Abstract
This reply to the Symposium on Stefano Guzzini (ed.) The return of geopolitics in Europe?, answers the criticisms by John Agnew, Jeffrey Checkel, Dan Deudney and Jennifer Mitzen. It justifies (1) its specific definition and critique of geopolitics as a theory – and not just a foreign policy strategy; (2) its proposed interpretivist process tracing; (3) the role of mechanisms in constructivist theorizing and foreign policy theory; and (4) its usage of non-Humean causality in the analysis of multiple parallel processes and their interaction. At the same time, it develops the logic of the book’s main mechanism of foreign policy identity crisis reduction.
The return of geopolitics in Europe? (Guzzini, 2012c) engages a series of research agendas: the analysis of European security in the 1990s, the content and role of geopolitical theory, the usage of social mechanisms and process tracing in interpretivism and the theorization of micro-dynamics in constructivist theories of international relations (IR). It did not start that way (Guzzini, 2003). The more the research advanced, the more these different agendas connected. To a considerable extent, this is due to the discussion within the research group. Although the six framework and theory chapters addressed in this Forum were written by myself, they have been informed by an ongoing conversation. Far from simply providing case studies on different countries, the other authors in the book are all theorists in their own right with quite diverse sensitivities (Astrov and Morozova, 2012; Behnke, 2012; Bilgin, 2012; Brighi and Petito, 2012; Drulák, 2012; Kuus, 2012). Their empirical analyses were no mere application, but spurred reflections that fed back into the general framework; inversely, the discussion around the framework also affected the different paths of their own research (see e.g. Behnke, 2013; Bilgin, 2017; Kuus, 2014).
The present Forum bears witness to the book’s approach to thinking empirics, methodology and theory at the same time. John Agnew (in Agnew et al., 2017), although generally sympathetic, sees crucial factors missing. Daniel Deudney (in Agnew et al., 2017) takes me to task over the very understanding and tradition of geopolitics. Jeffrey Checkel (in Agnew et al., 2017) assesses the methodological proposal of interpretivist process tracing with the teaser of excessive inductivism. Jennifer Mitzen (in Agnew et al., 2017) would have preferred more theoretical engagement with anxiety, identity and security in constructivism before making the methodological jumps.
My reply roughly follows the logic of the book. I will start with Deudney, who got lost, perhaps not ‘on Earth’, but somewhere in Chapter 2. Since he alone misunderstands the research design and choices of the book, most of his remarks are unfortunately irrelevant. Yet, they allow me to situate the logic of the book before addressing Agnew’s, Checkel’s and Mitzen’s incisive comments on theory and methodology.
Which geopolitics?
Deudney (in Agnew et al., 2017: 10) claims that the omissions, absences and silences profoundly shape ‘in negative ways the insights this book advances about the nature of foreign policy crises, the role of “geopolitical” claims in the politics of this crisis’. In one stroke, he reduces foreign policy identity crises to simple foreign policy crises and, hence, eliminates the major focus of the book. As he sees it, I omit the pedigree of contemporary geopolitical thought – from Colin Gray’s strategic culture to his own work – that shows how geopolitical and liberal, democratic and republican thought can be combined. I do not deal with the central role of technology in geopolitical thought, neither at its height (Mackinder) nor in its role at the end of the Cold War with the nuclear fear heightened by Chernobyl. Finally, all turns towards Russia again, when I miss how justified Russia’s foreign policy (now suddenly) ‘identity crisis’ is in the wake of the gentlemen’s agreement undermined by a cheating West. 1 Instead, I concentrate on Ratzel, the pre-Nazi soil and land (Lebensraum) theorist.
The critique follows an established scheme. People criticize geopolitics, either because they do not know it, or because they have put on ideological blinders. Geopolitics is the victimized pet enemy of the righteous but ignorant liberal idealist. If only Deudney had read the chapter carefully, not to speak of the other chapters that establish the puzzle and framework of analysis, he would have noticed that the purpose of this chapter had to be entirely different than assessing a geopolitical explanation of the end of the Cold War or the crisis in Ukraine. He may have been intrigued by me making a controversial defence of Ratzel from easy charges of proto-fascist expansionism, and noticed that I do not omit but actually use Gray’s strategic culture, yet as a way to show the difference with the proposed framework (Guzzini, 2012a: 51–52). This said, his neglect of the main purpose of the study and even of the argument of the chapter in case makes it necessary for me to clarify the issue.
The first chapter (Guzzini, 2012f) establishes the background to the puzzle and sets the stage for the definition in Chapter 2. To (geopolitical) realists, the revival of geopolitical thought in Europe was not unexpected, since the end of the Cold War was accompanied by the Balkan wars - an argument along Mearsheimer’s ‘Back to the Future’ (1990). To political geographers, the revival was also foreseeable. If ‘geopolitical language can be recognized by the occurrence of words referring to boundaries and the conflict between territorially bounded interests’ (Dijkink, 1996: 5), then the territorial reshuffling at the end of the Cold War had to increase geopolitical talk in Europe almost by definition. In my view, the realist line begs the question, whereas the political geographers help to identify the puzzle more clearly.
There are, in fact, two realist arguments. One is to say that ideas only reflect the surrounding political and material conditions. Since the end of the Cold War was accompanied by local territorial wars in Europe, geopolitical theories would become prominent (again) compared to the relatively stable Cold War (in Europe). Such an argument begs the question, though. The end of the Cold War was an epochal and unexpected change in the global security order. It is hardly self-evident to consider the end of global bipolarity, in whose name massive wars had been fought, of lesser material importance for the development of ideas than a civil war confined to some parts of the Balkans. A second realist argument implies that any conflict is a point for realism, applied here: any territorial conflict is a point for geopolitics. Witness Deudney’s discussion about the territorial conflict/geopolitics in East Ukraine. Yet, this move ‘conflates description and explanation’ (Wendt, 1995: 76), political strategy and theory, the level of action and the level of observation. Diplomatic solutions are no point for liberal theories, and conflicts themselves prove nothing; all theories have and need explanations for both. Again, it begs the question.
Political geographers have been more to the point. Yet, if geographic imagination expectedly revives in moments of territorial change, it is not clear why some countries that got new borders (unified Germany and the Czech Republic) saw no or only an ephemeral revival of geopolitics (for the German case, see also Sur, 1995). Similarly, why would such revival take the form of ‘neoclassical geopolitics’ (Bassin, 2004: 621) with explicit references back to classical geopolitics and its explanatory primacy given to human and physical geographic factors? This then is the puzzle of the book: why would the revival of the most materialist and determinist line of theories occur in times of a new peace that nobody had anticipated?
Hence, the book is not about how geopolitical thought explains the end of the Cold War and its aftermath, but, in reverse, how the latter could revive geopolitical thought. In fact, what seems to have eluded Deudney’s reading is that geopolitical thought is not used primarily for the explanation of events, but as an empirical object of analysis. Ideas are real; they can be studied. Here, identifying the object in ‘neoclassical geopolitics’ with its strong determinacy, and not some watered-down ‘geography matters’, is doubly important: it is least expected in times of peaceful change, and it alone can be expected to have (self-fulfilling) performative effects where our way of seeing and conceiving the world not only passively registers, but also interacts with the social world.
But then, why defend Ratzel? This controversial move responds to the usual line of defence by geopolitical writers: being victims of guilt by association. If only, so the story goes, one would cut off the German branch, neither necessary nor typical, the geopolitical tradition would appear less tainted, its discussion more even-handed. Ratzel stands for the quintessential difference: seeing the state as a (growing) organism (1896a, 1896b). Without that biological analogy and the Social Darwinism it may inspire, geopolitics can have a respectable place in the history of ideas.
This critique may be partially correct for Haushofer (1924, 1934), despite his conflict with the Nazi focus on race instead of space and the eventual strategy of attacking Russia, which Haushofer opposed (Bassin, 1987). However, for Ratzel, allegedly the main culprit, it applies much less. He warns about taking racial distinctions seriously, which are but ‘deceptive garments’ (täuschende Gewänder), misleading the ‘superficial observer’ (Ratzel, 1882: 469). For him, humankind is fundamentally unitary in its anthropology – as well as in its destiny, which would see the increasing fusion of peoples into a common mankind (Ratzel, 1882: 177). Just as much as his education in zoology made him think about biology, it made him wary of it (Hunter, 1986). Indeed, at some point he openly cautioned against the use of a biological analogy (Ratzel, 1897: 12–13).
Inversely, the loose usage of organicism and of Social Darwinism (‘survival of the fittest’) does not distinguish the German from other classical geopolitical positions (Jean, 1995: 17–19). Organicism can be witnessed even in Mackinder’s famous speech (Mackinder, 1904: 422). Social Darwinism had two important non-German inspirations. On the one hand, it reached the Anglo-American world through the organicism of Herbert Spencer, who coined the term ‘survival of the fittest’, and whose volumes sold a staggering 368,755 copies between 1860 and 1903 in the USA (Hofstadter, 1944: 34). On the other hand, it is fundamental in the idea of laissez-faire economics, that is, a libertarian-conservative vision of progress, where the system itself, in its ‘natural’ selection most effective when left alone, would ensure that those unfit eventually be left out or perish, in capitalism’s ‘perennial gale of creative destruction’, as Schumpeter (1975 [1942]: 84) later put it. This derivation from liberal economics explains why E.H. Carr (1946: 46–50), one of the few early IR writers to critically engage Social Darwinism, places it oddly (or wryly) within the utopian, not the realist tradition (for a critical assessment of classical liberalism’s relationship to imperialism and scientific racism, see Hall and Hobson, 2010).
Therefore, Chapter 2 turns around and defines the classical geopolitical tradition by central core attributes alone. I develop four (Guzzini, 2012e: 27–28). Firstly, geopolitics stresses the interconnectedness of the world that produces a ‘Ganzheit’ (totality), that, for one present-day geopolitical writer, presents ‘the ultimate object and justification’ of geopolitical theory today (Parker, 1985: 2). Here, the organism metaphor functions as a proxy for the holism of classical geopolitics, but it can have other functional equivalents, as in system theory (besides Parker, see e.g. Cohen, 1991, 2003). Secondly, it insists on the finiteness of the world in which no conflict can be exported, no easy compensations achieved. Thirdly, it mobilizes neo-Malthusianism that provides the pessimistic determinacy: in a finite world with limited resources, seeking expansionism is the default position of any state, and its realization is necessary whenever an opportunity arises. This is, finally, connected to the Social Darwinism of an existential struggle for national/cultural or racial primacy.
Even granted that the classical geopolitical tradition is connected to Social Darwinism, is it fair to charge the neoclassical turn with classical environmental determinism, here understood as the explanatory primacy of geographic factors? It is not fair for those scholars, mainly in political geography, who analyse geography in its perceptions and institutional context (Lacoste, 1978) or, as done in critical geopolitics, focus on the politics of geography rather than the other way round. It may be unfair for those who use it as a framework of analysis, a ‘geopolitical method’ that refuses determinacy, although not when this slides into a ‘geostrategy’ to predict/the conditions of border disputes (see the slide in the conclusion of Foucher, 1988).
Such slides are my main justification. They are systematic. For, in their purpose of offering persuasive arguments and practical advice, geopolitical arguments confront a dilemma (Guzzini, 2012e: 36–41). If geopolitics merely means that we need to take physical and human geographic factors seriously, we all agree. This is neither distinct nor determinate: it cannot show how much geography matters, let alone how much it matters ‘in itself’, as geopolitical theory has it. If, then, it wishes to justify its distinctiveness, and legitimates recourse to its theory, it systematically glides into treating geography as the primary determinant – yet not providing a theoretical justification for it, since it rejects the charge of ‘environmental determinism’. In other words, geopolitical theory is either indistinct, indeterminate and correct but trivial, or distinct and determinate but wrong. This irresolvable dilemma explains what Carlo Jean (1995: 8, 20), a geopolitical thinker himself, calls the unavoidable geopolitical temptation of scientism and determinism (which he indeed does not avoid himself, as I show in the book).
It is important to understand this dilemma to appreciate what distinguishes realism from geopolitical theory (Guzzini, 2012e: 28–35). Whereas geopolitical theory is part of the realist family, not all realists do geopolitical theory. 2 There is a more diplomatic understanding of realism. Such understanding can be informed by the indeterminacy of rational human behaviour (Aron, 1962), a sense of practical prudence against the self-fulfilling character of the worst case (Wolfers, 1962) and self-restraint ‘to deal with the most fundamental problem of politics, which is not the control of wickedness but the limitation of righteousness’ (Kissinger, 1954: 1029). The overlap occurs when realists argue instead in terms of worst-case scenarios, most famously exemplified by the US Secretary of Defense Donald Rumsfeld’s (2002) plea to defend against the ‘unknown unknowns’, which, incidentally, is the one thing we apparently do know (and, hence, uncertainty is rationally domesticated into ultimate risk). Here, a fundamental determinacy creeps in, which mobilizes a ‘strategic’ version of realism, what I call ‘realism’s military gaze’.
For the effect of invoking geopolitical terminology is a militarization of politics. It reverses Clausewitz’s famous dictum. Rather than making war the prolongation of politics with other means, it makes politics the mere prolongation of war with other means. Such a pernicious reversal was a constant risk during the Cold War (Aron, 1976); a revival of geopolitics would return to it. It also has a second effect, namely the homogenization and essentialization of geographies, including collective identities, spatially and culturally defined. Talk is not cheap.
For these reasons, the empirical object of analysis, neoclassical geopolitics, is defined as…
…a policy-oriented analysis, generally conservative and with nationalist overtones, that gives explanatory primacy, but not exclusivity, to certain physical and human geographic factors (whether the analyst is open about this or not), and gives precedence to a strategic view, realism with a military and nationalist gaze, for analysing the ‘objective necessities’ within which states compete for power and rank. (Guzzini, 2012e: 43)
The question was then to find out whether and why it was revived in several European countries in the 1990s and hypothesize some of its effects. The case studies of ‘smaller European languages’ (Deudney), which found a revival in four out six countries, analysed geopolitical thought in terms of this narrower definition. If referred to in this revival, Deudney, Gray or Huntington would be part of the analysis.
The framework of analysis
Agnew and Mitzen bring a series of qualifications to the framework of analysis. Although both may sometimes push me from a vantage point that could have been another book, I think they are generally justified in doing so for the sake of the wider intellectual discussion to which this book belongs. Let me start with a short summary before locating their criticisms.
The book’s general framework of analysis is inspired by constructivist foreign policy analysis. This pertains to at least three ideas. Firstly, social action and interaction is significantly, but not exclusively, constituted by social practices of recognition (Guzzini, 2012d: 265–266). 3 Secondly, security imaginaries (Weldes, 1999) and their embedded identity discourses are a privileged place for understanding foreign policy traditions and strategic options (see also Mälksoo, 2012b). Such understandings are the condition of possibility for behaviour, but do not directly determine behaviour (Hopf, 2002: Chapter 1). Thirdly, the social world is characterized by reflexivity, something that Hacking (1999) calls ‘looping’ or ‘interaction’ effects, and that are crucial for understanding self-fulfilling effects.
The central unit of analysis of the book is Jutta Weldes’s (1999) security imaginary (for more details of the following, see Guzzini, 2012a: 52–55). It is conceptualized in the book as a country’s foreign policy tradition, a repository that carries lessons of history, repertoires and scripts (or plots) out of which representations of the self (biographies) and of the world of IR are created. The security imaginary circumscribes a specific social field, in which social reality (always combining matter and ideas) interacts with our understandings. Social reality does not come with a single meaning-tag attached; discursive formations evolve through social practices enabled or constrained by social reality.
A security imaginary is not homogeneous. A foreign policy tradition is not about a single shared opinion. It is a system of references that frames and authorizes certain positions as legitimate parts of the debate. The security imaginary shares a common sense that provides the understandings upon which foreign policy experts can then disagree. This is part of ‘the concession to a social universe that one makes by accepting to become acceptable’ (Bourdieu, 2001: 114).
A security imaginary includes multiple identity discourses. Those identity discourses usually include perceptions on how others perceive the self. It is no stable state, but itself the snapshot in an ongoing process of identification with an ever-precarious fixing (Campbell, 1992: 11 and passim). The prevalence of one particular identity plot in the discursive formation and the potential continuity of the latter are the result of dynamic processes, not of stasis.
Since such identity discourses are in a constant re-articulation, a crisis is not to be understood as just any externally induced change of a discourse ‘at rest’. A crisis refers to those moments in which debates cannot update or combine repertoires to fix tensions, either internally or externally induced, or, when doing so, end up questioning the imaginary’s tacit assumptions, its common sense. In the historical context of the study, that is, the end of the Cold War and its immediate aftermath, the book focuses on the expectation of such crises to be happening when repertoires have to be mobilized for creating a new, or responding to a newly questioned or newly acquired, self-understanding or external role recognition in world affairs.
The country studies in the volume reflect these cases. All six countries were prompted to rethink their role in international affairs, partly because they were countries in new borders, or indeed new countries, or because their international role was so closely tied to the Cold War dynamics that its absence produced a conceptual vacuum. Hence, we had countries where political geographers would expect the emergence of a geopolitical debate and potentially the revival of geopolitical thought. Yet, in two of those countries, namely the unifying Germany and the now split Czech Republic, no revival happened.
Finally, identity crises do not automatically trigger a new understanding that privileges a recourse to geopolitical thought. Even if geopolitical thought can be alluring for fixing identity, it does not necessarily happen (it is functional, but not necessary). For this to happen, the book discusses a series of ideational, institutional and rhetorical factors that could facilitate that effect, such as the ideational path dependence of a materialist political culture in a country, the political economy of expertise production, including the importance of military elites or the lacking financial or status independence of academia and media, and the role of nationalism in political debates.
Agnew’s critique sets in here. He appreciates the definition of neoclassical geopolitics to be open enough to account for the more geocultural than strictly territorial versions prevalent for some decades. He supports the focus on identity discourses, and that understanding their crises is embedded in a wider analysis. Yet, he sees some factors and some effects of a geopolitical revival missing. For one, he thinks that a country’s previous historical standing in the world is an important factor for understanding the revival. Secondly, he would see these crises as driven by not only the end of the Cold War, but also by certain effects of globalization more generally. Finally, he thinks that neoclassical geopolitics not only depoliticizes by remilitarizing public debate as in the reversal of Clausewitz, but also by systematically projecting ‘the causes of our discontent beyond state borders. … Action from over the horizon dictates our fate. Foreigners are to blame’ (in Agnew et al., 2017: 8). 4
The first two critiques are informed by Agnew’s (2003) well-argued reference to larger historical developments for understanding the role and effect of geopolitical thought, including the history of national rank in the world political hierarchy. This factor, which he sees as salient for the Russian case, is one that the book discusses in terms of ‘the ideology of great or dissatisfied power’ (Guzzini, 2012a: 61–64). Rather than being a separate factor, we included it within the security imaginary itself, where it would be part of identity discourses.
A second deep point of his has to do with process. If the book is right in its insistence on imaginaries as processes, then any historical starting date for the analysis will always have something arbitrary. Historical processes that led to the revival did not start in 1989 (or 1991 for Russia). This point is very consequential, since it is not just that structural factors of the global political economy were neglected in the analysis, although they are, but that by opening up to these macro-processes, the setup in terms of a process from A to B is to be rethought. I agree with the thrust of the critique and will take this up when the process analysis is introduced in the last section and in the conclusion.
A third and crucial point raised by Agnew is the role of nationalism in constituting geopolitical thought and in triggering its revival. The nation is a victim to be defended against all ills, which originate elsewhere (even if that ‘elsewhere’ ends up being ‘here’; see, for instance, Combes, 2017). A prominent role of nationalism does not contradict the book’s theoretical setup, since state identity discourses are by definition tied to ideas of the nation (see also Berenskoetter, 2014). The book explicitly dealt with the role of nationalism when it qualified the definition of neoclassical geopolitics for not being necessarily tied to conservative circles and thought – just as Social Darwinism could be found across the political spectrum, including with liberals and socialists (Hawkins, 1997). Indeed, geopolitics appears as the foreign policy ideology of nationalism. Yet, Agnew is surely right that this connection to nationalism could have warranted more thought, because it is theoretically tricky and because it could provide a better grip for understanding the distinctive content of individual geopolitical traditions (as critically noted by Klinke, 2013; Sidaway, 2014).
Mitzen’s critique probes deeply into the causal logic of the framework of analysis. She argues that ontological insecurity is a cause that can have multiple effects (multi-finality). Concretely, there are also creative agency-empowering ways in which such insecurity may be resolved. In addition, the other way round, there are resolutions to ontological anxiety that have a similar agency-narrowing effect, but are not connected to geopolitical thought (equifinality). By reducing the outcome to a binary (rise or not of geopolitics), the framework impoverishes the theoretical understanding of anxiety and its resolution. This critique is informed by her own research agenda on ontological security (Mitzen, 2006), to which I also refer in the book. I will start by showing that there is more flexibility to my process tracing than she acknowledges, before using her pointers to open the more fundamental issues on identity theory and causality.
There are different ways to conceptualize identity. In the book, identity is a discursive phenomenon; that is, it is not analysed in terms of scaling up from individual to collective identity (with all the problems entailed). Rather, the identity of a country exists through its recognition by other actors in a diplomatic and public discourse that attributes often anthropomorphic expectations to collective actors for making sense of their behaviour. Identity is constituted in this ongoing communication and recognition. It is in the practices of actors, and it exists only as long as it is enacted through these practices.
Jutta Weldes (1999: 215–231) shows how this works, for instance, when she discusses how identity discourses and credibility were connected in the US decision-making during the Cuban Missile Crisis. Or, to use a recent statement in an interview with Henry Kissinger, ‘credibility for a state plays the role of character for a human being’ (interview in Goldberg, 2016; for a similar take by a Brazilian scholar-diplomat, see Lafer, 2004). Hence, ‘identity’ refers to those attributions of personality in discourses among political actors, their observers (media and academia) or the public at large within which the subject-position of the country in world politics is negotiated. This negotiation takes place in domestic and external fields of recognition (identity as self-understanding and external role recognition, respectively). Based on social theories of recognition, it is assumed that countries generally seek to achieve a consistent subject-position (even if that is one of an unpredictable actor) such as to allow for a credible self-understanding and external role. 5
For the argument of the book, an identity crisis is not understood as just any disturbance of a foreign policy tradition, but only those for which the existing repertoires and plots find no solution. This is tricky, of course. On the one hand, any tradition is in potential crisis and, hence, a resting self is always transitory. Moreover, as the analysis in the book shows for Russia and Turkey, and to some extent Italy, some foreign policy traditions are in almost constant crisis, since alleged solutions taken from past repertoires unravel almost as soon as they take grip. Indeed, precisely for being an easy but indeterminate fix, geopolitical thought only perpetuates those identity crises (Guzzini, 2012b: 233–236, 239–242). Yet, it is fair to say, and can be empirically shown, during some times more than others, the self-narratives in foreign policy discourses are unchallenged and/or are in harmony with the way internal or external publics (or fields of recognition) recognize them.
This starting point of identity in discursive formations is hence related to, but different from ontological security theory (OST), as used by Mitzen. Concomitantly, the central issue of an identity crisis and its resolution – that is, the first social mechanism of the analysis, ‘foreign policy identity crisis reduction’ – is somewhat analogous, but not the same as the reaction to existential anxieties, even if such anxieties can also include issues of identity.
A first difference is that dissonance reduction does not primarily respond to a human need as it does in OST, as in an emotional response to anxiety (see also Steele, 2007). Instead, it derives from the response to the conventional expectations of both domestic and international audiences for a coherent ‘credible’ identity to make sense of a state’s action. They are part of a social field that makes such attributions. Similarly, as in the Kissinger quote above, such reduction is deemed necessary in order to be able to project a coherent purpose of a country. As compared with OST, this is decidedly a more discursive take on identity (see e.g. Epstein, 2011; Hansen, 2006; Huysmans, 1998), although with a more sociological twist, since it is embedded in social fields of recognition.
That difference also shows in the understanding of Mitzen’s critique that anxiety can also encourage creative agency (see the excellent discussions in Berenskoetter, 2016; Rumelili, 2015a, 2015b) and not only narrowing agency, which she sees represented by the revival of geopolitical thought. Yet, as truly important as this argument is for OST, it differs in two ways from my approach. Firstly, the dissonance reduction in the book is empirically studied at the level of ideas themselves, not of foreign policy behaviour. 6 Secondly, the approach does not share the potential status quo bias of OST. Indeed, Weldes’ analysis of the security imaginary insists on the possibility of re-articulating these repertoires. It is something we would expect to happen, in particular, in times of crises where public demand for ‘story-telling’ is elevated, that is, where agents come up with new ways to rearrange the plots for making sense of a world in change (Krebs, 2015).
It is, therefore, not primarily in terms of agency-narrowing that one should understand the revival of geopolitical thought. Although Huntington’s geopolitical as geocultural re-articulation of the Cold War in terms of a ‘clash of civilization’ relies on old tropes (see e.g. Dalby, 1990; Ó Tuathail and Agnew, 1992), it puts them creatively together to provide a very powerful vision in a moment of conceptual disarray. 7 In other words, the response to anxiety in terms of OS is not the same as dissonance reduction in the book, both for the locus and conceptualization of identity and, consequently, for the actual mechanism that carries the weight of theorization.
This still leaves open Mitzen’s probing question about the causal process. Mitzen raises the issue of multi-finality, namely that the identity crisis could have several outcomes, not only the binary about geopolitical revival. I agree. In fact, the first mechanism I develop in the book is not about geopolitical thought at all, but specifies possible solutions for an identity crisis. As I write:
Such a dissonance would then trigger more concrete responses, which include: (1) denial that there is any dissonance to start with; (2) negotiation, that is, attempts to dissuade the other from the faulty identity vision (i.e. there is a perceived dissonance, but this is based on a misunderstanding); and (3–4) the acceptance of a real dissonance, which then spurs either attempts to change the international culture in such a way as to enable one’s own identity to fit (imposition), or efforts to redefine the identity as a way of adapting to external expectations or projections (adaptation). I posit these responses as components of the overall analysis of this discursive mechanism. (Guzzini, 2012d: 267)
These four outcomes fit a deductive matrix, distinguishing on the one axis whether foreign policy discourses acknowledge the need for identity discourses to change, and then whether they see the change to be happening within themselves or abroad (see Agnew in Agnew et al., 2017).
Multiple paths of foreign policy identity crisis reduction.
There is no reason to believe that these are the only ways, however. For instance, Pinar Bilgin and Ayşe Zarakol develop other historical answers to identity crises. Bilgin (2008) adapts Homi Bhabha’s (1994: 85–92) take on ‘mimicry’ as something that includes both emulation and camouflage. She develops the puzzle that secularism in Turkey is associated with security and argues that Turkey’s Kemalist turn to secularism was meant to imitate a picture of the West for protecting the new country from an always possible external intervention. This is a form of adaptation that is an active appropriation that does not, however, end up ‘being the same’. Zarakol (2010) also finds a different reaction to international socialization pressures beyond adaptation or imposition. In her analysis of Turkey’s and Japan’s refusal to accept what appears to them as a form of self-denial, we would, in fact, have a way to understand endemic identity crises as depicted in the book; for dissonance is only temporarily reduced, oscillating between showing the ‘good’ internal side and calling the hypocrisy of the external expectation. In her work on stigma management, she opens up the categories of adaptation and imposition by showing strategies of outright rejection without a chance to impose, or ressentiment without real adaptation (Zarakol, 2011: 104–107).
Their conceptualizations point to a neglected component in conceptualizing the processes of identity crisis reduction in the book. Not only are there always different identity discourses within a security imaginary, but such individual discourses are often intrinsically heterogeneous. This can be understood in terms of liminal identities (Norton, 1988), like ‘strangers’, both foreign and at home (for IR, see e.g. Mälksoo, 2012a; Rumelili, 2003, 2012), or as ‘hybridity’ stressed by postcolonial approaches (for a congenial IR application, see e.g. Morozov, 2015). As this shows, there are surely more, but not infinite, ways to dissonance reduction. The book contributes to the theorization of the mechanism; it is not its end. Hence, multi-finality is indeed crucial for my understanding of social/causal mechanisms.
However, that may only displace Mitzen’s critique: the argument is now too loose to be explanatory at all, neglecting alternative or parallel causal lines. The issue turns towards the very status of the explanation and its underlying take on causation. Mitzen rightly identifies a shift in the analysis from the early framework chapter to the last one. Whereas the early setup looks rather for the causes of effects (the geopolitical revival), by the final chapter the book looks for the indeterminate effects of causal mechanisms within the process itself (and not between variables). Her charge is therefore justified to the extent that if the book had stayed with its initial empirical puzzle only and not proposed a theory development, a different strategy would have been warranted. The book can still make a point for the role of identity processes in the revival, but this role is more circumscribed, since I introduce multi-final outcomes for the first mechanism with an open link between this mechanism to geopolitical revival. Mitzen’s concern about what is actually explained turns into a question of what can be explained in interpretivist process tracing.
Social mechanisms, causality and constructivist theory
Jeffrey Checkel too sees the need for more theorization. Although generally persuaded by the clarity and transparency of the way the mechanisms are established and interpretivist process tracing is conceptualized, he sees ‘constructivism at mid-life’ (crisis). For one, he ironically sees the book falling into the trap of the ‘tyranny of mechanisms’ where research programmes feed into an ever-growing list of non-cumulative mechanisms. This charge is ironic, since it echoes the critique of its methodological opposite, that is, quantitative correlational analysis: lots of little regularities, empirical and inductive, but no real theory. He does not see how these mechanisms could travel or be generalizable. No forest of theory for all the trees of mechanisms; at best, some meagre middle-range theory, as he teases. Secondly, he finds that the empirical study turns out to be of little difference to soft positivist approaches, in particular the ones that take historical sequence and time seriously. Finally, he wants to have more operational guidelines for the empirical analysis, good-practice community standards. I will concentrate on the first two and leave, just as he suggests, the third one for future discussion, including about the nature of such standards (since pushing for some type of standards undermines the very specificity of interpretivist research; it is not innocent).
Checkel sees the mechanisms as inductively derived. I think this take vitiates much of his argument. Just because the book does no deductive hypothesis testing, its empirical analysis is not purely inductive (which, as we know, is not possible anyway). Concentrating on the processes that go into identity formation and the self-fulfilling practices that constitute international cultural structures (Wendt, 1999: 184–189) are not just any kind of mechanisms; they are central for the micro–macro links in already established constructivist theorizing.
Moreover, his comments seem to suggest that there are only two ways to theory, the grand theories, which tend to lose empirical detail, and empirical generalizations, inductively derived. By not using the first way, the book would be compelled to use the second – and from there the critique of inductivism, non-accumulation and problematic generalizability follows. That is too narrow a view on theorizing, however. In the book, the empirical–theoretical link is not in terms of empirical generalization to the universe of cases, but in terms of how the empirics affect the conceptual lenses with which we approach any empirical analysis. In this case, it specifies the mechanisms, their possible origins or multi-final results and their role in the understanding of the processes with which the underlying social theories of recognition can be used to analyse our world. Theory comes before the induction.
Such a view on theorization also leads to a different understanding of generalization and accumulation. The accumulation does not take place in the aggregate of mechanisms that fill up our theoretical universe; it takes place in the refinement of the conceptual apparatus with which our theories engage the empirical universe (and where Mitzen justifiably pitched her critique). Those mechanisms do indeed travel, but not via generalization; they travel via a form of abstraction, similar to ideal types, which are embedded in a (here: constructivist) framework of analysis that will be applied to other cases or events (for a comparable argument within a thin soft rationalist framework, see Bengtsson and Hertting, 2014).
By not seeing the crucial role of ontological or constitutive theorizing for the social sciences in general (Guzzini, 2013), and for constructivism in particular, and by reducing our theorizing to the classical inductive–deductive scheme of (more empiricist) positivism, Checkel misses the dynamics of theorization in the book. Truth be told, the book takes much of this for granted when embarking on its constructivist analysis, and the embeddedness into underlying social theories is not thoroughly pursued. Still, when claiming to look for the micro-dynamics of constructivist theory from an interpretivist angle, no other type of theorizing would make sense.
Similarly, Checkel’s teaser of ‘middle-range theory’ mobilizes a more positivist understanding of the term (mid-range empirical generalization), which is not all what the middle-range can offer. The book does constitutive theorizing by developing micro-dynamics of constructivism and empirical theorizing by specifying the work of mechanisms. Doing so, it may actually not be too far from Merton’s initial call for the middle-range, which was never meant just to be a kind of inductive generalization of limited ambition (see also Tilly, 2010). Merton’s strategy was dual. On the one hand, he asked for middle-range ‘abstraction’ to be done in terms of ‘mechanisms’, which is how he understood self-fulfilling prophecies (Merton, 1948: 200) and that could very well include macro-sociological phenomena, as in Max Weber’s Protestant Ethic. On the other hand, this was to be done in connection with the development of a ‘conceptual scheme that is adequate to consolidate groups of special theories’, which he calls ‘paradigms’ and that develop both the underlying assumptions and conceptual meanings (Merton, 1968: 51, 69–72).
I would agree more with Checkel’s second remark that empirical analyses may not look all that different across the epistemological divide. I share his sense that there is something positive about this, precisely because meta-theoretical discussions can be misused for empty flag-waving and cosy camp-building, which Checkel has always rightly and consistently opposed. I would, however, reverse the implication that is usually drawn (albeit perhaps not by Checkel).
If it were true that the differences are not so big in empirical research, that different methodological understandings (about the types of theory, etc.) do not impede cross-epistemological communication, then surely our research and teaching would be happily integrating both sides of the divide. It seems, however, that not much has changed since Hayward Alker’s (2000) pointed invitation to do so (Hagmann and Biersteker, 2014). To push it a little further: rather than assuming that all empirical constructivist analyses are closet positivistic approaches, hence condoning the use of positivist criteria for their assessment, could it be that the positivists who take historical institutionalism and ideational approaches seriously are closet interpretivists? What if the differences were not large, because it is so hard to be a consistently positivist researcher when doing problem-driven, not method-driven analysis of the kind that opens the black box of historical processes and individual preference formation? What, for instance, if it turns out that the underlying understanding of causality cannot be a Humean constant conjunction?
For this is precisely what the conceptualization of social mechanisms proposes in the book. It is inspired by understandings of causality that go beyond Humean constant conjunction (Guzzini, 2012d: 256–264). This is a point that has been pushed for quite a while by critical realists in IR (e.g. Kurki, 2008; Patomäki, 1996; Wight, 2006) and also by researchers interested in historical processes (e.g. Lebow, 2014; Suganami, 2008). 8 The book’s take starts from Jon Elster’s definition of social mechanisms as ‘frequently occurring and easily recognizable causal patterns that are triggered under generally unknown conditions or with indeterminate consequences’ (Elster, 1998: 45, 2007: 36). In this wording, leaving open either origins or effects, social sciences cannot aspire to general laws (where both conditions and consequences are known) but offer more than pure contingency (where none would). As Hedström and Ylikovski (2010: 53) write, this excludes approaches that define ‘causality in terms of regularities (such as Hume’s constant conjunction theory or many probabilistic theories of causation)’. Yet, the latter was initially even proposed by George and Bennett (2005: 225–227), a take now revised (Bennett and Checkel, 2015: 6–7).
It seems to me that the interpretivist implications of going beyond probability (which leaves the logic of constant conjunction unchanged, but qualifies its usage in the social world) and of taking non-linear temporality seriously (Abbott, 1988) are not sufficiently acknowledged, if at all (for more detailed discussion on determinacy, see Guzzini, 2016; Lebow, 2014: 92ff.). So does the introduction of a recent special issue invite scholars to focus on their ontology first, already distinguishing mechanisms for being either deductively or inductively derived, and either determinist or probabilistic (Trampusch and Palier, 2016)? This prompts a punctual reminder by Renate Mayntz (2016) that these distinctions miss the more fundamental issues on causality opened by contemporary ontological debates on contingency and time. If, however, we start from such ontologies, then there is a way to think causality in terms of ‘causal reconstructions’ (Mayntz, 2002), which is characterized by its multi-finality with contingent links, its non-linear process and mechanism character, historicity (taking sequence seriously) and complexity (which is not just stochastic). It is an explanation but without regularity, and yet not just a description since it includes abstracted and transferable aspects (see also her excellent discussion in Mayntz, 2004). 9
The book explores a version of process tracing that includes such open causality and that is interpretivist, historical and multi-layered. It is interpretivist for two reasons. Firstly, the starting moment of the process is not given by a mere external choc (the end of the Cold War), but by how such events are understood in foreign policy discourses. Secondly, by looking at self-fulfilling mechanisms, it includes a reflexive element in which such interpretations, in turn, affect world politics. It is historical, because it takes sequence and time seriously (Grzymala-Busse, 2011), which means that explanatory factors along the path cannot simply be aggregated to increase total variance. This is obviously a point that has been made for a long time by some historical sociologists (Tilly, 1998, 2001) and historical institutionalism (Pierson, 2000) or comparative-historical analysis, when the latter is defined by a focus on macro-configurational explanation, the pursuit of problem-driven case-based research and a methodical commitment to temporally oriented analysis (see this recent conceptualization in Thelen and Mahoney, 2015).
The process tracing is also multi-layered (for want of a better term), showing how autonomous processes evolve and interact with each other during the period an analyst has decided to study. Rather than assuming one single process line that various factors punctuate, ‘the focus is not … on the trajectory of a single process. It is on the temporal intersection of distinctive trajectories of different, but connected, long-term processes’ (Aminzade, 1992: 467). As Falleti and Lynch (2009: 1156–1158) show, such processes can be conceptualized within layers in an analysis that cuts a certain time period out of them for the sake of studying a theory-informed empirical puzzle. 10 Configurational analysis (see also Jackson, 2006) is laid out into time, where some processes change at higher speed than others (from bottom to top in Figure 1).
Social mechanisms in a time-layered interpretivist process model (reprinted with permission from Guzzini, 2012d: 273).
Figure 1 illustrates two sequences with one mechanism each, connected (or not) by the rise of geopolitical thought. The first mechanism has a multi-final outcome, which, in turn, is dependent on the effects of the other relevant processes (political culture, political economy of expertise and political struggles/mobilization) for making the revival of geopolitical thought possible (for a summary of the book’s hypotheses and findings, see Guzzini, 2012d: 248–249). Geopolitical revival is not a functional necessity. The book unfortunately only has cases where identity crises and revivals are connected. Yet, the case of Sweden could be one where the end of the Cold War produced an identity crisis, since the country’s neutral identity (Malmborg, 2001) made less sense, and yet no geopolitical revival ensued. The process tracing is not committed to a regularity, but to finding out how and whether the mechanism can be seen in the individual cases. The second mechanism, which is not developed in detail in the book, starts from a geopolitical revival and then applies it as a self-fulfilling mechanism, a ‘vicious circle of essentialization’, which affects the culture of anarchy (for the same type of mechanism, see ‘securitization’ in the Copenhagen School; see Guzzini, 2011).
Just as the choice of these two mechanisms is not accidental, but central for constructivist theories, the choice of geopolitics as the focus is also not fortuitous. Neoclassical geopolitics provides a crucial link in the process. On the one hand, it responds to identity processes, being about state properties such as culture, territory, borders and the national imagination. On the other hand, its dual dynamics of essentializing identities and militarizing politics have an effect on the cultures of anarchy, constituted by the collective self-understandings and the definition of the relevant ‘roles’ in the international sphere of recognition. In this case, they would be encouraging a more Hobbesian culture of anarchy.
Yet, the overall outcome can be influenced by the same type of self-fulfilling mechanism but with different content. There can obviously also be de-essentializing and demilitarizing identity discourses. The post-national European Union (EU) is sometimes seen in this light, and desecuritizing strategies such as German Ostpolitik work that way, too. The revival of neoclassical geopolitics was the empirical puzzle that, for its central place in the two sequences (start or end), prompted a theory development. However, it would surely not claim to be the only set of ideational practices affecting cultures of anarchy.
Conclusion
In my reading, Checkel’s critique is based on a limited understanding of the role of theories in the social sciences, and the place the two theoretically central mechanisms of the book have in and for constructivist theory. I did not answer his methodological critique, which would then have to start from that different take on theorizing and explanations. Similarly, besides acknowledging her points, I think that my discussion has shown that neither the mechanisms nor the underlying causality work in the way Mitzen assumes in her critique of missing multi-finality and theory. Theorizing foreign policy identity is not subsumed under OST (see also Browning and Joenniemi, 2017). If so, their theoretical relationship is, however, not sufficiently explored, which I take to be one important implication of Mitzen’s critique. That would include, for instance, the question of how these mechanisms relate to the social psychology of identity processes (for different takes, see Kinnvall, 2004, 2017; Lebow, 2016) and of nationalism, mentioned in Agnew’s critique.
That leaves Agnew’s main point. Precisely because I have developed the analysis in terms of multiple simultaneous processes, and precisely because I do analyse security imaginaries/foreign policy discourses within their institutional and political context (for such a critique of discursive approaches, see Laffey, 2000), his critique is meant to push me/us further by including an economic factor of the longue durée. The issue is how to include it. It could be yet another layer with the practices of the global political economy, which would affect the process under scrutiny. The book does not, in principle, exclude such an addition; yet so far, it theorizes the effects of capitalist practices within the other processual layers. Similar to Bourdieu, capitalist processes are a background process that affect all fields; their analysis is hence done within those fields and their interaction. This way to deal with them fits the short time-span focus of the book, but it does not cover the macro-historical shifts that Agnew may have in mind.
I am grateful to this Forum for having pushed the book’s conceptualization of foreign policy identity crisis and its proposed interpretivist process tracing with social mechanisms further, not least by exposing its limits and omissions. Its political implications remain. The book traces how the end of the Cold War and the overall peace in Europe could lead to a militarization of international politics and essentialization of identities via foreign policy identity crises and the rise in geopolitical thinking. As paradoxical as it may sound, peace made militarization possible – well before 9/11. Such militarization and essentialization of foreign policy discourses has affected the interpretation of international events since the 1990s, and has prepared the ground (among other factors) for securitizing societies and militarizing actual foreign policies, such as in NATO expansion and in the Crimea.
Footnotes
Acknowledgements
This reply has profited from comments I received when presenting the book’s theses at lectures at Comenius University (Bratislava), the San Tiago Dantas Graduate School in International Relations (São Paulo), the Universities of the Basque Country (Bilbao), Coimbra, Cornell, LSE, Minnesota, Montréal (UdM), Pavia, Stockholm, UNA (Universidad Nacional Costa Rica), as well as the Collegio Carlo Alberto (Turin), the Swedish Institute of International Affairs (Stockholm) and in particular from a workshop dedicated to its theses at the Université Libre de Bruxelles and the discussion following my Dasturzada Dr Jal Pavry Memorial Lecture 2016 at Oxford University. My special thanks go to Pinar Bilgin, Barbara Delcourt, Andrew Hurrell, Ian Klinke, Frédéric Mérand, Kacper Szulecki, and Sid Tarrow. I also wish to express my gratitude to Riksbankens Jubileumsfond for financing a sabbatical at Cornell University during which this reply was researched and drafted.
Notes
Author biography
