No CrossRef data available.
Conclusions and Extensions: Toward Mid-Range Theorizing and Beyond Europe
Published online by Cambridge University Press: 18 October 2005
Abstract
This article reflects on three sets of issues raised by the research in this special issue. First it summarizes and critiques the core analytical claims and main scope conditions for socialization as developed by the authors. Then it examines how a critical socialization micro-process—persuasion—fits with a thin rationalist argument. I suggest that agents who are more deeply socialized may be more strategic in their behavior precisely because they are true believers. Thus the presence of strategic behavior does not undermine the possibility of persuasion. It may be an especially obvious consequence of persuasion. Finally, I examine how the authors understand socialization and identity change. Here I argue that the content of identity needs to be unpacked further into four dimensions—constitutive norms, social purpose, relational beliefs, and cognitive worldviews—in order to more fully test how much, and what type of, identity change occurs as a result of socialization. Finally, the article looks at how the findings in the European case might be extended to, and compared with, empirical evidence from other parts of the world, in particular Asia.Thanks to Jeff Checkel, Michael Glosny, and two anonymous referees for comments and criticisms.
Information
- Type
- Research Article
- Information
- Copyright
- © 2005 The IO Foundation and Cambridge University Press
The purpose of this article is to reflect on the core analytical claims that the other articles in this special issue make about the processes and conditions under which socialization of state agents occurs in Europe, and to ask how these arguments could be extended, clarified, retested, and transported to other types of institutions in other parts of the world. Frankly, this is a daunting task. At best I can offer some suggestions about how to push some arguments further and why it might be worth doing so. In the first section of this article, therefore, I summarize the authors' core claims about scope conditions under which socialization is possible, and suggest where these can be extended, compared, and clarified. The second section looks more closely at one of the core mechanisms that establish the added value of both this project and, more generally, the concept of socialization in understanding international relations (IR): persuasion. It suggests how a focus on persuasion actually may suggest the need for more focus on strategic, or “thin rationalist,” behavior. The third section looks at how the treatment of identity in this volume might be extended theoretically. The final section looks at issues involved in extending the claims of this collection of articles to other parts of the world.
Core Analytical Claims
This is a diverse set of articles. At first glance, there appear to be some fairly large differences in whether the empirical evidence does or does not show socialization inside international institutions. Some authors argue that socialization in perhaps its truest sense—the internalization of new group norms through persuasion and/or communicative action—can be found in European institutions, though under certain scope conditions. Others appear skeptical about such claims and suggest that adaptation to the preferred norms of an institution (or of the actors who control the institution) is strategic, optimizing behavior. These apparent differences are useful because when considered together they suggest a rather long and rich list of scope conditions under which socialization can be hypothesized to work and not to work. This is, perhaps, collectively the major contribution of this research. To date, with some more recent exceptions,1
there has been a tendency to view strategic-choice effects and socialization effects as “either/or.” An actor either adopts some set of norms deliberately so as to maximize some material payoff (for example, the economic benefits of inclusion in Europe; political survival in a government where others endorse these new norms), or the actor internalizes new values and attitudes semiconsciously through more or less hardwired responses to specific environmental stimuli in specific social-psychological environments, or through a reasoned reevaluation of the inherent normative goodness of old values. These articles collectively point out, however, that progress in understanding how actors change their interests and their behavior inside institutions will come from understanding the interaction of strategic behavior with social-psychological socialization mechanisms. This understanding is likely to come from so-called middle-range theories of socialization. These theories are both the most practical to test (because many have been developed in other subfields already) and the most likely to explain what IR theory has heretofore generally viewed as large macro changes in states as generalized wholes.My argument here is that the skeptics actually help to refine the scope conditions under which persuasion is more likely to be operative. I do not believe that the skeptics intend to show or conclude that persuasion socialization is conceptually or empirically inconceivable. Elsewhere Kelley notes, for example, that the scope conditions for this kind of socialization are narrow, and thus that this particular process is relatively rare.2
She suggests that persuasion is only likely when there is low domestic opposition to an institution's persuasion efforts. In her case—the convergence of domestic legislation concerning minority rights in newly liberated Eastern European states with European Union (EU) standards—the issue at stake is highly volatile, is contested, and touches on the identities of majority and minority populations. Linkage politics worked, persuasion alone did not. This is perhaps not so surprising. The arenas for effective persuasion, according to articles in this volume, are those in which intensive, face-to-face interaction is the primary mode of interaction. When national policies are the object of change, and when the political process involves massaging myriad interests of politicians and bureaucrats, this kind of small-group intensive interaction is unlikely to be the avenue of socialization in the first place. When the object of change is, however, the preferences, norms, and values of individual bureaucrats and key politicians, persuasion processes apparently do work.3See Lewis, this volume. The case in Gheciu, this volume, also fits this pattern, even though the objects of change are ultimately national policies (civil-military relations). Her dependent variable is effects on those participating in NATO/Partnership for Peace (PfP) seminars (her informants). The link from this variable to national policies is less clear (for example, to the Romanian Democratic Convention) and it would not be surprising if strategic incentives were also used by newly socialized players to change positions of others in the policy process—from bureaucrats to politicians and legislatures.
In this volume, Hooghe is also skeptical that persuasion-based socialization has much impact inside international institutions. For Hooghe, socialization within a state tends to trump international socialization because it occurs prior to the latter. Socialization within states, of course, does not necessarily undermine the internalization of pro-European norms. These norms may be inculcated at the national or subnational level through the internalization of pro-European ideologies or the development of a concept of identity that can incorporate both national and international allegiances. Hooghe's results suggest first that persuasion may in fact occur, but mainly at the national and subnational level (hence the internalization of political ideologies or national identities that support international norms). So the concept of socialization is still important in explaining outcomes. Second, she suggests that precisely because primacy matters, those who enter international institutions at a relatively young age are more likely to internalize the institution's norms than those older participants who were more socialized in national institutions. Whether youth is a conjunctural cause along with noviceness, or whether youth means a greater likelihood of socialization in pro-European norms at the national level, is unclear. But either way, the existence of socialization as a process that creates preferences is not disputed by Hooghe's research.
The Schimmelfennig argument is less clear, however, about the possibility of socialization. At first glance, Schimmelfennig rules out the possibility that persuasion and/or communicative action have any causal impact in his cases. It is possible, of course, that he has picked two exceptions that do not undermine “the rule.” But I think his cases implicitly reveal some interesting questions about the scope conditions for persuasion.
As one scope condition, Schimmelfennig implies that on issues related to identity change, where groups have developed deeply ingrained concepts of self and other, persuasion and communicative action are less likely to work. From some theoretical perspectives on identity (for example, social identity theory), this should not seem too surprising. In identity conflicts, ingroups are likely to be especially vigilant about the defection of members from the value systems that are deemed to characterize the ingroup. Internal dissent threatens the cohesion and positive valuation of the ingroup. “Deviants” will be considered threats to group norms, and there is considerable evidence that deviants are more often the targets of ingroup disciplining and coercion than devalued outgroup members.4
This makes the ingroup more hypervigilant about, and resistant to, perceived efforts by norm entrepreneurs to “persuade” members of the group to reevaluate their attitudes toward outgroups. In addition, ingroups are likely to interpret these efforts at large-scale persuasion to be exercises in blaming them for the conflict. Identity blaming—the problem is X because your group has Y characteristics (as opposed to the problem is X because individual Y has done Z)—is also associated with the polarization of views and with resistance to the message from outside norm entrepreneurs.5Romer et al. 1997. Granting that issues such as political rights in highly polarized identity environments are too sticky and too resistant to attempts to inculcate ideas that appear to threaten the dominant ingroup, this, of course, does not rule out the role of persuasion in inculcating the identity norms that make resistance to persuasive attempts to change these norms effective in the first place. This underscores the obvious, but not trite, point that socialization theories may help explain resistance to resocialization at time t + n. In other words, the absence of evidence of socialization at time t + n does not refute the phenomenon of socialization at time t. One does not have to immediately cede the floor to a strategic material argument in the absence of evidence of resocialization.
Relatedly, Schimmelfennig also hypothesizes that persuasion requires the presence of a “liberal-minded” interlocutor. This is a testable argument about persuasion. But it could be explored and refined further. One direction would be to ask what exactly is a liberal-minded interlocutor? Why are liberal-minded actors more susceptible to persuasion than illiberal ones, especially on issues related to identity?6
Moravcsik and others have noted that human rights norms (indeed any new norms) are more likely to diffuse among those whose values are closest to those of the norm entrepreneur. But stated this way, the hypothesis technically does not limit norm diffusion to liberal targets, only like-minded targets.
But this argument about liberalism, tolerance, and persuasion could also draw on older research on the relationship between authoritarian personalities, conservative ideology, cognitive simplicity, and resistance to new information. Put bluntly, authoritarian personalities and ideologies are likely to have simplistic cognitive constructs, to be more resistant to disconfirming information, and to embody low levels of trust toward others, all conditions that are not conducive to persuasion.7
See Feldman and Stenner, 1997; and Jost et al. 2003.
Another possibility is that liberalism, at least in its Anglo-American form, also stresses the authority of law and legal precedent. Here the effect of precedent goes beyond merely reminding actors of the legal costs of norm violation.8
Rather, in terms of persuasion, one hypothesis might be that legal precedent is authoritative, and authoritative sources of ideas are more likely to be persuasive than nonauthoritative sources.9See Axsom, Yates, and Chaiken 1987; Cialdini 1985 and 1987; Jorgensen, Kock, and Rorbech 1998; Mutz, Sniderman, and Brody 1996; Rule and Bisanz 1987; and Zimbardo and Leippe 1991.
Cialdini 1985, chap. 3.
This is a roundabout way of saying that buried underneath the claim that liberal actors are more susceptible to persuasion are at least three possible mechanisms—reason and respect for alternative evidence, the authoritativeness of legal precedent, and consistency.
There is always the possibility, of course, that Schimmelfennig's empirical work has fundamentally undercut the major premise of the project. He got it right, and most of the other authors did not. They have observed processes that are not really there or are irrelevant or secondary, accounting for little of the variance especially when placed in a “linear regression equation”-type model that includes strategic incentives. I believe it is premature to concede this possibility. The other articles provide fairly convincing evidence from a wide range of methods—interviews, participant-observation, case analysis, and quantitative analysis—that persuasion and social influence effects are apparent in these institutions to different degrees. One way of testing whether Schimmelfennig's claims undermine the findings of the project is to apply his methods/starting point to other cases, and vice versa. It might be worthwhile, then, to go over the same cases using different arguments to piece together the process. This might mean, for instance, taking Gheciu's dependent variable (convergence in attitudes and policies toward civil-military relations) and asking: “What does Lewis' work on intensity/quality have to say about which part of this process of change in this dependent variable?” or “What does Shimmelfennig have to say about the role of material (or political) rewards, or the political strategies that ‘socialized’ agents inside these countries used to push new policies through parliaments and bureaucracies?” Alternatively, a microanalysis of Schimmelfennig's cases might focus on whether there is any variation at the individual and small group level in the propensity for one identity group to accept the norm of protecting the interests of another identity group. Do hypotheses about face-to-face interaction, deliberative institutional environment, intensity, exposure, authoritativeness of messenger, and so on, work at the microlevel in his cases?
Having referred to some of the scope conditions that, together, the articles in this volume reveal, let me summarize some of the important ones and add questions and suggestions for extensions. These overlap to some degree with those identified in the Zürn and Checkel article, though we develop different criticisms and extensions. My first condition maps onto their “properties of institutions” condition. My second, third, and fourth conditions map onto, but unpack further, their “properties of interaction between socializing and socialized agent.” My fifth condition corresponds to their “properties of political systems.” My sixth condition is similar to their “properties of the issues or objects of contention.”
Institutional Design Features
There are characteristics, both formal and informal, of institutional design that contribute to socialization. As developed in the articles, these include the following.
Degree of agent autonomy
As Lewis notes about the EU's Committee of Permanent Representatives (COREPER), national representatives inside these meetings appear to have a fair amount of leeway to push forward, or retreat from, a national position. There appears to be a relatively weak norm of supervision by the “principal,” even though the permanent representatives are supposed to be state agents par excellence in establishing the agenda of ministerial meetings. This leeway allows representatives to present a position, knowing that it is unlikely to sell, and to back down without fear of repercussions from, or monitoring by, the principal. It is unclear how precisely this feature of the institution evolved. Additional hypotheses need to be developed about when an institution allows this kind of agent autonomy. It may be an issue-specific feature of institutions. Institutions in which the agenda is highly technical or on a relatively new topic, where the principal is unsure or less aware of what its interests might be, are more likely to witness this kind of autonomy. Of course, this autonomy may have little to do with institutional design or issue-area. It may be nation-specific. In one of Lewis' cases, the agent's (for example, the French ambassador's) autonomy came from his seniority over his foreign minister.
Norms of consensus, trust, reciprocity, and compromise
These norms help ensure that national agendas cannot be pushed so far that the institution logjams. In a sense, these norms presuppose that the legitimacy of the institution is so strong that state agents are unwilling to undermine it even on issues where they have strong and divergent state preferences. It is unclear, however, how these norms evolve from deliberate institutional design choices to start with. There is probably a relationship. For example, as one Chinese diplomat explained it to me in an interview in 1999, it took this official some time to adjust from the style of discourse in the United Nations (UN) Conference on Disarmament to the style of the Association of Southeast Asian Nations (ASEAN) Regional Forum. In the Conference on Disarmament, the primary arms-control treaty negotiation body of the UN, the language is often that of national interest, and the debating style is parliamentary. The purpose of the institution is to negotiate treaties, hence the discourse is generally about bargaining over the distribution of goods. The ASEAN Regional Forum (ARF) is less formal. It is a deliberative institution, and one in which the provision of information, in the form of Confidence Building Measures (CBMs), is one of its major purposes. Thus the language is more consensus-focused, the atmosphere more flexible. In addition, perhaps like COREPER, the norm of flexible consensus is quite strong. This means that chairs of ARF meetings (who must be from an ASEAN state) have a legitimacy that allows them to declare a consensus even when it is unclear one actually exists. State agents refrain from challenging a chair's authority. So one design-related difference may be distributive versus deliberative institutions. The latter is more likely to develop norms of consensus, trust, reciprocity, and compromise. At first blush, COREPER does not seem to fit this hypothesis because it is designed to set distributive-like policies for the EU, but also arguably has a deliberative function—to consider agenda issues for further discussion.
Organizational characteristics
A number of authors have pointed out that organizational features of the institution in which actors are being socialized may be important conditions for the effectiveness of the process. There are three features in particular:
- Organizational primacy: secondary organizations are less likely to
socialize actors once their identities are established by primary (usually
national) organizations (Hooghe). This suggests, however, that primacy
effects, rather than most recent experience (recency effects), are
critical in socialization. The literature is divided on this, however, and
it would be worth testing these assumptions further. In effect, what are
the scope conditions for this scope condition; that is, when does primacy
and when does recency matter in organizational socialization?12It may be, of course, that some degree of hybridity is the result of socialization in primary and secondary organizations, in which case, any follow-on to this volume will have to develop specific hypotheses about hybridization.13
In the cognitive and social psychology literature, the size of the temporal gap between first and last messages and the responder's recall is considered a scope condition for primacy or recency effects. Analogizing to organizational socialization, the amount of time spent by an actor in the primary and secondary organizations, and the amount of time between occupation in the primary and secondary organizations, may be critical to determining which has the greatest impact on some measurement of the actor's values.
Related to the issue of hybridity is the possibility that different institutions and institutional purposes can have interactive effects, whereby there is a cumulative effect from participation in multiple but related institutions. Hooghe suggests that “it is extremely difficult—nigh impossible—for an international organization to substantively shift the views of mature recruits.” Hooghe, this volume. However, in some instances agents are not participating in only one institution, but in a number of related institutions and subinstitutions as well. All this participation might have a cumulative effect. For example, the ASEAN Regional Forum only meets once a year. But its formal intersessional meetings occur a few more times a year, its related Track II activities (unofficial or nonofficial) meet even more frequently. All of these meetings generally have similar purposes, namely promoting ideas of cooperative security and multilateralism. In many instances the same people often show up to some or all of these activities. So in the case of Asian security institutions and their socialization effects, at least, one would have to take into account the myriad, related institutional environments in which officials are involved.
- Organizational function: the more compatible secondary sectoral organizations are with primary (national) sectoral organizations, the less likely there will be any resocialization—indeed the more likely the persistence of older values and attitudes. This is a variant of the argument in, say, the human-rights norms literature, in which actors with more compatible values are more likely to influence each other. The resocialization, however, is likely to be marginal because the actors share so many values in common.
- Organizational specialization: variation in the purposes of institutions will have different effects on the efficacy of socialization. Thus socialization effects could promote identification with a geographic space or with a problem-solving ideology, depending on the type of specialization in the institution. It would be worthwhile to develop more general hypotheses about whether these different specializations have systematic effects on socialization. Institutions that specialize in issues relating to identity and identity change (minority rights) may be less effective as socializing environments than other kinds of specializations, however.14
- Organizational identity: the institution needs to have a coherent
identity with a clearly linked set of normative characteristics. Without
this feature, the socialization effects of the institution will be
diluted, or indeed nonexistent. This point comes across in an important
comment by one of Hooghe's informants about obstacles to
socialization inside the European Commission: “national tensions are
transferred to the Commission, and that makes it impossible to have
everybody agree ex ante on common objectives. There are no common
objectives…. This is still a relatively young, expanding, and
maturing institution, which has not yet found its own identity.”15
Hooghe, this volume.
Style of Discourse and Shared Language
Lewis raises the interesting question of whether merely a shared way of expressing ideas may help socialize new members of a group. While he does not develop this point theoretically, there is certainly a large literature that claims, at its most radical, that language constructs reality, and at its less radical, that language shuts off or delegitimizes some kinds of arguments more than others.16
For the classics in the linguistic construction of thinking, see Whorf 1956; and Sapir 1956. See also Gumperz and Levinson 1996; Dalby 1990; Doty 1993; Ochs 1986, 2–3; Waever 1990; Searle 1992; Bourdieu 1991, 49; Yee 1996, 94–97; Fierke 2002; and Fairclough 2003, 21–24.
Mimicking is a mechanism whereby a novice initially copies the behavioral norms—including discursive practices—of the group to navigate through an uncertain environment (in this regard it might be considered along with Checkel's notion of cognitive role playing and Zürn and Checkel's discussion of cognitive dissonance as part of a process of cognitive short-cutting that socialization can induce). But mimicking differs from rational strategic adaptation. It is a somewhat efficient means of adapting to uncertainty prior to any detailed ends-means calculation of the benefits of doing so. That is, mimicking, unlike strategic emulation, does not mean searching for and copying exemplars, or the most successful actor in the group. It is rather a satisficing first step designed simply to be able to participate in the group by following its most basic rules, even before the actor has a clear sense of what its interests might be that need maximizing.17
To some degree, the concept complements Checkel's notion of “cognitive role playing.”
Mimicking can be reinforced by status concerns. That is, if one does not join an organization, then one may not just suffer a loss of material benefits, but may also be viewed by others as out of fashion, behind the times, and thus missing out on a status-enhancing experience. This can lead to choosing behaviors that may not be beneficial if the actor only behaved according to private information (because the private information is discounted in an uncertain environment).18
Hirshleifer 1995, 191.
To be sure, mimicking stretches somewhat the concept of socialization, because pro-group behavior is only indirectly an effect generated by the nature of the social environment. Rather, it is a survival strategy in a particular social environment. That is, while mimicking is distinct from exogenously induced threats or punishments, and is not characterized by individual efforts to optimize long-run material well-being, it is not in the same class of causes as persuasion and social influence. The latter two are mechanisms that motivate. In mimicking, the mechanism that motivates can be survival under uncertainty.
On the other hand, choosing which groups to mimic involves a degree of prior identification. Moreover, mimicking prosocial behavior can lead to the internalization of norms through repetition. Or, alternatively, by mimicking an actor goes on record as behaving in some particular way. It may then be loathe to deviate from this precedent for status and image reasons. Many of the procedural constraints, work habits, and standard operating procedures, including discursive practices, that actors develop to minimally function inside an institution come from mimicking the behavior of others in the group. These can then limit legitimate forms of participation in the institution.19
See Holland 1996, 29; Cialdini, Kallgren, and Reno 1991, 203–4; Hirshleifer 1995; and Biddle 1985, 162.
James notes that communicative action involves agreement on the medium through which discussion of conflict takes place. This agreement—like learning a common foreign language—comes before all other communication—see James 1998, 7–11. Hence socialization, whether via persuasion or social influence, requires a basic level of communicative cooperation to begin with. I do not disagree with this argument. But there is no reason why this basic agreement need be achieved through discussion and argumentation that two or more equal, empathetic actors eventually find convincing. Mimicking the language, procedures, and habits—the parameters of this medium of communication—can lock an actor seeking to survive under uncertainty into such agreement as well.
Related to mimicking is the question of behavioral templates and how these affect actor behavior. Gheciu implies that much of the “teaching” or persuading in her cases appears to involve agents from the North Atlantic Treaty Organization (NATO) offering legal, organizational, and normative templates for reorganizing civil-military relations. So what is doing the socializing—the way in which these templates are offered and discussed by NATO? Or the templates themselves? There is perhaps an over-determination problem here: when a “teacher” says, “this is the way you should do X, if you have identity Y” (normative persuasion) and also says, “here is a method for acting Y-like” (template), it may be unclear which has the greatest impact on pro-group behavior.21
One of the U.S. Coast Guard's jobs is to provide templates for developing countries to set up coast guards. These templates include draft legislation governing the surveillance of territorial waters. For a lot of small littoral states, their coast guards are their navies. So the U.S. Coast Guard literally provides models for setting up claims to being an ocean state with a navy. In Gheciu's case, then, is it the face-to-face process of teaching or the existence of templates that is primarily responsible for eliciting new norms of behavior?
Exposure
Both Lewis and Gheciu raise the question of the intensity, duration, and quality of exposure to counterattitudinal messages as critical variables explaining the success of persuasion. There is some dispute over precisely how intensity or duration are independent of other elements of exposure, such as the autonomy of the agent, the authoritativeness of the persuader, the content of the message, and the cognitive characteristics of the persuadee, among other possibilities.
The quality of exposure is also somewhat illusive in meaning. But there appear to be at least three subelements to the “quality” variable. One is the consistency of the message. Another is social proof. A third is a focus on consistency between verbal commitment and action, a consistency that would highlight the hypocrisy of noncompliance with new norms. These latter two elements, however, imply a degree of optimizing, “logic of consequences” calculation by the target of the message. That is, as the literature on consistency suggests, there can be two ways in which actors decide to conform to a norm. In the first, actors compare the required behavior against a list of behaviors considered appropriate to the identity the actors believe they hold. Consistency theory suggests that it is the internal cognitive discomfort that hypocrisy generates that leads to changes in behavior. The second possibility is that conformity arises from calculating the relative status benefits (back patting) and status costs (shaming, opprobrium) of acting in ways consistent with the new norms.
How these more detailed micro-processes might be tested is a tough empirical problem. It probably requires moving more systematically into the details of social group interaction where individual and personality type are the critical units of analysis. There is likely to be a certain distribution in the sensitivity to appeals to consistency and social status across individuals inside institutions. This distribution will determine whether there is large- or small-scale compliance, and whether there are obvious tipping points in the group as to how many others have to first demonstrate compliance before the rest of the group does.22
This draws from the threshold model of social action in Granovetter 1978, 1420–43. The empirical problem is measuring the distribution in willingness to comply before exposure to the institution and its normative message. Modeling this process would also require taking advantage of agent-based modeling techniques.
Noviceness
In Gheciu's article, the noviceness of those being exposed to counterattitudinal messages (to whom she refers colorfully to “baby generals”) is a critical scope condition. Hooghe finds the same thing, except that noviceness combines with youth to encourage socialization for officials in the European Commission. This is believable because these are actors most likely to experience “primacy effects,” in which early experience and information will have out-of-proportionate effects on inferences drawn from later experiences and information.23
See Choi 1993, 52–53.
See, for instance, the argument in Chafetz, Abramson, and Grillot 1996. The authors argue that the debate in the Ukraine in the 1990s over whether to sign the Non-Proliferation Treaty (NPT) was in large measure a debate over whether Ukraine's identity was that of a great power (hence it could legitimately keep and develop nuclear weapons), or a middle European power (in which case it should denuclearize and join the NPT).
But while novices may be more susceptible to persuasion and communicative action, this is not the only route to socialization. Novices are also more likely to mimic. As I noted above, while it is strategic behavior, mimicking is not necessarily optimizing behavior. At one level it is survival behavior, doing what is minimally necessary to survive in a new environment by copying what most other actors are doing. At another level, however, it can be behavior that maximizes self-valuation by demonstrating competence in a new environment. Some literature on organizational culture argues that actors have a desire to appear to be competent in their own eyes. This is done by adopting the trappings of competence in their field of endeavor—the discourse, symbolic practices, and even physical appearances that are presumed to indicate competence. This is sometimes called auto-communication.25
Thus mimicking and auto-communication may feature more prominently in the conformist behavior of novices than in nonnovice actors inside an institution.Domestic Variables
One of the important features of this project on European socialization is its sensitivity to the domestic or national political constraints on the socialization of state agents inside international organizations. Clearly the pace, process, and outcome of socialization inside one environment will be affected by socialization in a previous environment. Actors are usually not blank slates when they enter new social environments, though the degree to which “fresh and beautiful characters” can be written on new entrants will vary.26
To paraphrase Mao Zedong's comment about peasants being similar to blank sheets of paper.
National socialization versus international socialization
A number of articles (Beyers, Gheciu, Hooghe) raise an interesting, although understudied question: Why is it that national allegiances and/or allegiances to national-level organizational and bureaucratic interests remain strong in the face of supranational socialization processes? Beyers invokes three institutional reasons connected to socialization in national bureaucracies as well. Yet the answer is not as obvious as one might expect. After all, there are powerful discourses inside some national bureaucracies in many of these countries in favor of further integration, and in favor of the notion of “European-ness.” Moreover, many of these individuals spend considerable time away from national and organizational environments when they operate in European bureaucracies.
So what is it about these particular national levels of allegiance that seems so powerful? An answer to this question might suggest why the project finds a “European identity deficit.” In principle there are a range of hypotheses (not all of which will apply to European states today, though this will depend on the demographics of some state agents in European institutions)27
Some of these are raised in Hooghe, this volume.
- Primary and primacy effects: state agents' first and most intensive period of socialization occurs inside the main institutions of state socialization (for example, education systems) and from exposure to the myriad symbols of state sovereignty—national anthems, passports, political participation in national rather than supranational collective expressions of a will (voting), and so on. Thus it should not be surprising that the strongest allegiances are to the state and its definitions of interest.
- Intensity effects: socialization by institutions at the state level is more powerful because of the time spent in them and by the often zero-sum presentation of national allegiances (one cannot be both patriotic and internationalist).
- Boundary policing: it should not be surprising that policing ingroup
identity is easier whether at the national and/or national
organizational level or at the supranational or international
organizational level. It is common in most types of ingroups to have
formal and informal methods for punishing those with excessively liminal
identities.28These mechanisms can include discourses that draw clear lines between the traits that the ingroup is expected to exhibit, in contrast to those of the outgroup. These discourses are signaling devices: those who do not exhibit these traits or who exhibit hybrid, liminal traits, will be exposed to peer pressure, physical coercion, threats, sanctions, and so on. Different groups have different membership rules.29 The concept of treason, and its legal, economic, political, and personal sanctions, is a long-standing policing tool used against liminal identities at the national level. Different states have different definitions of treason. But in general, many have come to expect violations of national allegiances to be among the least easily forgiven of many kinds of social identities, and the kind of violation that brings down the power of the state on the heads of individuals and groups.30
See Burke and Tully 1977, on the effects of sex stereotyping on ostracism and bullying.
In contrast, in most liberal democracies defection from what in the past were quite rigid identity groups—for example, religious allegiances, sexual orientations, and class—does not carry these sorts of state-sanctioned penalties, at least not as much as it has in the past.
But this raises an implication for this project, underexplored in the articles here. Perhaps one of the reasons for an underdeveloped supranational identity is that European-level mechanisms for policing identity are more or less absent. There are no powerful European-level institutions, for instance, that “punish” agents for holding liminal or cross-national and cross-regional identities.
This raises another possibility. Perhaps there is little policing by either national institutions or European institutions of the identities of those who participate in supranational organizations precisely because no contradiction is seen between having memberships in both groups. This is perhaps the true innovation in European identities—a discourse that reduces the incompatibility between the national and the supranational levels of identity. Indeed, arguably the essence of Western Europe's liberalism is that multiple identities are legitimate. This leads to the proposition that the key identity change for most state agents inside European institutions is not the emergence of a supranational identity as much as the emergence at the national level of identities that do not view a supranational identity as a particularly competitive “other.”
There is a third possibility (raised by Hooghe's article), namely a hybridization process whereby national and international socialization, combined, produce variants of each level of norms. Acharya calls this “localization”: norm-takers “build congruence between transnational norms—including norms previously institutionalized in a region—and local beliefs and practices.”31
Acharya 2004, 241.
The state-society relationship
Schimmelfennig suggests that when populations are weakly mobilized (because of underdeveloped party systems), political elites are under less bottom-up pressure to conform to demands for normative changes from other states or international institutions. Transnational mobilization is less likely on these sorts of normative issues. By default, then, material incentives and disincentives are likely to be more effective in compelling political elites to change domestic policies. Moreover, variation in compliance through socialization will also depend on how liberal or illiberal these political elites are to begin with. It would be worth exploring these hypotheses further, however, because I think the causal arguments could go in a number of different directions. Mansfield and Snyder argue, for instance, that in newly democratizing states with underdeveloped party systems, the population can actually be overmobilized, particularly through appeals to nationalism.32
This can, over time, constrain political elites (via blowback in nationalist ideologies). Following this line of argument, the reason for resistance to nonmaterial incentives for normative conformity may have more to do with society-level resistance to socialization attempts than political elite resistance. Conversely, a weak society, and low political mobilization, may give political elites the political space to endorse and push through new norms (embodied in legal and political reforms, say) that are being promoted by outside actors.33See Risse-Kappen 1995.
Issues
The articles in this volume look at socialization in a wide variety of issue-areas in which European institutions have some jurisdiction—minority rights, civil-military relations, domestic elections, EU agenda setting, and so on. As I noted, the evidence for socialization, especially of the persuasion/communicative action kind, looks weakest on identity and ethnic conflict issues. Yet if Gheciu is right, it seems that these socialization processes worked quite well when it came to something as fundamental to political reform inside states as how the coercive arm of the state relates to the political arm (for example, civil-military relations in Eastern Europe). So it would seem worthwhile to categorize issue-areas more systematically according to their hypothesized effects on socialization.
In a sense, the articles in this special issue together provide a rich list of scope conditions, a list of potentially relevant independent and intervening variables that could be tested on a shared set of dependent variables. While this project is important precisely because of the wide range of socialization phenomena it is trying to explain, pushing the causal hypotheses further would require one or both of two different projects: (1) a project in which these hypotheses and scope conditions were more systematically retested on the dependent variables in the various articles in this volume (for example, testing for Gheciu's micro-processes on Schimmelfennig's dependent variables); and (2) a project that tests each of these independent and intervening variables on a shared dependent variable.
This latter type of project would have to be clearer about what the relevant dependent variables were for testing the effects of socialization processes. It is clear that socialization research is at an early stage. Like the early stage of contractual institutionalist research, the goal is more or less to show that it happens, that it has ontological validity, in a sense, and, somewhat more vaguely, that it “matters” because one is seeing changes in preferences and interests that could not be explained otherwise. But, as with institutionalist theorizing, the next step is to show how much it matters and why. By definition, then, the next stage will involve looking at a range of different dependent variables to see how far socialization extends into different arenas and types of human political behavior. Thus there is nothing wrong with socialization theorizing dealing with a range of different dependent variables. I can think of several discrete dependent variables and one highly complex dependent variable (more properly, an “emergent property”).34
All of these, of course, could at some point be independent variables as well.
The list of discrete dependent variables might include the following (some of these come from individual papers):
- The effects of socialization on the identities and preferences of individual, human agents/actors. This is essentially the dependent variable of many of the articles in this volume.
- The foreign policies of states (in which case the link between individual socialization, mediated by state foreign policy processes, and state policy becomes a critical variable). In a sense, Schimmelfennig looks at this dependent variable.
- The evolution of the international institution itself—its organizational structures, normative structures and ideological message, and its performance as an agent in the socialization of human actors.
- The quantity and quality of interstate cooperation and integration in some functional area.
- Variation in regional organizational/normative structures around the globe.
- Variation in the institution's capacity to channel or focus
“bottom-up,” social-movement pressure on states.35
For instance, UN summit meetings in the 1990s were increasingly designed to allow nongovernmental organization (NGO) pressure to be placed on state delegations. The ASEAN Regional Forum and its NGO Track II partner, the Council on Security Cooperation in the Asia-Pacific, however, are highly elite-based, with no avenues for mass “peace and security” NGOs to participate.
The “emergent property” captures, basically, how the integration of all these discrete variables affects agent-structure constitution across time and space. That is, perhaps the end point of this research should be to close the loop that connects institutions-as-normative-agents vertically to the teaching and learning at the human agent level within states, and from there to the effect of horizontal interaction of multiple agents at the single state level, and from there to the creation of “hybrid” identity/interests in particular states, and from there to the effects of these new hybrid interests on interstate relations, and from there to the effects of these interstate interactions back to the organizational and normative structure of institutions at time t + n.
Many of the articles in this project are focused on how participation in institutions changes identities, interests, normative models, notions of appropriateness, and so on, in individuals as national agents. In a sense, this project thus picks up where Finnemore's “teaching” left off and looks at the kinds of resistances (individual, organizational, ideological at the state level) to teaching inside institutions. This is important, but it is only one part of the entire structure-agent relationship that is of interest to the sociological turn in IR.
Actually conducting empirical research that captures the “emergent property” effects of socialization within institutions is, of course, the hard part. In principle it would require, I think, large-scale cooperative projects where the life cycle of one institution (and its normative structure) is followed for a few iterations. Individual researchers would look at the institution and its ideology at time t; the institution's impact on socializing state-level agents at t + 1; state-level agent responses at t + 2; state-level agent interaction within a state's foreign policy process at t + 3 (agent-principal interaction); interstate interaction at t + 4; state-institution interaction at t + 5; and changes in institution and ideology at t + 6; with this process repeated over a few iterations.
There are costs for untenured scholars to engage in this kind of collaborative work. It would be important, I think, to present to the IR theory community a bold, long-term research plan showing precisely where such disparate studies as work on socialization in security ideologies in Eastern Europe and work on the Europeanization of a Belgian bureaucrat fit into a project on the mutual constitution of agents and structures.
The Question of Persuasion and ‘Thin Rationalism’
As I noted earlier, persuasion is, in a sense, the trump card for demonstrating the added value of socialization approaches. This is because, if shown, persuasion suggests that preferences and identities do change and that they change in the absence of material incentives and disincentives. This does not undermine logic of consequences arguments about identity or preference change; it only establishes that there are places where argument and emotion help explain changes in actor motivation. But persuasion may also be fairly rare. Obviously at some point agents have to internalize a set of norms and values. It just is not clear that the independent effects of persuasion as opposed to consequentialist calculus can be separated out all that neatly, even if one adopts the micro-focused analysis that some of the articles do here. Checkel suggests that the articles in this volume show the value of a “thin rationalist”–constructivist synthesis.36
Checkel, this volume.
But I think it is also important for socialization theorists to point out that the thin rationalism they have in mind is in many respects fundamentally different from that used by more mainstream institutionalism. Contractualism has almost entirely neglected social thin rationalism in favor of material thin rationalism. Yet when one speaks of “normative pressure” in IR, one is talking about conformity for rational reasons but under conditions that are entirely “social.” The “sociality” of this thin rationalist calculus comes in two forms.
First, actions (conformity to group norms) designed to maximize social status markers may be a “consequentialist” choice, but such a choice can only be made in a social environment in which members of a community are rewarding appropriate behavior with status markers. There has to be some common understanding between the actor and the community about the value of these markers as well.37
I outline the social influence argument and the role of social back patting and shaming in more detail in Johnston 2001.
Second, because actors tend not to engage in audience shopping—they tend not to easily shift loyalties from group to group depending on which group maximizes status markers—this suggests a certain stickiness that is not captured by the materialist thin rationalism of contractualism. This stickiness is likely rooted in a sense of obligation to the social group or, perhaps more correctly, in at least some degree of identification with the values of the group. Otherwise, as noted, one would expect little stickiness in the search for status maximization. Thus, in the case of status maximization at least, the consequentialist calculus is still fundamentally different from side payments or sanctions of a material form provided by one actor or one institution acting as an agent. Expressed differently, social status cannot be a private side payment as technology transfers can be. Social status has to come from a community, not an autonomous privileged or powerful actor on the side. One could get a gold star from a friend for one's homework. But this will have no status value unless the star is bestowed publicly and comes endorsed by a valued social group. Moreover, one has to have some modicum of identification with this social group so that one does not go off looking for other groups that will give one larger or more golden stars (to press the metaphor perhaps too far!).
Another issue related to the status of “thin rationalism” in the study of socialization concerns strategic agency. No research project can do everything at once. But any follow-on to the agenda set by this volume might explicitly test for both the amount of strategic agency in the socialization story and the amount of unintended socialization in any strategic agency story. For instance, one might actually expect deeply socialized actors to be especially strategic in their behavior. Fundamentalists and true believers often have a particularly amoral approach to strategy, given their strong beliefs in the correctness of their cause and in the utter lack of legitimacy in the cause of others. At the most extreme, if one believes that the outgroup is devoid of moral capacity and that it thus threatens one's existence, then any and all means are justified in dealing with the outgroup. I am not suggesting that true-believer European supranationalists are Leninists when it comes to tactics and strategy. I am simply suggesting that the more deeply committed one is to a principle, oftentimes the more likely it is that one will behave strategically in an “ends justify the means” way. One might actually expect to see a U-shaped relationship, with degree of strategic calculation on the y axis, and true opportunism to true belief on the x axis. One sign of both shallow and deep socialization in new values may in fact be highly strategic behavior designed to promote these values. Therefore, strategic behavior, per se, does not necessarily undermine the possibility of prior socialization.
The Question of Socialization and Identity Change
The “strong” or maximum constructivist claim about pure socialization is that persuasion leads not just to change in causal arguments, or even principled ideas, but in identities and then in interests. Often the invoked “theory” of identity that links identity change to behavior is role-theoretic, that is, actors literally act out behaviors considered appropriate to the new roles they have internalized. While the articles in this volume go farther in delving into the mechanisms behind socialization than any study of which I am aware, I think it is important to push further into the mechanisms of identity change and normative conformity. This requires more specificity, first, about the logical range of identity change possible. The articles as a whole suggest that actors, after exposure to an international institution, could discover that:
- Their primary identification is with the institution's interests first, and with national interests second.
- There is no fundamental conflict between identification with the institution's interests and with national interests, as was believed before.
- Their primary identification is with a new ideology and its associated practices.
Next, one would want to ask: what are the mechanisms that link particular institutional designs with particular types of identity change? As Checkel notes, for instance, the duration and intensity of exposure to an institution may actually lead to two distinct routes to identity change.38
Checkel, this volume.
Cialdini 1985, chap. 3.
There are at least three issues here. The first is what kind of identity change one should expect from participation in international institutions. Unpacking the content of identity, one could argue that identities are composed of four somewhat overlapping components or dimensions: normative or constitutive beliefs and practices, beliefs about the social purposes of the group, beliefs about the group's relationship to other groups, and cognitive processes or worldviews.40
Which dimensions of identity content is socialization in institutions more likely to change? Which behavioral practices do actors believe are constitutive of group identity? How do actors define the purposes of the extant ingroup? The entativity of other social groups with which the agent interacts? Or the cognitive models with which the agent understands how the world works? Gheciu's research suggests, for instance, that socialization mainly involves changing constitutive norms, and perhaps cognitive worldviews. In Hooghe's case, one of the main dependent variables is preferences about supranationalism—in essence, peoples' beliefs about how the EU should be constituted. In Beyers' case, the focus on role identity combines elements of change in constitutive norms and social purposes. In Lewis' case, identity change inside COREPER appears to mean a change in constitutive norms and in cognitive worldviews. But one could also imagine that a group's constitutive norms, social purposes, and cognitions may remain static, while beliefs about who the outgroup is could change. One could also imagine a change in constitutive norms without a change in social purpose or in perceived outgroups. The point is simply that if the socialization literature continues to work with a relatively unproblematized and blunt notion of identity, it is possible that null findings for one element of identity (such as constitutive practices) might obscure positive findings for some other elements of identity (a Type II error). One would thus run the risk of miscoding the effects of institutions on actors inside of them.The second issue concerns the theories of action that link identity to behavior. Role theory is not the only way to think about the link between identity and action. Social identity theory provides a different theory of action. The authors in this special issue use different theories of action. Beyers uses a role-theoretic notion of identity. Gheciu implies some kind of social identity theory in which identity expresses taken-for-granted characterizations of self and other. Yet these are two different theories with two different kinds of empirical evidence that “identity” matters. Role theory stresses the relevance of the content of identity to the kinds of actions considered appropriate. Social identity theory (SIT), in its basic form, downplays the importance of the content of identity in favor of the competition-inducing effects of having an identity, regardless of its content or characteristics. The former basically proposes that norms of appropriate behavior provide notions of self-efficacy and self-esteem. The latter suggests that evidence of identity change would involve changes in conceptions of “ingroupness,” new definitions of boundaries, changes in self-valuation as new allegiances are created, and action motivated by a desire to maximize group valuation.
So, if one focuses on how socialization affects identities and, in turn, behavior, ideally both the role-theoretic understanding of this relationship and SIT-based understandings should be tested against each other. It may be that SIT is actually empirically wrong when applied to European institutions. But it seems to be a powerful argument when one looks at the effects of socialization in other kinds of institutions (national, military, clubs, gangs, and so on). One wonders what it is about European international institutions that makes the effects of induction into their ideological and normative environment best captured by role-theoretic and not SIT-based processes.
The third issue has to do with the methods for observing identity change. The conclusion is this: regardless of whether one is conducting interviews, surveys, or textual content analysis, the trick is to isolate the characteristics that respondents believe are central to the group identity they believe they have, the intensity with which they identify with the group, the degree of difference they believe exists between own-group characteristics and other-group characteristics, and the appropriate behaviors associated with these different identity memberships.
Showing appropriateness is an exceedingly hard thing to do. As it is used in constructivist literature, appropriateness can mean at least three different things. First, it can mean that behavioral alternatives were never actively considered by agents with one identity, while agents with another might actively consider them. If one were using cognitive mapping, this would mean that because of internalized notions of appropriateness, entire cause-effect relationships would simply not show up in the maps because the actor was not consciously thinking about alternative ways of achieving some desired end.
Second, appropriateness could mean that behavioral alternatives are considered for a moment but are heavily discounted because they are considered useless or normatively objectionable (or they are considered useless because they are normatively objectionable).
The third possibility is that a behavior is chosen because it is considered appropriate for maximizing social rewards. The actor knows from experience that there is a severely limited range of behaviors that will be rewarded by the group socially; thus, alternatives are never considered. Behavior is automatic and optimizing at the same time.
These three different possibilities basically capture the difference between deep socialization/internalization (taken-for-grantedness where actors are simply unaware of their habits of group conformity) versus somewhat deep internalization (appropriateness where inappropriate options are consciously rejected) versus social influence.41
This is somewhat similar to Hopf's distinction between logics of habit, the logics of appropriateness, and the logics of consequences. See Hopf 2002. The main difference is that at least one manifestation of socialization—social influence—works only because the influencer and the influenced view the bestowal and acceptance of status markers in instrumental ways. The influencer uses them to extract compliance; the influenced uses them to satisfy a psychological need for status recognition. In this regard, I would agree with Checkel, who notes that socialization mechanisms should not a priori exclude elements of strict rationality and bounded rationality. See Checkel, this volume.
Hopf 2002, chap. 2.
One final comment about identity: Does one really need to show that supranational loyalties have emerged to show that socialization in institutions has been effective? I am not a Europeanist, but my sense is that among the ideological messages that European institutions are trying to promote, particularly in the East, are liberal values of tolerance and diversity. This means, I assume, that evidence of a changed identity could also include new understandings of national identity, not just supranational identity. Identification with “European-ness” could be manifest in acceptance that multiculturalism is a legitimate part of one's national identity; a strengthened commitment to one's local democratic institutions. This point is taken up by Beyers. There are two components to this argument, however: (1) that changes in national identity come from intensive participation in European institutions; and (2) that changes in national identity come before participation in these institutions—that national-level policies of integration actually change national-level conceptions of identity, such that European supranationalism becomes compatible with national identity. Indeed, some of the new work on nationalism argues that national identity emerges from and/or alongside local identities that accommodate national allegiances as a natural extension, rather than the elimination, of local identities.45
One way of getting at this question of whether national-level socialization is incompatible with the emergence of supranational identities would not be to look for evidence of socialization in supranational values, but to look for evidence (or the absence) of national policing against supranational values. As noted earlier, social-psychological work on how ingroups harass liminal identities suggests that powerful, though indirect, evidence of identity change is whether the ingroups detect and punish, police, or discipline these people. So, too, national bureaucracies may be alert to the possibility of supranational identity formation. These bureaucracies should only be so if they perceive there to be a tension or competition between the two. If there are no such efforts, then perhaps this means there is no perceived contradiction between the two levels of identification.
Empirical Extensions Beyond Europe
One could argue that European institutions constitute a relatively easy case for showing the processes and effects of socialization. The conventional wisdom has it that they are the most complex, most legalistic, most bureaucratized, most focused on the active inculcation of liberal values across national boundaries, and so on. If so, do the findings suffer from a sui generis problem? On the one hand, it would seem obvious that this is potentially a major problem, because nowhere in the world are interstate relations so bound up with international institutional commitments. The degree of European integration at the interstate, substate, and nonstate levels creates an overdetermination problem, depending on one's dependent variable. If the dependent variable is, for instance, “identification with European-ness” or “preference for multilateral solutions to national problems,” there are so many routes through which individuals, even in government positions, could be exposed to the ideology and identity of “European-ness” (personal contact; travel; national education systems; experience in the EU; experience in those portions of the national bureaucracies designed to interact with the EU, the media, and so on, the intensity and extensiveness of which exceed any other region in the world except perhaps North America), that it would be hard to disaggregate the separate effects of all these possible socialization experiences at the microlevel. In addition to these multiple channels of socialization, the low level of interstate security threats suggests that identity socialization at the national level will be a weaker counterweight to supranational socialization than in other parts of the globe.
On the other hand, this sui generis feature of the European case could help generate at least one powerful theoretical argument. That is, precisely because Europe, compared to other regions and their institutions, may be an “easy” or “most likely” case for socialization arguments, if socialization arguments, in essence, fail in the European case, this is a fairly telling blow to the whole theoretical approach. If not in Europe, where?
Because socialization arguments, for a number of the authors in this volume, do not fail in the European context, the obvious question is, where else might they work? Unfortunately the field of comparative regional institutional studies is almost nonexistent in IR. Few specialists in the international relations and international institutions of different regions of the globe have sat down to compare whether socialization processes are limited to the “rich” institutional life of European interstate relations. This cross-regional dialogue has just begun in the academic community with a project on “Crafting Cooperation: The Design and Efficacy of Regional Institutions in Comparative Perspective,” a project jointly sponsored by Harvard University and the Institute for Defence and Strategic Studies in Singapore. The project brings together people who specialize in both IR and area studies to examine similarities and differences in Asian, European, Latin American, African, and Middle Eastern regional institutions. It is too early to tell whether any systematic cross-regional differences exist, in part because in some regions institutional design is dynamic. Southeast Asian institutions (ASEAN ARF, for instance)—widely viewed as deliberately “underinstitutionalized” and based on informal, elite-based consensus-driven deliberation on issues that exclude domestic governance questions—may be changing in more intrusive and bureaucratized directions. Some Latin American institutions are as formally complex as some European institutions, yet the practice of solving disputes inside them often tends to rely on informal elite interaction.46
On the ASEAN Regional Forum, see Tan et al. 2002. On Latin America, see Domínguez 2002.
My first reaction, as an Asianist, to the question of whether there might be useful comparisons or extensions of the arguments in this volume to institutions in Asia, was envy. First, there is virtually no one doing work that is similar theoretically or empirically to this project on European institutions. The best work in this regard is being done by Acharya. It is hard for one individual, however, to go into the level of empirical detail and to employ the range of theoretical and methodological tools that, combined, these articles provide. There is, as far as I am aware, no systematic effort to examine how intensity or quality of exposure to ASEAN or the ARF, for instance, has affected the attitudes and allegiances of key participants, or how issue-area or institutional design may have a systematic effect on the propensity of actors to endorse values distinct from their national government. The systematic study of institutions and socialization in the Asia-Pacific has been underdeveloped.
My work focuses on the mechanisms of mimicking, persuasion, and social influence, with an empirical focus on China's participation in international security institutions such as the ASEAN Regional Forum, the Comprehensive Test Ban Treaty negotiations, the Conference on Disarmament in Geneva, the Protocol II of the Conventional Weapons Convention, and the Ottawa Treaty on antipersonnel landmines. It has taken me about 120 interviews over about eight to ten years to gain a partial understanding about some of the issues raised in this volume's articles. These interviews, and my analysis of discourses of multilateralism over time in Chinese foreign policy, suggest that one may be seeing in some instances similar processes to those described by Gheciu and Lewis. That is, there is some evidence that those individuals most directly exposed to intensive social interactions with individuals and groups that are strong proponents of regional security dialogues in the Asia-Pacific, for example, are more likely to have a positive attitude toward multilateralism than those who have not had this exposure. The nature of these security dialogues—deliberative, exploratory, and, for some, multilateralist activists an opportunity to “teach” new norms of security—seems to facilitate this socialization process.47
For a summary of this research, see Johnston 2003.
Still, to be honest, the richness and robustness of these conclusions about Chinese participants in regional security dialogues leaves something to be desired compared to the data on Europe. Access to bureaucrats who can be systematically interviewed or sampled for survey-style protocols is hampered by the fact that many of the relevant countries in the Asia-Pacific are authoritarian regimes in which foreign policy issues are off limits to public and academic debate.
Having said that, I think there is a reasonable chance of doing cross-regional surveys and interviews on institutions and identity, comparing Europe with nonauthoritarian Asia-Pacific, North American, and South American institutions. Depending on the variables of interest, these contrasts could control for external threat environment, internal social-political stability, level of development, the temporal space that a region is in, the macro-historical diffusion of institutional models, colonial history and postcolonial ideology, as well as formal and informal rules of the institutions.
It is quite possible, of course, that what appear to be broad interregional differences in institutional design and effects disappear if one looks at specific processes within institutions across regions. I was struck, for instance, by potential similarities between the ASEAN Regional Forum and COREPER in terms of the effects of small-group interaction, particularly the role of persuasion in effecting changing views of “national interest.” Indeed, I think that Lewis' research methods in analyzing COREPER might be usefully reproduced in an analysis of the ARF and its Track II companion, the Council on Security Cooperation in the Asia Pacific. Another interesting study, inspired by Hooghe's distinction between national (or primary) and international (or secondary) affiliations, might be cross-regional comparisons in attitudes toward regional institutions between national bureaucrats in foreign ministries with no exposure to or responsibility for these institutions, diplomats whose responsibility is to participate in these institutions, and the professional Track II academics and think tank members whose main responsibility is to generate ideas for the Track I multilateral institutions in the region. In the Asia-Pacific region, at least, one would expect the first group to exhibit the lowest level of support for multilateralism, and the last group to exhibit the highest level.
I am also intrigued by potential comparisons among the effects of socialization processes on state agents from countries that are not far removed from the political and strategic culture of Marxist-Leninist systems—undemocratic and hard realpolitik. Here, comparisons between Gheciu's Eastern European cases and Chinese, Vietnamese, or North Korean diplomats would be instructive. Obviously the starting conditions are different—China is not officially eager for interactions on civil-military relations, or seminars on conflict management from the Organization for Security and Cooperation in Europe (OSCE). But the processes Gheciu describes, if transportable in some politically acceptable fashion, seem to be prime facie evidence in favor of intensive military-to-military engagement between the United States and its allies and China.
The major problem in doing any of these comparisons, of course, is access to diplomats or state agents. In the ARF case, for instance, some delegations allow academic participant-observers (Canada and Australia), while others (China) clearly do not.
Conclusion
One of the standard, usually valid, criticisms of any new analytical approach in international relations is the academic version of “where's the beef”—what does it tell us about the empirical world and the empirical puzzles that have been central to IR as a subfield? Theorizing about socialization in IR has been criticized for being long on meta-theory but short on testable hypotheses and empirical analysis. This criticism has not just come from those who are fundamentally skeptical of the “sociological” turn in IR, but also from its sympathizers.48
See Checkel 1998.
This volume has responded to this criticism, and admirably, in my view. To do this it has gone to mid-range theorizing, eclectically borrowing hypotheses from other subfields and testing them on a “most likely case” to see whether it is worth going further with the concept of socialization. In general, this is a smart research strategy for a novel approach in IR. It avoids alienating much of the field with another debate about meta-theory.49
I do not think that these meta-theory debates should stop, however. There is a danger that the field, having started with debates that assumed the complete incommensurability of rationalist and constructivists approaches, may swing to the other extreme, to the view that there is complete complementarity. At least in terms of research practice, rationalists have tended to start from a materialist ontology; constructivists from an ideational one. It is hard to see how these two can be easily reconciled. The fault line, then, is not so much rationalism versus constructivism, but the rationalist tendency to default to maximizing material interest versus the constructivist tendency to default to ideational appropriateness as the keys to understanding human behavior. These two starting points may indeed be incommensurate and it is important to understand, therefore, the implications of both for IR research.
Specifically, the articles show the value of using mid-range theories about exposure effects (for example, primacy, recency, intensity), persuasion, social influence, cognitive role playing, and identity for analyzing the real-world policy implications of the social effects of human interaction inside different institutional environments. These articles have outlined a range of scope conditions under which each of these different mechanisms are hypothesized to work—conditions that range from the nature of state-society relations to the degree of “noviceness” of actors to the length of time actors have been exposed to new social environments.
In a sense, the articles collectively show that what is called a sociological turn is really a sociological and psychological turn. It remains to be seen whether this particular application of psychology in IR will meet the same fate as the application of psychology in comparative foreign policy analysis—namely being looked down upon by the field as a whole in favor of pristine deductive theorizing. But I suspect not, for at least three reasons. First, the socialization research agenda, while focused thus far on fairly narrow and discrete empirical cases (for example, attitudes toward civil-military relations in Eastern Europe, for example) actually addresses perhaps the most fundamental issue in IR, that of the origins and evolution of actor preferences.
Second, the socialization approach is, I believe, sustained methodologically and ontologically, by the growing realization in the social sciences in general that outcomes are often the nonlinear emergent properties of multiple adaptive agents. In other words, whether it is falls under the rubric of complexity theory, agent-based modeling, or parallel processing, there is a growing willingness to accept that endogeneity is a “fact” of politics and IR, and thus agent-structure interaction is a more insightful, if still hard to execute, mode of analysis than the traditional linear regression-type mode. (I include in this latter mode the standard comparative case study method as well.) This ontology requires micro-analyses, precisely what these articles do by focusing on the study of individuals in small groups. Perhaps this is the start of applying the agent-based modeling concept to case-based empirical analysis. (I will return to this in a moment.)
Third, the socialization approach—particularly as developed in this volume—does not exclude a priori any “theory” of human action—such as strategic optimization, strategic satisficing, normative appropriateness, identity role playing, and habituation. While those working with this approach are most likely to be labeled constructivists or fellow travelers, this obscures the fact that they are more inclusive about motivation, hence about the methods of analysis and the range of empirical data used in IR, than both what is usually labeled “the mainstream” (versions of realism, and international political economy) and what is labeled the “marginalized” (postmodernism).
If this volume is a good start to the empirical analysis of socialization mechanisms, what is the next step? I do not intend to recap all the criticisms and suggestions for extensions that Checkel, Zürn and Checkel, and I have listed in our articles.50
See Checkel, this volume; and Zürn and Checkel, this volume.
First, socialization research needs to focus more on comparison with other regions, other periods of time, and other socialization environments. For instance, beyond how European institutions might encourage socialization to be more supportive of institutional cooperation or new ideas about democratic notions of civil-military relations, scholars need to look more at how different levels of resistances to socialization might lead to hybrid or localized normative change.51
See Acharya 2004.
Second, socialization research needs to think more about methods. This goes beyond investigating the value, for instance, of testing one author's propositions on another author's cases, as I suggested earlier about the articles in this volume. It includes a more fundamental look at how to access the most difficult, though most central, piece of the socialization puzzle: how does one show changes in how agents understand their identities and preferences after they have been exposed to socialization environments? How does one show appropriateness or habituation in decision making? How does one show change in the content of agent identities? How does one show the depth of socialization (how does one show changes in the tradeoffs that agents would make under realistic conditions before and after socialization experiences)? There may be some potentially promising methods outside of the traditional toolbox of mainstream international relations (for example, cognitive mapping to show the absence of certain choices in the cause-effect arguments of actors after their socialization; semantic-differential analysis of changes in the content of agents' beliefs about their identity, among others).52
For examples of how these and other methods might be applied to the analysis of identity and identity change, see Abdelal et al. 2005. Some useful tools have actually been used in comparative foreign policy analysis for some time (for example, cognitive mapping and other forms of cognitive-linguistic analysis, and content analysis), but many IR scholars are not routinely exposed to these methods.
Third, socialization research needs to think more about research design. I do not mean by this simply the problems of case selection, the selection of appropriate methods, and alternative hypothesis testing. What I mean is how to research the endogeneity at the heart of socialization claims—agents interacting with agents producing changes in normative structures, which in turn resocialize or differently socialize agents and so on. This kind of research will rest on two key steps. The first is to figure out the prior distribution among relevant agents of their susceptibility to persuasion, in their sensitivity to image/status markers, of their willingness to brave policing by the ingroup in order to hold on to liminal identities. This distribution will affect the speed, depth, and content of socialization, that is, whether one should expect a strict reproduction of prior norms, their hybridization or localization, or their complete rejection. How does one measure this prior distribution?
The second step is to figure out how to empirically test this endogeneity process. The analogical method is agent-based modeling using computers. How does one organize a research project whereby all relevant agents and their iterated interactions with each other and with the emergent property (normative structure) are traced empirically systematically over time and space? The empirical data requirements are intimidating to say the least. Pragmatically it would require multiple researchers with multiple skills working on the same research puzzle. It would, in short, require lab-like research workshops, producing multiauthored research products. Unfortunately the social sciences are not set up for rewarding this kind of research financially or in terms of career opportunities, especially for untenured faculty. Atomized, individual research on a case of socialization here or a case of socialization there only covers one small part of the entire endogeneity processes.
The bottom line, then, is that moving forward in analyzing socialization in international relations requires a much more fundamental rethinking of how research is conceptualized and conducted in IR. Indeed, it requires the resocialization of a field.
References
REFERENCES
Abdelal, Rawi,
Yoshiko M. Herrera,
Alastair Iain Johnston, and
Rose McDermott.
2005.
Identity as a Variable.
Paper presented at the Harvard Identity Project Conference,
December, Cambridge, Mass.
Acharya, Amitav.
2004.
How Ideas Spread: Whose Norms Matter? Norm Localization and
Institutional Change in Asian Regionalism.
International Organization
58 (2):239–75.CrossRefGoogle Scholar
Alter, Karen.
2002.
Delegating to International Courts: Why, How, and the Problems in
Doing So.
Paper presented at the Conference on Delegation to International
Organizations, Park City, Utah.
Available at
〈http://www.internationalorganizations.org/alter.PDF〉.
Accessed 22 June 2005.
Axelrod, Robert, ed.
1976.
The Structure of Decision: The Cognitive Maps of Political
Elites.
Princeton, N.J.:
Princeton University Press.
Axsom, Danny,
Suzanne Yates, and
Shelley Chaiken.
1987.
Audience Response Cues as a Heuristic Cue in Persuasion.
Journal of Personality and Social Psychology
53 (1):30–40.CrossRefGoogle Scholar
Biddle, Bruce J.
1985.
Social Influence, Self-Referent Identity Labels and Behavior.
Sociological Quarterly
26 (2):59–86.CrossRefGoogle Scholar
Bourdieu, Pierre.
1991.
Language and Symbolic Power, edited by
John B. Thompson.
Cambridge, Mass.:
Harvard University Press.
Broms, Henri, and
Henrik Gahmberg.
1983.
Communication to Self in Organizations and Cultures.
Administrative Science Quarterly
28 (3):482–95.CrossRefGoogle Scholar
Burke, Peter J., and
Judy C. Tully.
1977.
The Measurement of Role Identity.
Social Forces
55 (4):881–97.CrossRefGoogle Scholar
Chafetz, Glenn,
Hillel Abramson, and
Suzette Grillot.
1996.
Role Theory and Foreign Policy: Belarussian and Ukrainian Compliance
with the Nuclear Nonproliferation Regime.
Political Psychology
17 (4):727–57.CrossRefGoogle Scholar
Checkel, Jeffrey T.
1998.
The Constructivist Turn in International Relations Theory.
World Politics
50 (2):324–48.CrossRefGoogle Scholar
Choi, Young Back.
1993.
Paradigms and Conventions: Uncertainty, Decisionmaking, and
Entrepreneurship.
Ann Arbor:
University of Michigan Press.
Cialdini, Robert.
1985.
Influence: The New Psychology of Modern Persuasion.
New York:
Quill Books.
Cialdini, Robert.
1987.
Compliance Principles of Compliance Professionals: Psychologists of
Necessity. In
Social Influence: The Ontario Symposium on Personality and Social
Psychology, Vol. 5, edited by
Mark P. Zanna,
James M. Olson, and
Peter Herman,
165–84.
Hillsdale, N.J.:
Lawrence Erlbaum.
Cialdini, R. B.,
C. A. Kallgren, and
R. R. Reno.
1991.
A Focus Theory of Normative Conduct: A Theoretical Refinement and
Reevaluation of the Role of Norms in Human Behavior. In
Advances in Experimental Social Psychology, Vol. 24, edited
by
Mark P. Zanna,
201–34.
New York:
Academic Press.
Confino, Alon, and
Ajay Skaria.
2002.
The Local Life of Nationhood.
National Identities
4 (1):7–23.CrossRefGoogle Scholar
Dalby, Simon.
1990.
American Security Discourse: The Persistence of Geopolitics.
Political Geography Quarterly
9 (2):171–88.CrossRefGoogle Scholar
Domínguez, Jorge I.
2002.
Latin America's Long March toward International Cooperation:
The Jigsaw-Puzzle Approach to the Design of Regional Institutions.
Paper presented at the Crafting Cooperation: A Comparative Study
of the Design and Efficacy of Regional Institutions conference, October,
Harvard University, Cambridge, Mass.
Doty, Roxanne Lynn.
1993.
Foreign Policy as Social Construction: A Post-Positivist Analysis of
US Counterinsurgency Policy in the Philippines.
International Studies Quarterly
37 (3):297–320.CrossRefGoogle Scholar
Fairclough, Norman.
2003.
Political Correctness: The Politics of Culture and Language.
Discourse and Society
14 (1):17–28.CrossRefGoogle Scholar
Fearon, James.
1999.
What Is Identity (as We Now Use the Word)?
Unpublished manuscript,
Stanford University,
Stanford, Calif.
Fearon, James, and
Alexander Wendt.
2002.
Rationalism vs. Constructivism: A Skeptical View. In
Handbook of International Relations, edited by
Walter Carlsnaes,
Thomas Risse, and
Beth Simmons,
52–72.
London:
Sage.
Feldman, Stanley, and
Karen Stenner.
1997.
Perceived Threat and Authoritarianism.
Political Psychology
18 (4):741–70.CrossRefGoogle Scholar
Fierke, K. M.
2002.
Links Across the Abyss: Language and Logic in International
Relations.
International Studies Quarterly
46 (3):331–54.CrossRefGoogle Scholar
Granovetter, Mark.
1978.
Threshold Models of Collective Behavior.
American Journal of Sociology
83 (6):1420–43.CrossRefGoogle Scholar
Gumperz, John, and
Stephen Levinson, eds.
1996.
Rethinking Linguistic Relativity.
Cambridge:
Cambridge University Press.
Hirshleifer, David.
1995.
The Blind Leading the Blind: Social Influence, Fads, and
Informational Cascades. In
The New Economics of Human Behavior, edited by
Mariano Tommasi and
Kathryn Ierulli,
188–215.
Cambridge:
Cambridge University Press.
Holland, John.
1996.
Hidden Order: How Adaptation Builds Complexity.
Boston:
Addison-Wesley.
Hopf, Ted.
2002.
Social Constructions of International Politics: Identities and
Foreign Policies, Moscow, 1955 and 1999.
Ithaca, N.Y.:
Cornell University Press.
James, Michael Rabinder.
1998.
Communicative Action and the Logics of Group Conflict.
Paper presented at the 84th Annual Meeting of the American
Political Science Association, September, Boston.
Johnston, Alastair Iain.
2001.
Treating International Institutions as Social Environments.
International Studies Quarterly
45 (4):487–515.CrossRefGoogle Scholar
Johnston, Alastair Iain.
2003.
Socialization in International Institutions: The ASEAN Way and
International Relations Theory. In
International Relations Theory and the Asia-Pacific, edited
by
Michael Mastanduno and
G. John Ikenberry,
487–516.
New York:
Columbia University Press.
Jorgensen, Charlotte,
Christian Kock, and
Lone Rorbech.
1998.
Rhetoric that Shifts Votes: An Exploratory Study of Persuasion in
Issue-Oriented Public Debates.
Political Communication
15 (3):283–99.CrossRefGoogle Scholar
Jost, John T.,
Arie W. Kruglanski,
Jack Glaser, and
Frank J. Sulloway.
2003.
Political Conservatism as Motivated Social Cognition.
Psychological Bulletin
129 (3):339–75.CrossRefGoogle Scholar
Kelley, Judith.
2004.
International Actors on the Domestic Scene: Membership
Conditionality and Socialization by International Institutions.
International Organization
58 (3):425–57.CrossRefGoogle Scholar
Kull, Steven.
1988.
Minds at War: Nuclear Reality and the Inner Conflicts of Defense
Policy Makers.
New York:
Basic Books.
Mansfield, Edward, and
Jack L. Snyder.
2002.
Democratic Transitions, Institutional Strength, and War.
International Organization
56 (2):297–337.CrossRefGoogle Scholar
Marques, Jose,
Dominic Abrams, and
Rui Serôdio.
2001.
Being Better by Being Right: Subjective Group Dynamics and
Derogation of Ingroup Deviants When Generic Norms Are Undermined.
Journal of Personality and Social Psychology
81 (3):436–47.CrossRefGoogle Scholar
Mutz, Diana C.,
Paul M. Sniderman, and
Richard A. Brody.
1996.
Political Persuasion: The Birth of a Field of Study. In
Political Persuasion and Attitude Change, edited by
Diane C. Mutz,
Paul M. Sniderman, and
Richard A. Brody,
1–16.
Ann Arbor:
University of Michigan Press.
Ochs, Elinor.
1986.
Introduction. In
Language Socialization Across Cultures, edited by
Bambi B. Schieffelin and
Elinor Ochs,
1–13.
Cambridge:
Cambridge University Press.
Perell, Paul.
1987.
Stare Decisis and Techniques of Legal Reasoning and Legal
Argument.
Legal Research Update, Nos. 2, 3.
Available at:
〈http://legalresearch.org/docs/perell.html〉.
Accessed 22 June 2005.Google Scholar
Risse-Kappen, Thomas, ed.
1995.
Bringing Transnational Relations Back In: Non-State Actors, Domestic
Structures, and International Institutions.
Cambridge:
Cambridge University Press.
Romer, Daniel,
Kathleen H. Jamieson,
Catharine Riegner,
Mika Emori, and
Brigette Rouson.
1997.
Blame Discourse versus Realistic Conflict as Explanations of Ethnic
Tension in Urban Neighborhoods.
Political Communication
14 (3):273–91.CrossRefGoogle Scholar
Rule, Brendon Gail, and
Gay L. Bisanz.
1987.
Goals and Strategies of Persuasion: A Cognitive Schema for
Understanding Social Events. In
Social Influence: The Ontario Symposium on Personality and Social
Psychology, Vol. 5, edited by
Mark P. Zanna,
James M. Olson, and
Peter Herman,
185–206.
Hillsdale, N.J.:
Lawrence Erlbaum.
Sapir, Edward.
1956.
Culture, Language and Personality, edited by
D. G. Mandelbaum.
Berkeley:
University of California Press.
Searle, John.
1992.
The Social Construction of Reality.
Cambridge, Mass.:
MIT Press.
Tan, See Seng,
Ralf Emmers,
Mely Caballero-Anthony,
Amitav Acharya,
Barry Desker, and
Kwa Chong Guan.
2002.
A New Agenda for the ASEAN Regional Forum.
Singapore:
Institute for Defence and Strategic Studies.
Waever, Ole.
1990.
The Language of Foreign Policy.
Journal of Peace Research
27 (3):335–43.CrossRefGoogle Scholar
Whorf, B. L.
1956.
Language, Thought and Reality, edited by
J. B. Carroll.
Cambridge, Mass.:
MIT Press.
Yee, Albert S.
1996.
The Causal Effects of Ideas on Policies.
International Organization
50 (1):69–108.CrossRefGoogle Scholar
Zimbardo, Philip G., and
Michael R Leippe.
1991.
The Psychology of Attitude Change and Social Influence.
Boston:
McGraw Hill.
You have
Access
- 95
- Cited by