Integrative pluralism, security studies, and IR

Our first task is to clarify how an integrative pluralist approach would conceive of IR theory in general. Given its emphasis on integration at ‘the concrete explanatory level’, its focus is clearly on explanatory theory.Footnote ¹¹ Tim Dunne, Lene Hansen, and Colin Wight define explanatory theory as an attempt ‘to explain events by providing an account of causes in a temporal sequence’.Footnote ¹² This view is in opposition to a view of theory, central to many grand theoretical designs, as ‘simplified pictures of reality’ or ‘maps’.Footnote ¹³ The former notion of theory is close(r) to the one adopted here,Footnote ¹⁴ but of these dominant currents, the flaws with grand theoretical ‘maps’ are most in need of addressing.

As David Lake observes, existing grand theoretical traditions each have ‘a unique set of core assumptions’ that ‘embody different visions of world politics as inherently conflictual, more cooperative, or the “open” product of actors’ own actions’.Footnote ¹⁵ Theorists then often reify these assumptions – conflating the ‘map’ with the world as a whole.Footnote ¹⁶ These assumptions become so entrenched that theorists working within a particular ‘ism’ feel compelled to reject inclusion of relevant considerations simply because doing so would mean the theory no longer fit within the ‘ism’.Footnote ¹⁷ The reason grand theorising looks like this, we suspect, is the presumption among many IR theorists, typically attributed to Lakatos, that theory must begin with a single deductive logic stemming from a ‘hard core’ of parametric assumptions as well as a ‘negative heuristic’ – ‘rules that prohibit certain kinds of theorizing’.Footnote ¹⁸ In John Mearsheimer's formulation, for example, realism's hard core assumption is that ‘the structure of the international system forces states … to act aggressively toward each other’; its negative heuristic demands the view that international institutions represent a ‘false promise’.Footnote ¹⁹ International cooperation has little place in the theory.

What's more, existing grand theory, identified as including ‘realism’ or ‘liberalism’, are not theories at all but ‘research traditions’, which are not internally coherent.Footnote ²⁰ Realists, for example, are divided into offensive realist, defensive realist, and neoclassical realist camps whose arguments are sharply distinct.Footnote ²¹ ‘Liberalism’ may similarly refer to those like Robert O. Keohane who focus on international institutions, those like Andrew Moravscik who emphasise domestic political processes, or theorists of the democratic peace who may consider both.Footnote ²² Constructivism also is not a theory but, in Chris Brown's formulation ‘a set of dispositions towards social reality … [focused] on ideas, values, norms and practices’, with much of the recent work particularly emphasising practices.Footnote ²³

An integrative pluralist perspective views these traditions as a repository of theories, each possibly valuable for their abstract identification of causal factors available to be employed in explanations. This is quite similar to Andrew Bennett's vision for a ‘taxonomy of theories on social mechanisms’ that ‘mirrors the three leading “isms” in the IR subfield’, then used to foster more sophisticated explanatory theories at the middle range.Footnote ²⁴ Where integrative pluralism goes further is by viewing that latter level of theoretical development as simultaneously an effort to explain concrete phenomena and as the testing ground for broader theories that structure the taxonomy. Indeed, one goal of integration at the explanatory level could be progress at the grand theoretical level, defined as integration-informed modifications to theories that leave us better able to draw from it for future explanations. Ultimately, this should involve grand theory going beyond the straitjacket of the ‘isms’ to integrate multiple causal factors and processes to depict international relations in a coherent way.

The basis for this programme is a fairly innocuous notion that too often gets lost in the array of theoretical approaches available to the IR researcher. While accepting that we will never know with certainty that we have gotten things just right, it would seem that for any healthy science, whatever degree or kind of pluralism that pertains should at least not prevent us from being able to sift through ‘knowledge claims in the hope of discarding those that fail to provide a valuable contribution to the overall stock of knowledge’.Footnote ²⁵ Put more positively, it seems sensible to ground the contact between different theories on the goal of developing better explanations of international political phenomena. That notion, though, is much more easily said than done.

As noted, Wight draws from the philosopher Sandra Mitchell, whose integrative pluralism has been featured prominently in philosophy of science debates about the best way to embrace pluralism.Footnote ²⁶ Mitchell argues that because biology is a field marked by ‘multicomponent, multilevel, evolved complex systems’, theoretical pluralism is something ‘the nature of nature demands’.Footnote ²⁷ Causes are multiple, interactive, and are emergent at different levels of reality. Additionally, specific instances of outwardly like phenomena can involve different causal pathways, and historical change can generate novelty. Furthermore, we can never expect the conceptual representations we develop to be fully secure or stable because of the ‘combination of our [limited] cognitive abilities and the purposes for which we intend to use … knowledge’.Footnote ²⁸ Therefore, we should expect to have multiple legitimate ‘grand’ or general theoretical perspectives because we need ‘a means of providing more comprehensive and multi-dimensional accounts of complex phenomena’.Footnote ²⁹ The complexity of biological life reveals that though we may have a cogent general theory from which we can draw a fairly parsimonious explanation for a particular phenomenon, once we attempt to adequately explain instances of it we will almost always find that a number of other supposedly competing theories may be crucial to fleshing out the processes at work.

If this is the case then the standard way of conceiving of progress in science, with general theories – or the major theories that constitute them – being ‘tested’ against one another, will not only not lead to some chimerical theoretical unification, but will be detrimental to our ability to understand the world. As Mitchell puts it, a ‘search for the one, singular, absolute truth must be replaced by humble respect for the plurality of truths that partially and pragmatically represent the world.’Footnote ³⁰ This means that theories ‘are not always competing; they are sometimes compatible and complementary’,Footnote ³¹ a point very much in line with Bennett's vision for ‘structured pluralism’ in IR. As Bennett contends, ‘scholars have constructed IR's leading “paradigms” around groups of theories about kinds of causal mechanisms that are in fact not mutually exclusive.’Footnote ³² Unlike Bennett's approach, though, integrative pluralism provides a pathway by which we should expect grand theoretical refinement and possibly even cumulation.

Both Wight and Mitchell are somewhat vague on how this would proceed because, as Mitchell emphasises, ‘the strategy for integrating diverse theories and explanations will not be algorithmic.’Footnote ³³ Integrative strategy will depend heavily on the specifics of the phenomena we examine, the angle of our investigation into them, and the always fallible assessments we make about how those specifics should guide the search for salient causes. And yet Wight boldly asserts that as we go about the process of attempting to integrate theories at the concrete explanatory level ‘some theories may not survive … (or) may undergo substantial change and modification’.Footnote ³⁴ Fundamentally this is because, as Mitchell puts it, however ‘complex and however many contributing causes participated, there is only one causal history that, in fact, has generated the phenomenon to be explained’.Footnote ³⁵ In our efforts to uncover that one causal history – which, of course, we can never be completely certain we have found – the consideration of evidence acts to regulate the pluralism within any scientific domain. How, though, to enact this vision of scientific practice to the specific field of IR?

Wight's attraction to Mitchell's work, with its frank acknowledgment that science seeks, however haltingly, to uncover ‘only one causal history’, is clearly connected to his commitment to a realist philosophy of social science.Footnote ³⁶ Scientific realism is defined by the view ‘that we are warranted in taking the principles of our current theories as true and the entities they postulate as real, and not merely as “useful”’.Footnote ³⁷ The critical realist form of scientific realism that Wight adopts, developed with the social sciences specifically in mind – and to which IR theorists seem most amenable – has been prominently used in the recent past by IR theorists in efforts to reorient the field.Footnote ³⁸ More generally, critical realism aims for a ‘synergy between social theory and philosophy’,Footnote ³⁹ and as such has generated a ‘broad alliance of social theorists and researchers trying to develop a properly post-positivist social science’.Footnote ⁴⁰ This alliance shares an insistence that because what can be known must at some level be a function of what is we should begin with ontological considerations when developing our methodological choices.

All of which is to say, since Mitchell clearly developed integrative pluralism by way of considering what biological ‘nature’ demanded per scientific progress, exporting it to security studies and IR must involve a consideration of the ‘nature’ of our subject matter. Talk of nature in the social world should of course rightly be viewed with a sceptical eye; it would seem to denote some unchanging element or wellspring from which to explain social life. Given both the broad space/time variation on offer empirically, and the ethical problematic of seemingly denying the human potential for fundamental change, we do not wish to be misunderstood when referencing such a task. Critical realist philosophy opens space for such talk from a postpositivist register by reference to the concept of emergence: while all social phenomena certainly have some foundation in the natural world, aspects such as psychological and social dynamics cannot be reduced to those elements.Footnote ⁴¹ Any ontology of the social realm must account for the very existence and seemingly infinite richness of social life itself, such that the major constants are contingency, complexity, and systems that are always open to exogenous influence. That shares quite a bit with Mitchell's description of biology, though perhaps with a (much) heavier dose of contingency and historical change on offer. Indeed, Roy Bhaskar, the founder of critical realism, developed thoughts on ‘integrative or structured pluralism’ as the approach to social science that would properly take into account the emergent ‘distinctions’ and yet simultaneous ‘connections between the various objects of scientific inquiry’ in the social world.Footnote ⁴² We turn now to critical realist insights into the social world to inform a more detailed understanding of how integrative pluralism might manifest in security studies and IR.

Integrative pluralism for security studies and IR

Social theorists draw from critical realism as an ‘underlabouring’ resource, for instance in working from ‘transcendental argument’ – asking ‘what must be true in order for x to be possible’ – to think ontologically about some social sphere, then considering how that should impinge on our epistemological strategies.Footnote ⁴³ Bhaskar began by asking what the basic ontology of the world must be given the natural sciences’ many successes, and from there constructed a vision for social-scientific practice more consciously drawn from that ontological assessment.Footnote ⁴⁴ While we do not seek to sweepingly ‘reclaim reality’ for IR theory by way of a fully developed critical realist foundation,Footnote ⁴⁵ we here take seriously the ‘underlabouring’ function of the perspective to flesh out integrative pluralism for security studies and IR, specifically.

There are no doubt numerous variations on the way that an integrative pluralist methodology might be deployed in IR, and numerous ways that critical realist philosophy and social theory might be drawn upon to inform that deployment. Here we begin developing a vision for integrative pluralism by way of the two most basic insights that grew out of critical realism's combination of a commitment to a mind-independent reality and the contention that the best scientific practice has successfully captured aspects of that reality: reality is marked by deeply stratified and emergent layers, and events thus arise from underlying and often unobservable but still real causal forces. Scientific investigation, then, should seek to go not only ever deeper in uncovering those often readily unapparent causal forces, but the explanation of social phenomena involves mapping the way they intersect, sequence, and conjoin to give rise to events of interest. Thus rather than simply correlating that which can be observed – which in practice usually means quantified – critical realism advises that we conceive of explanation as the mapping of ‘complexes’ of different things that constitute the key generative aspects of a phenomenon of interest.Footnote ⁴⁶ In short, because we are explaining a world that is ‘structured, differentiated, stratified, and changing’,Footnote ⁴⁷ the enterprise of IR theory must embrace increasingly complex, multicausal explanations or else it will stagnate.

From this baseline, critical realists contend that causes are constituted by the internal dispositions of whatever ‘things’ – whether ‘forces, powers, mechanisms, or sets of relations’ – make ‘things happen or “trigger” events’.Footnote ⁴⁸ These causal ‘things’ may, depending on the nature of and approach to a phenomenon, rest at various levels of the social strata, and consist of what are traditionally thought of as either ‘material’ or ‘ideational’ characteristics.Footnote ⁴⁹ This in some way complicates our theoretical task, pointing towards the need for security studies and IR, concerned they often are with the most macro of social phenomena, to consider together the panoply of things that have an effect in the world. IR theory broadly conceived would be a constant process of at once isolating the tendencies of causal mechanisms, processes, relations, or simply brute material facts and modelling how, when operating together across the various layers of the social strata, they conjoin agents and structures and give rise to, and thus explain, an event of interest. Explanation would by necessity require what Wight calls ‘epistemological opportunism’,Footnote ⁵⁰ which likely translates into the creative deployment of process tracing,Footnote ⁵¹ mixed methods, and comparative strategies that go beyond most similar or different cases and instead compare between causes operating in sometimes quite different environments.Footnote ⁵² At the same time, elided is the mandate to search endlessly and unfruitfully for correlations in a changing and complex world,Footnote ⁵³ as well as for the chimerical ‘covering laws’ – whether something like ‘utility maximization’ or the ‘Thucydides’ Trap’Footnote ⁵⁴ – meant to explain them. Identification of a correlation might spark the search for a cause of interest, but ‘given the contingency and flux of the social world, where multiple mechanisms are also constantly interacting’, causes have tendencies that can influence a range of phenomena in a range of different ways depending on the other causes at work.Footnote ⁵⁵

What does this mean for our practice of integrative pluralist research? At one level, following this path obviously means de-emphasising the importance of parsimony in theorising and embracing greater theoretical complexity. As Wight puts it, we must ‘move beyond a model of science that views simplicity, coherence and reduction as primary goods. We need models of science able to incorporate the chaotic complexity of the international system.’Footnote ⁵⁶ Scholarly specialisation makes this possible: we do not all have to be political psychologists or political economists, for example, to recognise that we might need to draw from both. Explanations of international politics therefore need to be open to theoretical fields that cover causal forces ranging from the tendencies of individuals to more familiar ‘material’ and ‘ideational’ macro-structuring conditions. Pragmatically, this means theories more readily open to specifying the sociological and psychological causes within their operations. Molecular biology, for example, must operate within the constraints of physical and chemical laws. In analogous fashion, while each IR scholar need not be simultaneously an expert in psychology, sociology, anthropology, economics, and comparative politics, they should make sure that the causes they hypothesise are at least compatible with, if not derived from, the best available knowledge from those other disciplines.

What we have in mind, then, is a field where normal research practice involves drawing from broad theories of world order, as well as theories of social processes from other fields, and integrating them as evidence calls for in order to explain concrete phenomena. This should most directly lead to the development of synthetic ‘middle-range’ theories.Footnote ⁵⁷ The explanations themselves ‘will be, almost by definition, more local and less global to reflect the context richness of the behavior of multilevel, multicomponent systems’.Footnote ⁵⁸ From there, though, the content of such ‘local’ explanations should, as we determine how their proposed causes operate with others in certain contexts, help refine broader theories, even world order theories. This is especially important for a social science as broadly conceived as IR. As Bhaskar stresses in his own development of integrative pluralism, the historical ‘flux of social life’ means that there must be a much more intensive conversation between what he calls ‘“pure” theory’ and the ‘concrete explanatory work’ that draws from it; the causes identified in the former ‘may be themselves only relatively enduring’.Footnote ⁵⁹ Integrative pluralist explanatory practice is thus simultaneously an attempt to find a more satisfying and precise explanation than on offer from more grand theoretical repositories, and a testing ground for the viability of the dynamics enshrined within those theories.

We might, for instance, consider middle-range theories, and even alternative grand theories, as a source of either supplementary hypotheses or conditioning hypotheses for any particular grand theory. Democratic peace theory, for example, increased in nuance and cogency as its explanations withstood significant criticism,Footnote ⁶⁰ but it also need not necessarily be viewed as a strict alternative to realism. Rather, it might also be a conditioning hypothesis for realism, positing that the security dilemma does not work the same way among democracies as it does in relations involving other regime types. Put quite simply, at every level of generality, theoretical development should proceed through attempts at integration, moving from such explanations back to the theories drawn from, and then repeating that process.

The basic idea of integrating theory in this way is not new, of course: in After Hegemony, for example, Robert Keohane tried to show how one can start from realist premises to show how international institutions can be efficacious.Footnote ⁶¹ The discipline as a whole, however, failed to follow the example of this neorealist-neoliberal synthesis and look for additional ways to draw alternative grand theories together instead of separating them into warring camps.Footnote ⁶² It is clear, then, that for various reasons not only is integrative pluralist theoretical explanation difficult, but broader disciplinary impact is especially so. Fortunately, as the next section illustrates, it is definitely not as if the former effort is foreign to IR theorising, especially in its security studies core.

Integrative pluralism by other names in security studies

It is not our intention to argue that there are prominent examples of research in IR that perfectly deploy the integrative pluralist playbook outlined above. What we wish to show here is that in the following examples the aspects of the research that do the most to distinguish it, that help it develop a compelling explanation and generate promising theoretical advances, are aspects that dovetail with an integrative pluralist approach. We also do not mean to argue or imply that in our call to see a field that more explicitly adopts an integrative pluralist approach we would have all research converge on a common methodology. The objects we study, the intentions we have for the knowledge produced about them, and our intractable epistemological limitations mean that even if every IR scholar conducted their work with integrative pluralism as their orientation towards theory, we would still expect to see a wide range of different modes of scholarship. Indeed, it is for that reason this section offers examples of integrative pluralism in action both from work that most would see as within the neopositivist mainstream, as well as examples from scholars that situate themselves within the varied postpositivist or critical security studies side of the discipline. That they all evince integrative pluralist tendencies, the wager is that with a more conscious focus on this model of IR research practice not only might the works themselves have been improved, but the field would have certainly been able to leverage them towards greater impact on the general theoretical landscape.

A prominent example of mainstream integrative pluralist theorising at the middle-range level is Jack Snyder's Ideology of the Offensive. Footnote ⁶³ In particular, Snyder's explanation of the catastrophic military planning in pre-First World War Europe identifies three mechanisms that may influence policymaking in the military or, indeed, in any bureaucracy, each operating at a different level of the social strata: rational calculation, doctrinal simplification, and motivated bias. To some extent, Snyder notes, military planners actually do what rationalists and realists posit, logically weighing the costs and risks of alternative options in the context of the strategic situation facing them. However, ‘below’ the structural conditions of a strategic relationship, military planners sometimes skew their analysis in ways that favour the institutional interests of the military, resulting, in his cases, in ‘offensive strategy as an institutional defense’ – just the explanation liberal theorists might suspect.Footnote ⁶⁴ The third set of factors Snyder considers – motivated bias, and especially wishful thinking – are psychological; they can work to exaggerate the effects of institutional interests, resulting in blatant departures from any sort of rational decision-making.

Snyder brings these considerations together into an account of ‘organisational ideology’, emphasising the interaction of strategic, institutional, domestic political, and psychological factors. In France, for example, military leaders in the last prewar years were so concerned about domestic political threats to their organisational ideology – essentially, the preference for a professional rather than a conscript army – that they shaped war plans to require it. They were only able to justify their plans, in Snyder's account, by refusing to believe intelligence reports about the size of the attack the Germans were planning.Footnote ⁶⁵ In sum, the French war plan of 1914 makes no sense without recognising how French military leaders were influenced not only by the German threat, but also by the organisational ideology into which they were socialised, by the domestic political threats to the policies that ideology dictated, and by the psychological biases that were generated by those conflicting strategic and political concerns.

Snyder thus proposes an integrative pluralist account, showing how multiple causes – strategic, psychological, and institutional – conjoin to create organisational ideologies. The discipline's response to Snyder's work demonstrates the difference between existing disciplinary practice and our vision of integrative pluralism. Realists, basing their approach on realism's negative heuristic confining theorising to material factors, responded to Snyder's and some other, similar studies to launch a fruitless search for the influences of the material ‘offense-defense’ balance.Footnote ⁶⁶ This ignored Snyder's central point that what mattered was not whether military technology made the offense stronger than the defence, but whether psychological and organisational factors caused military planners to believe that it did. Snyder's book title is, The Ideology of the Offensive. If they had followed the practice of integrative pluralism, Snyder's realist colleagues would have accepted the fact of institutional and psychological effects on military planning, and then reflected on how other areas of policymaking may be similarly influenced.Footnote ⁶⁷ The result might have been an early start on a three-way synthesis of neorealist, neoliberal, and political psychological insights in explaining international security.

While a few scholars made efforts in this direction, the discipline as a whole did not. One distinguished example of such efforts is the research programme connected with Peter Katzenstein's account of security policy in postwar Japan. Indeed, Katzenstein explicitly makes part of the integrative pluralist case, countering the simplistic ‘losers become pacifists’ argument by emphasising the need to ‘specify the political mechanisms by which [Japan's] purported transformation [to pacifist values] took place’.Footnote ⁶⁸ Katzenstein's explanation focuses on a combination of normative and institutional factors – that is, it draws from both constructivism and liberal institutionalism. The causes are political and social, including informal networking processes that generate the social consensus so highly prized in Japanese social life. Bureaucrats, politicians, businessmen, and journalists’ informal contacts help them to work out the meanings of social norms – most importantly, in this case, those of Japan's ‘peace constitution’. Constitutive norms are then institutionalised into the structure of the Japanese state to generate the striking continuity of Japan's relatively demilitarised security policy. The contrast with pre-Second World War security policy is explained in part by the difference in institutions, especially the removal of the rule that gave the army a veto power over the formation of prewar Japanese governments.

Notably, though Katzenstein has explicitly endorsed analytical eclecticism,Footnote ⁶⁹ his work also points the way towards the more ambitious goal of theoretical integration advocated by integrative pluralism. He argues that the ‘simple and all-encompassing explanations’ generated by realism and liberalism are simply ‘wrong’ about Japanese security policy. Instead of simply junking the ‘isms’, however, he argues for integrating them with the ‘institutional perspective’ laid out in his book: essentially, liberalism's rationalism provides a ‘theory of action’, while his norms-plus-institutions account provides an equally necessary ‘theory of interests’.Footnote ⁷⁰ Logically, it makes perfect sense to combine constructivist theorising about interest and identity formation with a rationalist (realist and/or liberal) theory of action. The main barrier to integrating Katzenstein's constructivist institutionalism with, say, Keohane's liberal institutionalism or even Waltzian neorealism at the grand theoretical level is the discipline's conviction that doing so is somehow inappropriate.Footnote ⁷¹

Turning to more explicitly critical scholarship, Laura Sjoberg's Gendering Global Conflict: Towards a Feminist Theory of War offers a different approach to pursuing theoretical integration. Ironically, Sjoberg disavows the goal of a convergence of views; she advocates instead a ‘dialogic approach’ based on the understanding that disagreement ‘is the substance of feminisms’.Footnote ⁷² Despite that starting point, however, Sjoberg does a great deal to show how various feminist insights can be not merely put in dialogue with ‘mainstream’ approaches to international relations, but can be used to modify them.

Perhaps most interesting is Sjoberg's extension of feminist theory to theorising about the international system as a whole. The key point is that the structure of the international system is characterised by gender hierarchy, with the patriarchal structure serving as an ordering principle; in this system, units are positioned as either dominant or subordinate.Footnote ⁷³ Logical consequences follow. First, ‘we can expect states … to see dominance as a goal’, as this is what masculinist norms demand. Furthermore, because power is defined in gendered terms as ‘power-over’, it follows that ‘the accumulation of power is necessarily competitive and zero sum, making conflict likely, if not necessary.’ Also explicable is state pursuit of a masculine understanding of honour, as illustrated by Belgium's self-destructive decision to resist the German invasion of 1914.Footnote ⁷⁴ Thus from the one assumption of patriarchal structure, Sjoberg derives expectations of dominance-seeking, power-seeking, and honour-seeking state behaviour – a picture that is arguably both more logically sound and more descriptively accurate than Waltz's attempt to deduce state goals from the assumption of international anarchy. From our perspective, Sjoberg's intervention is a first-rate illustration of the second step of integrative pluralism's two-step process of theory-building: taking the results of existing middle-range theorising to reflect on the implications for grand theorising, both within existing paradigms and beyond.

Pushing furthest on the scholarly evidence that integrative pluralism may be something the ‘nature of nature demands’, it is even possible to see it at work in poststructuralist scholarship that is often seen as incompatible with explanatory social science. In one of the clearest, most theoretically comprehensive, and best executed poststructuralist works of the past two decades, Lene Hansen's Security as Practice presents poststructuralism in a way that remains open to the integrative pluralist impulse.Footnote ⁷⁵ The foundation for an integrative pluralist reading of this work is in the clarifying presentation of poststructuralism's emphasis on discourse.

Hansen is at pains to emphasise the ‘material aspect of poststructuralism's conception of discourse’, and the way that ‘objective’ material facts can be ‘incorporated’ into analysis by way of the discursive presentation of them. What's more, like integrative pluralism generally the book employs a baseline ontological conception that shapes the epistemological concentration on discourse as a form of political practice. Hansen starts from the assumption that identity is relationally constructed, and at once shapes and is shaped through practice.Footnote ⁷⁶ Hansen does argue that poststructuralism precludes causal analysis, but her conception of causality is that of the neopositivist vision of distinct dependent and independent variables rather than the causal complexes view offered above. The notion that ‘identities are simultaneously a product of and the justification for foreign policy’ sits comfortably within an integrative pluralist approach. Indeed, Hansen articulates a poststructuralist version of concrete explanatory integration, whereby the impact of a particular foreign policy discourse is gauged not only by a deep analysis of its internal content, but by assessing the role of ‘the broader social and political context within which it is situated’. The main difference from how less critically-oriented scholars might approach the integration of various causal factors is that for Hansen the assessment of them is accessed largely through their discursive representations, rather than positing some objective sense of ‘military capabilities, institutional pressures’, and so on.Footnote ⁷⁷

Hansen applies this foundation to explain key aspects of the early to mid-1990s Bosnian War, adeptly integrating various causes that, abstracted from their representations in discourse, can be found in various theoretical traditions. For instance, as Hansen traces the Western policy discourses on Bosnia from 1992–5 we are also given a rich explanation of what amounted to often confused and counterproductive actions. The standard telling of this period usually emphasises how Europeans viewed the conflict as a civil war to simply be ended, no matter the perceived winners or losers, and the US viewed it as a case of Serb aggression against Bosnian Muslims. European peacekeeping could not translate into an effective peace plan without US support, and US bombing of Serb targets could not occur with European peacekeepers on the ground. In Hansen's analysis, this stalemate takes the form of competing discourses and their ability to, within a particular context of other forces, maintain a stable balance between identity and policy. The European position is shown to be discursively effective because it articulated a sense of European responsibility for depoliticised civilian victims, and laid responsibility for the conflict simply on ‘Balkan leaders’. Its identity / policy construction was ‘remarkably resilient’ because it simplistically explained each new atrocity as just something that happens in the Balkans, and delegitimised US plans for airstrikes that might endanger ‘civilians’ on all sides – all while maintaining wide support among European publics. In the American context, this discourse competed with a ‘Genocide’ frame that demanded immediate action, and confronted deep-seated aversion to putting ground troops at risk. The president justified the ‘stay out’ policy using the same ‘ancient hatreds’ aspect of the ‘Balkan’ discourse adopted by Europeans. Though Hansen poses her analysis in terms of discursive constructions, it can also be understood in terms of concepts of domestic and alliance politics, social psychology, human rights norms, and the uncertainty induced by a unipolar international system.Footnote ⁷⁸

Obviously, more needs to be said about how research like this should serve to inform the broader IR theory landscape. Indeed, the nuances in the explanations above, reflective as they are of the complexity of our world, make it difficult to develop any standard guide or set of principles for doing so. Indeed, from an integrative pluralist perspective it would be inappropriate to do so. The best we can offer is a more in-depth illustration of how this can work, for which we turn now to a more extended discussion of the development of another example of integrative pluralism in operation – one that is useful because it speaks fairly directly to the major theoretical traditions in security studies and IR: Stuart Kaufman's symbolic politics theory.

Realising integrative pluralism in security studies (and IR): The illustrative example of symbolic politics theory

Symbolic politics theory offers an explanation of some of the core problems in the field of security studies – and indeed of IR more broadly – namely, the amount of conflict or cooperation among ethnic political communities. It is thus a particularly useful illustrative example of the potential for integrative pluralism in the field. Specifically, it explains why contacts among ethnic groups may lead either to a ‘politics of protection’ leading to war, a ‘politics of redistribution’ leading to contention, a cooperative ‘politics of distribution’, or a ‘politics of submission’ leading to repressive peace.Footnote ⁷⁹ The core difference between ethnic conflict and IR is that the political communities in the former case are not all legally sovereign. While some theorists may argue that this factor excludes ethnic conflict from consideration as an IR phenomenon, we contend the opposite: the parties to ethnic conflict should be seen as acting out the varied processes of sovereignty, such as formation and solidification, which mark the core reality of international politics. Some parties actually are ‘sovereign’ in a formal sense, others seek to become so, while substantive sovereignty can vary in either case. Likewise, the power disparities and myriad sources of power on display mirror the complexities most IR scholars seem to agree must be accounted for in explaining international politics. Indeed, it has already been extended towards explaining international conflict and cooperation.Footnote ⁸⁰ Moreover, this focus speaks to IR theory's need to account not only for international conflict but also for phenomena like civil war, the overwhelmingly most common form of war in recent decades.Footnote ⁸¹

Symbolic politics theory is explicitly multidisciplinary and transtheoretical. It starts by transcending the divide between the individual and social ‘levels of analysis’ by identifying the mechanisms that convert individual attitudes and emotions into social action. To do so, symbolic politics theory draws explicitly upon and links the implications of findings from neuroscience, psychology, anthropology, and sociology as well as political science and IR. One effect of this move is to follow the advice of critical theory – and of much recent theorising on nationalism – and problematise the nation,Footnote ⁸² asking when and why individuals with common ethnocultural characteristics sometimes politicise their common identity, mobilising to make nationalist demands about sovereignty or political power. Considered from a different point of view, this is at the same time an account of how ethnic groups solve the collective action problem, or fail to do so. Critically, explaining the transition to collective behaviour is not only about individual attitudes and behaviour, but – drawing on mobilisation theory – even more about social ties, networks, and organisations, as well as the role of social agents in framing messages, activating social networks, and brokering ties across disparate networks.Footnote ⁸³ The causes, in short, are not only psychological; they are also relational and intersubjective.

Why so complicated a story? What is wrong with parsimony? The short answer is that the parsimonious explanation was inadequate. Kaufman's first significant work on ethnic conflict attempted to apply game theory to explain the outbreak of the Yugoslav civil war, but found that all of the explanatory power came in the process of preference formation; strategic interaction as game theory conceptualises it played very little role.Footnote ⁸⁴ In Modern Hatreds, Kaufman therefore turned to explaining ethnic groups’ preferences, focusing on the importance of nationalist myths and fears, and on the roles of demagogic leaders in appealing to those myths and stoking those fears.Footnote ⁸⁵ Strategic interaction – the security dilemma – is still there, but Kaufman sided with Jack Snyder and Robert Jervis in identifying ethnic security dilemmas as being typically driven by predatory motives rather than uncertainty-under-anarchy.Footnote ⁸⁶

The result is a theory drawing a bit from realism (security dilemma dynamics), a bit from constructivism (leader discourse working within existing narrative structures), and a bit from political psychology (the fears and ‘modern hatreds’ of the title). Yet this model, too, proved inadequate: if it explained the whole picture, for example, then those who mobilise for ethnic conflict would be those most strongly motivated by emotions such as hate and fear. But this is not the case. In Nationalist Passions, therefore, Kaufman introduced social organisation as one of the key mechanisms in the latest iteration of the theory.Footnote ⁸⁷ Ethnonationalist conflict, Kaufman found, simply cannot be explained without attention to nationalist narratives, individual predispositions and threat perceptions, social organisation, and the way leaders frame issues.

This and rather than or feature of the theory is the point. To start with, it builds on the classic psychological view that predispositions such as prejudice result from socialisation – for example, from exposure to narratives of group or national identity.Footnote ⁸⁸ Psychological predispositions are thus in part the results of processes emphasised by constructivists, but once inculcated they influence how individuals respond to future discourses. As a result the predispositions particularly relevant for understanding ethnic (and international) conflict dynamics are not conceived as fixed conditions for all humans. Rather, they vary systematically across individuals and populations, and this variation has been shown to have specific causal implications. They also change over time in response to social influences. The bottom line is that Kaufman found that a superior explanation of ethnic conflict required a prominent place for both predispositions and discourses, and that as such he was asserting their ontological status as real and, depending on the context, causally efficacious.

This integrated theory also puts its various causes into motion, allowing it to deal with the complexity of different iterations of ethnic conflict, and setting it up to speak back to the broader IR theory landscape. For instance, predispositions are only ever actualised in a social environment replete with other causal forces. Among the most important contextual factors are threat perceptions or ‘fears’, which work to activate and reinforce predispositions such as prejudice. Predispositions and fears, however, do not determine individual behaviour; still less do they determine collective behaviour of social groups or states. Turning attitudes into action requires mobilisation theory's insights into the key factors that drive social mobilisation, especially organisation (including formal institutions and informal networks) and framing. A key causal mechanism in this linkage is the way leaders use framing to mobilise followers by appealing to their predispositions and fears.Footnote ⁸⁹ The causal action is in the interaction between factors at different levels of analysis.

To summarise, the most important causes posited by symbolic politics theory to explain ethnic conflict are as follows:Footnote ⁹⁰

1. 1. Socialisation: group narratives shape symbolic predispositions; hostile group narratives promote prejudice.

2. 2. Discrimination: symbolic predispositions such as prejudice incline prejudiced individuals to respond with hostile feelings to outgroup behaviour.

3. 3. Provocation: perceived threats induce individuals to become more hostile to the source of the perceived threat.

4. 4. Framing: leaders use frames to appeals to predispositions such as prejudice and to stoke emotions such as fear to build political support for aggressive action against the outgroup.

5. 5. Networking: leaders use their social ties to appropriate institutional structures to mobilise followers for the leader's preferred action.

These causes are not merely additive but mutually interactive. Prejudice and fear reinforce each other. Aggressive nationalist framing not only benefits from fear; it also feeds it. The resonance of the frame enhances organisational strength, but at the same time organisational strength is essential to turn a popular cause into a popular movement. Finally, over the long run, framing either alters or reproduces the group narratives that shape popular symbolic predispositions.

It is also important to note that from the perspective of the original underlying theories, each of these causes is in fact a synthesis of more than one related process. This synthesis allows symbolic politics theory to reflect back constructively on these underlying theories. Discrimination, for example, is a broad mechanism that can refer to any kind of cognitive or motivated bias. In explaining ethnic conflict, the two most important underlying theories are Social Identity theory (SIT) and prejudice theory. SIT posits that individuals have a mild bias in favor of virtually any in-group of which they are a member.Footnote ⁹¹ Prejudice theory explains the stronger tendency of individuals to discriminate against members of those groups they particularly dislike.Footnote ⁹² Symbolic politics theory suggests that both processes may be operating at the same time.

Because of its basis in multiple theoretical traditions, symbolic politics theory can also draw from some of them to suggest alterations in others. In framing theory, for example, Amos Tversky and Daniel Kahneman originally posited a purely cognitive mechanism for its operation, and much political psychological literature on framing has kept this original understanding.Footnote ⁹³ Kaufman's findings show, however, that a framing theory that includes an emotional component has more explanatory power in accounting for political behaviour.Footnote ⁹⁴ Indeed, the emotional interpretation explains cognitive failures of the sort Tversky and Kahneman first identified. This is not to suggest the original understanding of framing theory was inaccurate. But we can say that the inclusion of emotional mechanisms allows our explanation to go deeper – a position also compatible with Kahneman's later work.Footnote ⁹⁵

Mobilisation theory, too, might benefit from a dose of the psychological insights deployed in symbolic politics theory. Making this move is admittedly controversial: Charles Tilly, for example, ‘rebels’ against ‘the claim that individual and collective dispositions explain social processes’.Footnote ⁹⁶ That position may be pragmatically useful, but attention to the partially socially constructed nature of interests and identities does not negate the evidence for psychological sources of motivation. At the concrete empirical level, Kaufman's findings show that both sets of factors matter. Within cases, as in Rwanda, it is indeed primarily social ties that determine who mobilises.Footnote ⁹⁷ Between cases, however, the evidence shows that such mobilisation is only likely where predispositions and threat perceptions strongly motivate it. Attention to predispositions can help mobilisation theorists understand where mobilisation is likely to occur.

The necessity of thinking about these causes as interactive is also at the heart of the way we might conceive of broader IR theories or paradigms and the relationship of symbolic politics to them. For IR realists or liberals, for instance, the message is that their logics are contingently correct. The central mechanism of realism is provocation – the tendency of agents to respond aggressively to threat. However, as symbolic politics theory demonstrates, individuals with different predispositions respond variably to threat, causing them to follow different strategic logics.Footnote ⁹⁸ Liberalism, in contrast to realism, focuses on the mechanism of networking, elaborating the ways in which institutions – either state institutions or international ones – channel international behaviour in ways that may be cooperative or conflictual. Some liberals add the influence of ‘ideas’ – that is, the mechanisms of socialisation and discrimination.Footnote ⁹⁹ Looked at this way, the ‘neo-neo synthesis’ is a theoretical enterprise that combines at least four of the five causes highlighted by symbolic politics theory – only framing is mostly absent – though they are conceived as purely rational in nature, with their emotional power stripped out.Footnote ¹⁰⁰

On a more positive note, the success of symbolic politics theory suggests that social mobilisation theory's cousin, relational theory, might provide an especially promising perspective for theorising the international system. In an important intervention, Patrick Jackson and Daniel Nexon argued that states should be conceived as simply bundles of social relations, as is the international system itself.Footnote ¹⁰¹ Thus social ties constitute the international system in both senses: not only do social ties make the international system – both those that form the state and those that tie states together – but collectively these social ties are the international system. The implications of these insights are just starting to be explored.Footnote ¹⁰² The key argument is that social ties have causal power, primarily by serving as pathways for potential cooperation or conflict management as well as for competition or conflict. Hypothetically, therefore, the denser the web of social ties binding together any set of organised groups in the international system, the greater the prospects for cooperation.Footnote ¹⁰³

Of course, IR theory should not be limited to the causal complex that constitutes symbolic politics theory. For example, while states are not very much like firms in a marketplace, market mechanisms are real and important in IR.Footnote ¹⁰⁴ Positioned practices, though dependent on other mechanisms such as socialisation, may also be seen as emergent phenomena with causal powers of their own. The broader message is that IR theory should focus on the interplay between processes that may be seen as operating at different levels of analysis. Just as in physics light must be understood as both a particle and a wave, a symbolic predisposition may be simultaneously an individual psychological attitude, an instantiation of a socially constructed intersubjective meaning, and an expression of Gramscian cultural hegemony.

Much like the domain that inspired the integrative pluralist vision for science, the picture of the international system that emerges from this understanding is analogous to the interaction of organisms in an ecosystem. Like an ecosystem, the international system is founded upon the interaction of living things with each other and with their natural environment, and it can only be understood by recognising the multiplicity of beasts that inhabit it. These creatures, furthermore, relate to each other in a wide variety of ways, including predator-prey, symbiotic, reproductive, and herding relations. Analogously, the international system is home to a diverse menagerie of creatures, from al-Qaeda to Google, the United Nations and Zambia. In their interactions, these international actors variously come into existence, cooperate, compete, fight, destroy or swallow rivals, and die.