email icon Email this citation

CIAO DATE: 11/00

Reassessing the Logic of Anarchy: Rationality versus Reflexivity

Ewan Harrison

Department of Politics
University of Edinburgh

International Studies Association
41st Annual Convention
Los Angeles, CA
March 14-18, 2000



Abstract

Drawing on the insights of constructivism and recent attempts to develop a liberal model of international relations, this paper compares neorealism, institutionalism and liberalism in terms of their competing conceptualisations of the composition and dynamics of the international system. It argues that prominent institutionalists have mistakenly interpreted neorealism as a rationalist model of the international system in order to develop their own arguments countering its central propositions. Rather than relying on rational actor assumptions, neorealism instead adopts a reflexive logic focusing on processes of identity construction and socialisation arising from the generative consequences of international anarchy. Thus whilst neorealism and liberalism differ fundamentally in terms of their assessment of the nature and long term consequences of international anarchy, they may actually be closer to each other in terms of their conceptualisation of system composition and dynamics than neorealism and institutionalism. On the basis of this argument, the paper rejects a ‘linear’ understanding of the relationship between neorealism, institutionalism and liberalism in which institutionalism is seen as providing the point of synthesis on a spectrum than runs between neorealism and liberalism. Instead, it proposes a ‘triangular’ understanding of the relationship between these three models of the international system in which the debates between neorealists and institutionalists constitute its rationalist leg, and debates between neorealists and liberals constitute its reflexivist leg. Understood in these terms, liberalism’s focus on the generative consequences of international anarchy and the overall trajectory of identity construction and socialisation within the international system is likely to be more productive in the search for a general synthesis of competing theories of international relations than institutionalists’ attempts to refine rationalist models.




Drawing on the insights of constructivism and recent attempts to develop a liberal model of international relations, this paper compares neorealism, institutionalism and liberalism in terms of their competing conceptualisations of the composition and dynamics of the international system. It suggests that prominent institutionalists have mistakenly interpreted neorealism as a rationalist model of the international system in order to develop their own arguments countering its central propositions. Rather than relying on rational actor assumptions, neorealism instead adopts a reflexive logic focusing on processes of identity construction and socialisation arising from the generative consequences of international anarchy. Thus whilst neorealism and liberalism differ fundamentally in terms of their assessment of the nature and long term consequences of international anarchy, they may actually be closer to each other in terms of their conceptualisation of system composition and dynamics than neorealism and institutionalism. On the basis of this argument, the paper rejects a ‘linear’ understanding of the relationship between neorealism, institutionalism and liberalism in which institutionalism is seen as providing the point of synthesis on a spectrum that runs between neorealism and liberalism. Instead, it proposes a ‘triangular’ understanding of the relationship between these three models of the international system in which the debates between neorealists and institutionalists constitute its rationalist leg, and debates between neorealists and liberals constitute its reflexivist leg. Understood in these terms, liberalism’s focus on the generative consequences of international anarchy and the overall trajectory of identity construction and socialisation within the international system is likely to be more productive in the search for a general synthesis of competing theories of international relations than institutionalists’ attempts to refine rationalist models.

The paper examines debates between neorealists, institutionalists and liberals over the mechanisms through which the anarchic structure of the international system is constituted (reproduced/transformed). All three theories agree that international anarchy is constituted through the interaction of units. However, they identify different mechanisms through which units within the international system interact. One way in which units can interact is on the basis of rational behaviour. Units which behave rationally “possess consistent, ordered preferences, and ... calculate the costs of benefits of alternative courses of action in order to maximise their utility in view of these preferences” (Keohane, 1984:27). The notion that actors behave rationally in their interactions exogenises questions about the characteristics of units within the international system, since units’ conceptions of their ‘interests’ are given by assumption (Wendt, 1992:398). An alternative means through which units can interact is on the basis of reflexive behaviour. Actors which behave in a reflexive manner “not only monitor the continuous flow of their activities and expect others to do the same for their own; they also routinely monitor aspects, social and physical, of the contexts in which they move” (Giddens, 1984:5). Actors are therefore viewed as competent, skilled and knowledgeable about the complex environments in which they move, but they are not necessarily rational in the sense that they expect to maximise their expected utility from patterns of social interaction. For example, even what may appear from a rational-choice perspective to be ‘irrational’ social behaviour involves a high degree of reflexive attention and awareness by purposeful social agents. The idea that actors are capable of reflexive behaviour involves making statements about their basic identity as agents. To adopt Wendt’s definition, identity refers to “relatively stable, role specific understandings and expectations about self” from which actors define their portfolio of ‘interests’ (Wendt, 1992:397). That is to say, ‘interests’ depend on identities which are either actively or passively constructed by knowledgeable actors on the basis of intersubjective meanings and purposes acquired from their institutionally defined roles (Wendt, 1992:398).

The paper is organised into three sections examining the neorealist, institutionalist and liberal models respectively. Firstly, each theory’s assumptions about the nature of international anarchy are analysed to highlight differences in its ontological status across them. Secondly, the way each theory conceptualises the constitution of the international system is examined to identify whether a rationalist or a reflexivist logic is adopted. Finally, the theories are compared with respect to their predictions about both the sources of state ‘power’, and the levels and type of institutionalisation expected within the international system. A fourth section of the paper will consider the relationship which exists between the three models set out, the possibilities for a general synthesis between them, and the implications of these considerations for substantive research programmes within the discipline.



The Neorealist Model

The key variable identified by the neorealist model is the configuration of capabilities between states. This conclusion is drawn from the specification of Waltz’s theory of international politics. Waltz’s model begins by sharply distinguishing between the structure of domestic and international political systems. For Waltz, domestic political systems are centralised and hierarchical whilst the structure of the international system is anarchic due to the absence of a central organising authority (Waltz, 1979:88). Waltz argues that, because of its anarchic deep structure, the international system becomes a self help one because “(w)hether units live, prosper or die depends on their own efforts” (Waltz, 1979:91). Under these circumstances “(f)ear of unwanted consequences therefore stimulates states to behave in ways that tend towards the creation of balances of power” (Waltz, 1979:118). The balance of power may therefore be seen as a condition of structural “equilibrium” in the distribution of capabilities within the international system (Waltz, 1979:121). It emerges in an unintended way as a “constraining and dispensing force” on the behaviour of units (Waltz, 1979:69).

As Dessler notes, Waltz’s theory relies on a positional model of structure which focuses on the role of unintended consequences of interaction arising from a unit’s relative position within a set of relationships (Dessler, 1989:448-451). Within such a model, “the units precede the system, and through action generates structure. ... (Thus whilst) in Waltz’s causal explanatory scheme, structure is seen as constraining and disposing state action .... ontologically speaking, it is the interaction of units that creates the structure” (Dessler, 1989:449). Wendt similarly rejects Ashley’s argument that Waltz is an unreformed structuralist, arguing instead that Waltz adopts an individualist definition of structure (Wendt, 1987:341; Ashley, 1986:238-242). Waltz claims that neorealist theory emphasises that causation runs from both structures to states and from states to structures (see figure 3.1 in Waltz, 1979:40). However, as Dessler notes, properly speaking the creation of system structure through unit interaction is postulated ontologically, whilst the constraint imposed on interaction is explained theoretically by neorealism (Dessler, 1989:449 fn 33).

The implication of Waltz’s reliance upon a positional model of structure is that the system he envisages takes on a reproductive rather than a transformative character. For Waltz, unintended consequences of interaction cause patterns of behaviour to reoccur even against the wishes of the system’s members. In this view, structure is defined as an arrangement of actors which acts as an external constraint on state action (Dessler, 1989:449). It is due to his adoption of a positional model that Waltz’s theory privileges the causal role played by system structure in the constitution of the international system (Wendt, 1987:340-344). Waltz does allow scope for the transformation of the international system if its ordering principle shifts from anarchy to hierarchy (Waltz, 1986:342). However, this must be seen as unlikely because of the anti-hegemonial character of the balance of power mechanism (Waltz, 1986b:343 and 340-341). As Ruggie has identified, in Waltz’s model “continuity (in the form of the persistence of anarchy) ... is the product of premise even before it is hypothesised as an outcome” (Ruggie, 1986:152). Waltz’s is a problem solving theory that “takes the world as it is, with the prevailing social and power relationships within it ... as the given framework for action”, rather than a critical theory capable of directing attention towards possible sources of historical transformation (Cox, 1986:208).

Despite the reproductive role played by structure in neorealism, there is remains a strong dynamic quality to Waltz’s model which he is careful to emphasise (Waltz, 1986b:341-344). Due to his stress on the role of unintended consequences, Waltz argues that “(i)n systems theory, structure is a generative notion ” (Waltz, 1979:72, emphasis added). For Waltz, “structure acts as a selector” (Waltz, 1979:73), and over time will eliminate units which fail to respond to the imperatives of the system. Waltz’s stress on the generative qualities of international anarchy leads him to emphasise the “pervasive” importance of competition and socialisation within the international system (Waltz, 1979:74). For Waltz, socialisation is a process of emulation, and competition a process of elimination (Waltz, 1979:74-77). Waltz sees these forces at work in the international system through its structural effects. States face strong structural incentives to emulate the effects of the most successful and become ‘socialised’ to the international system. If they fail to respond to these incentives, competition will, over time, tend to eliminate them as actors. This has both internal and external consequences for the organisation of states. Internally, they will adopt equivalent organisations. Externally, states will adopt structurally similar repertoires of behaviour. Competition and socialisation within the international system encourage the homogenisation of units such that they are functionally alike, and promote its convergence around a balance of power equilibrium.

Waltz’s suggestion that the structure of international anarchy is generative has important implications for the way he conceptualises the role played by the state in the constitution of the international system. Waltz is conventionally interpreted as assuming that states behave rationally in response to incentives provided by system structure. Most notably, Keohane, a prominent institutionalist, adopts this interpretation of neorealism. He holds that “(f)or Waltz ... (t)he link between system structure and actor behaviour is forged by the rationality assumption, which enables the theorist to predict that leaders will respond to incentives and constraints imposed by their environments. ... Otherwise, state behaviour might have to be accounted for by variations in the calculating ability of states” (Keohane, 1986:167; see also Keohane, 1984:83-84). However, in making this claim Keohane knowingly counters Waltz’s own assertion that his theory “requires no assumptions of rationality or of constancy of will on the part of all of the actors. The theory simply says that if some do relatively well, others will emulate them or fall by the wayside” (Waltz, 1979:118; see also Keohane, 1986b:201 fn 15). Waltz does hold that his theory requires the assumption that at least some states seek to ensure their survival (Waltz, 1979:91 and 118). However, he does not assume that they pursue this strategy rationally. Instead, he relies on the generative consequences of the anarchic structure. In responding to his critics, Waltz stresses “the importance I accord, and Keohane denies, to the process of selection that takes place in competitive systems”, and asserts that “(w)e should keep the notion of ‘selection’ in a position of central importance” (Waltz, 1986:330 and 331). Waltz does make the qualification that he does not differ with Keohane over the issue of rationality “except semantically” (Waltz, 1986:330). However, in the light of more recent developments in international relations theory, it is possible that Waltz’s judgement about the semantic nature of his differences with Keohane on this issue may have been somewhat premature.

In particular, Buzan et al have elaborated further on the role played by the state in neorealist theory. They argue that the state in neorealist theory is best characterised as a reflexive rather than a rational actor. They propose that implicit in Waltz’s model is an incipient ‘top down’ theory of the state (Buzan, 1993:117-119). Whilst recognising that the international system is individualist in origin in neorealist theory, they identify that “(o)nce states are coacting ... Waltz is quite clear that the structure of the international system does have an important bearing on the development of the state. This is because ... anarchy generates a competitive environment” (Buzan et al , 1993:117). For this reason Waltz observes the importance of the imitation of successful practices of others within the system, a process which “plays an integral part in the process that maintains the balance of power and ... sustains the anarchic system” (Buzan et al , 1993:117-118; see also Waltz, 1979:127-128). Thus “agents of the state, like Janus, are required to look in two directions simultaneously. They confront two sets of structures: one internal and the other external” (Buzan et al , 1993:120). Buzan et al use this observation to argue that Waltz’s view of the unit-structure relationship strongly echoes a depiction of social agents as “knowledgeable and reflexive, having not only a sophisticated view of the world and how it is structured, but also the ability to monitor their actions in the light of this knowledge” (Buzan et al , 1993:107).

There is an important distinction between an understanding of the state as a reflexive actor, and an understanding of it as a rational actor. Buzan et al use the example of the behaviour of motorists at traffic lights to illustrate this point. They note that there is a complex set of social institutions that shape the behaviour of motorists which “are observed not because of the sanctions imposed on deviants, but because of the reflexive ability to see what would happen in situations of strategic interaction if the structures did not exist. ... So it follows that a structure constitutes a structure only because of the behaviour of the agent, which in turn is intimately bound up with knowledge of the structure” (Buzan et al , 1993:107). In this way, patterns of behaviour may persist not because of rational calculations of interest, but because of the essential recursiveness of social life, i.e. the way in which social activity is ‘carried on’, at least at the level of practical consciousness, by reflexively aware actors. Dessler makes the same point when he notes that whilst neorealism suggests that the distribution of material capabilities amongst actors motivates their behaviour, the anarchic structure implicitly generates rule following activity on the behalf of units (Dessler, 1989:459-461). Thus “(r)ules, which give shape and meaning to rationality and thereby make survival possible, are a necessary (if theoretically suppressed) component of structure in the neorealist model” (Dessler, 1989:461). Distinguishing empirically between a social structure sustained by rational action and one sustained through reflexive self-regulation on behalf of actors is difficult because the outcomes in the form of patterned behaviour over time is identical. However, these theoretical considerations affect the way central features of the international system envisaged by neorealism are conceptualised.

For example, one way of viewing the balance of power is that states rationally respond to the incentives and constraints generated by their relative position within the international system at any given point in time. However, Jervis proposes that that instead the balance of power may be viewed in terms of negative feedback produced through the international system envisaged by neorealism (Jervis, 1997:131-143). In this conception, “the balance is systemic in that the norms have been internalised through socialisation as the actors watch and interact with their peers” (Jervis, 1997:136). The most compelling examples of reflexive self-regulation in operation in the international system have occurred when major states have behaved ‘irrationally’ in response to the incentives and constraints generated they face. Specifically, important actors have been eliminated as significant units within the international system because of their failure to imitate the practices of potential military challengers. An obvious case is the Anglo-Dutch conflict of the Seventeenth century. The failure of the Dutch to prepare for war led to their defeat and elimination as a major power. On a rationalist interpretation, the Dutch behaviour is difficult to account for using the neorealist model. Although the ‘rational’ option of building up its military capabilities was open to Holland, it failed to take this course of action. However, understood in terms of the effects of negative feedback and the international system’s capacity for reflexive self-regulation through selection and socialisation, the outcome of the Anglo-Dutch conflict is easily accounted for in neorealist terms. Moreover, as Waltz is well aware, even a very small number of cases such as the one discussed above are enough to provide an important example for other states in the system. For this reason they are extremely effective in ensuring the reproduction of a competitive environment over time. Thus, the distinction between rational action and reflexive self-regulation on behalf of actors is of crucial importance for the way in which the composition and dynamics of the international system are envisaged.

In particular, the salient point is that neorealism does indeed have a theory of the state, and specifically a theory of state identity construction and socialisation within the international system. It postulates that states are reflexively aware agents capable of internalising appropriate patterns of behaviour and deploying this understanding knowledgeably in contexts of action. To appropriate Wendt’s language, within the neorealist model, the state is able “1) to have a theoretical understanding (however inaccurate) of its activities, in the sense that it could supply reasons for its behaviour; 2) to reflexively monitor ... its behaviour; and 3) to make decisions” (Wendt, 1987, p.359). Drawing on Keohane’s interpretation of Waltz, Wendt himself explicitly concedes the view that neorealism presents a rationalist model of the international system (Wendt, 1992:391, fn 2). However, more recently Wendt et al have acknowledged that “Waltz was implicitly talking about identity when he acknowledged that anarchic structures tend to produce like units” (Wendt et al , 1996:34). Wendt et al do make the valid qualification that whilst “Waltz allows for what he calls ‘socialisation’ and ‘imitation’ processes ... he envisions the shaping of the behaviour of pregiven actors” (Wendt et al , 1996:41). However, they are merely drawing attention to the way in which neorealism adopts a positional model of structure in which states are primitive. The essential point is that in the neorealist model states should be understood as being reflexive rather than rational actors remains, and takes on crucial significance when comparing neorealism to institutionalist and liberal models of the international system.

Tensions in Waltz’s account of the unit-structure relationship become most apparent when the status of neorealism’s predictions about the behaviour of states within the international system are scrutinised. Superficially, it seems straightforward to arrive at general statements about the character of the international system that may be associated with Waltz’s model. Power is “defined in terms of the distribution of capabilities” within the international system (Waltz, 1979:192). It is understood in relative rather than absolute terms, relating back to neorealism’s positional model of structure. The predominant source of power for states will be the relative gains in terms of actual or potential capabilities to be made in their interactions with other units (Grieco, 1993). In terms of neorealism’s predictions about the levels and type of institutionalisation that will take place within the international system, Waltz stresses that the decentralised character of the international system does not mean that it is not ordered in important ways. Instead, “patterns emerge and endure without anyone arranging the parts to form patterns or striving to maintain them” (Waltz, 1979:77). For Waltz, therefore, the international system becomes ‘institutionalised’ around the balance of power. Levels of institutionalisation occurring above and beyond the balance of power will be limited. The existence of effective international organisations will ultimately be contingent on shifting distributions of power between states (Waltz, 1979:88). For this reason international politics is merely “flecked with particles of government” (Waltz, 1979:114).

However in the light of recent debates over the status of neorealism as a theory of foreign policy, it is becoming increasingly clear that neorealism is fundamentally ambiguous about the extent to which international structure will determine international behaviour and outcomes. Waltz is insistent that neorealism represents a theory of international politics which operates at a structural level of analysis, rather than as a theory of foreign policy which offers predictions about the determinate behaviour of units (Waltz, 1996). Yet, as Elman cautions, “refraining from making such predictions would diminish neorealism’s usefulness considerably” (Elman, 1996:60). Waltz asserts that international political theory can explain states’ behaviour only when external pressures come to dominate the internal disposition of states (Waltz, 1996:57). Thus the key question becomes how to explain the variation in the influence of system structure upon outcomes over time. There are therefore fundamental ambiguities about neorealism’s predictions that may be traced to its inability to specify clearly the relative influence of unit and structure over the foreign policy behaviour over time.

In summary, three aspects of the neorealist model have been highlighted. Firstly, it develops a positional model of structure, and thereby takes on a reproductive rather than a transformative logic. Secondly, neorealism is not, as is widely supposed, committed to the assumption that states are rational actors. Instead, it relies on the generative consequences of international anarchy in combination with the idea that units are reflexive agents which behave knowledgeably in relation to their environments. Thus neorealism develops a theory of state identity construction and socialisation within the international system, albeit one which operates within a pre-given set of structural parameters. Thirdly, neorealism predicts that power in the international system resides in relative gains from inter-state interaction. Under these conditions, levels of institutionalisation in the international system will be low, and the behaviour of states will become ‘institutionalised’ around the balance of power. However, due to the tensions between neorealism’s conceptualisation of unit-structure relations introduced by Waltz’s acceptance of the notion that international anarchy is not fully generative, the neorealist model is fundamentally ambiguous about the status of its predictions about international behaviour and outcomes.



The Institutionalist Model

The key variable identified by the institutionalist model is the configuration of information and institutions within the international system. By focusing on the systemic impact of information and institutions, institutionalists seek to demonstrate that international co-operation is possible from the same set of core assumptions about the nature of the international system as neorealism. Institutionalist theory therefore purports to accept neorealist assertions about the nature of international anarchy, and to share in common with neorealism the assumption that states behave rationally. The essence of the institutionalist argument is therefore that neorealism’s conclusions do not follow from its premises.

Institutionalists purport to accept neorealist claims about the impact of anarchy (Keohane, 1984:26). However, they also presuppose the existence of complex patterns of interdependence which create extensive common interests amongst advanced industrial states (Keohane, 1984:6; see also Keohane and Nye, 1989:23-37). According to institutionalists, the problem of cooperation between advanced industrial states can be modelled by game theory in terms of a Prisoner’s Dilemma. Within such a game, two participants could benefit from cooperation but fail to do so due to the patterns of incentives generated by the structure of strategic interaction that exists between them (Keohane, 1984:68). In making this claim, institutionalists “concede to neorealists the causal powers of anarchic structure” (Wendt, 1992:392). Institutionalism concentrates on system process, understood in terms of allocative and bargaining behaviour within a given power structure (Keohane and Nye, 1989:21). However, because insitiutionalists “lack a ... theory of how ... (systemic transformation) occurs ... (they) must privilege realist insights about structure” while advancing their own claims (Wendt, 1992:393).

Nevertheless, having asserted that Prisoner’s Dilemma is an appropriate model for theorising relations between advanced industrial states, institutionalists go on to build their critique of the neorealist model of international relations. As has been noted, institutionalists adopt a particular interpretation of neorealism in which the state is regarded as a rational actor. Drawing on theories of rational choice, Keohane suggests that whether or not defection is the dominant strategy within a Prisoner’s Dilemma depends on the number of times the game is played. If the game is played once or only a small number of times, defection will be the dominant strategy for rational egoists. However, if the game is played iteratively by the same players, cooperation between the participants can rationally emerge if the participants value future rewards (Keohane, 1984:75). This is because the iterative playing of the game generates a shadow of the future (Axelrod and Keohane, 1993:91). Thus within an iterated Prisoner’s Dilemma, mutual cooperation will develop around a strategy of ‘tit-for-tat’ between egoists (Keohane, 1984:76). Cooperation between states becomes institutionalised around norms of specific reciprocity , in which “specified partners exchange items of equivalent value in a strictly delimited sequence” (Keohane, 1989b:134). Having made these claims, Keohane goes on to suggest that iterated rather than single play Prisoner’s Dilemma is the more appropriate model for understanding the routine conduct of diplomacy between advanced industrial states. This is because negotiations between these states on issues of mutual concern “take place continuously and are expected to continue indefinitely into the future” (Keohane, 1984:76).

Keohane also holds that international regimes can play a functional role in facilitating cooperation among states which act as rational egoists. Although regimes do not substitute for reciprocity, they can play an important role in encouraging states to engage in this behaviour (Axelrod and Keohane, 1993:110). They can be functional for states seeking to cooperate for mutual gains for two principle reasons. Firstly they can reduce the transactions costs of coming to an agreement, for example, by making it cheaper for governments to meet to negotiate (Keohane, 1984:90). Secondly, regimes can help reduce uncertainties by providing information which shape states’ expectations of future behaviour. For example regimes can reduce asymmetries in information, by helping states insure compliance and raising the costs of deception and irresponsibility (Keohane, 1984:97). Under anarchy regimes will be difficult to create since it is difficult for states to establish legal liability through regimes (Keohane, 1984:89). However, once brought into existence (for instance, by a hegemonic power) they will be functional for states and will therefore tend to endure (Keohane, 1984:100-101). Keohane goes on to explore the implications of relaxing the assumption of rationality by examining the implications of assuming ‘bounded rationality’ (Keohane, 1984:110-120). He also discusses the possibility of limited redefinitions of actor interests away from pure egoism (Keohane, 1984:120-130). However, he stresses that the move away from the assumption of rational egoism takes his analysis beyond a functional theory of regimes (Keohane, 1984:132). Thus Keohane’s approach never abandons the assumption that states behave as rational actors, and indeed requires this assumption in order that it can be seen as progressively expanding neorealism’s explanatory power.

Institutionalism’s rationality assumption has important theoretical implications for how it conceptualises the composition and dynamics of the international system. Specifically, institutionalism does not, and indeed cannot, provide a theory of state identity formation within the international system . Keohane explicitly acknowledges that “rationalist theories contain no endogenous dynamic (because) ... preferences are assumed to be fixed” (Keohane, 1989c:171). This commitment to a rational choice framework makes institutionalist theory distinct from both neorealist and liberal models of the international system. As has been explained, neorealism does make an attempt to provide a theory of identity construction and socialisation within the international system and as such adopts a view of the state as a reflexive rather than a rational actor. Similarly, whilst institutionalism shares with liberalism a concern with explaining the formation of international institutions and regimes, unlike a liberal model of the international system, it allocates no role in this process for the importance of cognitive human characteristics such as the capability for autonomous self-reflection and choice, intentionality and the possibility of social learning for world politics (Keohane, 1989c:168-170). Instead, states simply respond mechanistically to functional incentives to cooperate. 1 It is on the basis of these tensions in institutionalists’ analysis of international institutions that Moravcsik argues that functional regime theory, referred to by Keohane himself as neoliberal institutionalism or simply as neoliberalism is not properly a liberal theory of international relations (Moravcsik; 1997:536-537). Wendt makes the same point when he distinguishes institutionalism’s “weak liberalism” which brackets processes of interest and identity formation, and “strong liberalism” which encompasses insights about learning and cognition (Wendt, 1992:393-394). Indeed, Keohane explicitly concedes the validity of such arguments when he notes that “(I)n view of the links between insititutionalist thinking and both liberalism and realism, I now prefer the simple appellation ‘institutionalist’ to ‘neoliberal institutionalist’ or ‘liberal institutionalist’” (Keohane, 1993b:298 fn 3).

Nevertheless, institutionalism does make relatively clear predictions about both the sources of power for states, and the level and type of institutionalisation expected within the international system. The predominant source of power for states in the international system, at least where relations are conducted in the context of complex interdependence, will be the absolute gains to be made through states’ functional incentives to co-operate (Baldwin, 1993:5-6). Keohane and others have accepted that pursuit of absolute gains is to some extent conditional, and there are debates about the precise nature of these conditions (Keohane, 1993:278-283; see also Powell, 1993). However, Keohane is clear that institutionalism’s focus on absolute gains constitutes a key prediction that may be tested against patterns of behaviour in the emerging international system (Keohane, 1993:284-291). Institutionalisation is likely to take place between advanced industrialised states only, where there are strong potential incentives to cooperate. Where patterns of interdependence between states are poorly developed and/or highly asymmetrical, potential incentives to cooperate will be weaker and regimes will tend not to form in the absence of a hegemon. The type of institutionalisation expected will be that which is organised around norms of specific reciprocity, reflecting states rational incentives to cooperate to receive mutual benefits on a tit-for-tat basis. Finally, the level of institutionalisation expected by institutionalists under conditions of complex interdependence is best described as moderate or between advanced industrial states only. This is in contrast to both the comparatively low level of institutionalisation predicted by neorealism and the comparatively high level of institutionalisation expected by liberalism in a late stage of development of the international system (see below).

In summary, three aspects of the insititutionalist model of the international system have been highlighted. Firstly, like neorealism insititutionalism takes on an essentially reproductive rather than a transformative logic. Secondly, unlike neorealism, institutionalism is committed to the assumption that states behave as rational actors. As such it lacks a theory of identity construction and socialisation within the international system. Finally, institutionalists predict that the predominant source of power in the international system will be the absolute gains they can make from international cooperation, that the level of institutionalisation in the international system will be moderate (between advanced industrial states only), and that the type of institutionalisation that will occur will be organised around norms of specific reciprocity.



The Liberal Model

The key variable identified in the liberal model is variation in the configuration of state preferences (Moravcsik, 1997:513). Preferences are defined as fundamental social purposes which provide the basis for the strategic calculations made by governments (Moravcsik, 1997:513). They may be understood as reflecting an underlying political identity derived from sets of domestic political values and ideas, commercial/economic interests and political institutions (Morvacsik, 1997:524-533). However, liberalism’s stress on the importance of states’ preferences in determining international outcomes should not be understood as undermining its ability to account for the impact of international anarchy on state behaviour. Indeed, one of the distinctive features of liberalism is that it develops a theory of state identity construction based on a very particular view of the nature and long term effects of the structure of international anarchy.

Liberal theory builds upon the insights of constructivism, a point which has been acknowledged by advocates of both these persuasions (Wendt, 1992:394; Moravcsik, 1997:539-540). Specifically, liberalism adopts a transformational model of structure (Dessler 1989:452-454). Within a transformational model, the dynamics of the international system can push it towards fundamental change rather than simple reproduction. This requires a very particular conceptualisation of the mechanisms through which the international system is constituted. Units and structures “are understood as ‘codetermined’ or ‘mutually constituted’ entities” which are of equal and irreducible ontological status (Wendt, 1987:339). Structures exist “only through the medium of the agents and practices they constitute” (Wendt, 1987:360). Thus structure is conceptualised as enabling as well as constraining and as internal to actors rather than external to them because “state action is possible only if there exist the instruments through which that action can in fact be carried out” (Dessler, 1989:453). In the transformational model, structure is more than an arrangement of actors. Rather, anarchy is conceptualised as a set of rules and resources which are recursively drawn upon by units in an ongoing process of development and change within the international system. In this context, rules refer to frameworks of meaning and resources to physical attributes through which units communicate with one another (Dessler, 1989:452-454). Thus the ontological status of anarchy in liberal theory differs fundamentally from neorealist and institutionalist models.

The implications of liberalism’s adoption of a transformational model of structure are two fold. Firstly, liberalism is not committed to a narrowly state-centric view of the international system. If structure is viewed purely as an arrangement of actors, the historical context within which units operate is not relevant to its characteristics. By contrast, if structure is viewed as a set of rules and resources drawn upon recursively by units in an ongoing process of historical change, then the social, economic and political characteristics of the units become important for the overall trajectory of development produced in and through the structural effects generated by the system as a whole. As Moravcsik notes, in contrast to neorealism and institutionalism, the view of system structure adopted by a liberal model of the international system “forges a direct causal link between economic, political and social change and state behaviour in world politics” (Moravcsik, 1997:535). In particular, liberalism stresses the significance of modernity in terms of the development of world society for overall patterns of change within the international system. Thus “(o)ver the modern period the principles of international order have been decreasingly linked to dynastic legitimacy and increasingly tied to factors directly drawn from ... liberal theory” in the form of sets of domestic political values and ideas, commercial/economic interests and political institutions compatible with liberal state preferences (Moravscik, 1997:535). Liberalism’s conceptualisation of structure in terms of rules and resources means that it adopts an inherently ‘critical’ approach to international anarchy.

Secondly, liberalism adopts a conception of anarchy which is, in principle, always open to transformation by its constitutive units. As Wendt remarks, “social structures are only instantiated by the practices of agents. The deep structure of the states system ... only exists in virtue of the recognition of certain rules and the performance of certain practices by states” (Wendt, 1987:359). However, a liberal theory of international relations also goes beyond the constructivist view that “anarchy is what states make of it” (Wendt, 1992). As Huntley identifies, it also specifies a set of determinate historical and sociological dynamics which encourage the emergence and perpetuation of these tendencies in the form of the long term effects of competition and socialisation within the international system (Huntley, 1996:57-62) 2 . Liberalism’s conceptualisation of the mechanisms through which the international system is constituted can be represented as set out in the matrix in Figure 1 below. The top left box of the matrix represents a system which is constituted along the reproductive lines suggested by neorealism in which units have limited discretionary power because regardless of their intentions outcomes will remain confined to relatively limited ranges. The bottom left box of the matrix represents an international system in which the structure of anarchy has generated unintended consequences which have encouraged system transformation through competition and socialisation. For example, Huntley discusses the ways in which liberal states will tend to outperform non-liberal states in long term competition, and similarly identifies liberal states’ capacity to be trusted as casting a long ‘shadow of the future’, thereby unwittingly socialising them to the liberal peace (Huntley, 1996:57-58) 3 .

Figure 1: System Composition and Dynamics

  Unintended
consequences
Intended
consequences
System
Reproduction
Neorealism XXXXXX

(Fundamentalism)
System
Transformation
Liberalism 1

Unintended consequences arising from anarchic structure
Liberalism 2

Arising from conscious agency and choice - ‘cultural change



INTERACTION
(Reflexive component of system organisation)

Thus within the liberal model, anarchy is “‘both a constraining and dispensing force’ and a force for transformation and progress” (Huntley, 1996:61). Liberalism provides a framework able to provide an explanatory account of systemic transformation which complements constructivist insights concerning the ontological status of anarchy in international relations theory (Moravcsik, 1997:540). It therefore both draws upon and moves beyond the claim of constructivists that “there is no logic of anarchy apart from the practices that create and instantiate one structure of identities and interests rather than another” (Wendt, 1992:395). In this sense, a liberal theory of international relations has affinities with neorealism in as much as anarchy is viewed as having important generative consequences. Yet, unlike the neorealist model, liberalism “is able to offer a conception of structure that accounts for its own transformation” (Huntley, 1996:61). Within the liberal model there is a definite ‘logic of anarchy’, but rather than this logic being reproductive as neorealists suggest, it is fundamentally transformative.

It is the bottom right hand corner of the matrix which is the most distinctive element of a liberal model, stressing the importance of ‘ cultural change ’ for the operation of competition and socialisation (see Wendt et al , 1996). Unlike the unintended effects of anarchy, cultural change involves conscious human agency and intentional choice, and as such is crucially important at the margins. Liberalism therefore understands competition and socialisation as operating through the dynamic interplay of state agency and anarchic structure. As liberal culture expands globally, states’ conscious and intentional choices reinforce the powerful long term unintended effects of competition and socialisation. This process will continue until a genuinely stable equilibrium has been reached as a peaceful federation of liberal democratic states on a global scale is approximated. The role played by cultural change within the liberal model means that although it adopts the same modalities of system reproduction suggested by neorealism - competition, socialisation and equilibrium - it nevertheless understands these central characteristics in very different terms. It draws out the full implications of liberalism’s reliance on a transformational model of structure for its understanding of system composition and dynamics.

The place that cultural change has in liberalism’s conceptualisation of competition and socialisation draws attention to important differences and similarities it has with other systemic theories. Its most distinctive feature is that it adopts a more differentiated view of the composition of the international system than neorealism and institutionalism. Whilst neorealism focuses on structure and institutionalism on process, liberalism focuses on unit-structure interaction within the international system. It suggests that an international system is not necessarily constituted within limits tightly set by its structure. Instead, it is the reflexive component of system organisation which is of greater overall historical significance for the trajectory of the international system (see Figure 1). Nevertheless, important parallels between the neorealist and liberal theories remain which set them apart from institutionalism. Firstly, unlike insititutionalism but like neorealism, liberalism offers a theory of state identity construction and socialisation which is directly related to its conception of the generative effects of anarchy. As Huntley identifies, the liberal model parallels neorealism in as much as it envisages powerful “homogenising” influences in which “competition and socialisation ... (work) dynamically, promoting convergence towards the rule of law” both within and between states (Huntley, 1996:57 and 59). Secondly, because of their shared reliance upon socialisation dynamics, liberalism shares with neorealism a conception of the state as a reflexive rather than a rational actor. Units’ abilities to be ‘socialised’ presupposes their ability to internalise appropriate patterns of behaviour and deploy this understanding knowledgeably in contexts of action. Unlike neorealism, state identity is not shaped within pregiven parameters because of the transformative model of structure adopted by liberal theory. Within liberal theory, the ability of actors to engage reflexively with their environment is critical to driving cultural change and systemic transformation. However, differences between neorealism and liberalism in terms of their conceptualisation of the fundamental nature and long term consequences of anarchy should not detract from their shared view of the state as a knowledgeable, reflexive agent.

Moravcsik makes the argument that a liberal model relies on the assumption that states are comprised of aggregates of individuals and groups which are “on the average rational” (Moravcsik, 1997:516 and 517 fn 7). This leads him to the view that “collective state behaviour should be analysed as a two-stage process ... . States first define preferences - a stage explained by liberal theories of state-society relations. Then they debate, bargain, or fight to particular agreements - a second stage explained by realist and institutionalist (as well as liberal) theories of strategic interaction” (Moravcsik, 1997:544). As has been highlighted by Wendt, the rationalist two-step provides a highly under-developed account of cultural selection (imitation and social learning) in the operation of competition and socialisation (Wendt, 1999:324-336). It underestimates the importance of cultural change, the bottom right hand cell of the matrix in Figure 1, for driving system transformation. It precludes the possibility that, under certain conditions, liberal peace may develop socialisation effects that are worth more than the sum of its constitutive parts, and which extend beyond the extant body of liberal states. As Wendt points out, this is because the scope conditions of rationalist models are limited to situations where “it is plausible to expect that interests and identities will not change over the course of an interaction” (Wendt, 1999:367). Where the socialisation effects generated by the liberal peace are significant enough to encourage states at the margins of the system to engage in cultural change, these scope conditions would not hold. To appropriate Wendt language, there is “reason to think that states will alter their identities over the course of an interaction” (Wendt, 1999:367). Yet this idea has been central to the sociological strand of liberal thinking on international relations from at least Kant onwards (Huntley, 1996). Liberalism must therefore draw upon the constructivist insight that the state is a reflexive rather than a rational actor to flesh out its model of socialisation within the international system.

Cultural change and the socialisation of states to the liberal peace becomes particularly acute with maturation of the international system and the emergence of a relatively stable core of liberal states at a global level. As Huntley identifies, liberalism’s distinctive conceptualisation of competition and socialisation allows it to identify the ways in which a dynamic expansion of a ‘Pacific Federation’ of liberal states is encouraged through recursivity effects within the international system. The forces of competition and socialisation generate a strong, “dialectical ‘causal loop’” of self-reinforcing positive feedback (Huntley, 1996:59). The socialisation effects generated by the liberal core encourage states at the margins of change to adopt shifts in their underlying identity. This in turn feeds back into the properties of the system by further strengthening the liberal core. This stimulates further cultural change and shifts in state identity as positive feedback gathers historical momentum within the system. The system as a whole will achieve reflexivity, understood as a capacity for critical self reflection that helps actors overcome a false sense of determinism (Wendt, 1999:375). Reflexivity as a property of the international system must be clearly distinguished from the reflexive monitoring of action on the behalf of states within it. The reflexive monitoring of action is a generic, transhistorical feature of the state as an actor within the liberal model. It implies the general ability of states to become socialised to patterns of behaviour, and deploy this understanding knowledgeably in contexts of action. By contrast, reflexivity is a very particular property of the international system during its late or mature stages of development, and is achieved only when the long term transformative effects of competition and socialisation have become firmly established.

An international system which has achieved reflexivity will be characterised by two distinctive features. Firstly, units within the system acquire high levels of autonomy of action . That is to say, their foreign policy preferences will become critically important to the determination of behaviour and outcomes at a systemic level. It is this which lies behind states’ increased abilities to engage in cultural change and learning through a process of critical self reflection within the international system. In and of itself, however, this does not necessarily ensure that states will engage in cultural change or adopt shifts in their identity compatible with socialisation to peaceful liberal norms. The very same capacity for autonomous action which allows states to engage in cultural change and social learning also allows them the latitude not to engage in this behaviour. Nevertheless, an international system characterised by high levels of reflexivity is equally distinctive in terms of a second characteristic. In particular, the structure of interaction it embodies is overwhelmingly dominated by liberal states. As Moravcsik identifies, within the liberal model it is the overall configuration of state preferences in the international system which is the key variable in explaining behaviour and outcomes (Moravcsik, 1997:523). When reflexivity within the international system has been achieved, the configuration of preferences will be overwhelmingly dominated by the liberal core. This is because of the prior homogenisation produced through the long terms effects of anarchy envisaged by liberal theory. The dominance of the international system by the liberal core alters the overall trajectory of socialisation within the international system. It places pressure on states at the margins of cultural change within the international system to adopt shifts in their identity compatible with the norms of the liberal peace. Through recursivity effects, this in turn increases levels of systemic reflexivity, thereby further heightening states’ capacity for autonomous action within it and the pace of cultural change of behalf of units.

The most distinctive features of an international system which has achieved reflexivity may be revealed through the identification of the causes of violent behaviour within it. In a system in which levels of reflexivity are high, the top right hand corner of the matrix in Figure 1 above becomes important. It is possible to identify such behaviour as being fundamentalist in origin. Fundamentalist behaviour exists where a state peristently articulates a non-liberal identity, even when levels of reflexivity in the international system are high. It pushes the international system away from transformation towards peaceful inter-state relations and back towards simple reproduction. Moreover, this behaviour persists even though the state concerned may be severely punished by the dominant liberal core, and may even risk elimination as a serious actor within the system. Under these conditions, violence is directly motivated by states intentions and runs against the overall trajectory of socialisation within the international system. This contrasts markedly with the causes of violence in the neorealist model, which are unintended and in alignment with then socialisation effects produced by system. Wendt discusses the possibility that “(e)ven if the pressure for states to become ‘friends’ is strong ... egoistic identities commitments might not give way” (Wendt, 1999:340). 4 However, the possibility of fundamentalist behaviour implies something more than simply a failure to engage in cultural change and social learning when the system has achieved reflexivity. It suggests that the very autonomy of action acquired by actors within such systems may itself be the direct cause of a violent counter-reaction by actors unable to internalise significant cultural change.

Related to its conception of system composition and dynamics, liberalism develops an associated understanding of the nature of state power within the international system (Moravcsik, 1997:523). Where levels of reflexivity are high, the autonomy of action acquired by states means that information about their underlying strategic intentions is the critical factor in the determination of behaviour and outcomes. Under these circumstances, being transparent about their foreign policy preferences becomes a highly effective way for states to influence the overall trajectory of international change. Clearly conveying preferences reflecting a benign set of strategic intentions to other states allows them to establish their relations with others on the basis of trust. Trust may be identified as a form of institutionalisation, reflecting norms of diffuse reciprocity . In contrast to specific reciprocity, diffuse reciprocity depends upon meeting obligations by “conforming to generally accepted standards of behaviour” (Keohane, 1989b:134). Neorealism suggests, and institutionalism accepts, that anarchy inherently breeds mistrust between states, regardless of their intentions. By contrast, liberalism suggests that when levels of reflexivity are high, trust becomes the predominant source of power for states within the international system.

Here, liberalism may again draw fruitfully upon constructivist insights. Sustaining the trust of others allows states to establish a basis for common action with other states founded upon “feelings of solidarity, community, and loyalty and thus for collective definitions of interest” (Wendt, 1994:386, emphasis added). Actors contribute their share or behave well towards others “not because of ensuing rewards form specific actors, but in the interest of continuing satisfactory overall results for the group of which one is a part, as a whole” (Keohane, 1989b:146). Such behaviour “need not imply altruism ... (but) can consist of standards of behaviour which are widely regarded as legitimate; they do not necessarily embody ethical principles that override self-interest” (Keohane, 1989b:147). Thus acting on the basis of interests derived from a collective identity “does not mean that actors are irrational or no longer calculate costs and benefits but, rather, that they do so on a higher level of social aggregation. This discourages free riding by increasing ... the willingness to bear costs without selective incentives” (Wendt, 1994:386). For example, collective identity can form an important basis for “states to make commitments to multilateral action against non-specific threat ... by increasing the willingness to act on ‘generalised principles of conduct’” (Wendt, 1994:386). Thus when levels of reflexivity in the international system are high, sustaining the trust of others is a source of power for states. It allows states to act in ways which they would not otherwise be able to. In particular, states will be able to form more deeply institutionalised ties with one another than neorealism and institutionalism predict is possible.

It is important to distinguish between the type of information flow stressed by liberalism, and those discussed by institutionalists. If a states’ strategic preferences are benign, transparency helps promote trust between states, a form of diffuse reciprocity providing a basis for collective identity and solidarity founded on mutual obligations. This is the type of information flow stressed by liberalism. By contrast, the type of information flow stressed by institutionalism promotes and facilitates only specific reciprocity, and only reduces probelms of misperception. The information exchanged between the actors does not relate to their underlying strategic intentions towards one another, but rather reflects a process of strategic bargaining for tit-for-tat exchange. This distinction is important to draw attention to for two reasons. Firstly, it highlights the primacy of state preferences as the key variable in the liberal model, in contrast to institutionalism’s focus purely on non-preference related information. Secondly, the distinction is of fundamental theoretical significance, and yet has been glossed over in the debates between neorealists and institutionalists. Baldwin identifies ‘intentions versus capabilities’ as one of six key “focal points” for discussion between these approaches (Baldwin, 1993:7-8 and 4). However, this perhaps more accurately reflect institutionalists’ tendencies to hold personal sympathies for liberal arguments about the nature of the international system rather than being derived logically from institutionalist theory.

Related to its predictions about the sources of state power, liberalism also provides predictions about the levels and type of institutionalisation which will take place between states when levels of reflexivity are high. In the system’s late or mature stages of development, levels of institutionalisation will be high in contrast to the moderate and low levels of institutionalisation predicted by institutionalism and neorealism respectively. Contrary to institutionalism’s prediction that institutionalisation will be strongest between advanced industrialised states, liberalism suggests that states’ incentives to institutionalise their behaviour will be strongest at the margins of cultural change. States within the core have already internalised cultural change and so face relatively low incentives to become socialised to the system’s dominant norms. It is therefore between the core and the periphery that socialisation effects generated by the system as a whole operate most powerfully. Whilst these states may be formally non-liberal, they face intense systemic pressures to internalise shifts in their political identity compatible within the overall trajectory of international change. This is because the structure of interaction within the international system is overwhelmingly dominated by the liberal core. Liberal theory predicts not only higher absolute levels of institutionalisation within the international system than insititutionalism , but also that there will be higher levels of institutionalisation at the margins of system change.

In terms of the type of institutionalisation predicted, liberalism suggests that, in the late stages of development, relations between states will become organised around norms of diffuse reciprocity. The global international system will be overwhelmingly dominated by a security community based on the practice of trust between of liberal democratic states. This has important implications for the form of relations between states. In particular, it implies that states will begin to negotiate and sustain their relationships on the basis of mutual autonomy, and thereby in terms of equality between partners . This follows by logical extension of Keohane’s definition of the general concept of reciprocity. Keohane suggests that both specific and diffuse reciprocity share two generic features. Firstly reciprocity implies contingency because it involves conditional action. Secondly, reciprocity implies at least a rough equivalence of benefits. Within specific reciprocity, these features are only weakly developed on a tit for tat basis. However, within diffuse reciprocity these features become more strongly developed. Exchange becomes contingent upon parties taking on obligations towards one another on the basis of independent and autonomous choice, such that the relationship becomes a partnership rather than merely contractual in nature. In turn, this implies increased equivalence in the sense that parties to exchange regard these obligations as mutual.

In summary, three aspects of liberal theory have been highlighted. Firstly, liberalism adopts a transformational model of structure in which anarchy is understood as a set of rules and resources recursively drawn into an ongoing process of development within the international system. Secondly, it is not committed to the assumption that states are rational actors. Like neorealism, liberalism relies on the generative consequences of anarchy in combination with the idea that units are capable of acting as reflexive agents. Consequently liberalism provides a theory of state identity construction and socialisation within the international system. Thirdly, liberalism offers a distinctive set of predictions about the international system in a late stages of development. It holds that under these conditions trust will become the dominant source of power for states. The level of institutionalisation associated with a system in this stage of development will be high (including at the margins of cultural change), and the type of institutionalisation which takes place will be organised around norms of diffuse reciprocity.

Having outlined each of the three models, Table 1 summarises their central attributes:

Table 1: Summary Of Theoretical Framework

  Key Variable State-as-actor Status Of Anarchy Source of State Power Level of Institutionalisation Type Of Institution-alisation
Neorealism Configuration of Capabilities Reflexive Reproductive Relative capabilities Low Institutionalisation around balance of power
Institution-alism Configuration of Information and Institutions Rational Reproductive Functional incentives to cooperate Moderate
(Between core states only)
Specific Reciprocity
Liberalism Configuration of State Preferences Reflexive Transformative Trust about underlying strategic intentions High
(Including core-periphery margin)
Diffuse Reciprocity


Towards A General Theoretical Synthesis

The final section of this paper analyses the relationship between neorealist, institutionalist and liberal theories and the prospects for a general synthesis between them. Discussion is framed around a comparison of what shall be referred to as the ‘linear’ and the ‘triangular’ conceptualisations of the relationship between the three theories. Each of these conceptualisations of the relationship between the three theories suggest the possibility of a general synthesis. However, they differ over the terms on which such a synthesis can be achieved. In turn, this has important implications for the content of substantive research programmes within the discipline of International Relations after the Cold War.

Prominent institutionalists have characterised institutionalism in terms of a ‘mid-point’ between the two extremes of neorealism and liberalism. Such an analysis implies a ‘linear’ conceptualisation of the relationship between neorealism, institutionalism and liberalism. Institutionalism is seen as providing the point of synthesis on a spectrum that runs between neorealism and liberalism as depicted in Figure 2 below:

Figure 2: The Linear Model

NR                           I                           L
Rationalism       Reflectivism

The linear model suggests the intuitively plausible argument that neorealism is closer to institutionalism in terms of its understanding of the composition and dynamics of the international system than are neorealism and liberalism. On the basis of the understanding of neorealism he adopts, Keohane reasons that neorealism and institutionalism share a rationalistic approach to international relations theory (Keohane, 1989c:160). By contrast, although institutionalism and liberalism have a common concern with the origins and persistence of international institutions, their approaches differ in as much as the former is rationalist and the latter is reflectivist in orientation (Keohane, 1989c:171). Thus, on Keohane’s account, whilst neorealism and institutionalism share a rationalistic approach to the study of international relations, and institutionalism and liberalism adopt a shared concern with international institutions, neorealism and liberalism have no theoretically significant commonalties. Keohane therefore arrives at the conclusion that “institutionalists accept a version of liberal principles that eschews determinism and that emphasises the pervasive significance of international institutions without denigrating the role of state power” (Keohane, 1989a:11). Understood in these terms, institutionalism is uniquely positioned to incorporate the strongest elements of neorealism and liberalism into its analysis.

Keohane’s opinions on the relationship between schools of thought have led directly to his assertions on future research within the discipline. He suggests that future research should attempt to refine rationalist models, both conceptually and through empirical research (Keohane, 1989a:12-16). Conceptually, debates should focus on the conditions under which states will be concerned about relative and absolute gains (Keohane, 1993b:273-283). Empirically, researchers should seek to test neorealist against institutionalist theory in the context of the emerging post-Cold War international system. Specifically, Western Europe provides the most fruitful context in which to test competing rationalist theories of international relations because this is a region where neorealism and institutionalism develop divergent hypothesises about international change (Keohane, 1993:272-273; see also Mearsheimer, 1990). The thrust of his argument is that attempts to refine rationalist models are likely to demonstrate that insititutionalism provides the most fruitful terrain for a general synthesis of competing theories of international relations.

However, it is possible to develop a different conceptualisation of the relationship between neorealism, institutionalism and liberalism which challenges Keohane’s conclusions. The starting point for this alternative approach is to question institutionalist’s interpretation of neorealism as a rationalist model. Instead of making rational actor assumptions, neorealism develops a theory of state identity construction and socialisation arising from the generative consequences of international anarchy. In this sense, neorealism adopts a reflexivist logic and is closer to liberalism than it is to insistitutionalism’s rational choice framework. Rather than conceiving of the relationship between neorealism, institutionalism and liberalism in terms of a linear spectrum, it is more accurate to conceptualise the relationship in triangular terms. Within such a conceptualisation, debates between neorealists and institutionalists constitute the rationalist leg of the triangle because neorealism’s commitment to a positional model of structure means that it is potentially amenable to a rationalist interpretation, as insistitutionalists have attempted to exploit. By contrast, debates between neorealists and liberals about the generative consequences of anarchy and the overall trajectory of state identity construction and socialisation within the international system constitute the reflexivist leg of the triangle. 5 The triangular model of the relationship between neorealist, institutionalist and liberal theories is illustrated in Figure 3 below:

Figure 3:The Triangular Model

NR






I                            L

Such a conceptualisation of the relationship between realism, insititutionalism and liberalism calls into question the validity of insistitutionalists’ claims that their own theory provides the strongest basis for subsuming the research programmes of competing approaches. Instead, it suggests that liberalism’s focus on the generative consequences of anarchy and the overall trajectory of identity construction and socialisation within the international system is likely to be more productive in the search for a general synthesis of competing theories of international relations than institutionalists’ attempts to refine the rationalist models. Liberalism’s most important characteristic in this regard is the claim that, unlike neorealism, it “provides a plausible theoretical explanation for variation in the substantive content of foreign policy” (Moravcsik, 1997:534, emphasis added). Neorealist assumptions about the international system hold when states have been socialised to adopt certain preferences. A liberal model of international relations invites analysts to scrutinise the conditions under which those preferences are prevalent (Moravcsik, 1997:517). Two such conditions suggest themselves: situations in which levels of reflexivity in the international system are comparatively low due to the long term effects of competition and socialisation being poorly established; and situations in which levels of reflexivity in the international system are comparatively high, but in which states adopt non-liberal fundamentalist political identities in their interactions. This indicates that liberalism offers a “ general account of variation in socialisation ” (Moravcsik, 1997:540). It demonstrates that “liberal theory remains important, even primary, even in what are currently considered ‘least likely’ cases, for example, where there exist direct threats to national security, high levels of interstate conflict and large numbers of non-liberal states” (Moravcsik, 1997:541). Indeed, the ability of a liberal model to offer an account of variation in socialisation within the international system answers the call made by Waltz for “a single theory capable of explaining the behaviour of states, their interactions, and international outcomes” (Waltz, 1996:57). A liberal model subsumes the research programme generated by neorealist analysis, and as such satisfies the criteria for theory evaluation against which Waltz himself suggests neorealism should be assessed (see Waltz, 1997:914).

The account of variation in socialisation offered by a liberal model of systemic dynamics is illustrated in Figure 4 below. Curve A depicts a perfect positive correlation between the level of institutional density and the contingency of system structure. ‘Institutional density’ represents an aggregate of both levels and type of institutionalisation. 6 ‘Contingency of structure’ adopts the neorealist definition of structure, and refers to the extent to which behaviour is motivated by the distribution of material capabilities between actors within the international system. Curve A illustrates the ways in which levels of institutional density will rise as the international system achieves higher and higher levels of reflexivity over time. Curve B suggests a possible modification to the basic liberal model which could incorporate some of the insights offered by the institutionalist school of thought. Under the scenario envisaged in Curve B, certain levels of homogenisation and interdependence between states are considered a prerequisite to the emergence of a relatively stable core of liberal democratic states. Once this point has been reached (time period t2 below), then socialisation effects can begin to operate effectively, and levels of institutionalisation will rise more rapidly. The overall trend will thus produce a concave curve such as the one depicted. Thus, as Moravcsik identifies, the liberal model is unique in terms of its ability to explain “when and why the assumptions about state preferences underlying realism and institutionalism hold, whereas the reverse is not the case” (Moravcsik, 1997:543).

Figure 4

Variation in Socialisation


(Adapted from the matrix suggested in Wendt et. al. , 1996:38) 7

The account of variation in socialisation provided by the modified liberal model of the international system illustrated in Curve B above highlights the significance of a point of criticality in a developmental sequence of change in the international system. Criticality may be identified as a point at which a system’s emergent properties switch, resulting in a dramatic change in patterns of alignment within it (see the terminology adopted by Jervis, 1997:12-17 and 210-252). The type of system effects envisaged by the liberal model are therefore dramatically different to those identified by neorealism. The neorealist model envisages linear system effects, in which negative feedback loops promote stability through homeostatic reproduction and path dependent change (Jervis, 1997:92-146). By contrast, the system effects envisaged by the liberal model are non-linear becuase there is “a disproportion between the magnitude of the cause and the results, which will depend on the system as a whole” (Jervis, 1997:146). After the point of criticality, what were initially small and insignificant changes within the system suddenly produce an enormous change in the system’s overall characteristics. A simple example from the physical world is provided by the type of spinning top which so designed that, once it spins at above a certain velocity, its centre of gravity shifts so that it flips and spins on what was previously its head. Within the liberal model of the international system, such a point of criticality is provided by the emergence of a stable core of liberal democratic states at a global level at time period t2 on Curve B in Figure 4. Up until this point, the existence of democratic states within the system may have some significant effect on international behaviour and outcomes, but does not result in a fundamental shift in the emergent properties of the system as a whole. After this point, however, the structure of interaction within the system as a whole changes. The socialisation effects generated by the core begin to dramatically transform overall patterns of alignment and change within the entire system. 8

It is the change in the structure of interaction within the international system which is the key to identifying the point of criticality within it. Wendt has recently identified four ‘master variables’ capable of driving structural transformation within the international system: interdependence, homogeneity, common fate, and self-restraint. He discusses the ways in which each of these factors are likely to be present in an international system characterised by a strong convergence of liberal preferences (Wendt, 1999:343-366). However, he also recognises that although these master variables are necessary conditions for producing structural transformation, they are not sufficient ones. This is because structural transformation requires “not just that identities change, but that their frequency and distribution cross a threshold at which the logic of the structure tips over into a new logic” (Wendt, 1999:365). Identifying changes in the structure of interaction within the international system supplies the ‘missing link’ in Wendt’s account of structural transformation. This is because changes in the structure of interaction are the only changes capable of fundamentally altering the system’s ordering principle . With the emergence of a stable core of liberal states, the ordering principle of the international system shifts away from anarchy and towards a network form of organisation. Networked systems are characterised by a decentralised authority structure (Wendt, 1999:307). Unlike anarchic systems, there is an authority structure provided by shared cultural norms of the liberal core. This provides a de facto rule of law that limits what states can legitimately do to advance their interests. Unlike hierarchical systems, there is no central authority to enforce or police this authority structure (Wendt, 1999:307-308). Thus the change in the structure of interaction after the point of criticality has been reached represents a fundamental shift in the basic ordering principle of the international system. It reflects a shift away from anarchy as the irreducible feature of international politics, and towards a network of states centred on the liberal core.

The most obvious point of historical reference for the emergence of a critical mass of liberal democratic states is provided by the recent collapse of the Soviet Union. As Huntley identifies, “(t)he strikingly rapid collapse of communist ideology and Soviet authority marks an historic threshold” for the development of the international system (Huntley, 1996:67). Liberalism is uniquely positioned to both explain this development, and account for why it might constitute a point of enormous significance for the overall course of historical change in relations between states. Firstly, the liberal model could plausibly suggest that the collapse of the Soviet Union was not only a cause of system change, as suggested by neorealism, but was itself symptomatic of overall historical trends within the international system. As Huntley argues, “the demise of Soviet Style communism is perfectly compatible with the long term patterns of change in international politics” provided by a liberal understanding of competition and socialisation (Huntley, 1996:67). Secondly, a liberal model provides a vision of the emerging post-Cold War international order which is “sharply divergent” from the predictions offered by established theoretical frameworks (Huntley, 1996:62). As Huntley identifies, the “new era of world politics now emerging seems likely to provide telling tests between opposing expectations” of competing theories of international relations (Huntley, 1996:72).

In particular, the general synthesis provided by the liberal model suggests a very different research programme to that proposed by institutionalists. Within a European context, the research programme is broader than institutionalists anticipate. Liberalism, as well as neorealism and institutionalism, develops divergent hypotheses regarding the overall nature of change in the post Cold War Europe. As has been identified, these hypotheses encompass the sources of power for states and the levels and type of institutionalisation. Secondly, liberalism suggests that the emergence of a critical mass of liberal states alters the configuration of preferences within global the international system. The structure of interaction within the global international system becomes overwhelmingly dominated by the liberal core, with the consequence that the democratic peace will begin to produce powerful socialisation effects. These effects are liekly to place pressure on non-liberal states at the core-periphery margin to engage in cultural change by adopt shifts in their underlying political identity. Thus liberalism suggests that the emergence of a stable liberal core in the wake of the collapse of the Soviet Union has implications for the international system which extend beyond the confines of Western Europe. It affects the international system at a truly global level by altering its emergent properties and shifting the overall trajectory of its historical development. Prominent neorealists have identified China and Japan as emerging Asian powers which are likely to engage in balancing behaviour within this region as part of a more general shift towards global multipolarity (Waltz, 1993; Layne, 1993). The post Cold War foreign policy behaviour of key East Asian states identified by neorealists provide critical tests for liberalism’s predictions about the overall trajectory of change in the post Cold War international system. States such as Japan and China are at the margins of international change in as much as they interact in a regional context in which complex patterns of interdependence are not clearly established, and there are large numbers of formally non-liberal states. As such these cases represent ‘least likely’ cases for the strong predictions offered by the liberal model. However, even under these conditions, liberalism makes predictions about both the levels and type of institutionalisation and the sources of power between states in the East Asia region which go well beyond what might be anticipated by institutionalists, even compared to institutionalist’s predictions about state behaviour within a European context. This suggests that the collapse of the Soviet Union and the end of the Cold War has much more dramatic implications for research within the discipline than institutionalists anticipate. Indeed, liberalism indicates that the collapse of the Soviet Union ought to be considered a development of truly profound significance for the study of international relations.

In summary, it is possible that prominent institutionalists have misconstrued the nature of the relationship between schools of thought in the discipline, and thereby underestimated the possibilities for a general synthesis of their insights. Institutionalists have proposed a linear conceptualisation of the relationship between the three theories. Understood in these terms, attempts to refine rationalist models provide the most fruitful research programme for the discipline of international relations. It follows that the critical focus of applied research after the Cold War is the West European region. However, it is possible to dispute institutionalists’ conclusions by questioning its interpretation of neorealism as a rationalist theory. Instead, neorealism may be interpreted as adopting a reflexive logic in virtue of its focus on the generative consequences of anarchy and the overall trajectory of identity construction and socialisation within the international system. This implies a triangular conceptualisation of the relationship between the three theories in which debates between neorealists and institutionalists constitute its rationalist leg, and debates between neorealists and liberals constitute its reflexivist leg. Understood in these terms, liberalism’s focus on the generative consequences of anarchy and the overall trajectory of identity construction and socialisation within the international system is likely to be more productive in the search for a general synthesis of competing theories of international relations than institutionalists’ attempts to refine the rationalist approach. This is because of the ability of liberalism to offer a fully generative model of the international system providing an account of variation in socialisation which can subsume the research programmes generated by both neorealism and institutionalism. Such conclusions have important implications for substantive research within the discipline, indicating that the collapse of the Soviet Union has much more profound implications for the study of the global international system than institutionalists have indicated. In particular, the foreign policy behaviour of key states in East Asia as well as in a European context will provide critical tests for identifying the overall trajectory of change in post Cold War international relations. In concluding his recent book, Wendt suggests that the central question faced by the field of International Relations today is “can the states system achieve reflexivity ... and what are the implications if it does?” (Wendt, 1999:375). Liberalism hypothesises that since the collapse of the Soviet Union and the emergence of a critical mass of liberal states at a global level, the international system has achieved reflexivity, and this has resulted in a dramatic shift in the pace of cultural change and the overall trajectory of socialisation within it.



Conclusions

One of the most distinctive features of debates within the discipline of international relations over the 1980’s and 1990’s has been the attempt to pin down systemic approaches to the field of enquiry. Systemic theory offers the promise of providing parsimonious, general explanations of patterns of behaviour in international politics, and as such potentially provides the discipline with the logical rigour and explanatory insight that its status as a modern social science demands.

However, the debates which have taken place between neorealists and institutionalists have become increasingly unproductive, both theoretically and empirically. Even major contributors to the debate have long recognised it has “tended to obscure rather than clarify”, and has become increasingly “sterile” (Powell, 1994:313 and 344). Yet the reason for this failure is not that institutionalists’ attempts to take on neorealism on its own terms is a fundamentally misconceived endeavour. A more sophisticated approach to identifying the deficiencies of institutionalist critique is to recognise that institutionalists have developed an uncritical understanding of the relationship between schools of thought. They adopt a linear model of the relationship between neorealism, institutionalism and liberalism. In this view, institutionalism is seen as providing the point of synthesis on a spectrum that runs between neorealism and liberalism and as such is the model most likely to be able to reconcile the competing claims of alternative positions. Whilst superficially plausible, such a view of the relationship between schools of thought glosses over important similarities between neorealism and liberalism as systemic approaches to international relations. Indeed, it is possible that it has been institutionalism’s commitment to rationalism which has actively blocked the development of a model of the international system fully compatible with liberal assumptions. A more nuanced approach proposes a triangular understanding of the relationship between the three models of the international system, with debates between neorealists and institutionalists constituting its rationalist leg, and debates between neorealists and liberal constituting its reflexivist leg. Such a conceptualisation of the relationship between the three models allows a fuller appreciation of the complex linkages which exist between them. Moreover, it highlights the way in which liberalism’s distinctive focus on the generative consequences of international anarchy allows it to offer an account of variation in socialisation within the international system which is capable of subsuming the combined research programmes generated by neorealist and institutionalist models.

Finally, these theoretical considerations have important implications for the content of substantive research programmes within the discipline in the post Cold War period. Proponents of neorealism, institutionalism and liberalism have all identified that the collapse of the Soviet Union has provided an historic opportunity to test their competing theoretical claims. Neorealists have developed a broadly pessimistic analysis of the nature of international political change after the Cold War, predicting a return to conflictual multipolarity on a global scale. The nature of the responses to the challenge laid down by neorealists derived from institutionalism and liberalism varies considerably. On the basis of their argument that attempts to refine rationalist models are likely to provide the most promising basis for future research, prominent institutionalists have suggested that the discipline should focus on testing a relatively limited and extraordinarily narrow range of propositions about the conditionality of relative gains seeking behaviour, and to do so primarily within a West European context. In making these claims, however, institutionalists have glossed over the way in which the collapse of the Soviet Union, perhaps one of the most significant developments in world politics over the past half a century, was not predicted by either neorealist or institutionalist international relations theory. As such, they have failed to identify the way in which the collapse of the Soviet Union and the emergence of a relatively stable body of formally liberal democratic states at a global level has raised fundamental questions about the long term historical dynamics embodied by the international system. By contrast, liberalism’s distinctive claim to subsume the insights of the neorealist and institutionalist models suggests that it is capable of both accounting for this development in terms of the long run effects of competition and socialisation, and of providing a much richer and ultimately more challenging agenda for research into the overall nature of post-Cold War international change. International relations specialists can and must learn from the debates between neorealists and institutionalists that have taken place over the 1980’s and 1990’s in order to revitalise systemic theory for the post-Cold War period. However, to do so they need to engage critically with what have become deeply ingrained assumptions about systemic theorising, and think imaginatively about the overall trajectory of international political change at the turn of the century.



Bibliography

Ashley, R. K. (1986) "The Poverty Of Neorealism". In R.O. Keohane (Ed.) Neorealism And Its Critics pp. 255-300. New York: Columbia University Press.

Axelrod, R. and Keohane, R.O. “Achieving Cooperation under Anarchy: Strategies And Institutions”. In D.A. Baldwin (Ed.) Neorealism And Neoliberalism: The Contemporary Debate pp. 85-115. New York: Columbia University Press.

Baldwin, D.A. (1993) “Neoliberalism, Neorealism and World Politics”. In D.A. Baldwin (Ed.) Neorealism And Neoliberalism: The Contemporary Debate pp. 3-28. New York: Columbia University Press

Buzan, B. (1993) From International System To International Society: Structural Realism And Regime Theory Meet The English School. International Organisation 47:327-352

Buzan, B. et. al. (1993) The Logic Of Anarchy: Neorealism To Structural Realism . New York: Columbia University Press.

Cox, R.W. (1986) "Social Forces, States And World Orders: Beyond International Relations Theory". In R.O. Keohane (Ed.) Neorealism And Its Critics pp.204- 254. New York: Columbia University Press.

Dessler, D. (1989) What’s At Stake In The Agent-Structure Debate? International Organisation 43:441-474.

Elman, C. (1996) Cause, Effect And Consistency: A Response To Kenneth Waltz. Security Studies 6:58-61

Giddens, A. (1984) The Constitution Of Society:Outline Of The Theory Of Structuration. Cambridge: Polity Press

Grieco, J.M (1993a) “Anarchy And The Limits Of Cooperation:A Realist Critique Of The Newest Liberal Institutionalism”. In D.A. Baldwin (Ed.) Neorealism And Neoliberalism: The Contemporary Debate pp. 116-142. New York: Columbia University Press.

Huntley, W.L. (1996) Kant’s Third Image: Systemic Sources Of The Liberal Peace. International Studies Quarterly 40:45-76.

Jervis, R. (1997) System Effects: Complexity In Social And Political Life . Princeton: Princeton University Press

Keohane, R.O. (1984) After Hegemony: Cooperation And Discord In The World Political Economy . Princeton: Princeton University Press.

Keohane, R.O. (1986b) “Theory Of World Politics: Structural Realism And Beyond”. In R.O. Keohane (Ed.) Neorealism And Its Critics pp. 158-203. New York: Columbia University Press.

Keohane, R.O. (1989a) “Neoliberal Institutionalism: A Perspective On World Politics”. In R.O. Keohane International Institutions And State Power: Essays In International Relations Theory pp. 1-20. London:Westview Press.

Keohane, R.O. (1989b) “Reciprocity In International Relations”. In R.O. Keohane International Institutions And State Power: Essays In International Relations Theory pp. 132-157. London: Westview Press.

Keohane, R.O. (1989c) “International Institutions: Two Approaches”. In R.O. Keohane International Institutions And State Power: Essays In International Relations Theory pp. 158-179. London: Westview Press

Keohane, R.O. (1993) “Institutional Theory And the Realist Challenge After The Cold War”. In D.A. Baldwin (Ed.) Neorealism And Neoliberalism: The Contemporary Debate pp. 269-300. New York: Columbia University Press.

Keohane, R.O. and J.S. Nye, (1989) Power And Interdependence: World Politics In Transition (Second Edition). Boston: Little,Brown.

Layne, C. (1993) The Unipolar Illusion: Why New Great Powers Will Rise. International Security 17:5-51.

Mearsheimer, J.J. (1990) Back To The Future: Instability In Europe After The Cold War. International Security 15:5-56.

Moravcsik, A. (1997) Taking Preferences Seriously: A Liberal Theory Of International Politics. International Organisation 51:513-553

Powell, R. (1993) “Absolute And Relative Gains In International Relations Theory”. In D.A. Baldwin (Ed.) Neorealism And Neoliberalism: The Contemporary Debate pp. 209-233. New York: Columbia University Press.

Powell, R. (1994) “Anarchy In International Relations Theory: The Neorealist- Neoliberal Debate”. International Organisation 48:313-344.

Ruggie, J.G. (1986) "Continuity And Transformation In The World Polity: Towards A Neorealist Synthesis". In R.O. Keohane (Ed.) Neorealism And Its Critics pp. 131-157. New York: Columbia University Press.

Waltz, K.N. (1979) Theory Of International Politics . New York:Random House.

Waltz, K.N. (1986) "Reflections On 'Theory Of International Politics': A Reply To My Critics". In R.O. Keohane (Ed.) Neorealism And Its Critics pp. 322-346. New York: Columbia University Press.

Waltz, K.N. (1993) The Emerging Structure Of International Politics. International Security 18:44-79.

Waltz, K.N. (1996) International Politics Is Not Foreign Policy. Security Studies 6:54-57.

Waltz, K.N. (1997) Evaluating Theories. American Political Science Review 91:913- 917.

Wendt, A.E. (1987) The Agent-Structure Problem In International Relations Theory. International Organisation 41:335-370.

Wendt, A.E. (1992) Anarchy Is What States Make Of It:The Social Construction Of Power Politics. International Organisation 46:391-425.

Wendt, A.E. (1994) Collective Identity Formation And The International State. American Political Science Review 88:384-396.

Wendt, A.E. (1999) Social Theory Of International Politics . Cambridge:Cambridge University Press

Wendt, A.E. et. al. (1996) “Norms, Identity And Culture In National Security”. In P.J. Katzenstein (Ed.) The Culture Of National Security: Norms And Identity In World Politics . New York: Columbia University Press.




Endnotes

Note 1: It might be suggested that the institutionalist model can be re-read to contain a reflexive dimension because of the iterative nature of the games it assumes are played by actors. In particular, institutionalism assumes that actors value future rewards, and this could be understood as a form of ability to acquire knowledge and reflexive awareness. However, it is important to recognise that the knowledge of actors in the institutionalist account is given by assumption, such that they merely respond to patterns of incentives generated in their external environment. It makes no assertions about their basic identity as actors because their knowledge is not internalised through a process of socialisation. Thus the iterative games played by actors in the institutionalist model does not require them to behave reflexively and engage in learning. Instead actors merely respond mechanistically to external incentives. Back.

Note 2: Huntley’s claim that international anarchy has generative consequences of a transformative nature and may play a role in the spread of the liberal peace has been under-explored in subsequent literature. Back.

Note 3: It is important to carefully distinguish between the long shadow of the future discussed by Huntley in his discussion liberal states’ capacity to be trusted and the shadow of the future discussed in institutionalist international relations theory. The former is based on trust and the diffuse practice of reciprocity. Peaceful relations between states emerge as the unintended consequences of state’s shared or collective political identity as actors. By contrast, the shadow of the future in institutionalist theory is not based on trust, promotes only specific reciprocity between units, and does not involve socialisation to a shared identity. For further elaboration on this point, see Wendt’s discussion of the difference between behavioural and constitutive effects of interdependence and ongoing interactions (1999:344-349) Back.

Note 4: Similarly, Moravcsik has identified that “socialisation towards convergent norms stem from convergent domestic institutions and ideas. Liberal institutions and norms may be particularly conducive to the promotion of peace and co-operation, but the argument also implies that the convergence of nonliberal values ... may also have significant effects” (Moravcsik, 1997:540). Back.

Note 5: The third leg of the triangle identifies the links which exist between institutionalism and liberalism. In particular, both institutionalism and liberalism stress the importance of institutions for the constitution of the international system. However, institutionalism and liberalism make very different assumptions about the fundamental nature of social actors. As such the links which exists between these schools of thought are not directly relevant to the overall line of argument being developed by this paper, and shall not be discussed here. Back.

Note 6: It is important to aggregate levels and type of institutionalisation into a single measure for institutional density because historically there have been hegemonial international systems embodying fairly high levels of institutionalisation, but which remained highly asymmetrical and at best operated around the limited practice of reciprocity (see Keohane, 1984) Back.

Note 7: Interestingly, in his recent book Wendt turns the matrix into a graph, and displays an interest in progress along its diagonal over time (Wendt, 1999:254 and 313-369). In so doing, he has taken an important step towards developing a general account of account of variation in socialisation. However, he is careful to caution that different pathways other than the diagonal may be taken by the system or by different actors within the system. This is suggested both by Curve B and by the possibility of fundamentalist behaviour by units in the late stages of development of the international system. Wendt also explores other situations which highlight the way in which displaying an interest in only the diagonal mistakenly conflates the role of shared ideas and co-operation (Wendt, 1999:254-255 and 259-312). Back.

Note 8: The overall picture which emerges from this synthetic approach resonates strongly with systemic dynamics envisaged by Buzan in his attempt to render compatible structural realism and regime theory with the insights offered by the English School of international relations theory (Buzan, 1993). Buzan attempts to provide a developmental model of systemic change linking structure, process and interaction as components of system organisation. Furthermore, to achieve this he adopts a critical stance towards conceptualising the relationship between schools of thought often considered distinct in their approaches to international relations. The work of at least one major analyst therefore provides a recent precedent for the enterprise engaged in this paper. Back.