Bridging the Rational-Choice / Constructivist Gap? Theorizing Social Interaction in European Institutions*
Jeffrey T. Checkel**
ARENA and University of Oslo
The debate between rationalists and social constructivists, which has played out in the broader discipline for many years, has finally reached studies of integration (Christiansen, Joergensen and Wiener 1999; Moravcsik 1999a; Aspinwall and Schneider 2000a). While much still separates these schools, there are accumulating signs this gap is closing -- at least within international relations theory (International Organization 1998). My memo contributes to this emerging dialogue by theorizing social interaction within European institutions.
If one defines social interaction as a process during which fundamental agent properties can change, then it is clear that rational choice or rational-choice institutionalist arguments have little to say in this regard. Premised upon a strong form of methodological individualism, they reduce interaction to strategic exchange among actors with pre-social givens (Kratochwil and Ruggie 1986; 707). One might think that constructivists, with their mutually constitutive ontology, would be all about process and social interaction, with preferences endogenous to the latter. Surprisingly, this has not been the case, especially among mainstream constructivists in the US. Instead, much like the rationalists from whom they so sharply distinguish themselves, these scholars have employed a variant of "as if" reasoning. They typically argue that fundamental agent properties have been reshaped by prevailing social norms, but fail to theorize or empirically document the process of social interaction through which this occurs; agents act "as if" their behavior becomes rule-governed (Checkel 2000a; see also Risse, Ropp and Sikkink 1999, 29). 
Given this state of affairs, the good news from Europe is that both soft rationalists and soft constructivists are beginning to theorize this missing element of social interaction. For constructivists, this has meant a much-needed (re)turn to questions of agency and decisionmaking (Checkel 1999; Risse 2000); for rationalists, it has meant a new emphasis on language and communication -- in ways that go beyond cheap talk and signalling games (Underdal 1998, 20-23; Schimmelfennig 1999). 
The memo proceeds as follows. I begin by asking what it means to theorize social interaction from a constructivist perspective, noting points of contact/overlap with rationalist arguments. Next, I explore how one could empirically operationalize such an approach. Finally, I suggest why theories of social interaction are important for students of European institutions, as well as for institutional theorists who study the EU.
Theorizing Social Interaction
"Socialization," "social learning," "deliberation," "rule-driven behavior," and the like -- these are the buzzwords of choice for constructivists, many sociological institutionalists and numerous non-rationalist students of the EU (Caporaso and Jupille 1999, 433). These terms all imply a social process through which agent properties and preferences change as a result of interaction. Unfortunately, much of the research employing them has emphasized end states, where socialization is complete or rule-governed behavior a given. Interaction thus drops out of the analysis. As a result, the causal mechanisms and motors underlying such processes have been neglected (Johnston 1998; Checkel 1999, 5-9; Risse, Ropp and Sikkink 1999, chapter 1).
Whatever the buzzword, language and communication play central roles in such analyses. For example, the social interaction that occurs within EU institutions is often characterized as a process of argumentation and deliberation (Joerges and Neyer 1997); moreover, international norms, we are told, socialize states by empowering new social actors who then seek, with principled arguments, to persuade and shame political elites (Price 1998). Given this seemingly ubiquitous role for language, it is odd that constructivists and sociological institutionalists systematically fail to theorize it.
Here, I suggest one route to such theorization: work in social psychology and communications research on persuasion and argumentation. In considering this literature, however, one should keep in mind an important distinction -- namely, the fundamental difference between manipulative and argumentative persuasion. The former is asocial and lacking in interaction, often concerned with political elites manipulating mass publics and has a long tradition -- extending back to studies by William Riker. With its individualism and emphasis on strategic agency, persuasion of this sort figures prominently in the work of several rational-choice scholars (Moravcsik 1999b, 272, 281; see also Riker 1986; Evangelista 1999). 
In contrast, argumentative persuasion is a social process of interaction that involves changing attitudes about cause and effect in the absence of overt coercion. It is thus a mechanism through which preference change may occur. More formally, it is "an activity or process in which a communicator attempts to induce a change in the belief, attitude or behavior of another person ... through the transmission of a message in a context in which the persuadee has some degree of free choice." Here, persuasion is not manipulation, but a process of convincing someone through argument and principled debate (Brody, Mutz and Sniderman 1996, chapters 1, 5-6; Lupia and McCubbins 1998, chapter 3).
For sure, the persuasion literature is not without its own limitations. In particular, much of this work, owing to its disciplinary roots in social psychology, proceeds inductively. One result has been a failure to develop middle-range theory specifying scope conditions. Below, I advance five such conditions for when agents should be especially open to argumentative persuasion and preference change; however, given the inductive approach and occasional contradictions within the literature, these should be viewed as preliminary.
H#1 Argumentative persuasion is more likely to be effective when the persuadee is in a novel and uncertain environment and thus cognitively motivated to analyze new information (Zimbardo and Leippe 1991, 225).
H#2 Argumentative persuasion is more likely to be effective when the persuadee has few prior, ingrained beliefs that are inconsistent with the persuader's message. Put differently, agents with few cognitive priors who are novices will be more open to persuasion (Brody, Mutz, Sniderman 1996, 161-62; Gibson 1998, 833, 835; see also Stryker 1980; Aspinwall and Schneider 2000b, 31).
H#3 Argumentative persuasion is more likely to be effective when the persuader is an authoritative member of the in-group to which the persuadee belongs or wants to belong (Joergensen, Kock and Roerbech 1998, 287; Johnston 1998, 16-25).
H#4 Argumentative persuasion is more likely to be effective when the persuader does not lecture or demand, but, instead, "acts out principles of serious deliberative argument" (Joergensen, Kock and Roerbech 1998, 297; see also Brody, Mutz, Sniderman 1996, 154).
H#5 Argumentative persuasion is more likely to be effective when the persuader-persuadee interaction occurs in less politicized and more insulated, in-camera settings (Elster 1991, 46-47; Idem 1998, 109-111).
This focus on argumentative persuasion begins to operationalize the roles of communication and social interaction implicit but undertheorized in the work of constructivists, sociological institutionalists and students of the EU. The result is to expand our analytic toolkit for explaining why and how social actors change their preferences. In some cases, they do so by learning new interests via non-strategic communication and persuasion. The ontology and understanding of social reality here is not individualist, but relational (Ruggie 1998, 4); I take seriously dynamics of social interaction. Agents thus "decide" by puzzling and arguing.
I should emphasize that the approach sketched above is not method-driven. It develops scope conditions recognizing that preference change caused by social learning and persuasion often does not occur. This leaves plenty of analytic space for rationalist arguments and strategic bargaining games, as I have demonstrated elsewhere (Checkel 1999).
The Methodological Challenges
A central -- and largely correct -- criticism of constructivist approaches to the study of the EU has been their theoretical and even meta-theoretical orientation. That is, tough questions of research methodology and operationalization have been neglected (Moravcsik 1999a; Caporaso and Jupille 1999, 436). In a larger work-in-progress, I address these issues in some detail. The challenge, in brief, is to measure persuasion and (possible) preference change. 
My basic method is process-tracing, where one seeks to investigate and explain the decision process by which various initial conditions are translated into outcomes. I operationalize the method through use of three techniques.
First, I interview participants in contemporary policy debates, seeking to ascertain their awareness of emerging European norms on membership and citizenship (my focus in the project), and, more important, whether/why they comply with the norms' prescriptions. In all instances, I utilize a similar interview protocol, starting with the general, then moving to more specific questions. These are designed to tap an individual's basic beliefs about citizenship/membership and what might be motivating him/her to change them. On the latter, I give interviewees several possibilities, including both their own cognitive uncertainty as well as external social pressure. I suggest answers that address material incentives, as well as identity concerns.
In designing the interviews, I sought to capture both temporal and intersubjective dimensions. On the former, I have interviewed and then re-interviewed the same individual at two different points in time (so-called panel samples) whenever possible. This not only allows for establishing a degree of interviewer-interviewee rapport, which is crucial for in-depth questioning; it also enables me to assess the validity of interviewee accounts. (Were they consistent over time? If they changed, then why?) Intersubjectively, I asked interviewees "to step outside" their individual thought processes and characterize their social interaction context. Especially for those who had been in one-on-one or small group settings, I suggested four possible ways to portray the dynamics within them: coercion, bargaining, emulation, persuasion/arguing. Interviewees were then asked to rank order the various possibilities, and to consider whether their rankings changed over time.
Second, as a supplement and check on interview data, I carry out a content analysis of major media and specialist publications (international law journals, reports produced by the NGO community, for example). This allows for checking the beliefs and motivations of particular individuals who were both interviewees and participants in public debates.
Third, I consult official documentary records. For example, one of my cases in the larger project explores the supranational process through which the Council of Europe (CE) came to promote revised European citizenship norms; this included confidential meetings in Strasbourg (the Council's home). I therefore sought and gained access to official summaries of those sessions. 
While it ain't rocket science, these techniques -- taken together -- do multiply the observable implications of my approach and allow me to triangulate when assessing the degree to which, and through what mechanism(s), agent preferences change as a result of interaction (see also Aspinwall and Schneider 2000b, 37). This use of process-tracing along with a consideration of counterfactual explanations, where appropriate, allow me to minimize reliance on "as if" assumptions at the agent level. I thus shrink the black box surrounding the social interaction context.
The last comment raises an important issue. Empirically, can one really disentangle preference change driven by persuasion and social learning from strategic adaptation in the face of changed incentives, or from passive, cognitively simplifying imitation? One answer to this query is methodological. My use of multiple, process-oriented techniques allows for a reconstruction of actual agent motivations; equally important, they introduce a degree of cross-checking. A strategically dissimulating interviewee who was just "feeding me a line" about being persuaded would likely offer different motivations and justifications in other, more private or public settings. Put differently, consistency across contexts, or what has been called the "norm of consistency," is a strong indicator that an agent sees him/herself in a genuinely persuasive interaction (Elster 1991, 19-20). Likewise, a cognitive-miser/emulating agent should, across various settings, offer little substantive argumentation or reasoning in response to questioning, for emulation was simply an economic way of reducing uncertainty in the environment (DiMaggio and Powell 1991, 69-70).
A second response questions the role of assumptions in theory building. Ever since Robert Merton's pioneering work in the late 1940s, social theorists have argued that the middle-range frameworks of interest here can only be constructed by elaborating social mechanisms that shrink the gap between "input" and "output" (Merton 1968, chapter II; Hedstroem and Swedberg 1998, 7-9, 25). In turn, this requires one to minimize use of the "as if" assumptions that play such important roles in both rational choice and (more surprisingly) constructivist studies. As one theorist has argued:
the fact that we can construct an "as if" story in any situation to reconcile behavior to a self-interest explanation does not mean that self interest should be our default position either, unless we can establish that story is more compelling as an account of actual motivations than that offered by other theories (Hurd 1999, 392 - emphasis in original).
For sure, the choice-theoretic critique of those who study preference formation is a well-taken and cautionary reminder of the difficulties involved in the enterprise. However, criticism should not be allowed to become dogma, especially if one's concern is to model and explain the social world as it really works (Johnston 1998).
So What?/Who Cares?
Why should any of this be of interest to students of the EU, or to institutional theorists? Regarding the former, the answer is straightforward. For over three decades, scholars of integration have attempted to theorize and document how state elites, in insulated settings, may adopt multiple identities and in some cases redefine their interests through processes of social interaction within EU institutions. While early (neo-functional) work along these lines was flawed by serious methodological problems (Pollack 1998; Martin and Simmons 1998, 735-36), a new generation of research goes some way in correcting them (Beyers 1998; Egeberg 1999; Eriksen and Fossum 2000, among others). 
Despite these advances, much work remains. In particular, key concepts -- socialization, learning, deliberation -- remain frustratingly underspecified (Checkel 2000b for review and details). The result, methodologically, is a severe lack of clarity on how researchers are drawing valid inferences on the degree of preference change (or learning, or whatever) within European institutions. This is a pity, for the EU's densely institutionalized structure would seem an ideal laboratory and "social soil within which actors' preferences might be transformed" (Caporaso and Jupille 1999, 440). The approach outlined above is one possible way to rectify such problems.
What about institutional theorists? Why should they care? After all, there is not a whole lot of explicit institutional analysis in my approach; the focus is much more on agency. Instead, institutional factors are backgrounded -- for example, in H#2 (the normative and institutional context shaping the cognitive priors an agent carries into interaction), and H#5. However, my biases are intentional. Sociological institutionalism and its close relation -- constructivism -- will remain incomplete absent a greater balance between structure and agency in their accounts (Zucker 1988; Risse 2000, respectively; see also Aspinwall and Schneider 2000b, 9, 36, 39-40). Indeed, any move from meta-theory to substantive theory would seem to require just such a synthesis.
For sure, there is no such thing as a free lunch -- even in institutional studies of European integration. There are tradeoffs to consider if the subfield moves in the directions suggested in this memo. The middle ground I advocate -- of a contextualized understanding of agency based on a relational ontology, which nonetheless strives to be theoretically and methodologically self-conscious -- will never satisfy proponents of ontological and theoretical purity, be they hardcore rationalists or post-structuralists. The "ontological terrorists," in Stephano Bartolini's apt phrase, will not be happy (although see Economist 2000).
Yet, this middle ground does capture the reality of institutional dynamics within complex polities such as the EU. When my students read and discuss Schimmelfennig on rhetorical action and enlargement, Risse on arguing and bargaining and the end of the Cold War in Europe, or Checkel on manipulative/argumentative persuasion in European institutions, they often have two responses. A first is along the lines of "wow, these guys care about theory, but still manage to convey both complexity and nuance in their empirical stories." A second, more analytic response goes as follows: "Is Schimmelfennig [or, whoever] a rationalist, a constructivist, or both? We can't tell!" And those latter queries, in themselves, might be important signs of progress.
Aspinwall, Mark and Gerald Schneider, Editors. 2000a. The Rules of Integration: The Institutionalist Approach to European Studies. Manchester: Manchester University Press.
Aspinwall, Mark and Gerald Schneider. 2000b. "Same Menu, Separate Tables: The Institutionalist Turn in Political Science and the Study of European Integration." In The Rules of Integration: The Institutionalist Approach to European Studies, edited by Mark Aspinwall and Gerald Schneider. Manchester: Manchester University Press.
Beyers, Jan. 1998. "Where does Supranationalism Come from? The Ideas Floating through the Working Groups of the Council of the European Union." Paper presented at the Third Pan-European Conference on International Relations, Vienna (September).
Brody, Richard, Diana Mutz and Paul Sniderman, Editors. 1996. Political Persuasion and Attitude Change. Ann Arbor: University of Michigan Press.
Caporaso, James and Joseph Jupille. 1999. "Institutionalism and the European Union: Beyond International Relations and Comparative Politics." Annual Review of Political Science 2:429-44.
Chayes, Abram and Antonia Handler Chayes. 1995. The New Sovereignty: Compliance with International Regulatory Agreements. Cambridge: Harvard University Press.
Checkel, Jeffrey T. 1999. "Why Comply? Constructivism, Social Norms and the Study of International Institutions." Manuscript under review (20 October).
-------. 2000a. "Strange Bedfellows? IR Theory and Socialization -- A review of The Power of Human Rights: International Norms and Domestic Change." Comparative Political Studies (Forthcoming).
-------. 2000b. "Constructing European Institutions." In The Rules of Integration: The Institutionalist Approach to European Studies, edited by Mark Aspinwall and Gerald Schneider. Manchester: Manchester University Press.
Christiansen, Thomas, Knud Erik Joergensen and Antje Wiener, Editors. 1999. "The Social Construction of Europe." Journal of European Public Policy 6 (Special Issue).
DiMaggio, Paul and Walter Powell, Editors. 1991. The New Institutionalism in Organizational Analysis. Chicago: University of Chicago Press.
Economist. 2000. "The Future of Economics: In the Long Run, is the Subject Dead?" Economist (4 March).
Egeberg, Morten. 1999. "Transcending Intergovernmentalism? Identity and Role Perceptions of National Officials in EU Decision-Making." Journal of European Public Policy 6 (September): 456-74.
Elster, Jon. 1991. "Arguing and Bargaining in the Federal Convention and the Assemblee Constituante." In Rationality and Institutions, edited by Raino Malnes and Arild Underdal. Oslo: Universitetsforlaget.
-------, Editor. 1998. Deliberative Democracy. NY: Cambridge University Press.
Eriksen, Erik Oddvar and John Fossum, Editors. 2000. Democracy in the European Union: Integration through Deliberation? London: Routledge.
Evangelista, Matthew. 1999. "Norms, Heresthetics and the End of the Cold War." Paper presented at a Research Workshop of the Program on New Approaches to Russian Security, Boston (May).
Fierke, Karin and Antje Wiener. 1999. "Constructing Institutional Interests: EU and NATO Enlargement." Journal of European Public Policy 6 (December): 721-42.
Gibson, James. 1998. "A Sober Second Thought: An Experiment in Persuading Russians to Tolerate." American Journal of Political Science 42 (July).
Haas, Ernst. 1990. When Knowledge is Power: Three Models of Change in International Organizations. Berkeley: University of California Press.
Hedstroem, Peter and Richard Swedberg, Editors. 1998. Social Mechanisms: An Analytical Approach to Social Theory. Cambridge: Cambridge University Press.
Hurd, Ian. 1999. "Legitimacy and Authority in International Politics." International Organization 53 (Spring).
International Organization. 1998. "International Organization at Fifty: Exploration and Contestation in the Study of World Politics." International Organization 52 (Autumn).
Joergensen, Charlotte, Christian Kock and Lone Roerbech. 1998. "Rhetoric that Shifts Votes: An Exploratory Study of Persuasion in Issue-Oriented Public Debates." Political Communication 15: 283-299.
Joerges, Christian and Juergen Neyer. 1997. "Transforming Strategic Interaction into Deliberative Problem-Solving: European Comitology in the Foodstuffs Sector." Journal of European Public Policy 4 (December).
Johnston, Alastair Iain. 1998. "Socialization in International Institutions: The ASEAN Regional Forum and IR Theory." Paper presented at the Workshop on "The Emerging International Relations of the Asia-Pacific Region," University of Pennsylvania (May).
Kratochwil, Friedrich and John Gerard Ruggie. 1986. "International Organization: A State of the Art on the Art of the State." International Organization 40 (Autumn).
Lupia, Arthur and Mathew McCubbins. 1998. The Democratic Dilemma: Can Citizens Learn What They Need to Know? Cambridge: Cambridge University Press.
Martin, Lisa and Beth Simmons. 1998. "Theories and Empirical Studies of International Institutions." International Organization 52 (Autumn).
Merton, Robert. 1968 . Social Theory and Social Structure. NY: The Free Press.
Moravcsik, Andrew. 1998. The Choice for Europe: Social Purpose and State Power from Messina to Maastricht. Ithaca: Cornell University Press.
-------. 1999a. "Is Something Rotten in the State of Denmark? Constructivism and European Integration." Journal of European Public Policy 6 (Special Issue): 669-81.
-------. 1999b. "A New Statecraft? Supranational Entrepreneurs and International Cooperation." International Organization 53 (Spring).
Pollack, Mark. 1998. "Constructivism, Social Psychology and Elite Attitude Change: Lessons from an Exhausted Research Program." Paper presented at the 11th International Conference of Europeanists, February.
Price, Richard. 1998. "Reversing the Gun Sights: Transnational Civil Society Targets Land Mines." International Organization 52 (Summer).
Riker, William. 1986. The Art of Political Manipulation. New Haven: Yale University Press.
Risse, Thomas, Stephen Ropp and Kathryn Sikkink, Editors. 1999. The Power of Human Rights: International Norms and Domestic Change. Cambridge: Cambridge University Press.
Risse, Thomas. 2000. "'Let's Argue!': Communicative Action in World Politics." International Organization 54 (Winter).
Ruggie, John Gerard. 1998. Constructing the World Polity: Essays on International Institutionalization. New York: Routledge Press.
Schimmelfennig, Frank. 1999. "The Double Puzzle of Enlargement: Liberal Norms, Rhetorical Action and the Decision to Expand to the East." Oslo: ARENA Working Paper No.15, June.
Stryker, Sheldon. 1980. Symbolic Interactionism: A Social Structural Version. London: Benjamin/Cummings.
Underdal, Arild. 1998. "Explaining Compliance and Defection: Three Models." European Journal of International Relations 4 (March).
Zimbardo, Philip and Michael Leippe. 1991. The Psychology of Attitude Change and Social Influence. NY: McGraw Hill.
Zucker, Lynne, Editor. 1988. Institutional Patterns and Organizations: Culture and Environment. Cambridge: Ballinger Publishing.
Zuern, Michael. 1997. "Assessing State Preferences and Explaining Institutional Choice: The Case of Intra-German Trade." International Studies Quarterly 41 (June).
[*] Memo prepared for a workshop on "Institutionalism and the Study of the European Union," University of Washington, March 2000
[**] ARENA, University of Oslo
P.O. Box 1143, Blindern
0317 Oslo, Norway
Tlf: (47) (22) 85 56 97
Fax: (47) (22) 85 78 32
 Put differently, lacking theories of process, constructivists and, especially, sociological institutionalists offer suggestive correlational arguments.
 The adjective "soft" denotes scholars driven more by empirical puzzles than by the ontological purity of their arguments.
 Within integration studies, Schimmelfennig's "rhetorical action" -- the strategic use of arguments -- builds upon this manipulative, strategic understanding of persuasion. Schimmelfennig 1999. See also Fierke and Wiener 1999.
 Such questions will also be the main focus of an international workshop on "Socialization and Identity Change: Theoretical and Methodological Issues," to be held at the University of Oslo this coming June under the auspices of the European Union's Fifth Framework Program.
 On the legitimacy and feasibility of these three techniques, see Hurd 1999, 382, 390-92; and Zuern 1997, 298-302. Also see Moravcsik 1998, 77-85, on "hard" and "soft" primary sources.
 This work on the EU thus intersects with a long tradition of research in both IR theory and negotiation analysis that emphasizes the socializing affects of international organizations and institutions on the actors who participate in them -- Haas 1990; Chayes and Chayes 1995, chapter 1, for example.
[Date of publication in the ARENA Working Paper series: 15.05.2000]