A genealogy of metatheory in IR: how ‘ontology’ emerged from the inter-paradigm debate

Abstract

Drawn from the philosophy of science, metatheory is generally defined in International Relations (IR) as ‘theorizing about theory’ by examining a theory’s ontology or epistemology. Yet, despite being available to IR since at least the 1940s, metatheory ‘exploded’ into IR only in the late-1980s. What explains this sudden proliferation of metatheory in IR’s literature? This article answers this question by conducting a genealogy of metatheory in IR. It begins by offering a four-step heuristic guide for conducting a genealogical analysis. Then, bracketing IR’s traditional historical narratives, it problematizes a tacit practice undergirding IR’s present use of metatheory: theorizing about the ‘world’ using the philosophy of science. Tracing and interpreting the transformations of this practice through scholarship from the 1940s to the present reveals how metatheory emerged unexpectedly from what is now considered to be an outmoded and forgotten event: the inter-paradigm debate (IPD) of the mid-1980s. The IPD transformed what had always been conceptualized as a single, dynamic world for IR theory, into plural, static, theoretical worlds best (meta)theorized through a concept of ontology drawn from scientific realism. In sum, this genealogy demonstrates the implicit power that explicit theoretical practices wield over IR scholars, tacitly conditioning the discipline’s conceptual possibilities and limits.

After occupying a prominent place in recent theoretical debates in the discipline of International Relations (IR), metatheory is now fading from the spotlight. Drawn from the philosophy of science, metatheory, or ‘theorizing about theory,’ elucidates how ontologies and epistemologies – ‘the stuff of metatheory’ – undergird the implicit theoretical boundaries, methodologies, and explicit and substantive content, of first-order knowledge in the discipline (Reus-Smit 2013, 590). For many scholars, metatheory is deservedly commonplace within IR’s literature (Jackson 2010; Kurki and Wight 2013, 14; Smith 2013; Chernoff 2013). Indeed, since the late-1980s ‘a wave of second order or meta-theorizing in the field’ (Wendt 1991, 383) has transformed IR’s ‘central questions – of the proper or appropriate character of IR theory and its value – [into] questions of meta-theory that can be answered only by considerations in the philosophy of the social sciences’ (Chernoff 2005, 2). Although considered by this group as an ‘indispensible foundation for competent scholarly activity’ (Neufeld 1993, 56), other prominent IR theorists lament and bracket this metatheoretical wave from their own consideration of ‘the real world’ (Walt 1998; Lake 2013). These critics argue that metatheory does not increase IR’s theoretical scope and power, but instead magnifies a widening gulf between IR theory and real-world political problems and practices (Wallace 1996, 304; Wesley 2001; Sil and Katzenstein 2010, 2; Lake 2011, 465–80). Monteiro and Ruby, for example, claim that metatheory attempts to reconcile IR’s theories upon ultimately irreconcilable conceptual foundations. This consigns IR to a Sisyphean, self-referential, and increasingly detached and self-indulgent ‘philosophy of science debate’ (Kurki 2009; Monteiro and Ruby 2009, 18). A recent forum into ‘The end of International Relations Theory?’ ¹ reiterated how this ongoing debate has, over the course of decades, fostered a growing atmosphere of ambivalence and detachment concerning grand and purely (meta)theoretical inquiry. As a result, calls for IR to move away from endless ‘paradigm wars’ to greener or more productive theoretical pastures grounded in hypothesis testing and middle-range theory testing (Lake 2013) or scientific realism (Bennett 2013; Walt and Mearsheimer 2013), are now being heeded. While in previous decades theorizing was ‘the core of the discipline of IR,’ today ‘this has ceased to be the case; the continuous bickering and flag-waving having in fact become an obstacle to it’ (Guzzini 2013, 522).

This raises an important question, however. Although it is becoming common for scholars to declare that IR’s paradigm wars are over and that the ‘wave of metatheory’ is now rolling back (Dunne et al. 2013), there is also a simultaneous awareness – and an acceptance among users or proponents of the philosophy of science – that although we ‘can bracket metatheoretical inquiry,’ notes Reus-Smit, this ‘does not free one’s work, theoretical or otherwise, of metatheoretical assumptions’ (2013, 590). We are all participants in the philosophy of science debate in some way. Yet, if IR is a discipline in which ‘Nothing seems to accumulate, not even criticism’ (Waltz 2010, 18), and the philosophy of science has been used sporadically in IR since the 1940s, then why did metatheory explode in IR only in the late-1980s (Wendt 1991; Chernoff 2005, 1; Smith 2013), as this ‘indispensable’ foundation, a ‘heart,’ and a ‘forefront of our analyses’ in IR (Wight 2006, 290; Jackson 2010; Arfi 2012, 191)? Even as it fades from explicit debate, might implicit (meta)theoretical assumptions still affect the discipline in ways undetected?

This article answers these questions in a new way, by conducting a genealogy of metatheory in IR. Rather than repeating and further entrenching commonplace and traditional disciplinary histories and narratives concerning IR’s philosophy of science debate, metatheory, and the theoretical development of IR writ large (e.g. see Holsti 1985; Schmidt 2002; Wight 2002; Jackson 2010), a genealogy works differently. Its value lies in providing an unfamiliar and unexpected account of how and why we have come to ‘think’ in the present – whether as a proponent or critic – about theory and metatheory the way we commonly or axiomatically do. Its target is our present moment, not the past. This ‘radical historicism’ works by establishing ‘delegitimizing, denaturalizing perspectives on the processes of subject constitution and construction’ (Saar 2002, 237). In other words, the point of a genealogy is to uncover the surprising history about the foundations of our own naturalized style of thinking. By revealing how everyday concepts previously assumed to be banal or without history actually emerged into being unexpectedly and contingently from forgotten problems and events, conclusions previously un-thinkable prior to the genealogy taking place, may thus emerge. A dual de-naturalization occurs here: first, of the historical narratives previously taken for granted; that there are always many differing perspectives, possibilities, and undetected threads of history to trace and explore. Second, of our ourselves in the present moment; we are forced to question how and why we think and act as we do, when concepts forming our sense of self and our world – assumed as immutable or natural – divulge their historical and random constitution. A genealogy asks anew how we are continuously constituted as historical subjects, and thus how we may think otherwise, rather than just looking back at how we have acted to constitute one uniform historical path.

The genealogy conducted in this article proceeds by diagnosing, tracing, and then interpreting the effects of an innocuous everyday practice that IR scholars now engage in when (meta)theorizing: using the philosophy of science to theorize and conceptualize the ‘real world’ or ‘worlds’ in their IR theory or theories. It carefully analyzes and interprets the tacit shifts and transformations in IR’s way of thinking and embodying this subtle (meta)theoretical practice in its literature. In doing so, a new historical understanding of metatheory’s explosion in IR emerges, through an event now forgotten or discredited by conventional disciplinary narratives as an ‘anachronistic’ (Patomäki and Wight 2000, 214) and ‘singularly pointless’ affair (Brown 2007, 409), yet what is revealed here to be an indispensible yet neglected transformation in IR theory: the inter-paradigm debate (IPD) of the mid-1980s (Banks 1984a; 1984b; 1985a; 1985b). Typically, the IPD’s tripartite paradigmatic framework of realism-liberalism-structuralism is dismissed as an obsolescent pedagogical tool (Dunne et al. 2013; Guzzini 2013, 531); a bygone ‘informative metaphor for telling the history of the discipline’ as a negligible stopgap between the second great, and rationalist-reflectivist and post-positivist, debates (Wæver 1997, 15; 1998). Although some discussion remains over whether or not there are paradigmatic wars still ongoing in IR today (Wight 1996; Jackson and Nexon 2009; Bennett 2013; Jackson 2013; Ringmar 2014), as well as over the proper or improper usage of Kuhnian terms such as paradigms and ‘incommensurability’ (Wight 1996; Wæver 1997; Guzzini 1998; Jackson and Nexon 2009, 2013), there is indeed a general acceptance that in the 1990s ‘the paradigms of the [IPD] dissolved’ into relative obscurity (Knutsen 1997, 277). Like the IPD, many now claim that metatheory has ‘gone into hibernation in the aftermath of the third/fourth debate,’ and thus scholars may once again gain ‘unmediated access to a real world’ through something akin to middle-range theories, analytical eclecticism, or scientific realism (Dunne et al. 2013, 415).

This genealogy reveals a surprising story about how IR arrived at its current (meta)theoretical impasse. Its key is not to debate the specific theoretical mechanics on the surface of each substantive theory, but to analyze the background conceptual contexts or grounds upon which they are formed. If we concern ourselves not with IR’s traditional narratives and historiographies, but with the subtle transformations and shifts visible in the everyday knowledge practices sustaining them, then on an implicit and conceptual level the effect of the IPD becomes immense. As will be argued below, the IPD was a unique conceptual event that transformed what IR had always hitherto theorized as a dynamic and singular ‘real world’ of competitive IR theories, into a sui generis and atemporal plurality of static and coterminous ‘metaphysical worlds’ of theory (Banks 1984a, xiii; Sandole 1984). In other words, IR’s background context of theorizing a single real world was suddenly and unexpectedly divided into multiple, coexistent worlds of theory. This new conceptual territory demanded forms of knowledge and comparison that could thus articulate, theorize, and differentiate between the fundaments of these plural yet intersecting worlds. It was a theoretical dialogue never before required in IR, thinking not in terms of what first-order grand theory would dominate, but in terms of theories about theories: meta-theories. It was at this moment that ‘ontology,’ a metatheoretical concept rarely needed nor used in IR before the IPD, suddenly acquired an unprecedented cogency and capability to demarcate and conceptualize IR’s contextual background of coexisting and ‘competing visions of how the world is and how it should be’ (Wight 2006, dustjacket). Only after the IPD, therefore, could the metatheoretical notion so common today, that ‘All theories presuppose a basic ontology’ (Wendt 1999, 2; Wight 2006, 4) actually make sense within the discipline. With this new style of thinking then naturalizing the use of metatheory, and with the steady importation of ‘ontology’ as first embraced in Bhaskarian scientific realism, the oft-noted ‘explosion’ of metatheory in the discipline after the late-1980s was facilitated. What this genealogy reveals, therefore, is that today’s philosophy of science debate, metatheory, scientific realism, and even IR’s widespread use of the concept of ‘ontology,’ are vestigial remnants or echoes of the implicit conceptual space opened by the IPD. It reminds us as scholars that, as theoretical practices might appear to fade away into the dustbin of history, they may still affect and orient how own thinking operates on an implicit level in our present moment.

This article proceeds as follows. First, it briefly reviews contemporary disciplinary histories proffered in IR in order to establish what general and commonplace assumptions concerning metatheory are prominent today. This fulfills the need for a genealogy to ‘start from an analysis of the present, which serves as the point of departure’ for a new historical interpretation (Bartelson 1995, 77). Second, this article offers a heuristic guide for conducting a genealogy in the style of philosopher Michel Foucault, outlining four crucial stages or steps in any genealogical analysis: problematization, practice, rationality, and emergence. Common to each step is the careful interpretation of the background ‘rationality,’ or the tacit style of thought undergirding the historical practice in question. Third, it engages in an empirical examination and interpretation of past theoretical practices in IR, spanning from the 1940s until the present. This careful tracing uncovers the unexpected materialization of the rationality supporting our everyday notions of metatheory today, in the event of the IPD. Finally, it concludes with a brief discussion of how the unique (metatheoretical) rationality emerging from the IPD opened the conceptual space for a specific variant of ‘ontology’ derived from scientific realism to become naturalized in IR, even as the IPD and later ‘paradigm wars’ appeared to fade away. In sum, this article hopes to promote the use of genealogical methodologies in researching IR, its disciplinary history, metatheory and ontology, and in future critiques and re-examinations of the temporal and conceptual limits within which all IR theories – and hence, even the arguments set forth here – are always admittedly posed.

IR’s textbook histories of metatheory

‘I am not a professional historian; nobody is perfect.’ (Michel Foucault, 1982 as quoted in Megill 1987, 117).

The social sciences have had a long and extensive engagement with the philosophy of science (Wight 2002, 26–29), and there are many historical explanations for why metatheory became incorporated into IR when and how it did. IR is an extremely broad and complex field, with a multitude of authors, topics, theories, journals, etc., always in flux. To delve into each is far beyond the scope of this article, which aims to trace and untangle only one thread of IR’s massive knot of interwoven practices. It remains important, however, to outline IR’s textbook narratives here, in order to situate the genealogy provided below and to highlight the significance of this article’s differing methodology and conclusions.

First, it is commonly asserted today that metatheory reflects how the discipline has become ‘increasingly unified by a self-conscious aim on the part of its practitioners to make it a ‘science’’ (Hollis and Smith 1990, 16). Metatheories contribute to IR’s ongoing desire to feel that ‘the field as a whole is progressing’ towards a ‘scientific status’ (Jackson 2013, 9–11; Knutsen 1997; Gunnell 1995; Schmidt 1998, 2002, 10). Second, as Wight points out, the problem of ‘naturalism’ has always concerned the desirability, definition, and applicability of science to IR, from the time of Thucydides, to Hobbes, and up to the present day (2002, 23, 27). Hence, the use of the PoS and metatheory today should not surprise us, since it keeps IR in sync with developments in the philosophy of science. Third, due to the positivist and post-positivist theoretical debates of the 1980s, IR scholars were ‘forced’ to incorporate metatheory into their work to guard against fears of the discipline’s theoretical fracture in the face of nihilistic post-positivism, post-structuralism, post-modernism, and Critical Theory (Lapid 1989; Schmidt 2002, 10,16; Wight 2002, 33; Reus-Smit 2012, 528). Fourth, Wight also offers a variety of justifications for IR’s recent engagement with metatheory, ranging from assertions that increasingly refined levels of disciplinary self-reflection have ‘inevitably’ taken IR theory ‘into the terrain of the philosophy of science’ (Kurki and Wight 2013, 14); that metatheory was required for the discipline’s foray into the agent-structure debate (Wight 2006; Hollis and Smith 1994); and that, because the philosophy of science ‘had not yet emerged as a sub-discipline of philosophy,’ it was only once it became more established that it could be usefully imported into IR theory (Wight 2002, 24–25). Alternative and common explanations for IR’s use of metatheory also include the introduction of influential publications to the disciplinary milieu, such as Allison’s Essence of Decision (Hollis and Smith 1990, 54, 147–55; Allison 1971), Waltz’s Theory of International Politics (TIP) (Wæver 2010, 314; Waltz 1979), Onuf’s World of Our Making (Wendt 1991; Onuf 1989), Hollis and Smith’s Explaining and Understanding International Relations (Wendt 1991), and Wendt’s Social Theory of International Politics (Gunnell 2011, 1458; Hollis and Smith 1990; Wendt 1999); the theoretical and substantive vacuum opened for new scholarship and theorizing catalyzed by the end of the Cold War (Monteiro and Ruby 2009; Knutsen 1997, 6, 277–80; Holsti 1985, 131); the erosion of overtly normative theory in the late-1980s (Guzzini 1998, 193); the concomitant rationalist-reflectivist or ‘fourth debate’ (Wæver 1997, 20; 1998); the rise of constructivism as a middle-way or ground (Adler 1997); and the search to replace the faltering ‘hegemony’ of realism’s paradigmatic yet anomalous ‘research programme’ (Vasquez 1998, 213, 240–2). Considered from today’s present standpoint, each of these explanations presents a cogent case for how and why metatheory entered IR when it did.

With so many historical narratives to choose from, why bother conducting this genealogy? What new value does this article provide, that these common and (generally) accepted narratives do not? In short, they are presentist accounts. Each offers a detailed history of IR’s past from the perspectival vantage point of our current ‘end of IR theory.’ Reaching backwards into the past, they embrace a top-down form of conceptualizing history in which the stability and importance of the referents selected at each inquiry’s outset – such as IR’s great events, debates, authors, texts, and even theoretical concepts themselves – are projected backwards in time, accidentally determining their developmental trajectory from today’s standpoint. This is conducted with the intention of shedding additional light on a referent’s checkered past, but without challenging the fundamental assumptions justifying its inquiry – and conceptual boundaries – in the present. However, as a bottom-up or nominalist tracing of practices, a genealogy works very differently. Rather than adding new twists to a linear historical narrative, it adopts a radical historicist methodology that inverts the focus of the narratives above, so as to become a history of the present ‘in terms of its past’ (Bartelson 1995, 7; Foucault 1991). Rather than looking backwards, a genealogy looks in a mirror. For any genealogist, despite our best intentions, our current historical (and subjective) perspective is never a neutral nor objective judge of past nor present experience or knowledge; it is akin to an a priori conceptual framework that orients and predetermines what forms and concepts of knowledge, truth, and practice, will be relevant for us – and for our historical inquiries – before any analysis can even begin (Hamilton 2014). This article’s genealogy, therefore, aims to make visible this subjective framework. It circumvents IR’s historical and subjective presentist bias by refusing to acknowledge the historical continuity of any present referents or concepts. It ‘has not its task to tell what actually happened in the past’ by describing or elucidating entire topics, disciplines, or worldviews, nor does it prescribe what is ‘good or bad’ metatheory from an assumedly learned, enlightened, or atemporal normative standpoint. Its task, rather, is ‘to describe how the present became logically possible’ by focusing ‘only on those episodes of the past which are crucial to our understanding of what was singled out as problematic in the present’ (Bartelson 1995, 8). By uncovering the forgotten and implicit episodes and problems of the past that could never be spotted by glancing backwards in time, and then by interpreting the transformations in our own knowledge and rationality that these moments engendered, this article offers IR a new and unexpected narrative regarding how and why we now think of IR, metatheory, and ontology, the way we do today. Histories of the past should not change, but histories of the present will transform and shift in accordance with the constitution of subjectivity that engages them in that moment. It this ethos to always question ourselves anew, rather than to reify a lost true past, that animates a genealogy.

Conducting a genealogy: problematization, practice, rationality, emergence

Most commonly associated in IR with philosopher and historian Michel Foucault, a genealogy is a way to analyze, historicize, and denaturalize commonplace or immutable practices, subjects, objects, referents, and truths. In showing how forms of knowledge and meaning that are considered to be universal or obvious are actually temporal and historical, a genealogy thereby opens the possibility for transcending the limits and boundaries that these concepts create in our thought. If an ahistorical universal is historicized and thus reconceptualized as being temporal and malleable, the possibility to think and do otherwise than what came prior to the genealogy, is created (Walters 2012, 118; Saar 2002, 233; Bevir 2008; Foucault 1991, 2000a). The essence of a genealogy lies in momentarily circumventing two fallacies orienting our historical and subjective perspective, and which were hinted at above when discussing the historical narratives of IR: presentism and finalism. Presentism occurs through the unintended projection of our contemporary discursive meaning(s) into the past, thereby mistaking present referents as being coexistent or contiguous with the past. Finalism occurs by accidentally (pre)determining or assuming a current referent as developing along a progressive or linear teleological path, from a kernel or germ in the past into its present and immutable form (Bartelson 1995; Dreyfus and Rabinow 1983). Unbeknownst to us, these implicit predilections invariably determine how our present knowledge and thought conceptualizes our world(s). Labeled here as presentism/finalism, it is from these twin fallacies that every successful genealogy must momentarily escape (Bevir 2008; Foucault 1991).

As Biebricher has noted, it is impossible to review every single one of the numerous studies invoking genealogy, and Foucault’s own inconsistent use of the term makes a genealogical method ‘difficult to pin down’ (2008, 365). Simply put, there is ‘no single genealogical method’ (Walters 2012, 7). That being said, this article will now offer a brief heuristic guide for conducting a Foucauldian genealogy, using four concepts that it argues are essential to genealogical analytics: problematization, practice, rationality, and emergence. As such, this article is not a declaration of what every genealogy was, is, should, or can be, but is a recipe or toolkit that scholars may draw upon in the future when conducting their own.

Problematization

A genealogy begins by selecting a commonsensical or naturalized truth – a referent from the present – to be ‘problematized’ (Foucault 2000b; Hoy 2008, 276–81; Koopman 2013). This involves the challenging task of considering its current, universal, or taken-for-granted form, as a contemporary solution to a forgotten problem. What we have to do with our everyday assumptions, or the ‘banal facts’ of our referent’s obvious truth, is to ‘try to discover – which specific and perhaps original problem is connected to them.’ (Foucault 1983, 210) The genealogist’s task, therefore, becomes ‘to grasp the general form of problematization that has made [the referent] possible’ for us to think about in the present, in its current ‘historically unique form’ (Foucault 2000a, 318). With no universals allowed, the initial question of a genealogy becomes: What problem once existed to have allowed this referent to be thought of, or to emerge into being, in the axiomatic way we think of it today? This article has selected metatheory in IR as a referent to be problematized. Hence, what underlying problem allowed metatheory to emerge in IR, and to assume the form(s) in which scholars understand it today? What is the forgotten question to which metatheory was once an answer?

Practice

Second, every genealogy should select an empirical practice within which its problematized referent is commonly embodied or materialized in thought. For Foucault, thought – or thinking – is always in process; an endless interplay of delimiting truth/falsity, right/wrong, good/bad, etc., and it is what constitutes our ‘knowledge’ about anything. Knowledge, as a collection of accrued thought(s), channels and manifests thought into ‘action,’ or behavior that becomes visible on the surface of a practice. A material, socially meaningful pattern or ‘system of action’ is thus inhabited by this ongoing interplay of thought and knowledge (Foucault 2000c, 201). In being enacted empirically ‘in and on the material world,’ a practice thereby makes visible the discursive background conditions that imbue its associated referents with meaning (Adler and Pouliot 2011, 4). A practice makes intangible knowledge visible. If interpreted carefully, therefore, a material practice may act as a window into the constitution of subjectivity underlying it. A genealogy thus proceeds through ‘the problematizations through which being offers itself to be, necessarily, thought – and the practices on the basis of which these problematizations are formed’ (Foucault 1992, 11, emphasis original). For example, when an IR scholar writes a text, they are engaging in a specific material practice, imbued with a particular form of immaterial truth or knowledge, that is manifested empirically in that text in response to a problem (Neumann 2002; Onuf 1989, 15; Smith 2013, 31). In IR, ‘theory can in itself be a form of practice, … [and likewise] all practice is predicated on the basis of some or other theory’ (Smith 2013, 31). A theoretical practice, recorded textually, is still a practice. It may be analyzed genealogically, because text is ‘logically prior to the objects with which it deals’ and may thus be treated autonomously (Bartelson 1995, 8).

Agreeing with Wight that privileging the traditional historical narratives outlined above over the (meta)theoretical debates themselves has ‘the effect of trivializing the debates and misses the point that there are real and causally effective patterns of disagreement within the discipline and beyond’ (2006, 1–2; Kurki 2009), this genealogy selects a (meta)theoretical practice that is common in popular IR texts engaging metatheory today: using the philosophy of science and metatheory to conceptualize and theorize a ‘real world,’ the world, or ‘worlds.’ For example, Jackson notes how (meta)theories in IR are now differentiated by their ‘philosophic ontology,’ or how researchers conceptualize what they study when they make ‘wagers’ concerning how they ‘[connect] to the world’ through a specific ‘mind-world hookup’ (2010, 37, 28, 29; Wight 2006. Also, for only a small sample, see Aradau and Huysmans 2014, 598; Monteiro and Ruby 2009, 23; Wæver 2010, 298; Smith 2013, 8; Wight 1996, 294; Hollis and Smith 1990, 40; Jackson, 2010, 2013; Brown 1997, 4). At first, this subtle practice of using the philosophy of science and metatheory to conceptualize IR’s world(s) might appear to have no history, or to be too banal or everyday to be worthy of analysis. Hence, and as will be demonstrated below, it is exactly this type of presentist/finalist assumption that makes it an excellent practice for a genealogical analysis, by teasing-out the tacit rationality sustaining and underlying its use.

Rationality

Prior to delving into past texts, a genealogist must always interpret what their chosen referent and practice’s contemporary ‘rationality’ is. Generally put, a rationality is the everyday background or discursive frame through which thought filters, elaborates, accrues, and organizes knowledge. It is the historically specific style of thought that orients the way(s) in which reality is rendered thinkable and debatable for subjects in a given time and place (Miller and Rose 2008, 16). A rationality determines the historical possibility of what surfaces in every practice:

One is not assessing things in terms of an absolute against which they [practices] could be evaluated as constituting more or less perfect forms of rationality, but rather examining how forms of rationality inscribe themselves in practices or systems of practices, and what role they play within them, because it is true that ‘practices’ do not exist without a certain regime of rationality (Foucault 2002b: 229).

Once a genealogist interprets the rationality undergirding their current problematized practice, they may then compare it with rationalities of the past. Without this careful and detailed historical comparison, a genealogical critique is impossible. ‘[W]e are thinking beings,’ Foucault claimed, and so we act ‘on the specific ground of a historical rationality. It is this rationality, and the life and death game that takes place in it, that I’d like to investigate from a historical point of view’ (2002a, 405). This is also what makes a genealogy a nominalist analytic capable of escaping the fallacies of presentism/finalism affecting IR’s dominant historical narratives. Prior to a genealogical analysis taking place, it is impossible to anticipate from one’s present standpoint where, how, and when, past rationalities will overlap, transform, and emerge into what we have mistakenly assumed is the natural or immutable referent we are so familiar with and have thus problematized. As noted above, a widespread and subtle practice engaged in by IR scholars is the use of metatheory to conceptualize IR’s world(s) (see Jackson 2010, 195, 212). There are common conceptualizations of a singular ‘world’ and plural conceptualizations of many ‘worlds,’ so that ‘One world, it seems, is not enough’ (Michels 2013). We may label the rationality supporting this practice as a rationality of worlds, since each metatheory or ‘philosophical ontology’ now ‘worlds’ in a distinct yet coexistent way (Jackson 2010, 37, 197). Thus, when did this rationality of plural worlds in IR emerge in practice, and in response to what underlying problem? What is the underlying conceptual connection between metatheory and a rationality of plural, ontological worlds?

Emergence

Finally, as the genealogist carefully documents, traces, and interprets the surface of past practices, the contents of the practice and the rationalities supporting them will initially seem obscure or even irrelevant to one’s presentist/finalist perspective. Why? Lost or foreign styles of thought will naturally appear as alien or banal when considered from one’s present(ist) vantage point. Yet, this step demands patience from the genealogist, for while they are parsing through these seemingly lost or irrelevant rationalities, an important and transformative ‘event’ should occur at an unanticipated or unexpected moment. This event is the unexpected ‘emergence’ of one’s own contemporary rationality into being. As Foucault stressed, this emergence is not an origin, which implies a metaphysical and primordially pure and truthful essence that ‘assumes the existence of immobile forms that precede the external world of accident and succession’ (1991, 78). Rather, detecting the emergence of one’s contemporary rationality should be embraced as a new opportunity for the analyst to interpret the complex circumstances surrounding this event. What forgotten problem of thought created the conceptual space that this rationality filled? What combinations of concepts fused together into this rationality? If it was once emergent and new, what contributed to its becoming ossified and naturalized until appearing as normal, universal, or unquestionable in the present? (e.g. see Walters 2012, 132; Foucault 1991; Foucault 2000c, 201; Koopman 2013, 15). By interpreting and uncovering the unexpected historical emergence of our present rationality, the genealogist thereby gains a new understanding of how their present subjectivity and perspectival standpoint came into being and became thinkable. Once these buried conceptual conditions are unearthed and explored, an enhanced capacity for a new analysis of the referent, practice, and a critique of the present concepts sustaining them, is facilitated.

A genealogy of metatheory in IR

Having now problematized a contemporary referent (i.e. metatheory in IR), selected a subtle and everyday practice found within it to trace (i.e. the use of the philosophy of science to theorize and conceptualize IR’s world(s)), and interpreted the rationality underlying this practice today (i.e. a plurality of worlds), we may now analyze past texts so as to trace when and how our own rationality emerged in practice. Considering the massiveness of IR’s literature, and in order to conduct an operable genealogy within the scope of this article, this genealogy begins in the post-WWII era when our problematized referent and rationality are found to be completely absent from the selected practice.

In 1949 Dunn noted that IR’s research methodology was shifting from forms of philosophic speculation dealing with the world as a whole, to a parceled and systematic form of observation and analysis akin to the natural sciences (1949, 83). If research methodologies captured these political divisions existing in the world, then IR as a ‘policy science’ could thus be established through the rigorous classification, logic, and testing, of verifiable hypotheses (1949, 92). Yet, difficulties in IR’s aspiring policy science began to emerge in the early 1950s when the internal ‘perspective’ of the IR scholar became recognized as a problem. Dunn had cautioned in 1949 that the ‘unconsciously acquired value patterns’ of the IR scholar meant that there was one external world, but many different ways or internal perspectives through which scholars viewed it (1949, 92). IR could only be ‘both a science and an art’ when it accounted for this problem of ‘perspectivism’: how and why ‘people … see the social world in entirely different ways’ (Icheiser 1951, 311). If not, IR would be consigned to the ‘most serious occupational disease of social scientists,’ the tacit ignorance of how the ‘unconscious’ values and biases contained within each perspective differed (1951, 316; Thompson 1955, 740). Hence, Wright claimed that IR had to emulate the science of ‘Einstein, Planck, and Heisenberg,’ because ‘A unified discipline of international relations implies a generally accepted conception of the human world’ (1955, 484). This uniform ‘picture’ was ‘the Weltanschauung accepted by an individual or group,’ and was best captured by an agglomerative synthesis of the mental ‘images’ unique to each IR scholar’s perspective (1955, 484, 492–95).

With this problem, a space opened for psychological decision-making theories aiming to unify these disparate psychological images and perspectives. ‘It is what we think the world is like’ from many imagined viewpoints, ‘not what it is really like, that determines our behavior,’ noted Boulding. ‘If our image of the world is in some sense ‘wrong’ … it is always the image, not the truth, that immediately determines behavior. We act according to the way the world appears to us, not necessarily according to the way it ‘is.’’ (1959, 120). The point here was to use the philosophy of science to aid decision-making theories in unifying disparate images and perspectives, into one objective world (McClosky 1956, 283). Comprising ‘a filter through which we pass our data’ in IR, a young scholar by the name of Kenneth Waltz highlighted three overlapping yet inherently ‘contradictory’ images in Man, the State, and War (1959, 10, 12). He also stressed that each ‘picture of the world’ was only partial and thus overlapping, but still examined ‘the same world, the same range of events,’ albeit with ‘a different ambit’ (1959, 6). Three internal images shared one concept of the world: ‘In a manner of speaking, all three images are a part of nature,’ wrote Waltz, because ‘… seldom does an analyst, however wedded to one image, entirely overlook the other two’ (1959, 160). The world was stable, external, and immutable, while scholars were not.

Interpreting the rationality underlying IR theory at this time, we see a disjuncture between many individual perspectives and images, and the conceptualizing of a singular objective or external world. IR scholars had not yet divided into factions based upon their respective scientific or philosophic images or theories, but shared the frustrations of a perspectivism that saturated the entire discipline with fractured viewpoints or mental representations. There was no philosophy of science divide nor debate in IR at this point. Even into the early 1960s, IR scholars expressed ‘caution against being excessively concerned with ‘theorizing about theory’ because this ‘peril’ presented ‘the danger of visualizing a forest without any trees’ (Claude 1960, 264; Singer 1960, 431). Discussions concerning ‘theorizing about theory’ referred only to the epistemological scope and scale of perspectival mental representations, and the aim to achieve a single ‘grand theoretical scheme, embracing the entire reality of international relations within an agreed intellectual framework’ (Claude 1960, 264, 265; Fox 1964). From philosophers of science commenting on IR (Kemeny 1960) to IR theorists (Fox 1967; Boulding 1967), discussions centered on avoiding pure or metatheory in IR, because, unlike the physical sciences, IR would always be a collection of fractured and subjective representations of reality. McClelland wrote disapprovingly that these perspectivist biases among IR scholars fostered ‘numerous individualistic conceptualizations about ‘realities’ and the study of those phenomena’ (1960, 306), attributing these multiple subjective perspectives to a lack of mental organization or ‘consubjectivity’: a common frame of reference for ‘reality-identification’ that formed ‘the total social reality’ of ‘the empirical world that we investigate’ (1960, 307, 331; Farrell and Smith 1967, vi).

Notably, although IR’s ‘second great debate’ has been described elsewhere (see Schmidt 1998; Schmidt 2002; Wight 2002), IR’s traditionalists and behavioralists shared the same underlying rationality: both theorized and conceptualized only a single, external, ‘real world.’ Bull accused behavioralists of hiding behind ‘a world of intellectual constructs’ to escape from ‘political reality’ (1966, 365, 370–2), while Kaplan claimed that, unlike the traditionalist’s ‘style of story telling’ (1961, 468), behavioralist models demanded a higher degree of sophistication to ‘fit’ the ‘real world.’ (1966, 1) Behavioralists lauded the logical positivism popular in the philosophy of science for its perceived capability to exchange the philosopher’s ‘worlds of [utopian] knowledge’ or thought, for the singular ‘empirical world of politics, that has been made possible by science and technology.’ (Leiserson 1967, 243, 244, 248. Also, see Alker 1966, 639; Hamlett and Neal 1969, 282; Thompson 1967, 150).

As the second great debate grew stale, IR scholars became frustrated with grand theories aiming to unify the discipline’s images under one ‘underlying conception of the world [that] has remained essentially the same’ (Holsti 1971, 169). Yet, the underlying rationality remained: a singular world. As Jervis wrote in The Logic of Images, to ‘investigate the world from the perspective of the projection of images’ (1970, 18) meant to turn away from superfluous philosophies of science and grand theory: ‘one of the best routes to international relations theory … [lies now] in the attempt to see what the world would look like if only a few dominant influences were at work.’ (1970, 16; Rosenau 1976, 1) Eschewing ‘grand theory’ to move inside the perspectives of the state or the individual, scholars could now view the world ‘within limited frameworks, paradigms, or hypotheses’ (Holsti 1971, 169, 172; Young 1972, 188). Allison bemoaned ‘The fact that almost as much time has been spent lamenting the lack of theory, and theorizing about theory in international relations, as has been invested in producing substantive work,’ implying ‘that the “discipline” is retarded’ (1971, 273). He theorized three self-contained and logical models, each acting as a ‘conceptual lens’ and ‘common frame of reference’ for how individuals behaved within national governments. Why? Not to unify ‘a full description of the world,’ but to ‘pick up pieces of the world in search of an answer’ (1971, 279, 249). Underneath the practice of IR, there was a rationality still conceptualizing one world, with no ‘metatheory’ as we know it today, being used.

As IR sought to theorize within disparate images or perspectives, the philosophy of science’s concept of the ‘paradigm’ became common. Popularized by Thomas Kuhn in 1962, despite frequent usage in American political science literatures (see Truman, 1965), paradigms were not widespread in IR until the mid-1970s (for early exceptions, see Fox 1967, 100; Morgenthau 1967, 26; Allison 1971, 32). Defined in 1974 by Lijphart, a paradigm was ‘a way of seeing and interpreting the world’ that combined a philosophical and metaphysical ‘world view’ with sociological sets of ‘scientific habits’ shared by research communities (1974a, 55, 43). After Keohane and Nye contrasted their ‘world politics paradigm’ against the ‘classic’ or ‘state-centric paradigm’ (1971, 332, 345), the paradigm slowly became the ‘conceptual framework’ that made ‘a real difference’ in determining how IR scholars conceptualized their ‘real world.’ (Young 1972, 188) Paradigms were now used to bring clarity to a discipline self-consciously ‘without a general concurrence on a paradigm that would serve to explain the changes that the international system has undergone.’ (Morse 1976, xvi) For instance, in a 1974 article tellingly titled ‘International Politics in the 1970s: A Search for a Perspective,’ Fagan and Puchala advocated a ‘new image’ of a security politics paradigm over previous ‘images of the world out there, [because images] produce fewer insights than they once did’ (1974, 247, 248). According to Lijphart, Kuhn’s description of scientific development entailed ‘normal science’ as securing a dominant paradigm’s worldview. But this process was dynamic, giving way to paradigmatic anomalies, competitions, debates, and eventually, replacement by a new dominant paradigm (1974a, 58). Lijphart proclaimed that a ‘traditional paradigm’ (i.e. realism) had held sway over IR’s puzzle solving from the days of Thucydides into the 1960s, but that now a competing worldview in the form of a ‘behavioral paradigm’ was predestining a ‘paradigm shift’ (1974a, 69). Being both ‘wider than and prior to theory,’ this paradigm shift would again restore a single conceptual mode of ‘seeing and interpreting the world’ for IR, quelling its longstanding theoretical and perspectival disparity (1974a, 43, 62).

Despite Phillips (1974, 187) noting that predicting a paradigm shift ‘is a difficult question, one which is not answerable directly by the philosophy of science,’ a boom of paradigmatic competition in IR then occurred. IR theorists competed intensely, inventing and prophesizing a variety of paradigms – from transnational, to state-centric, to cobweb – they claimed would soon assume the mantle of the discipline’s new dominant, ‘accepted,’ or ‘future paradigm’ (Rosenau 1979, 135, 143. Also see Leurdijk 1974, 54; Banks 1978; Rochester 1978). Even with this increasing paradigmatic conflict and competition on the explicit surface of IR theory, like the 1940s, 1950s, and 1960s before it, in the 1970s IR’s underlying rationality remained the same. From Reynolds describing how the role of the philosophy of science in IR was to determine that ‘reality’ and ‘validity’ is found in and ‘consists of our world-view’ (1973, 48), to systems thinkers taking opposite epistemological stances on the ‘organized complexity’ of reality, asserting that systems were not mere ‘mental abstractions … [because] systems exist in the real world’ (Little 1977, 281), IR’s theorizing remained within a rationality of a single ‘real world.’ Paradigms simply warred on its surface to claim it.

By 1980, many scholars anticipated that IR was on the verge of transitioning to a ‘post-behavioralist’ paradigm that would finally return IR to a uniform and normal science (Mitchell 1980, 37; Little 1980, 24; Lijphart 1981, 233, 238–45; Banks 1979, 266). When this post-behavioralist paradigm failed to materialize, IR became a ‘discipline now in a state of flux’ because each competing paradigm asserted ‘a particular view of the world’ as sacrosanct (Little 1984, 7; Holsti 1985 vii, 1, 4–5). In an almost frantic mission to ‘search for a dominant paradigm’ (Mitchell 1980, 42, 44), scholars combined them into patchworks of various frameworks and syntheses, combining structuralist, pluralist, and world society paradigms (Pettman 1981, 39, 47), updated versions of Martin Wight’s three traditions of realist, rationalist, and revolutionist paradigms (Weltman 1982, 40, 41), and ‘cross-paradigmatic fertilization’ between traditional, behavioral, and radical/dialectic paradigms (Alker and Biersteker 1984, 125, 136). The aptly-titled In Search of Theory cautioned IR not to hastily replace its faltering realist paradigm, because it was ‘better to have no map of the world than to be guided by one that is badly flawed’ (Mansbach and Vasquez 1981, 487; 1983, 149, 153). Even up to the mid-1980s, therefore, these confused paradigmatic battles embodied a rationality that used the philosophy of science to conceptualize a single world, now waiting to be claimed by the winner of the paradigmatic war to become IR’s dominant future-paradigm. ‘Where is all this [paradigmatic competition] taking the field? Will it remain characterized, and perhaps paralyzed, by multiple definitions of ‘reality’? Will one of the candidates for the throne succeed either in replacing or co-opting realism?’ (Sandole 1985, 222; Brown 1981) A question raised here, interestingly, is why these bipartite and tripartite frameworks did not transform IR’s theoretical practice of theorizing the world through paradigms, into that which undergirds metatheory today: a style of thought conceptualizing plural worlds. Should not, for instance, have Maghroori and Ramberg’s claim that IR’s third great debate was occurring between globalist and realist paradigms (1982, 14), have catalyzed an IPD, or spurred a transformation in IR’s theoretical rationality towards metatheory?

A genealogical analysis ‘means making visible a singularity at places where there is a temptation to invoke a historical constant, … an obviousness that imposes itself uniformly on all.’ (Foucault 2002b, 226) In our case, this transformative singularity, or what Foucault also described as the ‘event’ sought by every genealogical tracing, emerges unexpectedly in 1984. Like Little, Holsti, and many others, Banks believed the realist paradigm was ‘crumbling,’ but that ‘Arguments over [paradigmatic] anomalies caused more confusion than confidence.’ While agreeing with Maghroori and Ramberg’s claim that IR’s third great debate was occurring (1984b, 15), Banks saw their thinking as too narrow. Each paradigm ‘intended to convey a world view more basic than theory,’ with different actors, forces, and effects (1984b, 15), and hence the fundamental tenets of each paradigm could never be merged to form any super-paradigm to claim IR’s throne. They were simply ‘incommensurate’ (1984a, 15, 1984b, 19, 1985a). So, instead of continuing IR’s longstanding pattern of prophesizing and fomenting dynamic and grand theoretical transitions into an imminent paradigmatic unification or a normal science, IR had to change how theories related to one another. The ‘appropriate judgment to be made is not that of deciding which [paradigm] is right and which is wrong. Rather it is a matter of recognizing how and why it is that thinkers within each paradigm belong to different mental worlds’ (Banks 1984a, xiii). Crucially, the temporality of theorizing these worlds now changed: it was not a dynamic race for conceptual domination, but a static IPD. With this bizarre claim, Banks posited paradigms as plural theoretical worlds coexisting in a safe and atemporal dialogue, rather than as the aggressive and zero-sum competitive race to replace realism with one grand winner as IR’s new, dominant paradigm. Rather than one theory or paradigm being right and another wrong, the IPD’s practice of theorizing coexistent yet ‘different metaphysical worlds’ (Sandole 1984, 42) now transformed one of the most important problematics driving IR theory since the 1940s – namely, ‘What was to be the newly enthroned general theory [paradigm] of international relations?’ (Banks 1984b, 13) – into a new type of theoretical problem and practice. With each coexistent paradigmatic worldview now frozen in its place, plural theoretical worlds became, suddenly, paused in a draw. How, therefore, could IR conceptualize, theorize, and delimit the fundaments of plural metaphysical worlds? (See Banks 1985a, 10–11; 1985b, 219–220; Hoffman, 1987; Whitworth, 1989; Lapid, 1989; Smith, 1989; Smith, 1992; Wight, 1996).

Although glossed-over and even ignored by the contemporary histories and narratives noted above, it is here that this genealogy detects the emergence of our contemporary theoretical rationality of plural worlds. The IPD was the germ, the catalyst, of IR’s present use of metatheory. As Goodman wrote, ‘by constructing a static image’ from dynamic experience, such a new pattern of ‘radical ordering’ ‘participates in worldmaking’ by opening new conceptual boundaries or spaces (1978, 13, 14). ‘The movement is from unique truth and a world fixed and found to a diversity of right and even conflicting versions or worlds in the making’ (1978, x). After the IPD occurs, we see IR scholars grappling with its new, underlying problematization of theorizing plural worlds, in new ways. From their own standpoint in the 1980s, these scholars were not able to fully articulate how or why they were doing so at the time. ² For example, Holsti lamented how ‘Meta-Theoretically-Driven-Theory’ suddenly arose from new ontological and epistemological considerations in the late-1980s, leading to ‘dead ends’ because of a failure to enhance a basic understanding of substantive international relations (1989, 260). Yet, ‘The field [of IR] is maturing,’ Holsti also asserted, ‘because there is an increased recognition and acceptance of multiple realities, and hence of multiple theories’ (1989, 260, 261). Similarly, while Onuf wrote that his work was ‘a tentative first step toward that new [disciplinary] paradigm’ (1989, 22), he constructed his ‘world of our making’ through the nominalism of Goodman, who also asserted that ‘no one ‘world’ is more real than others. None is ontologically privileged as the unique real world’ (1989, 37). Indeed, Onuf’s World of Our Making is a prime example of a scholar’s explicit rejection of the IPD’s substantive content, while answering its implicit problematization. ‘My position, and Goodman’s, as I understand it,’ Onuf claimed, grants ‘existential standing to plural worlds’ because ‘We construct worlds we know in a world we do not’ (1989, 37, 38). IR’s image of one fixed world was now a plurality of worlds for IR theory to engage.

As this new theoretical rationality took hold, an explicit theorization of multiple realities and worlds established itself in IR’s literature (e.g. see Wight 1996). With it, a new type of theory entered IR: theories about theoretical worlds, or metatheory. Guzzini, one of the few IR scholars in the 1990s to note that the IPD ‘provided the opening round’ of metatheoretical discussion in IR, wrote that ‘its main effect was to freeze one historical moment in the development of the discipline into allegedly immutable categories’ (1998, 109). However, Guzzini joined the chorus of scholars and historians discounting the IPD because its substantive content – its tripartite framework – fostered a theoretical conservatism by legitimizing ‘business as usual,’ reifying rigid tripartite categories that misused the concept of Kuhnian ‘incommensurability’ (1998, 191–3; 1993, 446; Holsti 1985). ³ Ironically, Guzzini also admitted that he was unable to explain the resilience of this ‘seemingly irreducible’ IPD in the literature, because its being ‘riddled with conceptual difficulties’ should have rendered it facile or quickly supplanted by more accurate metatheories (1998, 118). It was declared ‘to be pretty well over’ by 1989 (Windsor, 1989, x), and references and pontifications concerning the IPD, and the eventual domination of IR theory by a single paradigm, were disappearing as well. In their place, scholars such as Lapid (1989), Holsti (1989), George (1989), and Smith (1989) now debated the familiar historical narratives outlined above, discussing ‘paradigmatism’ or perspectivism as facilitating relativistic post-positivist scholarship and normative theories (Guzzini 1998). Onuf declared that he ‘did not believe a paradigm theory for IR is forthcoming’ and that he did ‘not believe it was possible’ (1989, 14). Slowly fading from general discussion, the IPD became known as a paradigmatic pluralism that IR was consigned to live with (Wæver 1996, 155; 1998. Also, see Guzzini, 1993, 1998; Wight, 2002; Schmidt, 2002; Jackson, 2008, 2010). Today, it has been largely shelved; relegated to textbooks as a bygone pedagogical aid, or simply ignored and forgotten (Guzzini 2013).

Looking back at IR’s past practices from today’s presentist/finalist vantage point, there appears to be no question that the IPD and its subsequent paradigm wars have largely disappeared from explicit disciplinary debates. Bennett claims they were ‘ultimately distracting and even counterproductive,’ and hence ‘there is a widespread sense that [theoretical] progress has arisen in spite of inter-paradigmatic debates rather than because of them’ (2013, 460). Even in the 1990s, critics lambasted the IPD’s ‘notion of ‘different worlds’ … [as] something of a de rigueur’ in IR theory, with its erroneous use of Kuhnian incommensurability dooming it through a ‘foundational fallacy’ privileging a nominalist relativism over the epistemological certainty of one material world (Wight 1996, 294; Wendt 1998, 107). Yet, from a presentist/finalist vantage point, what these narratives omit is how their critiques take place on the implicit conceptual grounds that first congealed and emerged in the IPD. This event formed and grappled with a new conceptual problem hitherto unexplored in IR: how best to think and to compare the theoretical fundaments of a static plurality of theoretical and metaphysical worlds. As will be explored below, not even Banks (nor detractors of the IPD, and later, metatheory itself) were aware of this transformation in IR’s background or everyday rationality at the time. Instead, while the explicit contents of the IPD’s tripartite framework faded and were even ridiculed (Wight 1996), and the discipline’s embrace and later allergy towards metatheory developed, the rationality underlying both the IPD and metatheory evinced tremendous theoretical changes occurring implicitly in the discipline. As theorizing plural theoretical worlds demanded theories about theories, a specific metatheoretical concept pertaining to theorizing world(s) became embraced and naturalized. It is, perhaps ironically, what is de rigeur in IR today: ontology.

Ontology: filling the gaps between the IPD’s worlds

IR’s present use of metatheory is associated with the practice of theorizing the basic fundaments of IR’s world(s) using the concept of ontology (see Jackson, 2010). ⁴ ‘It is almost unthinkable among social scientists nowadays to envision a formulation of social theory that does not posit an ontological beginning point … that one takes to be the foundations of the (world-)view being explored or posited’ (Arfi 2012, 191). In other words, ontology is now a common, obvious, or uncontroversial term in IR. Its use – both in (meta)theoretical discussions, and in common parlance – is fairly ubiquitous (for exceptions to the term, see Chernoff 2005; Gunnell 2011). Ontologies identify the ‘building-blocks and fundaments’ underlying every referent and theory in IR (Guzzini 2013, 533; Jackson, 2010). They describe the being of entities within our theories, and thus how these beings exist and relate to one another. ‘[W]e all have ontologies,’ notes Wendt, because ‘we all make assumptions about what exists in the world: dogs, cats, and trees’ (1999, 22, 370). These ontologies order our different ‘visions of how the world is and how it should be’ (Wight 2006, 2). In a recent commentary on the status of metatheory in IR, Reus-Smit notes that ‘ontology is not considered an intractable mire in the same way as epistemology,’ reflecting the tendency of IR scholars ‘of diverse persuasions advocating a focus on ontology over epistemology’ (2013, 595). Ontology has thus secured an inoffensive place in IR’s lexicon. Indeed, to read that ‘Theories are ontology-building, ontology-defining, and [that] claims regarding the uniqueness or otherwise of ontological statuses must be theoretically based,’ expresses nothing controversial in IR today (Ringmar 2014). Yet, what explains the apparent normality and widespread use of this concept? It is easy to attribute it to the rise of metatheory since the late-1980s, in that ‘a specific trend has evolved which advocates a renewed focus on matters of ontology as a way to overcome or at least to reconceptualise the divides within the field of IR’ (Michels 2009, 397). However, if IR theory has a penchant for borrowing concepts from the philosophy of science (Gunnell 2011, 1457), and if ontology was commonly used in philosophical texts for centuries, and in the philosophy of science since (at least) the 1940s (e.g. see Margenau 1941; Bergmann 1950; Maxwell 1962; Hanneborg 1966; Giere 1973), then why was ontology very rarely used in IR prior to the IPD? For instance, Waltz’s (1979) brilliant exposition of the philosophy of science in TIP reminds us that discussions of ontological beginning points or divisions were not necessary nor prevalent at this time. But, should not ‘ontology’ have featured prominently in the paradigm wars of the 1970s–80s, and especially during the comparisons of theoretical worlds occurring in the IPD itself? Upon analysis, the concept is absent. This is despite the fact that, prior to the IPD, there were countless situations that would have warranted and demanded the use of ontology by IR’s current standards. For a small example: perspectivism’s ‘multiple realities, multiple perceived’ were structured by ‘taxonomic classifications,’ but not ontologies (McClelland 1960, 320). Young claimed that every IR scholar ‘views the real world in terms of some conceptual framework … [yet determining] what a person regards as worth explaining’ was simply impracticable (1972, 188). Presaging the recent work of Guzzini and Jackson almost verbatim, Mitchell exclaimed during the paradigm wars that IR’s puzzle was finding the ‘fundamental’ ‘conceptual element establishing the basic units of which reality is deemed to consist,’ through the ‘picking out and labeling [of] particular phenomena to serve as basic building blocks for the development of theoretical assertions about the nature of … the referent world’ (1981, 40, 41, 3–40. Also, see Lijphart 1974b; Smith 1979, 241-242). Even Banks declared during the throes of the IPD that ‘The most confusing usages [within each paradigm] occur at the level of specific concepts – the building blocks of any theory’ (Banks 1984a; 1984b; 1985a, 13; 1985b). Still the concept of ontology was never invoked by these scholars, even when their contentious debates were steeped in the philosophy of science for decades while comparing the fundaments of worldly paradigms. ⁵ Why this past ontological silence when the concept roars so loudly today?

As Smith once wrote, IR’s ‘silences are often its most significant feature. Silences are the loudest voices’ (1995, 2). Ontology was silent in IR until the IPD transformed its everyday rationality of theorizing, by opening a conceptual gap between static worlds that required theories about the constitution and interrelation of theoretical worlds – metatheory – to fill it. What concept best theorized this new metatheoretical problematique concerning the differing fundaments or building blocks between atemporal ‘metaphysical’ worlds? As Roy Bhaskar wrote, ‘the philosophy of science abhors an ontological vacuum’ (1975, 30), ⁶ and so the metatheory of scientific realism and its concept of ‘ontology’ entered IR in ways never before possible (see Holsti 1989; Dessler 1989; Lapid 1989; Wendt 1987). As noted by Gunnell, in 1987 Wendt became the first IR scholar to successfully entrench metatheories concerning ontology into IR using scientific realist concepts of ontological differentiation (2011, 1458), with the ultimate aim of elaborating the fundamental ontological differences between theories. Giving credit to Maclean’s (1981) book chapter for being one of ‘the only discussions of scientific realism in international relations’ up to that point, Wendt stressed the importance of researching the ‘ontologically primitive units’ of any agents, structures, or theories under analysis because the structural theories he was discussing were most ‘strongly conditioned by a more fundamental difference of ontology’ (1987, 336–337). Wendt was closely followed by Dessler’s scientific realist argument that ‘a theory’s ontology (the substantive entities and configurations a theory postulates) is both the basis of its explanatory power and the ultimate grounding of claims it may have to superiority over rival theories (1989, 444). Coming prior to the IPD in 1981, Maclean’s use of scientific realism’s metatheoretical ontology fell flat when placed within a background rationality of one shared world. Ontological differentiations were just not relevant at this time. After the IPD, however, claims that scientific realism opposed ‘Kuhn’s view that paradigms create “different worlds”, because IR scholars ‘hook on to the world’ through meaning accrued by language regulated and determined by existents in a single, ‘mind-independent, extra-linguistic world’ (Wendt 1999, 53–7), were ideal solutions to IR’s new theoretical problematization. Styles of ontological metatheorizing, as embraced by scientific realism, capitalized on this implicit conceptual vacuum of plural worlds. It exploded in IR scholarship thereafter, and throughout the 1990s, to the present day (for only a small sample, see Wendt 1987, 1991, 1998, 1999; Dessler 1989; Hollis and Smith 1990, 1994; Schmidt 1994, 357;Chan and Jabri 1996; Wight 1996, 2006; Smith 1995; Jackson 2008).

Although this literature is massive and cannot be summarized here, since the 1990s, IR’s substantive and explicit theoretical content grew tired of compartmentalized paradigmatic debate. The philosophy of science was soon used to gauge the perils and possibilities of paradigmatism, theoretical synthesis, pluralism, eclecticism, and middle-range theories – yet, all while indirectly taking into account the ‘metatheoretical challenge’ setting forth ontologies, epistemologies, and methodologies, for thinking both about a world, and the ‘different worlds’ of IR theories (Jackson 2008; Smith 1995; Michel 2013). The point is that not that all IR scholars now using ‘ontology’ are avid Bhaskarians, nor scientific realists. The point is that as the IPD faded from explicit discussion, its underlying rationality – its style of thinking IR – became increasingly entrenched, as evidenced through IR’s widespread use of ontology and (metatheoretical) worlds today. To refer to scientific realism or to ontology, therefore, is to incidentally summon ghosts of the IPD. For instance, Bennett laments how ‘the IR field’s lingering paradigm wars are another manifestation of out-of-date and problematic views on the philosophy of science’ that are distracting and counter-productive. Yet, for Bennett, ‘The way out of the cul-de-sac of paradigms and research programs lies … in a parallel development in the philosophy of social science in the last several decades: the scientific realist turn toward theories about causal mechanisms as the locus of scientific explanations’ (2013, 462, 465). The irony here is that IR’s contemporary use of ontology and its proclivity to embrace scientific realism as a way out of IR’s ‘sterile, second-order’ metatheorizing, is rolled back upon itself to attack the very conceptual foundation that first established it, and upon which it still stands. To cognize and even to reject metatheoretical debates and concepts by recognizing and theorizing about world(s), ontology, scientific realism, or metatheory, highlights the existence of, and dependence upon, the tacit and shared underlying problematization and rationality that emerged in the IPD. Despite its de rigeur or normalized status today, ‘ontology’ did not fall from theoretical heaven as a pure concept referring to being(s). It is dependent upon an implicit and shared background rationality that allows it to be articulated and made mutually intelligible. Hence, regardless of what side of ontology or the philosophy of science debate one stands on, scholars such as Wendt and Wight, or Smith and Jackson, could not debate metatheory nor ontology without sharing these conceptual rules of the game.

By tracing the emergence of this knowledge, this genealogy has revealed not its permanence or immutability but the recent and unexpected twists it has taken in IR as incidental spinoffs of the IPD and Bhaskarian realist philosophy. For example, Jackson’s noted claim that the ‘contemporary philosophical usage’ of ontology is divided between scientific and philosophic ontologies (2010, 28), is not universally shared across the social sciences, as might be assumed. Instead, it is derived from a specific IR article by Patomäki and Wight (2000), who – for reasons examined above – took this division and definition of ontology from Bhaskarian scientific realism. ‘[V]iewed from an ontological perspective current understandings of IR take on an altogether different hue,’ they write, to be engaged through the ‘benefits of metatheoretical inquiry to IR’: an explicit opposition to an implicit ontological problem of a relativism of worlds (2000, 215, 216). It is thus no wonder why scholars such as Brown note that ‘the exchange between our discipline and the rest of the social/human sciences is pretty much one way, and not in our favour’ (2013, 484). Theories and concepts we in IR assume to be universal, innocuous, or without history – such as ontology – contain their own implicit boundaries, limits, and histories. They have been shaped and steered in unique but forgotten ways that affect what can or cannot be made thinkable when they are cognized and used. ⁷ Agreeing with Brown, therefore, IR’s present proclivity towards a form of ‘ontology’ that has Bhaskarian roots and has since been developed and shaped specifically within IR, may accidentally encourage conceptual isolation rather than inter-disciplinary conversation. Indeed, it is in revealing these surprising facets of the history of our present constitution that genealogical analysis adds its value. Where IR goes from here is not for a genealogy to declare, although it opens unexpected paths that future researchers may explore.

Conclusion

As the recent philosophy of science debate illustrates, IR scholars have grown accustomed to, and perhaps even weary of, metatheory. Is it a passing fad of theorizing, or the latest manifestation of IR’s endemic science-envy? Is it a conceptual hangover from the post-positivist debates of recent decades, now fading into a middle-range pluralism or an analytic eclecticism? Or, is metatheory now an inexorable and unavoidable aspect of every IR theory, regardless of whether scholars choose to acknowledge its presence or not? With IR’s so-called ‘great debates’ presaging its future, the philosophy of science debate may soon be collecting dust as a brief theoretical skirmish within the discipline’s long developmental historiography. Despite where one stands, debating first-order problems of IR through second-order lenses of ontology and epistemology appears now to fall within the specialized purview of avid meta-theorists, and not of IR researchers generally. As international crises, thinkers, and the discipline itself moves on, so goes metatheory. … or, so it might appear to us from within our present perspective.

Beneath the surface of IR’s receding metatheoretical wave rests a much more complex and tangled conceptual history. It is a history that shapes the present and everyday concepts that are knowable and commensurable for all IR scholars – from students, to researchers, to (meta)theorists alike – by providing the background contexts, concepts, theories, and forms of knowledge, that make IR thinkable and communicable. No concept or practice in IR is immutable or universal, even those appearing banal or without history at first. These, too, result from forgotten paths tread to answer problems of thought hidden long ago. Yet, they may still be problematized and traced from our present vantage point so as to tell us something about how they came to be, and how they affect us in important ways. IR is a broad field, woven together by a patchwork of such implicit concepts and knowledge practices that have congealed and concretized over time into the discipline lying before us today. The point is not that there is a single historical thread that ties it all together, nor one true ideal or objective history of a concept or practice that we must unearth for all posterity. Rather, with a genealogical eye, the goal is to shed new light on how the concepts making our present thinkable and knowable at all, have formed. The value here is to understand how we have come to think the way we do in the present moment, and to accept with humility that this moment will change. Just as thought is always ongoing and in process, and thus changes, so does its history.

With this genealogical ethos in mind, this article has aimed at untangling only one of IR’s many theoretical threads. After outlining four crucial steps in every genealogical analysis, it applied these to IR’s philosophy of science debate by conducting a genealogy of metatheory in IR. Rather than reading our present and metatheoretical concepts backwards in time to focus on IR’s faithfulness or accuracy regarding the contents of theories of the philosophy of science, or revisiting the quality or worth of great debates or moments celebrated in our textbooks, it eschewed this presentism/finalism. Instead, it engaged in a careful and patient nominalist historical analysis of a seemingly obscure or irrelevant practice: theorizing the world(s) using the philosophy of science. This yielded surprising results that were both unanticipated and opposed to the taken-for-granted concepts and histories of IR’s accepted historical narratives. The style of thought, or background context for knowledge – the ‘rationality’ – undergirding metatheory and the philosophy of science debate today, emerged from the ridiculed and forgotten event of the IPD in the mid-1980s. The new and implicit problem the IPD posed to IR scholarship – theorizing and comparing the fundaments of static and plural metaphysical worlds – opened an unparalleled conceptual space that metatheory, and the scientific realist variant of ‘ontology,’ then filled in ways never before possible. This tacit style of thinking then became entrenched in IR’s literature, as today’s widespread use of ontology indicates. Even as the IPD itself faded from discussion, therefore, the original problem of knowledge connected to it provided an implicit foundation for new theories and knowledge practices to develop.

In sum, metatheory did not ‘explode’ into IR because of an inexorable or steady scientific development in the discipline or in the philosophy of science, nor even from IR’s longstanding aims to theorize the building-blocks of other theories. Instead, metatheory entered IR contingently and unpredictably as an explicit answer to the implicit problematization set out by the IPD. We must, therefore, be forever cautious of our tendency to reflect upon our present theories, histories, and concepts, with a presentist/finalist lens that ignores their inexorable historicity and contingency. We are historical beings, and so we must analyze and diagnose the nature of our present by problematizing ourselves, as solutions to problems of thought that remain buried; not in a lost past, but within our everyday practices and rationalities in the present. Through this article’s genealogy of metatheory in IR, and the heuristic guide it has set forth for conducting a genealogical analytic, it is hoped that new and critical ways of thinking about IR’s theoretical practices, theories, debates, and concepts, may be historicized and reconceptualized in new ways going forth. All of our thought and theory in IR accrues and ossifies on the sediment of such ghostly echoes and forgotten problems. It is up to future IR scholars to investigate their effect on us by taking nothing for granted. The present moment is always continuing and transforming, and hence its history must always be pursued anew. This resilient questioning of ourselves is the painful but rewarding price to stave off complacency and ignorance in the face of time. Indeed, as in the case of IR after the IPD:

With false hope of a firm foundation gone, with the world displaced by worlds that are but versions, with substance dissolved into function, and with the given acknowledged as taken, we face the questions [of] how worlds are made, tested, and known.

(Goodman 1978, 7)