Introduction
What is the subject matter of the discipline of International Relations (IR), particularly if seen in relation to its closest neighbouring disciplines: in IR’s case Sociology, Politics/Political Science, and History? This seems to be a straightforward, if quite basic, question to ask of any academic discipline. Yet in IR, this question is rarely addressed (let alone answered) in a direct fashion. We argue in this contribution that systematically engaging with it is a worthwhile exercise required to focus on central research questions for IR as a discipline. The argument is based on the conviction that any discipline functions and innovates by always asking basic questions about itself anew, and that doing so also serves as a boundary-drawing exercise in the demarcation practice of forming (fleeting and always contested) ‘disciplinary identities’. The purpose of this article is to inquire whether it is possible, particularly with a view to neighbouring disciplines, to legitimately delineate one or many realms as the discipline’s prime subject.Footnote 1 The purpose of this article is not to provide an agenda for what IR should do (either collectively or as individual scholars). It is a process piece, something a discipline needs to do as part of reproducing itself and keeping up to date (see more below on the practice of boundary drawing).Footnote 2
In order to address the issue of delineating the discipline’s subject, the next section first raises a number of problems associated with both identifying the subject matter of IR and ‘labelling’ the discipline in relation to competing terms and disciplines. The following section approaches the issue by first of all dealing with the questions of whether, and to what degree, IR takes its identity from a confluence of disciplinary traditions or from a distinct methodology.Footnote 3 The section after that then turns to the two possibilities we see that would lead to identifying IR as a discipline defined by a specific realm in distinction to other disciplines. The first possibility refers to the identification of the ‘international’ as a specific realm of the social world, differentiated from other such realms in functional terms. The second possibility refers to IR being about everything in the social world above a particular scale, that is, IR in effect being similar to sociology in claiming the entire social world as its subject matter, however in contrast to sociology only in those parts or aspects of it that are ‘macro’ enough.Footnote 4 The difference between both possibilities might not seem extensive at first glance, but it might have consequences for how we understand and study international relations. The final section therefore discusses whether and to what degree these two possibilities are mutually exclusive of each other, where they could possibly overlap, and, most importantly, what the consequences of leaning more towards either one of them are for the study of International Relations.
While this article is based on the conviction that it is important to address the issue of what ‘IR’ is or should be all about, and while we come up with two possible answers in this respect, we do not claim that these are the only possible answers. The present argument is not an exercise in trying to fix a disciplinary identity, but reflects the insight that identities are always in the process of being negotiated in boundary-drawing practices against relevant others. In cases of academic fields or disciplines the relevant others most likely are neighbouring academic fields or disciplines. This also means that we are looking primarily at boundary processes, or system-environment distinctions, within the system of science. This is something distinctly different from looking at ‘ontologies’. In fact we subscribe to the view that this is one of the most over- and partly misused term in IR debates in the sense that there is a lot of reflection on ontologies, but little on ontology and its appropriateness as a key concept for contemporary social science.Footnote 5 One consequence is that a concept coined in and for the observation of premodern society is extensively used for observing complex social arrangements that are distinctly different from premodern society.Footnote 6 When we argue in the following that what IR is or should be about needs to be seen through the lens of specific realms, this is precisely not primarily an argument about ontology. This is not to dismiss philosophy of science or methodological issues in this regard. It would be possible to explore our argument about the subject matter of IR in these terms, yet we claim that for our specific purposes this is hardly necessary. We are in fact grateful to one of the reviewers of an earlier version of this article for remarking that the ‘authors skate gracefully along the surface of their subject. Yet the ice is thin, philosophically speaking, the water beneath cold and dark. I would be happier if the authors at least acknowledged the dangers of skating on thin ice.’ We are happy to do this, yet also state that we continue to prefer skating over diving deep into cold and dark waters, which is what many debates on ontology in IR feel like. To stay within this figure of speech, we aim for means of better ways of identifying those parts of the lake with safer ice conditions, while being fully aware of the dangers that lurk underneath (this incidentally being an important difference to so-called ‘neo-positivism’ in which skaters claim that they are on safe and dry land all along).
Labels, disciplinarity, and a programmatic approach
In a discipline not known for an inimical relation towards introspection, seeking to define its ‘subject matter’ will certainly raise a few eyebrows. There is indeed no lack of ‘state of the discipline’ overviews, of heated fights on grand theories or on methodology. In contrast to these debates, fairly little explicit discussion takes place on the long-standing and quite basic question of whether IR’s primary subject should indeed be seen as being about an, however broadly defined, area of ‘international relations’, ‘international politics’, ‘world politics’, globalisation, etc. In the context of the so-called ‘end of IR theory’ debate, a good example is Patrick Jackson’s and Daniel Nexon’s thoughtful contribution on scientific ontologies as being at the heart of IR theory, ‘which is to say, a catalog – or map – of the basic substances and processes that constitutes world politics’. While there obviously is a great variety at play in this constitution what is not called into question at all is that there ‘is’ something that can legitimately be called world politics. Jackson and Nexon in fact do not raise the theme at all (let alone why it should be ‘world politics’ rather than ‘international politics’ or ‘international relations’). We raise this point here in order to argue that even in the most advanced forms of contemporary IR theorising, the IR tradition mostly continues with asking what world politics is ‘made of’, yet somehow simply assumes that there ‘is’ something distinguishable there in the first place.Footnote 7
The uncertainty about IR’s subject matter is tracked by variations in the way the discipline is labelled. ‘International Politics’ is perhaps the narrowest name, implying that the subject matter is just the macro-scale end of Political Science focusing on relations between states and covering foreign policy, strategy and security, international organisations, and the politics of the international system.Footnote 8 ‘World Politics’ is barely wider, still a subsection of Political Science, but carrying the implication that there is more politics at the global level itself, including transnational politics and global governance. This framing can lead to a division of labour with ‘Politics’, between an ‘International’ side and a domestic or ‘Government’ side, sometimes realised in different departments, as at the LSE. ‘International Relations’ occupies the middle ground, being comprised of two rather vague terms: ‘international’, which at least hints at actors other than the state; and ‘relations’ which opens the possibility of more than just politics. ‘Relations’ can cover economic and societal as well, opening up the possibility for IR as a multidisciplinary construction. ‘International Studies’ is broader still, bringing language and culture into the frame. Perhaps the broadest term is ‘globalisation’/‘global studies’, which eschews both ‘international’ and ‘relations’, and underlines a multisectoral perspective in which interactions of all sorts, and especially economic and common-fate ones, are given prominence. Most of these definitions prioritise the unit level, especially the state, and in so doing they privilege relations among the units as being the essence of what is meant by ‘world’ or ‘international’. But globalisation makes clear that the primary referent is the global system/society itself, in other words a material and social construction above the units. This distinction goes some way towards explaining the tensions between globalisation and the other formulations of IR, and why globalisation is sometimes in competition with IR both institutionally and conceptually. Although some concepts prominent in IR, such as most notably ‘international system’ or ‘international society’ appear to mark system-privileging approaches, more often than not they are in fact used as mere reference to the systemic environment in what essentially are actor-privileging approaches. Much of what we argue in the following is that a definition of IR, while not having to solely focus on system and a systemic level, must include a characterisation of IR’s subject matter in difference to its environment, that is, some boundary practice.
In what follows we will for convenience use ‘IR’ as a generic label to cover all of this, although we will return to these distinctions as necessary. It is notable that most of the big academic associations in the field use the broadest label, ‘International Studies’, in their names: for example, ISA (North America), BISA (UK), EISA (Europe), NISA (Nordic), CEEISA (Central and East European), KISA (Korea), MISC (Mexico), WISC (world). Exceptions to this practice are the Japan and Taiwan Associations of International Relations (JAIR and TAIR). As yet globalisation figures little at this institutional level, although it can be found attached to various institutes and programmes within some universities. Titles of journals are more eclectic. ‘International’ is prominent in most, but challenged by ‘world’ and ‘global’; and ‘relations’, ‘studies’ and ‘politics’ are all widely used. Globalisation is also in play in a few journal titles.
This labelling question exposes another oddity of IR as an academic discipline. Most social science disciplines are, as their names suggest, defined by function or the type of activity/behaviour studied: economics, law, politics, psychology, and sociology. This scheme suggests, wrongly, that the social world is neatly carved into distinct functional realms, for each of which a specific academic discipline claims prime competence and responsibility. Rather, the more puzzling observation is that this both is and is not the case. On the one hand, modernity is characterised by functional differentiation. Politics, law, the economy, science, etc. form distinguishable, though interconnected, parts or aspects of the social world. This functional differentiation is partly mirrored in science as well, as in Mathematics, Physics, Chemistry, and Biology.Footnote 9 This supposed division of labour does not stop some from seeing their particular discipline as somehow the master one, that, Lord of the Rings-like, binds all the others. Within the Social Sciences, Economics, Sociology, and up to a point, Politics are all prone to this sort of intellectual hubris. One expression of this is the way in which different disciplines pursue functional differentiation internally. Thus one finds innumerable hybrids and trespassings: political sociology, political history, the sociology of education, corporate law, international law, the sociology of law, etc., etc. One can track this all the way back to the initial formation of the social sciences, mostly during the nineteenth century, and how, for example, the study of politics gradually differentiated itself from the study of law.Footnote 10
Whatever label it uses, IR conspicuously does not seem to fit the model of functional differentiation. It is mostly defined not by function but by scale: its defining focus is on humankind and a global social, and mainly political, system as a whole. That makes it either, on the narrow view, a subdiscipline of Political Science, or on the broader view, an imperial discipline that seeks to colonise and integrate the macro side of all of the other social sciences and History. History also does not fit the functional model because it is defined not by function but by time. So History and IR have in common (a) that they are not defined by function, and (b) that they therefore necessarily trespass on the territory of other disciplines that are defined by function. This creates a tension that is very visible in academic life. Does the subject matter of History and IR get broken up and absorbed into the functional disciplines (for example, IR as ‘international/world politics’; economic historians being in economics departments, etc.) or do History and IR float free, and in some senses above the functional disciplines in the social sciences, pursuing their temporal and spatial missions across the whole range? Adding in scale (or space) and time, opens up yet more prospects for the internal differentiation of disciplines. Historical Sociology, for example, combines function and time, International Law combines function and scale, International Political Economy combines two functions and scale, International Historical Sociology combines scale, function and time. Disciplines thus not only differentiate themselves from each other, but also try to reproduce the others within themselves, reflecting an ongoing tension between the need to understand the whole, and the attraction of specialising in the parts. In this sense IR and History act the same as the more functionally defined disciplines, constructing internal differentiations that reach towards the whole.
As Buzan and Little note, this evolution is visible in how IR has unfolded over the past decades:
There is a narrow, somewhat traditionalist view that IR is mainly about relationships among states. This view largely locates IR within political science, confining its scope to the sub-area of ‘international’ or ‘world’ politics. Such a description might have fit early Anglo-American IR, but gradually the understanding of the subject has broadened, albeit with the ‘political’ element remaining at the core. Since the late 1950s, English School (ES) thinkers took both history and ‘international society’ seriously, and from the late 1970s onward, economics made its way back into the IR agenda. The ending of the Cold War saw an explosion of interest in sociological questions of identity, and in moral and legal questions of human rights. Over the last few decades, consciousness has thus grown that the object of study of IR is an international system which is not just a politico-military construct, but also an economic, sociological and historical one.Footnote 11
Posing the question of whether IR can legitimately claim one or more realms as its prime subject is motivated by a wealth of engagements between disciplines and by disciplinary developments that have taken place over recent years. Thus, for example, while IR has (re)developed a strong interest in historical international relations and the making of modern international relations,Footnote 12 the approach of global history has taken up a wide range of issues on the formation of global modernity and its political orders as well.Footnote 13 This immediately raises the issue of potential differences either or both in the subject matter studied or in the methodology applied to it. Similarly, while over the years IR has developed a growing interest in sociological approaches, many sociological approaches – most notably on ‘global’ or ‘world’ society – at least nominally seem to cover similar ground, raising the question of overlaps and differences here as well. In the sometimes bizarre world of academic identity and organisation, strange things happen. Historical Sociology, for example, is closer to IR in both respects than it is to either Sociology or History. And while sociologists have busied themselves ‘bringing the state back in’, so trespassing on the ground of Political Science, the ‘English School’ within IR has busied itself with exploring the subject of international society (more accurately interstate society), a concept either ignored or rejected by the bulk of Sociology which has only been able to think about world society.Footnote 14
Posing the question about the subject matter of IR in this sense means posing systematic-analytical questions. These are neither purely empirical nor purely normative, though they can never abstract fully from both. The empirical question would be about what is (or has been) studied in IR. The normative one would be about what should be studied, and to what purpose. The definition of the subject matter could then be a function of specific normative preferences or of envisaged practical use. The systematic-analytical question probed here, however, is about what subject matter can usefully and distinctively be studied by IR in the light of other disciplines’ approaches.
Skipping through journals, conference programmes, and book catalogues, IR seems to be a discipline of a rather hermaphroditic nature. On the one hand, there is a wealth of scholarly activity that deals with a specific subset of a functionally defined realm of the social world, that is various aspects of international politics as a specific form of politics. On the other hand, and particularly when it comes to ‘thinking big’ about issues of global society, the making of modern international relations, a capitalist world-system, etc., IR frequently ventures into directions where it becomes quite difficult to say what is not included in its subject matter, and how IR then differs most notably from Sociology as a discipline with more universalist theoretical ambitions to understand the social world.
While not without reference to empirical description and normative argument, the question explored here is about the delimitation of the realm underpinning the claim of disciplinarity. It is not about the question of whether IR ‘is’ a discipline in the first place.Footnote 15 Rather, it is about the claim that engaging with the subject of a discipline provides a marker of difference from other disciplines. As Rosenberg argues, in the social sciences disciplines legitimise themselves by being rooted in ‘a specific feature of social reality’ that enable them to function in a division of labour with other disciplines.Footnote 16 The issue at stake here then is the question of self-identity, which has troubled IR for so long, and makes it a bit embarrassing to be a member of it when facing disciplines with clearer and more robust identities. In other words, how can IR more effectively rationalise, and therefore ‘sell’ itself?
Engaging with these issues requires one to develop a position on the question of whether it makes sense at all to speak about ‘IR’ as a single ‘thing’, where this singularity could be asserted not only on the grounds of the existence of academic departments, people identifying with it professionally, conferences, journals, etc., but on grounds of identifying a common subject. There seems to be a remarkable gap between, on the one hand, IR textbooks that answer this question in the affirmative (this being their raison d’être), and, on the other hand, explicit reflections on this question that usually tend to come up with a sceptical answer.Footnote 17 Nonetheless, we are convinced that engaging with such questions, and particularly the question of what a discipline should be about, is a central task for any discipline in the sense that disciplinarity relies and is always in the process of being (re)constituted through cultural practices of boundary drawing. In this context, it is important that these cultural practices are strongly influenced by science, rather than, most notably, the economy or politics.Footnote 18
The subject matter and the traditions of IR
It could be said that the present contribution continues a ‘core’ IR activity in the sense that IR from its beginnings as an academic enterprise in the early twentieth century has always been about delimitation. The interwar years saw not only important moves of institutionalisation, but also intensive debates on the ‘nature and subject-matter of international relations considered as an academic discipline’.Footnote 19 These debates invariably concentrated on delimiting International Relations from other established disciplines. While obviously the relation to, and delimitation from, International Law played a major role in this regard, the possibility of IR becoming ‘boundless’ also was discussed as an issue, with hints to the danger that with a wide understanding of IR’s subject matter, ‘the study of International Relations would be practically identical with the study of Sociology in its widest sense’.Footnote 20 The quite obvious corollary of these debates on delimitation was problematising the difficulties of relating to other disciplines.Footnote 21 Indeed, both boundary-drawing exercises and emphases on the difficulties of relating to other disciplines underscore that IR is not a mere ‘field’ or ‘area’ of academic inquiry, but has strongly engaged in processes of the disciplinary organisation of scientific knowledge. Although an on-going practice that never really is ‘finished’ – and both, for example, perennial debates on whether to treat IR as a mere subdiscipline of Political Science or as a discipline in its own right,Footnote 22 as well as the present article attest to that – it is probably safe to say that disciplinary closure in the sense of both substantial definitions and institutional consolidations reached a first apex in the 1950s (and early 1960s), after the forceful interruption of the interwar process described above by the Second World War.
The history of IR has been written in many different ways and need not be repeated here.Footnote 23 The interesting thing to observe in this context, however, is that in this process of disciplinary consolidation, there always remained a considerable unease regarding the proper subject matter of IR. In order to illustrate our point we use Quincy Wright’s (1955) The Study of International Relations as an example, which probably remains the most elaborate early treatise that dwells specifically on some of the issues introduced in the previous paragraph.Footnote 24 The uncertainty about international relations as a subject and a field is aptly summarised in the preface to that book, when Wright observes that ‘there might seem to be little doubt that international relations exists. Yet there is some doubt on this point or, at least, on the sense in which it exists.’Footnote 25 He subsequently goes on to demonstrate that doubt pertains to both the subject area of International Relations (‘International Relations as a Condition’, Chapter Two) as well as to its status and development as a discipline (Chapters Three and Four).
While on these issues Wright echoes many of his contemporaries (and later writers), it is Chapters Five and Six of his book that warrant closer inspection for our purpose, that is the chapters in which he deals with ‘The Root Disciplines of International Relations’ (Chapter Five) and ‘Disciplines with a World Point of View’ (Chapter Six). In the former chapter, Wright is quite straightforward in identifying International Law, Diplomatic History, and Military Science as ‘root disciplines’ of IR, on which he remarks that as a discipline it ‘has developed synthetically and this has militated against its unity’.Footnote 26 His remarks become rather less straightforward, however, and immediately attest to the difficulties surrounding this subject matter, when he discusses the role of International Politics. While all of the root and other disciplines have dealt with International Politics, Wright seems to claim that International Politics constitutes a specific discipline that in practice is usually treated as a subdiscipline of Political Science. The important point here is the observation that International Politics has developed ‘as a distinct discipline in the modern world’.Footnote 27 The emerging discipline of International Relations, as a ‘synthetic discipline’ in this sense is composed of a number of root disciplines and a range of modern disciplines geared to studying things that are somehow ‘international’ (international politics, international trade). But the sense clearly is that International Relations is a synthetic and encompassing discipline when it comes to the international, and particularly cannot be seen as limited to International Politics.
This impression is reinforced in the chapter on ‘Disciplines with a World Point of View’. This chapter distinguishes between disciplines that take the division of the world into nation-states as a starting point, and those that take a view of the world as a whole. While initially introduced as a normative difference, the enumeration of disciplines in the chapter turns that into a mostly analytical difference. While this is not the place to discuss in detail all of the ‘disciplines with a world point of view’ discussed by Wright (and including, for example, also biology, operational research, and pacifism), it is most notable that he includes World History and Sociology under that rubric. Regarding the latter, he observes that although it has concerned itself more with smaller-scale phenomena, it has the potential to say more about larger-scale issues such as international relations.Footnote 28 As a result of these observations, Wright argues for further synthesis, meaning that the discipline of International Relations should integrate insights from the mentioned disciplines. Although not stated explicitly, from the thrust of his argument it is clear that this is to serve the purpose of establishing International Relations as a discipline with a world point of view. The unity of the synthetic discipline, in distinction to other disciplines with a world point of view, is then clearly derived in substantive terms: ‘It should be the function of a discipline of international relations to analyzed the entities, processes, forces, and relations in the international field’.Footnote 29 International Relations, then, is the discipline that specialises in studying the international field as a distinct realm of the social world. The question then becomes, of course: what is that field, and how is it delineated from others?
Wright himself, in Part V of his book, proceeds ‘Toward A Unified Discipline of International Relations’. After outlining that each discipline as a form of knowledge is ultimately closely connected to worldviews and conceptions of what could be called a ‘social whole’,Footnote 30 he professes a range of ideas of the many and multidisciplinary ingredients for an encompassing and synthetic discipline of International Relations. However, while he goes to great lengths to discuss the dependence of scientific knowledge on particular worldviews and comes up with a unified, if internally diverse, scheme for a discipline of IR, he ultimately fails to specify the meaning of the ‘international’ as the reference point for the unity of his endeavor.
Revisiting some of Wright’s arguments here primarily serves the purpose of identifying a point where one can usefully take off again when thinking about a discipline of IR. This is of course not to deny that much ink and useful thought has been spent on the subject during the last sixty years. However, what we claim is that starting with some basic questions is still relevant, and that basic questions about a discipline cannot be answered by referring to internal disciplinary developments or introspection alone, but need to reflect on a discipline’s subject matter(s) in the light of the subject matter(s) of neighbouring disciplines, and particularly in the light of other disciplines with a ‘world point of view’.
It is not uncommon for there to be a relationship between methodology and disciplinarity. Functional disciplines often legitimise their specific claim to knowledge in methodology: Economics has econometrics, Sociology has an orientation towards classical sociological theories on the one hand, and survey data on the other, and even Political Science, especially its American variant, has attempted to validate itself by turning to positivist methods. History, IR’s companion in not being a functional discipline, rests its claim to disciplinarity not just on being about the study of things in time, but also on historiography. The behaviouralist turn in various other social sciences might also be understood as a move to stake methodological claims, though of course there is also a discipline-dissolving aspect to the adoption of a common set of ‘scientific’ methods: all reduced to branches of mathematics. IR might be in some trouble here. If its basic claim rests on scale, then it more or less cannot go for a single methodology, but has to accept that it is multidisciplinary and therefore multi-method, unless of course some method is particularly suited to the study of scale.
Put in positive terms, the argument developed here is that theoretical and methodological plurality is something that could and should be embraced, including a plurality that crosses some of the boundaries that restrict the social sciences. However, it proceeds on the basis of the view that every theoretical and methodological plurality can only flourish if there is at least a minimal operative sense about the delimitation of a discipline’s subject matter, in order to prevent an ‘anything goes’ attitude leading to a study of all kinds of ‘contingency without practice’.Footnote 31
That said, the argument pursued here is not meant to exclude inter- or transdisciplinary work. Quite to the contrary, it proceeds on the assumptions that inter- or transdisciplinary work requires an understanding of the disciplinarity of knowledge. We acknowledge, however, that the reproduction of disciplinarity as well as obstacles for inter- and transdisciplinary work often stem not from substantive or methodological issues, but from the institutional realm underpinning the claim of disciplinarity. More often than not, it is people, departments, grant applications, and jobs that count in this respect more than anything else.
It is in this sense that we probe two different understandings of the subject matter of IR. The first of these argues that IR is about a specific realm, a part of social reality which can be clearly delineated and that warrants being studied comprehensively and ‘synthetically’ – this includes the possibility that the specific realm of IR is defined in terms of functional differentiation after all. The second argues that IR is not primarily about a clearly demarcated realm of the social world, and therefore different from clearly functionally defined disciplines such as Economics or Law. Rather, it is first and foremost about thinking about things on a macro scale. It is first about synthesis and, more than other disciplines, about the ‘world point of view’. We will at first introduce both understandings, not denying that these are highly stylised introductions, due to them being partly extractions of thoughts (if sometimes in the form of implicit assumptions and understandings) in our earlier work. In the next section, we will reflect on each position’s character as a world point of view and its relation to neighbouring disciplines.
Two paths to singularity
Our core question is about what particular parts or aspects of the social world warrant treatment by a specialised academic discipline of IR. We want as much as possible to avoid getting dragged into philosophical quagmires around the meaning of terms, because following that path would bring the boundary drawing exercise to a standstill. Thus, for example, while it might be possible to describe the international as a ‘part’ of the social world, this includes a reference to the philosophically highly problematic parts-whole scheme. Identifying it as a ‘level’ carries similar philosophical problems. We prefer the analytically more inclusive, and arguably philosophically less well-defined, terms of ‘realm’ and ‘macro-scale’ as we think that analytical and theoretical innovation thrives on such a practice of philosophically more inclusive theoretical specificity.
With this and the discussion above in mind, the next step turns to the issue of disciplinary boundaries in relation particularly to neighbouring disciplines, probing two lines of thought towards the possibility of delineating the boundaries of IR. In a first line of thought, we will probe the idea that IR is best identified as a discipline about a specific realm, ‘the international’, a part of social reality that can be clearly delineated and that warrants being studied comprehensively. In a second line of thought we will probe the idea that IR is not primarily about a clearly demarcated realm of the social world, but rather first and foremost about integrating the macro-scales of other disciplines via the concepts of ‘international system’ and ‘international society’. In the conclusions, we will weigh the merits and disadvantages of both lines of thought in relation to each other, also outlining the degree to which they are compatible with each other.
IR’s subject, Mk. 1: The ‘international’ as a specific realm of the social world
It is probably fair to describe IR’s disciplinary origins – combining insights from other disciplines, yet at the same time distancing themselves from those – as being built on some core idea that international politics forms a distinct realm in the world of politics in general (and, inter alia, in a wider social world). While this never meant that international politics was seen as a social reality completely (that is, causally) disconnected from domestic politics, the economy, and international law, it meant that the relations between states were governed by a distinct logic or dynamic not to be found in the domestic realm, thus constituting IR as some discernable sphere of its own. The existence of a distinct international realm is supported within IR by the longstanding discussion of an inside/outside differentiation based on the inside being a realm of order and progress, and the outside being a realm of anarchy and stasis/repetition.Footnote 32 While this of course means generalising to an extreme degree, it is probably also correct to describe the evolution of IR as a long process of widening its scope of inquiry. Attention to actors other than states, to various ‘levels’ of analysis, to a range of philosophical traditions, and to themes beyond the traditional confines of security and power politics to some degree would seem like an almost natural result of disciplinary institutionalisation and specialisation. However, this development entails the risk of becoming boundless: even a superficial inspection of book titles and journal articles under the ‘IR’ rubric reveals that over time hardly any actor, philosopher, or realm of the social world was able escape ‘discovery’ by an IR scholar.
When talking about a specific ‘realm’ as the ground for distinguishing IR, there obviously is a vast array of explicit proposals or implicit assumptions that have been made in this respect. IR could, in that sense, be about specific things that states do in relation to other states (as realism would have it when it comes to power); it could be the maze of political and economic entanglements in the world (as IPE would have it, as long as these entanglements do not remain confined to processes taking place solely within the boundaries of single states); or it could be about basically any- and everything, as long as it somehow involves something that spans or crosses a state boundary (which is what IR as a discipline, taken together, mostly feels like). Of course, there is an almost unlimited number of possibilities defining a subject matter in such a way.
Nonetheless, one part of all definitions of IR’s subject matter more than others has withstood the test of time, and that quite simply is the ‘international’. Whether seen as a substance itself – ‘the’ international – or as an attribute of something else: no IR without the ‘international’. This observation even holds for the problematisation of the international, be it in relation to its historical contingency or appearance, or in critical approaches regarding its disciplining power (both in terms of social practice and in terms of knowledge). One may not like the ‘international’ at all, but ‘doing IR’ seems to imply having to live with it one way or the other.
Taking such an open-ended and ever-present use of the ‘international’ as a common bond would seem to be a minimalist definition. However, any step further would call for a narrower definition of a specific ‘international’ (even if only in the negative form of problematising such definitions without a positive alternative statement). In addition, what all such definitions also require is to actually say how a specific international is distinguished from its environment.
IR arguably has not been tremendously good thus far in providing such a basic distinction. Some approaches, such as idealism and political realism are quite strong in this respect given their concentration on power political and (in the case of idealism) legal and social relations between sovereign states. This ‘traditional’ understanding of what ‘international relations’ is about is one that all but the hard realist faction in the discipline would find to be very limited analytically. Yet it certainly is the most unproblematic one in terms of parsimony: the realm of IR here is no more and no less than relations between sovereign states. It should be added, however, that such views of the international are rarely made very explicit, let alone are those espousing them often reflective about them. Almost all allusions to an international ‘system’ suffer from the fact that they usually do not define the system by identifying a relevant system/environment distinction, but by rather simplistic understandings around the ‘regular and patterned interaction’ motif. They build a sense of international system from the bottom up by looking at the interactions of units, and in the process neglect to specify both what the environment of this system is, and with what other systems it coexists.
Perhaps the most systematic contemporary attempt to come up with a definition of ‘the international’ that remedies the shortcomings and pitfalls discussed above is Justin Rosenberg’sFootnote 33 project to use the theory of uneven and combined development (UCD) deriving from Trotsky, precisely, and in his view for the first time, to theorise ‘the international’. He sees ‘the specific feature of social reality’ captured by ‘the international’ as being the coexistence within that social reality of more than one society, which he shorthands as ‘multiplicity’.Footnote 34 Rosenberg argues that UCD is a powerful way to theorise ‘the international’, in that the political and economic dynamics of its operation necessarily generate ‘the international’ in the sense of multiple units linked together in ways that reproduce the differentiation among them. This formulation provides the basis for IR’s place in the division of labour of the social sciences. It addresses the ‘environment’ question of within what ‘the international’ is located: it is one aspect of the social whole, and it is constrained and shaped by geography and technology. It also delivers on providing new perspectives and research agendas: whereas Kenneth Waltz sees a tendency towards homogenisation and ‘like units’, Rosenberg sees dynamics of differentiation.Footnote 35 What is less clear is how Rosenberg’s formulation of ‘the international’ plays into the ‘parts versus wholes’ division within the discipline. Its emphasis on multiplicity seems to privilege parts, but at the same time there is a major role for combination and dialectics that leans strongly towards the global social system.
For Rosenberg, unevenness is universal and always present in international systems, though its degree may vary a lot. There are two drivers for this universal rule: (1) the geographical separation of political units within a diverse physical environment; and (2) the differential impact on units of being combined. Combination is intrinsic to the definition of international system, and can also vary a lot. In the premodern world, combination varied mainly with geography, which facilitated it in some places (most notably the availability of sea and river transportation routes) and obstructed it in others (for example, land and climate barriers). Agrarian technology made a difference to the degree of combination (for example, the quality of ships and knowledge of navigation, up to a point the construction of roads) but did not dominate it. By contrast, in the modern world combination is heavily determined by industrial technology. Under the impact of steamships, railways, highways, aircraft, spacecraft, and electronic means of communication from the telegraph to the Internet, geography and climate fall away, and combination intensifies rapidly, and probably permanently.Footnote 36 Combination therefore increases directly with the third element of UCD: development. Combination is both a homogenising force (for example, around a dominant idea or model such as Westernisation or ‘standard of civilization’) and a differentiating one (for example, the dialectics of varied impacts and responses in the process of diffusion from core to periphery).
The spread of a new power configuration produces diverse outcomes. Each social order that encounters the new configuration has its own way of adapting to it. The ‘whip of external necessity’Footnote 37 produced by a new power configuration is often coercive, occurring through force of arms, as in the surge of Western imperialism during the nineteenth century. But intersocietal dynamics also take the form of imitation. Some societies do not take on the new configuration at all, or do so only weakly, either because of internal resistance to the social changes it required, or because of attempts by leading-edge polities to maintain inequalities between them by denying access to elements of the transformation: for example, China during the nineteenth century. Others succeed in developing indigenous versions of the new configuration: for example, Japan during the later nineteenth century. ‘Late’ developers are not carbon copies of the original adopters, but develop their own distinctive characteristics. In this sense, the interactions between different social orders produce not convergence, but (often unstable) amalgams of new and old: German industrialisation was not a replica of British development, and Chinese development has distinctive ‘Chinese characteristics’. The scale and intensity of combination within the international sphere has increased, meaning that every society becomes less self-contained and more exposed to developments elsewhere. As societies become larger in scale and more functionally differentiated, differences between them are accentuated and interactions between them intensified. The mutual constitution of combination and unevenness is thus intensified by development, producing larger, more complex social orders bound together in denser, more interdependent ways.
Rosenberg’s account tells us how ‘the international’ arises, and how it works. In that sense, it is both a clear advance on realism, which just assumes the international into existence along with states, and a radically alternative vision to neorealism’s ‘like units’ assumption. Rosenberg’s ‘international’ is not just political, but also deeply social, economic, and physical. It makes a powerful case for the distinctiveness and importance of the international as a social realm.
Against the background of these observations, it is clear that coming up with a delineation of the subject area of IR is not in itself a problematic exercise. However, one might legitimately wonder whether and to which degree even a more complex attempt to define the ‘international’ as in the works of Rosenberg can in the end be successful if such a definition is supposed to serve as a foundation for underpinning a disciplinary delineation for IR. UCD is quite close to theories of modernisation and thus might be claimed by Sociology. UCD might also be seen as an outright version of International Political Economy, though that raises the question of whether IPE is inside IR (as is mostly the case institutionally) or something more encompassing than IR. Rosenberg might well claim that UCD settles this old question by forcing IR and IPE to integrate.
But even ‘the international’ is not without its problems as a delineation for a discipline of IR. Mathias Albert’s recent A Theory of World Politics identifies ‘world politics’ as a distinct realm of the social world, as one specific subsystem of the political system of world society. One could take the view that this actually provides for a definition in the way looked for in the present context: IR would then simply be the discipline concerned with everything that takes place within the system of world politics (as a ‘narrow’ definition), or somehow in relation to it (as a ‘wide’ definition). Such an approach also fulfills the requirement of a definition that it needs to account for how the subject matter as a part of the social world forms a distinct part of that social world and on what (historical and systematic grounds) it is thus distinguished (or differentiated).Footnote 38 The problem that comes up immediately then is that while by all means A Theory of World Politics is an ‘IR book’ in institutional terms, it is only partly so in terms of content: in fact, it arrives at its definition of world politics almost exclusively through the reading of sociological and historical, and not IR literature. Would this then lead to the conclusion that History and Sociology necessarily remain the more general disciplines, with IR bound to be more ‘specialised’, and forever remaining incarcerated in the ‘prison of political science’?Footnote 39
IR’s subject, Mk 2: Things on a macro-scale: the international system
The previous section was concerned with defining IR through a realm of the social world that by and large was defined through being distinct from other parts of that social world. Regardless of whether this distinction historically and theoretically is seen to be the result of, for example, a functional differentiation within a wider system of politics, or a segmentation within the political system that allows for an ‘international’ to emerge in the first place: the important point is that there is a distinct realm that is not wholly covered by other social science disciplines – although they might be interested in aspects of it.
The second way of delineating IR’s subject discussed above would in a strict sense also pertain to identifying a specific realm in the social world. However, this realm would not be delimited from other realms in terms of, most notably, functional differentiation or segmentation. It cuts across these forms of social differentiation and includes all those parts of the social world that are structurally relevant. The realm of ‘international relations’ would then be seen purely in terms of levels. The subject of IR would, in that sense, be all the subjects of disciplines defined by functional differentiation, but only those parts of those subjects that pertain to a certain scope (variously defined as ‘systemic’, ‘global’, ‘international’, ‘macro’, etc.). IR would be the discipline for things on a macro-scale. Without necessarily pegging such an understanding of macro-scale things to specific concepts of systemic quality, Waltz’s transposition of an analogy from economics to the study of international politics provides a useful heuristic here. Waltz remarks that just like understanding the (world) economy (as an issue of macroeconomics) cannot function by analysing individual firms (as the prerogative of microeconomics), the study of international system cannot succeed on basis of studying (the behavior of) individual states.Footnote 40 We have something similar in mind with the second definition. The two main and fundamental differences between our understanding and Waltz’s first pertains to the fact that the realm studied is not limited to (international) politics, but cuts across all kinds of social relations (for example, economic, legal, scientific, etc.). And our definition leaves open the possibility that both differentiations among actors, and variations in interaction capacity, do matter on a structural level.
As we are interested in a possible definition of IR, it almost goes without saying that ‘levels’ here can only mean levels in a basic structural sense. This is not a point of mere scholastic sophistication. A definition of a discipline’s realm through a ‘pure’ level of analysis (if there can be such a thing at all), in the end privileges a methodological definition. Levels of analysis by definition pertain to a specific perspective, a specific angle of looking at the social world. They do not make explicit statements about what the most relevant structural levels are.Footnote 41 This does not mean that this would be an easy task. Structures are easy to talk about, but at the same time they are notoriously difficult to observe and describe empirically. Meaningful analysis in this respect cannot be limited to the description of the existence of specific structures, but requires an account of processes that lead to their constitution and transformation.Footnote 42
We suspect that there is no specific threshold of what in that sense counts definitely as macro-scale for constituting the proper subject matter of International Relations, particularly as larger structures can be influenced and ‘punctuated’ in their transformation by seemingly ‘small’ developments and events.Footnote 43 We would rather think of this threshold as a kind of analytical ‘burden-of-proof’ rule: that it is necessary to make it plausible that some issue or question has relevance on a structural level. The important point is that a reference to a structural level and analytical rule of being macro-scale is completely indifferent to functional specification – the structures in question can be legal, political, economic, etc. It is relatively indifferent to historical scope. And it is at least somewhat indifferent to geographical scale (with a tendency towards a larger extension, but bearing in mind that important things often start at very specific localities).Footnote 44
Although it might be impossible to identify an exact ‘scale’ in this respect, the term ‘international’ might in fact serve as a useful heuristic here in two respects. On the one hand, structural effects need to be observed by a significant number of observers on a recurring basis in order to become structural in the first place. It is thus, at first, not even necessary that any interaction spanning nation-state boundaries takes place. Some local politics and events might have structural effects if reported and remembered worldwide, whereas – vice versa – much international interaction will not arouse much interest or will be forgotten in the public quickly. It is, however, quite likely that there is a positive correlation between interaction intensity and scale on the one hand, and the intensity of its observation on the other. Put differently: it is more likely, but not necessary, that international interactions have structural effects on a larger scale rather than local interactions having them.
On the other hand, however, the ‘international’ can serve as a useful reminder that not all structural effects need to be ‘global’, and that in fact quite a lot of structure formation takes place in regional or other non-global (but not necessarily regionally bundled) international systems.
It could be argued that such a description comes relatively close to what IR actually ‘does’, if what is published in academic journals and discussed at academic conferences is to serve as a useful guide in this respect. However, we would like to point out that scale here should be important. Using the illustrative analogy drawn from Waltz again, one could say that while international business practices are certainly worthwhile studying, they would meet the macro-scale test only if they can be seen to have an impact on the world economy (which, for example, is mostly not the case in small or medium-sized companies, even though they might be very active on international markets). Equally, the many cases of an unequal distribution of income and other assets between groups or even states would only meet the test if it contributes to an overall picture of global inequality or is of a scale that it obviously has effects way beyond local contexts (for example, the boundaries between rich and poor in the Mediterranean, the US-Mexico border, etc.). Taking this illustration closer to one of the perennial discussions of the discipline, that is the question of whether foreign policy analysis is or should be a part of IR, the answer is obviously that this depends primarily on whether the policy decisions analysed can be argued to have significant structural effects (although in many cases this might often only be ascertained with the benefit of historical hindsight).
It should be clear by now that seeing IR as being about macro-scale things does offer a heuristic, but not a clear-cut definition of what is definitely ‘in’ and what is definitely ‘out’ under such a definition. The macro-scale does not, however, refer to the issue of whether we are talking about ‘grand theory’ in this respect, that is, comprehensive conceptual and theoretical frameworks for making sense of the worlds of IR.Footnote 45 Although both aspects often are related, they are not necessarily so. The difference here is neatly captured in a distinction that is less prominent in English than, most notably, in German, that is the distinction between ‘social theory’ (‘Sozialtheorie’) and ‘theory of society’ (‘Gesellschaftstheorie’). While the latter is often translated into ‘social theory’, there is a marked difference here. Social theory in whatever guise refers to fundamental characteristics of sociality. It can take its starting point in individual face-to-face interaction settings, assumptions about the rationality of actors, signs and signification, the relation between humans and things, etc. In that sense, any theory of society will invariably have an underlying social theory, but it takes its starting point in the assumption that macro-scale social structure formation has taken place and is mostly interested only in those parts of social theory that contribute to understanding society.
Two worlds of IR and their consequences
To summarise the two worlds: the first is where IR is concerned about a specific realm that is distinguished from other parts of the social world through forms of social differentiation and the quality of social relations (where approaches explicitly drawing on theories of social differentiation would emphasise the former, and the concept of uneven and combined development would emphasise the latter, as precondition for differentiation to take place). IR would be the discipline that studies that specific realm. The second world is a world viewed through the lens of levels – not only analytical levels, but levels of social structure. IR here would be the discipline that studies what goes on at the higher levels of the social world, quite irrespective of functional specification. In analytical terms the first world seems to require specialisation, and an in-depth knowledge about a specific (if admittedly still rather huge) part of the social world. It would not discount the connections between that part of the social world and other realms (most notably the economy and law), but would always see these connections through the lens of relevance for, and the question of how they feed into, the dynamics of world politics or the international. The second world seems to take the opposite (and in this sense also ‘wider’) analytical approach, basically considering everything in the social world irrespective of the specific quality of the international or functional specification, if only it reaches macro-scale. Both worlds also entail specific delimitations of what constitutes the relevant social environment for the analysed subject. If the first world is seen in Albert’s terms, the relevant social environment would primarily be everything in the political and other functionally differentiated systems of society that is not ‘world’ or ‘international’ politics. If it is seen in Rosenberg’s terms, geography is the primary environment, plus those aspects of social reality that are not driven by UCD. In the second case it would be all processes in the social world beyond a specific macro-scale.Footnote 46
Albert’s option points more to a subject that could be seen to be located somehow ‘within’ political science, whereas Rosenberg’s is necessarily multi-disciplinary, though quite how far the scope of UCD reaches is not wholly clear. Does it cover the social totality or are there parts of social reality outside it? The second world would probably see IR as a far broader and encompassing form of social science.
We deliberately do not come down in favour of one or the other option, and acknowledge that they might not exhaust the possibilities for defining the subject matter of IR. We do claim, however, that these are two very promising and relatively coherent options that any alternative account trying to provide similarly comprehensive understandings of the subject matter of IR would have to engage with. Whatever ‘world’ of IR is being favoured will ultimately depend on both specific cognitive interests as much as on the varying personal sense of urgency to provide an understanding of the field that defines daily professional proceedings. The message that this contribution seeks to convey is that of a delineation of IR as a discipline asking a range of quite fundamental questions: how are relevant social systems delineated from their environments and what are these environments? How is a social whole structured, differentiated, carved up into ‘parts’, ‘levels’, (micro-/macro-) ‘sizes’? Which of those systems, parts, levels, sizes are or should be the focus of IR as a discipline? What do we mean when we talk about concepts like, for example, ‘structure’ and ’levels’ not in abstract philosophical terms, but in relation historically specific configurations and in there embeddedness in (which always entails a distinction from) a social environment?
We are aware of the fact that these basic questions are simply put yet complicated to deal with, as much as we are aware of the fact that they are not at all novel questions: in the history of IR as a discipline they are actually quite old questions. Asking these questions, and asking them anew repeatedly is not at all the prerogative of the present contribution. It is in fact what the widely discussed ‘end of IR’-debate has all been about,Footnote 47 it is what the so-called (debates on) ‘great debates’ have been about, and it is what the various constructivist or practice turns and twists have mostly been about. The present contribution asks these old questions again, yet also gives possible answers categorically different from most of those in the earlier debates. We propose these answers should be ‘substantive’ in the sense that it is the reference to a specific ‘realm’ of the social world (whether defined in terms of social differentiation or scale) that could best hold the discipline of IR together aspirationally, while retaining a high degree of theoretical (and philosophical) openness. We acknowledge the fundamental paradox of this exercise, as it entails and in fact fuels an ongoing process of the simultaneous disciplinary integration of knowledge (at least in aspiration), while at the same time relying and reproducing the disciplinary delimitation of knowledge. It is, however, the permanent negotiation of, and finding new resolutions for this kind of paradox that drives the evolution of scientific knowledge, rather than the discovery of novel things that ‘are there’.
Acknowledgements
We would like to thank Justin Rosenberg as well as the reviewers of this article for their very helpful comments.
