Aalberts and Van Munster. From Wendt to Kuhn

From Wendt to Kuhn: Reviving the ‘Third Debate’

in International Relations

Tanja E. Aalbertsa and Rens van MunsterbaDepartment of Political Science, Leiden University, Postbus 9555, Leiden 2300 RB,

The Netherlands.

E-mail: [email protected] of Political Science, University of Southern Denmark, Campusvej 55, Odense C 5000,

Denmark.

E-mail: [email protected]

Constructivism is often identified as the legitimate occupant of the middle groundbetween rationalism and reflectivism that emerged from the Third Debate ininternational relations (IR) theory. Indeed, the rationalist–constructivist debate isalready being framed as the next dominant debate with the IR community. Thispaper evaluates the bridge-building project as initiated by Alexander Wendt, andtakes issue with the via media as proposed by the so-called conventionalconstructivists. It is claimed that the rationalist–constructivist debate has beenlimited to a discussion of ontology, which has brought about a contradictionbetween ontology and epistemology. Returning to the pressing epistemologicalissues that were put on the table by reflectivist scholars, this article refocuses thecurrent debate by taking up the Kuhnian link between substance and science. Itelaborates a view of science as a communal practice built on intersubjectiveconventions and argumentative procedures. This leads to an alternative conceptionof the middle ground as a communicative space.International Politics (2008) 45, 720–746. doi:10.1057/ip.2008.26;published online 1 August 2008

Keywords: constructivism; epistemology; Kuhn; Third Debate; rationalism;reflectivism

Introduction

In international relations (IR) academia, constructivism has, amongst others,been claimed to be ‘one of the most important theoretical developments of thelast decades,’ an ‘inescapable phenomenon,’ the ‘major points of contestationfor IR scholarship’ and ‘the officially accredited contender to the establishedcore of the discipline’ (Katzenstein et al., 1998, 646; Guzzini, 2000, 147; Smith,2001, 226; Zehfuss, 2002, 2; for similar claims, see Walt, 1998; Adler, 1997,2002; Fearon and Wendt, 2002; Ba and Hoffman, 2003).1 No doubtconstructivism owes much of its current appeal to its promise to provide a‘middle ground,’ ‘via media,’ or to ‘bridge the gap’ between rationalist and

International Politics, 2008, 45, (720–746)r 2008 Palgrave Macmillan Ltd 1384-5748/08

www.palgrave-journals.com/ip

reflectivist approaches to world politics (Adler, 1997; Wendt, 1999, 40, 47;2000; Smith, 2001, 242). Yet, constructivism in its main variant has not lived upto its original promise to bridge these two sides. Instead, in the course ofpositioning constructivism within IR theory, the bridge-building project hasmoved its focus to the debate between rationalist and constructivists. Manynow frame the rationalist–constructivist debate as the dominant debate withinthe IR community (see e.g. Katzenstein et al., 1998; Price and Reus-Smit, 1998;Fierke and Jørgensen, 2001; Reus-Smit, 2001), and even refer to it as theFourth Debate (Fearon and Wendt, 2002, 129). However, while Fearon andWendt take these approaches to be at the center of the discipline, they present asceptical view about the idea of rationalism vs constructivism, and claim thatmore than a debate there is (the beginning of) a synthesis at hand.2

In this paper it is argued that the shift of the middle ground in the directionof the rationalist side has had serious implications for the content of thedebate. More specifically, it is claimed that the rationalist–constructivist debatehas been limited to a discussion of ontology that, we argue, has moved too fastbeyond the pressing epistemological issues that were put on the table by thereflectivist scholars of the Third Debate. Mainly concentrating on therationalist side of the theoretical spectrum, constructivism gives the wrongimpression that the yawning gap between rationalism and reflectivism can bereduced to a cleft between rationalism and constructivism that hardly needs abridge in order to be crossed. This can be illustrated as shown in Figure 1.

We want it to be clear from the outset that we do not claim that allconstructivists are part of the current middle ground. We are well aware of thelines of contestation that divide constructivists themselves.3 Our point is ratherthat the most dominant position currently identified as the middle ground isconstructivist scholarship that lies closest to rationalism — hence the popularlabel of conventional constructivism (Hopf, 1998, see also Katzenstein et al.,1998, Price and Reus-Smit, 1998). Indeed, our aim is to point out thatconstructivist epistemological insights that are closer to reflectivism aresidestepped in the current rationalist–constructivist debate, and that this isnot without consequences.

Returning to the issues of the Third Debate, this paper seeks to refocus thecurrent debate, firstly, by showing the inconsistencies of conventionalconstructivism as the alleged bridge between these poles and, secondly, by

ConventionalConstructivism

Original Middle Ground

RATIONALISM CONSTRUCTIVISM REFLECTIVISM

Figure 1 The Third Debate

Tanja E. Aalberts and Rens van MunsterFrom Wendt to Kuhn

721

International Politics 2008 45

proposing an alternative conception of the middle ground as an communicativespace rather than a bridge. Whereas the latter implies a third way or synthesisbetween two extremes, the picture of the middle ground as a communicativespace has a different aim and favors open dialogue and reflectivity (see alsoLapid, 2003). Our argument proceeds in three stages. The next section willbriefly discuss the Third Debate in order to recall the main issues at stake inthat debate. Pointing out that some of the better-known constructivists saw inconstructivism a potential to resolve the issues of the Third Debate, the thirdsection sets out to evaluate their claim to provide a viable middle groundbetween rationalism and reflectivism. Recontemplating Kratochwil andRuggie’s (1986) warning about contradictory ontologies and epistemologieswithin IR theory, the fourth section will return to some of the metatheoreticalissues of the Third Debate theorists by arguing that we need to refocusour attention to epistemological issues. While being sympathetic towardsreflectivist objections against epistemology, we argue that ‘substance’ and‘science’ are interrelated (Kuhn, 1970) and in addition elaborate scienceas a scholarly practice which stresses the importance of communicativecompetence.4 In this account we take issue with both poles of the originaldebate. The middle ground the way we conceive it rejects the either/or choiceof positivism vs absolute relativism, and acknowledges the role of persuasionand argumentation as important factors in deciding on competing know-ledge claims.

The Science and Substance of the Third Debate

In the historiography of the discipline of IR theory there is some controversyabout the so-called Third Debate. Apart from the question whether it is correctto conceive of the discipline as evolving through a succession of Great Debatesin the first place, the Third Debate causes a problem in particular as it isunclear what the opposite sides of this debate exactly are (Maghroori andRamberg, 1982; Wæver, 1996; Schmidt, 2002; Wight, 2002). To avoidambiguity about which Third Debate we are talking, we will frame ourdiscussion in the rationalism vs reflectivism dichotomy (Lapid, 1989). The storyof its origin is renowned and dates back to Robert Keohane’s presidentialspeech to the International Studies Association, delivered at the 29th AnnualConvention in 1988. He juxtaposed what he labelled rationalism (entailingboth neorealism and neoliberalism) to reflectivism and claimed the need toexplore possibilities for a synthesis in research programs. This set the stage forthe notorious Third Debate in IR theory as discussed by Lapid (1989; see alsoAshley and Walker, 1990; George and Campbell, 1990). Whether or not thereever was such a debate, that is, whether the two sides indeed ever got talking, is


722


a question we will not address here in detail. Three points are crucial for ourdiscussion. First, that in the 1980s critical reflections on mainstream IR theoryemerged, which introduced important and pressing metatheoretical issues intoIR theory (second-order theorising). Second, that this reflexivity was of suchdisposition that it entailed something akin to Kuhn’s notion of paradigms —that is, it includes both substance and science (see also Walker, 1989). As suchit was more fundamental than any of the previous Great Debates as, inprinciple, it touched upon the identity of the academic field of IR (Puchala,2000). And finally, that the bare middle ground in between the opponents ofthe debate was cultivated by constructivism in the bridge-building projectinitiated by Alexander Wendt (1992, 1999, see also Wendt, 1987).

Lapid (1989) states that the Third Debate emerged in the late 1980s, parallelto the trend to move away from empiricist–positivist orthodoxy in the socialsciences in general. Whereas IR theory had been charged for remaining anintellectual backwater of the main approaches of Western social theory,5 thediscipline got involved in a discussion about positivism vs post-positivism,which led a substantial part of the IR community to a critical re-examinationof the ontological and epistemological foundations of their scientificendeavors. Whereas the picket lines of this debate might not be as apparentas in the earlier Great Debates (Puchala, 2000) as it concerns a basket of (meta-theoretical) issues, the main point of contention is a fundamental distinctionbetween natural and social sciences. For the insight had gained momentumthat the very intersubjective nature of the social realm renders it problematic toadopt methodologies of natural science to explain the social world. Thus theprinciple of unity of science (or naturalism) was called into question byreflectivist scholars. This rejection of naturalism in turn bears upon theadherence to (other aspects of) positivism in IR theory, as this metatheoreticalstance stems directly from natural science. Following from the unified view ofscience, positivism can be characterized by three additional assumptions: (i) abelief in nature-like regularities in the social realm, which can be studied alongdeductive-nomological and inductive statistical models;6 (ii) a distinctionbetween facts and values and belief in theory-neutrality of facts; (iii) empiricalvalidation or falsification as certification of real enquiry (empiricism) (Smith,1996, 16). These latter two points concern the existence of an objective truth,which is accessible and can serve as the ultimate Archimedean point for theappraisal of research. Ultimately this relates to the demarcation of science andnon-science, which illustrates the high stakes involved with the debate. Indeed,the wager is ‘trailblazing ideas about the nature and progression of knowledge’(Lapid, 1989, 238).

Contrary to the celebrated objectivism as one of the hallmarks of scientificresearch, reflectivism claims that it is problematic, if not utterly impossible, toapply absolute objectivism as advocated by positivism. Facts do not speak for


723


themselves, but are always theory-laden, which makes the discovery of a newfact ‘necessarily a complex event, one which involves recognizing both thatsomething is and what something is’ (Kuhn, 1970, 55, italics in original). This isnicely captured by Kratochwil:

y different from empiricism, which assumes that ‘things’ show themselvesin an unadulterated fashion because experiments are conceptualized aspointed questions posed to nature to which the latter has to respond, wehave to realize that ‘nature’ cannot answer because it needs a language tocommunicate. ‘Truth’ can, then, no longer be a property of the ‘world outthere’ but has to be one of ‘statements about the world.’ Such knowledgemeans that we cannot test our ideas against reality as all our questions tonature are already phrased in a theory (or language); we test only theoriesagainst other theories. (Kratochwil, 2003, 124)

In response to Kratochwil’s depiction of empiricism as a somewhat naıveposition, it could be objected that empiricism is better understood as a scepticalepistemology that starts from observables rather than theoretical assumptionsabout reality that may appear dogmatic because they escape direct experience(the existence of God, for example). Yet, while empiricists accept that facts aretheory laden when it comes to theory-building (which is why they generallyprefer induction over deduction), they mistakenly assume that when it comesto theory-testing facts actually do speak for themselves (cf. Guzzini, 2000, 157).

A second objection to Kratochwil’s statement has been formulated bycritical realists, who lament the purportedly anti-realist ontology ofreflectivism. In their view, to argue that reality is only accessible throughlinguistic categories (theories) equals the idealist claim that there is no worldexternal to thought (Wight, 1999, Patomaki and Wight, 2000, 217). Such acritique, however, is misleading and wrongly reduces reflectivism to idealism.7

Yet, reflectivism does not deny that a material world exists outside our heads;it merely adds that this world cannot be known without our intersubjectiveframeworks and interference. As objects of knowledge, brute facts dependupon discursive practices. Or as Onuf aptly summarizes the relationshipbetween brute facts (‘things’) and discursive practices: ‘We construct worlds weknow in a world we do not’ (Onuf, 1989, 38). Consequently, we cannot escapethe interpretive moment (Price and Reus-Smit, 1998, 271) and objective realityis an oxymoron (Puchala, 2000, 139).

This in turn renders problematic both the absolute distinction betweenobject and subject, and the value-neutrality of the observer. We are alwaysinside our discursive frameworks, which shape us and which are shaped by ourpractices. All of this makes theories extremely powerful as they delineate bothwhat can be known and what is sensible to discuss. This is further reinforced by


724


the positivist commitment to empirical validation, as that limits at forehandwhat the discipline should and can study. Positivism presents a ratherstraightforward picture of what kind of stuff exists (hence can be studied) in IRand what not: ‘positivism in international relations y has essentially been amethodological commitment, tied to an empiricist epistemology: together theseresult in a very restricted range of permissible ontological claims y [namely]only in so far as they are empirically warranted’ (Smith, 1996, 11–16, 17).Given that data are only data within a certain theoretical framework, such apositivist–empiricist stance runs the risk of resulting in a vicious circle, whichcan hardly claimed to be productive to our scientific endeavor.

Hence, the post-positivist project appreciates how the very nature of thesocial realm as consisting of ‘facts that are only facts by human agreement’(institutional facts, Searle, 1995, 12) impinges upon methodologies andepistemologies of science (see Hopf, 2007). It should be noted that the closerelationship (interdependency) of ontology and epistemology counts fornatural (‘brute’) and institutional facts alike. However, the problems are mostmanifest and particularly acute when the objects of study are facts by humanagreement. By their very nature they have to be described (can only be) withina framework of shared representations. In this context post-positivism stressesthat epistemology and ontology are interlinked and inseparable. This runsparallel to Kuhn’s notion of a paradigm as an encompassing notion, entailingboth substance and science:

y paradigms differ in more than substance, for they are directed not only tonature but also back upon the science that produced them. They are thesource of the methods, problem-field, and standards of solution accepted byany mature scientific community at any given time. As a result, the receptionof a new paradigm often necessitates a redefinition of the correspondingscience. Some old problems may be relegated to another science or declaredentirely ‘unscientific’yThe normal-scientific tradition that emerges from ascientific revolution is not only incompatible but often actually incommen-surable with that which has gone before. (Kuhn, 1970, 103, italics added)

Since Kuhn was mainly writing about scientific progress in the natural sciences,the term paradigm is deployed here as a heuristic tool that can shed some lighton current developments and controversies within the IR discipline. In thiscontext it is important to note that a paradigm links questions about thefundamental entities in the universe (ontology) to epistemological andmethodological concerns with regard to which questions can be legitimatelyasked about these entities and what techniques can be employed in seekingsolutions. As Kuhn argues, ‘paradigms provide scientists not only with a mapbut also with some of the directions essential for map-making’ (Kuhn, 1970,109). Because of this emphasis on the intimate relationship between substance


725


and science, we find his notion of paradigms useful to discuss what we take tobe the crux of the Third Debate: the interrelatedness between ontology andepistemology (Kratochwil and Ruggie, 1986).

Because of the linkage between substance and science, Kuhn (1970, 93)claims that there is no supra-institutional framework to settle revolutionarydifferences. Hence, in case of a revolution the choice between competingparadigms cannot be settled by the evaluative procedures and criteria ofnormal science. After all, these are dependent upon a particular predominantparadigm, and it is precisely that paradigm which is put into question.Consequently, Kuhn rejects a notion of scientific progress on the basis ofindependent checks against (given) reality (Archimedean points of reference),arguing that in the final analysis, ‘there is no standard higher than the assent ofthe relevant community’ (Kuhn, 1970, 94).

This dual, inclusive notion of science and substance in paradigms nicelycaptures the high stakes involved with the Third Debate. Rather than leadingto a genuine debate with an open attitude, in practice the Third Debate resultedin fence building from both sides of the debate. In contrast to Lapid’s overlyoptimistic representation of the endorsement of ‘a more reflexive intellectualenvironment, involving talking and listening’ on the basis of arguments(Lapid, 1989, 250–251), mainstream theorists continued to emphasize positivistcriteria as a precondition for a dialogue. Hence, far from the depiction ofa ‘neutralisation of the once intimidating bite of the positivist ‘‘anti-scientific’’label’ (Lapid, 1989, 246), Keohane made very clear that if there was any roomfor dialogue, let alone persuasion, it had to be on the terms of the positivistorthodoxy (Keohane, 1988; cf. Katzenstein et al., 1998). Any other argumentswere deemed non-scientific and hence could not even be considered to bepersuasive in the first place (on the basis of their science rather than theirsubstance so to speak). Hence, while in principle the stakes were so high as totouch upon the very identity of IR as a discipline, in practice the effect was lessdramatic. The issues reflected upon were often dismissed as metatheoreticaldiversions and the majority of the discipline continued with business as usual.8

From their side, reflectivists were in turn predisposed to reject any sort ofcriteria, as there are no a priori vantage points from which the truth can beknown. Indeed, reflectivism is often merely defined in terms of its critique of themainstream, that is defined by its ‘anti’s’ (i.e. anti-naturalism, anti-empiricism,anti-objectivism, etcetera) and refusal of dominant positivist commitments(George and Campbell, 1990, 269). As such, the debate might be more aptlydescribed as one between mainstream vs dissidents (Puchala, 2000), which indeedhooks up with the self-designation of the reflectivist side (see Ashley and Walker,1990; George and Campbell, 1990; cf. Keohane, 1988, 392–293).9

Such are then the contours of the middle ground in which constructivism isoperating and the gap it sets out to bridge. Alexander Wendt took up the


726


challenge to find a via media between rationalist and reflectivist approachesand became the pioneer of the bridge-building project.10 He has set the tone inthe sense that constructivism is now often interpreted in terms of Wendt’sreading of the approach (Klotz, 2001). For Wendt the distinctiveness ofconstructivism is considered to lie in its theoretical approach (centrality ofnorms and ideas, intersubjectivity of social reality), not so much in itsepistemological and empirical research strategies (see also Checkel, 1998,Katzenstein et al., 1998). The epistemological stance of Wendt’s constructivismcan be determined by examining two epistemological questions: (i) whether it ispossible to develop causal laws about social reality; and (ii) whether there is anobjective basis for knowledge claims (Fearon and Wendt, 2002, 57). Wendtanswers both these questions with ‘yes,’ and as such shares a positivistepistemology with rationalism. Given the move away from its critical roots,Hopf (1998) labels this type conventional constructivism (see also Price andReus-Smit, 1998). According to Checkel (1998, 327) the very combination ofan intersubjective ontology with a mainstream epistemology makes (conven-tional) constructivism the obvious tenant of the middle ground. Wendt, too,addresses the separation between ontology and epistemology explicitly:

In some sense this [strong belief in science] puts me in the middle of the ThirdDebate, not because I want to find an eclectic epistemology, which I do not,but because I do not think an idealist ontology implies a post-positivistepistemology y Rather than reduce ontological differences to epistemolo-gical ones, in my view the latter should be seen as a third, independent axisof the debate. (Wendt, 1999, 40, italics added)

We rebut and argue instead that it in fact constitutes an erroneous demarcationwith regard to substance and science, as a result of which constructivism hasnot fully lived up to its promise to provide a viable via media.

Delinking Substance and Science: The False Promise of the CurrentMiddle Ground

If the debate between rationalism and conventional constructivism is the(depleted) successor of the Third Debate between rationalism and reflectivism,Keohane should be pleased with the outcome of the challenge he posed to‘middle ground academics,’ that is, to develop a research program that canendorse a synthesis with rationalistic and reflective approaches. Arguably,conventional constructivism does exactly this. However, with its commitmentto a positivist epistemology of social sciences, the bridge-building project hadclear boundaries from the outset. In addition, as we will argue in this section,this delinking of substance from science does not come without costs.


727


The separation of ontology from epistemology is not an innocent move andwe would claim that such a belief in positivism and the concomitant bracketingof epistemology actually puts conventional constructivists out of (or at least atthe outer boundary of ) rather than at the via media between rationalism andreflectivism. With epistemological differences controlled for, in the conven-tional reading a constructivist ontology is arguably compatible with so-called‘normal science’ of rationalist approaches to IR (cf. Jepperson et al., 1996;Katzenstein et al., 1998; Adler, 1997, 2002). While the adherence ofconventional constructivists to a social ontology differs significantly frommainstream conceptions of politics in substance, they claim that thisontological stance still fits within the positivist criteria of science. Thus,despite the normative overtones of the notion of the middle ground whichindicates a more reflexive environment involving bridge-building rather thanfence-building (Reus-Smit, 2001, 223), the conventional constructivist projecthas implicitly, and at times more explicitly, contributed to sustaining thepolemic fences between rationalists and reflectivists by defining positivismas an essential value shared by the scientific community of IR theorists(cf. Jepperson et al., 1996; Katzenstein et al., 1998). The point is that positivists(both of the rationalist and conventional constructivist kind) do notregard these critical viewpoints as part of the academic spectrum, but explicitlytake post-positivist approaches to lie outside the boundaries of (normal)science.

When taken on its own terms, moreover, the conventional constructivistembracement of the positivist practices of social science suffers from internalcontradictions and conundrums. On the one hand, it is maintained that theworld is socially constructed; on the other hand, conventional constructivistanalyses take a certain ‘reality’ as given from which their inquiries start. Theopenness of a social ontology is confronted by an epistemological andmethodological closure that privileges the status quo as the benchmark againstwhich theoretical assumptions and claims can be checked (see for instanceZehfuss, 2002; Hopf, 2007). Thus, whenever these conventional constructivistsgo empirical, they seem unable to stick to their innovative insights about theconstruction of social reality. They seek to design testable theories (and as suchlive up to Keohane’s challenge), but ultimately these tests do not fit their ownconstructivist commands. Hence insofar as their research design does notfollow logically from their research question, constructivists are unable toanswer the questions they ask in an internally consistent way (Pennings et al.,1999). What is at stake here, of course, is that conventional constructivism —given its commitment to a positivist science — cannot deal properly with thefact that actors are always already part of a reality, which can only beinterpreted in relation to our practices, which in turn help constituting thatreality (Zehfuss, 2002, 255).


728


The fact that conventional constructivists have a hard time dealing with thisis a direct consequence of a positivist notion of a ‘correspondence theory oftruth.’ That is, they assume truth to be out there in the world, and it is the jobof scholars to discover theories that correspond with that world. As such thesestatements are to be tested against reality. This of course is problematic whenthat reality is socially constructed and reconstructed by our very practicesand language. Facts do not speak for themselves, but are construed withinour theoretical and discursive frameworks, as reflectivists and criticalconstructivists have convincingly claimed, and language is not a neutralmedium to mirror the independent reality (cf. Fierke, 2002, see Wittgenstein,1953; Laclau and Mouffe, 1985). The fact that constitutive processes cannotbe satisfactorily dealt with in a positivist fashion bears upon the possibilityfor causal reasoning. Causality implies a one-way relationship between anobjectively given cause (reality out there) and some measurable effect, whereasa constructivist ontology of mutual constitution between structures and agentswould in fact defy the articulation of monological causal relationships. This isalso reflected in the standard conventional constructivist solution foroperationalizing mutual constitution: double bracketing (see for instanceFinnemore, 1996; Checkel, 1998; Barnett and Finnemore, 1999). It entails thatone first considers the impact of one level (e.g. structure), while holding theother constant (e.g. agency) and vice versa. In practice, however, this is anunsatisfactory way out as it reduces constitution to causation and contributesto a distorted view of agency and/or structure.11

On the one hand, if they bracket agency, constructivists risk ending up with astructural determinist notion of action where the causal logic of mainstreamapproaches is simply replaced or complemented with explanations that startfrom ideas rather than material factors (see also Finnemore and Sikkink, 1998,Fierke, 2001). However, the empirical conception of norms as causes forbehavior is inconsistent with the intersubjective underpinnings of constructi-vism (in all its variants) as it reduces norms to objective, structural, rather thanintersubjective constraints for social behavior.12

Agency-based interpretations, on the other hand, have a propensity to treatagents as being under-socialized in the sense that these neglect the constitutiveeffects of structures on actors’ identities (Campbell, 1998). Thus, while startingfrom an intersubjective framework, in the course of their empirical analysessome conventional constructivists end up providing cognitive explanationsfor certain actions (to put it more bluntly, they seek to disclose intersubjectiverealities by ‘looking inside people’s heads’).13 As such it differs little frommore idealist forms of rationalism that add some ideational assumptionsto the utility functions of individual actors (cf. Goldstein and Keohane,1993; Jupille et al., 2003). From the ontological commandments ofconstructivism, however, such individualist conceptions of social action should


729


be avoided. A genuine constructivist approach acknowledges that what‘goes on inside heads’ (i.e. private language) is not independent fromhistorical contexts (i.e. public language). Consequently, constructivistsshould not conceptualize ideas and beliefs as something private or individual,but as publicly available mentalities of thought and action that haveevolved over time through interaction and socialization. Thus, although theimpact of intentions is an empirical question that cannot be assumed away,constructivism deals with the intersubjective rather than subjective structuresof reality.

To sum up, its narrow definition and metatheoretical choices precludeconventional constructivism from the ability to live up to its promises. Itsexplicit commitment to a positivist epistemology does not only marginalizeother critical scholarship, it also is an inherently unstable position on its ownterms as it results in a contradiction or clash between ontology andepistemology (Kratochwil and Ruggie, 1986; see also Kratochwil, 2007b).The result is some ‘diluted constructivism’ (Fierke and Jørgensen, 2001, 6).Contrary to Wendt’s suggestions, science and substance are interrelatedindeed, and metatheory is not mere diversion. Hence we need to refocus thedebate and practice what we preach by taking seriously the consequences of asocially constructed reality on the level of methodology and epistemology(Hopf, 2007). Rather than seeking to provide a synthesis between rationalismand reflectivism and rejecting both the positivist notion of objectivism and thealleged relativism of reflectivists, the middle ground project should focus on anapproach to science that explicitly acknowledges that scientific standards, too,are of our making. In the next section, we will consider how a reformulation ofthe middle ground can contribute to a more reflexive IR environment free fromparadigmatic exclusions.

Refocusing the Debate: Exploring the Communicative Middle Ground

Bearing in mind the failed attempt of conventional constructivism to reconcilerationalism and reflectivism, this section seeks to refocus the meaning ofthe middle ground project. We are not the first to propose to refocus thedebate, however. In a thought-provoking article, Patomaki and Wight (2000)have tried to solve the dilemma of the middle ground by displacing the debatebetween rationalism and reflectivism in favor of a critical realist problem-field.In their view, both rationalism and reflectivism are based upon an anti-realistontology in which ontological arguments are derived from epistemologicalones. But because every theory of knowledge presumes a theory of the worldabout which knowledge can be gained, Patomaki and Wight suggest that itwould be more fruitful to start from ontology rather than epistemology: ‘In


730


this respect we want to reverse a long-standing Western philosophical dogma;that of the privileging of epistemological questions over ontological ones’(Patomaki and Wight, 2000, 215). Hence, turning to critical realist philosophy,they claim that research should depart from the assumption of an underlyingmaterial reality with causal effects that are intransitive to those who would liketo know and study them (Patomaki and Wight, 2000, 224). From thisontological commitment, they argue, a critical realist problem-field of apluralist science free of paradigmatic boundaries can emerge:

In summary the critical realist ‘problem-field’ we advocate can be said to becommitted to ontological realism (that there is a reality, which isdifferentiated, structured and layered, and independent of mind), epistemo-logical relativism (that all beliefs are socially produced and hence potentiallyfallible), and judgmental rationalism (that despite epistemological relati-vism, it is still possible, in principle, to provide justifiable grounds forpreferring one theory over another). (Patomaki and Wight, 2000, 224)

While we agree with the critical realist notion of ‘epistemological relativism’and ‘judgmental rationalism,’ we disagree that this could or should be based ona critical realist ontology. First of all, the critical realist argument derives muchof its force from a wrong assumption, namely that reflectivism is anti-realist.As we argued above, this erroneously reduces reflectivism to a branch ofidealism and rejects it on those terms. A more important point perhaps is that acritical realist ontology is not without problems either. Its notion of ontologyas existing of ‘underlying structures, powers, and tendencies that exist, whetheror not detected or known through experience and/or discourse’ (Patomaki andWight, 2000, 223) somehow assumes that the quality and identity of things,objects, subjects and practices are given outside the interpretive practicesthrough which we know them. As such, critical realism seems to adhere, at leastimplicitly, to an essential view of the world that stands in sharp contrast to theinsight that the world is socially constructed.14

Rather than looking for an approach (conventional constructivism, criticalrealism) that can fill the gap, or build the bridge, we propose to conceive ofthe middle ground as a communicative space (Lapid, 2003, 130–131).Concentrating on the link between science and substance on the one hand,and science as a communal and conversational practice on the other, we takeup on Lapid’s plea for engaged pluralism (see also Kratochwil, 2003). Themiddle ground conceived as engaged pluralism does not, as conventionalconstructivism, put forward a single epistemology as the basis for synthesisbetween different approaches. But neither does the absence of a commonepistemology necessarily need to devolve into defensive and polemical stancesbetween different paradigmatic camps. In order to further elaborate this,we focus on rules of conduct, which are analysed along two dimensions:


731


(i) (communicative) dispositions and (ii) (scientific) conventions. Together,these not only constitute the necessary conditions to enable genuine discussionsand effective communication across paradigmatic boundaries, they are alsoconstitutive of IR scholars’ identity as part of the wider social scientificcommunity.15

Communicative Dispositions

At first sight, the notion of a communicative middle ground betweenparadigmatic communities seems to run counter to Kuhn’s incommensurabilitythesis. Because science and substance come in an inextricable mix, thestandards for assessment can only be within a particular research tradition. Assuch it can guard against the danger that approaches are excluded because theyhave an alternative epistemology from the mainstream (Smith, 2003, 143).With Keohane’s address still in mind, this seems to be a valid point. But whileit may protect against epistemological gatekeeping, it is a thin borderlinebetween the plea for self-assessment on the one hand and self-referentialismon the other (cf. Wight, 2002). Self-assessment could lead to the protectionof theories from fundamental criticisms that originate in other researchtraditions, allowing ‘everyone to construct his own little whole — his own littleparadigm, his own little practice, his own little language-game — and thencrawl into it’ (Rorty, 1980, 317). Taken to the extreme, then, theincommensurability thesis appears to suggest that the only alternative to adominant epistemology is that of endless paradigms that live together in asituation of indifferent tolerance between isolated islands of paradigmaticcommunities (Kratochwil, 2003, 126).

However, Kuhn never intended his incommensurability thesis to serve as astraw man for an isolated science, either in terms of ‘flabby, anything-goespluralism’ or ‘fortress-like, incommensurable pluralism’ (which, according toLapid (2003) is the outcome of the Third Debate so far). While it is true thatparadigmatic communities were Kuhn’s main units of analysis, he did notintend them to sit back and indulge in the safe environment wherecommunication is ‘relatively full,’ professional judgement ‘relatively unan-imous,’ and ‘competition y usually quickly ended’ (Kuhn, 1970, 177). In replyto the prevalent misinterpretation of his incommensurability thesis (see Wight,2002), Kuhn conceives of academia as a layered community of whichparadigms constitute the most inner layer, but exist as part of a largerscientific community. This resembles Oakeshott’s notion of societas, which is aform of association in which the members do not relate to each other so muchin terms of common actions and goals but, rather, in terms of their loyaltytowards one another and in terms of their recognition of certain conditionsthat regulate their intercourse (Oakeshott, 1975, 201–203).16 Hence Kuhn


732


speaks of these conditions as values that supersede paradigmatic ties and ‘domuch to provide a sense of community to y scientists as a whole’ (Kuhn,1970, 184). This non-exclusionary notion of layered communities, wheremembership of one research community (e.g. rationalism or reflectivism) doesnot exclude membership of, and hence loyalty to a larger community (that ofthe IR discipline as a whole), is a fundamental premise for a communicativeground and engaged pluralism. While the idea of engaged pluralism may soundsomewhat idealistic — in view of the gate-keeping role of editorial boards,the existence of hidden colleges, citation cartels and the hard competitionfor research funding and tenure — it is also the case that specific researchersdo not only argue their own position, but also pass informed judgement onother positions, precisely because they are involved in an enterprise thatoverrides paradigmatic boundaries! In an attempt to make these ‘conditionsof intercourse’ across paradigmatic boundaries more concrete, the final part ofthis article elaborates rules of conduct, in terms of (i) communicativedispositions (conduct as ‘behavior’) and (ii) scientific conventions (conductas ‘instruction’).

In his rejoinder to the misuse of his incommensurability thesis, and againstthe dubious upshot that the impossibility of a universal language licensesscholars to ‘have their paradigm and eat it,’ Kuhn appeals directly to thedisposition of scholars as participants in the scientific community. In a helpfulanalogy, he likens research traditions to language communities.17 While perfectcommunication does not exist by lack of a universal, neutral language, hemaintains that discussion about paradigmatic viewpoints to some degree isalways possible because of a possibility of translation. Hence, to facilitatecommunication across paradigmatic boundaries, he argues that participantsmust become translators in order to ‘experience vicariously something ofthe merits and defects of each other’s points of view’ (Kuhn, 1970, 201–202).The directions to his students for their introduction to academia areinstructive:

When reading the works of an important thinker, look first for the apparentabsurdities in the text and ask yourself how a sensible person could havewritten them. When you find an answer, y when these passages makesense, then you may find that more central passages, one you previouslythought you understood, have change their meaning’ (Thomas S. Kuhn, TheEssential Tension, Chicago, 1977: xii, quoted in Rorty, 1980, 323)

Hence the presence of difference need not and should not lead to indifference.At the same time, although the metaphor of translation is a useful image tostart from, it risks imposing a one-dimensional picture of communication.Translation shifts most of the burden to the audience and undervalues thecrucial interactive dimension of engaged pluralism. In itself translation is thus


733


not enough but needs to be complemented with that of dialogue, albeit partialand imperfect, between a claimant and its audience. This in turn requires anopen attitude, that is, the acceptance of the other as being different and equalat the same time. That is, while the different parties are considered in all theirdifferences, communication requires certain openness to the other and her/hisviewpoints. As such we need to move beyond the stereotypes that blind us(Hermann, 1998), in order to make genuine engagement possible: ‘The momentof equality in conversation occurs at the point in which a participantacknowledges not only the limits of their own knowledge, but also thepossibility that the other participant(s) may be able to bring to light new waysof seeing or understanding, which are of equal or greater validity’ (Shapcott,2001, 234). This includes an open attitude towards the possibility that the othermay reveal limitations and possible distortions of our analyses andperspectives, that is, an attitude of critical self-reflection (Kratochwil, 2003,126). With regard to (constructively) criticising each other’s work, one shouldbe able to ‘put on different hats,’ that is, be able to take another’s perspectiveand as such judge an analysis/theory on its own premises. This in turn feedsback into Kuhn’s idea of translation.18

A communicative attitude of a social scientist in terms of a readiness totranslate and constructively engage with incompatible worldviews is thus animportant rule of conduct for a flourishing academic community and aprerequisite for a pluralist middle ground. However, for genuine and effectivecommunication to take place and not lapse into a plurality of monologueswhere people talk across each other, there has to be some common frame ofreference. It is important to notice, though, that this does not entail an attemptto reintroduce foundationalism through the back door: the focus is onintersubjectively shared conventions, rather than exogenous standards(Archimedean points) and given principles to validate truth claims. Arguingthat there is a common framework for reference is not the same as claiming agiven rationality on which communication can be grounded. Even Popperacknowledged that rules for rational appraisal depend on shared norms thatguide social interaction between scientists (see Stokes, 1998).19 Neither doesthis entail that ‘rationality’ can only be defined within paradigms, and (any)standards can only be defined within a research tradition, as Smith (2003)wants to argue. Instead, question-driven research on the basis of engagedpluralism simply lays bare what has been general practice all along: that scienceis a communal enterprise with agreement or consensus as the only ‘standard’ toevaluate conflicting knowledge claims (Kuhn, 1970, 1977; Rouse, 1987;Kratochwil, 2000, 89; Pollins, 2007).20

In this sense the understanding of science as a social and communicativepractice does not only refer to the fact that interaction should be at the centerof the scientific enterprise, but also and foremost to the acknowledgement that


734


scientific ‘standards’ are of our making — hence we refer to them asconventions.21 These are not epistemological and not limited to paradigmaticcommunities as such and, crucially, while structuring the debate they arethemselves not exempted from discussion. For example, while they take issuewith Steve Smith’s post-positivist rejections of ‘extra-paradigmatic’ standards,Harvey and Cobb notice that the standards Smith (2002, 37) explicitly appealsto himself (coherence, argumentation, knowledge of literature, empiricalevidence, persuasiveness, innovation, and standards of writing) ‘are virtuallyidentical to those that are widely accepted’ by the mainstream (Harvey andCobb, 2003, 145). We argue that rather than a sole focus on ontological and/orepistemological consensus or synthesis, the middle ground should consist of acommunicative space where standards or conventions such as these are explicitpart of the dialogue, too. This not only should guard against epistemologicalgatekeeping, but also against paradigmatic hiding and huddling (cf. Lebow,2007, 19–21), as it is these conventions that render us part of the widercommunity of social scientists. Indeed, it is such conventions that constituteour very identity as academics.

Scientific Conventions

At the risk of dropping too fast from a metatheoretical to a more practical,applied level, this brings us to the second dimension of rules of conductthat enable translation and effective communication: scientific conventions.These conventions bear witness to the fact that validity is a communicativeprocess that does not in the first place refer to the correspondence betweentheory and facts. As facts are always theory-laden, observations alonecan never prove the invalidity of a truth claim and the process of validationis to a great extent something that takes place between the scholar and her/hisaudience (Pollins, 2007; cf. Kuhn, 1977). Hence, truth claims are not rightor wrong; they are more or less correct depending on how well a scientistsucceeds in arguing the validity of his argument to her or his audience.Similarly objectivity is neither a feature of the world (given in facts), nor acharacteristic of an individual researcher (his/her opinion); rather it is aninherently social phenomenon, guided by the shared rules of academia. Butwhile less abstract, these rules are not less important as following thesescientific conventions makes it possible for an argument to be recognized as avalid scientific argument by other members of the scientific community. Or asKratochwil argues: ‘Truth is y not only contingent on some theoreticalframework and some taken-for-granted or background knowledge y but isalso derived from argumentative procedures’ (Kratochwil, 2000, 90; cf.Kratochwil, 2007b, 13). Argumentative procedures or procedural rules openup the space for dialogue without requiring or claiming consensus and/or


735


synthesis on the level of ontology or epistemology, and hence are not limited tothe rule ‘follow positivist principles in your research’ (Wight, 2007a, 47). Thefinal part of this section therefore takes up Kratochwil’s (2007b) call forinvestigating criteria that lend force to our assertions by preliminarilydiscussing three of such conventions. In this way it is an attempt to movebeyond the discussion about the importance of dialogue and open up the floorfor a frank discussion about such values, standards, rules of conduct — orwhatever one wishes to call them — by which we can explore our identities asacademics.22

A first rule that can enhance the validity of a particular truth claim is thatof retrace-ability. This criterion does not refer so much to the results ofa particular research as the research process itself. As such, it needs to becarefully distinguished from the criterion of reliability or replicability. Thelatter presume that the result of a research can always be replicated if the sameprocedure is applied again (see King et al., 1994). Such a view, however, is onlypossible if one accepts an epistemological stance that acknowledges theexistence of an external reality that can be apprehended outside its discursivemediation, and in a sense denies the role (intervention) of the researcher.Accepting the existence of different epistemologies and aiming for opendialogue, the notion of retrace-ability rather seeks to engage scholars in adebate about the integrity of the research process itself. Retrace-ability definedas the ability to follow the steps of a researcher thus enables effectivecommunication about how a scholar logically proceeds and is, ultimately, aquestion of transparency. For translation and dialogue to be possible, it isimportant that a scientist publicly documents the crucial steps of the researchproject.23 Hence retrace-ability is an important precondition for engaging in adiscussion about any of the two following criteria: internal validity and scopeconditions.

Comparative to the issue of retrace-ability, the criterion of internal validitydoes not refer to some external reality, but concerns the ways in which ascholar moves from a particular research question via a particular researchdesign to the answers of this research question (Pennings et al., 1999). In otherwords, in the context of our argument ‘validity’ should not be conceived interms of approaching the Archimedean point of objective truth, but rather asvalid argumentation in relation to the intersubjective scientific conventions.Rather, it is ‘internal’ to the research (design) in as much as it requires that aresearcher in fact measures what (s)he claims is being measured. Questions ofinternal validity thus refer to the use of concepts and the linkage betweenquestion/theory and evidence.24 It enables a discussion about whether or notour claims are consistent with the evidence provided for these claims. Inaddition, the rule of internal validity to a certain degree permits communica-tion about persuasiveness of different accounts. The truth claim of an


736


internally inconsistent, incoherent account of world politics that does not positits added value vis-a-vis alternative approaches that seek to interpret similarphenomena is obviously less persuasive than a well-argued and internallyconsistent account.

A third communicative condition that enhances effective communication isthe specification of scope conditions. This is a difficult criterion. Scopeconditions are often taken to refer to the domains of application of a certaintheory (Hermann, 1998, 617; Jupille et al., 2003, 21). The idea is that eachtheory has its domain of application which, when put together in the largerpicture, can cumulatively illuminate aspects of the empirical world. However,such a view is problematic as it presupposes that there is only one givenempirical world ‘out there.’ Similar to the notion of replicability, it does notappreciate that facts are theory-laden and that different theories see andconstitute different worlds. However, the idea of specifying scope conditionscan also be interpreted in a slightly different way. In this alternative sense, itsimply serves to increase the falsifiability of a truth claim by promoting adiscussion about the generalizability of a truth claim across different contextsin time and space (Pollins, 2007).25 Appreciating that interpretations arealways partial and open to alternative understandings, scholars can never-theless arrive at conditional generalizations. Or as Price and Reus-Smit put it:‘The issue is not that we either have generalizations or we do not, but that weneed to be conscious about the degrees of abstraction involved in ourstatements about politics and the degree to which our claims and storiestravel from the unique and historically contingent to other contexts’ (Price andReus-Smit, 1998, 275). The specification of scope conditions thus contributesto a communicative process between the researcher and her/his audience aboutthe validity of the proposed constitutive or causal relationships. The questionof how to specify the scope conditions knows no single answer, but one waycould be what King et al. (1994, 208–229) refer to as increasing the number ofobservations. Their advice on this account may help to enhance the likelihoodthat the relationship observed in one context also will be valid in anothercontext (for a more extensive discussion, see Hopf, 2007).

In conclusion, then, the specification of certain conventions that enabledialogue about the validity of different truth claims does not seek to ground IRin a single epistemology or foundational standards. If researchers are genuinelydisposed towards the communicative demands of a middle ground, andacademic community at large, dialogue is a possibility.26 While it may becountered that genuine dialogue ultimately requires a sense of trust andbonding, we would argue with Kuhn that conceiving ourselves as member ofthe wider societas of IR academics (and social scientists) at a minimum createsa certain (obligation of) loyalty. In addition, we argue that within an academicclimate trust should be built on academic disposition such as the ones discussed


737


above (i.e. because one feels to be criticized on a fair basis) rather than thecloseness and safety of paradigmatic bonds. Although some will find dialogueless ambitious than synthesis (Moravscik, 2003), the conditions specifiedabove should allow a cross-paradigmatic discussion about truth claims, whileat the same time ensuring that various empirical agendas can grow on differentontological and epistemological bases. As such the rules of conduct discussedabove can facilitate dialogue within and about the conventions of social science.

Conclusion

‘Understanding the social character of science can be liberatingy’ — we couldnot agree more with King, Keohane and Verba (1994)! In a way, this nicelysummarizes the underlying premise of this paper. It has sought to outline thatconceiving science as a social practice can be a cathartic move for a disciplinethat at times seems stifled by calls for a common epistemology on the one handand pleas for complete paradigmatic autonomy on the other as dubiousoutcomes of the Third Debate. The conceding attempt of conventionalconstructivism to bridge the gap has hardly succeeded. For one thing, it has putforward a very narrow conception of the middle ground, paying only dimattention to the epistemological questions raised in the Third Debate andresulting in what has been labelled a mere ‘facelift of the mainstream’ (Guzzini,2000). In addition, and more importantly, within conventional constructivismscience and substance do not match up, which, we argued, leads to inherentcontradictions.

This paper suggested a re-exploration of the middle ground betweenrationalism and reflectivism in order to overcome the deadlock betweenepistemological gatekeeping and indiscriminate tolerance. As an alternativeavenue, the paper put forward an procedural attitude towards science thatdismisses both the positivist view that scientific assertions are supposed tomarshal universal assent (Kratochwil, 2003, 124), and the reflectivist stancethat it is (virtually) impossible to make any knowledge claims, given the lack ofexternal reference points from which to judge and be judged upon. The way weconceive it, both these positions ignore two crucial observations: (i) thatontology and epistemology come in an inextricable mix; and (ii) that science isa communal and conversational enterprise. Our view of the middle ground hassought to make this explicit. It argued for an engaged pluralism, entailing theview that warranted knowledge claims can only be made if scientists are willingto engage in a dialogue about standards of scientific arguments. To continuethe quote by King, Keohane, and Verba: ‘ y [a]s long as our work explicitlyaddresses (or attempts to redirect) the concerns of the community of scholarsand uses public methods to arrive at inferences that are consistent with rules of


738


science and the information at our disposal, it is likely to make a contribution’(King et al., 1994, 9, italics added). This is indeed what we argue, albeit withone crucial qualification, namely that the rules of science are part of the socialpractice too, and science as such does not equal positivist standards asobjective methods to unravel the Truth out there, waiting to be discovered.Rather, the conventions discussed here consist of ‘professional standards,’standards set by and for the profession, which allow us to move beyondepistemology and which constitute the very boundaries of our identity as IRscholars and, ultimately, social scientists.

A revival of the Third Debate as advocated here may turn out to bebeneficial by (re)confirming that there are multiple pathways to knowledgeabout IR, with a common focus on edification (Puchala, 2000, 141–142).Conceived as such, the middle ground can indeed provide a powerful challengeto epistemological gatekeeping and indiscriminant tolerance, which precludesdiscussion about quality of research. It can help both rationalists andreflectivists forward by providing a procedural framework that can guidevarious empirical agendas (in the broad sense).

About the authors

Tanja E. Aalberts is Assistant Professor for International Relations at theDepartment of Political Science, Leiden University (The Netherlands). Shereceived her Ph.D. from the Vrije Universiteit, Amsterdam. Her thesis onpolitical-legal discourses of sovereignty was granted the Jean BlondelDissertation Award 2007 by the ECPR. She has published in the Journal ofCommon Market Studies and the International Journal for the Semiotics of Law,as well as several international edited volumes.

Rens van Munster is Assistant Professor of International Politics at theDepartment of Political Science, University of Southern Denmark (Odense).He is interested in the politicization of immigration as a security issue and hecurrently works on questions of risk, catastrophe and the war on terror. He has(co-)published in journals such as European Journal of International Relations,International Relations and the Internal Journal for the Semiotics of Law. He isthe co-editor of a special issue of Security Dialogue on ‘Security, Technologiesof Risk, and the Political.’

Notes

1 An earlier version of this paper was presented at the Fifth Pan-European International

Relations Conference ‘Constructing World Orders’ of the ECPR/SGIR The Hague, September

9–11, 2004.


739


2 See also Checkel (1997), Finnemore and Sikkink (1998, 2001), Jupille et al. (2003), Risse (2002),

and Sterling-Folker (2002). For a genuine debate between constructivism and rationalism, and

an elaboration of a constructivist research program in positivist terms, see Checkel and

Moravcsik (2001).

3 For typologies of different types of constructivism, see for instance Katzenstein et al. (1998),

Adler (1997), Hopf (1998), Ruggie (1998).

4 Especially poststructuralists have questioned the idea of epistemology as rules on how to

produce secure, let alone objective, knowledge (cf. Wæver, 1997, 16–17). Also, for them

the ontological does not refer so much to ‘what is’ but is invoked mainly as a way of criticizing

the onto-political (see e.g. Campbell, 1998; Edkins and Pin-Fat, 1999). However, even if the

concern is with the onto-political (the way central ontological categories such as sovereignty or

the state are politically constructed), it is still necessary to document carefully the dominant

discourses through which the onto-political is produced, resisted and reified. In other words,

poststructuralists too cannot escape the question of what constitutes good research in the

examination of the onto-political (see Price and Reus-Smit, 1998; Milliken, 1999). See also the

fourth section.

5 Mervyn Frost (1986), Towards a Normative Theory of International Relations, Cambridge:

Cambridge University Press; quoted by George and Campbell (1990). Lapid makes a similar

remark when he talks about the tragedy of international relations, as ‘they proved incapable of

either fruitfully adopting or decisively rejecting the grail of positivist science’ (Lapid, 1989, 246).

6 This was also the point of contention in the so-called Second Debate between traditionalist/

historians and behavioralists/scientists (cf. Schmidt, 2002). However, this ‘history vs science’

controversy did not really disturb the safe grounds of positivist epistemology, and focused on

methodological issues instead (cf. Lapid, 1989). In this account, Wæver’s (1996) explicit

distinction between the interparadigm debate and the more philosophical debate between

rationalists and reflectivists has indeed been useful to draw attention to the latter. Nevertheless,

it risks a too clear separation of substance and science. This issue is pursued below.

7 The strategy to reduce reflectivism to idealism and to reject it on those terms has been repeatedly

deployed by IR theorists (see Mearsheimer, 1995; Østerud, 1996; Wight, 1999). On the difference

between reflectivism and idealism, see Laclau and Mouffe (1990).

8 Keohane (1988, 382) reproached reflectivist scholars for leading the discipline into a ‘debate at

the purely theoretical level,’ which distracts researchers from the subject matter, world politics,

in favor of a ‘programmatically diversionary philosophical discussion.’

9 It should be noted that whereas Lapid equals the rationalist/reflectivist debate with positivism vs

post-positivism, the ‘dissident agenda’ entails more than a post-positivist stance. George and

Campbell (1990, 270) identify the following interdisciplinary elements of critical analysis: (i)

inadequacy of positivist/empiricist approaches to knowledge; (ii) focus on actual process of

knowledge construction in repudiating external sources of understanding (i.e. Archimedean

points); (iii) focus on linguistic construction of reality; (iv) focus on question of subjectivity, that

is, construction of meaning and identity.

10 See Wendt (1992) for a first contribution to the bridge-building project, but notably Wendt

(1999) for a full-fledged articulation of a social theory to international politics.

11 For a detailed discussion of the uneasy tensions of structure/agency, and in particular of mental

causality and rump materialism in ‘middle-ground’ constructivism see also Kessler (2007).

12 In turn this makes it virtually impossible to theorize change (which was one of the original

promises of constructivism and its main critique against neo-realism, see e.g. Wendt, 1992).

Moreover, such a reading of norms ignores another of Wendt’s premises, that is, the need to move

beyond the level of behavior and take effects on identity and, ultimately, interests into account (see

e.g. Wendt, 1999). Finally, it ignores Kratochwil and Ruggie’s (1986, 767) famous critique that it

is problematic to conceive of norms as causes, because they are counterfactually valid.


740


13 The resort to cognitive psychology is advocated by Checkel (1998, 2001a). See also Johnston

(2001). To understand how socialization works, many constructivists study arguing dynamics,

focusing on the role of persuasion, deliberation, shaming and rhetorical action (see, among

others, Risse et al., 1999; Risse, 2000; Checkel, 2001b; Schimmelfennig, 2001). Although

substantial differences exist among them, most authors — drawing upon laboratory-like, non-

contextual experiments of communication — go on to specify the conditions under which actors

are more likely to persuade other actors of the appropriateness of new norms. The question is,

however, whether socialization can be ‘measured’ accurately if one only focuses upon arguing

dynamics as such, hence ignoring issues of embeddedness in broader discursive frameworks.

Kessler (2007), in this regard, distinguishes between ‘thin’ and ‘thick intersubjectivity.’

14 See also a recent exchange between Friedrich Kratochwil (2007a, b) and Colin Wight (2007a, b)

on these issues.

15 This parallels Poppers later conception of the scientific community as ‘a set of persons who

share certain techniques such as measurement procedures, methodological commitments and

presumptions of what constitutes ‘‘good practice’’ in a given field’ (Karl Popper (1972),

Objective Knowledge: An Evolutionary Approach, Oxford University Press; quoted by Lebow,

2007, 17).

16 In distinction to societas, Oakeshott refers to universitas as a mode of association that postulates

the idea of a substantive or common purpose. This is the type of association characteristic for

Kuhn’s paradigmatic community which pursues a set of common goals, hence enjoying

relatively full communication and unanimous professional judgement (Kuhn, 1970: 176–178).

17 For a similar analogy, see Hermann (1998) who compares different approaches to distinct kind

of grammatical structures and IR to the Tower of Babel. See also Guzzini (2001).

18 In terms of engagement and open dialogue between opposing perspectives, a recent exemplar is

Checkel and Moravcsik’s (2001) forum about constructivism vs rationalism in European Union

Politics. A similar endeavor was undertaken by Fierke and Nicholson (2001) in their

conversation on the notion of games from a linguistic, respectively rational choice perspective.

However, as discussed above, we take issue with the emphasis on synthesis, which seems to be

the aim in both cases.

19 Kratochwil (2000, 92) argues that Popper’s stance in his later work does not differ that much

from Kuhn’s. In The Myth of the Framework Popper argues the virtue of disagreement, where

communication does not depend on a shared (theoretical) framework, but rather on the respect

of certain ethical principles. See also Lebow (2007).

20 This has also been the stake of a recent debate on the merits of a pragmatic approach to theory

building, as proposed by Kratochwil (2007b) in his Tartu lecture. See the special issue of the

Journal of International Relations and Development 10(1): 2007.

21 For instance, Foucault remarks that ‘those texts that we now would call scientific — those

dealing with cosmology and the heavens, medicine and illnesses, natural sciences and geography

— were accepted in the Middle Ages, and accepted as ‘‘true’’, only when marked with the name

of their author. ‘‘Hippocrates said’’, ‘‘Plinty recounts’’, were not really formulas of an argument

based on authority; they were the markers inserted in discourses that were supported to be

received as statements of demonstrated truth’ (Foucault, 1991, 109).

22 In The Essential Tension Kuhn (1977, 322) lists five indicative criteria that partly overlap:

‘accuracy, consistency, scope, simplicity, and fruitfulness.’ Whereas these count as a set of values

to guide and shape theory development, these criteria are rather imprecise on their own terms —

which is the(ir) very point: ‘What the tradition sees as eliminable imperfections in its rules of

choice I take to be in part responses to the essential nature of science’ (Kuhn, 1977, 330). In

other words, this only illuminates the social character of the scientific endeavor — because

scientists may differ about the relative weight of the criteria, and for instance their application to

concrete cases, science is not a determinate process dictated by ‘rational, unanimous choice’ and


741


a given formula (Kuhn, 1977, 326, quoted by Suganami, 2007, 31). The discussion about

‘scientific’ criteria is as much part of the academic endeavor ‘of our making’ as the substance of

particular research projects themselves.

23 This includes conventions with regard to citations, footnotes and bibliographies (see also

note 21).

24 For instance, the operationalization of mutual constitution of much conventional constructivist

work is problematic in this regard. While these scholars are to be praised for their efforts to deal

with the structure/agency problem in empirical research, the proposed solution of ‘bracketing’

so far does not seem very satisfying for it ends up measuring the impact of structure upon agents

on the one hand, and the impact of agents on structure, but cannot grasp the dynamic of mutual

constitution. This is even more problematic in case of those studies that add a temporal

dimension to the analysis. Alternatively, such analyses result in privileging the agent over the

structure, or vice versa.

25 Pollins (2007, 93–94) usefully distinguishes between falsifiability and falsificationism. Whereas

he rejects the latter because of its positivist overtones, he embraces the ideal of falsifiability.

Falsifiability can be defined simply, as King et al. (1994, 19) do, as ‘the principle that theories

should be stated clearly enough so that they could be wrong.’

26 To a certain extent this possibility is reflected in how we teach students about the historiography

of our discipline. Being detached to the enmities of the debates, they are often encouraged to put

on different hats in order to give specific theories their best shot, and dissect internal

inconsistencies at the same time. On parochialism in teaching, see Alker and Biersteker (1984).

References

Adler, E. (1997) ‘Seizing the Middle Ground: Constructivism in World Politics’, European Journal

of International Relations 3(3): 319–363.

Adler, E. (2002) ‘Constructivism and International Relations’, in W. Carlsnaes, T. Risse and B.A.

Simmons (eds.) Handbook of International Relations, London: Sage.

Alker, H.R. and Biersteker, T. (1984) ‘The Dialectics of World Order: Notes for a Future

Archeologist of International Savoir Faire’, International Studies Quarterly 28(2): 121–142.

Ashley, R.K. and Walker, R.B.J. (1990) ‘Introduction: Speaking the Language of Exile: Dissident

Thought in International Studies’, International Studies Quarterly 34(3): 259–268.

Ba, A. and Hoffman, M.J. (2003) ‘Making and Remaking the World for IR 101: A Resource for

Teaching Social Constructivism in Introductory Classes’, International Studies Perspectives 4(1):

15–33.

Barnett, M. and Finnemore, M. (1999) ‘The Politics, Power and Pathologies of International

Organizations’, International Organization 53(4): 705–732.

Campbell, D. (1998) National Deconstruction: Violence, Identity, and Justice in Bosnia,

Minneapolis: University of Minneapolis Press.

Checkel, J.T. (1997) ‘International Norms and Domestic Politics: Bridging the Rationalist-

Constructivist Divide’, European Journal of International Relations 3(4): 473–495.

Checkel, J.T. (1998) ‘The Constructivist Turn in International Relations Theory’, World Politics

50(1): 324–348.

Checkel, J.T. (2001a) ‘Taking Deliberation Seriously’, ARENA Working Paper no. 14, ARENA/

University of Oslo, Oslo.

Checkel, J.T. (2001b) ‘Why Comply? Social Learning and European Identity Change’, International

Organization 55(3): 553–588.

Checkel, J.T. and Moravcsik, A. (2001) ‘A Constructivist Research Program in EU Studies?

(Forum Section)’, European Union Politics 2(2): 219–249.


742


Edkins, J. and Pin-Fat, V. (1999) ‘The Subject of the Political’, in J. Edkins, N. Persram and

V. Pin-Fat (eds.) Sovereignty and Subjectivity, Boulder: Lynne Rienner.

Fearon, J. and Wendt, A. (2002) ‘Rationalism v. Constructivism: A Skeptical View’, in

W. Carlsnaes, T. Risse and B.A. Simmons (eds.) Handbook of International Relations, London:

Sage.

Fierke, K.M. (2001) ‘Critical Methodology and Constructivism’, in K.M. Fierke and K.E. Jørgensen

(eds.) Constructing International Relations: The Next Generation, London: M.E. Sharpe.

Fierke, K.M. (2002) ‘Links Across the Abyss: Language and Logic in International Relations’,

International Studies Quarterly 46(3): 331–354.

Fierke, K.M. and Jørgensen, K.E. (2001) ‘Introduction’, in K.M. Fierke and K.E. Jørgensen (eds.)

Constructing International Relations. The Next Generation, New York: M.E. Sharpe.

Fierke, K.M. and Nicholson, M. (2001) ‘Divided by a Common Language: Formal and

Constructivist Approaches to Games’, Global Society 15(1): 7–25.

Finnemore, M. (1996) National Interests in International Society, Ithaca: Cornell University Press.

Finnemore, M. and Sikkink, K. (1998) ‘International Norm Dynamics and Political Change’,

International Organization 52(4): 887–917.

Finnemore, M. and Sikkink, K. (2001) ‘Taking Stock: The Constructivist Research Program in

International Relations and Comparative Politics’, Annual Review of Political Science 4: 391–416.

Foucault, M. (1991) ‘What is an Author?’, in P. Rabinow (ed.) The Foucault Reader, London:

Penguin.

George, J. and Campbell, D. (1990) ‘Patterns of Dissent and the Celebration of Difference: Critical

Social Theory and International Relations’, International Studies Quarterly 34(3): 269–293.

Goldstein, J. and Keohane, R.O. (eds.) (1993) Ideas and Foreign Policy: Beliefs, Institutions, and

Political Change, Ithaca: Cornell.

Guzzini, S. (2000) ‘A Reconstruction of Constructivism in International Relations’, European

Journal of International Relations 6(2): 147–182.

Guzzini, S. (2001) ‘The Significance and Roles of Teaching Theory in International Relations’,

Journal of International Relations and Development 4(2): 98–177.

Harvey, F. and Cobb, J. (2003) ‘Multiple Dialogues, Layered Syntheses, and the Limits of

Expansive Cumulation’, International Studies Review 5(1): 144–147.

Hermann, M.G. (1998) ‘One Field, Many Perspectives: Building the Foundations for Dialogue’,

International Studies Quarterly 42(4): 605–624.

Hopf, T. (1998) ‘The Promise of Constructivism in International Theory’, International Security

23(1): 171–200.

Hopf, T. (2007) ‘The Limits of Interpreting Evidence’, in R.N. Lebow and M. Lichbach (eds.)

Theory and Evidence in Comparative Politics and International Relations, Houndmills: Palgrave

Macmillan.

Jepperson, R.L., Wendt, A. and Katzenstein, P.J. (1996) ‘Norms, Identity, and Culture in National

Security’, in P.J. Katzenstein (ed.) The Culture of National Security: Norms and Identity in World

Politics, New York: Columbia University Press.

Johnston, A.I. (2001) ‘Treating International Institutions as Social Environments’, International

Studies Quarterly 45(4): 487–516.

Jupille, J., Caporaso, J.A. and Checkel, J.T. (2003) ‘Integrating Institutions. Theory, Method, and

the Study of the European Union’, Comparative Political Studies 36(1): 7–41.

Katzenstein, P.J., Keohane, R.O. and Krasner, S.D. (1998) ‘International Organization and the

Study of World Politics’, International Organization 52(4): 645–685.

Keohane, R.O. (1988) ‘International Institutions: Two Approaches’, International Studies

Quarterly 32(4): 379–396.

Kessler, O. (2007) ‘From Agents and Structures to Minds and Bodies: Of Supervenience, Quantum,

and the Linguistic Turn’, Journal of International Relations and Development 10(3): 243–271.


743


King, G., Keohane, R.O. and Verba, S. (1994) Designing Social Inquiry: Scientific Inference in

Qualitative Research, Princeton: Princeton University Press.

Klotz, A. (2001) ‘Can We Speak a Common Constructivist Language?’, in K.M. Fierke and K.E.

Jørgensen (eds.) Constructing International Relations: The Next Generation, London: M.E.

Sharpe.

Kratochwil, F. (2000) ‘Constructing a New Orthodoxy? Wendt’s ‘Social Theory of International

Politics’ and the Constructivist Challenge’, Millennium 29(1): 73–101.

Kratochwil, F. (2003) ‘The Monologue of ‘‘Science’’’, International Studies Review 5(1): 124–128.

Kratochwil, F. and Ruggie, J.G. (1986) ‘International Organization: A State of the Art on an Art of

the State’, International Organization 40(4): 753–775.

Kratochwil, F.V. (2007a) ‘Of Communities, Gangs, Historicity and the Problem of Santa

Claus: Replies to my Critics’, Journal of International Relations and Development 10(1):

57–78.

Kratochwil, F.V. (2007b) ‘Of False Promises and Good Bets: a Pleas for a Pragmatic Approach to

Theory Building (the Tartu Lecture)’, Journal of International Relations and Development 10(1):

1–15.

Kuhn, T.S. (1970) The Structure of Scientific Revolutions, Chicago: University of Chicago Press.

Kuhn, T.S. (1977) The Essential Tension, Chicago: Chicago University Press.

Laclau, E. and Mouffe, C. (1985) Hegemony and Socialist Strategy: Towards a Radical Democratic

Politics, London: Verso.

Laclau, E. and Mouffe, C. (1990) ‘Post-Marxism Without Apologies’, in E. Laclau (ed.) New

Reflections on the Revolution of Our Time, London: Verso.

Lapid, Y. (1989) ‘The Third Debate: On the Prospects of International Theory in a Post-Positivist

Era’, International Studies Quarterly 33: 235–254.

Lapid, Y. (2003) ‘Through Dialogue to Engaged Pluralism: The Unfinished Business of the Third

Debate’, International Studies Review 5(1): 128–131.

Lebow, R.N. (2007) ‘Social Science as an Ethical Practice’, Journal of International Relations and

Development 10(1): 16–24.

Maghroori, R. and Ramberg, B. (eds.) (1982) Globalism versus Realism: International Relations’

Third Debate, Boulder: Westview Press.

Mearsheimer, J.J. (1995) ‘The False Promises of International Institutions’, International Security

19(3): 5–49.

Milliken, J. (1999) ‘The Study of Discourse in International Relations: A Critique of Research and

Methods’, European Journal of International Relations 5(2): 225–254.

Moravscik, A. (2003) ‘Theory Synthesis in International Relations: Real Not Metaphysical’,

International Studies Review 5(1): 131–136.

Oakeshott, M. (1975) On Human Conduct, Oxford: Clarendon.

Onuf, N.G. (1989) World of Our Making: Rules and Rule in Social Theory and International

Relations, Columbia: University of South Carolina Press.

Østerud, Ø (1996) ‘Antinomies of Postmodernism in International Studies’, Journal of Peace

Research 33(4): 385–390.

Patomaki, H. and Wight, C. (2000) ‘After Postpositivism? The Promises of Critical Realism’,

International Studies Quarterly 44: 213–237.

Pennings, P., Keman, H. and Kleinnijenhuis, J. (1999) Doing Research in Political Science. An

Introduction to Comparative Methods and Statitiscs, London: Sage.

Pollins, B.M. (2007) ‘Beyond Logical Positivism: Reframing King, Keohane and Verba’s Designing

Social Inquiry’, in R.N. Lebow and M. Lichbach (eds.) Theory and Evidence in Comparative

Politics and International Relations, Houndmills: Palgrave Macmillan.

Price, R. and Reus-Smit, C. (1998) ‘Dangerous Liaisons? Critical International Theory and

Constructivism’, European Journal of International Relations 4(3): 259–294.


744


Puchala, D.J. (2000) ‘Marking a Weberian Moment: Our Discipline Looks Ahead’, International

Studies Perspectives 1: 133–144.

Reus-Smit, C. (2001) ‘Constructivism’, in S. Burchill, Linklater, A, Devetak, R, Donnelly, J,

Paterson, M, Reus-Smit, C. and True, J. (eds.) Theories of International Relations, Basingstoke:

Palgrave.

Risse, T. (2000) ‘‘‘Let’s Argue!’’ Communicative Action in World Politics’, International


Risse, T. (2002) ‘Constructivism and International Institutions: Toward Conversations across

Paradigms’, in I. Katznelson and H.V. Milner (eds.) Polticial Science. State of the Discipline,

New York: W.W. Norton.

Risse, T., Ropp, S.C. and Sikkink, K. (1999) The Power of Human Rights: International Norms and

Domestic Change, Cambridge: Cambridge University Press.

Rorty, R. (1980) Philosophy and the Mirror of Language, Oxford: Blackwell.

Rouse, J. (1987) Knowledge and Power: Toward a Political Philosophy of Science, Ithaca: Cornell

University Press.

Ruggie, J.G. (1998) ‘What Makes the World Hang Together? Neo-utilitarianism and the Social

Constructivist Challenge’, International Organization 52(4): 855–885.

Schimmelfennig, F. (2001) ‘The Community Trap: Liberal Norms, Rhetorical Action and the

Eastern Enlargement of the European Union’, International Organization 55(1): 47–80.

Schmidt, B.C. (2002) ‘On the History and Historiography of International Relations’, in

W. Carlsnaes, T. Risse and B.A. Simmons (eds.) Handbook of International Relations, London:

Sage.

Searle, J.R. (1995) The Construction of Social Reality, New York: Free Press.

Shapcott, R. (2001) Justice, Community, and Dialogue in International Relations, Cambridge:

Cambridge University Press.

Smith, S. (1996) ‘Positivism and Beyond’, in S. Smith, K. Booth, and M. Zalewski (eds.)

International Theory: Positivism and Beyond, Cambridge: Cambridge University Press.

Smith, S. (2001) ‘Reflectivist and Constructivist Approaches to International Theory’, in J. Baylis

and S. Smith (eds.) The Globalization of World Politics: an Introduction to International

Relations, 2nd edn, Oxford: Oxford University Press.

Smith, S. (2002) ‘Alternative and Critical Perspectives’, in F.P. Harvey and M. Brecher (eds.)

Millennial Reflections on International Studies, Ann Arbor: University of Michigan Press.

Smith, S. (2003) ‘Dialogue and the Reinforcement of Orthodoxy in International Relations’,

International Studies Review 5(1): 141–143.

Sterling-Folker, J. (2002) ‘Realism and the Constructivist Challenge: Rejecting, Reconstructing, or

Rereading’, International Studies Review 4(1): 73–100.

Stokes, G. (1998) Popper. Philosophy, Politics and Scientific Method, Cambridge: Polity Press.

Suganami, H. (2007) ‘Friedrich Kratochwil’s Pragmatic Search for a Theory of International

Relations’, Journal of International Relations and Development 10(1): 25–39.

Wæver, O. (1996) ‘The Rise and Fall of the Inter-paradigm Debate’, in S. Smith, K. Booth, and

M. Zalewski (eds.) International Theory: Positivism and Beyond, Cambridge: Cambridge

University Press.

Wæver, O. (1997) ‘Figures of International Thought: Introducing Persons instead of Paradigms’, in

I.B. Neumann and O. Wæver (eds.) The Future of International Relations. Masters in the

Making?, London: Routledge.

Walker, R.B.J. (1989) ‘History and Structure in the Theory of International Relations’, Millennium

18(2): 163–183.

Walt, S.M. (1998) ‘International Relations: One World, Many Theories’, Foreign Policy 110: 29–47.

Wendt, A. (1987) ‘The Agent-Structure Problem in International Relations Theory’, International



745


Wendt, A. (1992) ‘Anarchy is What States Make of It: The Social Construction of Power Politics’,

International Organization 46(2): 391–425.

Wendt, A. (1999) A Social Theory of International Politics, Cambridge: Cambridge University

Press.

Wendt, A. (2000) ‘On the Via Media: A Response to the Critics’, Review of International Studies

26(1): 165–180.

Wight, C. (1999) ‘Metacampbell: The Epistemological Problematics of Perspectivism’, Review of

International Studies 25(2): 311–316.

Wight, C. (2002) ‘Philosophy of Social Science and International Relations’, in W. Carlsnaes,

T. Risse and B.A. Simmons (eds.) Handbook of International Relations, London: Sage.

Wight, C. (2007a) ‘Inside the Epistemological Cave All Bets are Off ’, Journal of International

Relations and Development 10(1): 40–56.

Wight, C. (2007b) ‘A Response to Friedrich Kratochwil: Why Shooting the Messenger Does

Not Make the Bad News Go Away!’, Journal of International Relations and Development 10(3):

301–315.

Wittgenstein, L. (1953) Philosophical Investigations, Oxford: Blackwell.

Zehfuss, M. (2002) Constructivism in International Relations: The Politics of Reality, Cambridge:

Cambridge University Press.


746


Aalberts and Van Munster. From Wendt to Kuhn

Documents

rens van munster

cornell university press

constructivist research program

middle ground project

cambridge university press

critical realist problem

pressing epistemological issues

constructing international relations