Concept formation in historical International Relations

Abstract

For a long time, International Relations’ (IR) spatiotemporal parameters could be summarized in two words: Western and modern. The recent growth of Historical IR has contributed to a dramatic extension of these spatiotemporal coordinates, extending the discipline’s gaze to places and times that had previously been neglected. A key challenge that this ‘historical turn’ presents for IR theory, and that we take up in this article, is the proper formulation of analytical concepts capable of ‘travelling’ across different places, cultures, languages and time periods. So far, the existing Historical IR literature on concepts has focused primarily on the historicization of existing categories (especially through conceptual history and genealogical methods) and has not yet addressed the problem of concept formation head-on. In this article, we take up this challenge by theorizing the role of ‘travelling concepts’ in Historical IR – concepts that are neither entirely bound to their specific context nor entirely divorced from it. Returning to Max Weber’s work on ideal types, we suggest that travelling concepts are not only possible for historicist work, but necessary, and that it is therefore important for Historical IR scholars to be open and reflexive about their use of such concepts. By rejecting essentialism and recognizing the historical situatedness of the scholar, Weber encourages us to view the entire lexicon of Historical IR as an evolving archive for conceptual historians of the future to explore.

Introduction

For a long time, the spatiotemporal parameters of the field of International Relations (IR) could be summarized in two words: ‘Western’ and ‘modern’. The recent growth of Historical IR ¹ has contributed to a dramatic extension of these spatiotemporal coordinates: global histories have centred non-Western actors and world orders as legitimate objects of analysis (e.g. Kang, 2010; Neumann and Wigen, 2018; Pardesi, 2017; Phillips, 2021; Phillips and Sharman, 2015; Spruyt, 2020; Zarakol, 2022); disciplinary histories have unveiled the plural origins of IR as an intellectual field (e.g. Long and Schmidt, 2005; Thakur and Vale, 2020; Vitalis, 2017); and work on medieval and ancient history has extended the discipline’s gaze backward by centuries (e.g. Costa Grynaviski and Steinsson, 2023; Lopez, 2020; Mukoyama, 2023; Zarakol, 2022) or even millennia (e.g. Neumann and Glørstad, 2022; Powel, 2024). All told, the ongoing ‘historical turn’ has contributed to a radical diversification of the regions and times that IR scholars study.

Besides expanding the purview of the field, the ‘historical turn’ has unsettled the concepts through which IR scholars traditionally organized and theorized about their subject. Genealogical inquiries and conceptual histories have denaturalized and provincialized many of the analytical categories that mainstream IR has viewed as universally applicable, including core concepts such as ‘state’, ‘sovereignty’ and ‘anarchy’. At the same time, the inclusion of a wider range of historical case studies has led scholars to redefine existing terms in different (and sometimes contradictory) ways. As a result, some worry that the Eurocentric vocabulary that had provided a shared reference point for disciplinary dialogue for much of the 20th century is being displaced by a cacophony of historically and regionally specific conceptualizations. Commenting on these trends, Butcher and Griffiths (2017) memorably describe IR as caught ‘between Eurocentrism and Babel’.

Such concerns are not new to IR or political science. ² More than half a century ago, in the neighbouring field of comparative politics, Giovanni Sartori (1970) called it ‘the travelling problem’. The scope of comparative politics in the middle of the 20th century was rapidly expanding, similar to the ongoing globalization of IR. The number of newly independent polities had multiplied, and political participation had widened. Sartori noted then that comparative politics faced a pressing need for ‘conceptual tools that are able to travel’ (Sartori, 1970: 1034). ³ The existing political vocabulary, formulated to describe the political experience of the West, ‘was not devised for world-wide, cross-area travelling’ (Sartori, 1970: 1034). The simplest solution, of course, was to broaden the meaning of existing concepts to make them more inclusive. But Sartori worried that simply expanding the range of application of existing categories would lead to ‘conceptual stretching’ or ‘vague, amorphous conceptualizations’ that undermined the ability of political scientists to develop robust generalizations (Sartori, 1970: 1034).

Sartori’s diagnosis back then is reminiscent of Butcher and Griffiths’ (2017: 329) aforementioned more recent concerns about the ‘conceptual ambiguity’ and ‘contradictory definitions’ generated by the inclusion of new historical cases within IR’s horizon. Both also offer similar solutions to the travelling problem, focused on the standardization of concepts. Thus, Sartori (1970) writes that political scientists must establish ‘disciplined use of terms and procedures of comparison’ (p. 1052). Similarly, Butcher and Griffiths (2017) seek to develop ‘a baseline and unified vocabulary’ to ensure ‘rigorous comparison across systems’ (p. 328). Without a ‘consistent vocabulary’ to describe political systems, they argue, scholars will struggle ‘to accumulate knowledge across systems and differentiate the general from the local and the similar from the unique’ (Butcher and Griffiths, 2017: 330).

Unsurprisingly, such calls to establish a unified analytical vocabulary have not found much resonance in the subfield of Historical IR. Sartori’s ‘scientific’ understanding of concepts as ‘data containers’ has been criticized for reifying and essentializing analytical categories, mistakenly presuming that it is possible to arrive at culture-neutral definitions of concepts, and ignoring the historical situatedness of the scholar (Bevir and Kedar, 2008: 509–511; Berenskoetter, 2017: 164–167; see also Berenskoetter, 2016). Similarly, the framework put forward by Butcher and Griffiths has been described as an anachronistic ‘throwback’ to a time when ‘parsimonious scientism and ahistoricism’ were the gold standard of scientific research (Barder, 2017: 3). These are important critiques that point to the limits of the ‘scientific’ approach to concept formation. And yet, the challenge of concept formation does not disappear just because a specific formulation thereof is found wanting. Whether we like it or not, we are always involved in the active conceptualization of our subject matter and the application of analytical concepts across different historical, geographical, cultural and linguistic contexts. Insofar as concepts are the means through which we grasp the world, concept formation is an inevitable part of what we do: concepts ‘provide analysts with an understanding of what is “out there” and in doing so help to grasp relevant phenomena by naming and giving meaning to its features’ (Berenskoetter, 2017: 152). Our goal in this paper is therefore twofold.

First, and most fundamentally, we seek to initiate an overdue methodological conversation in Historical IR by foregrounding the overlooked problem of concept formation. We may have generated a wealth of conceptual histories problematizing the anachronistic application of existing theoretical categories, but this has not been accompanied by a systematic rethinking of how to use concepts analytically while doing historical comparisons. For example, the two recent articles by Kessler and Leira (2024a, 2024b) on concepts in Historical IR are devoted mainly to conceptual history and conceptual change and have little to say about how to articulate travelling concepts for comparative historical work. Previous discussions about how to ‘do’ historically inflected IR focus either on epistemology – usually advocating for varieties of relationalism (e.g. Barnett and Zarakol, 2023; Hobson and Lawson, 2008; Lawson, 2012; Schlichte and Stetter, 2023), postcolonial and decolonial theorizing (e.g. Bilgin, 2016a; Çapan, 2020; Chipato and Chandler, 2022; Sharma, 2021), various strands of Marxism such as Uneven and Combined Development (UCD; e.g. Rosenberg et al., 2022; Rosenberg and Kurki, 2021; Teschke, 2014), or Global Historical Sociology (GHS; e.g. Go and Lawson, 2017) – or they focus on methods more narrowly, such as etiquettes for archival, genealogical or big-data-driven work (e.g. Lemke et al., 2023; Thies, 2002; Vucetic, 2011).

The second goal of our article is to suggest one way of approaching concept formation if the goal is to balance a historicist sensibility with analytical power. Max Weber’s methodological writings on ideal types, we argue, offer an understanding of concept formation that is more line with the interpretivist and historicist assumptions of Historical IR than the scientific methodology of scholars such as Sartori. To be clear, our aim here is not to formalize a blueprint for how all Historical IR scholars ought to formulate their concepts or provide a rigid prescription about what makes a concept ‘good’ (cf. Gerring, 1999). Such an endeavour would entail precisely the kind of scientific approach to concepts that most Historical IR scholars, in our view, justly regard with scepticism. Instead, we proceed by explicitly delineating implicit practices of concept formation that are already present across Historical IR. ⁴ Recent works in Historical IR have not shied away from comparing resemblances and divergences across historical cases, nor from using and modifying analytical categories such as ‘hierarchy’, ‘order’ and ‘sovereignty’ in productive ways. Given that all Historical IR scholarship trades in the analytical use of concepts in one way or another, we need a clear statement about how Historical IR can approach conceptual work beyond conceptual history. And while we have our eye trained primarily on getting better purchase on historical periods, a more careful consideration of the problem of concept formation should also help IR theory stay more relevant into the 21st century (see also Kessler and Leira, 2024b).

The article is organized into four sections. In the first section, we briefly review how Historical IR has criticized the mainstream’s uses of history and problematized the presumed universality of existing analytical concepts. We liken the mainstream approach to ‘tourism’ in that it assumes conceptual travel is easy, requires little to no preparation and comes back from such travel having confirmed its preconceived biases about IR theories. The historical turn has articulated severe criticisms of such conceptual ‘tourism’ and has generally been much more sceptical about the possibility of conceptual travel. However, too much scepticism of analytical concepts can also hinder the ambitions of Historical IR, which aims to write new international and global histories for IR theory, to compare across different time periods and to use new historical material for devising theories for the 21st century. Indeed, the second section of the article shows that travelling concepts are in fact already present throughout Historical IR scholarship. Historical IR already implicitly operates with concepts that travel, at least to some extent – it just has not addressed the issue of concept formation explicitly. To address the issue of concept formation head-on, the article’s third section recovers Weberian ideal types for Historical IR. We suggest that Weber’s methodological writings on ideal types outline an approach to concept formulation that is sensitive to historicism but also aims at generalizability. In the fourth section, we deal with the common criticisms of the ideal type approach and argue that they do not damn the whole enterprise. We conclude by arguing more attention needs to be paid to concept formation in Historical IR and identifying the main advantages of the ideal-typical method for this purpose.

From history to historicism in IR

History has always been present in IR scholarship, but the way in which the relationship between history and theory has been understood by different corners of the discipline has varied significantly (see, for example, Hobson and Lawson, 2008; MacKay and LaRoche, 2017). For the mainstream, history tends to signify a dataset of empirical material against which theories can be tested. Theory is thus separated from history, with the latter serving as ‘a potentially rich stream of data, producing lessons and acting as test cases for deductively derived hypotheses’ (Hobson and Lawson, 2008: 423; see also Lawson, 2012). The conceptual corollary of these assumptions is that analytical categories are presumed to travel without much difficulty, and to work seamlessly in any historical period or geography. Let’s call this approach conceptual ‘tourism’ in that it is similar to visiting foreign countries only to take selfies at the most known landmark, to eat at global chain restaurants, to sightsee via guided tours and to come back home having confirmed their pre-existing opinions about the world.

Conceptual ‘tourism’ can happen for several reasons. First, and perhaps most fundamentally, the ability of IR’s analytical categories to appear timeless owes much to the hegemonic position that the major European languages have occupied in the global political economy of knowledge production. Since the 19th century, the topography of conceptual transfer has been highly uneven, with concepts moving more from European to non-European languages than the reverse (see, for example, De Swaan, 2001; Hill, 2013; Schulz-Forberg, 2014). One consequence of these asymmetries is that European political concepts have acquired the status of theoretical ‘universals’ that are presumed to be capable of relatively seamless travel, while concepts from other parts of the world are perceived as geographically or culturally rooted. In effect, European categories become ‘the prisms through which the diversity of the world has been refracted back into a singularity of concepts’ (Menon, 2022: 25; see also Heiskanen, 2021).

Second, on a more practical level, it is very easy for approaches that mine history for case studies for hypothesis testing to be misled by false equivalences that seem to exist between historical and present-day entities that bear the same name. Such false equivalences are also often pushed by modern-day national projects that want to read history through a particular identity lens. For example, modern Chinese official historiography projects a unified, unchanging ‘China’ back in time (Hui, 2023). This makes it seem that China has always existed as a nation-state. Most countries have their own version of this presentism. It is very difficult to read the past outside of a national (or regional or religious) identity that we care about today. Abstracted to the level of IR theory, these national identity narratives manifest as a ‘methodological nationalism’ that adopts the nation-state as the natural unit of social scientific analysis (see Wimmer and Schiller, 2002). If one does not think about it too much, it becomes easy to assume that nations and states have always existed or that they have not changed much through history.

The third path to conceptual ‘tourism’ in the use of history in IR is cherry-picking time periods and cultural settings that are either inherently similar to our own and/or verifying our present-day beliefs, creating a selection bias effect. There are many examples of this in mainstream IR theory, but an obvious one is the study of ‘great power’ politics in realism only by looking at classical Greece and European history from the 17th century onwards. As Ruggie (1983) observed, there is no good justification for this, as neorealism – for instance – has no explanation for the transformation of political units from the medieval period to the modern era. Others have pointed out that ‘power transition’ theory no longer works the moment one steps out of European history and looks at Asian history (Kang and Ma, 2024). In some cases, even cherry-picked historical examples prove to be a poor fit for our theories, glossing over contextual specificities and distorting past events in the service of contemporary debates. Textbook readings of the Concert of Europe are a good example of this type of distortion (e.g. Schroeder, 1994). Attempts to apply the so-called ‘Thucydides trap’ to contemporary US–China relations are another (e.g. DiCicco et al., 2020; Welch, 2003).

Such criticisms of the mainstream approach are not entirely new, even if they have not always focused on concepts. There have always been more critical and interpretivist voices in IR that have been quite sceptical of our ability to generalize about world political dynamics. These alternative approaches are defined not only by an engagement with history but also by their commitment to historicism: ‘a philosophy that emphasizes the importance of history in understanding, explaining, or evaluating phenomena’ (Bevir, 2017: 1; see also Barkawi et al., 2025; Cello, 2018; Schlichte and Stetter, 2023; Zarakol, 2023). The historicist emphasis on the importance of historical context, including the historical context within which IR scholars themselves operate, leads them to be suspicious of formal theory. Where the mainstream uses historical evidence to construct general and often immutable laws, historicists are more attuned to ‘the contingent, disruptive, constitutive impact of local events, particularities and discontinuities’ (Lawson, 2012: 207). As a result, such approaches generally eschew the desire to compare across time periods or to derive generalizable IR theories from the past to shed light on present-day dynamics.

The historicist position has entailed a rejection of mainstream IR’s ‘isomorphic transhistorical categories’ (Hobson and Lawson, 2008: 428) and a scepticism towards conceptual ‘tourism’. Instead of seeking to formulate new travelling concepts, historicist work has mostly explored how ‘the meanings of key concepts evolve or are reconstructed over and over again’ (Ferguson and Mansbach, 2008: 310; see also Kessler and Leira, 2024a). Thus, analytical categories such as ‘sovereignty’ (Bartelson, 1995; Costa Lopez et al., 2018; Onuf, 1991; Osiander, 2001), ‘territory’ (Branch, 2013; Goettlich, 2019; Ruggie, 1993), ‘anarchy’ (Donnelly, 2015), ‘state of nature’ (Jahn, 2000), ‘society’ (Owens et al., 2015), ‘ethnicity’ (Heiskanen, 2025), ‘foreign policy’ (Leira, 2019), ‘world community’ (Bartelson, 2009), ‘powers’ both ‘middle’ and ‘great’ (de Bhal 2023; Keene, 2013a; Robertson and Carr, 2023), ‘mercenary’ (Riemann, 2021, 2024), ‘intervention’ (Keene, 2013b) and finally the ‘international’ itself (Bartelson, 2023) have been shorn of their universal veil and unmasked as contingent products of particular historical forces. In these genealogical readings, concepts are often presumed to be closely attached to the historical context within which they were produced and are best studied in these contexts.

Conceptual histories and genealogical inquiries typically seek to side-step the problem of concept formation by refraining from ‘explicit, active conceptualization on the part of the scholar’ (Kedar, 2007: 344). ⁵ As sympathetic as we are to such work, it has some limitations. First, the scepticism towards generalizability and comparison contains within it the danger of reductionism and essentialism. If it insists that things cannot be compared, the historicist camp may end up perpetuating the notion that historical institutions have particular cultural essences – or, in conceptual terms, that different languages, cultures or regions have their own incommensurable concepts (see, for example, Barkawi et al., 2023). There is thus a way in which contextual sensitivity can come around full circle to become reductionist as well. Second, the contextualist methodologies typically advocated by such approaches can end up closing a lot of historical periods and places to further investigation by IR. The study of prehistoric IR, for instance, necessarily entails the use of anachronistic analytical categories in the absence of written sources (e.g. Neumann and Glørstad, 2022). Such methodologies could also foreclose the study of vast parts of the world where the macro-history is not yet settled enough for contextual work, where primary documents are not available or where written materials do not resemble their European counterparts. These methodologies thus could end up perpetuating another type of Eurocentrism (Zarakol, 2022). ⁶

Third, and finally, not even the most ardent historicist can avoid the problem of concept formation entirely. If one takes seriously (as we do) the historicist wager that the meaning and understanding of concepts is historically contingent and context-bound, this does not only preclude the possibility of seamless conceptual travel: it also makes conceptual travel impossible to avoid. ‘When we use a present word/concept with reference to a word/concept of the past (recent or distant)’, as Alonzi (2023: 675) explains, ‘we perform a linguistic act that is unavoidably anachronistic’ (see also Fasolt, 2004; Lianeri, 2014). Likewise, when we use a word/concept of one language or culture with reference to a word/concept of another, we perform a linguistic act that is unavoidably anatopistic (the spatial equivalent of anachronism). Conceptual travel across different spatiotemporal contexts is thus not only possible, but also an inevitable part of historical research, including Historical IR. As we demonstrate in the next section, travelling concepts are a silent presence throughout Historical IR scholarship: ubiquitous, yet rarely remarked upon.

The silent presence of travelling concepts in Historical IR

Historical IR is a broad tent. It includes many different approaches to history and varied methodologies, ranging from historical sociology to histories of international political thought and all that lies between. Different positions along this spectrum exhibit varying strengths of historicism, but all share a baseline scepticism towards the transhistorical analytical categories of the mainstream. In this section, by re-reading key contributions to Historical IR scholarship, we aim to show that travelling concepts feature across the entire breadth of the subfield.

On the more historical sociological end of the Historical IR spectrum, some early work explicitly sought to come up with concepts that travel while rejecting the conceptual ‘tourism’ of the mainstream. An influential example is Ferguson and Mansbach’s conceptualization of ‘polity’ as an alternative to ‘state’. Whereas the concept of the state connotes a territorialized form of political authority specific to the modern era, the concept of polity enables the comparison of many different polity types – tribes, empires, corporations, city-states and so on. This, Ferguson and Mansbach contended, would allow the concept of polity to travel to different places and times in a way that the concept of state was not able to (Ferguson et al., 2000; Ferguson and Mansbach, 1996, 2008). ⁷

Another early discussion of the problem of concept formation can be found in Buzan and Little’s International Systems in World History. The key concept in question was ‘international system’. Noting the presentist, ahistorical, state-centric and Eurocentric tendency of IR scholars to equate ‘international system’ with ‘Westphalian system’, Buzan and Little (2000) endeavoured to free this concept to historical travel (pp. 18–22). After a detailed survey of both IR and World History literature, their chosen strategy was to ‘disaggregate the idea into sectors, and refer separately to international military-political systems, international economic systems, and international socio-cultural systems’ (Buzan and Little, 2000: 95). The traditional conceptualization of the international system in terms of the Westphalian model was thus displaced by a more abstract framework based only on the ‘interaction capacity’ of its units. Exactly what kind of interaction was involved, and exactly what kind of units composed the system, was left open, to be filled in by careful historical analysis. Thus, in the same way that Ferguson and Mansbach’s conceptualization of ‘polity’ allowed for a plurality of polity types, Buzan and Little disaggregated the concept of ‘international system’ into a multiplicity of subtypes.

The analytical categories of ‘international system’ and ‘interaction capacity’ were taken up again by Buzan in The Global Transformation (2015), co-authored with Lawson. The reception of the book in Historical IR circles testifies to the subfield’s hesitancy towards travelling concepts. While Buzan and Lawson’s work received praise for offering a new narrative of the birth of the international system and debunking the Westphalian myth, it was also criticized for not adequately historicizing the concepts upon which it relied. Thus, Owens (2016; see also Owens, 2015; Owens et al., 2015) called for a deeper historicization of the category of the ‘social’ that underpinned the book’s methodological apparatus, while Bilgin (2016b) cautioned against the unthinking application of theoretical concepts derived from European social theory to the rest of the world. Responding to these challenges, Buzan and Lawson (2016: 509) acknowledged that ‘concepts and categories of thought should always be subject to the denaturalizing impulses of historicization’. However, they gave no explanation as to how this historicist impulse is to be reconciled with IR’s need for concepts capable of travelling across different historical contexts.

A similar reluctance to address the problem of concept formation can be found in more recent work on GHS, a research programme that grew out of the works discussed above. In their programmatic introduction to a recent edited volume on GHS, Go and Lawson (2017) explained that GHS consists of ‘engaging fully with transnational and global histories, while occupying a register at one remove from such studies through the overt deployment of conceptual abstractions, analytic schemas, and theoretical frames’ (p. 5). It is this need for ‘conceptual abstractions, analytic schemas, and theoretical frames’ that brings GHS to the problem of concept formation: if the existing analytical toolkit of IR theory is problematic, through what concepts should the world be grasped instead? Go and Lawson defer the answer to this question, writing that

as yet, historical sociology has not fully elaborated the concepts and theories that could be used in a systematic analysis of transnational and global processes. This is because, as with other branches of sociology, much historical sociology has been hindered by internalism and methodological nationalism. (Go and Lawson, 2017: 7)

Nonetheless, we can glean some preliminary answers from their remarks. For example, one of the criticisms that Go and Lawson (2017) raise against mainstream IR is that it ‘occludes differences between polities (such as empires and nation-states)’ and ‘fails to distinguish between types of international order (such as imperial and sovereign orders)’ (p. 17). Implicit in Go and Lawson’s argument is the assumption that ‘polity’ is a better traveller than ‘empire’ or ‘nation-state’ and that ‘international order’ is a better traveller than ‘imperial order’ or ‘sovereign order’.

The above remarks are also true for much of recent Historical IR work that falls under the banner of GHS. All assume that certain concepts can travel, but rarely do they explain their method of concept formation or explain why the concepts of present-day IR also work in the past (even as they display historicist sensibilities in their approach to their cases). To give a few examples, both Sharman (2019) and Spruyt (2020) lean strongly on the concept of a ‘global international system’, not to mention ‘empire’. Similarly, Phillips (2021) works with both ‘empire’ and ‘hierarchy’. Phillips and Sharman (2015) also use ‘international order’ alongside ‘international system’; they discuss ‘sovereign states’ in contrast to ‘empire’ and compare both across time and space. Reus-Smit (2018) discusses the possibility of comparing ‘cultural diversity’ across different ‘international orders’ (see also Phillips and Reus-Smit, 2020). We do not mean to suggest there is anything wrong with these works (to the contrary) or their use of concepts or their definitions. We merely mean to draw attention to their silence about how, when and why concepts travel.

Clearly, scholars located towards the historical sociological end of the Historical IR spectrum regularly invoke and apply travelling concepts in their work. But is this also the case among scholars with a stronger historicist commitment? An initial glance through the literature might suggest that the answer is ‘no’. For instance, Bruneau has challenged the applicability of the ‘polity’ concept to early modern Europe, arguing that this period is better understood through the concept of ‘crown’, which he describes as a form of ‘political authority that does not conform to our idea of polity’ (Bruneau, 2023: 715). Costa Lopez (2020) goes further, questioning the usefulness of the very concept of ‘political authority’ for understanding politics in medieval Europe. Nevertheless, these works are littered with analytical concepts that travel across contexts, including ‘international order’ in Bruneau and ‘authorization’ in Costa Lopez. Rather than representing a wholesale rejection of conceptual travel, these scholars work on different scales and with different analytical frames that could just as well be deployed to problematize contemporary assumptions of the nation-state as a coherent unit.

Let’s wrap up this section by considering what should be the ‘hardest’ case for our argument: outright conceptual histories dedicated to the historicization of existing categories. As it turns out, we find travelling concepts even here. Consider, for instance, Leira’s conceptual history of ‘foreign policy’. On the one hand, Leira’s article provides a ‘denaturalization’ of present understandings of foreign policy by revealing the way in which this concept coalesced in the eighteenth century (Leira, 2019: 188). In Leira’s analysis, ‘foreign policy’ ceases to be a transhistorical analytical category and is instead shown as the contingent outcome of specific historical forces to do with European state formation. On the other hand, Leira’s historicization of foreign policy rests upon a series of adjacent concepts that make it possible to chart the emergence of foreign policy in particular: concepts such as ‘government’, ‘monarch’, ‘boundary’, ‘interaction’ and ‘power’ are dispersed throughout the analysis and remain largely unchanged, forming the ground for the historicization of ‘foreign policy’. Leira himself appears to acknowledge this in his conclusion, when he writes, ‘Boundaries might be a universal feature of human group relations, but foreign policy, understood as a practice concept, is not’ (Leira, 2019: 195). A conceptual history of ‘boundary’, ‘human’ or ‘group’ would no doubt problematize this statement, but this is not a weakness of Leira’s argument. The point is, simply, that historicization must always take place against a plane of concepts that are held constant.

To recap our argument so far: In the previous section, we demonstrated that Historical IR’s critique of conceptual ‘tourism’ has entailed a commitment to historicism and a scepticism towards the articulation of travelling concepts. Indeed, one of the most important contributions of Historical IR has been to denaturalize and provincialize the analytical categories that the discipline’s mainstream had simply taken for granted. At the same time, we have argued that historical work cannot avoid the problem of concept formation entirely. As we have sought to demonstrate in this section, the full spectrum of Historical IR scholarship is replete with analytical concepts that are presumed to be capable of travel, even as the conditions for forming such concepts remain underspecified. In the following section, we seek to fill this gap with Weber’s ideal types.

Travelling concepts as ideal types: returning to Weber

To develop a historicist approach to concept-formation, we want to return to Max Weber’s work on ideal types. Weber is an apposite reference point for Historical IR, as he developed the notion of the ideal type against the backdrop of the Methodenstreit – a debate among German-speaking economists over the use of history and the role of general theory in the social sciences. Attempting to chart a middle path between historicism and positivism, Weber developed the notion of the ideal type as a critique of naïve historicist empiricism, which assumes that historical explanation is possible without recourse to abstract conceptualization. Against the historicists, Weber insisted on the necessity of abstract conceptualization and general theory: our knowledge of historical reality is never entirely presuppositionless, but always laden with theoretical assumptions that allow us to translate the past into the present. Given the discursive nature of knowledge, general concepts and theories in historical research are not only possible, but necessary. At the same time, Weber rejected the representational conception of truth espoused by the positivists: concepts were not partial pictures that could be fitted together to form a perfect representation of empirical reality or produce law-like generalizations, but heuristic devices through which the empirical world was made knowable to the researcher (Beiser, 2011: 521–528; Drysdale, 1996: 77–78). Weber’s writings on ideal types thus allow us to bring together an historicist sensibility with the possibility of theoretical generalization and conceptual travel.

Commentators have identified at least two (Swedberg, 2018: 182) or three (Weinert, 1996: 75) distinct phases in Weber’s methodological writings on ideal types. As a result, Weber’s notion of the ideal type can be interpreted in many different ways. Some have emphasized Weber’s affinity to the interpretivist tradition and described ideal types as ‘hermeneutic concepts’ (Kedar, 2007; see also Bevir and Kedar, 2008). Focusing on his later, more ‘scientific’ works, others have argued that the ultimate goal of Weberian ideal types is to generate ‘law-like statements’ that enable ‘relatively value-free causal analyses of historical events’ (Rex, 1977: 165). Attempts have also been made to align ideal types with the tenets of critical realism (Patomäki, 2002: 109) and poststructuralism (Rabinow, 2003: Chapter 2; Koch, 2007: Chapter 2). Our aim here is not to mediate between these competing claims or to offer a definitive interpretation of Weber, but to expand on what we see as the most valuable features of Weberian ideal types for concept formation in Historical IR. We focus mainly on Weber’s 1904 essay ‘“Objectivity” in Social Science’, partly because it contains his most detailed discussion of ideal types, but mainly because it is directly concerned with the challenges of historical research and thus of direct relevance to Historical IR. ⁸

What, then, is an ideal type? In his 1904 essay, Weber offers the following definition:

An ideal type is formed by the one-sided accentuation of one or more points of view and the by the syntheses of a great many diffuse, discrete, more or less present and occasionally absent concrete individual phenomena, which are arranged according to those one-sidedly emphasized viewpoints into a unified analytical construct (Gedankenbild). In its conceptual purity, this mental construct (Gedankenbild) cannot be found empirically anywhere in reality. It is a utopia. (Weber, 1949: 90)

An ideal type, Weber insists, ‘is “ideal” in the strictly logical sense of the term’ (Weber, 1949: 92, see also 98–99). In other words, rather than gesturing at some kind of normative model, the ideal type is a deliberate simplification of reality – an abstraction – that enables the analyst to pick out relevant and significant features of the empirical world for analysis and comparison (Weber, 1949: 93).

Weber considers ideal types ‘indispensable’ for both ‘heuristic’ and ‘expository’ purposes (Weber, 1949: 90). In their heuristic function, ideal types serve as interpretative devices that enable scholars to delimit the infinite richness of empirical reality:

If the historian [. . .] rejects an attempt to construct such ideal types as a ‘theoretical construction’, [. . .] the inevitable consequence is either that he consciously or unconsciously uses other similar concepts without formulating them verbally and elaborating them logically or that he remains stuck in the realm of the vaguely ‘felt’. (Weber, 1949: 94)

A number of commentators have emphasized the similarities between Weberian ideal types and Herbert Blumer’s ‘sensitizing concepts’ (Rosenthal, 2012: 202–203; Swedberg, 2016: 28). ‘Sensitizing’ concepts lack ‘specification of attributes or benchmarks’ and instead give ‘the user a general sense of [. . .] guidance in approaching empirical instances’ (Blumer, 1954: 7). This sets them apart from ‘definitive concepts’. Embedded in our background knowledge and disciplinary structures, sensitizing concepts do not prescribe the conclusions of the analysis but ‘suggest directions along which to look’ (Blumer, 1954: 7; see also Bowen, 2006). Sensitizing concepts are thus always already there, whether we openly acknowledge them or not. Understood as sensitizing concepts, Weberian ideal types are interpretative devices that guide the direction of analysis.

This brings us to the important role of value-relevance and cultural significance in making sense of an infinitely complex reality. ‘Order is brought into this chaos’, Weber (1949) writes, ‘only on the condition that in every case only a part of concrete reality is interesting and significant to us, because only it is related to the cultural values with which we approach reality’ (p. 78). In contrast to scientific approaches that judge the merit of a concept on the basis of its supposed correspondence to empirical reality, Weber here acknowledges the historical situatedness of the scholar. The phenomena that are selected and accentuated in the construction of ideal types are always and necessarily influenced by the historical standpoint of the observer. However, this does not lead Weber into a nihilistic relativism whatsoever: the cultural significance of the ideal type may render it historically relative, but it also imposes constraints on this relativism insofar as the ideal type must be communicable to a wider (scientific) community (Drysdale, 1996: 82–84; see also Kedar, 2007: 340).

If the heuristic function of ideal types comes at the beginning of the investigation, their expository function comes at the end: the construction of an analytical narrative that is as clear and as consistent as possible. Importantly, this does not require an ideal type to be empirically accurate. Insofar as an ideal type is a ‘utopia’ that cannot be found anywhere in the real world, it will necessarily fail to provide a perfect representation of reality – and this is precisely the point. As Jackson (2010) explains, it is the ‘failure’ of the ideal type to accurately capture empirical reality that enables the formulation of a richer analytical narrative,

either by introducing case-specific coincidental factors to help explain why a case does not look exactly like its idealization, or by drawing on multiple ideal types (and perhaps additional coincidental factors) in order to account for the complexity of the observed outcome. (p. 155)

The ideal type is thus not only the condition of possibility of the analytical narrative, but also the limit that leaves any such narrative open to critique and reinterpretation. Kedar (2007) refers to this as the ‘self-deconstructing’ quality of the ideal type, whereby the ideal type ‘anticipates its own problematization as the inevitable outcome of every contact with the infinite diversity of human existence’ (p. 333). In short, the ideal type is the condition of (im)possibility of historicity. ⁹

The heuristic and expository functions of the ideal type underscore the pragmatic underpinnings of Weber’s methodology. For Weber, the value of an ideal type is never judged by its empirical accuracy, but rather by its usefulness: ‘whether, once applied, the ideal type is efficacious in revealing intriguing and useful things about the objects to which it is applied’ (Jackson, 2010: 146). A similar pragmatism applies to the naming of ideal types. On the one hand, given that an ideal type is a deliberate simplification of empirical reality, Weber considered it generally prudent to avoid highly politicized or ambiguous terms used in everyday language (Weber, 1949: 108). This line of argument resonates with Brubaker’s (2004) more recent exhortation to separate ‘categories of analysis’ from ‘categories of practice’. On the other hand, Weber acknowledged that it was unfeasible to invent a new technical term every time one encountered a new phenomenon (Weber, 2001: 63). The redefinition or refinement of existing categories, or their combination or qualification, can also be valid ways forward (Swedberg, 2016: 23–25). Equally, Weber recognized that it was not possible to define every single concept that appeared in the analytical narrative:

Hundreds of words in the historian’s vocabulary are ambiguous constructs created to meet the unconsciously felt need for adequate expression and the meaning of which is only concretely felt but not clearly thought out. In a great many cases [. . .] their ambiguity has not been prejudicial to the clarity of the presentation. (Weber, 1949: 92–93)

The question of which concepts should be systematically named and defined is, again, pragmatically determined by the needs of the analytical narrative and the significance of the particular phenomenon under scrutiny (Weber, 1949: 93).

To sum up, Weber’s writings on ideal types outline an approach to concept formation that avoids essentialism, acknowledges the transiency of our analytical categories and incorporates the historical situatedness of the scholar into its methodology. This sets Weber apart from positivists such as Sartori, who reduce the problem of concept formation to a problem of classification: for Sartori, the end goal of concept formation is the construction of ‘a filing system provided by discriminating, i.e., taxonomic, conceptual containers’ (Sartori, 1970: 1052). While Weber acknowledges the analytical value of classifications and typologies, he takes care to emphasize that ideal types are not ‘simple class or generic concepts’. ‘The goal of ideal-typical concept-construction’, Weber writes, ‘is always to make clearly explicit not the class or average character but rather the unique individual character of cultural phenomena’ (Weber, 1949: 101).

The methodological differences between Sartori and Weber have important implications for how they understand conceptual travel. By treating concepts as value-neutral ‘data containers’, Sartori turns conceptual travel into a zero-sum tug-of-war between a concept’s intension and extension, where the only way to extend a concept’s range of applicability is by reducing the number of attributes associated with the concept (Sartori, 1970: 1041). For Sartori, a concept can travel where one or more of its essential attributes can be found in a particular empirical case; therefore, the fewer attributes a concept has, the more effectively it can travel. If a concept’s extension is increased without diminishing its intension, we commit the sin of ‘conceptual stretching’. By contrast, a Weberian understanding of conceptual travel does not require the same essential attributes of the concept to be present across all of the cases to which it is applied. After all, the value of an ideal type is judged not only by its ability to accurately represent empirical reality, but also by its usefulness, which may stem as much from its ‘failures’ as its ‘successes’. More often than not, additional context-specific descriptions will be necessary to clarify divergences from the ideal type. Rather than requiring a common set of attributes to be present across multiple cases, therefore, the Weberian understanding of conceptual travel is more akin to the Wittgenstein’s notion of ‘family resemblances’ (Kedar, 2007: 338–339; see also Bevir and Kedar, 2008: 513). The different cases to which an ideal type can be applied can manifest a variety of similarities or relationships that are not necessarily reducible to any essential commonalities. Instead, ‘we see a complicated network of similarities overlapping and crisscrossing’ (Wittgenstein, as cited in Kedar 2007: 338). The ability of an ideal-typical concept to travel across different historical contexts consequently has very little to do with the relationship between the concept’s intension and extension. It is based, rather, on the careful analysis of the contextual specificities and relationships of different cases, and the ability of the ideal type to illuminate those cases.

Uses and abuses of ideal types in IR

We began this article by highlighting the Eurocentric and modernist underpinnings of mainstream IR’s conceptual toolkit. To what extent are the methodological writings of Max Weber – a European sociologist of modernity – an appropriate remedy to such issues? Recent scholarship has highlighted the imperialist influences in Weber’s work (Boatcă, 2013; Zimmerman, 2006). Among IR scholars, Weber’s ideal types have also been criticized for perpetuating methodological internalism and methodological nationalism, reproducing a singular conception of society and obscuring (colonial) relations of power that cut across different social formations. Among Marxist scholars, this argument is evident in the growing literature on UCD. ¹⁰ For example, Matin (2013: 7) rejects ideal types on the grounds that they ‘presuppose and theoretically reproduce the ontological singularity of the social because otherwise their utility as comparative yardsticks would be compromised’. Similarly, in their UCD-inspired take on the origins of capitalism, Anievas and Nişancıoğlu (2015) criticize the tendency of ideal types ‘to keep history at arm’s length from theory’ (p. 97). Against the societal singularities supposedly implied by Weberian ideal types, UCD emphasizes the relational connections between multiple interacting social formations: it would be

a mistake to conceive of any form of combined development as a kind of ideal type with which others can be compared and contrasted. Rather, there is no uniform type of combined development, but only a multiplicity of differentiated forms and trajectories. (Anievas and Nişancıoğlu, 2015: 49)

This line of attack against Weberian ideal types comes across most powerfully in Bhambra’s work on ‘connected sociologies’. The major ideal types of modern social theory, Bhambra argues, were originally devised to explain the rise of European capitalist modernity and remain profoundly Eurocentric. Even though ideal types are supposed to be heuristic devices open to change, what tends to happen in practice is the multiplication of ideal types alongside one another in a way that preserves the Eurocentrism of the original (Bhambra, 2014: 146–147). A salient example of this tendency is the literature on ‘multiple modernities’ where the European version of modernity remains the point of reference. ¹¹ Because different regions and viewpoints are deemed to require their own ideal types, these individual ideal types become isolated from critique and a general understanding of the phenomenon is foreclosed (Bhambra, 2016: 337–338). Echoing the UCD literature mentioned above, Bhambra’s critique of ideal types is thus fundamentally a critique of methodological internalism: ‘the function of ideal types is to separate some events and “entities” from others and to represent their internal relationships, thereby making other entities and events mere contingencies from the perspective of those relations’ (Bhambra, 2016: 347).

Against the additive tendencies of Weber’s ideal types, Bhambra proposes a ‘connected histories’ or ‘connected sociologies’ approach as an alternative. This approach proceeds from the assumption that ‘events are constituted by processes that are always broader than the selections that bound events as particular and specific to their theoretical constructs’ (Bhambra, 2014: 3–4). It is an argument for the ‘reconstruction of concepts’ (Bhambra, 2014: 4) by shifting from a ‘national’ framework of concept formation to a framework that recognizes ‘intertwined histories and overlapping territories’ (Bhambra, 2014: 155). To make this case, Bhambra draws on decolonial scholarship and, in particular, Mignolo’s writings on the ‘de-linking’ of conceptual formations from Western modernity (Bhambra, 2014: 134–136). As Mignolo (2007) explains, ‘one strategy of de-linking is to de-naturalize concepts and conceptual fields that totalize a reality’ (p. 459). Overturning the epistemological imperialism of the West requires changing ‘the terms (concepts) as well as the content (histories) of the conversations on modernity/coloniality’ (Bhambra, 2014: 135).

Bhambra (2014) is probably correct to suggest that the ‘standard account of ideal types’ serves to ‘separate some events and “entities” from others’, with a focus on their internal relationships rather than the connections between them (p. 4). It is also true that describing a concept as an ‘ideal type’ can be a sly way of glossing over historical inaccuracies and oversimplifications, and can thus serve as an excuse for the perpetuation of Eurocentrism or other biases. What Bhambra’s critique shows beyond doubt, moreover, is that the ideal-typical conception of modernity that has dominated Western social sciences glosses over colonial and imperial relations, that the cultural and political context within which historical sociology is conducted has dramatically changed since the early 20th century, and consequently, that the ideal-typical concepts that structured sociological knowledge no longer possess the heuristic value they perhaps once did. Bhambra’s is a powerful – and arguably devastating – critique of the specific ideal types that have prevailed in Western sociology; what it is not is a refutation of ideal-typification as such. Indeed, while Bhambra presents her argument as a general critique of Weber’s methodology, her argument is based mostly on secondary sources and takes aim primarily at the literature on ‘multiple modernities’ that she presents as the bête noire of ideal-typification. A closer reading of Weber’s work reveals important similarities between his ideal-typical account of concept formation and Bhambra’s manifesto for a connected sociology.

To begin with, Bhambra (2014) insists that the connected sociologies approach ‘recognizes a plurality of possible interpretations and selections, not as a “description”, but as an opportunity for reconsidering what we previously thought we had known’ (p. 4). This has striking echoes of Weber’s insistence on the transiency of ideal types, discussed in the preceding section. Rather than providing an empirically accurate ‘description’ that can be fixed once and for all, an ideal type is an interpretative device that always remains open to challenge: ‘The great attempts at theory-construction in our science were always useful for revealing the limits of the significance of those points of view which provided their foundation’ (Weber, 1949: 105–106). Furthermore, the key concept of ‘coloniality’ that informs Bhambra’s work appears to have the heuristic function of an ideal type – one derived primarily from the Latin American experience of modernity, the main reference point of decolonial scholarship. In Blumer’s terms, we might say that the concept of coloniality ‘sensitizes’ the researcher to issues that had been neglected by the disciplinary mainstream. This would appear to corroborate Weber’s contention, discussed in the previous section, that historical scholarship ‘must use concepts which are precisely and unambiguously definable only in the form of ideal types’ (Weber, 1949: 92).

Bhambra’s critique of ideal types thus appears to fall back on a different set of ideal types despite itself. By contrasting the ideal-typical method of classical sociology to her connected sociologies approach, Bhambra is effectively setting up a typology composed of two ideal-typical approaches to sociological research. In a process that could be described as ‘one-sided accentuation’, the additive and separating logic of the ‘multiple modernities’ literature is taken as the ideal-typical manifestation of the Weberian method and contrasted with Bhambra’s own connected sociologies approach. Put differently, there appears to be a ‘meta’ problem here, where Weber’s method appears not only as a particular item within Bhambra’s framework, but also as the general form of her framework as a whole. Bhambra’s empirical work also seems to rely on ideal-typical contrasts and distinctions. In her revisionist history of the British welfare state, for example, Bhambra contrasts her own ‘imperial’ reading of this story to the prevailing ‘national’ one: we must rethink the British welfare state as ‘an imperial state and not just a national state’ (Bhambra, 2022: 13). What are the ‘imperial state’ and the ‘national state’ if not ideal-typical constructs, partial analytical cuts shaped by considerations of cultural significance and value-relevance, informed by certain theoretical presuppositions, to present a coherent analytical narrative? Part of the challenge with navigating these overlaps is that Bhambra never formalizes an alternative process of concept formation to rival Weber’s; her project, thus far, has done more to dismantle existing conceptualizations than to reconstruct alternatives. ¹²

This ‘meta’ problem is not limited to Bhambra’s critique of ideal types. In his well-known typology of IR methodologies, for instance, Jackson positions Weber’s ‘analyticist’ approach as one of four possible methodological approaches, the other three being neopositivism, critical realism and reflexivism. Does this mean that ideal-typification is but one among four methodological possibilities, the other three not relying on ideal types? It is actually not quite so straightforward. As Suganami notes in a response to Jackson, the latter’s two-by-two matrix gives the somewhat misleading impression that the four methodologies ‘are all on the same plane’ (Suganami, 2013: 264). In fact, as Suganami (2013) convincingly argues, critical realism, analyticism and reflexivism arise, respectively, from ontological, epistemological and political critiques of neopositivism (pp. 264–265). This asymmetry is reflected on the ‘meta’ level in the fact that Jackson’s two-by-two typology is itself an application of the analyticist method of ideal-typification (Jackson, 2010: 37–38). Weber’s method is thus not only one particular methodology within Jackson’s matrix, but also the general logic through which a plurality of methodologies becomes legible and comparable.

The same problem repeats itself in MacKay and LaRoche’s (2017) attempt to organize different philosophies of history among IR scholars. Inspired by Jackson’s two-by-two matrix, MacKay and LaRoche organize IR scholarship along two axes: (1) a linear, nonlinear or multilinear conceptions of history, and (2) a conception of history as either predictable or unpredictable. The result, as MacKay and LaRoche (2017) explain, is a typology composed of ‘six ideal-typical categories’ (p. 205). Once again, as in the work of Bhambra and Jackson, it is only through a process of ideal-typification that a plurality of possibilities become comparable. This does not mean that this plurality is somehow an illusion and that, in reality, there is only one ‘true’ philosophy of history. The point is, rather, that the very possibility of comparing different philosophies of history appears to be predicated upon a process of ideal-typical concept formation that enables the analyst to identify the different philosophies of history to be compared.

In this section, we have sought to defend Weber’s ideal-typical approach to concept formation from charges of methodological internalism and Eurocentrism. While recent critiques have highlighted the limitations of the ideal types inherited from classical sociology, we do not believe these critiques discredit Weber’s method as such. To the contrary, we have suggested that some of the arguments presented by his critics often remain compatible with the Weberian approach. This could be taken to vindicate Aron’s (1964: 71) judgement that the ‘“ideal type” should be regarded less as a distinct variety of concept than as a generic name for all the concepts used by the cultural sciences’ (see also Kedar, 2007: 345). At a minimum, the apparent difficulty of escaping Weber underlines the importance of taking his approach to concept formation seriously.

Conclusion

We opened this article by discussing the diversification of IR’s spatiotemporal parameters beyond the modern Western world. What is today recognized as the subfield of Historical IR has played an especially important role in this diversification, bringing new spaces and times within the purview of the discipline. Thus far, however, this diversification of IR’s subject matter has not been accompanied by a serious rethinking of the conceptual apparatus through which IR scholars study the world. We have argued that Weberian ideal types offer a valuable resource for making sense of the kind of active conceptualization that Historical IR scholars are engaged in. To be clear, our aim here is not to provide a new approach to concept formation that all Historical IR scholars should adopt. Rather, in the spirit of Weber himself, our goal has been to reconstruct the practices of concept formation that are already present in Historical IR – to make explicit the conceptual work that Historical IR scholars of many different stripes already seem to be doing.

Approaching concept-formation in Historical IR through Weber’s writings on ideal types has a number of important implications. First, it encourages us to acknowledge the role of conceptual abstraction in historical inquiry and to reflect on the pragmatic value that we hope to gain from our analytical concepts. If, as we have argued, the use of ideal-typical concepts and anachronisms is inevitable in historical research, then it is best to be as explicit and open about this as possible. We have shown above that not even conceptual history can entirely avoid such abstractions. Meanwhile, the historical sociology approaches to Historical IR become entirely impossible without travelling concepts, so researchers might as well be explicit about how they are making their concepts travel. In this article we have suggested that many of these travelling concepts can be understood as Weberian ideal types. We thus urge the field to consider more carefully how a Historical IR that is relational, historicist and respectful of primary source research can nevertheless develop and employ travelling concepts – concepts that are neither entirely bound to their specific context nor entirely divorced from it.

Second, the approach suggested in this article forces us to query the conditions under which ideal-typification becomes either Eurocentric or otherwise problematic. Here, we find less fault with the conception of ideal types per se, and more with the failure to query the ideal types inherited from classical European sociology. As we suggested above and as recent Historical IR research on especially Asia has shown, there is no inherent reason why ideal types need to be Eurocentric. This was a flaw of execution in the past as opposed to something always dictated by this method (on this question see also Collins, 1997).

Third, much of Historical IR even in the historical sociologist vein thus far has focused on connected narratives and histories, which is partly why we have been able to afford not to think about analytical concept construction. Yet, not everything we may want to study in history is necessarily connected. Some political dynamics recur independently across time and space, and we need the right toolkit to think about such dynamics comparatively. This is probably going to be a growing research area in Historical IR in the next decade as we start from concepts first and compare/contrast, but without the Eurocentrism of previous comparative approaches. Specifically, the Weberian approach to conceptual travel encourages us to think about cross-case comparisons not in terms of a zero-sum trade-off between conceptual intension and extension, but in terms of the narrative illumination of family resemblances across different historical contexts.

Fourth, the Weberian approach encourages us to view the lexicon of Historical IR as an archive for conceptual historians of the future to explore. Precisely because ideal-typical concepts are a means rather than an end, they are necessarily provisional. Concepts that once appeared to travel well and generate theoretical insights may one day cease to be able to do so. As the context within which we as scholars operate changes, so too must our concepts. Progress in Historical IR therefore does not mean conceptual fixity, but to the contrary, ‘the perpetual reconstruction of those concepts through which we seek to comprehend reality’ (Weber, 1949: 105).

Finally, the above also puts into a different light the seemingly endless debates about IR’s lack of a secure disciplinary identity, inability to settle on a grand theory, and unrequited borrowing of concepts from neighbouring fields (see, for example, Buzan and Little, 2001; Kristensen, 2016). From a certain point of view, these features appear as shortcomings, as signs of the adolescence of the discipline. Yet, if we take Weber’s insight into the nature of historical inquiry seriously, they take on a more positive light, as a sign of the discipline’s continued vitality. In Weber’s eloquent words,

there are sciences to which eternal youth is granted, and the historical disciplines are among them – all those to which the eternally onward flowing stream of culture perpetually brings new problems. At the very heart of their task lies not only the transciency [sic] of all ideal types but also at the same time the inevitability of new ones. (Weber, 1949: 104)

All of this underlines the importance of Historical IR for the wider discipline. Not only is it here to stay but it may indeed be the thing to take IR into the 21st century.