12. Norms, Identity and Their Limits: A Theoretical Reprise

Paul Kowert and Jeffrey Legro

It is neither the borders nor the men who make a nation; it is the laws, the habits, the customs, the government, the constitution, the manner of being that comes from all of this. The nation is in the relations of the state to its members: when its relations change or cease to exist, the nation vanishes.

-- Jean-Jacques Rousseau Correspondance générale

Unlike Rousseau, many contemporary students of international politics treat the material facts of a nation's existence-its physical capabilities, technological achievements, geographical location, and so on-as the final arbiters of political outcomes. Not only are the laws, habits, and customs that Rousseau emphasizes considered epiphenomena, but they are notoriously elusive subjects for social scientific inquiry. Volumes such as this one therefore necessarily devote considerable effort to justifying their own approach and subject matter. The authors of the preceding essays seek to show in great detail, and in a wide variety of settings, that custom is often more important than capability and that social prescription often super sedes material self-interest. This essay critically reviews these efforts and considers the payoff for international relations theory of a focus on norms.

Norms, including the laws, habits, and customs to which Rousseau refers, are social prescriptions. While such prescriptions take many forms, this volume focuses first on norms that regulate the behavior of important actors in international politics. ₁   Norms are thus tied to actors. "New Thinking" in the Soviet Union was a norm, as was "hard realpolitik" in Mao's China. Both prescribe behavior for national policy-making elites (and, on another level, for nations). The changing collective beliefs about the use of nuclear or chemical weapons, the legitimacy of military intervention, and even the spread of democracy are all examples of widespread and evolving political norms. These too prescribe and regulate the practice of agents in international politics.

We wish, however, to maintain one important distinction among the different types of norms discussed in this essay. In much the same way that Wendt has distinguished between structure and agent in international relations theory, we will distinguish between norms as the regulative cultural content of international politics and identities-regulative accounts of actors themselves. ₂   Another example may help to clarify this distinction. Charles Kindleberger has argued that the rise of free trade in mid-nineteenth-century Europe came not from changing material interests but rather from the spread of a laissez-faire ideology. ₃   This ideology itself consists of a web of normative claims (about the efficacy of new trade practices, for example). But it also incorporates an account of the identity of actors who may legitimately participate in and govern trade (the legitimate practices of sovereign nations, for example, continued to differ from those of colonies). Identities, therefore, are prescriptive representations of political actors themselves and of their relationships to each other. We thus divide social prescription into these two categories: prescriptive accounts of actors themselves (identities) and behavioral prescriptions for the proper enactment of these identities (behavioral norms). ₄   When maintaining this distinction is unimportant, however, we will occasionally use the latter term (norms) in the more general sense described above to indicate social prescriptions including regulative accounts of actor identity.

This essay raises two questions about social prescriptions in world politics. First, what are the consequences of behavioral norms and identities? What, in other words, is the payoff of the "sociological turn" in international relations theory away from individualism and materialism toward a focus on collective interpretation? The central claim of this volume is that international relations theory cannot afford to ignore norms. This essay assesses that claim by considering the impact of norms on the interests, beliefs, and behavior of actors in international politics. Demonstrating that impact does not invalidate other theories of international relations. Rather, it points to analytical blind spots and gaps in existing accounts. In so doing, it not only casts light into the shadows of existing theory but raises new questions (and offers new explanations) as well.

If norms are important, a second question naturally emerges: Where do the norms themselves come from? While the preceding essays devote considerable effort to answering the first question, they rarely address the second one. But they nevertheless offer some insights into potential answers. The second section of this essay therefore identifies three possible avenues to norm building. For the most part, these "sources" of norms remain ill-defined, incompletely theorized, and understudied. But if one accepts the central contention of this volume-that norms matter to the conduct of international politics-then the origins of norms is a natural subject for further study.

The third section of this essay considers several other difficulties with ongoing efforts to investigate the role of norms in international politics. While the practitioners of international politics (national leaders, for example) may believe they know norms when (and if) they see them, identification and categorization remain difficult problems for social scientists. Moreover, the fluidity of social norms and the complex interplay between physical reality and interpretation complicate such scholarship enormously. And the fact that at least some political actors are aware of norms and actively seek to manipulate them further confounds the efforts of scholars. This essay concludes, therefore, with an overview of the promise and perils of a focus on norms in international politics.

How Norms Matter

Norms have attracted attention in disciplines across the social sciences, including psychology, sociology, anthropology, economics, and political science. Such wide interest undoubtedly stems in large part from a desire to explain otherwise perplexing behavior. Why do actors adhere to social rules even in situations where these rules may run counter to their own material interests? ₅   What, more generally, is the fabric that holds pluralistic societies together? And when will these societies, facing each other across national boundaries, adhere to a set of (admittedly looser) rules for international conduct? These are critical concerns for students of international relations.

The international arena is often characterized as having a minimalist order because it is "anarchic"-that is, it lacks a sovereign to enforce rules, leaving only appeals to armed force to resolve clashes of interest between states. ₆   Scholars of post-World War II international relations, especially in security affairs, consequently tend to downplay the role of norms. Realists focus primarily on material capabilities and argue that norms, where they exist, merely ratify underlying power relationships. And while (neo)liberal theorists more often accord an independent role to norms, they nevertheless concentrate on explicit contractual arrangements (such as those embodied in regimes) intended to resolve collective action problems. This volume, however, argues that norms play a much broader role in world politics, shaping both cooperation and conflict in ways that are invisible to theories that focus either on material structural forces or on individual choice.

The argument here is not that approaches such as realism or liberalism are "wrong." Rather, it is that the micro- and macrofoundations of these perspectives are not equipped to account for the full range of political norms and their consequences. Indeed, the central assumptions of such theories direct attention away from the cultural variables on which this volume focuses. Thus, a useful starting point for this section on the consequences of norms is with these very gaps in existing approaches. Such gaps are important because they point directly to the most important effects of norms-and because they highlight the relevance of norms for mainstream international relations theory. This section considers first, therefore, the "economic" approach that provides the dominant microfoundation of both neorealist and neoliberal theory. It then turns to the formal structuralism that serves as a macrofoundation to both bodies of theory. ₇

Neoclassical microeconomics, in both its orthodox and its institutional versions, offers an important account of the microfoundations of social order. In the orthodox approach, social behavior and outcomes are a product of the rational choices of individuals who maximize their satisfaction (utility) by efficiently matching available means to their desired ends. The main factor that conditions actor choice is change in "prices" based on market forces (supply and demand) that tend to stabilize in equilibrium. ₈

In international relations theory, several approaches adopt this economic individualism. Strategic and game theoretic perspectives on conflict and cooperation employ a generic version of the economic approach to understand the dynamics of strategic choice and the circumstances (costs) that shape actor decisions. ₉   More generally, realist (and particularly neorealist) international relations theory rests on orthodox microeconomic foundations. Neorealists thus view states as generic entities, like firms, that respond rationally to costs in an international "market" defined by a distribution of capabilities among states. Robert Gilpin, to take a prominent example, explicitly adopts microeconomic language and analogies. ₁₀

While it may offer considerable advantages, the orthodox economic approach and its application to international relations tend to marginalize the importance of both behavioral norms and actor identity. In this spare version of economic individualism, norms play no independent role apart from the strategic choices of actors, since explanation focuses on agents that respond rationally to the "objective" environment (i.e., shifting prices in the markets for security and welfare). ₁₁   And actors themselves are posited as unitary, calculating, self-interested agents. ₁₂   When conceptions of actor identity and roles cleave to these minimalist assumptions, then orthodox economics can only provide an extremely sparse account of behavioral variations in actors facing similar constraints. The neoclassical model possesses no theory of differences in goals.

Another, less orthodox variant of neoclassical theory directs greater attention toward the institutional framework of individual choice. This school accepts the rational egoism of orthodox theory but accords institutionalized rules a special role in resolving problems of aggregation and coordination. These rules facilitate repeated interaction and, as Michio Morishima has argued, may contribute to long-run economic success. ₁₃

When individual preferences do not aggregate to a consistent "collective" interest, as discussed by Kenneth Arrow, these institutions may provide a framework of order that permits collective decision making. ₁₄   Similarly, institutions may coordinate expectations and actions in situations where there are multiple equilibria. A focus on institutions thus permits the theorist to examine mechanisms for dealing with the costs of uncertainty, information acquisition, and transactions rather than assuming a perfect and frictionless market. ₁₅

The study of regimes is the most prominent example of the institutionalist microeconomic approach in international relations theory. Neoliberal institutionalists explicitly acknowledge the collective rules (norms) that constrain and enable individual choice, but they continue to treat actor identities and interests themselves as preexisting and fixed. ₁₆   And to the extent that they are considered, norms (embodied in institutions) derive exclusively from rational egoistic choice. ₁₇   Their origins are thus limited to the preexisting preferences of agents, and their consequences tend to reflect this constraint. Identity thus remains marginalized, even in more expansive neoliberal institutionalist arguments. ₁₈

In contrast to the bottom-up view of economic individualism, structural approaches adopt a top-down logic. Structure can be conceptualized as having two aspects: material and social. The material aspect of structure refers to the relative position of subgroups within a society and the distribution of material capabilities among them. Social structure refers to the cultural context of actor behavior-the dominant beliefs and understandings that characterize a society. ₁₉   Most theories of international relations, as already suggested above, have placed far greater emphasis on the material aspects than on the social attributes of structure. And this tendency is evident in structural theories at both the international and the domestic levels of analysis.

At the level of the international system, neorealism's materially grounded view of structure has played an influential intellectual role, particularly in security studies. ₂₀   For neorealists, the structure of the interna tional system is sketched primarily along a single dimension: the distribution of power, defined in terms of material capabilities. Realism assumes that states generally perceive this balance in an accurate fashion and respond accordingly to secure their own relative advantage wherever possible. ₂₁   And if they fail to do so, they are likely to fall victim to a process of "natural selection" at the international level. Thus the system as a whole rewards realist adaptation. In this approach, customs and beliefs far removed from the distribution of power are relatively unimportant. Norms, where they matter at all, matter only at the discretion of (or in service to) the power structure. ₂₂   Neoliberal institutionalism, on the other hand, more readily allows for the possibility that norms, conventions, and principles may over time become uncoupled from "material" structure, thus exerting a limited independent influence. Even this approach, however, usually focuses on formal treaties and institutions rather than on less formal social expectations. And it grants norms an independent influence only when the (materially) functional imperatives of the international system are relatively modest. Many neoliberal theorists have thus been particularly reluctant to focus on informal structure and norms in the area of security studies. ₂₃

At the level of the nation-state, analysts have also relied on material or formal notions of structure and functional approaches to the problem of order. For example, Peter Gourevitch, Peter Katzenstein, and Stephen Krasner have all traced foreign economic or security policies to variations-derived from differences in material capabilities-in state political structures. ₂₄   Others, such as Barry Posen and Jack Snyder, point to similarity in organizational structures (e.g., the pursuit of maximum autonomy and size) to account for similarities in military doctrine. ₂₅   But, as with structural theories at the international level, these domestic structural theories make scant room for cultural norms to exert an independent effect. ₂₆   Conceptions of both the actors and the environment are, in these models, functionally derived from materially "objective" structures.

Both the economic individualist and the structural functionalist approaches are silent, therefore, on what this volume identifies as critical. The authors of the preceding essays find considerable variation in the preferences and character of actors in international politics so that, even when actors face similar constraints, they may respond in different ways. And bringing social, normative structures into sharper focus greatly facilitates an account of behavior in those situations that neorealists and neoliberals find perplexing. The authors of this volume thus explicitly problematize the assumptions-economic rationalism and formal, material structuralism-of the dominant approaches to international relations thought and are consequently able to shed new light on a variety of ways in which norms matter. Their arguments can be grouped into three categories, focusing on: (1) the effects of norms on interests, (2) the ways norms shape instrumental awareness of links between interests and behavior, and (3) the effects of norms on other normative structures (including actor identities). These three effects are shown schematically in figure 12.1 (see also the related discussion and diagram in Jepperson, Wendt, and Katzenstein's essay 2).

Of course, not every author in this volume emphasizes all three of these effects: the essays trace a variety of different paths through figure 12.1. The remainder of this section considers these paths in more detail, focusing on the three effects of norms listed above.

Interests

Theories of norms address gaps in economic rationalist accounts most directly by examining (rather than merely assuming) the construction of actor interests. In fact, norms may even shape an actor's interests or preferences in ways that contradict the strategic imperative of the international environment (as specified, for example, in realist balance of power theories) or the functional need to cooperate. In Finnemore's analysis of military intervention, for example, states may intervene to accomplish humanitarian objectives even when no obvious economic or strategic rationale is present. As she points out, a realist could identify few interests that explain the commitments Western nations have made in Somalia or Cambodia. But theories that allow for a fuller range of national goals-including humanitarian goals-are more readily able to account for evolving patterns of intervention. Similarly, Eyre and Suchman find evidence that developing countries-which may face very different strategic problems or threats-will tend to buy very similar types of weapons. These choices, they argue, have little to do with the external threats these states face. Instead, global norms of modernity shape the interests of national elites, directing their attention toward symbols of status and power such as advanced (but often inappropriate) weaponry.

Domestic-level norms may also shape state interests in ways that contradict the material international structure. Kier argues that the interwar preferences of both civilian and military leaders for military doctrine (either offensive or defensive) were defined more by internal culture than by the external balance of power. While the military aspired to professionalism, French civilians focused on how military policy affected their own domestic power rather than on external threats. And Risse-Kappen shows that norms associated with the spread of democracy, in a later period of European history, once again shaped preferences-this time promoting a new European alliance structure (nato). Again, this argument contrasts sharply with the realist expectation that states should ally against the strongest power regardless of ideological or political considerations. Finally, Johnston describes a remarkably constant parabellum strategic culture in China that, he argues, has produced a consistent set of Chinese interests despite changes in China's strategic position with respect to other powers. In fact, Johnston finds that this strategic culture determined the character of China's involvement in external politics, thus reversing the primacy that realism usually accords to the international system.

Instrumentality

Norms affect not only actor interests but also the ways actors connect their preferences to policy choices. More precisely, norms shape the instruments or means that states find available and appropriate. In other words, norms shape actors' awareness and acceptance of the methods and technologies on which they might rely to accomplish their objectives. Some means, of course, may be ignored simply because they are outside the knowledge set of the actors involved, however functional or technologically viable they might otherwise be. Jacob Viner argues, for example, that wealth-seeking states failed for some time to understand that free trade actually served their goals better than mercantilist autarky. ₂₇   And, in security studies, others have pointed to a "cult of the offensive" before World War I that biased choices of military doctrine at a time when defense (as ample evidence from the Boer and Russo-Japanese wars indicated) was functionally more appropriate. ₂₈

Even when actors are aware of a wide array of means to accomplish their policy objectives, they may nevertheless reject some means as inappropriate because of normative constraints. Finnemore's discussion of military intervention suggests, for example, that while states now have broader humanitarian goals that extend to more parts of the world, they simultaneously confront increasing normative constraints on the enactment of these goals. Unilateral intervention, even to accomplish humanitarian objectives, is seen as less and less appropriate. Ignoring this norm would have both international and domestic consequences that most leaders are unwilling to risk. And not only do such prohibitions shape actors' adoption of means, but their self-understanding (identity) also determines which means are acceptable. Price and Tannenwald thus argue that the widespread tendency to distinguish nuclear and chemical weapons from "conventional" weapons, and to prohibit their use, stems in important ways from norms of "civilization" (including prescriptions for a "civilized state," a particular identity). A civilized state, so the argument goes, cannot adopt these means-even in warfare. A rationalist and functionalist framework, they contend, cannot explain such restrictions when nothing about "unconventional" technologies distinguishes them from any other means of warfare. Burying the enemy alive in trenches and caves, or killing by immolation with a flamethrower, is hard to distinguish functionally (or even morally) from death by gas. Indeed, the World War II firebombing campaigns in Japan were every bit as brutal and even more lethal than the atomic bombs used later. Nor can realism account for the reluctance of states to use unconventional weapons, despite the military advantage they might have produced, in those cases where one combatant had no ability to respond in kind (as in Korea, Vietnam, or the Persian Gulf war). The taboo against chemical, biological, and nuclear weapons thus emerges as a subective and socially determined phenomenon.

Domestic norms have also helped to define the means that actors consider acceptable and effective. Kier's analysis of interwar France indicates that the norms of military organizational culture, in combination with France's broader political-military culture, had a definitive impact on whether offense or defense was chosen as the guiding principle of military doctrine. Thus norms shape not only civilian and military preferences (as noted above) but also the means for accomplishing these goals. While an offensive military doctrine was objectively possible (and Kier argues that the French military was well aware of German developments in offensive doctrine), the prevailing organizational wisdom among French officers was that short-term conscription afforded no means to attain the operational goals of an offensive strategy. French officers believed themselves constrained, therefore, to a defensive strategy. Similarly, in both Germany and Japan, Berger finds that internal cultural norms dictated an antimilitarist defense policy. ₂₉   These countries still desired security, but the perceived lessons of World War II highlighting the costs of unilateral defense prohibited the development of military tools that were affordable and that, from a realist viewpoint, would enhance security in an anarchic, uncertain world where even allies cannot be trusted completely. Once again, these cultural norms do not simply parallel systemic constraints as neorealism or neoliberalism might suggest; instead, they reflect internal social values and ideas.

Normative Structure

Not only do norms affect actor interests and awareness of instrumentality, but they also affect other normative structures. A good deal of this volume, in fact, is preoccupied with this kind of normative effect (which is circumscribed by the dotted line in figure 12.1). Behavioral norms, for example, may encourage certain national identities. Diffuse, underlying norms at one level may shape specific norms at another level. ₃₀   These effects are complex and frequently appear to be circular, but interactions among norms have profound consequences for the other effects of norms (on interests and instrumentality) already examined.

In their discussion of interactions between norms, several authors in this volume focus on the ways in which identity shapes prevailing rules for behavior. Barnett shows, for example, that pan-Arabism strongly influenced Arab national identities and inter-Arab politics, promoting a specific pattern of strategic ties that required consultation, placed limits on the use of force to resolve intracommunity conflicts, and excluded Western participation. And while the Persian Gulf war signaled an apparent end to whatever pan-Arabism remained in the region, the reemergence of statist identities paved the way for new patterns of sovereign behavior and new alliances with the West. In another case of alliance politics, Risse-Kap pen argues that democratic identity promoted reciprocity and consultation within the nato alliance. Even the trying circumstances of the Suez crisis, during which these norms were violated, served only to redouble the efforts of nato partners to build better institutions for cooperation. ₃₁

Not only does identity shape rules for behavior within the political environment, but behavioral norms can also interact powerfully with conceptions of identity. Sociologists and anthropologists have noted that some identities are challenged so severely by certain behavioral prescriptions that individuals would rather risk death than permit the two to remain in conflict. For example, Clifford Geertz describes a Balinese elite who marched into Dutch machine-gun fire rather than accept a way of life (and its attendant norms) that included foreign domination. ₃₂   Samurai of ancient Japan committed seppuku in the face of lost honor. And in times of starvation, elderly Eskimos committed suicide rather than burden their families. ₃₃   In each of these cases, individuals chose to negate their own identity rather than permit it to conflict with prevailing behavioral norms. ₃₄   And in each case the outcome is, to say the least, puzzling when viewed from the perspective of economic rationalism.

In international relations, sovereignty is the quintessential norm that demarcates global political space into nation-states, thereby conferring legitimacy on some actor conceptions (nationhood, for example) and not on others (such as supranational movements). Indeed, much of this volume focuses on the way prescriptive accounts of the political environment shape conceptions of identity. German and Japanese antimilitarism, as Berger indicates, owes much to the common experience of these two countries as they emerged from World War II (and to the prescriptive consequences that this experience entailed). Of course, their experiences were not identical. Nor were the roles that they adopted: Japan became a merchant state, while Germany sought closer ties to the West through formal alliances and institution-building. But in both cases the collective lessons of wartime defeat became embedded in political-military cultures that have since prevented the development of strong military resources to match the international position of these states.

The preceding discussion has noted that norms at several different levels may affect interests and instrumentality. This observation introduces a distinction-which is rarely treated explicitly in the empirical essays of this volume-between "levels of norms." And this distinction suggests, in turn, that norms may interact across these levels. Consider, for example, the vision of systemic structure adopted by hegemonic stability theory or Kenneth Waltz's discussion of systemic polarity, both of which propose a direct link between the distribution of power in the international arena and peace. ₃₅   Despite the materialist orientation of such theories, the structures that they feature involve considerable normative content. A nation-state system, whatever its polarity, requires a conception of the world as meaningfully divided according to state boundaries. And theories that focus on power require some collective knowledge about what national capabilities are meaningful and about how they are distributed. Differences in the content of these underlying structures may thus cause the hegemony of one nation to differ considerably from that of another, as John Ruggie has argued in a comparative analysis of Dutch, British, and American hegemony. ₃₆   And, as several authors in this volume argue, these underlying normative structures greatly influence specific behavioral norms and identities.

Herman suggests, for example, that Soviet perceptions of international structure were multifaceted (and certainly not the product of material structures alone). The mezhdunarodniki were informed more by the underlying norms of detente (such as the Basic Principles Agreement) than by those of competitive bipolarity. These structures pushed the New Thinkers toward a reconceptualization of Soviet identity (as a less "revolu tionary," more "normal" state). And their more liberal interpretation of international structure, in turn, also encouraged the transfer and consideration of new prescriptions for foreign and domestic security policies-of new behavioral norms, in other words. Interestingly, while detente may have encouraged greater pluralism in Soviet domestic politics for a time, Herman also argues that underlying normative structures (of "normalcy") eventually promoted isomorphism and convergence around New Thinking. Several other authors make similar arguments. Eyre and Suchman trace the acquisition of symbolically modern weapons by developing countries to underlying structures of national sovereignty. And Price and Tannenwald's account of weapons taboos relies strongly on preexisting structures of national sovereignty, "civilization," and technology. Before a weapon can be considered unconventional, the distinction between conventional and unconventional weapons in the context of national self-help must be meaningful-it must be institutionalized within existing global structures. And given this context, weapons taboos are not only possible but functional since they serve to demarcate civilized and uncivilized nations (another institutionalized distinction). Thus diffuse normative structures and institutions (e.g., "It is meaningful to speak of 'conventional weapons' and 'civilized states'") permit specific identities and behavioral norms (e.g., "We are a civilized nation" or "This weapon is not conventional"), resulting finally in specific means knowledge and behavioral prohibitins (e.g., "We will not use nuclear or chemical weapons").

To summarize, then, norms matter in a wide variety of ways. Diffuse normative structures such as sovereignty or even "civilization" shape the particular identities of actors on the international stage, as well as the rules for enacting these identities. The character of actors also influences the way they interpret the rules that apply to them. And, conversely, these rules limit the types of identities that are viable. This interlocking web of norms, in turn, shapes the particular interests of political agents. And it shapes their beliefs about the available (and best) instruments for achieving their goals. Neorealist and neoliberal theories of international politics offer no comparable model for specifying the interests of actors, nor can they account for persistent differences in interpretation of seemingly objective political structures. For all of these reasons, the sociological turn in international relations theory described here holds considerable promise. Of course, the sociological approach is not without its challenges. But before a consideration of some of these difficulties, a prior question merits discussion: What processes generate social norms?

The Sources of Norms

The irony of the criticism of neorealist and neoliberal theory voiced here is that, while the authors of this volume often take existing international relations theory to task for treating the construction of political actors and their preferences as exogenous, the essays that make up the body of this book tend to treat their own core concepts as exogenously given. To be sure, they acknowledge, often explicitly and at some length, that actor identity and behavioral norms are socially constructed. But this is generally a starting point, from which the essays proceed to focus on the impact of these social constructions (which is, after all, their primary concern). Arab, European, Chinese, Japanese, and Soviet political identities are all shown to have a great impact on strategic behavior in international politics . . . an impact that rationalist theory tends to ignore. And in some cases, the historical development of these identities is described, occasionally in detail. But about the process of identity construction, the authors have relatively little to say. And without any theory of how such identities are constructed and evolve, this research struggles to contribute more to an understanding of political behavior than the work that it criticizes.

The next task for scholars such as these is to take their own criticisms seriously and to develop more explicitly theoretical propositions about the construction of sociopolitical facts-the process of building collective norms and political identities. ₃₇   At least some of the authors in this volume might resist this task on the grounds that such generalizations are impossible. Indeed, while the essays are both "descriptive and explanatory," as Jepperson, Wendt, and Katzenstein observe in essay 2, they often seem to view the generative processes of norm building and evolution as unique to the cases they examine. And yet there is considerable material in these essays from which to construct plausible hypotheses about norm creation that might be applied to other cases.

This section takes up the task of identifying candidates for generalizable theory-not about the ways that social constructions cause (or enable, or give meaning to) certain political outcomes but about the sources of norms themselves. Where do collective political norms come from? How are the identities of states, military organizations, alliances, and other international actors constructed? In this section, we seek to distill, from the essays in this volume and from the broader social science literature, some tentative answers to these questions. In fact, the previous section has already pointed to one source of norms: other norms (possibly at other levels). But it is difficult to generalize about this source of norms because of their historical specificity and because of the potential for circular reasoning ("norms cause norms"). Therefore, we will not revisit the interaction between norms discussed above but focuses instead on three other processes that generate, maintain, and change political norms: (1) ecological, (2) social, and (3) internal.

Ecological processes derive from the pattern of relations between actors and their environment. Social processes stem from the relations between actors themselves. And, as the term implies, internal processes spring from the internal characteristics of actors. Internal processes, therefore, are reductive, while both ecological and social processes are irreducible to individual actors (and thus, in some sense, systemic).

Ecological Processes

Ecological processes result from the patterned interaction of actors and their environment. In some cases, actors confront a rapidly (or dramatically) changing environment. ₃₈   In other cases, continuities are more obvious; and in still other cases, the characteristics of the environment are unclear. The third category of cases, which we will consider first, point to the role of ambiguity in social knowledge. Ambiguity is not a function of particular actors or of the environment itself in the abstract. A political situation appears ambiguous only to actors, depending on their relationship to it. Arguments about the role of ambiguity in the creation of social norms have an old pedigree. In 1939, Muzafer Sherif described a series of experiments that suggested that collective, normative agreement was much easier in ambiguous settings (e.g., those in which a clear frame of reference is lacking) than in more clear-cut situations. ₃₉   His argument finds some confirmation in this volume. Herman proposes, for example, that in the Soviet Union the vision articulated by Gorbachev "was sufficiently vague to permit political forces with relatively moderate agendas to claim allegiance to a common ideal even as they strived to head off any radical change." ₄₀   At the outset, at least, politicians with different agendas could all claim to be New Thinkers. And while he is not so explicit, Barnett provides numerous clues that pan-Arab identity, to the extent that it existed, relied on a similar ambiguity to smooth over many important differences among Arab nations. When events such as the Baghdad Pact and, later, the Persian Gulf war brought these differences into sharper relief, the weak norms (and identities) of Arabism suffered greatly.

But the effects of ambiguity are . . . ambiguous. The authors of this volume take pains to point out that almost every aspect of international politics requires considerable social interpretation, no matter how incontrovertible materialists may take it to be. What counts, therefore, as an unambiguous international situation? By what metric can social scientists determine that one norm is more ambiguous than another? Put another way, how can differences in the degree of normative ambiguity be mea sured? And should one assume that ambiguous norms are always easier to establish than highly specified ones? After all, prohibitionary norms regarding the use of nuclear weapons flourished only after nuclear weapons were clearly distinguished from other weapons. In this case, clarity rather than ambiguity hastened the institutionalization of the nuclear taboo. The ways in which ambiguity relates to norm and identity construction thus remain in need of elaboration.

A second ecological process argument addresses the problem of norm and identity maintenance rather than their emergence. This argument stresses the passage of time and continuity in the environment, positing simply that iteration strengthens norms, quite apart from any other efforts to reinforce or undermine them. In other words, the longer a norm goes unchallenged, the more it tends to "solidify." Finnemore places great emphasis, for example, on the growth and solidification of humanitarian norms (including, but not limited to, norms of humanitarian intervention such as multilateralism) over "the past fifty or one hundred years." ₄₁   And Price and Tannenwald link iteration to both norms and identities; compliance with weapons taboos over time, they contend, both strengthens non-use norms and "reinforces the identity of states." ₄₂   While the iteration hypothesis seems plausible and is appealing in its simplicity, it seems equally possible that the opposite might sometimes occur: norms or identities, although unchallenged, might simply "fade away" over time, pass out of style, or for a variety of reasons become irrelevant. What, for example, of the prohibition in the Hague Regulations of 1907 against the attack or bombardment of undefended towns, villages, dwellings, or buildings? Rather than being challenged directly, this norm has simply lost force with the passage of time and the development of modern mass armies requiring a "nation at arms." ₄₃   Since time is not equally kind to all norms, some further refinement of this hypothesis seems in order.

A third ecological process argument expressed in this volume addresses changes in norms. This proposition-that dramatic shocks in the environment (to the international system, for example) loosen commitments to existing identities and behavioral norms-is an institutionalist's version of Mancur Olson's argument in The Rise and Decline of Nations. ₄₄   But whereas Olson focused on the ways in which systemic shocks eliminate redistributive agreements and redefine material incentives within a society, this volume opens the question of how such agreements and incentives themselves are constructed. One way of putting the argument is to say that an obvious shock to the international (or domestic) political system provides political capital for the proponents of change within a political system. The shock "proves" that politics must be conducted differently . . . or thought about differently. In this volume, the shock hypothesis finds its strongest exposition in Berger's discussion of German and Japanese identities after World War II. Berger argues that a national identity or a political-military culture will change rapidly only when it is "challenged by a major external shock" and "placed under great strain." Otherwise, cultural change is likely to be slow and incremental. World War II, he contends, provided just such a massive shock to the political-military cultures of militarism in Germany and Japan. Similarly, Barnett suggests that the shock of the Persian Gulf war helped to crystallize the dependence of Gulf Arab states on the West, reinforced emergent statism, and thereby allowed what had previously been taboo-a security alliance with the United States.

One difficulty with this type of argument lies in defining what counts as a shock. In Berger's case, the impact of World War II seems so decisive that one would expect it to qualify as a systemic shock whatever the definition. This allows Berger to avoid definitional issues. But how shocking must shocks be? What is it about the Persian Gulf war that qualifies it as a major systemic shock? Moreover, the "rally 'round the flag" effect must be reconciled with this general argument. Some shocks, at least initially, reinforce rather than challenge existing collective beliefs. The Japanese attack on Pearl Harbor, for example, strongly reinforced American identity. On the other hand, it served to undermine norms of isolationism that until that time had been quite strong. Not only does greater attention to the circumstances under which shocks will reinforce rather than challenge existing norms seem desirable, but also the relationship of shocks at one level to norms at another deserves further exploration. Barnett alludes to this phenomenon in his discussion of the intifada, which not only provided a shock to domestic norms of democracy within Israel but, as a result, also called into question the alliance relationship of Israel to the United States.

Social Processes

Social process arguments of norm building take the form of generalizations about the way human beings, organizations, states, or other political agents interact. ₄₅   At least two such arguments appear in the foregoing essays. The first of these is also perhaps the most straightforward: norms are spread through a simple process of social diffusion. ₄₆   Although they do not elaborate, several authors allude to this process. Risse-Kappen, for example, finds that "transgovernmental networks" greatly facilitated the development of a common identity (and norms of reciprocity and consultation) within nato. Finnemore suggests that international organizations encouraged a similar process involving the spread of humanitarian norms, while Herman argues that New Thinking in the Soviet Union was "essentially homegrown . . . [and] nurtured in an oppressive system." ₄₇

In the most explicit discussion of diffusion, Eyre and Suchman suggest that affinities such as common heritage or language may serve to increase the chances for norm or identity diffusion. Unfortunately, they do not take up this argument in any detail in the empirical section of their essay.

As appealing and straightforward as the social diffusion hypothesis is, it leaves many questions unanswered. Are new collective understandings expected simply to seep across "transgovernmental networks" like ever-widening inkblots? If not, then what operational criteria will allow us to identify the most prominent paths for diffusion? And which ideas will be communicated across these linkages? Why humanitarian norms instead of self-help ones, particularly given the rise of highly nationalist identities in some areas of the globe? Even if these questions are addressed, a more fundamental problem remains. Jepperson, Wendt, and Katzenstein, following Durkheim, call attention to the difference between shared understandings and collective understandings-the latter presumably being the concern of this volume. ₄₈   But diffusion seems to explain the sharing rather than the "collectivizing" of norms. The process by which shared knowledge becomes a collective norm remains underspecified. While these difficulties deserve further attention, the empirical essays nevertheless make a strong case that social diffusion is responsible for the emergence of many of the norms considered in this volume.

A second social process account of norm building draws on "interaction role theory" and suggests that norms-and particularly identity-emerge from a process of in-group/out-group differentiation and social role definition. ₄₉   It may encourage New Thinking in the Soviet Union, for example, to be able to oppose it to the "old thinking" of Brezhnev and Gromyko. Identifying the out-group helps to reify its supposed values (old thinking), which may then be opposed to in-group values. Another statement of the role relation argument can be found in Barnett's discussion of security communities in the Middle East. Barnett argues that identity is profoundly social-a product of continuous debate within the community, of evolving historical narratives, and of competing social portrayals (some of which posed great challenges to leaders in the Middle East). Barnett thus goes beyond simple in-group/out-group distinctions to argue that social interaction may produce a variety of relational (and interrelated) identities as actors take up various roles with respect to each other. Israel's role as an American ally in the Middle East, for example, also promotes (arguably) a democratic Israeli identity. And the Iraq-Turkey Treaty of 1955 "sent shockwaves through the Arab World" because it placed Iraq in a new role vis-á-vis the West and thereby challenged its Arab identity-an identity that was itself founded on role relations and the perceived threat from the West (a homogeneous out-group) to Arab states.

Barnett's discussion of role relations points the way to a more complex account of roles in international politics. Despite the advantages of parsimony, a degree of complexity may be partly unavoidable, since nations (not to mention other political actors) can relate to each other in so many different ways. Kal Holsti offers a typology of no fewer than seventeen specific national role conceptions defined, at least in part, through social interaction. ₅₀   These national roles include: regional leader, independent, faithful ally, liberator, and defender of the faith. And, more recently, Richard Herrmann has proposed twenty-seven categories of national role images based on differential social comparisons of capability, culture, and threat. Herrmann does go on to argue, however, that only a small subset of these (notably: enemy, ally, degenerate, colonial, and imperial) is likely to be common. ₅₁   Finally, Martin Sampson and Stephen Walker tie these role conceptions to cultural norms in a study that speaks directly to Berger's discussion of militarism in Germany and Japan. Sampson and Walker offer a different comparison-between France and Japan-but reach much the same conclusion: that differences in national role conceptions (identity) explain decisions about militarization in these two societies. ₅₂   Typologies of national roles take an important step toward analytical generalization. But role relation arguments do not always offer clear predictions, perhaps because they emphasize a highly contingent process of role creation. It remains unclear which social roles will become linked together and whether or not such linkages must be functional. Nor is it obvious how role relation theory should account for different levels of analysis. It may well make sense, after all, to speak both of a single Japanese identity (which relies greatly on Japanese distinctiveness-a relational concept) and of many Japanese identities, some more nationalist than others (indeed, Berger makes such an argument for both Japan and Germany). Much the same could be said for the Israeli identities that Barnett examines. While two-level games have received considerable attention among international relations theorists, two-level roles have not.

Internal Processes

Unlike other paths to the generation of norms, internal processes operate within political actors. For this reason, scholars interested in the phenomena of collectives often ignore internal processes, usually with the dismissive rationale that psychological or rational choice arguments cannot account for the generation of truly social norms or knowledge structures. Judith Goldstein and Robert Keohane, in their recent book Ideas and Foreign Policy, thus take pains to distinguish their work from psychological approaches. ₅₃   And in his paper on the evolution of Soviet New Thinking, Herman takes both rational choice and psychological theories to task for their failure to integrate social and political processes into their accounts of learning. ₅₄   But despite this implied criticism, Herman's subject matter lends itself easily to an internal process interpretation. In fact, Douglas Blum has offered an account of the evolving New Thinking in the Soviet Union that parallels Herman's in many respects but explicitly emphasizes cognitive rather than social processes. ₅₅   While internal processes do not themselves operate at the level of the collective, their effects may nevertheless be felt at other levels. In fact, Johnston's reliance on cognitive mapping to construct an image of the parabellum strategic culture in Maoist China is a tacit adoption of this internal process premise. ₅₆   Limitations of space do not permit a full discussion here of the ways individual-level processes may result in collective political norms. But three examples should suffice to give some indication of the variety of internal process arguments.

The first of these examples emphasizes a psychological process and is closely related to one of the social process arguments discussed above. Role-relation hypotheses about the construction of norms emphasize a process of social interaction encouraging role definition. But the definition of social roles might also, as considerable research suggests, stem from cognitive or motivational processes within individuals. In fact, the reluctance of the authors of this volume to place emphasis on people's need for identity in social relations-a need so strong that they will invent in-group and out-group identities and differences even when there is no rational basis for doing so-makes sense only in the light of this volume's general preoccupation with social rather than individual processes. ₅₇   Marilyn Brewer's theory of "optimal distinctiveness," however, traces identity formation to the individual's need to perceive distinctions between self and other, and to justify one's own behavior. ₅₈   Taking another approach, John Turner argues that social identities emerge instead from cognitive miserliness. According to Turner, "the first question determining group-belongingness is not 'Do I like these other individuals?' but 'Who am I?' " ₅₉   Turner's self-categorization theory emphasizes the mental heuristics that cause people, for reasons of cognitive efficiency, to focus selectively on information that confirms simplifying group stereotypes. In the case of either cognitive or motivational arguments, the result is the same: stable in-group and out-group identities-precisely the sort of identity on which this book places so much emphasis. ₆₀   Another internal process argument focuses on the use and interpretation of language. It would, of course, be inappropriate to identify the "linguistic" path to norm construction solely with an internal process since neither the grammar nor the content of a language is reducible to individual speakers. But, by the same token, speech act processes begin their opera tion at the level of the individual. ₆₁   Livia Polanyi et al. argue, for example, that "we need comprehensive textual analyses if we wish to identify and explore cultural expectations of normal and abnormal conduct, good or bad self and other identifications, possible or impossible or necessary historical development within particular security cultures." ₆₂   Their objective, rephrased in the terms of this book, is a linguistic account of norms and identity. ₆₃   To this end, they propose a "linguistic discourse model" that identifies regularized processes and multiple levels of speech: speech events (gestures, facial expressions, and so on), words, phrases, clauses, sentences, multi-sentences, and finally texts. Textual instruction devices-including analogies, metaphors, and varied forms of negation or emphasis-all serve as definitional wayposts in the construction of a discourse. In a similar vein, both Friedrich Kratochwil and Nicholas Onuf argue that social meanings and institutions are constructed out of the practical linguistic rules that operate within individuals but that have profound normative, constructive effects. ₆₄   In a sense, these linguistic theorists go one step further than cognitive psychologists: they argue that people must not only make sense out of their world, but must then communicate those mental representations to others . . . and that the process of communication is a process of mking sense. This process inevitably, they would argue, produces norms (in fact, the process is the production of norms). And while the "semantic dimensions of the language" permit some normative constructions, they render other constructions unintelligible. ₆₅

Although psychological and linguistic theories are often set in opposition to rational choice theories-at least in the field of international relations-rigid distinctions of this type are illusory. The cognitive efficiency invoked by Turner's self-categorization theory might easily be restated in rational choice terms. And Hayward Alker and his colleagues working in the fields of discourse analysis, artificial intelligence, and computer modeling have already examined in some detail the links between story-telling, motivation, and economic activity in iterated Prisoner's Dilemma games. ₆₆   In these theories, strategic agency plays an important role. But the internal process version of agency emphasizes not the normative constructions that interact through the behavior of actors who subscribe to them, but rather the (exogenously given) goals of agents themselves. A more straightforward restatement of the argument in rational choice terms is that the efforts of utility maximizers to attain efficient outcomes encourage norm construction as a device that reduces transaction costs. ₆₇   There is, of course, a large body of neoliberal research (on declining hegemony and regimes) that has focused almost exclusively on this pathway to norm creation. ₆₈   For neoliberals, norms are strictly functional at their inception, serving the (innate or given) interests of strategic actors. ₆₉   They reduce transaction costs and facilitate agreement among political agents. ₇₀   And they help to encourage behavior congenial to the interests (and identity) of the hegemon that (neoliberals assume) created them.

Whatever the flaws of internal process accounts of norm building, it seems inappropriate to dismiss them merely because they are not (by definition) "social" theories of collective knowledge. As a wide variety of cognitive, linguistic, and economic research shows, norms may also arise from the aggregate effects of processes that operate within individuals. Moreover, such reductive arguments may have another practical advantage: they are often better specified and may lead more easily to generalizable predictions than do the sociological theories emphasized in this volume. Remedying the "weakness" of the latter theories is, of course, precisely the justification for books such as this one.

To be sure, internal process theories are not immune to criticism. A particularly important gap in the individualist approach is the problem of aggregating individual choices. The aggregation problem is common to psychological, linguistic, and rational choice approaches. But it has received particular attention from rational choice theorists, who point out that social order can be difficult to understand from the point of view of atomistic choice when strategic situations do not encourage stability or equilibrium. As noted above, Arrow has demonstrated that certain individual preference orderings yield no consistent social preference ordering. ₇₁   And Jon Elster argues that many strategic interactions produce either no equilibrium, multiple equilibria, or unstable equilibria. ₇₂   Both Arrow and Elster suggest that norms themselves may play a role in resolving these aggregation problems. Elster argues that norms afford predictability, and for Arrow they take the form of "consensus on social objectives" that align individual interests. In either case, the implication is that norms precede and constrain the strategic interaction of individuals. For this reason, Alexander Field argues that an economic approach cannot explain institutions. ₇₃   If they are to be productive, efforts to account for norms on the basis of internal processes must squarely address the aggregation problem.

The authors of the preceding essays provide ample reason to suspect that further investigation of the social construction of norms and identity would be fruitful. They offer an intriguing variety of proto-theories (schematically summarized in figure 12.2) about the genesis of norms and identities. While some of these theories have been tested in other disciplines (sociology and psychology, for example), they deserve similar attention from students of international politics.

Challenges in the Study of Norms

In addition to the issue of generalizing about norm construction, the sociological turn in international relations theory faces several other theoretical and methodological challenges. This section examines five of them. The first, confronted in some way by the authors of each empirical paper in this volume, is to decide on criteria for identifying and measuring norms. It is not always apparent how, from a social scientific perspective, one can be certain that a norm is present. A second problem-in some sense the reverse of the first-is that norms often seem to be all too present. In other words, the field may suffer from an embarrassment of norms. A third problem is that norms figure in efforts to account for both continuity and change, with occasionally confusing results. A fourth problem is the difficulty of specifying the relationship between the normative and the material worlds. And finally, the fifth problem is the potentially confounding effect of agency on normative analysis. The criticisms offered in this section do not reflect a desire to undermine or reject the sociological approach to international relations. In fact, as we have argued, many of these problems are shared by (and may be more intractable within) the mainstream rationalist research program. ₇₄   The following discussion is simply an attempt to stake out challenges that lie ahead. Addressing these challenges will improve research efforts that make up the "sociological turn."

Knowing Norms

This volume generally treats norms as collective beliefs regulating the behavior and identity of actors. But that definition leaves important questions unanswered. To what degree must norms be shared before they can be called collective? And on what does their regulative authority depend? To argue that certain norms are influential is to suggest that their effect may vary with their strength. Thus Price and Tannenwald indicate that, as the taboo against using unconventional weapons has become more robust, the use of these weapons has declined. But by what criteria can one assess norm robustness (independent of the very outcomes one seeks to explain)? Some have linked a norm's strength to its institutionalization. ₇₅   But this simply pushes the problem back to one of measuring institutionalization-no easy task itself. ₇₆   And the chore is complicated by the widely accepted notion that some behavioral violations of norms do not necessarily invalidate the norms. At a certain point violations clearly do begin to undermine norms. But how do we assess that point?

Rational choice scholars have faced similar problems in identifying and measuring actor preferences or interests. Relying on what actors say can be misleading because of the strategic role of deception in public statements. Judging interests according to behavior, on the other hand, invites circularity. The unhappy compromise is to rely on "revealed preferences"-desires exhibited in past behavior. But such a move assumes that preferences do not change (or at least that they have not yet changed). And as game theorists themselves emphasize, action is not a sure guide to preferences since it also reflects actors' assessments of strategic circumstances and constraints. ₇₇   In general, students of norms have taken two different approaches to identifying their subject matter. The first is to focus their attention on what actors do-that is, whether actor behavior complies with norms or not (and, if not, how other actors respond). But if some degree of deviation from norms does not necessarily imply their repudiation, then this "revealed norm" method is unreliable. ₇₈   A case of incest, for example, does not by itself invalidate prohibitions against incest. Likewise, in Risse-Kappen's study, the violation of an alliance consultation norm during the Suez crisis did not lead to a collapse of the Atlantic security community or to the illegitimacy of its norms. Nor have the many violations of the chemical weapons taboo led (so far) to its irrelevance. And, at any rate, this approach permits only the post hoc recognition of norms.

A second approach is to focus on what actors say-that is, how they justify or defend their actions. Because an interpretive approach sees norms as communication devices, perhaps their presence and efficacy are indeed better judged by the pleas for understanding, rationales, and normative justifications than by behavior itself. The problem here, as for the rational choice theorist, is to distinguish manipulation and deception from more "genuine" forms of communication. How should one interpret proclamations of support for nuclear nonproliferation when a nation simultaneously works to build nuclear weapons . . . or worse, claims that nuclear or chemical weapons are too horrible to use, even when preparations for attack are under way? At the very least, then, it seems necessary to study both rhetoric and behavior over time-an approach adopted by most of the essays in this volume.

Efforts to identify and measure norms also suffer from a bias toward "the norm that worked." Most studies of norms, including those in this volume, focus only on a single, specific norm-or, at most, on a small set of norms. Typically, the norms under consideration are effective norms or ones that seem to have obvious consequences. ₇₉   Yet, in order to understand how norms work, studies must allow for more variation: the success or failure, existence or obsolescence of norms. A companion study to Price and Tannenwald's examining weapons taboos that did not take hold would be instructive. ₈₀   But normative research has tended to overlook the emerging norms that never quite made it, the portfolio of identities that were never realized, and the international structures that might have been. These counterfactuals, analyzed in conjunction with comparable cases of success, should lead to a more substantial research program. ₈₁   Addressing these identification and measurement problems is a formidable task for work in this area-but not one that is necessarily intractable. And the particular virtue of the empirical essays in this volume, unlike research that overlooks behavioral norms and that fails to problematize actor identity, is that these essays have at least begun to confront these important variables.

An Embarrassment of Norms

An equally troubling issue for those working in the sociological tradition is not the difficulty of identifying norms but their ubiquity. Several of the empirical essays in this volume make it clear that norms are multifaceted and that many different identities can exist side by side in a collectivity. ₈₂   In one sense, this is an advantage of sociological analysis rather than a problem, since it allows a richer account of political behavior. Moreover, while rationalists must assume that preferences are ordered transitively in order to aggregate interests in a collectivity, the sociological approach faces no similar constraint with respect to norms or identities. But because multiple norms can influence actors-with competing or even contradictory prescriptions for behavior and for identity-it is difficult to predict which norms will be most influential. Without clear conceptual definition and convincing measures of norm salience, the consequences of norms are likely to be indeterminate. One can almost always identify, post hoc, a norm to explain a given behavior. ₈₃   Examples of this problem are manifold. Finnemore recognizes that increasingly dense organization in global politics has facilitated the rise of humanitarian norms, but she argues that these norms run counter to other important institutional structures such as sovereignty and self-help. It is clear neither in the abstract nor in the concrete terms of the case that Finnemore discusses whether humanitarian or "isolationist" norms will ultimately prevail. But whatever the outcome, one could easily point to norms as a cause! ₈₄   Or, to take another example, Barnett discusses a variety of competing regional, Arab, and statist identities. Each of these identities has different implications for behavior. But, once again, which will prevail?

Perhaps a resolution to this problem will come from a reintegration of different levels of analysis (or more specifically, "levels of norms") within the sociological approach. ₈₅   A neorealist would almost certainly argue that there is a hierarchy of global political structures such that norms must conform to certain structures even though they may violate others. ₈₆   When basic national interests are challenged, for example, a self-help system will punish the states that do not respond appropriately. Structures such as these are functionally dominant in a neorealist analysis. And, even if we do not accept neorealist functionalism, the need for a "replacement" framework relating collective beliefs at different levels to one another is clear. ₈₇   Otherwise, it is difficult to avoid the criticism that "anything goes"-that any behavior can be explained with reference to some norm. In reaction to this position, several of the authors in this volume, including Eyre and Suchman, Finnemore, and Price and Tannenwald, focus on normative structures that they take to be decisive at the international level. Others (Berger, Kier, and Johnston) accord more impact to domestic structures. The great variety of norms that these authors identify, however, makes a cross-level, integrative analysis a challenging task.

Continuity and Change

Another complication for research on political norms is the difficulty of accounting for both stasis and change. One of the central contributions of the sociological turn in international relations theory is its move away from static conceptions of fixed identities and interests. This volume problematizes both state and other collective identities. But while social interaction among actors may produce new collective identities, existing normative structures also shape the properties and behavior of agents. Consequently, the sociological approach pursues an explanation both for patterns of stability and for cases of change.

At the heart of the problem is the inertial force of cultural constructs such as norms: this inertia tells us why certain patterns persist, but not why they change. Of course, a culture may consistently demonstrate patterns of change, as Harry Eckstein contends in the case of modern societies, but then the problem becomes one of accounting for observed continuities (themselves a deviation from more fluid patterns). ₈₈   The problem, in other words, is accounting for both stasis and change in the same culture. To do so, normative theorists must usually appeal to ad hoc arguments about forces external to a given culture. Thus Berger focuses on the shock of World War II to explain the rise of new cultures in Japan and Germany. And while Arab and regional identities may explain continuity in the Gulf Cooperation Council, they cannot also explain its demise and the emergence of alliances with the West. So Barnett, like Berger, invokes an external shock: the Persian Gulf war.

While cultures may certainly have multiple themes (or subcultures) that are in tension, explanations for the hierarchy of these different cultural elements (including multiple identities) are poorly developed at present. Several approaches to this difficulty have been proposed. Ann Swidler sees culture, for example, as a "tool kit" from which actors choose different types of actions. But, as already noted, if the available tools allow for both any action and its opposite, then their explanatory role is suspect. Indeed, Swidler accords exogenous "structural constraints" and "historical circumstances" substantial influence in dictating which "strategy of action" is selected. ₈₉   Similarly, in this volume Kier grants considerable leeway to historical circumstance in shaping both the organizational culture of the French military and the fears of civilian elites. And yet a greater appreciation for the hierarchical integration of cultural themes suggests at least a degree of cultural and normative stability. Finnemore, although clearly concerned with change and evolution in humanitarian norms over a 150-year period, thus maintains that normative change may be guided by principles of logical consistency. Expanding definitions of "humanity" led naturally (although perhaps not unavoidably) to the expansion of associated human rights. Likewise, Eckstein finds evidence of a more uniform process at work in culture. He proposes that changes in culture occur in response to environmental changes and argues that these changes will ordinarily work to maintain existing patterns. ₉₀   He allows that rapid contextual changes and even the instrumental efforts of agents (such as revolutions) may lead to cultural discontinuities. But these sources of change cannot themselves be explained within the logic of culture itself.

Constructivists appear to offer another approach to the problem of change and continuity, relying on a conceptualization that views structures and agents as linked in a dialectical synthesis. The interaction among agents and between agents and structures both produces and reproduces these entities. ₉₁   The key issue, of course, is whether "reproduction" or "different" production (i.e., change) will occur. And to answer this question, constructivists once again appeal to exogenous historical conditions. While historical contingency is undoubtedly central to understanding specific events, it does not easily lend itself to theoretical generalization. ₉₂   In fact, this line of reasoning suggests that social scientific inquiry can offer nothing more than a historically descriptive and highly contingent account. Although this proposition is inconsistent with the overall tenor of this volume, it occasionally finds exposition in the empirical essays (in Price and Tannenwald's observation, for example, that the chemical weapons taboo owes much to a series of fortuitous events). But while some historians and interpretive theorists might find common ground in this approach, it seems too pessimistic and too restrictive.

The bulk of the studies in this volume suggest yet another approach to change and continuity. While a single culture may be inadequate to explain both change and stasis, an appreciation for levels of culture-that is, different facets of culture nested within one another-may help. The preceding essays suggest multiple levels of nested norms, ranging from organizational culture (Kier) to strategic culture (Johnston) to political-military culture (Berger, Kier, and perhaps Herman) to international technological and political norms (Barnett, Eyre and Suchman, Finnemore, Price and Tannenwald, Risse-Kappen). This solution is remarkably similar to the rationalist attempts to deal with anomalies by demonstrating that actors are maximizing their welfare in different nested games. ₉₃   Of course, the danger for realists of both traditional and sociological stripe is an endless appeal to nests within nests of interests or norms. But in general the authors in this volume adopt the position that norms and identity are constructed through regularized processes, often with relatively stable effects. Even their discussions of historical contingency do not imply that all conceivable outcomes were equally likely.

Material and Normative Worlds

While the sociological perspective rejects the realist preoccupation with uniquely material forces, students of norms cannot afford to ignore the material world. Norms do not float "freely," unencumbered by any physical reality. They are attached to real physical environments and are promoted by real human agents (though norms, of course, are not themselves material). But the relationship of normative to material structures is rarely examined or explicitly theorized, despite the likelihood that the influence of norms may be related to the characteristics of the material structures in which they are embedded or the qualities of the actors that adopt or promote them. Norms backed by the United States are likely to become more widespread and effectual than otherwise similar norms originating in Luxembourg. While the differing capabilities of these two nations are undoubtedly a matter of interpretation, it is difficult to ignore the overwhelming material contrasts.

Scientific, technological, political, economic, religious, and even artistic structures all involve both formal and informal, physical and interpretive components. While Kier argues in this volume that the French choice of strategic doctrine in the interwar period did not derive functionally from perceptions about the state of military technology (favoring either a defensive or offensive stance), the history of warfare offers many examples of how technology was decisive in other ways. When Americans rushed to develop nuclear weapons, their expectations were eventually rewarded by a Japanese surrender. ₉₄   Very often, as in the latter case, actors are keenly aware of such technological change, even though they may be unsure of its impact. And perhaps the material aspects of cultural and religious similarities between Israel and the United States help to explain the close relationship between these two nations. Or, to take another example, the economic and technological performance of the United States undoubtedly contributed strongly to the persuasiveness of Western values reflected in Soviet New Thinking (and helps to explain why the West itself seemed to derive so little from Soviet norms during the interactions that Herman describes). In each of these instances, it is hard to overlook the possible connection between material developments and normative changes. To suggest a role for material influences on norms does not imply that norms stem directly from physical capabilities as realism more or less expects. But neither do collective understandings exist in a material vacuum. And a more synthetic conceptualization of the interaction between the material and the interpretive worlds remains necessary. ₉₅   center> a name="492"> Agency and Norms

One of the most important effects of norms, as we noted earlier in this essay, is their influence on the interests of political actors. But we also found the reverse to be true: agents sometimes set out to manipulate or change norms. Actors may be well aware of the potential advantages accruing to those who control certain norms. Hans Morgenthau thus warned long ago that rationalizations and justifications should not be allowed to conceal the true nature of foreign policy. ₉₆   Identifying the "true nature" of foreign policy is no easy task. But some agents are clearly able to use norms in an instrumental fashion to further their own interests rather than simply being held captive to various normative mandates. ₉₇   In World War II, countries sometimes abided by restrictions on the use of force (e.g., strategic bombing, submarine warfare) not out of ethical restraint but instead to retain domestic or international support. At the same time, some states tried to arrange situations that would cast another party as a "violator" and thus justify otherwise prohibited measures in retaliation. ₉₈   In such cases, both support and violation of prohibitionary norms were functions of strategic calculations that had little to do with the norms themselves. In this volume, Eyre and Suchman, Johnston, and Herman all point to other ways that leaders use norms instrumentally to enhance their own positions. The development of New Thinking in the Soviet Union, for example, was not merely an academic exercise and cannot be divorced from the strategic behavior of the mezhdunarodniki who promoted these new understandings in an attempt to change the course of Soviet foreign policy and, perhaps, to improve their own positions. In this case, promoting a certain redefinition of the situation (as a "failed" Brezhnev-Gromyko foreign policy) served the political interests of the liberal reformers. ₉₉

Even an actor's own identity can be manipulated for instrumental reasons. Kier provides an excellent example of this process, noting the extent to which military organizations construct their identity for strategic purposes: battlefield effectiveness requires a self-conception incorporating professionalism, obedience, and esprit de corps. In this case, effectiveness relies on strategic behavior designed to encourage particular (military) identities. Or, to take another example, Barnett argues that Israel manipulated its own identity for strategic advantage. Israeli leaders emphasized Israel's Western and democratic character so that Americans would see "something in themselves when they saw Israel." American "identification" with Israel was critical, Israeli leaders realized, to the continued flow of financial support from the United States.

These examples illustrate a distinction, which is not always acknowledged in the empirical papers of this volume, between internal and external norms. ₁₀₀   Internal norms are representations of the environment or constructions of identity as it appears to an actor. External norms, on the other hand, are representations of situations or actors to others. ₁₀₁   Some might object to this distinction on the grounds that, because norms are a collective phenomenon, all norms must be external. At the level of the individual, this is true. An internal norm for an individual is not a norm at all but merely an idea. But many of the actors discussed in this volume are not individuals but collectives (foreign policy elites, militaries, and nations, for example). For such actors, identity is different from the presentation of identity-though both involve regulative content. And the fact that actors manipulate self-presentation (that is, external norms) strategically does not mean that such norms are irrelevant. On the contrary, these manipulations would be pointless if norms did not matter. ₁₀₂   The instrumental manipulation of norms poses a more fundamental problem, however, than distinguishing internal from external norms. If actors self-consciously manipulate external norms, then the social scientific study of norms, their origins, and consequences is greatly complicated. Not only do the variables in such an analysis interact, but some variables (agents) are directly aware of other variables. In such cases, many social scientific models must be applied with greater caution than usual. Linear regression models, for example, might be inappropriate for certain analyses. Such instrumental awareness may violate assumptions about the independence of error components in the linear model, and it is very likely to introduce additional problems of multicollinearity and serial correlation. Quasi-experimental research designs also suffer when their subjects "know too much." Researchers should be cautious when their subjects are not only aware of the "experiment" but also actively trying to change its parameters. For these reasons, the instrumental manipulation of norms poses special difficulties for social scientific analysis. This issue, in turn, foreshadows a more philosophical problem-the double hermeneutic (the interpretation by scholars of the interpretations of actors) in social science. ₁₀₃   Since scholarly inquiry is itself interpretation (thus the product of agency), can it presume to investigate interpretive phenomena such as norms without, by the nature of the scientific enterprise, altering their meanings? This is not the place for an extended discussion of the philosophy of social science, but a few concrete examples drawn from the empirical essays may help to clarify the point. It seems almost undeniable that the norms of New Thinking in the Soviet Union, parabellum strategic culture in Maoist China, or reciprocity among nato allies mean the same thing to the authors of this volume as they did (or do) to the relevant political actors themselves. Can the effects of sociopolitical norms then be stated precisely, or are these effects (like the norms themselves) so much a matter of interpretation that any attempt to measure them distorts them? The analogy in physics is the Heisenberg uncertainty principle. But while physicists no longer find efforts to locate electrons within the electron shell of an atom meaningful (because this cannot be known except by measurement procedures that themselves define the possible locations), this volume undertakes a closely analogous task in the domain of international politics. The practical objection, therefore, is that there should be a Heisenberg uncertainty principle of interpretive analysis. And if the nor mative interpretations of actors are themselves unknowable except by an equally interpretive procedure carried out by social scientists, then it is no longer clear who or what is being studied.

This is a problem that confronts all social science. ₁₀₄   Those working in the sociological tradition are neither more nor less prone to it (although they may be more likely to recognize it). The criterion for judging international relations scholarship is not therefore the extent to which it replaces interpretation with "objectivity." Rather, it is the extent to which it provides better historical accounts and raises new questions. ₁₀₅   And while the other essays of this volume are mostly concerned with assessing the plausibility of norm-based portraits of political behavior, this paper argues that they succeed in doing much more. Not only are they able to tell stories about international politics that realists, neoliberals, and neorealist structuralists (among others) have difficulty telling, but they also point the way toward numerous intriguing relationships between underlying cultural structures, norms, identities, interests, and behavior-relationships that students of international politics have scarcely begun to investigate.

As this essay has suggested, there are (if anything) too many rather than too few plausible generalizations that describe the causes and effects of norms. The practical task is therefore to sort out unrelated, rival, and contradictory hypotheses rather than to reject the whole enterprise on the grounds that social norms can operate in a wide variety of ways. One method of doing this is to search for regularized pathways between norms, identities, interests, and behavior. One need not subscribe to functionalism in order to search for such patterned behavior. Indeed, there is considerable danger that a na•vely teleological view of norm function will obscure both "deeper" causes and counterfactual hypotheses. Many of the authors in this volume consequently resist functionalist arguments. But whether or not social norms are functional, this volume finds evidence that they are in some cases quite stable and that when they change they do so in predictable ways (in response to shocks or ambiguities, for example). General accounts of the causes and consequences of norms, such as those offered in this essay, thus provide a basis for rejecting the pessimistic claim that every interpretation of political reality is unique in all respects and that theory is impossible.

In fact, this volume is ambitious in its approach to international relations theory. It rejects the extreme historical constructivist position that generalizations about how actors will interpret their environment (political or otherwise) are impossible. Moreover, it rejects the materialist position that the physical reality of this environment governs cultural interpretations of it. And it rejects the tendency of rationalist theory to assume fixed goals and identities of actors. To attempt theory simultaneously on all of these fronts-that is, theory within the upper-right quadrant of Jepperson, Wendt, and Katzenstein's figure 2.1-would go beyond ambition if the authors also rejected the insights of scholarship in the other quadrants. But while this volume identifies shortcomings in the micro- and macrofoundations of mainstream approaches, it does not reject realism, liberalism, or their structural variants out of hand. In a broad analytical sense, there is a complementary relationship between the sociological perspective and these more traditional approaches. Accounts (and theories) of norms and identity fill gaps where other perspectives fall short. Of course, this complementarity can be overstated. A realist would not argue that threatened states will buy weapons for their symbolic (rather than military) value or that they will avoid other weapons-chemical or nuclear-that might serve their interests. Nor would a realist expect a nation to select a military doctrine that satisfies certain domestic and organizational interests but that leaves the nation woefully unprepared for the next war. Eyre and Suchman, Price and Tannenwald, and Kier-to take three examples-do not merely supplement realism; to a point, at least, they challenge it. ₁₀₆   The payoff of the sociological turn in international relations theory, therefore, is not merely incremental (i.e., more detailed historical accounts than are otherwise possible). It redirects scholars' attention toward consequential variables and processes that might otherwise go unnoticed. Norms and identity are both facets of culture and, as such, have been invis ible to scholars who choose to investigate more tangible, less value-laden phenomena. This essay highlights the analytical and explanatory significance of social norms. The final essay of the volume argues that this work is also timely and particularly relevant for problems of the post-Cold War order.

---

The authors wish to thank the participants of the Social Science Research Council/MacArthur Workshops at Cornell University (February 5-7, 1993), the University of Minnesota (January 14-16, 1994), and Stanford University (October 7-9, 1994), as well as members of the Mershon Center Research Training Group (rtg) on the Role of Cognition in Collective Political Decision Making at Ohio State University for discussions that enriched this essay. We are also indebted to Ethan Cherin, Martha Finnemore, Peter Katzenstein, Nicholas Onuf, Diana Richards, Stephen Walt, Katja Weber, the other participants in this book project, and two anonymous reviewers for their helpful comments and advice. Kowert gratefully acknowledges support from the International Relations Department at Florida International University and from a National Science Foundation grant (DIR-9113599) to the Mershon Center rtg at the Ohio State University. Legro is grateful for support from the Department of Political Science at the University of Minnesota.

Note 1:  Jepperson, Wendt, and Katzenstein define norms, in the second essay of this volume, as "collective expectations about proper behavior for a given identity." This is consistent with our definition, but naturally one might distinguish among many different types of norms. Some regulate morality, others prescribe accepted (but not ethically required) practice, and still others regulate the character of actors themselves. The latter type of norm, which we term identity, is discussed below.

Norms may also differ in their effects, in the extent to which they are known and accepted as just, in the explicitness of their content or provisions, in the uniformity of their application, in the degree to which sanctions accompany violations, in the degree to which they are internalized, and in the modes of their transmission. See Judith Blake and Kingsley Davis, "Norms, Values, and Sanctions," in Robert E. L. Faris, ed., Handbook of Modern Sociology  (Chicago: Rand McNally, 1964); Robert E. Edgerton, Rules, Exceptions, and Social Order  (Berkeley: University of California Press, 1985); and Janice E. Thomson, "Norms in International Relations: A Conceptual Analysis," International Journal of Group Tensions 23 (1993): 67-83. The common analytical divide between constitutive and regulative norms, for example, is a distinction based on effect: constitutive norms create or grant properties while regulative norms specify the proper enactment of these properties. On this distinction, see David Dessler, "What's at Stake in the Agent-Structure Debate?" International Organization 43, no. 3 (Summer 1989): 454-58; and Friedrich Kratochwil, Rules, Norms, and Decisions: On the Conditions of Practical and Legal Reasoning in International Relations and Domestic Affairs  (Cambridge: Cambridge University Press, 1989). Others make a related distinction between descriptive and prescriptive norms. Because we focus in this article only on norms with prescriptive or regulative effects, we leave aside many of these distinctions. Thus even our treatment of identity focuses explicitly on the regulative (rather than merely constitutive) effects of norms. We concentrate, in other words, not on how collective understandings describe or constitute actors but on how they regulate the "proper" identity of actors.

Another related distinction also deserves brief mention. Essays 1 and 2-and several of the empirical essays (notably those by Berger, Johnston, and Kier)-devote attention to "culture" as well as to norms or identity. Because we focus on social prescription, we avoid the more general term culture, which may include a wide variety of social knowledge apart from (prescriptive) norms. We do not attempt, therefore, to capture every aspect of the arguments offered in the empirical essays. What is distinctive about this volume, in our view, is its emphasis on social prescriptions of both identity and behavior (see Katzenstein's essay 1 and its discussion of culture). For a recent account of culture as collective knowledge that does not  (necessarily) involve prescription, see Judith Goldstein and Robert O. Keohane, Ideas and Foreign Policy: Beliefs, Institutions, and Political Change  (Ithaca: Cornell University Press, 1993). For a multifaceted account of the impact of culture on international politics, see Jongsuk Chay, Culture and International Relations  (New York: Praeger, 1990). Back.

Note 2:  Alexander Wendt, "The Agent-Structure Problem in International Relations Theory," International Organization 41, no. 3 (Summer 1987): 335-70; see also Dessler, "What's at Stake in the Agent-Structure Debate." Back.

Note 3:  Charles Kindleberger, "The Rise of Free Trade in Western Europe, 1820-1875," Journal of Economic History 35 (1975): 20-55. Back.

Note 4:  This distinction corresponds closely to Jepperson, Wendt, and Katzenstein's distinction in essay 2 between portfolios of identities and regulations for the behavior of "already-constituted identities." It also parallels, very roughly, the distinction between the two axes of international relations theory that they discuss (see their figure 2.1). Theories that emphasize the cultural, regulative content of the environment in which actors move (the x-axis) are theories of norms. Theories that focus on the way the environment shapes cultural representations of actors themselves (that is, theories high on the y-axis) are theories of identity. Back.

Note 5:  See Jon Elster, The Cement of Society: A Study of Social Order  (Cambridge: Cambridge University Press, 1989); and Jack P. Gibbs, Norms, Deviance, and Social Control: Conceptual Matters  (New York: Elsevier, 1981). Scholars may have different interests in studying order-some hope to undermine it, others to strengthen it. For an interpretivist assessment of "order" as a social construct, see Nicholas Onuf, World of Our Making: Rules and Rule in Social Theory and International Relations  (Columbia: University of South Carolina Press, 1989), pp. 127-59. Back.

Note 6:  Hedley Bull defines order as "an arrangement of social life such that it promotes certain goals or values" (The Anarchical Society: A Study of Order in World Politics [New York: Columbia University Press, 1977], p. 4). Back.

Note 7:  These schools-neoclassical individualism and formal structuralism-represent different extremes of the schema developed in essay 2 (figure 2.1). The former ignores the way that structures (either material or normative) constitute actors, whereas the latter ignores the cultural content of the environment. Back.

Note 8:  Gary Becker, The Economic Approach to Human Behavior  (Chicago: University of Chicago Press, 1976); Jack Hirschleifer, "The Expanding Domain of Economics," American Economic Review  (1985), pp. 53-68. Back.

Note 9:  Duncan Snidal, "The Game Theory of International Politics," in Kenneth Oye, ed., Cooperation Under Anarchy  (Princeton: Princeton University Press, 1986); Arthur Stein, Why Nations Cooperate  (Ithaca: Cornell University Press, 1990). Back.

Note 10:  Robert Gilpin, War and Change in the International System  (Princeton: Princeton University Press, 1981). See also Robert O. Keohane, Neorealism and Its Critics  (New York: Columbia University Press, 1986); and Kenneth Waltz, Theory of International Politics  (Reading, Mass.: Addison-Wesley, 1979). Back.

Note 11:  Many of the arguments in this volume distinguish themselves from rationalism, but that does not imply that they espouse "irrationalism." Rather, the point is that "rationalist" arguments are based on narrow self-interest that is influenced primarily by material-economic forces. A sociological approach rejects the basis of realist rationality-that states are driven primarily by the international power distribution-as too limited. Instead it argues that a range of social phenomena help to constitute and constrain agents. Once an agent knows what it is and what it wants, it will move instrumentally (even in the sociological model) to achieve its goals within the material and normative constraints it faces. The virtue of the sociological approach is that it fills in some of the more interesting lacunae of rationalism (e.g., the character and goals of actors). Back.

Note 12:  Albert Hirschman, "Three Ways of Complicating Some Categories of Economic Discourse," Economics and Philosophy 1 (1985): 7-21; Robert Jervis, "Realism, Game Theory, and Cooperation," World Politics 40 (1989): 317-49; and Amartya Sen, "Rational Fools," Public Affairs and Philosophy 6 (1977): 317-44. Back.

Note 13:  For discussions of how rules facilitate repeated interaction, see Robert Axelrod, The Evolution of Cooperation  (New York: Basic Books, 1984); Andrew Schotter, "The Evolution of Rules," in Richard N. Langlois, ed., Economics as a Process: Essays in the New Institutional Economics, pp. 117-33 (Cambridge: Cambridge University Press, 1986); Robert Sugden, The Economics of Rights, Co-operation, and Welfare  (Oxford: Blackwell, 1986); and R. Sugden, "Spontaneous Order," Journal of Economic Perspectives 3 (1989): 85-97.

Morishima argues that the adoption of Confucian ethics during the Meiji period produced an economically functional civic code that contributed to Japan's long-run commercial success. Similarly, Ramsey McMullen attributes the decline of the Roman Empire to a failure by the Roman state to encourage social norms capable of reining in corruption. And Greif extends this argument to the commercial success of Western traders during the Middle Ages. See Avner Greif, "Reputation and Coalitions in Medieval Trade: Evidence on the Maghribi Traders," Journal of Economic History 49 (1989): 857-82; Avner Greif, "Institutions and International Trade: Lessons from the Commerical Revolution," American Economic Review 82 (1992): 128-33; Avner Grief, "Cultural Beliefs and the Organization of Society: A Historical and Theoretical Reflection on Collectivist and Individualistic Societies," Journal of Political Economy 102 (1994): 912-50; Ramsey McMullen, Corruption and the Decline of Rome  (New Haven: Yale University Press, 1988); and Michio Morishima, Why Has Japan "Succeeded"? Western Technology and the Japanese Ethos  (Cambridge: Cambridge University Press, 1982). For a discussion of these points and for an extended account of the economic role of social norms, see also Jean-Philippe Platteau, "Behind the Market Stage Where Real Societies Exist-Part I: The Role of Public and Private Order Institutions," The Journal of Development Studies 30 (1994): 533-77; and J.-P. Platteau, "Behind the Market Stage Where Real Societies Exist-Part II: The Role of Moral Norms," The Journal of Development Studies 30 (1994): 753-817. Back.

Note 14:  Kenneth Arrow, Social Choice and Individual Values  (New Haven: Yale University Press, 1951); Kenneth Schepsle, "Institutional Arrangements and Equilibrium in Multidimensional Voting Models," American Journal of Political Science 23 (1979): 27-59; and Kenneth Schepsle and Barry Weingast, "The Institutional Foundations of Committee Power," American Political Science Review 81 (1987): 85-104. Back.

Note 15:  Douglass North, Structure and Change in Economic History  (New York: Norton, 1981); D. North, Institutions, Institutional Change, and Economic Performance  (Cambridge: Cambridge University Press, 1990); and Oliver E. Williamson, Markets and Hierarchies: Analysis and Antitrust Implications  (New York: Free Press, 1975). Back.

Note 16:  See, for example, Robert O. Keohane, International Institutions and State Power  (Boulder: Westview, 1989); and Stephen D. Krasner, ed., International Regimes  (Ithaca: Cornell University Press, 1983).

This is not true in the most "sociological" versions of regime theory. For example, Oran Young defines institutions as "recognized practices consisting of easily identifiable roles, coupled with collections of rules or conventions governing relations among the occupants of these roles" (O. Young, "International Regimes: Toward a New Theory of Institutions," World Politics 39 [1986]: 107). Young thus explicitly addresses the identities that define agent roles, but it is unusual for neoliberal institutionalists to devote much attention to the constitution of actors themselves (as opposed to the rules that govern actor behavior). Back.

Note 17:  Axelrod, The Evolution of Cooperation; Keohane, International Institutions and State Power; Oye, Cooperation Under Anarchy; Philip Pettit, "Virtus Normativa: Rational Choice Perspectives," Ethics 100 (1990): 725-55; Edna Ullmann-Margalit, The Emergence of Norms  (New York: Oxford University Press, 1977). Back.

Note 18:  To be sure, some rational choice theorists have gone beyond their customary focus on strategy in an effort to model individual preferences and beliefs. Aaron Wildavsky argued, in his 1986 presidential address at the 82d annual meeting of the American Political Science Association, that international relations theorists (and neoliberal institutionalists in particular) should devote greater attention to preference formation. See A. Wildavsky, "Choosing Preferences by Constructing Institutions: A Cultural Theory of Preference Formation," American Political Science Review 81 (1987): 3-21. See also Vinod Aggarwal, "A General Theory of Preference Formation" (manuscript); also published, in a different version, as "Obiettivi, Preferenzie, e Giochi: Verso una Teoria della Contrattazione Internazionale," in Paolo Guerrieri and Pier Carlo Padoan, eds., Politiche Economiche Nazionale e Regimi Internazionali  (Milan: Franco Angeli, 1990); Vinod Aggarwal and Pierre Allan, "Modeling Game Change: The Berlin Deadline, Berlin Wall, and Cuban Missile Crises" (paper prepared for delivery at the 1990 annual meeting of the American Political Science Association, San Francisco, August 29-September 2, 1990); and Michael D. Cohen and Robert Axelrod, "Coping with Complexity: The Adaptive Value of Changing Utility," American Economic Review 74 (1984): 30-42. But there is no obvious rational explanation for these beliefs (or for the character of actors themselves). Offering such an account, in fact, would take rational choice theorists well into the domain of "substantive" rationality (Weber's Wertrationalitat)-a move that they have, so far, wisely resisted. Back.

Note 19:  Material structures, of course, are interpreted by actors and thus "shade into" social structures. Moreover, social structures themselves may range from the formal and legalistic to the informal and "customary." Even within virtually identical formal structures, behavior may vary widely depending on informal beliefs and practice. While formal laws prohibit jaywalking in many places, for example, the practice of jaywalking varies according to cultural practice-it is widespread in some areas and rare in others. Back.

Note 20:  Other systemic theories, based on different material notions of structure, also accord only a minor role to norms and social structures. Wallerstein's modern world system approach, for example, explains international relations as a product of the material nature of the global capitalist system that leads to a hierarchy of power (i.e., core, semiperiphery, periphery) with accompanying institutions (e.g., strong and weak states). Hegemonic stability theory likewise bases systemic structure on relative power. See Robert Gilpin, U.S. Power and the Multinational Corporation  (New York: Basic Books, 1975); R. Gilpin, War and Change in the International System; and R. Gilpin, The Political Economy of International Relations  (Princeton: Princeton University Press, 1987). And other cyclic theories trace structure to the long-term fluctuations of the international economy; see Joshua Goldstein, Long Cycles: Prosperity and War in the Modern Age  (New Haven: Yale University Press, 1988). Back.

Note 21:  Another variant of this argument is the "security dilemma," which posits an international structure of insecurity independent of actors' understandings (and relying, instead, on such putatively "objective" factors as the offense-defense balance). But this balance and the dilemma itself are affected by collective beliefs that vary independently of power, as in the cases of the U.S.-British condominium before World War I and the U.S.-European alliance after World War II against the Soviet Union. See Stephen R. Rock, Why Peace Breaks Out  (Chapel Hill: University of North Carolina Press, 1989); and Stephen M. Walt, The Origins of Alliances  (Ithaca: Cornell University Press, 1987). Back.

Note 22:  For example, see Waltz on socialization and emulation and Posen's account of the sources of nationalism: Kenneth N. Waltz, "The Emerging Structure of International Politics," International Security 18 (1994): 44-79; Barry Posen, "Nationalism, the Mass Army, and Military Power," International Security 18 (1993): 80-124. Back.

Note 23:  See Robert Jervis, "Security Regimes," in Krasner, International Regimes, pp. 173-94; R. Jervis, "From Balance to Concert: A Study of International Security Cooperation," World Politics 38 (1985): 58-79; and Charles Lipson, "International Cooperation in Economic and Security Affairs," World Politics 37 (1984): 1-23. Back.

Note 24:  Peter Gourevitch, Politics in Hard Times: Comparative Responses to International Economic Crises  (Ithaca: Cornell University Press, 1986); Peter J. Katzenstein, ed., Between Power and Plenty: Foreign Economic Policies of Advanced Industrial States  (Madison: University of Wisconsin Press, 1978); and Stephen D. Krasner, Defending the National Interest  (Princeton: Princeton University Press, 1978). See also Matthew Evangelista, Innovation and the Arms Race: How the United States and the Soviet Union Develop New Military Technologies  (Ithaca: Cornell University Press, 1988). Back.

Note 25:  Barry Posen, The Sources of Military Doctrine  (Ithaca: Cornell University Press, 1984); Jack Snyder, The Ideology of the Offensive  (Ithaca: Cornell University Press, 1984). Back.

Note 26:  An exception is Katzenstein's two-volume study of the "corporatist compromise" in the small European states. These states' common ideology of democratic corporatism, resulting from their peculiar vulnerabilities, nevertheless allows for considerable variation, ranging from Swiss, Belgian, and Dutch liberal corporatism to Austrian and Scandinavian brands of social corporatism. Peter J. Katzenstein, Corporatism and Change: Austria, Switzerland, and the Politics of Industry  (Ithaca: Cornell University Press, 1984); and Katzenstein, Small States in World Markets: Industrial Policy in Europe  (Ithaca: Cornell University Press, 1985). Back.

Note 27:  Jacob Viner, "Power Versus Plenty as Objectives of Foreign Policy in the Seventeenth and Eighteenth Centuries," World Politics 1 (1948): 1-29. Back.

Note 28:  Stephen Van Evera, "The Cult of the Offensive and the Origins of the First World War," International Security 9 (1984): 58-107; and Snyder, The Ideology of  . Back.

Note 29:  Kier and Berger (and Johnston) generally use the term culture rather than discussing specific social norms or prescriptions (see the first essay for further discussion of culture and its relationship to norms and identity). But in each of these essays, there is obvious concern for the effects of norms. The military's culture prescribes how the military ought to be constituted as a professional organization and how its members ought to behave. Kier also shows that these norms were subject to vehement dispute. The culture of the French Left, for example, included very different prescriptions for the role and identity (presumably bourgeois) of the military. Thus, while culture does not always involve prescription, the arguments in these essays do clearly involve cultural norms. Back.

Note 30:  See, for example, Robert Axelrod's discussion of "metanorms" in "An Evolutionary Approach to Norms," American Political Science Review 80 (1986): 1095-111. Back.

Note 31:  Risse-Kappen notes, of course, that these efforts to strengthen the ties binding nato members were not accepted unanimously. The French-American relationship, in particular, deteriorated from this point onward. But, by the time of the Bermuda summit in March 1957, the British and the Americans had already begun to restore their "special relationship."

For an interesting related discussion of possible conflicts between norms of democracy and deterrence within nato, see Josef Joffe, "Democracy and Deterrence: What Have They Done to Each Other?" in Linda Miller and Michael J. Smith, eds., Ideas and Ideals: Essays on Politics in Honor of Stanley Hoffmann, pp. 108-26 (Boulder: Westview, 1993). Back.

Note 32:  Clifford Geertz, Negara: The Theatre State in Nineteenth-Century Bali  (Princeton: Princeton University Press, 1980). Back.

Note 33:  Edgerton, Rules, Exceptions, and Social Order, p. 28. Back.

Note 34:  In each of these cases, it is not merely identity (as a Balinese, a Japanese, or an Eskimo) that is in conflict with a particular external threat. Rather, identity is in conflict with culturally accepted practice for meeting those threats. One cannot simultaneously remain a good Balinese and allow foreign domination. It is not the fact of Dutch rule that is at issue: it is the acceptance of this rule that is incompatible with the Balinese elite's identity. The conflict, in each of these cases, is directly between behavioral rules and actor identity. The suggestion of change in the former implies the negation of the latter. Back.

Note 35:  See Gilpin, War and Change in the International System; Waltz, Theory of International Politics. Back.

Note 36:  John G. Ruggie, "International Regimes, Transactions, and Change: Embedded Liberalism in the Postwar Economic Order," in Krasner, International Regimes, pp. 195-231. Back.

Note 37:  We substitute the term norm building for norm creation because norms are rarely (if ever) created de novo. Instead, they rely on preexisting cultural knowledge and institutions. Nevertheless, the importance of particular behavioral regulations and identities does rise and fall over time. In some cases, this process may occur so rapidly that a new norm appears to be "created," as in the case of prohibitions against the use of nuclear weapons. Undeniably, this norm drew on prohibitions against the use of other "unconventional" weapons. But since not all highly destructive weapons are taboo, it is not unreasonable to ask how this particular norm, pertaining to this weapon, came into being. Although we avoid the term for the sake of clarity, we therefore do not believe it is meaningless to ask how norms are "created." Back.

Note 38:  See Harry Eckstein, "A Culturalist Theory of Political Change," American Political Science Review 82 (1988): 789-804. Back.

Note 39:  Sherif projected a dot of light on a wall in an otherwise darkened room. The autokinetic effect makes the dot appear to move, and Sherif investigated the extent to which a confederate could influence judgments about this movement. But even in this scenario, there is nothing inherently ambiguous about the situation that Sherif created. It was ambiguous only given the autokinetic effect and the visual acuity of the human subjects that Sherif used in his experiment. See Muzafer Sherif, The Psychology of Social Norms  (New York: Harper, 1939). Back.

Note 40:  See Herman, essay 8 in this volume, p. 287. Herman also notes that while virtually all New Thinkers he interviewed invoked the idea of a "normal country" to describe their hopes for their nation's future, "not all of them meant precisely the same thing" (Herman, essay 8 in this volume, p. 294). Here again ambiguity serves a useful role. Back.

Note 41:  Finnemore also notes that the passage of time permits greater institutionalization in formal international organizations such as the United Nations; see Finnemore, essay 5, p. 160. Back.

Note 42:  Berger's finding that support for new political-military cultures in Germany and Japan increased slowly over time also fits well with the iteration hypothesis. See Price and Tannenwald, essay 4, and Berger, essay 9. Back.

Note 43:  Nor is it sufficient to explain away the decline of the "aerial bombardment of undefended areas" prohibition simply by pointing out that its frequent violation undermined the norm. Chemical weapons, after all, were widely used in World War I; yet the taboo against chemical weapons is now strong, while aerial bombardment has become permissible.

Some might even argue that this norm, while often violated, remains in effect. To be sure, attacking undefended noncombatants would appear to violate current norms of warfare. But it seems difficult to argue that attacking undefended cities or buildings is a similar violation in light of the role that so many civilian technologies (telecommunications, power generation, etc.) can play in modern war-fare. The aerial bombardment of Iraq during the Persian Gulf war-and particularly the American emphasis on precision-guided munitions targeted at specific buildings-illustrates this distinction. Back.

Note 44:  Mancur Olson, The Rise and Decline of Nations: Economic Growth, Stagflation, and Social Rigidities  (New Haven: Yale University Press, 1982). See also Stephen Krasner, "State Power and the Structure of Foreign Trade," World Politics 28 (1976): 317-47. On the twin shocks of World War I and the Depression, see Jeff Frieden, "Sectoral Conflict and U.S. Foreign Economic Policy, 1914-1940," in G. John Ikenberry, David Lake, and Michael Mastanduno, eds., The State and American Foreign Economic Policy, pp. 59-90 (Ithaca: Cornell University Press, 1988). And on the oil shocks of the 1970s, see G. John Ikenberry, "Market Solutions for State Problems: The International and Domestic Politics of American Oil Decontrol," in Ikenberry, Lake, and Mastanduno, The State and American Foreign Economic Policy, pp. 151-77. Back.

Note 45:  For an account of norm creation as a problem of strategic interaction, see Arthur Stein, "Normative Equilibrium and International Politics" (paper presented at the conference "Norms and Emulation in International and Domestic Behavior," Palm Springs, Calif., January 13-16, 1994). Back.

Note 46:  This notion has been developed in some detail by sociologists and by political scientists who focus on "epistemic communities." See Mary Fennell and Richard Warnecke, The Diffusion of Medical Innovations: An Applied Network Analysis  (New York: Plenum Press, 1988); Claude Fischer and Glenn Carroll, "Telephone and Automobile Diffusion in the United States, 1902-1937," American Journal of Sociology 93 (1988): 1153-78; Peter Hall, The Political Power of Economic Ideas: Keynesianism Across Nations  (Princeton: Princeton University Press, 1989); and "Knowledge, Power, and International Policy Coordination," International Organization 46 (special issue, 1992): Back.

Note 47:  Herman, essay 8 in this volume, p. 310. Back.

Note 48:  For a belief (or norm) to be shared does not necessarily mean that it is institutionalized or, in some other way, a property of a collective. On this point, see also Ronald L. Jepperson and Ann Swidler, "What Properties of Culture Should We Measure?" Poetics 22 (1993/94). Saying that a certain percentage of Americans have patriotic beliefs is different, for example, from saying that Americans as such are patriotic. Only the latter is a statement of a collective norm. But despite this discussion, it often seems difficult to distinguish between the two in practice. Is New Thinking the shared representation of a new foreign policy agenda among Soviet elites or is it a property of some collective, ultimately institutionalized in perestroika? Perhaps it is both. But, if so, it is not clear what evidence of its presence would help us to distinguish between its shared existence and its collective existence. Back.

Note 49:  Stephen Walker, ed., Role Theory and Foreign Policy Analysis  (Durham: Duke University Press, 1987); and S. Walker, "Symbolic Interactionism and International Politics: Role Theory's Contribution to International Organization," in Martha Cottam and Chih-yu Shih, eds., Contending Dramas: A Cognitive Approach to International Organization, pp. 19-38 (New York: Praeger, 1992). Back.

Note 50:  Kal J. Holsti, "National Role Conceptions in the Study of Foreign Policy," International Studies Quarterly 14 (1970): 233-309. Back.

Note 51:  See Richard K. Herrmann, Perceptions and Behavior in Soviet Foreign Policy  (Pittsburgh: University of Pittsburgh Press, 1985); R. K. Herrmann, "The Power of Perceptions in Foreign-Policy Decision Making: Do Views of the Soviet Union Determine the Policy Choices of American Leaders?" American Journal of Political Science 30 (1986): 841-75; and Richard K. Herrmann and Michael P. Fischerkeller, "A Cognitive Strategic Approach to International Relations: Theory and Practice in the Persian Gulf" (manuscript, 1994).

Herrmann argues that states, in the course of their interaction, make judgments about their relative capabilities (superior, comparable, or inferior), cultures (superior, comparable, or inferior), and interactions (threat, opportunity, or mutual interest). Different combinations of these judgments yield twenty-seven categories. A smaller number are likely to dominate, however, because some combinations make little sense. It seems unlikely, for example, that an "enemy" should simultaneously be viewed as posing a "high" threat and yet having "low" capabilities. One of the components of such a role image will, over time, be revised. Some will object to Herrmann's and other similar approaches to identity on the grounds that they overlook "intrinsic" components of identity, focusing instead only on relational aspects of identity definition. Whether identity is more an intrinsic property of actors or a function of their social interaction and environment is impossible to say. At the extreme, however, the contention that identity is fully and solely intrinsic precludes any investigation of its sources-they would consequently be held unknowable, as in rationalist theory, discussed above. This contention would have the effect, therefore, of foreclosing the very kind of inquiry that this volume (and this essay, in particular) undertakes. Back.

Note 52:  Martin W. Sampson and Stephen W. Walker, "Cultural Norms and National Roles: A Comparison of Japan and France," in Walker, Role Theory and Foreign Policy Analysis, pp. 105-22. Back.

Note 53:  Goldstein and Keohane, Ideas and Foreign Policy. Back.

Note 54:  Eyre and Suchman, Finnemore, and Risse-Kappen also distance their arguments from rational choice approaches. Back.

Note 55:  Douglas W. Blum, "The Soviet Foreign Policy Belief System: Beliefs, Politics, and Foreign Policy Outcomes," International Studies Quarterly 37 (1993): 373-94. Back.

Note 56:  See Johnston, essay 7 in this volume; also see Katzenstein's discussion in essay 13 of efforts by both rational choice and cognitive theorists to provide a microfoundation for the analysis of collective norms. Back.

Note 57:  On in-group and out-group identities, see Richard Cottam, Foreign Policy Motivation: A General Theory and a Case Study  (Pittsburgh: University of Pittsburgh Press, 1977); Paul A. Kowert, "The Cognitive Origins of International Norms: Identity Norms and the 1956 Suez Crisis" (paper presented at the 35th annual meeting of the International Studies Association, Washington, D.C., March 28-April 1, 1994); Penelope Oakes and John C. Turner, "Distinctiveness and the Salience of Social Category Memberships: Is There an Automatic Perceptual Bias Towards Novelty?" European Journal of Social Psychology 16 (1986): 325-44; Henri Tajfel, ed., Differentiation Between Social Groups: Studies in the Social Psychology of Intergroup Relations  (London: Academic Press, 1978); H. Tajfel, Human Groups and Social Categories: Studies in Social Psychology  (Cambridge: Cambridge University Press, 1981); and H. Tajfel et al., "Social Categorization and Intergroup Behavior," European Journal of Social Psychology 1 (1971): 149-78. Back.

Note 58:  Marilyn Brewer, "The Social Self: On Being the Same and Different at the Same Time," Personality and Social Psychology Bulletin 17 (1991): 475-82; and M. Brewer, "The Role of Distinctiveness in Social Identity and Group Behavior," in Michael Hogg and Dominic Abrams, eds., Group Motivation: Social Psychological Perspectives  (London: Harvester Wheatsheaf, 1993). Back.

Note 59:  John C. Turner, "Towards a Cognitive Redefinition of the Social Group," in Henri Tajfel, ed., Social Identity and Intergroup Relations, p. 16 (Cambridge: Cambridge University Press, 1982). For a comparative analysis of several other cognitive theories of group identification, see Richard R. Lau, "Individual and Contextual Influences on Group Identification," Social Psychology Quarterly 52 (1989): 220-31. Back.

Note 60:  This discussion of internal psychological processes has, for reasons of brevity, focused only on the microfoundation of identity. But cognitive and motivational psychologists have also undertaken numerous research projects investigating the construction of behavioral norms. See, among others, Jerry Burger, "Desire for Control and Conformity to a Perceived Norm," Journal of Personality and Social Psychology 53 (1987): 355-60; Martha Cottam, "Problem Representation After the Cold War" (paper presented at the 35th annual meeting of the International Studies Association, Washington, D.C., March 28-April 1, 1994); Daniel Kahneman and Dale Miller, "Norm Theory: Comparing Reality to Its Alternatives," Psychological Review 93 (1986): 136-53; Brian Ripley and Dwain Mefford, "The Cognitive Foundations of Regime Theory" (paper prepared for delivery at the 1987 annual meeting of the American Political Science Association, Chicago, 1987); Donald Sylvan and Stuart Thorson, "Ontologies, Problem Representation, and the Cuban Missile Crisis," Journal of Conflict Resolution 36 (1992): 709-32; and Stephen Walker, "The Motivational Foundations of Political Belief Systems: A Re-Analysis of the Operational Code Construct," International Studies Quarterly 27 (1983): 179-202. Jonathan Mercer has also employed Turner's self-categorization theory to argue that both the identities and the behavioral repertoires of nation-states are likely to be constructed in ways that mirror common realist assumptions; see J. Mercer, "Anarchy and Identity," International Organization 49 (1995): 229-52. Back.

Note 61:  Our claim here is more modest than it may seem. We do not contend that language can be reduced entirely to the behavior of individuals or that individuals are even "primary." It may largely rely, instead, on collective representations that exist completely apart from the thoughts or actions of any particular person. But the act of speech itself does begin with the human (biological and psychological) capacity for speech. In the following discussion, we briefly consider several regularities in the processes of speech that may serve to generate norms. Back.

Note 62:  Livia Polanyi et al., "Retelling Cold War Stories: Uncovering Cultural Meanings with Linguistic Discourse Analysis" (working paper, "Security Discourse in the Cold War Era," Center for Studies of Social Change, New School for Social Research, New York, 1993), p. 8. Back.

Note 63:  For related work, see Hayward Alker, "Fairy Tales, Tragedies, and World Histories: Towards Interpretive Story Grammars as Possibilist World Models," Behaviormetrika 21 (1987): 1-28; H. Alker et al., "Text Modeling for International Politics: A Tourist's Guide to relatus," in Valerie M. Hudson, ed., Artificial Intelligence and International Politics  (Boulder: Westview, 1991), pp. 97-126; and Sergio Alvarado, Understanding Editorial Text: A Computer Model of Argument Comprehension  (Boston: Kluwer Academic Publishers, 1990). Back.

Note 64:  Kratochwil, Rules, Norms, and Decisions; and Onuf, World of Our Making. Back.

Note 65:  On the semantic dimensions of language, see Kratochwil, Rules, Norms, and Decisions. Also see Onuf, World of Our Making, pp. 66-95. Neither Kratochwil nor Onuf would classify his own work under the rubric of "individual-level processes." To do so would be akin to suggesting that a language can be private  (a claim that Onuf specifically rejects; see Onuf, World of Our Making, pp. 47-49, esp. n. 12, and pp. 78-81). But our claim here is not that language or the social rules it engenders are reducible to individual behavior. Rather, they stem from a process of communication that, like processes of thinking, feeling, and (rationally) choosing, begins its operation at an individual level. For a related effort to bridge the gap between psychology and discourse, focusing on the concept of voice-which once again grows out of individual speech acts but which also acquires collective, cultural content-see the discussion of identity politics in Edward Sampson, "Identity Politics: Challenges to Psychology's Understanding," American Psychologist 48 (1993): 1219-30. Back.

Note 66:  Hayward Alker, Roger Hurwitz, and Karen Rothkin, "Fairy Tales Can Come True: Narrative Constructions of Prisoner's Dilemma Game Play" (paper presented at the annual meeting of the American Economic Association, Anaheim, Calif., January 5, 1993). Back.

Note 67:  See Platteau, "Behind the Market Stage Where Real Societies Exist-Part I"; Platteau, "Behind the Market Stage Where Real Societies Exist-Part II"; Sugden, The Economics of Rights, Co-operation, and Welfare; and Ullmann-Margalit, The Emergence of Norms. Back.

Note 68:  Robert O. Keohane, After Hegemony: Cooperation and Discord in the World Political Economy  (Princeton: Princeton University Press, 1984); and Krasner, International Regimes. Back.

Note 69:  Price and Tannenwald (essay 4 in this volume) give an example of the role of agency in creating a norm that turned out not to be very functional from the point of view of the agent. They observe that the efforts of the chemical industry to play up developments in chemical warfare in order to secure tariff protection facilitated the creation of a chemical weapons taboo-something that the industry evidently had not expected and that ultimately ran counter to its business interests. Back.

Note 70:  Deborah Avant, "The New Institutional Economics and the Origin of Norms of Warfare" (paper presented at the 35th annual meeting of the International Studies Association, Washington, D.C., March 28-April 1, 1994); Keohane, After Hegemony; North, Institutions, Institutional Change, and Economic Performance; and Katja Weber, "Hierarchy Amidst Anarchy: Transaction Costs and International Cooperation" (Ph.D. diss., University of California, Los Angeles, 1992). Back.

Note 71:  Arrow, Social Choice and Individual Values. Back.

Note 72:  Elster, The Cement of Society. Back.

Note 73:  Alexander Field, "Microeconomics, Norms, and Rationality," Economic Development and Cultural Change 32 (1984): 683-711. Back.

Note 74:  For a recent, general critique of the rationalist research program, see Donald P. Green and Ian Shapiro, Pathologies of Rational Choice Theory  (New Haven: Yale University Press, 1994). Back.

Note 75:  Keohane, International Institutions and State Power, pp. 4-5; Kratochwil, Rules, Norms, and Decisions, p. 62. Back.

Note 76:  This problem has also plagued the neoliberal institutionalist research program. It has been difficult to assess regime strength without referring to criteria that are conceptually related to regime effects. Discussions of different criteria for assessing institutionalization include: Keohane, International Institutions and State Power, pp. 4-5; Joseph Nye, "Nuclear Learning and U.S.-Soviet Security Regimes," International Organization 41 (1987): 375; and Roger Smith, "Institutionalization as a Measure of Regime Stability: Insights for International Regime Analysis from the Study of Domestic Politics," Millennium: Journal of International Studies 18 (1989): 234-36. For a related discussion in comparative politics, see Samuel Huntington, Political Order in Changing Societies  (New Haven: Yale University Press, 1968). Back.

Note 77:  Snidal, "The Game Theory of International Politics." Back.

Note 78:  See Blake and Davis, "Norms, Values, and Sanctions"; Edgerton, Rules, Exceptions, and Social Order; and Friedrich Kratochwil and John G. Ruggie, "International Organization: A State of the Art on an Art of the State," International Organization 40, no. 4 (Autumn 1986): 753-76. Back.

Note 79:  See, for example, James Mayall, Nationalism and International Society  (Cambridge: Cambridge University Press, 1990); James Lee Ray, "The Abolition of Slavery and the End of International War," International Organization 43 (1989): 405-39; Janice E. Thomson, Mercenaries, Pirates, and Sovereigns  (Princeton: Princeton University Press, 1994). Back.

Note 80:  In the case of prohibitions during World War II, see Jeffrey W. Legro, Cooperation Under Fire: Anglo-German Restraint During World War II  (Ithaca: Cornell University Press, 1995). Back.

Note 81:  A related bias in the study of norms is the "good norms" problem. Analysts tend to focus on those issues that are normatively desirable-e.g., the spread of democracy, the rise of human rights, the integration of world society, and prohibitions against the use of force. Yet undesirable norms are equally possible. Examples include norms of military autonomy and the use of force, economic domination, the acceptability of intrastate violence (e.g., civil war), and the disintegrative tendencies that exist in international politics (e.g., nationalism, religious exclusivity). These issues too deserve attention from the emerging sociological approach. By focusing on security affairs, this volume takes a step toward exploring a fuller range of normative consequences. But "bad" or threatening norms remain understudied. Back.

Note 82:  Diana Richards, "Interests and Identities in International Relations Theory" (paper presented at the annual meeting of the Midwest Political Science Association, Chicago, April 6-8, 1995).

The analogous problem for rationalist (economic individualist) international relations theory is the prospect of an "embarrassment of interests" in which any behavior can be explained by reference to some interest (often inferred ex post facto from the behavior). Back.

Note 83:  For a lucid discussion of these problems, see Stephen Welch, The Concept of Political Culture  (New York: St. Martin's Press, 1993). Back.

Note 84:  Finnemore might reasonably object that she never claims to make a deterministic argument about whether, when, or which norms will prevail. On the contrary, she specifically disavows such an intention because of the multiplicity of relevant norms and because of exogenous influences on state interests and behavior. Our point, therefore, is the same as Finnemore's-that multiple and conflicting international norms limit the prospects for predictive generalization. Back.

Note 85:  A constructivist might offer a much different solution to this problem. The meaning and force of rules, from a constructivist perspective, rely on the possibility of their violation. Affirming a norm thus requires its potential negation-otherwise the norm ceases to have meaning. In a more general sense, all norms must therefore explain their violation as well as their affirmation. David Campbell argues, in this vein, that the modern discourse of international security and foreign policy depends on (and therefore must define and create) insecurity and foreignness; see David Campbell, Writing Security: United States Foreign Policy and the Politics of Identity  (Minneapolis: University of Minnesota Press, 1992). From this perspective, the causal questions that concern the authors of this volume simply become irrelevant when understood in a larger interpretive context. Back.

Note 86:  Eyre and Suchman note, citing Anthony Giddens, that the current world system can actually be decomposed into several systems: an information system, a nation-state system, a capitalist economy, and so on. They warn that we should not exaggerate the extent to which these systems are integrated. See Eyre and Suchman, essay 3 in this volume; and A. Giddens, The Nation-State and Violence  (Berkeley: University of California Press, 1987). Back.

Note 87:  Swidler labels a less functionalist alternative to the traditional realist framework "sociological realism." See Ann Swidler, "The Ideal Society," American Behavioral Scientist 34 (1991): 563-80. Back.

Note 88:  Eckstein, "A Culturalist Theory of Political Change." Back.

Note 89:  Ann Swidler, "Culture in Action: Symbols and Strategies," American Sociological Review 51 (1986): 280. Back.

Note 90:  Eckstein, "A Culturalist Theory of Political Change." Back.

Note 91:  John G. Ruggie, "Continuity and Transformation in the World Polity: Toward a Neorealist Synthesis," World Politics 35, no. 2 (January 1983): 261-85; J. Ruggie, "Territoriality and Beyond: Problematizing Modernity in International Relations," International Organization 47 (1993): 845-65; Alexander Wendt, "Anarchy Is What States Make of It: The Social Construction of Power Politics," International Organization 46, no. 2 (Spring 1992), pp. 391-425; and Alexander Wendt, "Collective Identity Formation and the International State," American Political Science Review 88, no. 2 (June 1994): 384-96. Back.

Note 92:  As a possible first step toward such generalization, some theorists emphasize the way that small historical events can have enormous and persistent implications. Thus, an appreciation of historical contingency does not require that anything could happen. See, for example, the discussion of path dependence and punctuated equilibrium in Brian Arthur, "Competing Technologies, Increasing Returns, and Lock-in by Historical Events," Economic Journal 99 (1989): 116-31. Also see Stephen Krasner, "Sovereignty: An Institutional Perspective," in James Caporaso, ed., The Elusive State: International and Comparative Perspectives  (Newbury Park, Calif.: Sage, 1989). Back.

Note 93:  George Tsebelis, Nested Games: Rational Choice in Comparative Politics  (Berkeley: University of California Press, 1990). Back.

Note 94:  A determined constructivist might argue that, even in the case of nuclear weapons, technology was not initially decisive. Indeed, postwar U.S. presidents have often worried that nuclear weapons will not be sufficient to deter the threat of the day. We should also point out that Kier's argument applies equally to nuclear weapons-the technology does not clearly favor a particular military doctrine apart from the various (different) ways it is socially interpreted. Back.

Note 95:  To take one example of such a synthesis, Keegan offers an intriguing historical account of the interaction between cultures of warfare and technology; see John Keegan, A History of Warfare  (New York: Knopf, 1993). Back.

Note 96:  But he also cautioned against assuming that moral limitations play no role in international politics; see Hans Morgenthau, Politics Among Nations  (New York: Knopf, 1973), pp. 230-31. Back.

Note 97:  A nice statement of this point is Foucault's claim that "the successes of history belong to those who are capable of seizing these rules, to replace those who had used them, to disguise themselves so as to pervert them, invert their meaning, and redirect them against those who had initially imposed them." See Michel Foucault, "Nietzsche, Genealogy, History," in Paul Rabinow, ed., The Foucault Reader, pp. 85-86 (New York: Pantheon, 1984); also see Richard Price, "A Genealogy of the Chemical Weapons Taboo," International Organization 49 (1995): 73-103. Back.

Note 98:  Legro, Cooperation Under Fire. Back.

Note 99:  Herman would also argue that the Soviet reformers took political risks for which they occasionally paid a price (personal communication with the authors). We do not mean to suggest, therefore, that the norms they promoted can easily be construed as completely self-serving. Especially during the early years of the Reagan administration, the reformers left themselves open to harsh criticism from Soviet conservatives. Back.

Note 100:  See, however, Finnemore's distinction between identity and identification. Back.

Note 101:  On such "external" representations, see Erving Goffman, The Presentation of Self in Everyday Life  (Garden City, N.Y.: Doubleday, 1959); and A. Paul Hare and Herbert Blumberg, Dramaturgical Analysis of Social Interaction  (New York: Praeger, 1988). Back.

Note 102:  In fact, these manipulations may come to matter even to the manipulators; see Kevin McKillop, Michael Berzonsky, and Barry Schlenker, "The Impact of Self-Presentations on Self-Beliefs: Effects of Social Identity and Self-Presentational Context," Journal of Personality 60 (1992): 789-808; and Barry Schlenker and J. T. Trudeau, "The Impact of Self-Presentations on Private Self-Beliefs: Effects of Prior Self-Beliefs and Misattribution," Journal of Personality and Social Psychology 58 (1990): 22-32. For similar dynamics in "myths" and strategic culture, see Jack Snyder, Myths of Empire  (Ithaca: Cornell University Press, 1991); and Charles A. Kupchan, The Vulnerability of Empire  (Ithaca: Cornell University Press, 1994). Back.

Note 103:  Anthony Giddens, The Constitution of Society: Outline of the Theory of Structuration  (Cambridge: Cambridge University Press, 1984); and A. Giddens,

Note 99:New Rules of Sociological Method: A Positive Critique of Interpretative Sociologies  (Stanford: Stanford University Press, 1993). Back.

Note 104:  Thomas S. Kuhn, The Structure of Scientific Revolutions  (Chicago: University of Chicago Press, 1962). See also Hayward Alker, "The Return of Practical Reason" (unpublished review article, 1993); and Martin Hollis, The Cunning of Reason  (Cambridge: Cambridge University Press, 1987). Back.

Note 105:  What counts as a "better" historical argument is too philosophical a question to examine here in any detail. Simply put, however, "good" arguments are a matter of interpretation within the culture of social scientists. This certainly does not mean that all claims to historical "truth" are equally valid; social science offers many criteria for evaluating the performance of theories. These criteria are the social norms that make the professional study of international relations possible. Back.

Note 106:  Nor would a realist argue that states will enter into alliance relationships for reasons of cultural affinity or normative symbolism. Barnett and Risse-Kappen both contrast their approaches, therefore, to realist theories of alliance formation such as that proposed by Walt. See Walt, The Origins of . For their careful readings and critical comments of previous drafts, I would like to thank the members of the project; the participants at the Social Science Research Council/MacArthur Workshops at the University of Minnesota and Stanford University in 1994; and Emanuel Adler, Tom Christensen, Lynn Eden, Gunther Hellmann, Ron Jepperson, Mary F. Katzenstein, Robert Keohane, Jonathan Kirshner, Audie Klotz, Atul Kohli, Henry Nau, Jack Snyder, Janice Thomson, Stephen Walt, Alexander Wendt, and two anonymous readers for Columbia University Press. Back.

---

The Culture of National Security