Tübinger Arbeitspapiere zur Internationalen Politik und Friedensforschung

Nr. 34a

Henning Boekle,Volker Rittberger, Wolfgang Wagner

Norms and Foreign Policy:

Constructivist Foreign Policy Theory

Center for International Relations/Peace and Conflict Studies,

Institute for Political Science,

University of Tübingen

Address: Melanchthonstr. 36, D-72074 Tübingen

Phone: ++49 (0)7071 29-78372

Fax: ++49 (0)7071 29-2417

WWW Homepage: http://www.uni-tuebingen.de/uni/spi/ab2menu.htm

Copyright: H. Boekle, V. Rittberger, W.Wagner

Tübingen 1999

ISBN 3-927604-32-1

Editor: Volker Rittberger

Redaktion: Klaus Stodick
WWW-Layout: Jürgen Plieninger

1. Summary

2. Introduction

3. Outline of a Constructivist Theory of Foreign Policy
3.1. The Logic of Appropriateness: Norms as Independent Variables
3.2. Commonality and Specificity
3.2.1. Commonality
3.2.2. Specificity
3.3. Norms and Foreign Policy Behavior: Causal Mechanisms

4. Transnational and Societal Constructivism

4.1. Transnational Constructivism
4.1.1. The Structural Model of International Society
4.1.2. International Norms
4.1.3. The Role and Significance of International Institutions
4.2. Societal Constructivism
4.2.1. Norms Shared by Experts
4.2.2. Norms Shared by Society as a Whole
4.2.3. A Robust Societal-Constructivist Explanation
4.2.4. The Transfer of Domestically Valid Norms to Areas of Foreign Policy: The Domestic Analogy

5. Identification of Independent Variables

5.1. International Norms
5.1.1. International Law
5.1.2. Legal Acts of International Organizations
5.1.3. Final Acts of International Conferences
5.2. Identification of Societal Norms
5.2.1. Survey Data
5.2.2. The Constitutional and Legal Order of a Society
5.2.3. Party Programs and Election Platforms
5.2.4. Parliamentary Debates

6. Conceptualization of the Dependent Variable and General Constructivist Prediction
6.1. Norm-Consistent Foreign Policy
6.2. General Constructivist Prediction

References

Footnotes

1. Summary

During the last decade, constructivism has evolved as the main challenger to rationalist theories of international relations and foreign policy. As a social theory, constructivism refutes the basic assumption of rationalist theories that actors pursue their exogenously determined preferences according to a logic of consequentiality. Instead, in its explanation of foreign policy behavior constructivism assumes the working of a logic of appropriateness. Norms, i.e. value-based, shared expectations about appropriate behavior, are the independent variable of constructivist foreign policy theory. Norms shape actors' identities and preferences, define collective goals and prescribe or proscribe behavior.

Constructivist foreign policy theory draws upon two research traditions. Transnational constructivism emphasizes the influence of norms that are shared by international society or by subsets of that society as embodied by regional or function-specific international organizations. International law, resolutions of international organizations and final acts of international conferences are the indicators for international norms. Societal constructivism, on the other hand, stresses the importance of norms that are shared within domestic society. Indicators for societal norms are the constitutional and legal order, party programs and election platforms, parliamentary debates, and public opinion data.

In order to arrive at sound theory-based predictions about, and explanations of, German foreign policy behavior, constructivism must be able to identify ex ante the norms which pertain to the specific context of Germany's foreign policy behavior which it seeks to explain. Two properties of norms serve as criteria for an assessment of their relative strength. The first is the commonality of a norm, i.e. the degree to which it is shared among the units of a social system. The second criterion is its specificity, i.e. the clarity by which a norm discriminates between appropriate and inappropriate behavior. Thus, a norm must have at least a medium level of both commonality and specificity if a constructivist prediction or explanation is to be based on it.

Since constructivism posits that foreign policy actors abide by international and/or societal norms, it refutes the neorealist claim that German foreign policy behavior will change due to Germany's power increase in the wake of the end of the Cold War and of unification. According to constructivism, Germany's foreign policy is likely to change only to the extent to which the relevant norms have themselves changed.

2. Introduction [ 1 ]

The unification of Germany and the end of the East-West conflict represent a turning point which has brought the issue of continuity versus change to the center of research on German foreign policy. The period 1989-1991 is especially significant for the neorealist theory of foreign policy because, when seen from this perspective, it signifies a strengthening of Germany's power position which leads one to expect a change in German behavior toward "more power politics" (see Baumann/ Rittberger/Wagner 1998; Krasner 1993; Rittberger 1992; Mearsheimer 1990). The counter-argument, i.e. the expectation of continuity in German foreign policy, has been put forward, among others, by representatives of "constructivist" analytical approaches. They regard state action not as dependent on a state's power position but as guided by norms. From a "constructivist" [ 2 ] perspective, Germany’s continued enmeshment in a network of international (normative) institutions and its unaltered societal norms mean that a significant change in its behavior is not to be expected (see Katzenstein 1997; Anderson/Goodman 1993).

In our view, one shortcoming of this debate about German foreign policy after unification is that, as a rule, international and societal norms are assumed to have remained constant and that instances of continuity in German behavior are uncritically assumed to evidence the explanatory power of a constructivist theory of foreign policy. [ 3 ] However, if the values of the independent variable "norms" during the period under review are not assessed empirically, such conclusions remain unsubstantiated. For this reason, one of the main emphases of this paper is a consideration of the empirical assessment of international (5.1.) and societal (5.2.) norms affecting German foreign policy and of how to evaluate their explanatory power. Before doing this, however, the essential features of a constructivist theory of foreign policy will be presented and discussed (3.) with reference to both the transnational (4.1.) and the societal (4.2.) constructivist research traditions.

3. Outline of a Constructivist Theory of Foreign Policy

3.1. The Logic of Appropriateness: Norms as Independent Variables

The point of departure for a constructivist theory of foreign policy is its critique of the concept of utility-maximizing homo oeconomicus which is at the core of neorealist and utilitarian-liberal analyses of foreign policy (see Baumann/Rittberger/Wagner 1998; Bienen/Freund/Rittberger 1999). According to this concept, ideas, values or norms can only play a role as instruments for asserting and justifying given interests. Constructivist theory of foreign policy, by contrast, emphasizes the independent influence of these variables. According to the constructivist view, actors’ actions are guided by norms, i.e. by intersubjectively shared, value-based expectations of appropriate behavior. [ 4 ] The assumption of the independent influence of norms is incompatible with the concept of the self-regarding, rational homo oeconomicus. This concept is replaced by an actor concept described as homo sociologicus or role player (Hasenclever/Rittberger/Mayer 1997: 155; Schaber/Ulbert 1994). In the constructivist view, actors take decisions "on the basis of norms and rules on the background of subjective factors, historical-cultural experience and institutional involvement" (Schaber/Ulbert 1994: 142). When faced with various alternative courses of action, homo oeconomicus considers the anticipated consequences of his action in order to choose the alternative which will maximize his [ 5 ] self-regarding utility; homo sociologicus, on the other hand, bases his action on a "logic of appropriateness", which takes socially shared, value-based expectations of behavior as its point of reference. The logic of appropriateness states that

"behaviors (beliefs as well as actions) are intentional but not willful. They involve fulfilling the obligations of a role in a situation, and so of trying to determine the imperatives of holding a position. [...] Within a logic of appropriateness, a sane person is one who is ‘in touch with identity’ in the sense of maintaining consistency between behavior and a conception of self in a social role" (March/Olsen 1989: 160f.; see also Hasenclever/Mayer/Rittberger 1997: 155-157, Finnemore 1996a: 28-31 and Zürn 1992: 68-69).

The conceptualization of this logic of action as a logic of appropriateness has advantages in comparison with a conceptualization as value-rationality according to Max Weber who has defined value rationality as

"determined [...] by a conscious belief in the value for its own sake of some ethical, aesthetic, religious, or other form of behavior, independently of its prospects of success" (Weber 1968: 24f.).

On the one hand, the concept of the logic of appropriateness better expresses that what matters is not the value preferences of individuals, but the inter-subjectively shared, value-based expectations of appropriate behavior. On the other hand, constructivists hold that conscious belief in the intrinsic value of an action is not a necessary condition of norm-guided behavior. Within a social system, expectations of appropriate behavior can come to be taken for granted and their intrinsic value will not longer consciously be reflected by individual actors (see Finnemore 1996a: 23). Weber has called this mode of action, which is "determined by ingrained habituation" (Weber 1968: 25), traditional action, and has distinguished it from value-rational action. However, as constructivist theory of foreign policy seeks to include the influence of traditional cultural norms on states’ actions, there is no reason to exclude traditional action per definitionem. Thus, constructivist foreign policy theory considers both the value-based behavioral expectations that decision-makers are aware of and those that belong to cultural tradition and have come to be taken for granted. Both are included in the concept of the logic of appropriateness.

In the constructivist theory of foreign policy, social norms – defined as intersubjectively shared, value-based expectations of appropriate behavior – serve as independent variables for explanations of foreign policy behavior. Norms are distinguished from other ideational variables by virtue of their characteristics: (1) intersubjectivity, (2) immediate orientation to behavior (see Finnemore 1996a: 22f.; Florini 1996: 164), and (3) reference to values and counterfactual validity (see Hasenclever/Mayer/Rittberger 1997: 164f.; Goertz/Diehl 1992: 638f.; Kratochwil/Ruggie 1986: 767f.).

The characteristic of intersubjectivity distinguishes norms from individual convictions, and thus from ideas which have been described as "beliefs held by individuals" (Goldstein/Keohane 1993: 3). [ 6 ] Cognitive analytical approaches examine the influence of these individual and subjective "belief systems" (see Little/Smith 1988). Although the proponents of cognitive theories do not dispute the social origins of individual convictions and values, they regard the individual form of convictions held by individual decision-makers as exercising a decisive influence on foreign policy behavior. More or less explicitly, therefore, the convictions of individual decision-makers are ascribed a great degree of autonomy vis-a-vis their social environment.

One example of an attempt to provide a cognitive account of German foreign policy is Thomas Banchoff’s (1997) study of German EU policy after 1990. Banchoff ascribes the German federal government’s policy of bringing about further European integration to "Helmut Kohl's historical ideas" (Banchoff 1997: 66). However, one objection to this explanation is that "examining decision-making processes through individual motivation and cognition alone ignores the commonality of shared norms underlying dominant ideas or knowledge" (Klotz 1995: 32). Accordingly, it is not Kohl’s personal "belief system" that explains German EU policy but the social consensus on which that system is based and which it represents. Explanations which attribute a state’s foreign policy behavior to the "belief systems" of individual personalities are unsatisfactory in that they always raise the question of the social roots of individual convictions without themselves being able to answer it. Cognitive approaches will therefore not be considered in the present paper.

The second defining characteristic, that of immediate orientation to behavior, also distinguishes norms from ideas, values and ‘causal beliefs’. In addition, it distinguishes norms from world views (see Goldstein/Keohane 1993) and from principles in the sense of "beliefs of fact, causation, and rectitude" (Krasner 1983: 2). World views are comprehensive conceptions of reality which generally include ‘causal beliefs’ and ‘principled beliefs’ while at the same time transcending them. For generating actual expectations of behavior, however, they are too abstract. Principles, in Krasner’s sense, comprise the entire range of Goldstein and Keohane’s (1993) concept of ideas but differ from ideas in that they contain "descriptions of facts, goals and end-means relationships about which actors are in agreement" (Müller 1993a: 39; our italics). Therefore, while principles per definitionem have an intersubjective quality, they do not contain an explicit expectation of behavior even if they, as values, describe desirable or non-desirable conditions or goals. The statement "lying is bad" embodies such a general statement of value as a principle, while the commandment "Thou shalt not lie" is a concrete, socially shared, value-based expectation of behavior, i.e. a norm. As this example shows, norms do not contain any explicit value-judgement of an action (even though they are, of course, implicitly based on such a judgement); they merely contain the expectation, addressed to an actor, to perform an "appropriate" action or desist from an "inappropriate" one.

Finally, the third defining characteristic is that norms always involve a value reference and therefore have counterfactual validity. As explained above, the value reference is not necessarily the explicit but may merely be the implicit point of reference of an expectation of behavior: "[...] there are issues of justice and rights of a moral or ethical character" (Goertz/Diehl 1992: 638-639). This characteristic distinguishes norms from the type of non-value-related expectations of behavior that can arise, for example, from "causal beliefs". Because of their reference to values, norms possess a "compliance pull" independent from interests (Hurrell 1993; Franck 1990). This "compliance pull" does not have to cause uniform norm-compliant behavior by actors within a given context of action. Yet norms also possess a deontological quality. They are counterfactually valid so that the existence of a norm does not have to be called into question if it is occasionally violated (see Hasenclever/Mayer/Rittberger 1997: 164f.; Goertz/Diehl 1992: 638f.; Kratochwil/Ruggie 1986: 767f.). Of course, the assumption of the deontological quality of norms and their counterfactual validity must not be extended so far that any moral demand made at any time by any actor is uncritically ascribed the status of a norm, despite the fact that it is largely ignored in practice. We therefore need clear criteria as to when we can regard an expectation of behavior as a "norm" and thus as an independent variable for an explanation of foreign policy behavior along the lines of constructivist foreign policy theory.

3.2. Commonality and Specificity

A much-stated criticism of constructivist foreign policy theory is the fact that an actor is frequently confronted with many value-based expectations of behavior, with the result that a distinction between relevant and irrelevant expectations of behavior is made difficult or becomes arbitrary. Constructivists are therefore always at risk of "explaining" foreign policy ex post by choosing that expectation of behavior as an explanation which comes closest to the observed behavior to be explained (Legro 1997: 33). However, criteria for determining the strength of norms can be found in the constructivist literature, and these considerably increase the possibility of ex ante explanations. In the constructivist view, the strength of a norm (and thus the strength of its influence on (foreign) policy behavior) depends on two properties: on its commonality, i.e. on how many actors of a social system share a value-based expectation of behavior, and on its specificity, i.e. on how precisely a norm distinguishes appropriate from inappropriate behavior.

3.2.1. Commonality

The strength of obligation attached to a norm depends on the extent to which it is shared by the units within a social system. [ 7 ] We can speak of a high degree of commonality if all the actors in a social system, for example the member states of an international organization, share a certain value-based expectation of behavior. If a certain expectation is shared "only" by a majority of actors, then it possesses a medium degree of commonality. Low commonality prevails when only a minority of actors shares a certain expectation of behavior. In the last case, it is impossible to formulate a constructivist prediction for a state’s foreign policy because constructivists hold that a norm can only be ascribed influence on a state’s behavior if it can claim at least a medium degree of commonality (see Legro 1997: 35).

An increase in the commonality of norms goes along not only with their assumed impact on behavior but also with the robustness of a constructivist explanation. The lower the commonality of a value-based expectation of behavior, the greater the risk that this expectation is not an independent variable but that the effect of a previously ignored independent variable is manifesting itself in both the expectation of behavior and in the non-compliant behavior that can be observed. Unlike generally shared expectations of behavior, the expectations of certain social sub-groups frequently compete with the expectations of other sub-groups. The plurality of value-based expectations of behavior in a given social system therefore gives rise to the question as to why the value-based expectations of one group should be relevant and not those of others. However, this would mean that the expectations of behavior held by such sub-groups were themselves dependent variables. For only if it can be demonstrated that certain groups assert themselves over their rivals because of the norms they share will these norms truly be the explanatory variable (see also Goldstein/Keohane 1993: 11).

3.2.2. Specificity

The expectations of behavior resulting from norms are not always equally precise. To a considerable extent, their precise meaning depends on their explication, i.e. their formal expression, for instance, in written conventions (see Raymond 1997: 225; Franck 1990: 64 et passim). From a constructivist perspective, the impact of a norm depends not only on its commonality, but also on its specificity (Legro 1997: 34). A norm is highly specific if it clearly distinguishes between appropriate and inappropriate behavior. An unspecific expectation of behavior allows for a wide range of behavioral options which can be justified as appropriate, and will thus scarcely enable the actors within a social system to determine when a norm has been violated. Consequently, unspecific norms are unsuitable as a standard for appropriate behavior and therefore as an independent variable with which to explain foreign policy behavior.

3.3. Norms and Foreign Policy Behavior: Causal Mechanisms

In constructivist foreign policy theory, the logic of appropriateness forms the link between the independent variable "norms" and states’ foreign policy behavior. In the constructivist view, a norm’s impact will be greater the more actors within a social system share it and the more precisely it distinguishes appropriate from inappropriate behavior. Now, we still have to answer the question as to how norms can shape a state’s behavior – i.e., how norms are communicated to actors and are accepted and internalized by them as directions for action.

Constructivist theory of foreign policy answers the question as to the mode of action of norms fundamentally differently from neorealist or utilitarian-liberal foreign policy theory which link the impact of norms with the variables "power" or "interests". In neorealist explanations, norms only develop an impact on actors’ behavior to the extent that compliance with them can be enforced by powerful actors, or that they are complied with by weaker actors in anticipatory fear of sanctions (see Krasner 1993). From this viewpoint, it is not the norms themselves but the power behind them that causes the norm-compliant behavior that can be observed. Other authors who follow the utilitarian-liberal analytical approach (see Bienen/Freund/Rittberger 1999) have used models of interest mediation to conceptualize the impact of norms (e.g. Cortell/Davis 1996). According to these models, norms work by matching (exogenously determined) actors’ interests and therefore by serving actors as a "resource" with which to assert their interests. From this utilitarian-liberal perspective, accordingly, the independent variables are interests, not norms. Legitimations of behavior by recourse to norms are thus only "post-hoc rationalizations of self-interest" (Raymond 1997: 213). Some rationalist authors also point at the usefulness of norms for establishing and stabilizing cooperation (Gehring 1994; Axelrod 1986). This especially applies to situations in which behavior is not sufficiently determined by interests because several behavioral options promise the same benefit. Here, norms can function as "focal points" (see Garrett/Weingast 1993: 176-187, 203-206; Goldstein/Keohane 1993: 12) and are therefore not the independent variable but only an intervening variable.

According to constructivists, norms do not follow logically from actors’ interests, as is the case in rationalist models, but precede them. The effect of norms on behavior cannot be reduced to that of "constraints" or "incentives" in the sense that norms increase or reduce the cost of certain modes of behavior, that is to say, that norms have a merely regulative effect on actors’ behavior. In the constructivist view, norms also have a constitutive effect, i.e. "norms legitimize goals and thus define actors' interests" (Klotz 1995: 26). [ 8 ] By identifying certain goals as legitimate, norms act as "motives" (Klotz 1995: 26). As "motives", norms determine the goals towards which states should legitimately strive. The "motives" function ascribed to norms by constructivists manifests itself in that states define their interests in accordance with the goals that have been designated as legitimate.

In constructivist foreign policy theory, the effect of norms is attributed to socialization processes. In its original, sociological meaning, socialization is a "process in which a person grows into the society and culture surrounding him and, by learning social norms and roles, becomes an independent, competent social being" (Weiß 1986: 269). In the course of this process the actor internalizes the expectations of behavior imparted to him by his social environment. He

"acknowledges the institutionalized modes of thought and behavior as correct, makes them - literally – ‘his own’ and brings his interests and preferences into line with them " (Schimmelfennig 1994: 338; see also Müller 1993b).

However, the socialization process should not be conceived of as a one-way process to which the person being socialized contributes no preconceptions of his own. Rather, the person being socialized may well reflect on what he internalizes during the socialization process and even modify its content (see Schimmelfennig 1994: 339f.). As individuals can constantly be confronted with new decision-making situations in the course of their lives and hence need to learn new expectations of behavior or reinterpret those that they have already internalized, socialization is never complete but is a continuous process (Parsons 1951: 208). [ 9 ]

Compared with the process of an individual’s socialization into his social environment, the peculiar characteristic of the socialization process of foreign policy decision-makers [ 10 ] is that two analytically distinct socialization processes run simultaneously. Because foreign policy decision-makers are at the interface of two social systems, i.e. the international system on the one hand and the intra-national system on the other, they face two different groups of socializing agents and, consequently, go through two different socialization processes. Transnational socialization describes a process whereby government decision-makers internalize international norms, i.e. value-based expectations of appropriate behavior that are shared by states. Societal socialization refers to a process whereby government decision-makers internalize societal norms, i.e. value-based expectations of appropriate behavior that are shared by the citizens of their state.

(1) Transnational socialization: Authors belonging to the constructivist school take processes of transnational socialization within international society (see 4.1 below) as the basis for their argument that international norms determine behavior. States are the constitutive units of that society and therefore the most important socializing agents. In other words, the norms shared by the international society of states are regarded by their constitutive (i.e. state) members as standards of appropriate behavior. International organizations, by contrast, are not constitutive units of international society. However, they are significant as socializing agencies in that they represent associations of states. [ 11 ] As members, the states associated in international organizations undertake to work towards achieving the goals set by the respective organization and to comply with the norms established in founding charters and legal acts. In the transnational-constructivist view, international organizations are important as socializing agencies because they express value communities made up of states. In the constructivist view, states acknowledge the expectations of appropriate behavior formulated by international organizations as standards of appropriate behavior if they regard themselves as part of the value community of the member states and seek recognition as an equal member by the other member states. Because international organizations are regarded as proxies for value communities, they thus function as "norm teachers" (Keck/Sikkink 1998: 34; Finnemore 1996a, 1996b).

In addition to states and international organizations, transnational advocacy coalitions also play an important role in transnational socialization processes. These cross-border coalitions of societal actors, although not themselves constitutive members of international society, nevertheless contribute to the establishment of new norms and the diffusion and communication of existing international norms. In diffusing and imparting norms, transnational advocacy coalitions aim at the widest possible dissemination and acceptance of international norms. Besides this function in processes of transnational socialization, transnational advocacy coalitions are important in constructivism because they, on the one hand, act as "norm entrepreneurs" in that they develop further existing norms and help establish new ones (Finnemore/Sikkink 1998) and, on the other, demand and verify compliance with existing norms (Keck/Sikkink 1998). [ 12 ]

As constitutive entities of international society, states not only socialize but also are socialized because they are the primary addressees of internationally shared, value-based expectations of behavior (Armstrong 1994: 16ff.). States acknowledge the norms of international society as standards of appropriate behavior because their identity as states depends on their membership in international society (Armstrong 1994: 21, 24; Schimmelfennig 1994: 344). States only become sovereign when they are recognized as such by other states (Biersteker/Weber 1996: 3, 11-14; Thomson 1995; Jackson 1990). Moreover, continued recognition by the other constituent entities of international society also depends on their declaration of belief in collective goals, such as securing world peace (Claude 1966). International norms which define collective goals such as these and specify appropriate modes of behavior for pursuing them have a socializing effect on states because states are constantly concerned with their reputation as recognized (i.e. norm-compliant) members of international society.

The concept of reputation used by constructivists differs from the way reputation is understood in rationalist theories. In the latter, states are usually and primarily concerned with their reputation as reliable partners in negotiations or as allies (see Mercer 1996; McElroy 1992: 46-53). Constructivists hold that states seek to preserve and consolidate their reputation as legitimate members of international society (Franck 1990: 191). The difference in these concepts of reputation thus affects the way in which states can be sanctioned. Constructivist theories emphasize immaterial or even symbolic sanctions aimed at states’ status as legitimate members of international society (for example, the cancelling of diplomatic, cultural or sports contacts). In rationalist theories, emphasis is placed on material sanctions (such as trade embargoes) which have a negative effect on the pursuit of rational interests. [ 13 ]

The value-based expectations of appropriate behavior shared by states can be subject to change over time (Armstrong 1994: 12ff.). This is a result of the communicative processes in which value-based expectations of behavior are socialized and as a result of which the contents of socialization can change in the long term in the light of the situation-specific interpretation of these general expectations of behavior. [ 14 ]

(2) Societal socialization: While rationalist approaches emphasize the importance of public pressure for the effect of societal norms on decision-makers’ behavior (Checkel 1997: 476f.; see also Raymond 1997: 216; Cortell/Davis 1996; McElroy 1992: 43-46), constructivist foreign policy theory regards processes of societal socialization as decisive for the effect of societal norms. Both society as a whole and its sub-groups – in particular societal "advocacy coalitions" – are regarded as socializing agencies (see xa4.1. and 4.2 below) addressing expectations of appropriate behavior to the political decision-makers. From a constructivist view, there are three reasons why the behavior of foreign policy decision-makers is influenced by societal expectations of appropriate behavior. First, foreign policy decision-makers have already internalized societal expectations of appropriate behavior via the process of political socialization to which all the citizens of a state are subject. Second, before becoming representatives of their state in international society, politicians generally run through national political careers in the course of which they internalize more specific societal expectations of appropriate behavior. Third and finally, decision-makers behave consistent with societal expectations of appropriate behavior because this is in line with the way they see themselves as recognized representatives of their society in dealings with their international environment. If a government does not comply with the societal expectations of behavior addressed to it, it runs the risk of losing its recognition by society as its legitimate representative.

Foreign policy decision-makers are simultaneously subject to transnational and societal socialization processes. In international society, they are on the receiving end of expectations of appropriate behavior addressed to them by international society. Conversely, the nationally constituted society expects its representatives to satisfy societally shared expectations of appropriate behavior on the international level. If there are contradictory expectations of behavior of at least a medium degree of specificity and commonality on the international and societal levels, then a constructivist prediction is just as impossible as when these expectations of behavior are completely absent on both levels or do not reveal sufficient commonality and/or specificity for them to be regarded as significant from a constructivist point of view. This is because constructivist theory of foreign policy (yet) offers no criteria for determining whether foreign policy decision-makers are guided more by the expectations of behavior addressed to them by their international or their societal environment. If there are conflicting societal and international norms, a constructivist explanation is indeterminate because in such situations, foreign policy decision-makers are free to choose the norm which best justifies their behavior. Theoretically, therefore, it cannot be ruled out that actions are in fact guided by an interest with no normative base and are justified only ex post by recourse to a norm which matches the behavioral option chosen.

By the same token, if international and societal expectations of appropriate behavior match they reinforce each other. In such situations, foreign policy decision-makers comply with the expectations of behavior addressed to them because of their self-understanding as both representatives of members of international society and as representatives of their own society, which has delegated functional authority to them, vis-à-vis international society. Therefore, if there are the same expectations of appropriate behavior on both the international and the societal level, constructivist theory claims that its explanation of foreign policy is particularly powerful. However, an expectation of appropriate behavior does not need to be present on both levels in order to allow for the assumption that it guides foreign policy decision-makers. If there is an expectation of appropriate behavior with sufficient commonality and specificity on only one of the two levels, the degree of its internalization by foreign policy decision-makers (and thus its effect on their behavior) will be regarded as lower than if international and societal norms are congruent. Nevertheless, in these cases there is no reason to reject the usefulness of constructivist theory for explaining and predicting foreign policy behavior.

These considerations enable us to rank the predictive power of constructivist foreign policy theory based on international and/or domestic norms (Fig. 1). Constructivist theory claims high predictive power when there are congruent expectations of appropriate behavior of at least a medium degree of commonality and specificity on both levels. Medium predictive capability exists if there is an expectation of appropriate behavior on one level only. When international and societal norms contradict each other, it is just as impossible for constructivist foreign policy theory to make predictions as when norms are lacking on either level.

4. Transnational and Societal Constructivism

Our discussion of the various mechanisms of action has made clear that the independent variable "norms" can neither be allocated exclusively to the level of the international system nor to that of states and their societies, and that it can neither be thought of solely as a systemic characteristic nor as a unit characteristic. Rather, what is generally characteristic of the way norms work is that they can be taken up by actors within and outside states and be expressed as expectations of appropriate behavior addressed to a state’s foreign policy decision-makers. It is especially prominent norms such as the protection of human rights or free trade that are institutionally embedded both within states and on the level of international society.

Constructivists have devoted considerable attention to the question of the interaction between unit and systemic characteristics and have stressed the interdependence of the two. Indeed, certain authors regard this structurationist view of "agents and structures as mutually constituted or codetermined entities" (Wendt 1987: 350) as the defining characteristic of the constructivist research agenda (see Checkel 1998: 326). It would, however, be premature to conclude from the structurationist view of the agent-structure problematique that it is impossible within a research design to distinguish analytically systemic variables from unit variables and to examine their impact separately. [ 16 ]

Whether research focuses on the influence of international or societal norms depends above all on whether the aim of the research is to investigate the similarity of foreign policies given different interests (see, for example, Finnemore 1993) or differences in state behavior given identical international expectations of behavior (see, for example, Ulbert 1997). For an analysis of German foreign policy addressing the question of continuity or change, any decision to focus on either international or societal norms must appear arbitrary, especially as neither of these levels enjoys any theory-based primacy over the other. Moreover, focussing on either international or societal norms runs the risk of being blind to the reinforcing or counteracting influence of the other level. After all, as interwoven but analytically discrete social systems, German society and associations of states such as the European Union do not necessarily formulate the same expectations of appropriate behavior concerning German foreign policy. Only when the two levels are separated analytically, it is possible to discover differences between the expectations of German society and those of other states, but also interactions between them.

4.1. Transnational Constructivism

To a significant extent, transnational constructivism has its origins in a research tradition known as "reflexive institutionalism" (Schaber/Ulbert 1994; Keohane 1989b) or "sociological institutionalism" (Finnemore 1996a, 1996b). While the label "transnational" is justified by its assumptions about the processes of, and the actors involved in, the creation and diffusion of international norms, [ 17 ] the affinity of this research tradition with constructivism results from the fact that it is also based on the fundamental assumption of the social construction of reality. Unlike societal constructivism, however, transnational constructivism locates the influential expectations of appropriate behavior based on social constructs of reality on the international, not the societal, level. The basis for assuming that international norms guide behavior is the transnational-constructivist concept of the international system as an "international society", described in section 4.1.1. below. With reference to the general definition of norms set out above (3.1.), section 4.1.2. provides a definition of international norms. Finally, the role of international institutions in constituting, imparting and perpetuating international norms is discussed (4.1.3).

4.1.1. The Structural Model of International Society

Like neorealism, transnational constructivism also assumes that the actions of state actors on the international level are determined to a significant extent by the characteristics of their international environment. However, while the representatives of the neorealist school postulate that the material structures of the international system (distribution of power among states) determine actors’ behavior, the proponents of transnational constructivism stress the significance of immaterial structures (shared constructs of reality, institutions, norms) for actors’ behavior (Finnemore 1996: 15). Neorealism's anarchical international system is thus replaced in transnational constructivism by international society which, as a social system, both constitutes the identities and interests of its members and is reproduced by their practices (Ruggie 1998; Wendt 1994, 1991; Wendt/Duvall 1989). [ 18 ]

The point of departure for transnational-constructivist arguments for the behavior-guiding effect of international norms is the discovery that the practices of state actors on an international scale are characterized by a considerable level of similarity (isomorphism) (Finnemore 1996a: 6, 22; McNeely 1995: 2f., 20; Meyer 1987: 46-50). Now, neorealists also assume a certain degree of isomorphism in state behavior in international politics and attribute this to the necessity, experienced by all states in the anarchical international system, of securing their survival by self-help (see Waltz 1979: 93-97). However, constructivism attributes the isomorphous behavior of states to the influence of international norms. For if securing survival were the only motive for isomorphous behavior, states with differing characteristics and differing interests would only behave similarly in those areas that are of immediate relevance for securing their survival, but not in other areas which play no role, or only a very negligible one, for guaranteeing their survival. [ 19 ] As Finnemore (1996a: 30) writes:

"conventional theories treat preferences as inherent qualities of actors. Their proponents would expect different actors with different preferences to act differently. Similar action by dissimilar actors in the absence of constraint is anomalous under these theories. Such behavior is to be expected, however, within a social structural framework. International norms of behavior and shared values may make similar behavioral claims on dissimilar actors."

Of course, the finding that the behavior of state actors in international society is largely isomorphous does not imply that the culturally determined constructs of reality and values held by all individuals and social groups that it includes are homogeneous. For despite increasing transnational communications and political activities taking place outside of state control, international society is still not a "global (civil) society" constituted by human individuals. [ 20 ] The fact that international society is constituted by states perpetuates, at least partly, its cultural heterogeneity because states gain their legitimacy by appealing not only to the universal norms of international society but also to socially shared, culturally and traditionally rooted norms which justify their national independence. The processes of standardizing the behavior of state actors do not therefore occur uniformly and universally throughout all areas, but frequently within regionally and/or functionally contained social subsystems whose actors subscribe to a commonly shared social construction of reality and to shared values to a special degree (see Hurrell 1995). Apart from international society as a whole, these social subsystems also address expectations of behavior to those states which belong to them (see Raymond 1997: 226). Therefore, even those norms which are only institutionalized within these social subsystems must be considered in a transnational-constructivist analysis of foreign policy behavior.

4.1.2. International Norms

In accordance with the definition of norms as the intersubjectively shared, value-based expectations of appropriate behavior within a social system, international norms are defined as those expectations of appropriate behavior which are shared within international society or within a particular subsystem of international society by states, its constituent entities.

Like neorealism, transnational constructivism also assumes the absence of a superior coercive power on the international level but regards this as a further argument for the effect of international norms because given its absence, fear of punishment by such a coercive power cannot explain the wide extent of the isomorphous foreign policy behavior of states (cf. Wendt 1992). Furthermore, unlike utilitarian-liberal explanations of the effect of international norms, proponents of transnational constructivism point out that the complex interdependencies not only between actors, but also between different policy areas make it impossible for an actor to calculate the benefits to be gained from compliance with a certain norm (see Hasenclever/Mayer/ Rittberger 1997: 158-161; Hurrell 1993: 59). In the constructivist view, therefore, norm compliance is not linked to states’ "demand" for norm-regulated international cooperation arising from self-regarding interests. [ 21 ] Rather, it is the "supply" of expectations of appropriate behavior which significantly influences actors’ behavior within social systems. International society and its sub-systems are the "suppliers" of international norms. Usually, it is associations of like-minded states in international organizations which are regarded as value communities by constructivists. By recourse to the norms institutionalized in, and propagated by, them, these communities define their member states’ behavioral roles. Transnational advocacy coalitions play a significant role in communicating this supply of international norms. [ 22 ]

In principle, this definition of international norms as a value-based expectation of appropriate behavior, shared by international society or a sub-system and making behavioral claims on the norm addressees, already contains its operationalization for foreign policy analysis. The immediate consequence of the existence of such a norm is that all those to whom the norm is addressed are expected to behave norm-consistently. For the derivation of predictions for German foreign policy, this means that the systemic independent variable of transnational constructivism – international norms – can be transformed directly into an actor-related independent variable "expectations of behavior addressed to Germany". The strength of this expectation of behavior, however, is dependent on its commonality and specificity within the entire group of countries to whom the norm is addressed. When deriving a constructivist expectation of behavior for German foreign policy from an international norm, therefore, its commonality and specificity must be established. If a norm is only valid within a sub-system of international society, then Germany must be a member of this sub-system in order to be an addressee of its concomitant expectation of behavior. However, certain expectations of behavior can also depend on characteristics of states other than membership in a social system. For example, the expectation to grant development aid to developing countries is obviously not directed to all members of international society but only to the industrially developed nations. This means that in deducing transnational-constructivist predictions about German foreign policy, certain characteristics of Germany as a country may also have to be considered. This consideration may be necessary for determining whether Germany is one of the addressees of the international norm from which the prediction is derived.

4.1.3. The Role and Significance of International Institutions

Occasionally, there is confusion in the literature with regard to the conceptual difference between "norms" and "institutions". For example, sovereignty has been described as both a norm and an institution (see Finnemore/Sikkink 1998: 891; Finnemore 1996a: 16). However, if we consider the definitions of norms and institutions that have become broadly accepted, the terms can be clearly distinguished from one another. While a norm is always a single value-based expectation of an actor, we can follow Zürn (1992: 141) in defining a social institution as

"a permanent and consolidated pattern of behavior of a specific number of actors in specific, recurring situations. The patterns of behavior are based on a set of rules which define behavioral roles, give a meaning to activities, and influence actors’ expectations, thus themselves directing relationships between actors in the recurring situations."

Like all social systems, international society has also established social institutions. In the constructivist view, the existence of international institutions is one of the central elements characterizing international society and distinguishing it from the neorealist concept of an anarchical international system determined solely by the international distribution of power (see Buzan 1993: 330-336; Wendt/Duvall 1989). International institutions are to be understood as those social institutions which exist on the level of international society as a whole or of one of its sub-systems, constitute the roles and shape the behavioral patterns of their constitutive members. [ 23 ] As Zürn’s definition shows, they also represent sets of interrelated norms which, as a whole, constitute behavioral roles and give meaning to the concrete expectations of behavior attributed to these roles. Consequently, sovereignty is clearly an institution consisting of a bundle of norms and giving meaning to them (see Finnemore/Sikkink 1998: 891). [ 24 ]

By integrating interrelated norms into such norm sets establishing behavioral roles, institutions give a significance to individual norms which goes beyond that of constituting a certain goal of action and regulating behavior. States meet the expectations of behavior set by individual norms because they regard themselves as members of international institutions, and this requires that they perform the role ascribed to them. By integrating norms into norm sets constituting behavioral roles, institutions also grant permanence to individual norms because the redefinition of an individual norm’s expectation of behavior would require a redefinition of the entire behavioral role which the norm in question has helped to constitute (see Goldstein/Keohane 1993; Krasner 1983a,b).

Proponents of the transnational-constructivist research tradition interpret the political order of international society ("world polity", see McNeely 1995; Ruggie 1993; Meyer 1987) as a network of international institutions. The foundation of this order rests on the institution of sovereignty itself because it is through sovereignty that states as constitutive units of international society come into being, making cooperation (regulated and perpetuated with the help of specific institutions) between these units possible and necessary. At the same time, the norms subsumed under the institution of sovereignty specify the basic parameters of legitimate state action. Above the level of this "constitutional structure" of international society (Reus-Smit 1997), transnational constructivism locates specific international institutions such as international organizations and regimes. It applies equally to both types of institutions that actors’ ability to come together in them with all rights and duties presupposes their constitution as states. International organizations are formal associations of states with tasks that are partly issue area-specific and partly transcend issue areas. Unlike regimes, they also act as purposive collective actors made up of states (Rittberger/Zangl 1995: 26f.; Keohane 1989a: 3f.). Not least, their function is to formulate collective goals and to specify appropriate means for achieving them. In the emergence and communication of international norms, therefore, they play a central role which they partly also seek to fulfil by functionally integrating transnational advocacy coalitions. [ 25 ] International regimes are sets of principles, norms, rules and decision-making procedures which define the ends and means of action within specific policy areas and thus establish reciprocal reliability of expectation (Hasenclever/ Mayer/Rittberger 1997; Krasner 1983a). [ 26 ] From a transnational-constructivist view, the effect which the norms embedded in specific international institutions have on states’ foreign policy is a consequence of the deeper normative structure of the "world polity" or of its sub-systems because transnational constructivism assumes that states, like the other actors of international society, share the generally shared expectation that the actions of individual members of international society should match the shared expectations of appropriate behavior addressed to them (see Hurrell 1993: 59).

When using internationally shared expectations of appropriate behavior as a basis for deriving constructivist predictions for German foreign policy, reference will mainly be made to explicit and specific norms of both an international legal and non-legal nature which are embedded in issue area-specific or subject-specific international institutions (international regimes, international organizations). For they define the positive aims of state action and specify the means for their achievement whereas norms which can be ascribed to "constitutional structure" only define the properties a state must demonstrate in order to be able to be recognized and act as such, as well as what fundamental modes of behavior it should observe. However, the groups of states associated in certain organizations frequently only represent sub-systems of international society; the value-based expectations of appropriate behavior embedded in these organizations thus only pertain to their member states but not to international society as a whole (see Raymond 1997: 226; Hurrell 1995). When formulating constructivist expectations of behavior for German foreign policy, Germany’s membership or non-membership of these sub-systems must therefore be considered.

4.2. Societal Constructivism

The research tradition of societal constructivism stresses the dependence of foreign policy behavior on the norms existing in society. The main difference between the various analytical approaches hinges around the issue of whose expectations of behavior are considered as having decisive influence on foreign policy (Yee 1996: 69f.). While one group of authors concentrates on the influence of the value-based expectations of behavior shared by experts in a certain issue area (see Sabatier 1993, Adler/Haas 1992, Haas 1992, Hall 1989), other authors do not single out any societal sub-group as a bearer of norms but ascribe them to the actor ‘society’ in its entirety (see, for example, Engelmann/Knopf/Roscher/Risse 1997; Ulbert 1997; Barnett 1996). As experts in a certain issue area are a societal sub-group, the expectations of appropriate behavior that they share do not possess a high degree of commonality. Instead, the value-based expectations of behavior of these sub-groups are regarded as influential on a state’s foreign policy because they frequently manifest a high degree of specificity.

4.2.1. Norms Shared by Experts

Attempts to explain foreign policy behavior by reference to the value-based expectations of behavior shared by experts are most frequently made for those areas of foreign policy where decision-makers are insufficiently informed about the complexities of the issue area. To be able to make foreign policy decisions, decision-makers are thus dependent on expert advice.

In the literature, groups of experts are generally conceived of either as an ‘epistemic communities' (Haas 1992) or as ‘advocacy coalitions' (Sabatier 1993). Both approaches are based on the view that a group of experts not only shares knowledge about cause-effect relationships in a policy area (i.e. causal beliefs) but also have a "shared set of normative and principled beliefs" (Haas 1992: 3). The ‘advocacy coalition’ approach places more emphasis on the significance of "the normative and ontological axioms which determine an actor’s general political philosophy beyond the level of various policy areas" (Sabatier 1993: 133) than ‘epistemic community’ research does. While ‘epistemic community’ research emphasizes how value-based expectations of behavior are guided by the scientific ‘state of the art’, the ‘advocacy coalitions’ concept stresses the significance of value judgements inaccessible to scientific debate. [ 27 ] The concept of expert groups as ‘advocacy coalitions’ thus appears to be better suited to making use of this research tradition for our project to establish a constructivist theory of foreign policy. In the following, therefore, we will always use the term ‘advocacy coalitions’ when referring to a group of experts sharing principled beliefs about appropriate courses of action in a certain policy area.

Advocacy coalitions frequently compete with each other. An explanation based on expectations of appropriate behavior held by advocacy coalitions is thus faced with the question as to which expectation of behavior held by which advocacy coalition should be regarded as most important for a societal-constructivist explanation. In such cases, it is not possible to provide a robust societal-constructivist explanation on the basis of one advocacy coalition’s expectation of behavior because there is no criterion furnished by constructivist theory which would allow the analyst to decide which of several expectations prevails. Attempts to provide such an explanation would always be susceptible to allegations of ignoring other possible independent variables (see 4.2.3.). However, if there exists an advocacy coalition’s expectation of appropriate behavior concerning an issue area of foreign policy and if there is no competing view within the policy field in question, this expectation can be conceived of as a societal norm with medium commonality.

4.2.2. Norms Shared by Society as a Whole

Expectations of behavior, which can be said to be shared not only by individual societal groupings but by ‘society’ as a whole, can be ascribed high commonality. Common terms used to signify these norms shared by society as a whole are ‘(national) identity’ (see Ingebritsen/Larson 1997; Marcussen/Risse 1997; Jepperson/Wendt/ Katzenstein 1996: 33) and ‘(political) culture’ (see Hudson 1997; Ulbert 1997; Berger 1996; Risse-Kappen 1994).

Foreign policy analyses which explain behavior by recourse to expectations of behavior shared by society as a whole frequently point at these norms’ origins in collective historical experience. Thomas Berger, for example, has suggested that the historical experience of militarism and the Second World War has induced the conviction on the part of most societal actors in Germany and Japan that military power should only be used with constraint, if at all. The concomitant norms "are now integral parts of their countries’ post-1945 national identities" (Berger 1996: 318; see also Berger 1998). Because the norms shared by society as a whole are per definitionem not called into question within society to any significant extent, [ 28 ] societal constructivism assumes that they exert an especially strong influence on foreign policy.

4.2.3. A Robust Societal-Constructivist Explanation

As shown in section 4.2.1, the robustness of constructivist explanations increases in line with the level of commonality of the independent variable ‘norms’. If the norms shared by society as a whole compete with the norms of advocacy coalitions, societal constructivism assumes that the norms shared by society as a whole prevail and determine foreign policy. Advocacy coalitions’ expectations of appropriate courses of action are only regarded as relevant if there are no norms shared among society as a whole in a policy area. This is particularly the case in those foreign policy areas that are completely or largely removed from public scrutiny. It is, for example, plausible that there are no explicit expectations of appropriate behavior shared by society as a whole with regard to German foreign policy for keeping the North Sea or Baltic clean. Instead, reference can be made to a group of environmental experts as holders of an issue-specific expectation of appropriate behavior. In such cases, an explanation of foreign policy based on expectations of appropriate behavior shared by society as a whole would not be possible at all whereas an explanation based on the expectations of an advocacy coalition would be very robust. If there is no competing advocacy coalition within that society, there is no need either to explain why one coalition has asserted its expectations of behavior.

In order to establish whether a certain norm exists on the societal level, it must therefore first be ascertained whether there are expectations of German foreign policy behavior in this policy area shared by society as a whole. If this is not the case, the next step is to investigate whether there are advocacy coalitions in this area which formulate expectations of appropriate German foreign policy behavior. If several such coalitions exist and hold competing expectations of behavior, then it will not be possible to identify societal norms.

4.2.4. The Transfer of Domestically Valid Norms to Areas of Foreign Policy: The Domestic Analogy

Constructivists generally expect foreign policy to be influenced by value-based expectations of appropriate foreign policy behavior. On the societal level, however, the value-based expectations of appropriate domestic policy shared by society can also influence a state’s foreign policy. Proponents of societal constructivism assume that foreign policy decision-makers "want to see their international environment ordered according to the same values and principles governing their own political and social system" (Kittel/Rittberger/Schimmelfennig 1995: 68). One precondition for the influence of domestic policy-related, value-based expectations of appropriate behavior on foreign policy is that there are no foreign policy norms with sufficient specificity and commonality.

Anne-Marie Burley (1993b), for example, has suggested that US foreign policy in the years immediately following the Second Wold War was aimed at transferring the policy of the New Deal to the international economic order. According to Burley, the norm which characterized the New Deal was that of safeguarding citizens’ economic and social welfare by government intervention to correct market failure and was shared by large sections of US society. Thus, Burley argues that U.S. politicians sought to comply with the expectation, held by American citizens, that they would act in accordance with the norms of the New Deal when structuring the international economic order. These expectations thus influenced U.S. foreign policy.

A further example of the transfer of domestic norms to a foreign policy context can be found in the literature devoted to explanations of the ‘democratic peace’, i.e. the finding that democracies do not go to war with each other. The societal-constructivist explanation for this finding is that democratic governments strive to comply with the domestically valid norm of non-violent conflict settlement and thus also prefer non-violent behavior in to international conflicts with other democracies (see Weart 1994; Russett 1993). [ 29 ]

As the examples of the domestic settlement of conflicts in democracies and the New Deal show, norms that have emerged for domestic policy and initially only claim validity for this sphere possess a high degree of commonality. [ 30 ] In this regard, they are suitable for societal-constructivist explanations. However, the precondition is that norms which refer to domestic contexts allow for a distinction between appropriate and inappropriate foreign policy behavior. Whether or not this is the case depends on the specificity of these norms. Their influence on a state’s foreign policy will thus be the greater the more clearly societal expectations of behavior distinguish between appropriate and inappropriate behavior. If such expectations have only low specificity and do not exclude any form of behavior as inappropriate, an impact on a state’s foreign policy cannot be assumed.

5. Identification of Independent Variables

In this section we look at various indicators of international norms on the one hand (5.1.) and of societal norms on the other (5.2.). The various indicators help us to identify systematically the value-based expectations of appropriate German foreign policy on the international and societal levels. The identification itself, however, must be left to the application of constructivist foreign policy theory to issue area-specific case studies. In addition, account must be taken of the criteria of commonality and specificity, discussed in section 3.2, as their level determines whether an expectation of appropriate behavior is strong enough to accord to it a significant impact on foreign policy behavior.

5.1. International Norms

In section 4.2. above, international norms were defined as value-based expectations of appropriate behavior shared within international society. Many of these norms are codified in international law. Like any law, international law can be understood as "the expression of social and political values [...] of a community" (Burley 1993a: 211). From the constructivist point of view, therefore, international law is an important indicator of international norms. With a view to the identification of international norms, then, the following section will first discuss the norms of international law (5.1.1.). This will be followed by a discussion of the legal acts of international organizations as an indicator of international norms (5.1.2) and a further section will deal with the final acts of international conferences which can also present a significant indicator of the existence of internationally shared expectations of appropriate behavior (5.1.3).

5.1.1. International Law

Although violations of international law are not normally punished by either powerful states and groups of states or by international bodies, in particular the UN Security Council, states’ political practice is characterized by compliance with the law rather than by its violations (see, for example, Akehurst 1992: 2; Henkin 1968: 46). Constructivist authors conclude from this that a logic of appropriateness is at work in international society whose yardstick is provided to a considerable degree by the norms of international law (see, for example, Franck 1990; Kratochwil 1989). [ 31 ] As international society has no constitution equal to that of individual states, its "legislation" in the form of international law is largely decentralized (see Coplin 1969: 144). Article 38 of the Statute of the International Court of Justice (ICJ) names the four following sources of international law, which represent a hierarchy of norms:

1. international treaties;
2. customary international law;
3. "the general principles of law recognized by civilized nations" and
4. (as "subsidiary means") judicial decisions and "teachings of the most highly qualified publicists of the various nations".

"those rules of behavior that have so far been observed by the subjects of international law in their mutual transactions in general [...] or in particular, if this practice (‘state practice’) is joined by the conviction that there is a legal duty to comply with the objective rule (opinio juris vel necessitatis)." (italics in the original).

Since customary international law requires both that it is recognized by states as law and that there is a match between value-based expectations of behavior and actual behavior, its norms can generally be assumed to possess a high degree of commonality. Due to the exemplification of appropriate behavior by states’ repeated or even constant behavioral practice, the norms of customary international law generally also present a high degree of specificity. There are, of course, norms whose status as customary international law is disputed. The literature on international law can be an aid to interpretation, but when it is used to establish norms of customary international law there is a considerable danger of contradictions, that is, one author may regard a certain norm as customary international law while another may not. One solution to this problem may be found in legal acts of international organizations (e.g. UN General Assembly resolutions) which – at least when they have been adopted unanimously or by a great majority and repeatedly refer to a certain norm – are regarded by scholars of international law as evidence of customary international law. If such resolutions meet with broad approval and repeated reference is made to them, it can be assumed that they reflect states’ conviction of law (see Seidl-Hohenveldern 1997: 104; Akehurst 1992: 215f.). For constructivists, however, it is less significant whether or not an expectation of behavior is recognized as "law" because even norms whose legal character is disputed but whose expectations of behavior are recognized as a yardstick of appropriate behavior in international society display a high degree of commonality.

5.1.2. Legal Acts of International Organizations

International organizations are usually founded by states by means of a multilateral, international legal charter. As indicators of international norms, such charters thus belong to the category of international treaties. Further expectations of member states’ behavior can be made explicit by international organizations in legal acts. It does not matter whether the norm contained in such an act is legally binding or not because these norms in any case represent behavioral expectations that are regarded as standards of appropriate behavior within a sub-system of international society that can be considered a community of values (or, as in the case of the UN, a community of values encompassing almost all of international society) (see Keck/Sikkink 1998: 34). Generally, therefore, resolutions such as those of the UN General Assembly can indicate what expectations of appropriate behavior are shared within the social system constituted by the respective international organization. The criterion for judging the strength of these norms is therefore not their legal character but their commonality and specificity. The commonality of an expectation of appropriate behavior expressed in a resolution can be inferred from the way in which the resolution was adopted: if it is adopted unanimously, then the commonality of the norm it expresses can be classified as "high". Medium commonality exists if the majority of member states have approved the resolution in question. [ 33 ] Below this threshold, such document cannot be seen as presenting an expectation of appropriate behavior. The specificity of the expectations of behavior contained in these legal acts can also vary greatly and must be determined individually.

5.1.3. Final Acts of International Conferences

In recent years, international society has with increasing frequency resorted to large international conferences in order to formulate common goals and adopt action programs to realize them (e.g. the Vienna World Human Rights Conference, the Peking World Women’s Conference, the Copenhagen World Social Summit) (see in general Messner/Nuscheler 1996; Rittberger 1991). The final acts of these conferences are neither international treaties nor legal acts of an international organization but represent declarations of common political intent by the participating states. [ 34 ] With regard to the establishment of shared perceptions of problems, the definition of common goals of international society and the specification of suitable and appropriate means for their achievement, however, the significance of these acts cannot be neglected. In deriving constructivist predictions of foreign policy behavior, therefore, account must be taken of the norms expressed in them. The final acts of such conferences are generally accepted without a vote or by consensus. This would at first suggest that the commonality of the norms they contain should be regarded as high. However, it should not be forgotten that precisely because of the usual practice of consensual adoption, the expectations of behavior that such acts contain frequently represent compromise formulas with only low specificity, so it is often not possible to base a constructivist prediction on them alone. For deriving constructivist predictions, then, this indicator of international norms is more useful in combination with other indicators of international norms. This is all the more the case if world conferences confirm already existing norms. In these cases, the specificity of the expectations of behavior contained in such acts usually tends to be high because an international consensus is established on the specific expectations of behavior resulting from the confirmed norms, and at least in some cases this also leads to concrete action by states to satisfy these expectations of behavior. [ 35 ]

5.2. Identification of Societal Norms

To generate constructivist predictions for German foreign policy, an identification of the independent variable "norms" must also be made on the level of German society. Identifying societal norms is a task which societal constructivism has in common with political culture research. Societal constructivism can therefore benefit from the methodological considerations of this body of research.

In political culture research, there has been much debate about the advantages and disadvantages of qualitative-interpretative versus quantitative methods. Qualitative approaches dominated the field until the 1940s and can still be found in anthropologically inspired studies. [ 36 ] Qualitative methods include participant observation and the interpretation of documents ranging from text books to belletristic literature. The interpretation of the constitutional and legal order (5.2.2.), of the party political programs and election platforms of political parties (5.2.3.) and the analysis of general parliamentary debates about foreign policy (5.2.4.) draw on the tradition of qualitative research. Quantitative methods have dominated research into societal norms in Western democracies since technological innovations revolutionized data processing and made it possible to record and process huge amounts of data. Survey data are the archetype of quantitative political culture research which attempt to identify societal norms by means of standardized questionnaires of individual attitudes (5.2.1.). Quantitative and qualitative methods are not mutually exclusive but complement each other in the identification of societal norms (Bergem 1993: 55).

5.2.1. Survey Data

Approaches which assume that society influences foreign policy have also always been devoted to analyzing the role of public opinion in the foreign policy decision-making process (see, e.g. Holsti 1996, Risse-Kappen 1991, Russett 1990). One important reason for this is the proposition, which goes back to Bentham and Mill, that, if public opinion influenced foreign policy decision-making processes to a higher extent, this would make foreign policy more peaceful (Holsti 1996: 2f.). Most studies of the influence of public opinion on foreign policy are based on a cause-effect relationship which assumes that decision-makers act according to a consequential logic, i.e. their interest in re-election (see Russett 1990: 10). Societal constructivism, by contrast, assumes that public opinion influences foreign policy because it expresses societally shared, value-based expectations of behavior which foreign policy decision-makers, following the logic of appropriateness, take into account. Societal constructivism thus regards public opinion as an indicator of societal norms.

As an indicator of societal norms, the great advantage of survey data is their high degree of reliability. The reliability of surveys, i.e. of a standardized questionning of a representative section of the population and the statistical evaluation of the data collected, is acknowledged to be high if a large sample is used. [ 37 ] Critics, while scarcely doubting the high degree of reliability of findings, nevertheless cast doubt on their validity. Survey data are valid to the extent that they provide a survey "of public opinion as disposition to behavior and not [...] as non-committal, fleeting opinions" (Groß 1995: 18). Leaving aside this criticism of the validity of individual findings, there remain two fundamental criticisms of approaches which identify societal norms by taking recourse to survey data. On the one hand, fundamental methodological doubts have been expressed as to the suitability of surveys for identifying societal norms and, on the other, there is doubt as to the possibility of arriving at valid findings about societal attitudes towards foreign policy issues in particular.

This fundamental methodological criticism includes Lijphart’s warning of the danger of wrongly drawing individualistic conclusions if intersubjective phenomena are derived from subjective survey data: "a special danger in research based partly or wholly on survey data" (Lijphart 1980: 45). Proponents of a quantitative approach counter this by pointing out that political culture (and thus societal norms) "manifests itself on the individual level as values, convictions and attitudes" and "[can] for this reason be measured there" (Kaase 1980: 155; italics in the original). A similarly fundamental criticism is that many intersubjectively shared expectations of behavior "[consist of] ideas which are a matter of course, which the individual is often not even aware of, or is at best half aware of, and cannot therefore be made the subject of questionnaires" (Rohe 1994: 4; see also Finnemore 1996: 23). [ 38 ]

There are, however, more serious objections than these fundamental methodological criticisms of the suitability of poll research for identifying societal norms. These objections cast doubt on the possibility of survey data regarding attitudes to foreign policy issues being highly valid. The first reason given for this skepticism is the assumption that, compared with domestic policy issues, public opinion with regard to foreign policy issues is generally unstable and incoherent. [ 39 ] While studies have been able to relativize this proposition, even the proponents of societally centered explanations of foreign policy behavior concede that

"polls repeatedly reveal that the mass public [...] is poorly informed about the specifics of conflicts, treaties, negotiations with other nations, characteristics of weapons, foreign leaders, and the like" (Holsti 1992: 447).

The comparatively small public interest in problems of foreign policy affects the validity of survey findings. For the area of foreign security policy, Jürgen Gross concludes:

"If it can be assumed that most citizens are scarcely familiar with the topics of security policy, then it must be expected that the ‘instrumentation effects’ that are unavoidable in surveys and distort their representation of true opinion will show up especially strongly. If, in addition, the subject was of little importance for the respondents, then these effects would be reinforced still further" (Gross 1995: 17, see also Dobler 1989: 10f.).

The validity of the results of surveys thus depends on several conditions. First, findings are only valid if the respondents are familiar with the topic. In addition, the validity of findings depends on whether instrumentation errors have been minimized. Above all, this means that the suggested answers must be formulated very carefully. [ 40 ] Furthermore, surveys must have been conducted over a longer period of time. One single survey is not sufficient to identify societal norms rather than everyday political opinion. If these conditions are satisfied, survey data are a suitable indicator of norms on the societal level. In terms of research practice, however, these conditions can only rarely be expected. Societal norms will therefore generally have to be identified using other indicators.

5.2.2. The Constitutional and Legal Order of a Society

Intersubjectively shared, value-based expectations of political decision-makers’ behavior take on lasting influence when they are integrated into a society’s institutional order. In institutionalized form,

"the impact of ideas may be prolonged for decades or even generations. In this sense, ideas can have an impact even when no one genuinely believes in them as principled or causal statements" (Goldstein/Keohane 1993: 20).

In contrast to a rationalist perspective, societal constructivism assumes that the institutions of the constitutional and legal order are "not only neutral devices for the accommodation of different interests in the pursuit of common policies, but also provide symbolic guidance for society" (Jachtenfuchs 1995: 116). Societal institutions affect the preferences and actions of the members of society:

"Institutions provide moral or cognitive templates for interpretation and action. The individual is seen as an entity deeply imbricated in a world of institutions, composed of symbols, scripts and routines, which provide the filters of interpretation, of both the situation and oneself, out of which a course of action is constructed" (Hall/Taylor 1996: 939)

In modern societies, the constitutional and legal order serves to transform societal norms into specific rules for appropriate behavior. [ 41 ] In judicial decisions, moreover, these legal provisions are constantly being adapted to new situations. In the sociology of law, it has been pointed out that when the law is created by professional judges in such a way, the existing societal norms find their way into the legal order. It is precisely

"public law, from the catalogue of basic rights through to the law of public order, [which] is interspersed with imprecise legal concepts, discretionary rules and references to ideas of the common good, all of which involve the same problem of making their content specific" (Röhl 1987: 225).

As long as societal ideas do not conflict with the recognized principles of the legal order, a judge has to "proceed from dominant societal values" (Rehbinder 1993: 22) when making the content of legal norms specific.

A complex interplay thus exists between societal norms and the norms of the constitutional and legal order. On the one hand, "law consists of those parts of the social structure which have arrived at a particular consolidation" (Röhl 1987: 531) and integrates new developments only after some delay. On the other hand, the various references in legal texts to societal practice as a source of norms (‘morality’) ensure that the constitutional and legal order remains tightly linked with societal norms.

A society’s constitutional and legal order is a suitable indicator of societal norms because the stable societal norms are institutionalized in it and are transformed into specific rules for behavior. In addition, by means of ongoing judicial decisions, the constitutional and legal order remains tied to the norms shared by the society. From a perspective of practical research, a society’s constitutional and legal order is also a suitable indicator because it is public and therefore easily accessible for the scholar.

Thus, a two-stage procedure is necessary when using the constitutional and legal order to identify societally shared, value-based expectations of appropriate foreign policy behavior. First, one must examine the norms for foreign policy in the constitutional and legal order, expressed in the German Basic Law (Grundgesetz), post-constitutional legislation and the rulings of the Federal Constitutional Court. It should not, however, be expected that the constitutional and legal order will set standards for all areas of foreign policy. As a second step, therefore, according to the domestic analogy (see 4.2.4.), provisions must be identified in the constitutional and legal order which set standards for national behavior and which can be expected to guide foreign policy decision-makers in their interaction with the international environment.

5.2.3. Party Programs and Election Platforms

One of the tasks of political parties in a democracy is to articulate and aggregate (foreign) policy expectations in society in order to include them in the (foreign) policy-making process. Thus, party programs and election platforms are suitable indicators of societally shared, value-based expectations of appropriate (foreign) policy behavior (Jachtenfuchs/Diez/Jung 1997) because societal norms are also expressed in them. Although the programmatic work in parties is performed above all by party elites and active party members, the formulation of expectations of behavior is still closely linked with the norms of society as a whole. Programmatic documents in particular serve not only to bring party members to identify themselves with the party but also to present the party to outsiders, offering non-members the opportunity to identify themselves with the respective party's program. [ 42 ] As parties are always competing for the electorate’s support, their programs aim to take up and articulate, but also influence societal norms.

Because party programs (also) serve the purpose to induce party members to identify themselves with the party, there is an incentive to neutralize party-internal dissent by means of compromise formulas and thus to avoid party splits. However, because party programs are also a way of mobilizing the electorate’s support, there is also an incentive to express programmatic aims as specific expectations of behavior, allowing voters to recognize what aims the party is pursuing and in what issues it differs from its rivals. The formulation of programmatic statements in party programs is thus subject to conflicting incentives because of the dual function which party programs fulfil. Moreover, the party leadership is interested in keeping its options open in case it has to assume government responsibility. The consequence is that neither particularly specific nor completely unspecific formulations of expectations of behavior are to be expected in party programs and platforms. Because the aspect of giving party members something to identify with is especially pronounced in the formulation of statements of principle in party programs that are to be valid over a longer period of time, one is more likely to find formulations of societal norms with a low or medium degree of specificity in such documents. When drawing up election platforms, by contrast, the aspect of presenting the party to outsiders (i.e. mobilizing votes) predominates. In election platforms, therefore, medium or even highly specific formulations of value-based expectations of appropriate foreign policy behavior can be expected.

A party’s election results allow one to infer how broad the societal support is for the value-based expectations of appropriate foreign policy behavior articulated by that party. The programs and platforms of splinter parties do not have any significance as indicators of expectations of behavior shared by society as a whole. Whether a party has appreciable societal support for its platform can be seen from whether it is elected to the Bundestag on the basis of that platform. Under German electoral law, parties are represented in the Bundestag either if they receive at least 5% of the votes cast nationwide or win the direct parliamentary mandate in three constituencies. From the perspective of societal constructivism, this means that the parties represented in the Bundestag articulate societal expectations of behavior which are shared either by at least 5% of all voters or by a locally concentrated majority of the electorate. [ 43 ]

If societal value-based expectations of appropriate behavior are articulated only by individual, small parties, albeit represented in the Bundestag, then, according to societal constructivism, these norms must be regarded as having low commonality. Norms have medium commonality, by contrast, if these expectations of behavior are articulated by the CDU and SPD because, as parties representing large parts of the public, "they do not regard themselves as representing specific sections of the population but attempt to appeal to all citizens" (Rudzio 1996: 139). Their platforms thus express those societal expectations that are accorded wide societal support. [ 44 ] A high degree of commonality exists if expectations of behavior are shared by all parties represented in the Bundestag.

5.2.4. Parliamentary Debates

Creating publicity through debate is one of the Bundestag’s most important functions in the area of foreign policy (see Krause 1998: 138). Speeches in parliamentary debates thus can function as a further indicator of societal norms. According to the logic of action assumed by constructivism, MPs are not rational vote-maximizers but the "mouthpiece" of societal norms. MPs’ speeches are the expression of societally shared, value-based expectations of appropriate foreign policy behavior.

In Marijke Breuning’s (1997, 1995) comparative study of development policies, parliamentary debates serve as an indicator of different conceptions of national roles. [ 45 ] An analysis of the content of development policy debates reveals that the Netherlands conceives of its role as an "enlightened vanguard" while the UK sees its role as a "power broker". As the two states’ development policies match these role concepts, [ 46 ] such concepts can be seen as a powerful explanation. Given limited resources, it is justifiable to substitute a (qualitative) interpretation of debates for a time-consuming analysis of content.

As with parties’ political programs and election platforms, parliamentary debates can also reveal differences in the commonality of societal norms. In analogy to the evaluation of party programs, speeches in a parliamentary debate indicate a highly communal norm if speakers from all parties represented in the Bundestag articulate the same value-based expectations of appropriate behavior. If an expectation of behavior is put forward by just one of the two majority parties, then it has only medium commonality. An expectation of behavior articulated by just one majority party and/or by small parties does not possess sufficient commonality for one to expect it to have an influence on German foreign policy.

In identifying German society’s expectations of behavior, the indicator ‘parliamentary debates’ is, of course, not independent of the indicator ‘party programs and election platforms’ because in both cases the norms are propagated by political parties. As members of the Bundestag are generally subject to party discipline, it cannot be expected that the norms they articulate will openly contradict the norms set out in their parties’ programs. It is, however, plausible to assume that societal expectations of behavior are formulated more specifically in Bundestag debates than in party political programs and election platforms because speeches in Bundestag debates usually refer to specific policies in actual situations. The indicator ‘parliamentary debates’ may therefore be a more precise source of societally shared expectations of appropriate behavior than the indicator ‘party programs and election platforms’.

In Bundestag debates, members of the government also exercise their right to speak. However, the speeches of government members can be regarded as an indicator of societally shared expectations of behavior only to a limited extent because they also express official government policy and are thus part of the dependent variable. Even so, there is no risk of circular reasoning when speeches of members of the government and the parliamentary parties supporting them are understood as indicating societal norms and then used to explain government policy. From a constructivist perspective, value-based expectations of appropriate behavior which are only shared by the parties supporting the government do not possess sufficient (i.e. at least medium) commonality for one to expect them to have an influence on foreign policy decision-makers. Only if a value-based expectation of appropriate behavior is shared by a major party apart from the parties supporting the government, or indeed by all parties represented in the Bundestag, can it be expected to shape German foreign policy. [ 47 ]

6. Conceptualization of the Dependent Variable and General Constructivist Prediction

6.1. Norm-Consistent Foreign Policy

The constructivist theory of foreign policy posits that the actions of foreign policy decision-makers are guided by norms, i.e. by intersubjective, value-based expectations of appropriate behavior. German foreign policy which, from a constructivist point of view, is marked by international and/or societal norms can be described as norm-consistent foreign policy.

For constructivists, the prime issue is not whether the norms guiding action are international or societal in origin. Rather, constructivists point to the interplay between the expectations of behavior codified in international institutions and the norms that are shared within a society. For example, there is an international norm of promoting general free trade and dismantling trade barriers (for the following, see also Freund 1999). At the same time, the expectation that - with the exception of the agricultural sector - trade barriers generally be dismantled is also put forward on the societal level, for example by political parties and economic research institutes. In this case, international and societal norms thus converge. Accordingly, constructivists would expect the free trade norm to have a particularly significant influence on German foreign trade policy because it is shared on both the international and the societal level.

In such cases where the expectations of appropriate German behavior are shared on the international and societal level, constructivists expect a highly norm-consistent foreign policy. By contrast, constructivists will always expect foreign policy to be moderately norm-consistent if expectations of appropriate German behavior are significantly pronounced on one level only, i.e. there are norms with at least medium commonality and at least medium specificity either on the level of the international system or on the societal level.

Societal and international expectations of appropriate behavior can, however, also be contradictory. For example, there was no societally shared expectation in the Federal Republic of Germany at the beginning of the 1990s that the Bundeswehr be deployed for ‘out-of area’ missions (for the following, see also Baumann 1999; Duffield 1998: 173-221). Following the end of the East-West conflict, the societal norm of deploying the Bundeswehr in peace-keeping but not peace-making missions conflicted with the other NATO nations’ expectation that the German government should also provide troops for military enforcement action. Societal norms only began to converge with the expectations of behavior set by Germany’s partners in the Alliance after a Constitutional Court ruling of 1994 which declared that military action within systems of collective security, among which the court also included NATO, was compatible with German Basic Law. Up to that time, foreign policy decision-makers had been confronted with two conflicting value-based expectations of behavior addressed to them by two different social systems (NATO on the one hand and German society on the other). In such cases, a constructivist prediction is impossible. A constructivist theory of foreign policy will under such conditions be indeterminate.

6.2. General Constructivist Prediction

Whether the Federal Republic of Germany pursues a foreign policy that is highly, moderately or not at all norm-consistent will have to be studied for each area of German foreign policy individually. The assumption that international and societal norms had remained the same in spite of unification was initially widely held in the debate about German foreign policy (see Katzenstein 1997; Anderson/Goodman 1993). However, in the interest of a well-founded constructivist explanation of German foreign policy, this assumption must be tested empirically. However, the fact that the norms which are significant for foreign policy are firmly embedded in international and societal institutions, making their influence stable and rapid changes unlikely, seems to support this assumption. Nevertheless, profound changes such as the end of the East-West conflict and German unification present unexpected shifts in context even for firmly institutionalized norms and may thus lead to their substantial alteration or modification. A general prediction for German foreign policy after unification can therefore only be formulated with certain reservations:

References

Adler, Emanuel/Haas, Peter M. 1992: Conclusion: Epistemic Communities, World Order, and the Creation of a Reflective Research Program, in: International Organization 46:1, 367-390.

Akehurst, Michael 1992: A Modern Introduction to International Law, repr. 6th ed., London/New York: Routledge.

Anderson, Jeffrey J./Goodman, John B. 1993: Mars or Minerva? A United Germany in a Post-Cold War Europe, in: Keohane, Robert O./Nye, Joseph S./Hoffmann, Stanley (eds.), After the Cold War. International Institutions and State Strategies in Europe, 1989-1991, Cambridge, Mass.: Harvard University Press, 23-62.

Armstrong, David (1994): The Socialization of States. Paper Prepared for the ISA Annual Convention, Washington D.C., 28.3.-1.4.1994.

Axelrod, Robert 1986: An Evolutionary Approach to Norms, in: American Political Science Review 80:4, 1095-1111.

Banchoff, Thomas 1997: German Policy Toward the European Union: The Effect of Historical Memory, in: German Politics 6:1, 60-76.

Barnett, Michael 1996: Identity and Alliances in the Middle East, in: Katzenstein, Peter (ed.), The Culture of National Security. Norms and Identity in World Politics, New York: Columbia UP, 400-447.

Baumann, Rainer 1999: German Security Policy within and vis-à-vis NATO after Unification: Cases, Predictions and Findings, Universität Tübingen: unveröffentlichtes Manuskript.

Baumann, Rainer/Rittberger, Volker/Wagner, Wolfgang 1998: Power and Power Politics: Neorealist Foreign Policy Theory and Expectations about German foreign Policy since Unification (Tübinger Arbeitspapiere zur Internationalen Politik und Friedensforschung No. 30a), Tübingen: University of Tübingen, Center for International Relations/Peace and Conflict Studies.

Berger, Thomas U. 1996: Norms, Identity, and National Security in Germany and Japan, in: Katzenstein, Peter J. (ed.), The Culture of National Security. Norms and Identity in World Politics, New York: Columbia University Press, 317-356.

Berger, Thomas U. 1998: Cultures of Antimilitarism: National Security in Germany and Japan, Baltimore, MD: The Johns Hopkins University Press.

Bergem, Wolfgang 1993: Tradition und Transformation: eine vergleichende Untersuchung zur politischen Kultur in Deutschland, Opladen: Westdeutscher Verlag.

Bienen, Derk/Freund, Corinna/Rittberger, Volker 1999: Societal Interests, Policy Networks and Foreign Policy: An Outline of Utilitarian-Liberal Foreign Policy Theory (Tübinger Arbeitspapiere zur Internationalen Politik und Friedensforschung No. 33a), Tübingen: University of Tübingen, Center for International Relations/Peace and Conflict Studies.

Biersteker, Thomas J./Weber, Cynthia 1996: The Social Construction of State Sovereignty, in: Biersteker, Thomas J./Weber, Cynthia (eds.), State Sovereignty as Social Construct, Cambridge: Cambridge University Press, 1-21.

Breitmeier, Helmut/Rittberger, Volker 1998: Environmental NGOs in an Emerging Global Civil Society (Tübinger Arbeitspapiere zur Internationalen Politik und Friedensforschung No. 32), Tübingen: University of Tübingen, Center for International Relations/Peace and Conflict Studies.

Breuning, Marijke 1995: Words and Deeds: Foreign Assistance Rhetoric and Policy Behavior in the Netherlands, Belgium, and the United Kingdom, in: International Studies Quarterly 39:2, 235-254.

Breuning, Marijke 1997: Culture, History, Role: Belgian and Dutch Axioms and Foreign Assistance Policy, in: Hudson, Valerie M. (ed.), Culture and Foreign Policy, Boulder, Col.: Lynne Rienner, 99-123.

Brysk, A. 1993: From Above and Below: Social Movements, the International System, and Human Rights in Argentina, in: Comparative Political Studies 26:3, 259-285.

Bull, Hedley 1977: The Anarchical Society. A Study of Order in World Politics, Basingstoke: Macmillan.

Bulmer, Simon 1997: Shaping the Rules? The Constitutive Politics of the European Union and German Power, in: Katzenstein, Peter (ed.), Tamed Power. Germany in Europe, Ithaca/London: Cornell University Press, 49-79.

Burley, Anne-Marie 1993a: International Law and International Relations Theory: A Dual Agenda, in: American Journal of International Law 87:2, 205-239.

Burley, Anne-Marie 1993b: Regulating the World: Multilateralism, International Law, and the Protection of the New Deal Regulatory State, in: Ruggie, John Gerard (ed.): Multilateralism Matters. The Theory and Praxis of an Institutional Form, New York etc.: Columbia University Press, 125-156.

Buzan, Barry 1993: From international system to international society: structural realism and regime theory meet the English school, in: International Organization 47:3, 326-352.

Checkel, Jeffrey T. 1997: International Norms and Domestic Politics: Bridging the Rationalist-Constructivist Divide, in: European Journal of International Relations 3:4, 473-495.

Checkel, Jeffrey T. 1998: The Constructivist Turn in International Relations Theory, in: World Politics 50:2, 324-348.

Claude, Inis L. Jr. (1966), Collective Legitimization as a Political Function of the United Nations, in: International Organization 20:3, 367-379.

Coplin, William 1969: International Law and Assumptions about the State System, in: Rosenau, James (ed.): International Politics and Foreign Policy, New York/London: Free Press, 142-152.

Cortell, Andrew/Davis, James 1996: How Do International Institutions Matter? The Domestic Impact of International Rules and Norms, in: International Studies Quarterly 40:4, 451-478.

Czaplinski, Wladyslaw 1998: Minderheitenschutz durch Nachbarschaftsverträge oder durch politische Vereinbarungen? Eine Antwort aus polnischer Sicht, in: Heintze, Hans-Joachim (ed.), Moderner Minderheitenschutz. Rechtliche oder politische Absicherung? Bonn: Dietz, 160-177.

Dessler, David 1989: What's at Stake in the Agent-Structure Debate?, in: International Organization 43:3, 441-473.

Dobler, Wolfgang 1989: Aubenpolitik und öffentliche Meinung, Frankfurt/Main: Haag + Herchen.

Doxey, Margaret P. 1996: International Sanctions in Contemporary Perspective, 2nd ed., Basinstoke etc.: Macmillan/St. Martin's Press.

Duffield, John S. 1998: World Power Forsaken. Political Culture, International Institutions, and German Security Policy After Unification, Stanford, CA: Stanford University Press.

Engelmann, Daniela/Knopf, Hans-Joachim/Roscher, Klaus/Risse, Thomas 1997: Identität und Europäische Union: Die Diskussion um den Euro in Grobbritannien, Frankreich und Deutschland, in: König, Thomas/Rieger, Elmar/Schmitt, Hermann (eds.): Europäische Institutionenpolitik, Frankfurt/New York: Campus, 79-95.

Fagan, Richard 1969: The Transformation of Political Culture in Cuba, Stanford: Stanford University Press.

Finnemore, Martha 1993: International Organizations as Teachers of Norms: The United Nations Educational, Scientific, and Cultural Organization and Science Policy, in: International Organization 47:4, 565-597.

Finnemore, Martha 1996a: National Interests in International Society, Ithaca, NY: Cornell University Press.

Finnemore, Martha 1996b: Norms, Culture, and World Politics: Insights from Sociology's Institutionalism, in: International Organization 50:2, 325-347.

Finnemore, Martha/Sikkink, Kathryn 1998: International Norms Dynamics and Political Change, in: International Organization 52:4, 887-917.

Florini, Ann 1996: The Evolution of International Norms, in: International Studies Quarterly 40:3, 363-390.

Forschungsgruppe Menschenrechte (Sieglinde Gränzer/Anja Jetschke/Thomas Risse/Hans Peter Schmitz) 1998: Internationale Menschenrechtsnormen, transnationale Netzwerke und politischer Wandel in den Ländern des Südens, in: Zeitschrift für Internationale Beziehungen 5:1, 5-41.

Franck, Thomas M. 1990: The Power of Legitimacy among Nations, New York: Oxford University Press.

Freund, Corinna 1999: German Foreign Trade Policy: Cases, Predictions and Empirical Results, Universität Tübingen: unveröffentlichtes Manuskript.

Garrett, Geoffrey/Weingast, Barry R. 1993: Ideas, Interests, and Institutions. Constructing the European Community's Internal Market, in: Goldstein, Judith/Keohane, Robert O. (eds.): Ideas and Foreign Policy. Beliefs, Institutions, and Political Change, Ithaca, NY: Cornell University Press, 173-206.

Gehring, Thomas 1994: Der Beitrag von Institutionen zur Förderung der internationalen Zusammenarbeit. Lehren aus der institutionellen Struktur der Europäischen Gemeinschaft, in: Zeitschrift für Internationale Beziehungen 1:2, 211-242.

Goertz, Gary/Diehl, Paul F. 1992: Toward a Theory of International Norms. Some Conceptual and Measurement Issues, in: Journal of Conflict Resolution 36:4, 634-664.

Goldstein, Judith/Keohane, Robert O. 1993: Ideas and Foreign Policy. An Analytical Framework, in: Goldstein, Judith/Keohane, Robert O. (eds.): Ideas and Foreign Policy. Beliefs, Institutions, and Political Change, Ithaca, NY: Cornell University Press, 3-30.

Grob, Jürgen 1995: Die eingebildete Ohnmacht. Internationale Staatengemeinschaft und lokale Kriege, (Hamburger Beiträge zur Friedensforschung und Sicherheitspolitk 95), Hamburg: Institut für Friedensforschung und Sicherheitspolitik an der Universität Hamburg.

Haas, Peter M. 1992: Introduction. Epistemic Communities and International Policy Coordination, in: International Organization 46:1, 1-35.

Hall, Peter A. (ed.) 1989: The Political Power of Economic Ideas. Keynesianism Across Nations, Princeton, NJ: Princeton University Press.

Hall, Peter A./Taylor, Rosemary C. R. 1996: Political Science and the Three New Institutionalisms, in: Political Studies 44:4, 936-957.

Hasenclever, Andreas/Mayer, Peter/Rittberger, Volker 1997: Theories of International Regimes, Cambridge: Cambridge University Press.

Henkin, Louis 1968: How Nations Behave. Law and Foreign Policy, New York: Praeger.

Hofferbert, Richard I./Klingemann, Hans-Dieter 1990: The Policy Impact of Party Programmes and Government Declarations in the Federal Republic of Germany, in: European Journal of Political Research 18:3, 277-304.

Holsti, Ole R. 1992: Public Opinion and Foreign Policy: Challenges to the Almond-Lippmann Consensus, in: International Studies Quarterly 36:4, 439-466.

Holsti, Ole R. 1996: Public Opinion and American Foreign Policy, Ann Arbor, MI: The University of Michigan Press.

Hudson, Valerie M. (ed.) 1997: Culture and Foreign Policy, Boulder CO: Lynne Rienner.

Hurrell, Andrew 1993: International Society and the Study of Regimes, in:

Rittberger, Volker (ed.): Regime Theory and International Relations, Oxford: Clarendon Press, 49-72.

Hurrell, Andrew 1995: Regionalism in Theoretical Perspective, in: Fawcett, Louise/Hurrell, Andrew (eds.), Regionalism in World Politics. Regional Organization and International Order, Oxford: Oxford University Press, 37-73.

Ingebritsen, Christine/Larson, Susan 1997: Interest and Identity: Finland, Norway and European Union, in: Cooperation and Conflict 32:2, 207-222.

Jachtenfuchs, Markus 1995: Theoretical Perspectives on European Governance, in: European Law Journal 1:2, 115-133.

Jachtenfuchs, Markus/Diez, Thomas/Jung, Sabine 1997: Ideas and Integration. Conflicting Models of a Legitimate European Political Order. Paper for presentation at the Fifth Biannual ECSA Conference, Seattle, May 29th.

Jackson, Robert H. 1990: Quasi-States. Sovereignty, International Relations, and the Third World, Cambridge: Cambridge University Press.

Jepperson, Ronald/Wendt, Alexander/Katzenstein, Peter 1996: Norms, Identity, and Culture in National Security, in: Katzenstein, Peter (ed.): The Culture of National Security, New York: Columbia University Press, 33-75.

Johnson, Nevil 1978: Law as the Articulation of the State in Western Germany: A German Tradition Seen from a British Perspective, in: West European Politics 1:2, 177-192.

Kaase, Max 1980: Sinn oder Unsinn des Konzepts `Politische Kultur' für die vergleichende Politikforschung, oder auch: Der Versuch, einen Pudding an die Wand zu nageln, in: Kaase, Max/Klingemann, Hans-Dieter (eds.): Wahlen und politisches System, Opladen: Westdeutscher Verlag, 144-172.

Katzenstein, Peter J. 1993: Coping with Terrorism: Norms and Internal Security in Germany and Japan, in: Goldstein, Judith/Keohane, Robert (eds.): Ideas and Foreign Policy: Beliefs, Institutions and Political Change, Ithaca, N.Y.: Cornell University Press, 265-296.

Katzenstein, Peter (ed.) 1997: Tamed Power: Germany in Europe, Ithaca, NY: Cornell University Press.

Keck, Margaret E./Sikkink, Kathryn (eds.) 1998: Activists Beyond Borders: Advocacy Networks in International Politics, New York: Ithaca, N.Y.: Cornell University Press.

Keohane, Robert O. 1983: The Demand for International Regimes, in: Krasner, Stephen D. (ed.): International Regimes, Ithaca, NY: Cornell University Press, 141-171.

Keohane, Robert O. 1989a: International Institutions. Two Approaches, in: id. (ed.): International Institutions and State Power. Essays in International Relations Theory, Boulder, CO: Westview Press, 158-179.

Keohane, Robert O. 1989b: Neoliberal Institutionalism. A Perspective on World Politics, in: id. (ed.): International Institutions and State Power. Essays in International Relations Theory, Boulder, CO: Westview Press, 1-20.

Keohane, Robert O./Nye, Joseph S. (eds.) 1972: Transnational Relations and World Politics, Cambridge, MA: Harvard University Press.

Kittel, Gabriele/Rittberger, Volker/Schimmelfennig, Frank 1995: Staatenmerkmale und Aubenpolitik: Untersuchungsdesign und Hypothesen, in: Rittberger, Volker (ed.): Anpassung oder Austritt: Industriestaaten in der UNESCO-Krise. Ein Beitrag zur vergleichenden Aubenpolitikforschung, Berlin: edition sigma, 53-82.

Klingemann, Hans-Dieter/Volkens, Andrea 1997: Struktur und Entwicklung von Wahlprogrammen in der Bundesrepublik Deutschland 1949-1994, in: Gabriel, Oscar W./Niedermayer, Oskar/Stöss, R. (eds.): Parteiendemokratie in Deutschland, Opladen: Westdeutscher Verlag, 517-536.

Klingemann, Hans-Dieter/Hofferbert, Richard I./Budge, Ian 1994: Parties, Policies, and Democracy, Boulder, CO: Westview Press.

Klotz, Audie 1995: Norms in International Relations: The Struggle Against Apartheid, Ithaca, N.Y.: Cornell University Press.

Krasner, Stephen D. 1983a: Regimes and the Limits of Realism: Regimes as Autonomous Variables, in: Krasner, Stephen D. (ed.): International Regimes, Ithaca, NY: Cornell University Press, 355-368.

Krasner, Stephen D. 1983b: Structural Causes and Regime Consequences: Regimes as Intervening Variables, in: Krasner, Stephen D. (ed.): International Regimes, Ithaca, NY: Cornell University Press, 1-21.

Krasner, Stephen D. 1993a: Power, Polarity, and the Challenge of Disintegration, in: Haftendorn, Helga/Tuschoff, Christian (eds.): America and Europe in an Era of Change, Boulder, CO: Westview Press, 21-42.

Krasner, Stephen D. 1993b: Sovereignty, Regimes, and Human Rights, in: Rittberger, Volker (ed.): Regime Theory and International Relations, Oxford: Clarendon Press, 139-167.

Kratochwil, Friedrich V. 1989: Rules, Norms, and Decisions. On the Conditions of Practical and Legal Reasoning in International Relations and Domestic Affairs, Cambridge: Cambridge University Press.

Kratochwil, Friedrich V./Ruggie, John Gerard 1986: International Organization. A State of the Art on an Art of the State, in: International Organization 40:4, 753-775.

Krause, Joachim 1998: Die Rolle des Bundestags in der Aubenpolitik, in: Eberwein, Wolf-Dieter/Kaiser, Karl (eds.): Institutionen und Ressourcen (Deutschlands neue Aubenpolitik vol. 4), Munich: Oldenbourg, 137-150.

Legro, Jeffrey 1997: Which norms matter? Revisiting the "failure" of internationalism, in: International Organization 51:1, 31-64.

Lijphart, Arend 1980: The Structure of Inference, in: Almond, Gabriel/Verba, Sidney (eds.): The Civic Culture Revisited, Boston/Toronto: Brown & Co., 37-56.

Little, Richard/Smith, Steve (eds.) 1988: Belief Systems and International Relations, Oxford: Blackwell.

Lumsdaine, David Halloran 1993: Moral Vision in International Politics. The Foreign Aid Regime, 1949-1989, Princeton, NJ: Princeton University Press.

March, James G./Olsen, Johan P. 1989: Rediscovering Institutions. The Organizational Basis of Politics, New York: Free Press.

Marucssen, Martin/Risse, Thomas 1997: A Europeanization of Nation-State Identities? Conceptual Considerations and Research Design. Paper Prepared for the workshop `Theoretical, Methodological and Empirical Issues Involved in the Study of National Identities' at the EUI, 21st-22nd November 1997.

McElroy, Robert W. 1992: Morality and American Foreign Policy. The Role of Ethics in International Affairs, Princeton, NJ: Princeton University Press.

McNeely, Connie L. 1995: Constructing the Nation-State. International Organization and Prescriptive Action, Westport, CT: Greenwood Press.

Mearsheimer, John J. 1990: Back to the Future. Instability in Europe After the Cold War, in: International Security 15:1, 5-56.

Mercer, Jonathan 1996: Reputation and International Politics, Ithaca, NY: Cornell University Press.

Messner, Dirk/Nuscheler, Franz (eds.) 1996: Weltkonferenzen und Weltberichte: Ein Wegweiser durch die internationale Diskussion, Bonn: J.H.W. Dietz.

Meyer, John 1987: The World Polity and the Authority of the Nation-State, in: Thomas, George M./Meyer, John W./Ramirez, Francisco O./Boli, John (eds.), Institutional Structure: Constituting State, Society, and the Individual, Beverly Hills: Sage, 41-70.

Meyer, John W./Ramirez, Francisco O./Rubinson, Richard/Boli-Bennett, John 1979: The World Educational Revolution, 1950-70, in: Meyer, John W./Hannan, Michael T. (eds.), National Development and the World System. Educational, Economic and Political Change, 1950-1970, Chicago: The University of Chicago Press, 37-55.

Müller, Harald 1993a: Die Chance der Kooperation. Regime in den internationalen Beziehungen, Darmstadt: Wissenschaftliche Buchgesellschaft.

Müller, Harald 1993b: The Internalization of Principles, Norms and Rules by Governments: The Case of Security Regimes, in: Rittberger, Volker (ed.), Regime Theory and International Relations, Oxford: Clarendon Press, 360-388.

Noel, Alain/Thérien, Jean-Philippe 1995: From Domestic to International Justice: The Welfare State and Foreign Aid, in: International Organization 49:3, 523-553.

Onuf, Nicholas Greenwood 1989: World of Our Making. Rules and Rule in Social Theory and International Relations, Columbia, SC: University of South Carolina Press.

Pappi, Franz Urban 1986: Politische Kultur: Forschungsparadigma, Fragestellungen, Untersuchungsmöglichkeiten, in: Kaase, Max (ed.): Politische Wissenschaft und politische Ordnung, Opladen: Westdeutscher Verlag, 279-291.

Parsons, Talcott 1951: The Social System, London: Routledge & Kegan Paul.

Parsons, Talcott 1961: Order and Community in the International System, in: Rosenau, James N. (ed.): International Politics and Foreign Policy. A Reader in Research and Theory, New York: Free Press of Glencoe, 120-129.

Pye, Lucian W. 1962: Politics, Personality, and Nation Building, New Haven: Yale University Press.

Ramirez, Francisco O./Weiss, Jane (1979), The Political Incorporation of Women, in: Meyer, John W./Hannan, Michael T. (eds.), National Development and the World System. Educational, Economic and Political Change, 1950-1970, Chicago: The University of Chicago Press, 238-249.

Raymond, Gregory A. 1997: Problems and Prospects in the Study of International Norms, in: Mershon International Studies Review 41:2, 205-245.

Rehbinder, Manfred 1993: Rechtssoziologie, 3rd ed., Berlin/New York: de Gruyter.

Reus-Smit, Christian 1997: The Constitutional Structure of International Society and the Nature of Fundamental Institutions, in: International Organization 51:4, 555-589.

Richter, Emanuel 1990: Weltgesellschaft und Weltgemeinschaft. Begriffsverwirrung und Erklärungsversuche, in: Politische Vierteljahresschrift 31, 275-297.

Risse-Kappen, Thomas 1991: Public Opinion, Domestic Structure, and Foreign Policy in Liberal Democracies, in: World Politics 43:4, 479-512.

Risse-Kappen, Thomas 1994: Ideas Do Not Float Freely. Transnational Coalitions, Domestic Structures, and the End of the Cold War, in: International Organization 48:2, 185-214.

Risse-Kappen, Thomas 1995: Bringing Transnational Relations Back In: Introduction, in: id. (ed.): Bringing Transnational Relations Back In. Non-State Actors, Domestic Structures, and International Institutions, Cambridge: Cambridge University Press, 3-33.

Rittberger, Volker 1992: Nach der Vereinigung - Deutschlands Stellung in der Welt, in: Leviathan 20:2, 207-229.

Rittberger, Volker 1991: Konferenzen, in: Wolfrum, Rüdiger (ed.), Handbuch Vereinte Nationen, 2nd. rev. ed., Munich: C.H. Beck, 410-418.

Rittberger, Volker/Schrade, Christina/Schwarzer, Daniela 1999: Transnational civil society actors and the quest for security: Introduction, in: Alagappa, Mutiah/Inoguchi, Takashi (eds.), International Security Management and the United Nations, Tokyo: The United Nations University Press, 109-138.

Rittberger, Volker/Zangl, Bernhard 1995: Internationale Organisationen. Politik und Geschichte, 2nd ed., Opladen: Leske + Budrich.

Röhl, Klaus 1987: Rechtssoziologie, Cologne/Berlin/Bonn/Munich: Carl Heymanns Verlag.

Rohe, Karl 1994: Politische Kultur. Zum Verständnis eines theoretischen Konzepts, in: Niedermayer, Oskar/Beyme, Klaus von (eds.): Politische Kultur in Ost- und Westdeutschland, Berlin: Akademie Verlag, 1-21.

Rudzio, Wolfgang 1996: Das politische System in der Bundesrepublik Deutschland, 4th ed., Opladen: Leske + Budrich.

Ruggie, John Gerard 1993: Territoriality and Beyond, in: International Organization 47, 139-174.

Ruggie, John Gerard 1998: Introduction: What Makes the World Hang Together? Neo-Utilitarian and the Social Constructivist Challenge, in: Ruggie, John Gerard (ed.): Constructing the World Polity. Essays on International Relations, London/New York: Routledge, 1-44.

Russett, Bruce 1990: Controlling the Sword: The Democratic Governance of National Security, Cambridge, MA: Harvard Univ. Press.

Russett, Bruce 1993: Grasping the Democratic Peace. Principles for a Post-Cold War World, Princeton, NJ: Princeton University Press.

Sabatier, Paul 1993: Advocacy-Koalitionen, Policy-Wandel und Policy-Lernen: Eine Alternative zur Phasenheuristik, in: Héritier, Adrienne (ed.): Policy-Analyse. Kritik und Neuorientierung (PVS Sonderheft 24), Opladen: Westdeutscher Verlag, 116-148.

Schaber, Thomas/Ulbert, Cornelia 1994: Reflexivität in den Internationalen Beziehungen. Literaturbericht zum Beitrag kognitiver, reflexiver und interpretativer Ansätze zur dritten Theoriedebatte, in: Zeitschrift für Internationale Beziehungen 1:1, 139-169.

Schimmelfennig, Frank 1994: Internationale Sozialisation neuer Staaten. Heuristische Überlegungen zu einem Forschungsdesiderat, in: Zeitschrift für Internationale Beziehungen 1:2, 335-355.

Searle, John R. 1969: Speech Acts. An Essay in the Philosophy of Language, Cambridge: Cambridge University Press.

Seidl-Hohenveldern, Ignatz 1997: Völkerrecht, 9th ed., Cologne: Heymann.

Sikkink, Kathryn 1993: Human Rights, Principled Issue Networks, and Sovereignty in Latin Ameria, in: International Organization 47:3, 411-441.

Thomson, Janice E. 1995: State Sovereignty and International Relations: Bridging the Gap Between Theory and Empirical Research, in: International Studies Quarterly 39: 2, 213-233.

Tönnies, Ferdinand 1991: Gemeinschaft und Gesellschaft. Grundbegriffe der reinen Soziologie, 3rd ed., Darmstadt: Wissenschaftliche Buchgesellschaft.

Ulbert, Cornelia 1997: Ideen, Institutionen und Kultur. Die Konstruktion (inter-) nationaler Klimapolitik in der BRD und in den USA, in: Zeitschrift für Internationale Beziehungen 4:1, 9-40.

Verdross, Alfred/Simma, Bruno 1984: Universelles Völkerrecht. Theorie und Praxis, 3rd ed., Berlin: Duncker & Humblot.

Waltz, Kenneth N. 1979: Theory of International Politics, New York: Random House.

Weart, Spencer 1994: Peace among Democratic and Oligarchic Republics, in: Journal of Peace Research 31:3, 299-316.

Weber, Max 1968: Economy and Society. An Outline of Interpretive Sociology, ed. Guenther Roth/Claus Wittich, New York: Bedminster Press.

Weiß, Wolfgang W. 1986: Sozialisation, in: Schäfers, Bernhard (ed.), Grundbegriffe der Soziologie, 2nd ed., Opladen: Leske + Budrich, 269-271.

Wendt, Alexander E. 1987: The Agent-Structure Problem in International Relations Theory, in: International Organization 41:3, 335-370.

Wendt, Alexander 1991: Bridging the Theory/Meta-theory Gap in International Relations, in: Review of International Studies 17:4, 383-392.

Wendt, Alexander 1994: Collective Identity Formation and the International State, in: American Political Science Review 88:2, 384-396.

Wendt, Alexander/Duvall, Raymond 1989: Institutions and International Order, in: Czempiel, Ernst-Otto/Rosenau, James N. (eds.), Global Changes and Theoretical Challenges. Approaches to World Politics for the 1990s, Lexington: Lexington Books, 51-73.

Wildenmann, Rudolf 1992: Wahlforschung, Mannheim/Leipzig/Wien/Zürich: B.I.Taschenbuchverlag.

Yee, Albert S. 1996: The Causal Effects of Ideas on Policies, in: International Organization 50:1, 69-108.

Young, Oran R. 1989: International Cooperation. Building Regimes for Natural Resources and the Environment, Ithaca, NY: Cornell University Press.

Zürn, Michael 1992: Interessen und Institutionen in der internationalen Politik. Grundlegung und Anwendung des situationsstrukturellen Ansatzes, Opladen: Leske + Budrich.

[ 1 ]
This paper was written in the context of a research project on "German Foreign Policy after Reunification" funded by the German Research Association (Deutsche Forschungsgemeinschaft, DFG). Our special thanks are due to our colleagues Rainer Baumann, Derk Bienen, Corinna Freund, Simon Gajer, Daria Nashat, Dirk Peters and Christian Westhoff. For valuable comments on earlier versions of this paper we would also like to thank Sebastian Bartsch, Thomas Bernauer, Helmut Breitmeier, Andreas Hasenclever, Gunther Hellmann, Peter Mayer, Martin Mogler, Peter Moser, Thomas Nielebock, Frank Schimmelfennig, Hans Peter Schmitz and Bernhard Zangl. Our thanks are also due to Philip Mann who translated the German version into English.

[ 2 ]
The term "constructivist" has prevailed over other terms such as "reflexive" (see Keohane 1989, Schaber/Ulbert 1994) or "interpretative" (see Klotz 1995) and will therefore also be used by us in the present paper.

[ 3 ]
Even if there is constancy on the part of both the independent ("norms") and the dependent variable ("foreign policy behavior"), an influence of norms on behavior cannot be inferred with certainty from this correlation. A test of constructivist hypotheses is only possible if the independent variable reveals variance (such as the variable "German power" in the period between 1980 and 1998).

[ 4 ]
On definitions of norms, see also Legro 1997: 33, Finnemore 1996a: 22, Goldstein/Keohane 1993: 20, Parsons 1961: 120.

[ 5 ]
Our exclusive usage of the masculine form is based on considerations of practicability only and is not intended to represent gender discrimination.

[ 6 ]
For their part, ideas can include values in the sense of "principled beliefs held by individuals" as well as "causal beliefs".

[ 7 ]
This point is also made by Legro (1997: 35), who uses the term "concordance" to signify the "degree of intersubjective agreement".

[ 8 ]
The distinction between a "regulative" and a "constitutive" effect of norms, which was first made in linguistics in the context of speech act theory by John Searle (1969), is treated differently in the constructivist literature. Some authors regard regulative and constitutive norms as different categories. In their view, constitutive norms do not generate any specific expectations of behavior because their function lies in the constitution of actors’ identities, not in the regulation of their behavior (e.g. Klotz 1995). Others hold that each norm has a regulative and a constitutive function because, as a "motive", it constitutes a practice and hence constitutes a certain social role of the actor (z.B. Finnemore 1996a; Onuf 1989).

[ 9 ]
In sociology, the distinction between primary and secondary socialization is common. Primary socialization takes place when a child grows into its family and society and is therefore completed when the child becomes an adult. Secondary socialization, by contrast, signifies the continuing process of an actor’s internalizing standards of appropriate behavior. Secondary socialization is particularly significant for constructivist foreign policy theory while primary socialization only plays a role with respect to the growing into international society of "new" states such as the newly independent former Soviet republics (Schimmelfennig 1994) or with respect to the way in which new political elites grow into their role as the government of a country following a change in that country's domestic political system (Armstrong 1994: 25ff.).

[ 10 ]
As on the societal level, it is the individual foreign policy decision-makers themselves who, in their function as representatives of their state, internalize internationally shared expectations of behavior. However, as the internationally shared expectations of behavior to which foreign policy decision-makers are subject are directed towards the state they represent, one generally speaks of states being socialized (see Checkel 1997: 477; Raymond 1997: 216; McElroy 1992: 40-43).

[ 11 ]
In international law, international (governmental) organizations are described as "created" subjects of international law (e.g. Verdross/Simma 1984), indicating that, unlike states (which are, as it were, the "natural" subjects of international law), they are themselves not constitutive units of the international community because they are themselves constituted by a voluntary association of states.

[ 12 ]
Studies on the activity and significance of transnational advocacy coalitions emphasize their role as a link between expectations of behavior shared on the international level and those on the level of national societies. Transnational advocacy coalitions thus gain increasing importance in inverse proportion to the degree of receptiveness shown by governments to the expectations of behavior that their social environments address to them. By appealing to internationally shared expectations of behavior, transnational advocacy coalitions attempt to mobilize international pressure in order to incite the government in question to modify its behavior ("boomerang pattern", see Keck/Sikkink 1998: 12f.; see also the improved "spiral model" which illustrates the process from appeal to international norms by societal actors and transnational advocacy coalitions to eventual norm-consistent behavior by the government in question, in Forschungsgruppe Menschenrechte 1998).

[ 13 ]
For a discussion of the motives and effects of sanctions which defines and discusses the links between material and immaterial goals (albeit with an emphasis on material goals and effects), see Doxey (1996: 54-65).

[ 14 ]
The norm of the international protection of human rights is one example of this. The UN Charter proclaimed the universal protection of human rights as one of the main goals of this organization (Article 1 (3) and Article 55) but limited the UN's authority to intervene in states’ domestic affairs in order to protect human rights to situations jeopardizing world peace (Article 2 No. 7). Since the 1970s, however, the UN member states have increasingly adopted the view that gross violations of human rights by certain governments can at least be subjected to fact-finding investigations and public criticism by UN bodies even if they do not pose a risk to world peace. In the 1990s, finally, the UN Security Council also sanctioned military intervention to restore human rights and democracy without having explicitly stated that the situations in question posed any risk to world peace (e.g. in Somalia and Haiti).

[ 15 ]
As already mentioned, a norm must have at least medium commonality and specificity in order to be classed as "present" for the purposes of this ranking diagram.

[ 16 ]
Alexander Wendt, for example, states that "neither state agents nor the domestic and international system structures which constitute them should be treated always as given or primitive units; theories of international relations should be capable of providing explanatory leverage on both. This does not mean that a particular research endeavor cannot take some things as primitive: scientific practice has to start somewhere" (Wendt 1987: 349, our emphasis; see also Dessler 1989: 443f.).

[ 17 ]
The "classical" definition of transnational relations assumes the non-state constitution of at least one of the interaction partners (Keohane/Nye 1971: xii; Risse-Kappen 1995: 3). Here, our use of the label "transnational" follows other authors (e.g. McNeely 1995) in a sense which includes non-governmental and inter-governmental constitutive and communicative processes of international norms. We are using this term in order to underline the fact that these processes do not necessarily occur solely with the inclusion of governmental actors.

[ 18 ]
Our terminology follows the term "international society" which has been common in the English literature on international relations since Hedley Bull (1977), used to signify a "group of states, conscious of certain common interests and common values, [...] conceiv[ing] themselves to be bound by a common set of rules in their relations with one another, and shar[ing] in the working of common institutions" (Bull 1977: 13). Unlike Bull, however, and in line with the fundamental assumption of constructivism, we assume that norms constitute the identities and thus the interests of actors in international society, and not vice versa. In our usage, therefore, the term "international society" does not imply any dichotomous understanding of the terms "society" and "community" such as Tönnies's (1991) who defines "community" as organic and constituted by common values and mutual affection, but "society", by contrast, as constituted by interests and exchange (See Tönnies: 3f., 7, 34ff.; see also Richter 1990). As Buzan (1993) has shown, at least a social subsystem of international society, and theoretically international society as a whole, can gradually develop the features of a community among its members in the course of the traditional perpetuation of norm-guided interaction. For more detailed descriptions of the transnational-constructivist structural model of international society, see also Finnemore (1996a); Hurrell (1993); and Franck (1990). For a useful discussion of the conceptual differences between "international system" and "international society", see Buzan (1993: 331-333).

[ 19 ]
Finnemore (1996) has demonstrated the isomorphism of state behavior by means of case studies, for example, on the establishment of state scientific bureaucracies or on the norm of international humanitarian law that the wounded should enjoy treatment without regard to which war party they belong to. Other case studies on the institutionalization of education (Meyer/Ramirez/Rubin-son/Boli-Bennett 1979) or the establishment of women’s franchise (Ramirez/Weiss 1979) have arrived at the same result.

[ 20 ]
For definitions of "global (civil) society" and discussions of the relevance of the concept in International Relations, see Rittberger/Schrade/Schwarzer 1999: 111-115; Breitmeier/Rittberger 1999: 4-13; Buzan 1993: 336-340; Richter 1990).

[ 21 ]
On such "demand" approaches, see, among others, Keohane (1983). For a critique of the demand-based approach to explaining norm-guided interaction see McNeely (1995: 12f.) and Finnemore (1996a: 12).

[ 22 ]
Besides norm communication and diffusion, transnational advocacy coalitions can also play an important role in the creation of international norms, as has been shown by the examples of the worldwide struggle against South African apartheid (Klotz 1995) or the assertion of human rights and democracy in the countries of the Third World (Forschungsgruppe Menschenrechte 1998; Keck/Sikkink 1998: 79ff.; Brysk 1993; Sikkink 1993). However, for the generation of constructivist predictions for German foreign policy, these aspects of norm creation can be ignored.

[ 23 ]
For a definition of social institutions, see also Hall/Taylor (1996: 938). For a definition of international institutions, see McNeely (1995: 19), Keohane (1989a: 162f.) and Young (1989: 5f.).

[ 24 ]
Thus, for instance, the prohibition of external intervention in states’ domestic affairs is part of the complex of expectations of appropriate behavior bound together in the institution of sovereignty. The successive redefinition of the prohibition of intervention, and therefore of the institution "sovereignty", since the Second World War is also evidence of the historical contingency and changeability of the constitution of international society. In earlier times, for example, the way people were treated within the jurisdiction of a state was part of the "domestic affairs" protected by the proscription of intervention, whereas today the violation of human rights, at least on a serious, mass scale, is regarded as a legitimate concern of international society and no longer solely as the domestic affair of an individual state protected from intervention by virtue of the institution of sovereignty.

[ 25 ]
For example, the UN Economic and Social Council (ECOSOC) and its subsidiary bodies have established a "consultative arrangement" with non-governmental organizations working in their various areas of responsibility, thus granting NGOs legitimate participatory rights below the level of formal membership (see ECOSOC-Res. 1996/31 of July 25, 1996).

[ 26 ]
Keohane (1989a: 4) lists conventions as a further type of international institution and defines them as "informal institutions, with implicit rules and understandings, that shape the expectations of actors." Particularly when specific regulations (regimes) are missing, he ascribes a behavior-regulating effect to conventions. This implies that conventions can be ascribed to the level of "fundamental" institutions rather than to that of specific institutions. Moreover, Keohane’s understanding of conventions also raises problems concerning their distinction from norms. For example, the example of an international convention – reciprocity – cited by Keohane embodies in the first instance only the norm of repayment in kind. The convention of "reciprocity" only results when states take it for granted that there will de facto be repayment in kind and behave according to this expectation.

[ 27 ]
In U.S. clean-air policy, for example, Sabatier distinguishes between a ‘clean air coalition’ and an ‘economic efficiency coalition’ which differ above all in their view as to "what extent individual freedom in a market economy should be restricted in order to protect the health of ‘endangered sections of the population’" (Sabatier 1993: 133).

[ 28 ]
Exceptions prove this rule: within a society, there will, of course, be critics of the norms held by society as a whole. Typically, however, this criticism is countered by the argument that it has no roots in the basic values generally accepted by society. Criticism of the norms shared by society as a whole thus always bears the stigma of ‘wanting a different society’ or of ‘standing outside society’.

[ 29 ]
The example of the democratic peace is also fruitful in that the societal-constructivist explanation does not maintain that nationally valid norms can be transferred to international contexts at all costs. Non-violent settlement of conflict is regarded as the preference of democratic states which is put into practice as far as possible (preference over outcomes). Its realization will always come up against its limits when the conflicts are with non-democratic states prepared to use violence. In such situations, the societal-constructivist perspective will never expect democratic countries to pursue a foreign policy of ‘peace at any price’.

[ 30 ]
Further examples of societal-constructivist explanations of foreign policy behavior which assume a transfer of domestic norms along the lines of the domestic analogy can be found in Noel/Thérien 1995, Katzenstein 1993 and Lumsdaine 1993.

[ 31 ]
In international legal theory, the question has been discussed whether international law can be regarded as "law" at all. Followers of the positivist school of legal thought object that "law" requires both a legislator who is superior to the actors bound by the legal norms and an effective enforcement of its prescriptions. However, as there does not exist any body which is superior to states, international law cannot therefore be understood as "law" in the strict sense of the word. Followers of the sociological school of legal thought counter that even compliance with national law, which satisfies the above two conditions, only requires enforcement in exceptional cases. From a sociological perspective, law exists within a social system as soon as the legal character of a certain norm is recognized by its constituent units. If, as in constructivist theory, we accept the existence of "international society" as a social system, then there is no reason to deny the legal status at least of those norms whose character as law is recognized within this social system. For summaries of the arguments put forward by the followers of the different schools of legal thought in the debate about the status of international law as "law", see Seidl-Hohenveldern (1997: 4-12), Akehurst (1992: 1-11), and Franck (1990: 27-40). For an instructive discussion of the obvious affinity between the positions of positivist legal theory and the realist school of international relations with regard to the effect of international law on state behavior in international politics, as well as further descriptions of sociologically inspired counter-arguments from the spheres of international law and political science, see also Burley (1993a).

[ 32 ]
Of course, the law of international treaties allows the states parties to exclude certain treaty obligations or to modify their content by means of reservations. However, such reservations must not run contrary to the overall intent of the treaty. Following an ICJ judgement on the UN Convention on the Prevention and Punishment of the Crime of Genocide, this ruling replaced the previously valid ruling that all states had to agree to the reservations of individual member states, and was also set down in Articles 19 to 21 of the Vienna Convention of the Law of Treaties (Akehurst 1992: 129-131).

[ 33 ]
In view of the unequal regional distribution of voting rights in the UN General Assembly, however, medium commonality can only be inferred if the majority voting in favor of such a resolution is made up of states from different regions.

[ 34 ]
However, certain international conferences serve to negotiate and adopt international treaties. This applies e.g. to the UN Conference on the Law of the Sea and the Rome Diplomatic Conference which adopted the statute of the International Criminal Court in June/July 1998.

[ 35 ]
To illustrate this, Messner/Nuscheler (1996: 165) refer to the right to development. Originally proclaimed in 1986 in a UN General Assembly resolution which was not endorsed by Western industrial nations, it was then confirmed with these countries’ approval at the 1992 Rio Summit and the 1993 Vienna World Human Rights Conference. Its contents were modified in such a way that the focus on the human individual as the bearer of the right which had been demanded by the West was given clear prominence. In fact, Western industrialized nations have since then played an important role in the UN in implementing the (individual) right to development, for example, by means of their initiative for the appointment of a special rapporteur on the right to education by the UN Commission on Human Rights in 1998.

[ 36 ]
An early classic of qualitative-interpretative political culture research is de Tocqueville’s ‘Democracy in America’. Anthropologically inspired studies include Lucian Pye’s study on Burma (Pye 1962) and Richard Fagan’s study on Cuba (Fagan 1969).

[ 37 ]
The minimum number of respondents should be 2000 (Wildenmann 1992: 67).

[ 38 ]
Pappi (1986: 289) sees a further disadvantage of poll research in the fact that informal sanction mechanisms are not recorded. It thus remains unknown how strongly norm-deviant behavior is informally sanctioned. This means that information about the strength of norms is missing.

[ 39 ]
The most prominent proponents of this view are Walter Lippmann and Gabriel Almond (see the discussion in Holsti 1996).

[ 40 ]
In this context, for example, Gross states that medium positions must be suggested so as not to drive interviewees into polarized positions (see Gross 1995: 17). Of course, this gives rise to another error, namely, that interviewees group around the medium position precisely because they have no opinion but do not want to admit it. The problem of instrumentation errors is difficult to solve in this way.

[ 41 ]
The degree of significance accorded to the legal codification of societal norms varies from society to society. The Federal Republic of Germany is regarded as one of the states in which societal norms are legally codified to a great extent (see Bulmer 1997: 67f.; Katzenstein 1993: 276; Johnson 1978).

[ 42 ]
Societal constructivism thus views parties as being primarily marked by having programs and members. From the perspective of utilitarian liberalism, by contrast, it would be appropriate to regard parties primarily as competitors, their programs having no intrinsic value but serving solely as instruments to maximize votes (for a distinction between the two types of parties see Klingemann/Volkens 1997: 519).

[ 43 ]
For an analysis of societal norms following unification, it will be interesting to see whether a regional party becomes established in the new federal states (Länder) because this would mean that there are societally shared expectations with regard to German policy specific to the new federal Länder, supported by a relative majority concentrated there. Such a regional party is represented in the Bundestag in the form of the PDS. Precisely because the PDS represents constituents that only entered the political stage as a result of unification, the PDS and its platforms should in any case be considered as an indicator of societal norms.

[ 44 ]
Empirical studies have found a remarkably high degree of correlation between election platforms and a government’s actual behavior (Klingemann/Hofferbert/Budge 1994, and for Germany, Hofferbert/Klingemann 1990). These studies also emphasize the disproportionately great influence of the F.D.P. in formulating German foreign policy. The great influence of the F.D.P. is attributed to the fact that, as the factor tipping the scales when coalitions are formed between the parties represented in the Bundestag, it is a particularly assertive actor. The logic of the foreign policy-making process implied by this finding corresponds, of course, to that of utilitarian liberalism (see Bienen/Freund/Rittberger 1999). As an indicator of societal norms, however, the F.D.P. platforms have the same status as those of other small parties, such as Bündnis ‘90/DIE GRÜNEN.

[ 45 ]
"[...] the national role conception framework proceeds from the thesis that an understanding of how decision makers perceive the world stage and their state's part in the play will improve our understanding of their actions and increase our potential to accurately pinpoint the future course they might choose" (Breuning 1995: 236f.). While the variable ‘national role' used by Breuning cannot be equated with ‘societal norms’, Breuning feels that "[d]ecisionmakers are both products and representatives of their society." (Breuning 1997: 99). By means of an analysis of national conceptions of roles, it may hence be possible to draw conclusions about societal norms.

[ 46 ]
The share of Dutch GNP devoted to development aid is far higher than in the case of the UK. Moreover, the share of total UK aid given as tied aid is far greater than in the case of the Netherlands (Breuning 1995: 252).

[ 47 ]
The danger of circular reasoning is increased if the government is supported by a grand coalition because, then, the speeches of members of the two major parties supporting the government would have a sufficiently high degree of commonality for constructivists to expect them to shape government policy. However, as a grand coalition was not formed during the period being studied here, this methodically problematic case can be ignored.