494327 EJT19310.1177/1354066113494327European Journal of International RelationsGuzzini 2013 Article The ends of International Relations theory: Stages of reflexivity and modes of theorizing EJIR European Journal of International Relations 19(3) 521–541 © The Author(s) 2013 Reprints and permissions: sagepub.co.uk/journalsPermissions.nav DOI: 10.1177/1354066113494327 ejt.sagepub.com Stefano Guzzini Danish Institute for International Studies, Denmark Uppsala University, Sweden Abstract International Relations theory is being squeezed between two sides. On the one hand, the world of practitioners and attached experts often perceive International Relations theory as misleading if it does not correspond to practical knowledge, and redundant when it does. The academic study of international relations can and should not be anything beyond the capacity to provide political judgement which comes through reflection on the historical experience of practitioners. On the other hand, and within its disciplinary confines, International Relations theory is reduced to a particular type of empirical theory with increasing resistance to further self-reflection. Instead, this article argues that neither reduction is viable. Reducing theory to practical knowledge runs into self-contradictions; reducing theorizing to its empirical mode underestimates the constitutive function of theories, the role of concepts, and hence the variety of necessary modes of theorizing. I present this twofold claim in steps of increasing reflexivity in International Relations theory and propose four modes of theorizing: normative, metatheoretical, ontological/constitutive and empirical. Keywords conceptual analysis, intellectual history, methodological pluralism, nature of theory in the social sciences, Norbert Elias, sociology of International Relations Introduction Scholars of International Relations (IR) theory find themselves in a paradoxical position. On the one hand, being a theorist constitutes a core quality of the accomplished scholar, Corresponding author: Stefano Guzzini, Danish Institute for International Studies, Østbanegade 117, 2100 Copenhagen, Denmark & Dept of Government, Uppsala University, PO Box 514, 75120 Uppsala, Sweden. Email: sgu@diis.dk 522 European Journal of International Relations 19(3) at least to judge from our surveys. And yet, working with one of the isms has become considered outmoded, ‘been there’, if not outright harmful (Lake, 2011). Whereas theorists are tied to isms and their endless debates, research should move forward to get things done. More fundamentally, whereas in the past great theoretical debates provided the core of the discipline of IR, offering a common reference point in an otherwise everexpanding field of study, and a shared language with the world of practice (realism– idealism), this has ceased to be the case; the continuous bickering and flag-waving having in fact become an obstacle to it. The core comes now more by stealth through complete absorption, some say ‘normalization’, into (US) Political Science, where IR is simply the ‘external’ facet of public policy studies (for an early statement, see Milner, 1998). This article argues that this vision of theory is unwarranted and a consequence of failing to see the multiple modes of theorizing IR. There is much more to theorizing than isms. I will introduce four modes of theorizing: normative theorizing, meta-theorizing, constitutive/ontological theorizing and empirical theorizing. Those four modes of theorizing are neither unknown nor new. Yet, or so my argument goes, they have been neglected, if not cut out by a scissor movement coming from two opposite directions. First, it is common to hear that all the abstract language of academia (whether meta-theory or formal modelling and mathematization) has alienated IR from the world of practice. In its strongest version, this view claims that the ‘real’, or at least only relevant, knowledge is what has come down to us over centuries of practical selfreflection and political judgement. Being authorized to theorize world politics is something for which scholars have had to struggle by continuously challenging the allegedly superior knowledge of the practitioner. Yet, showing the distinctiveness of scientific knowledge from practical knowledge is constantly undermined by the sheer closeness to the field of political practitioners in whose language the analysis of world politics is authoritatively spoken, and by the socialization of analysts through (some) professional schools, attracting students who, in the school’s rite of passage to become leaders in their society, are often encouraged to belittle, if not de facto neglect, the ‘academic’ (i.e. useless) approach to knowledge. Second, even within their field, theorists find themselves under attack — at least some kinds of theorists. With the professionalization of the discipline, standard measures of ‘quality’ have produced a sometimes welcome, sometimes depressing, homogenization of research. Ever-increasing numbers of graduate students are educated (better than: ‘trained’) according to a ‘quantitative-followed-up-by-qualitative’ (meaning small-n) research design in which all there is to theory is reached when some robust empirical generalizations can be made under specified scope conditions. The article will address this dual challenge to IR theory, being squeezed out between practical knowledge and a specific version of empirical theory. To make my point, I will present the development of IR theory as historical steps of increasing reflexivity which cannot be undone. In the first section, I will try to show that there is no return to mere practical knowledge as all there is to theorizing IR, although I will later also insist on the need to understand that very practical knowledge as part of successful theorizing. I will present my argument by theorizing on the origins of the field of international expertise, analysing the way practical knowledge, and a specific one informed by the raison d’état, became the habitus of Court Aristocracy and later the diplomatic community. Western Guzzini 523 IR1 as field of study emerged not as a response to societal changes, as did other fields of systematic inquiry. In the early days, the discipline was not there to produce knowledge; already-existing (practical) knowledge produced its discipline. But, as my discussion of the early writings of Morgenthau which concludes the first section demonstrates, the resulting attempt to square the circle of practical knowledge in a scientific environment did not work. Although this argument supports the need to find a reflexive distance to the level of political action, and indeed a different language from the one in which world politics is spoken, I will then develop in the second section the need to further our understanding of different and equally important modes of theorizing which are not reducible to empirical generalizations. As Anna Leander (2011) so succinctly put it, we have to think of theorizing not as producing cookbooks, but instead as writing unfinished dictionaries, inside which a growing number of terms are in need of being continuously updated, in themselves and in their relation to each other. With this understanding of the different modes and ends of theorizing, the present state of IR can also be reframed. The classical isms may indeed no longer be the natural core of the discipline of IR, but that does not at all exclude the fact that theorizing can still be. Rather than seeing IR increasingly absorbed into Political Science, IR is conceived in terms of global politics, subsuming domestic and comparative politics. That is again no news to many scholars in International Political Sociology or International Political Economy (IPE). With these ends in mind, there is no end in sight for IR theory. Theory? No theorizing needed! Practical knowledge and the self-reflection of world politics Whether or not a debate between ‘realists’ and ‘idealists’ (or any equivalent labels) really constituted a great debate in the early days of the field of inquiry, the dichotomy has been constitutive for much of the self-understanding of its practitioners. The dichotomy reflected the two fields where expertise in IR could be grounded, namely, in the military and in diplomacy, as well as in the resources on which they relied, the study of history and ‘politics’ on the one hand, and of (international) law on the other, even if this divide was never absolute. This dichotomy structured much of the way international politics was to be understood. The lessons of the two world wars were paradigmatic in this regard. Were deterrence and escalation the wrong strategy after Sarajevo, so was appeasement equally erroneous in the 1930s (for two clear analyses of this dual lesson, see Jervis, 1976; Wolfers, 1962). The discourse of world politics got locked into its central binary opposition which, importantly, traded on a confusion of observational theories and foreign policy strategies. Here, ‘hawks’ are the ‘realist’ defenders of deterrence or containment and ‘doves’ the ‘idealist’/liberal proponents of engagement. This move had the pernicious effect of confusing description and explanation (Wendt, 1995): when things turned violent, this constituted both a hawkish description and validating proof for realism; 524 European Journal of International Relations 19(3) whereas when diplomacy succeeded, doves or idealism would score. But surely, ‘doves’ had ways to explain war, just as much as ‘hawks’ had ways to understand peaceful conflict resolution. The confusion between theory and (foreign policy) strategy, between explanation from a distance and maxims for action, is rarely found in other social sciences. That it regularly happens in the expert debate on international relations has a good reason: this binary simplification does not originate in the field of science, although it is reproduced there. Its origin is in the discourse of world diplomacy. Its conduit is the classical debate between political realism and idealism, and its conception of the superiority of practical knowledge. Whereas most social sciences are born out of the attempt of societies to reflect on and act upon their increasing differentiation and the development of the state with all its emerging functions, there is something peculiar to the self-reflection that led to IR as a discipline. Here, it is not with the distant view of science that social and political practice is improved; it is rather the other way round: it is through recourse to the lessons of practice that science is constituted. If the evolution of societies had made science necessary — for knowledge, control and for the legitimacy of rule — then the latecoming discipline of IR was to become the necessary detour to convince the new and enlarged world diplomatic society about, and thus preserve, the already-existing practical knowledge of its diplomatic and military elite. Science did not turn against tradition; tradition fitted itself into its science. The discipline was not there to produce knowledge; knowledge produced its discipline. This section will deal with the implications of this specific origin for theorizing in IR. Since my aim is to show the squeeze from practical knowledge in which IR theorizing finds itself, I will concentrate the discussion on the diplomatic lineage in the spirit of the raison d’état, and not on international law (although some legal positivists could easily join here).2 In the first step, I show the sociological and ideational underpinnings, that is, the field and epistemic habitus of that quite specific group which has come to define what European diplomacy talks and thinks: the ‘International’ of Court Aristocracy. In the second step, it will introduce how observers have made this view their own, how they have put this practical knowledge into the service of defining the specificity of the human/social sciences. It will finally indicate how that first-order reflexivity has become a crucial part of the early definition of IR, although it cannot overcome its internal tensions. For the sake of simplicity, I will present this through the work of Norbert Elias and a critique of Hans J. Morgenthau, respectively, passing quickly via Friedrich Meinecke. The remote origins of IR and the habitus of (absolutist) Court politics Norbert Elias has retraced the field and habitus of French Court Aristocracy in a way which sheds light on the identity of European diplomacy, its being a community in the first place, its way of seeing and doing things, and its shared practical knowledge.3 Elias’s argument is relevant in two ways. First, that Court Aristocracy is the bearer of a certain behavioural canon (what he calls a habitus) which survived stronger than anywhere else — indeed, almost in distilled form — in the diplomatic or foreign services of the Court. Second, even when the middle classes eventually took over those positions (by the early 20th century), Guzzini 525 they did not import their own canon, but adopted the existing pre-revolutionary aristocratic one, with some adaptations due to the ‘nationalization’ of politics (see also the discussion in Lebow, 2008: ch. 7). To start with the creation of an aristocratic norm-canon of diplomacy (‘the sociogenesis of norms’), Elias refers back to the establishment of a Court society, in his case, absolutist France. In the feudal bonds the aristocracy had to their rulers, their importance as knights and providers of armies guaranteed their autonomy. Yet, so Elias’s argument goes, rulers increasingly tried to rely on the lower classes for their armies. This change was accelerated by the shift in military technologies, where the advent of gunpowder and guns made the (aristocratic) cavalry an often powerless form of arms. Finally, with increasing financial needs, the Crown was incited to raise taxes, and/or to make public office available for money (Elias, 1969: 265–272). This development drove the old aristocracy into a dilemma between independence and prestige. If it kept going the old way, proud and autonomous from the Sovereign, its status was bound to decline, its privileges increasingly eroded. Worse, by being faithful to its own normative system, the aristocracy would run into debt to uphold its prestige, so fundamental to the habitus and hence recognition within its class, only to further precipitate impoverishment and loss of influence. If, however, they wished to uphold status and some influence, they would become increasingly dependent on the King and his largesse. The Court grew into a place where titles were sold, and favours offered, putting the King at the centre of a complex figuration of forces. Part of the aristocracy moved to the Court and stayed there permanently — the Versailles Court having at its maximum 10,000 residents according to one source in Elias. A new class, the Court Aristocracy, appeared. A new field of power was created, the Court. Politics in such a Court society were of a special manner. Lest one wished to forgo one’s status, it was an ‘inevitable’ field of social relations, both continuous in time and finite in scope. In such a world, any action, any word, fed immediately back into the totality of relations (Elias, 1969: 190). With no escape and high levels of interdependence, as Elias stresses, its members need advance carefully. This society taught an extreme control of personal feelings. Thoughts needed to be impenetrable; prudence was essential (Elias, 1969: 181 and 185, respectively). It also directed a total attention to things allegedly ‘external’, since any act in this web of relations could be read in terms of status, and hence even a slight shift in attention might indicate a sudden weakness (Elias, 1969: 179). The Court Aristocracy was obsessed with prestige, honour and status, and hence etiquette. Prestige was crucial for the very self-definition of Court Aristocrats; a threat to their honour would strike their identity as members of that society (Elias, 1969: 164). And so etiquette, far from being mere formality, became the self-representation of the Court society in which any shifts — subtle favours, possible slights — were registered (Elias, 1969: 174). At the Court, to employ a phrase used elsewhere, one had no friends, only interests. And so the Court society produced a habitus (and was reproduced by it) which was thoroughly based on self-interested action,4 which, in turn, was understood as always prudent, often duplicitous, potentially ruthless — although the very finiteness of the Court and the repetition of relations also put a break on such ruthlessness. Nothing was ever really forgotten. Consequently, so Elias writes, the habitus was what we now call 526 European Journal of International Relations 19(3) ‘diplomatic’ and has survived mainly in the foreign service, but increasingly also in the negotiation cultures of other (also private) international actors (Elias, 1969: 187). For, although the noblesse de robe increasingly took over major state functions (as judges, for instance), it did not touch the higher military and diplomatic services, in which part of the old noblesse d’épée was able to secure a monopoly position that kept its status intact (Elias, 1969: 325–326). Here, we find the undiluted habitus of the Court Aristocrat, the sociological base for what scholars in IR have come to call ‘diplomatic culture’, whose common idiom used to be French until the early 20th century. And the special knowledge of this ‘Aristocratic International’, as Morgenthau called it, was explicitly not reflexive or theoretical. Since the middle of the 18th century, the forms of knowledge considered relevant for the diplomatic service are to be found in memoirs, personal letters and correspondence, aphorisms and ‘maxims’ (like Rochefoucault’s), that is, the Court society’s ‘ongoing conversation’. Only with that form of practical knowledge, typical for the habitus of the Court Aristocrat, was a person considered suitable for the foreign service (Elias, 1969: 182, fn. 36 for the argument of this paragraph). With that special diplomatic community and its predispositions established, the second step in Elias’s argument is about the reproduction of its habitus. In the 19th century, social mobility put pressure on the relatively coherent social strata from which the diplomatic and military elite had conventionally been chosen. When the middle classes reached the leading positions in the state, their moral canon — either more strictly utilitarian and less prestige-oriented or more idealist — clashed with the aristocratic one. But, according to Elias, in particular in international affairs in which they had the least experience, and where the contradictions between their moral code and reigning realpolitik were greatest, the middle classes ended up simply adopting the maxims of the previous ruling groups. They took over the existing aristocratic habitus (Verhaltenskanon) ‘which, for no better word, one could call Machiavellian’, and which, in everyone’s unbridled pursuit of personal interests, had ‘left over a heritage of mistrust and fear from each other’ (Elias, 1989: 184). In this way the diplomatic culture reproduced not only itself, but also the international realm to which it supposedly was a practical answer. For it became a ‘self-perpetuating’ mechanism: the belief in the truth and inevitability of power politics is, according to Elias, itself one of the chief reasons for its perpetuation (Elias, 1989: 202). As such, the field can be considered remarkably stable. This middle classes’ adoption of the pre-existing maxims did, however, also lead to one significant adaptation. Power politics was no longer applied to dynastic relations between single sovereigns loyal to the(ir) state, but to relations in the name of sovereign collectives loyal to the(ir) nation (Elias, 1989: 185–189).5 This turned the relatively flexible postulate that the self-interest of a state is the last and decisive reference for action in international relations into ‘a categorical imperative, with deep roots not only in the emotions of the individuals, but also in their conscience, in their self- and we-image, their self- and weideal’ (Elias, 1989: 203). The aristocratic habitus of Court society thus became ‘the diplomatic culture’ characterized by self-interested prudence and nationalized power politics that was to socialize newcomers and rule international affairs at least until the early 20th century. All reflection needed was the ongoing conversation within the field. Guzzini 527 Interlude: Practical knowledge as an attempt to fuse political thought and history ‘wie es wirklich gewesen’ Friedrich Meinecke’s resurrection of the reason of state can serve as my bridge to contextualize the defence of practical knowledge even within a science. The tradition of the reason of state could form an important link, because it was both a foreign policy practice and the first-order reflection on that practice, that is, a reflection from within the field itself. It corresponds to the first wave of reflection which sets in during the Renaissance, in which previous social practices in foreign policy are explicitly framed, codified and in the process also to some extent erected as a model (or anti-model) (Elias, 1969: 408–409; 1989: 155, 179). Power politics à la Machiavelli was surely not something new to his age; but with the secularization of politics, its practice was explicitly reflected upon. In proposing this tradition in the interwar period, Meinecke attempts to ground practical knowledge as politics and vice versa. The basis would be a particular conception of historiography, departing from the spell of natural law and its concern with the ideal state, and thinking Staatskunst and History in direct parallel: Acting according to the reason of state reached relatively early a way of seeing and understanding which was akin to modern historical cognition. Modern historical cognition, in turn, profited also from the reason of state, from the attraction that emanated from the teaching of the interests of states, which was used as auxiliary practical science for the ‘art of government’ since the 17th century by those involved in the latter. (Meinecke, 1957 [1924/1929]: 22–23, my translation) Making the reason of state hence the privileged partner for establishing an empirical methodology for the history of ideas, such history also becomes the essence of and sedimented knowledge for the art of government. Statesmen and modern historians blend into each other in the quest to understand states and their interests in the motion of world history, an understanding acquired by looking through the eyes of the practitioner. The result is not universal knowledge, but practical ‘maxims’, which is the way Meinecke defined the reason of state. And since the emergence of modern diplomacy evolved in parallel with the discourse of the reason of state during the Renaissance (Meinecke, 1957 [1924/1929]: 176), these two discourses are of the same kind. Meinecke (1957 [1924/1929]: 100) sees in the teachings (Lehre) of the European balance of power nothing other than a detail of the general teachings (Lehre) of the reason of state. Yet, it is first and foremost that practical knowledge, and precisely not yet an attempt to turn it into a social science. Meinecke (1957 [1924/1929]: 174–175) is very explicit that a purely empirical and utilitarian study of the reason of state is necessarily limited, and general catalogues for the ideal behaviour of states are not possible. He ridicules the attempt to understand politics like ‘clock mechanics’ — and reads Hobbes in this tradition (Meinecke, 1957 [1924/1929]: 188). Indeed, Meinecke (1957 [1924/1929]: 245) says that by its very nature, a clear definition of the concept of reason of state is not possible. Nor is a calculus of the real interests of states always possible, since the dilemmas of political necessity escape a clear assessment and the interests are often 528 European Journal of International Relations 19(3) ambivalent themselves (Meinecke, 1957 [1924/1929]: 275). Looking at politics ‘as it really is’, Meinecke (1957 [1924/1929]: 165) sees the field of statesmanship like the classical world of tragedy: only with the sense of history and the experience of politics given by the reason of state can statesmen hope to acquire the art/craft of statesmanship (Staatskunst). It is a reflective knowledge, but ultimately cannot (and should not) be anything but practical. The impossibility of squaring the circle: Practical knowledge in a time of second-order reflection If Machiavelli stands for the self-awareness of a habitus, a first-order reflection, then the resurgence of the reason of state in the interwar period stands for the attempt to freeze both scientific knowledge and political practice at that level. A nascent Western discipline of IR could then have followed the way of other social sciences. If international relations had been conceived as an international society, one could have mobilized the scientific canons of Comte, Durkheim, Marx or Weber, who offered several matrixes for a scientific turn. But in IR, the call for the superior knowledge embedded in the practical tradition of the raison d’état was still alive. It was alive in the habitus of practitioners and now revived by the early epigones of the discipline. Yet the changing criteria for validity claims undermined the status of political judgement based on the sole appeal to experience and the lessons of history. This produced a strong tension, well illustrated by the early writings of Morgenthau. Here, Morgenthau (1946) showed himself in an uncomfortable legacy of Max Weber. But despite all his respect, in the end, Morgenthau could not follow Weber’s route to science. Theorizing as practical knowledge got stuck in a dilemma. Just like Weber, Morgenthau was exposed to three different traditions of theorization, one more normative, two more positive.6 They both rejected the classical normative tradition, too much concerned with what politics should be, rather than with what it actually was. In the positive tradition, they acknowledged the role of both the practical and ‘scientific’ tradition, nicely exemplified by Weber’s two famous lectures on ‘Politics as a Science and as a Vocation’ But whereas Weber ultimately saw a clear priority for sociology as a science, even if an interpretive one, Morgenthau was stuck with a priority for the practical tradition. This produces a certain dilemma. Weber could move to scientific justifications for his knowledge, Morgenthau’s Scientific Man vs Power Politics could do so only with qualifications. A first-order reflection gave way to a second-order reflection on that very core of knowledge that had been assembled and canonized beforehand. With the development of the social sciences as disciplines, an appeal to ‘historical experience’ is no strong argument. And so Morgenthau ended up making the central move: where Meinecke was still arguing against the very need for a ‘theory’ when claiming the superiority of practical knowledge (for both the historian and the practitioner), Morgenthau was trying to show that the maxims of practical knowledge are a scientific theory. From ‘no theory needed’, Morgenthau moved to ‘no new theory needed’. Morgenthau could declare that although Martin Wight’s (1966) ‘Why is there no international theory?’ was wrong (for Morgenthau, there is theory): Guzzini 529 Its fourteen pages contain more insights into the intellectual issues posed by theoretical concern with international relations than a whole shelf of books and articles that, following the fashion of the day, spin out theories of international relations and embark upon esoteric methodological studies on how to approach such theory-making. (Morgenthau, 1970 [1964]: 248) This quote clearly shows his two theoretical targets: the attempt to fashion theory in terms of testable models and the tendency to discuss methodology, which seems to include also theorizing in meta-theoretical terms (what he calls ‘theory-making’). Those misconceived ‘theories’, wrote Morgenthau, mistake ‘politics’ for something fully amenable to reason and measurement. The latter contradicts ‘the objective character of international relations’ and produces ‘dogmas’, a ‘kind of metaphysics, regardless in what empirical or mathematical garb it is clothed’ (Morgenthau, 1970 [1967]: 242, 243 and 246, respectively). By not talking truth to power in any relevant way, such theoretical research only serves to bolster the status quo (Morgenthau, 1970 [1967]: 247) and ultimately only serves the narrow interests and psychological self-satisfaction of scholars who do not dare to make statements whose closer contact with ‘political reality’ could disconfirm them (Morgenthau, 1970 [1964]: 261). Hence, Morgenthau’s aim was twofold. Against Wight, he wished to defend the possibility of an international theory, if redefined. Against behaviouralism (which he connected pêle-mêle with economic approaches, liberalism, utopianism), he wished to define its necessarily limited character. But when Morgenthau sketches the nature of possible IR theory, the argument becomes ultimately circular. For Morgenthau, it is a theory which goes beyond a philosophy of history, in that it makes explicit the theoretical assumptions upon which philosophically inclined historians (he refers to Thucydides and Ranke) have made their analysis, and then uses history to ‘demonstrate’ their validity (Morgenthau, 1970 [1964]: 251, referring to the same body of inspiration as Meinecke). Since politics is both contingent and rational, the limited ‘rational element in political action makes politics susceptible to theoretical analysis’, whereas, to conclude on a comfortably ambivalent position, its contingent element ‘obviates the possibility of theoretical understanding’ (Morgenthau, 1970 [1964]: 254). ‘Within these limits, a theory of international relations performs the function any theory performs, that is, to bring order and meaning into a mass of unconnected material and to increase knowledge through the logical development of certain propositions empirically established’ (Morgenthau, 1970 [1964]: 257). Indeed, in a suddenly arch-positivist answer to Wight’s critique that whereas domestic theory is about progress (and hence amenable to theory), IR is the realm of recurrence and repetition, Morgenthau wrote that it is precisely ‘this repetitive character of international politics, that is, the configurations of the balance of power, that lends itself to theoretical systematisation’ (1970 [1964]: 251). Theory meets geopolitics here, understood as the objectivist component of realist theorizing. To make this work, the raison d’état is not only in historical experience, but also in the nature of things. Hence, Morgenthau looked for regularities that can be empirically established and historically demonstrated and found them in the classical balance of power politics. But then why would he oppose the attempt to systematically test those regularities in controlled and often quantitative studies? He did see that any historical explanation necessarily relies upon theoretical assumptions which need to be made explicit. But then why 530 European Journal of International Relations 19(3) would he believe theoretical critiques of other theories to be useless, when they perform precisely that exposition and discussion of underlying assumptions? If empirical regularities cannot be established in a quantitative way because no historical case is really like any other, then also his own theory be established or justified, since the nature, nay, the existence, of the alleged regularities is then at stake. If, on the other hand, making theoretical assumptions explicit is the work of theorizing, then it is an unavoidable and fundamental part of IR theory, not idle self-centred talk. By denying a stronger role to theory, by not going the whole Weberian way, his approach falls back on positivist positions malgré lui and cannot defend itself well against the behaviouralist attack. As a result, and as shown elsewhere (Guzzini, 1998; 2004: 534–535, 546–548), practical knowledge when moving from a first- to a second-order reflexivity remains caught in the ‘conservative’ or justification/tradition dilemma, so named after Kissinger’s (1957: ch. XI) analysis of Metternich. Its classical defence no longer applies: it cannot just refer to the world ‘as it is’ and rely on its practical understanding by the responsible elites. But if it then defends itself in a ‘theoretical, that is, an objective, systematic manner’ (Morgenthau, 1970 [1967]: 254), practical knowledge at this level of reflexivity has no choice but to engage the scientific canon of the day, which it had left to others to define. Not redefining the core of theorizing itself, it ends up in a no-win situation: being consistent with itself, it should avoid a scientific defence, but that will no longer do; by attempting that scientific defence in positive and positivistic language (‘as it is’), practical knowledge does not, however, stay consistent with itself. This position always comes back to say that, although theory is needed, there is really nothing new under the sun; some amendments and systematization of the existing first-level reflection of the reason of state will do, as exemplified by the different developments of balance of power theory. It is a remarkable logical circle: since international relations are all recurrence and repetition, as Wight said, so are our theories; and so they must be according to theory as geopolitics, because so is ‘reality’. The logic of reality is said to impose practical knowledge, when it is practical knowledge that constructs this logic of reality in the first place without allowing a reflective distance to its own construction. In this circular move, the end(s) of theorizing are also the end of it, since no knowledge which goes beyond and against the nature of international politics, as asserted by Morgenthau, is tenable. In this view, any theory beyond is but the personification of the liberal rationalist hubris, any further debate about ‘theory-making’ only irresponsible scholastic narcissism. But however well this still resonates in policy circles within and outside academia, it gives no good reason to halt further theorizing. Which theory? Which theorizing? Scientific knowledge and the self-reflection of world political observation The role of IR theory can only be appreciated if we rid ourselves of two reductionisms. The first, mentioned above, concerns the confusion between practical and scientific knowledge. The second, and theme of this section, is the reduction of scientific knowledge to a narrow version of empirical theory. This section will argue that both Guzzini 531 movements together crowd out most of what theorizing is all about. Theory is not only the result of knowledge, but also the condition for the possibility of knowledge. Theorizing must cover both aspects, and the fact that not all knowledge is empirically determined. In fact, this section argues that it is better to confront and analyse the ubiquity of theoretical presuppositions than to exogenize them. In many contemporary research designs, theory is either the result of the study (the empirical generalization) or its given and external starting point (if that generalization informs a proposition put to a test). Theory, thus, is external to research design and divorced from methodology. But theory is not there only before or after — it never leaves us, as seen also by positivist philosophers of science. The argument proceeds in three steps. First, I will try to rescue an understanding of the Inter-Paradigm Debate (IPD) which, although shared by its proponents in the 1980s and later, seems to have been forgotten or reinterpreted, at least in parts of our field. Indeed, the IPD does not necessarily have to stand for stale debates, or for increasingly bitter turf wars, between realism, idealism and whatever other ism. It opened a stage not only for a second-order reflection on world politics at the level of the observer, but for its self-reflection and theoretical control by probing the assumptions of such world political observations. Reminding us of that space opened by the IPD, but unpacking the bundled paradigms, I propose four modes of theorizing which characterize our, and I guess any, field in the social sciences. Taking these modes of theorizing seriously asks, in turn, that we conceive of our core scientific communication as organized around concepts used for updating our knowledge in ever ‘unfinished dictionaries’ of the international. Revisiting the ‘Third Debate’7 In many corners of the discipline, the IPD has been shelved and is relegated to standard textbook presentations. But in my understanding, this follows from a narrow and impoverished form of the debate in which we simply throw isms at each other. Worse, the isms have been at times read in terms of ideologies, where IR theories would have to fit into conservatism, liberalism and whatever stood for radicalism. This move to reduce paradigms to ideologies is, however, telling. In a scientific field attuned to the idea that there can ultimately be only one truth, a plurality of equally justifiable theories is not easy to accept. In this understanding, an IPD, wherever it occurred, could only be a passing moment, eventually resolved by having a winner.8 If it dragged on, some scholars had to be either of inferior intellect or possibly dogmatic due to some normative commitment. When reason could not decide, it had to be the fault of values — and the longer the pluralism of theories lasted, the more bitter the fight, the more fenced the turf. Quite understandably, scholars would ask for a kind of truce in the form of eclecticism (Sil and Katzenstein, 2010) to move beyond such theory wars. Yet, this is not at all how the IPD (Banks, 1985; Holsti, 1985) or also the ‘Third Debate’ (Lapid, 1989a, 1989b) was seen by its proponents. They saw in it a liberating movement and an invitation to a healthy pluralism. Theorizing in IR had reached a deadlock, because the existing ways to conceive world politics — both the nature of politics 532 European Journal of International Relations 19(3) and what counts as the world — needed to ‘come to terms’ with social and historical changes. For instance, there had been attempts to subsume IR within a larger subject matter, namely, IPE (Cox, 1987; Gilpin, 1987; Strange, 1988). Moreover, the impossibility of finding a single winning theory was not due to the dogmatic ill-will of scholars, but to the fact that our theories are potentially underdetermined by evidence. In this context, the ‘Third Debate’ argued that the deadlock could only be managed and made fruitful for research if IR started a second-order theorizing on the underlying assumptions of theories and how they structure the understanding of the research subject, problematiques and techniques. Hence, when Jeff Checkel recently assessed pluralism in the discipline and admonished that ‘bridge-building’ had mainly produced middle-range theorizing which disregarded meta-theory and epistemology, as well as ‘the macro-level where material power and social discourse — say — fundamentally shape and predetermine the mechanisms playing out at lower levels’ (Checkel, 2013: 234ff.), then, oddly, the IPD is not over, but, in some parts of our field at least, has yet to begin. This does not mean that the IPD was or is unproblematic for reasons already exposed in the 1990s, as, for instance, for its conservative bias in the use of ‘incommensurability’ or for the risk of a shallow eclecticism picking out of ‘menus for choice’ without control for meta-theoretical consistency (Guzzini, 1992: ch. 10; 1993; 1998: ch. 8; Wæver, 1996; Wight, 1996). But that historical critique of the IPD concentrated not on the need to reflect on theoretical assumptions, which it easily agreed with. It aimed at the ways those assumptions were bundled, and how the relation between those packages was, in turn, understood. For, as can be imagined, by having unleashed debates on all possible types of assumptions — ontological, epistemological and normative — and by trying to tie them together in triads that would be recognizable to political debates, the IPD, just like the earlier English School formulation, packaged too much into this. In Hedley Bull’s critique, he writes that Wight was ‘too ambitious in attributing to the Machiavellians, the Grotians and the Kantians distinctive views not only about war, peace, diplomacy, intervention and other matters of International Relations but about human psychology, about irony and tragedy, about methodology and epistemology’ (Bull, 1976: 111). Hence, the problem is not the awareness and the theorization of these different levels of assumptions, but their ready-made packaging and superimpositions. If applied correctly, there is nothing wrong with comparing three different approaches even on a level of assumptions, as long as a narrow theoretical focus is used (see, e.g., Hasenclever et al., 1997). And, of course, probing those assumptions opens up the field to a plethora of other isms (e.g. rationalism, constructivism, naturalism, materialism, idealism) and ologies that are probably fundamental to all (social) sciences and humanities. Hence, although the isms debates may no longer constitute the core debates of the discipline, they are always there in the background even if we may now concentrate on the discussion of specific problematiques, rather than those isms as such (and they remain in the foreground in the discourse shared by practitioners and surrounding experts). More importantly, the focus on isms as testable explanatory theories was not all that the IPD was about, let alone being its most important aspect. The central contribution of that debate was to take the constitutive function of theories seriously, that is, looking at theories as conditions for the possibility of knowledge. It therefore moved beyond the first behaviouralist attempt to turn practical knowledge into a science by proposing Guzzini 533 a self-reflection of world political observation. Throwing the isms out of the core is understandable in the one-truth gladiatorial world, since such a context renders the debate almost inevitably dogmatic. Well understood, however, the IPD requires both a search for coherence, auto-critique and tolerance. The IPD’s plea for reflexive theorizing is needed now just as much as then. In the following section, I will suggest ways to unpack the different modes of theorizing that informed the IPD. Four modes of theorizing This section introduces four modes of theorizing: normative, meta-theoretical, ontological/constitutive and empirical. While presenting these traditions of theorizing in turn, the basic thrust of my argument is that no science can afford to leave one of them out, since they are connected and their connections are significant for each form of theorizing. The following distinctions are crucial for the second main claim of the article, namely, that one may see a decline in IR theory or plead for getting rid of all isms only because one, again, misconstrues the very nature of theory, thereby neglecting many types of theorizing. Moreover, this unpacking of the IPD in modes of theorizing allows us to see more combinations and possibilities for cross-paradigmatic collaboration and mutual learning, therefore rejoining the spirit of Sil and Katzenstein’s (2010) plea for eclecticism, but keeping the requirement of wider theoretical coherence and open reflexivity. Normative theorizing consists in applying the scientific criteria of moral and political philosophy to issues of international relations. As such, it is not just an analysis of ethical issues in world politics, but one which is done according to the rules of philosophical argument. This ensures that the accumulated knowledge on ethical (and ideological) issues is taken into account, limiting the reinvention of wheels. That does not imply that normative questions can be answered philosophically in this sense; but it imposes some discipline in the way substantive moral argument is conducted.9 The significance of normative theorizing is self-evident for political analysis (since all politics is about values). Our research problems need to be or are informed ultimately by major ethical and/or political value issues. Meta-theoretical theorizing provides the building blocks and fundaments upon which all theories are built. This includes: ontology, the assumptions of what ‘there is’; epistemology, the assumptions of what we can know; and methodology, the assumptions of how the former constitutive components of theories can meet the empirical. Besides this usual meta-theoretical triad, this includes also assumptions about the nature of time (linear or layered) or the understanding of history, which, in turn can have implications for methodology. Meta-theorizing usually takes a critical form when it checks out the consistency of those assumptions upon which existing theories are built. Such meta-theoretical checks can show where assumptions clash, and where theory formation is consequently incoherent (see in IR, Kratochwil and Ruggie, 1986). But meta-theorizing is also used in a positive form when scholars need to formulate their theories in the first place, in particular, for ontological theorizing (see below). For instance, despite all the interest Alexander Wendt may have in the meta-theoretical level as such, I think that his work (in particular, Wendt, 1999) is better understood as driven by an attempt to 534 European Journal of International Relations 19(3) construct a coherent theory of IR, for which he has no choice but to engage with the philosophy of science and social theorizing. A third type of theorizing I call ‘ontological’ for lack of a better word. One could call it also ‘constitutive’ since it is mainly about theorizing the central phenomena that constitute the field of inquiry (power, sovereignty, state, etc.). The term ontological theorizing is inspired by the way some of our major theoretical texts are constructed. For instance, Hedley Bull’s Anarchical Society: A Study of Order [sic!] in World Politics is a relentless analysis of ‘What “is” …?’ (Bull, 1977: 3, 101, 127, 162, 184, 200, respectively). Also, recent, more positivist concept analysts takes account of the ‘ontological view of concepts’ ‘because it focuses on what constitutes a phenomenon’ (Goertz, 2006: 5). Ontological theorizing is (as all theorizing) in some sense normative but differs from normative theorizing in its mode. The actual purpose of theorizing is not to establish theories of the common good; nor is it necessarily connected to the study of values. In fact, it constitutes exactly that type of theorizing which saw the day when social theories shed moral philosophy. In that regard, it is part of the positive turn that is typical of (Western) social sciences. It just does not, and cannot, leave philosophy and history out of thinking the empirics. Its focus on central concepts, as in Bull, is typical because they stand for the greater issues for which we do science in the first place. This would crucially involve the very understanding of the ‘international’, with all the concepts around anarchy, hierarchy, heteronomy, society, community and so forth, as well as fundamental and secondary institutions, such as sovereignty (for the variety of takes, see, e.g., AdlerNissen and Gammeltoft-Hansen, 2008; Bartelson, 1995; Krasner, 1999; Walker, 1993) or, indeed, the state. Theorizing such building blocks of more general understanding is obviously connected to empirical analysis. As in Max Weber’s Economy and Society (Weber, 1980 [1921–1922]: 1–30), the ‘fundamental sociological concepts’ with which the book opens are the very possibility of the sociological analysis which follows in the later chapters, as well as constituting the conceptual results of the empirical analyses he has been conducting (and all the others on which he relied) for much of his career and which defined the terms within which empirical theorizing can take place. ‘Concepts are about ontology. To develop a concept is more than providing a definition: it is deciding what is important about an entity’ (Goertz, 2006: 27, original emphasis). The content and relationship of these concepts to each other is then checked by meta-theorizing for their consistency, and assessed, in ontological theorizing, both for their historical and conceptual congruence and according to the research findings they are able to conceive of. It conceives of the relationship to the empirics in a different manner though. In fact, and contrary to Goertz’s use, ontological theorizing is not reducible to the empirical level. For Goertz, we decide on this basic ontological level of concepts in view of the ‘causal powers’ of its constituents. This leads him to think of ontological theorizing as something akin to Wendt’s (1998) constitutive (instead of causal) relations (Goertz, 2006: 62). But that mixes the levels of observation. Constitutive relations are on the level of empirical theorizing when factors do not relate to each other in a causal but constitutive way (‘by reference to the structures in virtue of which they exist’) as, for instance, in empirical analyses of identity formation. Ontological theorizing is at another level, Guzzini 535 visible often in the establishment of frameworks of analysis or typologies which are mainly concerned with the constitutive function of theorizing. Last but not least, there is empirical theorizing, by which I mean more inductively driven research agendas. These can involve large-n correlational analyses (which also lead to deductive hypothesis-testing), small-n comparisons, single case studies and the generalizations which these allow (and this includes also hypothesis generation; see George and Bennett, 2005). But empirical theorizing can also feature smaller units which can travel from very case-oriented studies to others, namely, social/causal mechanisms. Although causal mechanisms are often conceived as the causal link to explain an established correlation, they can also be used outside a positivist framework. Following Jon Elster’s (1998, 2007) and others’ lead to think of such mechanisms outside a correlational logic, indeed, outside a Humean understanding of causality in general (Hedström and Ylikovski, 2010), one can conceive of such mechanisms, duly redefined, as part of interpretivist process tracing with a limited (because contextually open) capacity to travel to other cases (Guzzini, 2012b: ch. 11). Finally, such empirical analysis can feature the very relationship between the way we conceive of the social world and the social world itself, where, so to speak, our theories or basic concepts become themselves the object of empirical analysis. This ‘performative’ analysis can be applied to IR theories like the democratic peace (Ish-Shalom, 2006), to analytical models and their interaction with social reality (for financial models and financial markets, see MacKenzie, 2006), or to the ways certain categories (such as ‘pariah’), if applied by the international community, interact with the foreign policy identity of international actors. It prompts the need for theorizing this reflexive relationship. As this discussion shows, there are indeed different modes of theorizing with their various ends of theories. But they are also clearly connected, often conditioning each other. Indeed, our understanding of world politics would be impaired if any of those logics of theorizing were allowed to lapse, and if we did not, although perhaps not continuously, try to overcome our normal (and necessary) specialization in order to follow the links between their respective findings. The unfinished dictionary of the ‘international’ In my understanding, taking the constitutive function of theories seriously and allowing for the different modes of theorizing in our field of inquiry puts concepts and their analysis at centre-stage for our scientific communication. Just as terms are co-constitutive of language, concepts are co-constitutive of theories; they are the words in which, but also for which, our theorizing is done. Not only are concepts the means to achieve theorizing, but theorizing is also a means to redefine our concepts. We constantly rewrite our dictionary. Such a view implies a wider understanding of ‘conceptual analysis’ than usually offered. Recently, the role of conceptual analysis has been more widely acknowledged (Gerring, 2001), in particular, in a renewed reception of the work of Giovanni Sartori (Collier and Gerring (2009) includes some of Sartori’s main statements on the matter). But there has been a tendency to concentrate on the technical side, on the definitional side so to speak, neglecting the social-historical and semantic self-reflection which needs 536 European Journal of International Relations 19(3) to accompany any analysis of concepts. Just like data, concepts do not speak for themselves; they have a history. Understanding their ongoing history is not just a means but also an end of our theoretical conversation, since it is a depository of our accumulated knowledge. Indeed, in a section entitled ‘the loss of historical anchorage’, Sartori writes: our understandings of meanings are not arbitrary stipulations but reminders of historical experience and experimentation. Most of our political concepts were shaped and acquired their meaning out of a survival of the fittest process.… Thus, political scientists and sociologists — let alone the layman — ignoring the authors of the past have freed themselves not only from the constraints of etymology, but equally from the learning process of history. (Sartori, 2009 [1975]: 62, original italics) Asking for such a wider understanding of the role of concepts would also, in turn, affect the way conceptual analysis is done. In my reading, by resolutely embedding conceptual analysis in intellectual history, Sartori’s point should be extended to oppose both instrumentalism and essentialism. Sartori’s quote is clear about the former. Despite the inevitable fact that any concept is a construction of the observer, not any construction goes. We cannot just instrumentally define our terms as we feel best for coding, with no concern for their historical and wider purpose. Besides the technical requirements (about extension and intension, the risk of conceptual stretching, etc.), there are the historical legacies and roles of a term. Taking them seriously means that we cannot play around the semantic field as we like: we would end up with a clean definition which simply misses the entire point of the research. Our concepts are living memory. But opposing instrumentalism does not make conceptual analysis go essentialist, either. It is not because we dig into the history of meanings and functions of a concept that we find their ‘essence’. Concepts cannot be really thought independently of their semantic context and their pragmatic use. They are part of semantic relations. Their analysis is inevitably part of an interpretation within that semantic web. Just as much as we need the individual concepts to decipher these relations, the opposite applies. Concepts may have a tradition, several traditions indeed, but no essence. In a sense, instrumentalism and essentialism are the two flip sides of the same coin. They both avoid keeping the tension open in the development of ‘data’, the tension which is the result of the interaction of the observer’s constitutive theorizing and the historical development within which it takes place. The essentialist resolves the tension by assuming an ultimately purely external anchoring point; the instrumentalist by pretending that the purist conceptual definition and the formalization of research permits getting rid of the interaction. This also affects our way of working with definitions. If they are part of our everexpanding and updated dictionary, if concepts are at the meeting point between the observer and the observed, their definition may often gain from being conceived in an open way so that the empirics can also feed back to it. In qualitative (interpretivist) studies, but not alone here, our understanding of the issue should improve over the analysis. But this almost inevitably implies that the definition of what makes the issue significant also shifts, its meaning shifts. One cannot start the analysis without a first definition, and many of Sartori’s precepts (plus more conceptual history and historical Guzzini 537 sociology) are a necessary start. But on its unfolding, the analysis will keep returning to those definitions. Hence, our dictionaries are more than mere tools for analysis. The concepts therein can be of different types. Illustrative of the four modes of theorizing, this could include a reference to specific phenomena in their evolution (e.g. ‘rights’ for normative theorizing or ‘the state’ for ontological theorizing), meta-theoretical categories (e.g. individualism), operational concepts (e.g. Standard Operating Procedures), empirical mechanisms (e.g. socialization) or even miniature theories or ideal-types (e.g. democratic peace). In my view, if we stay problem-oriented in our field, they are the main stuff of our communication. Although isms may no longer be the core, such theorizing inevitably is. These concepts and their discussion then also constitute what the dictionary is about. Any discussion about a field’s core will be about what is important for understanding the entity IR. The dictionary is about the ontology of the ‘international’. To me, it seems that this is not the reserved turf of the international practitioner (and many realists, but not them alone), which resurrects an internal–external divide hard to defend. Nor is it empirical Political Science writ large, where the divide is ‘overcome’ by denying any significant specificity to the historical development of international practices. There are and will be different ways to read and contribute to the dictionary. In my own research, the ‘international’ in IR is about ‘global politics’, which uses and rewrites the dictionary with a focus on power and governance, including impersonal rule (Guzzini, 2012a, 2013). Conclusion IR theory has to defend itself on two fronts. One front consists of the alleged superiority of practical knowledge that gave rise to the discipline of IR in the first place. The other front consists in a narrow understanding of the ends of theory, indeed, of theorizing. This article has argued that boxing theory in either a practical or narrowly scientific mould does no justice to the substance and ends of IR theorizing. The former errs by limiting the capacity to observe practice. Indeed, the very categories taken for granted in the already-existing practical knowledge need permanent reflection. They cannot be reified across time, whatever cyclical vision of history usually underpins such attempts. There is a need for controlled and distant observation. But if that means that scientific knowledge is needed, theorizing has to take into account the constitutive and instrumental function of theories, and hence the different modes of theorizing. The article suggests distinguishing between four modes of theorizing, all needed and all connected: normative, meta-theoretical, ontological/constitutive and empirical. The respective knowledge is checked according to different and specific criteria and logics in each of them. Concepts play a special role by linking up these different modes of theorizing, providing the common language (and translations) within which our progress in knowledge can take place. I have pursued this argument in two steps which were meant also to provide the different steps of reflexivity IR theory went through. And that leads me to a final and perhaps somewhat unexpected remark. In my first section, I characterized professional schools in IR as being very theory-adverse, at least in my own personal experience. This 538 European Journal of International Relations 19(3) is surely still often the case. But it is not necessarily so, and one can see some interesting shifts. In fact, some of those schools may want to combine the best of both worlds by being self-reflexive with practical and scientific knowledge at the same time. This is perhaps not so astonishing. If these programmes take their function of educating future elites seriously, they need to provide not only factual knowledge, but also the capacity to think and hence be self-reflexive and adaptive when facing new decision situations. Such capacity is often best acquired through theoretical courses focusing on the constitutive function of theories (and historical courses which provide another form of distance to one’s present-day assumptions). They increase the independence and autonomy of thinking. By definition, these programmes need to reflect on the ongoing changes in actual international practices and in the ‘official’ language in which this is bundled into practical knowledge. As such, the ideal is to see the future elite become bilingual, able to understand the language of practice and science from the inside, as it were, and selfreflective on both. Taking seriously the difference between practical and scientific knowledge and the richer pedigree of modes of theorizing may not only make the centrality of theory more visible, but contribute to rethinking the relation to international practice and practices. Acknowledgements This is a revised version of a paper presented at the 53rd annual convention of the International Studies Association, San Diego, 1–4 April 2012, in the context of the panel series on ‘The end of IR theory?’ An earlier version of the first section was written and presented at the conference ‘The Sociology of the Social Sciences 1945–2010’, Copenhagen, 9–10 June 2011. I wish to thank the Collegio Carlo Alberto in Turin for granting me a fellowship during which I carried out all the revisions. Finally, I gratefully acknowledge comments, help and criticisms from Emanuel Adler, Alexander Astrov, Jens Bartelson, Andreas Behnke, Jeffrey Checkel, Petr Drulák, Karin Fierke, Jef Huysmans, Peter Katzenstein, Audie Klotz, Anna Leander, Richard Ned Lebow, Iver Neumann, Nick Onuf, Hidemi Suganami, Alexander Wendt, two referees and the editors of EJIR. The usual disclaimers apply. Funding This research received no specific grant from any funding agency in the public, commercial, or not-for-profit sectors. Notes 1. The first section is indeed very Euro-centric, which is justifiable when analysing the evolution of what today is called the discipline of IR. But it does not claim that it covers the history of the inquiry into international affairs or politics understood globally, or that Western IR is all that there is to the present discipline, even if it did its best to colonize it. 2. As Alexander Astrov and Nick Onuf insisted, this is a major qualifier if one wanted to provide a more comprehensive genealogy of IR. 3. For an earlier use of Elias’s argument for IR, see Krippendorff (1985). 4. Elias (1969: 181) insists that this therefore precedes the bourgeois-economic self-understanding of self-interest as rational. This is a logical corollary to the position that rationality cannot be understood independently of the social configuration in which it is realized; doing otherwise would reify the concept (Elias, 1969: 190). Guzzini 539 5. This shift, also visible in the aristocratic elite’s loyalty away from their peers (whatever their origin) to the nation, is perhaps most remarkably illustrated in Jean Renoir’s movie La grande illusion (1937). 6. The excursus on Weber is dealt with in much more detail in Guzzini (2007a). The analysis of Morgenthau is adapted from Guzzini (2007b). 7. This section has profited from correspondence with Scott Hamilton and Patrick Jackson. 8. A point I borrow from Patrick Jackson, personal communication. 9. I am indebted to Hidemi Suganami for insisting on this distinction. References Adler-Nissen R and Gammeltoft-Hansen T (2008) Sovereignty Games: Instrumentalizing State Sovereignty in Europe and Beyond. London: Palgrave Macmillan. Banks M (1985) The Inter-Paradigm Debate. In: Light M and Groom AJR (eds) International Relations: A Handbook of Current Theory. London: Frances Pinter, 7–26. Bartelson J (1995) A Genealogy of Sovereignty. Cambridge: Cambridge University Press. Bull H (1976) Martin Wight and the theory of International Relations. The second Martin Wight Memorial Lecture. British Journal of International Studies 2(2): 101–116. Bull H (1977) The Anarchical Society: A Study of Order in World Politics. London: Macmillan. Checkel J (2013) Theoretical pluralism in IR: Possibilities and limits. In: Carlsnaes W, Risse T and Simmons BA (eds) Handbook of International Relations. Los Angeles, CA and London: Sage, 220–241. Collier D and Gerring J (eds) (2009) Concepts and Method in Social Science: The Tradition of Giovanni Sartori. London: Routledge. Cox RW (1987) Production, Power and World Order. Social Forces in the Making of History. New York: Columbia University Press. Elias N (1969) Die höfische Gesellschaft. Untersuchungen zur Soziologie des Königtums und der höfischen Aristokratie. Revised and enlarged version of Habilitationsschrift from 1933. Frankfurt/Main: Suhrkamp. Elias N (1989) Studien über die Deutschen: Machtkämpfe und Habitusentwicklung im 19. und 20. Jahrhundert. Frankfurt/M.: Suhrkamp. Elster J (1998) A plea for mechanisms. In: Hedström P and Swedberg R (eds) Social Mechanisms: An Analytical Approach to Social Theory. Cambridge: Cambridge University Press, 45–73. Elster J (2007) Explaining Social Behavior: More Nuts and Bolts for the Social Sciences. Cambridge: Cambridge University Press. George A and Bennett A (2005) Case Studies and Theory Development in the Social Sciences. Cambridge, MA: MIT Press. Gerring J (2001) Social Science Methodology: A Criterial Framework. Cambridge: Cambridge University Press. Gilpin R, with the assistance of Gilpin J (1987) The Political Economy of International Relations. Princeton, NJ: Princeton University Press. Goertz G (2006) Social Science Concepts: A User’s Guide. Princeton, NJ: Princeton University Press. Guzzini S (1992) The continuing story of a death foretold: Realism in International Relations and International Political Economy. EUI Working Papers 92/20. Florence: European University Institute 287pp. Guzzini S (1993) Structural power: The limits of neorealist power analysis. International Organization 47(3): 443–478. Guzzini S (1998) Realism in International Relations and International Political Economy: The Continuing Story of a Death Foretold. London and New York: Routledge. 540 European Journal of International Relations 19(3) Guzzini S (2004) The enduring dilemmas of realism in International Relations. European Journal of International Relations 10(4): 533–568. Guzzini S (2007a) Re-reading Weber, or: The three fields for the analysis of power in International Relations. DIIS Working Papers 2007/29. Copenhagen: Danish Institute for International Studies, 34pp. Guzzini S (2007b) Theorising International Relations: Lessons from Europe’s periphery. DIIS Working Papers 30/2007 2007/30. Copenhagen: Danish Institute for International Studies, 37pp. Guzzini S (2012a) The ambivalent ‘diffusion of power’ in global governance. In: Guzzini S and Neumann IB (eds) The Diffusion of Power in Global Governance: International Political Economy meets Foucault. Basingstoke: Palgrave Macmillan, 1–37. Guzzini S (ed.) (2012b) The Return of Geopolitics in Europe? Social Mechanisms and Foreign Policy Identity Crises. Cambridge: Cambridge University Press. Guzzini S (2013) Power, Realism and Constructivism. London and New York: Routledge. Hasenclever A, Mayer P and Rittberger V (1997) Theories of International Regimes. Cambridge: Cambridge University Press. Hedström P and Ylikovski P (2010) Causal mechanisms in the social sciences. Annual Review of Sociology 36: 49–67. Holsti KJ (1985) The Dividing Discipline: Hegemony and Diversity in International Theory. Boston, MA: Allen & Unwin. Ish-Shalom P (2006) Theory as a hermeneutical mechanism: The democratic peace and the politics of democratization. European Journal of International Relations 12(4): 565–598. Jervis R (1976) Perception and Misperception in International Politics. Princeton, NJ: Princeton University Press. Kissinger HA (1957) A World Restored: The Politics of Conservatism in a Revolutionary Era. London: Victor Gollancz Ltd. Krasner SD (1999) Sovereignty: Organized Hypocrisy. Princeton, NJ: Princeton University Press. Kratochwil F and Ruggie JG (1986) International organization: A state of the art on an art of the state. International Organization 40(4): 753–775. Krippendorff E (1985) Staat und Krieg. Die historische Logik politischer Unvernunft. Frankfurt/M.: Suhrkamp. Lake DA (2011) Why ‘isms’ are evil: Theory, epistemology and academic sects as impediments to understanding and progress. International Studies Quarterly 55(2): 465–480. Lapid Y (1989a) Quo vadis International Relations? Further reflections on the ‘next stage’ of international theory. Millennium: Journal of International Studies 18(1): 77–88. Lapid Y (1989b) The Third Debate: On the prospects of international theory in a post-positivist era. International Studies Quarterly 33(3): 235–254. Leander A (2011) Methodolgies in International Relations: Of cookbooks and unfinished dictionaries. Paper presented at the ABRI Annual Convention, São Paolo (Brazil), 20–22 July. Lebow RN (2008) A Cultural Theory of International Relations. Cambridge: Cambridge University Press. MacKenzie D (2006) An Engine, Not a Camera: How Financial Models Shape Markets. Cambridge, MA: MIT Press. Meinecke F (1957 [1924/1929]) Die Idee der Staatsräson in der neueren Geschichte. Werke — Band I. München: Oldenburg Verlag. Milner HV (1998) Rationalizing politics: The emerging synthesis of international, American and comparative politics. International Organization 52(4): 759–786. Morgenthau HJ (1946) Scientific Man vs Power Politics. Chicago, IL: University of Chicago Press. Guzzini 541 Morgenthau HJ (1970 [1964]) The intellectual and political functions of theory. In: Truth and Power: Essays of a Decade 1960–1970. London: Pall Mall Press, 248–261. Morgenthau HJ (1970 [1967]) Common sense and theories. In: Truth and Power: Essays of a Decade. London: Pall Mall Press, 241–248. Sartori G (2009 [1975]) The tower of Babel. In: Collier D and Gerring J (eds) Concepts and Method in Social Science: The Tradition of Giovanni Sartori. London: Routledge, 61–96. Sil R and Katzenstein PJ (2010) Analytic eclecticism in the study of world politics: Reconfiguring problems and mechanisms across research traditions. Perspectives on Politics 8(2): 411–431. Strange S (1988) States and Markets: An Introduction to International Political Economy. New York: Basil Blackwell. Wæver O (1996) The rise and fall of the Inter-Paradigm Debate. In: Smith S, Booth K and Zalewski M (eds) International Theory: Positivism and Beyond. Cambridge: Cambridge University Press, 149–184. Walker RBJ (1993) Inside/Outside: International Relations as Political Theory. Cambridge: Cambridge University Press. Weber M (1980 [1921–1922]) Wirtschaft und Gesellschaft. Grundriss der verstehenden Soziologie, 5th rev. edn. Tübingen: J.C.B. Mohr (Paul Siebeck). Wendt A (1995) Constructing international politics. International Security 20(1): 71–81. Wendt A (1998) On constitution and causation in International Relations. Review of International Studies 24(special issue): 101–117. Wendt A (1999) Social Theory of International Politics. Cambridge: Cambridge University Press. Wight C (1996) Incommensurability and cross-paradigm communication in International Relations theory: ‘What’s the frequency Kenneth?’ Millennium: Journal of International Studies 25(2): 291–319. Wight M (1966) Why is there no international theory? In: Butterfield H and Wight M (eds) Diplomatic Investigations. Cambridge: Cambridge University Press, 17–34. Wolfers A (1962) Discord and Collaboration: Essays on International Politics. Baltimore, MD and London: Johns Hopkins University Press. Author biography Stefano Guzzini is Senior Researcher at the Danish Institute for International Studies and Professor of Government at Uppsala University, Sweden. His recent publications include The Return of Geopolitics in Europe? Social Mechanisms and Foreign Policy Identity Crises (Cambridge University Press, 2012), The Diffusion of Power in Global Governance: International Political Economy meets Foucault (Palgrave, 2012, co-edited with Iver Neumann) and Power, Realism and Constructivism (Routledge, 2013).