remains to be demonstrated (Hanrieder Reference Hanrieder2008, 175–79).
As Habermas’ critics have pointed out, ‘ideal’ reasoning is not only an empirical fiction but also a philosophical chimera. If communicating means exchanging positions and standpoints, then how would one assume the wholly delocalized standpoint of ideal communication? What would be the incitement for, and what the ideational resources of, discourse (Larmore Reference Larmore1993, 322)? Rorty (Reference Rorty1996, 75) likewise criticizes the thinness of the rationality criterion, as ‘the “agreement of all competent judges operating under ideal epistemic conditions” is something we can never know whether we have attained. Unknowability and unconditionality go hand in hand’. It seems quasi-impossible to productively mobilize this harmonistic ideal in actual, conflictual political argument (Tietz Reference Tietz1993, 339–40), as universal validity ‘is too thin to help us change our minds about anything’ (Rorty Reference Rorty1996, 73).
As unconditional justification is essentially underdetermined, constructivist ‘observers’ of the better argument cannot but to fill in their moral convictions when identifying universally valid claims. ‘Idealization’ then means that speakers and/or observers universalize the validity of their values – a validity that still has to be justified on more substantive grounds (Larmore Reference Larmore2001, 110). We could conclude from this that the inevitability of standpoints encourages an explicit scholarly commitment to substantive norms such as human rights (Risse and Sikkink Reference Risse and Sikkink1999) or the rule of law (Deitelhoff Reference Deitelhoff2009). For Habermasian constructivists, however, committing to substantive arguments would undermine their positivist ambition to explain argumentative effects. Instead of being observers, they would deliberately assume the role of participants in the discourse they are analyzing.
Yet, the solution to base claims about the better argument on the seemingly neutral indicator of universality at best camouflages such moral standpoints. Moreover, a reification of specific standpoints as morally valid is no less self-defeating in that it violates the open-endedness of deliberation (Herborth Reference Herborth2007, 153–54). Habermas himself insists that the universalization principle (U) can impossibly supplement the discourse principle (D) 500, namely that only participants in a discourse are able and entitled to decide what is a convincing argument in a given context (Habermas Reference Habermas1983, 76; Holzinger Reference Holzinger2001, 252; Deitelhoff Reference Deitelhoff2006, 95–96). Hence, a ‘monological’ usage of (U) by seemingly impartial observers cannot be the basis of valid judgments (Habermas Reference Habermas1983, 76).
This dilemma of how to determine the better argument in moral controversies critically circumscribes the scope of arguing theory. Since a meaningful observation of arguing via ‘observing’ the success of the better argument is impossible, IR applications of TCA have increasingly scaled back the critical ambitions of the research program. Instead of explaining normative change, several Habermasians have reoriented their efforts toward explaining the diffusion of already defined norms (Herborth Reference Herborth2007, 153). They study persuasion as a mechanism of norm diffusion ensuring compliance with already established rules (Risse Reference Risse2007, 78). As Risse (Reference Risse2004, 306) claims, ‘one can engage actors – governments and private actors alike – in an arguing process to persuade them of the normative appropriateness of international rules and of the need to accept them as behavioural standards. Voluntary rule compliance would become a matter of persuasion’. With this compliance turn, the collective and critical fabrication of new rules and norms, the theoretical core of Habermas’ discourse theory, is no longer at issue. To the contrary, arguing no longer serves as a standard of democratic politics, but as an effective governance device for supranational institutions: through reason-giving and public justification, international organizations such as the World Trade Organization can compensate for their lack of democratic legitimacy and ensure compliance with their rules and rulings (Steffek Reference Steffek2003; see also Panke Reference Panke2010). This move paradoxically sacrifices the democratic project underlying Habermas’ theory, which critically relies on the open-ended contestation of normative orders.
Studying persuasion between realism and utopia
It is not by accident that several arguing researchers have started to move from authentic persuasion to effective compliance. The reorientation toward compliance research is one way out of a basic aporia inherent to the rationalist (in the philosophical sense of the word) communicative action program. Since genuine persuasion and universal truth are neither intelligible nor observable, scholars cannot but postulate specific norms as ‘persuasive’ in order to state that their acceptance is based on consent rather than coercion. Such persuasion research is good at reifying specific norms, but does not fare well in investigating normative change by persuasion. For where both social norms and actor motivations are problematized, scholars lack the analytical vantage point from which to assess the persuasiveness of moral claims.
Analytical compromises seeking to avoid this aporia usually fall on either the positivist side and abstain from normative claims or on the critical side, abstaining from explanatory claims. The positivist solution consists in treating persuasion as socialization, that is, in a deliberate reification of international norms (see section ‘The socialization move’). The critical approach defends the deliberative ideal and focuses on the conditions and distortions of moral discourse in international politics (see section ‘The critical move’). Both types of compromise demand strong – empirical or idealized – presuppositions and remain committed to the basic persuasion vs. coercion–dichotomy that leaves little space for normative complexity and argumentative practices other than persuasion.
The socialization move
As long as norms are socially contested and in flux, observers can hardly determine from outside whether the most ‘persuasive’ arguments have prevailed in a discourse. Hence, if an explanatory usage of the persuasion concept is to escape the dilemmas explicated in the above section, scholars need to fix instead of internalizing the meaning of valid norms. This corresponds to the move to compliance that Habermasians have made in recent years (see section ‘The treacherous certainty of universalism’). In this vein, empirical approaches to persuasion have been put forward that do not claim to explain normative change by persuasion, but focus on situations where social norms are rather unproblematic. This is the domain of socialization research.
Socialization applies where norms are not based on universal reason, but tied to particular group identities (Johnston Reference Johnston2001). The socialization of new actors into the liberal norms of the European Union (EU; Checkel Reference Checkel2001, Reference Checkel2005; Schimmelfennig Reference Schimmelfennig2001), the ‘international community’ (Risse and Sikkink Reference Risse and Sikkink1999), or other international organizations (Johnston Reference Johnston2001) is a process that works independent of material coercion. Socialization researchers emphasize that the depth of persuasion varies in processes of socialization. It can involve the deep internalization of norms, but this is not necessary and most likely not the most important path to group conformity (Johnston Reference Johnston2001, 495). Rather, ‘social influence’ plays a decisive part in socialization, that is, group pressure working through moral sanctions, shaming, exclusion, and ‘opprobrium’ (Johnston Reference Johnston2001, 488; Zürn and Checkel Reference Zürn and Checkel2005, 1052). The force of this mechanism thus does not rest on rational idealizations, but on concrete relationships – and the social authority of the persuader (Checkel Reference Checkel2001, 560–64; see Hurd Reference Hurd1999).
Testing for socialization is easier than testing for genuine persuasion: given that a rather stable social environment can be identified, pro-norm behavior that occurs after exposure to this social environment and in the absence of material coercion can be attributed to socialization (Johnston Reference Johnston2001, 510). The challenge for and motivation of socialization research thus consists in identifying the integrating force and resilience of particular social norms such as those embedded in the polity of the EU (Checkel Reference Checkel2001, Reference Checkel2005). Such an empirical perspective on norm effectiveness does not problematize but reifies particular norms. ‘Discourse’ in the Habermasian sense where henceforth unproblematic norms are questioned and reconsidered collectively is not part of this research program. Socialization research thus seeks to gain explanatory leverage at the price of an ‘empiricist truncation’ that reifies political communities and identities as objective givens. Yet, the ongoing contestation and ambiguity of European and international norms eludes such a generalizing perspective.
The critical move
An alternative conclusion from the normative–empirical dilemma of arguing research is to preserve the critical edge of arguing and adopt a reflexive stance toward the research ‘object’, international discourses. Such a normative–critical approach rests on the insight that communicative reason may be metaphysically justifiable, but cannot be reconstructed as an empirical law (Apel Reference Apel1989, Reference Apel1999). However, if rational discourse is not an empirical law, it still is a practical possibility. This is the conclusion that, for example, Benhabib (Reference Benhabib1986, 298–309) draws from her friendly critique of Habermas’ universalism. Reference BenhabibBenhabib maintains that the universalization rule is not a necessary part of discourse ethics and counterproductively introduces a ‘legalistic bias’ (1986, 310) into political discourses. Instead of justifying discourse ethics on incontrovertible rational grounds, Benhabib thus highlights its emancipatory potential and claims that ‘a certain anticipatory utopia, a projection of the future as it could be, becomes necessary’. This requires scholarly reflexivity and some theoretical activism, since ‘the theorist can no longer speak the language of evolution and necessity, but must conceive of him or herself as a participant in the formation of the future’ (Benhabib Reference Benhabib1986, 331).Footnote 15
This emancipatory purpose underlies critical scholarship on international discourse.Footnote 16 It uses the arguing ideal as an evaluative tool to assess the deliberative quality of international policies (e.g. Bjola Reference Bjola2005) or as an ideology-critical tool to identify obstacles to fair deliberation (Bohman Reference Bohman1997, 338–39, Reference Bohman2010, 13).Footnote 17 This requires attention to the discursive distortions that jeopardize the quality of argument and thereby genuine persuasion. Mostly, such distortions are sought in hidden motivations and the intrusion of power into discourse (e.g. Payne Reference Payne2001).
Distortion can also be inscribed in discursive settings with pre-defined normative outcomes, that is, where free and open argument is precluded. The negotiation process leading up to the treaty banning anti-personnel landmines is a case in point. The negotiation of the 1998 landmine treaty often serves as the prototype of a progressive ‘new multilateralism’, where moral persuasion wins out over power politics (Price Reference Price1998; see also Deitelhoff Reference Deitelhoff2009, 35). Yet, in fact, its institutional setting has been highly restrictive. Only states that committed beforehand to the goal of a total ban were allowed to participate in the ‘Ottawa process’, the venue through which the blockades of the United Nations (UN) system were circumvented (Dolan and Hunt Reference Dolan and Hunt1998, 37; Brem and Rutherford Reference Brem and Rutherford2001, 169–73). In this sense, the ‘new multilateralism’ exposes the gap between deliberative ideals and real-world consensus-making.
These ‘distortions’ of norm generation processes are of primary interest to the discourse ethical program that seeks to unveil (and remove) the obstacles to rational discourse. The flipside of this critical orientation is that everything that does not meet the requirements of rational discourse is no ‘real’ communication about norms, but at best strategically distorted. On the basis of the common dichotomy between coercion and ‘true’ persuasion (see also Crawford Reference Crawford2011), this perspective discounts all argumentative means besides ‘rationally convincing’ arguments as obstacles to reaching the discursive utopia. Hence, such a moralized approach to real-world argumentation is itself analytically distorted. It excludes from the analysis all those less ‘rational’ practices through which empirical actors settle normative conflicts in their non-utopian settings. This is the price of a normative commitment that may be useful in avoiding the normative–empirical fallacy, and yet only allows for highly circumscribed sociological insights into argument and normative change.
In sum, the analytical compromises presented in this section demand that scholars subscribe to specific international norms or evaluative standards in analyzing moral discourse. In a way, both are rationalizing and totalizing strategies. Positivist ‘socialization’ approaches reify certain norms as valid and accepted, while critical approaches apply evaluative standards to contexts where these standards may not be unanimously shared. Yet international politics can hardly be said to be marked by normative coherence and stability (Krook and True Reference Krook and True2010). Norms can be differently enacted (Wiener Reference Wiener2008), and their enactment in turn changes their very content (Sandholtz Reference Sandholtz2008). Most often, several competing standards can be brought to bear on world political issues. For example, non-profit organizations have to negotiate the call for political activism with the principle of neutrality (Barnett Reference Barnett2009), but also with the imperative of competitiveness and economic performance (Cooley and Ron Reference Cooley and Ron2002). International civil servants have to accommodate calls for economic efficiency and results-based management with the procedural requirements of supranational bureaucracy (Bauer and Knill Reference Bauer and Knill2007). ‘State-builders’ have to mediate between the bureaucratic requirements and practices of their ‘metropolitan’ headquarters and the normative orders of their target countries (Schlichte and Veit Reference Schlichte and Veit2007). These normative conflicts are the very reason why IR should be concerned with argument in world politics.
Yet such normative ambiguity and conflict eludes approaches that postulate the validity of specific norms and/or discursive standards. Both the socialization approach and the critical approach avoid the empirical-normative fallacy looming large in persuasion research. Yet they remain committed to the ‘persuasion vs. coercion’ paradigm and thereby reproduce and reify specific norms instead of exploring the argumentative repertoire brought to bear on world politics. What could be gained from a pragmatic abstention from such preconceptions will be sketched in the conclusion of this article.
Conclusion: for a pragmatic abstention from the better argument
This article has critically reconstructed the self-defeating implications of a ‘truth-seeking’ approach to persuasion in international politics. Drawing on Habermas bears the advantage that his communicative action theory explicates and thoroughly theorizes basic intuitions that underlie the concept of persuasion. However, the application of his theory in the IR arguing approach also brings scholars to the heart of the normative–empirical dilemma so acute in norms research. The article has shown that despite its moral and theoretical appeal, the ‘force of the better argument’ is a poor guide to argumentation research. As a sociological approach, the logic of arguing relies on a notion of authentic conviction that is both methodologically and ontologically tenuous. Moreover, its moral philosophical underpinnings are founded on an analytically empty presupposition: the notion of unconditional, universal truth. In its application to IR, the ‘better argument’ serves to reify particular norms as universal, instead of opening up space for deliberation.
Several Habermasian constructivists have responded to this dilemma by turning arguing into a compliance mechanism, and thus exchanged the emancipatory agenda of critical theory for a governance- and control-oriented research program. This compliance turn is one possible solution to the normative–empirical dilemma encountered by students of moral discourse. It is closely related to the agenda of international socialization research that treats particular norms as fixed and investigates the non-coercive means through which agents are brought to accept these norms. Through positivist reification, normative effects become researchable, yet at the price of bracketing normative conflict and contestation. The critical reaction to the arguing dilemma in turn renounces positivist reification and engages in political critique. Its emancipatory challenge to actual (distorted) discourses drives forward the discourse ethical agenda, but also requires a distorted view of argumentation: everything outside of true persuasion becomes mere coercion and manipulation.
The strategies of positivist reification and critical moralization are thus premised on the same ‘persuasion vs. coercion’ problematic that underlies research on the better argument. This dichotomy is based on the broadly shared assumption that argumentation follows a rational logic, a logic that scholars can reconstruct empirically and/or evaluate in normative terms. Yet, as this article has argued, such strong presumptions only allow for a highly rationalized or moralized view of argument. They forestall insight into the normative complexity of world politics and the practical creativity through which empirical actors settle normative conflicts. Moral agreements and orientations are constantly forged in international politics; yet approaching them through the normative lens of valid reasons alone a priori excludes a myriad of argumentative practices.
In order to transcend this dichotomy, scholars would have to pragmatically abstain from the a priori rationalization of argumentative practices. Taking seriously the means through which social actors come to moral agreements in the midst of ambiguity and contestation, a pragmatic abstention from normative claims could provide a new perspective on moral agreement as a genuinely practical achievement. Such a pragmatist stance would allow IR norms researchers to view effective argument not as a reasoned consensus, but first and foremost as a successful performance.
In this vein, recent pragmatist contributions to the sociology of norms show how a practice-centered approach to norms and morality transcends the legalism of communicative action theory. It is based on a pragmatist conception of social actors as acting rather than theoretical beings (Joas Reference Joas1996), a conception that is also gaining ground in IR theory (Hellmann Reference Hellmann2009; Friedrichs and Kratochwil Reference Friedrichs and Kratochwil2009). On the basis of this non-rationalist ontology, Hans Joas has argued that communication about values has an important emotive component, and often relies on successful narratives much more than on compelling reasons. Drawing particularly on American pragmatism, Joas therefore calls for a reorientation of social theory toward the role of collective emotions and cultural traditions in the negotiation of normative orders (Joas Reference Joas2000, Reference Joas2001).
Similarly, a performance-centered approach to moral discourse – in their terms: justification – has recently been developed by the so-called French school of pragmatism around Boltanski and Thévenot (Reference Boltanski and Thévenot1999, Reference Boltanski and Thévenot2006). This approach starts from the observation of an irreducible – and rationally irresolvable – plurality of moral standards. Accordingly, morally complex situations are not resolved through rational generalization and truth-seeking, but through the performance of critical ‘tests’ (Boltanski and Thévenot (Reference Boltanski and Thévenot2006, 40). These tests rely on the mobilization of manifold cognitive and material devices (‘things’) other than abstract reasons. This performativity perspective would allow IR norms researchers to integrate the status of various objects – blue helmets, the UN flag, and symbols (see Hurd Reference Hurd2002) – in the analysis of normative struggles. Likewise, cognitive devices such as evaluation techniques and statistical representations (Barnett and Finnemore Reference Barnett and Finnemore2004, 45–72) and representative tools such as amnesty international's practice of eyewitness accounts (Wilson Reference Wilson1997) could be analyzed as part of the critical repertoire of international actors.
A pragmatic abstention from the rationalizing view of discourse would thus accord worldly and non-rational dimensions of argumentative practice a more prominent analytical status. Pragmatist norms research would not so much be concerned with normative superiority (Deitelhoff Reference Deitelhoff2009, 35) but with successful practice. Even the invocation of power and authority might be accommodated within the pragmatist framework; not as an unreal distortion of discourse, but as an argumentative operation that can succeed or fail. Such a pragmatist move would also resonate with recent attempts at integrating the ontology of practices (Adler and Pouliot Reference Adler and Pouliot2011) and practical reason (Kornprobst Reference Kornprobst2011) into IR theory. Normatively speaking, it would take its critical standards not from compelling philosophical arguments, but from competent actors and their critical operations.
Acknowledgments
This article is based on a paper written for the workshop ‘The Politics of Talk in International Relations’ (Bremen, July 2010) organized by Jens Steffek in cooperation with Ron Krebs and Jennifer Mitzen. I thank the organizers and the workshop participants, in particular Nick Onuf, my colleagues at Munich University Judith Renner, Alex Spencer, Johannes Jüde, Christian Kreuder-Sonnen and Rainer Hülsse, Sebastian Büttner, and the reviewers and editors of International Theory for very helpful comments on various drafts of this article.
