Explaining the territorial trap: Revisiting the geographical assumptions of International Relations theory

Abstract

Territory and territoriality lie at the heart of both world politics and International Relations (IR) theory. In terms of IR theory’s geographical assumptions, one of the most influential studies to date has been political geographer John Agnew’s 1994 article on the ‘the territorial trap’ (TTT). While Agnew’s original insights and subsequent research has reached canonical status in political geography, mainstream IR scholarship has yet to fully engage TTT. Political geographers, in turn, have largely dealt with the consequences of TTT for our understanding of world politics. This study offers the first detailed account of the origins of TTT, which are hidden in broad daylight in IR’s own history. The origins of TTT and mainstream IR are intertwined in terms of two dynamics: the racist and colonial origins of IR, and the selective nationalistic ontology that dominated IR especially in the first half of the 20th century. The arguments offered in this study have a wide variety of implications for problematising the ways in which IR-as-epistemological-community approaches territory and territoriality as well as our understanding of the origins and evolution of the present-day global territorial order.

Not that long ago, numerous scholars declared the imminent death of territory.¹ More specifically, the end of the Cold War and the rise of globalisation throughout the 1990s emboldened the idea that the importance of territory per se and the salience of the so-called territorial state were both withering away. In the long run, the ‘birth of a post-territorial global order’ thesis proved to be premature. Simply put, ‘territory retains its allure’.² Territorial disputes persist, and governments are still extremely jealous of their own territories. The Russo-Ukrainian War, in turn, is a crude reminder that while some spectators may claim that wars of territorial expansion are a thing of the past, especially in Europe, the very facts of international politics do not agree with such claims. This is hardly surprising. Territory has not really made a ‘comeback’ in world politics; it simply never left.

International Relations (IR) scholarship’s interest in territory, especially territorial conflicts and disputes, blossomed during the course of the 1990s.³ The relevant research programme, to a large extent, has been organised around not specific theoretical foundations, but (quantitative) methodology.⁴ Research on territory’s role in global politics is not limited to IR. Political geographers have long studied the concepts of territory and territoriality in great detail.⁵ However, despite numerous intellectual common grounds, interdisciplinary dialogues between IR scholars and political geographers have been rather limited.⁶

One of the earliest attempts to initiate an interdisciplinary dialogue between IR and political geography is John Agnew’s ‘the territorial trap’ (TTT) intervention, published in 1994 in the Review of International Political Economy.⁷ In the article, Agnew offered a detailed analysis and constructive criticism of IR’s perspective on territory and territoriality, arguing that IR theory is built on certain implicit geographical assumptions that effectively limit its understanding of territory. Agnew identified three core issues: (1) unqualified reification of the so-called territorial state; (2) failure to conceptualise how boundaries emerge and function; (3) conflating ‘society’ and the ‘state’ in the context of the space–identity–politics nexus. Without recognising the flawed and misleading nature of these assumptions, Agnew argued, IR cannot move beyond the territorial blindfold that not only limits but also distorts its theoretical and historical vision. In that, Agnew’s territorial trap intervention pointed towards ‘the need to look at alternative and competing spaces of political practice’.⁸

Agnew’s intervention, however, had a paradoxical impact on the broader study of territory: TTT achieved canonical status in political geography but barely registered in subsequent IR research. The most salient factor that rendered TTT rather ‘invisible’ in IR entailed a core tendency: an overwhelming majority of IR scholars, even those who work on territorial conflicts, rarely define what territory and territoriality stand for in their research but instead treat them as self-obvious and ‘natural’ phenomena.⁹ The result is an exceptionally ambiguous understanding of territory, which also usually leaves out the ideational (and therefore intersubjective) aspects of the term.

Especially in the last decade, IR research on territory and territoriality took a constructivist turn, if partially.¹⁰ This line of inquiry does not take territories as mute and static objects but treats them as products of intersubjective ideas about the association between space, society, and politics: for territories to exist in any meaningful sense, human groups have to think of the ways in which space, society, and politics interact, and then act on these ideas.¹¹ In many ways, the recent ‘constructivist turn’ in the study of territory resembles a more cumulative version of the constructivist approaches to territory in IR during the late 1980s and early 1990s, exemplified by the works of scholars such as John Ruggie and Friedrich Kratochwil.¹² Arguably, just as the end of the Cold War motivated some IR scholars to explore the ideational aspects of territory, the multifaceted role that territory plays in an increasingly uncertain world is similarly motivating numerous scholars to unpack the socially constructed nature of territories. Compared with the 1990s, constructive and critical approaches have become more prevalent in IR scholarship, which itself has become, if relatively, more self-critical in the past decade or so. In this context, some 30 years after its publication, Agnew’s TTT intervention is becoming (even more) relevant to IR research on territory and territoriality.

Political geographers and (some) IR scholars have long examined the consequences as well as the resilience of TTT but have rarely dealt with the following question: how can we explain TTT? This is an important question for two main reasons. First, moving beyond TTT requires understanding why it exists in the first place. Second, an interdisciplinary approach that scrutinises TTT’s origins can help transform the ‘trap’ into a platform where scholars from different sides of the disciplinary fences can speak to one another in infinitely more constructive ways.

The origins of TTT are hidden in broad daylight in IR’s own history. The origins of TTT and mainstream IR are intertwined in terms of two dynamics: the racist and colonial origins of IR, and the selective nationalistic ontology that dominated IR especially in the first half of the 20th century. From a territorial perspective, the modern state system was built on colonialism- and racism-induced territorial hierarchies. The days of colonialism and blatant racism might be over, but their legacies still persist, not only in world politics, but also in IR. The nationalistic ontology that implicitly defined IR in the 20th century, in this context, privileged the nation-state form and its territorial underpinnings, establishing ‘nation-state’ as the territorial gold standard in world politics. Especially during the first half of the 20th century, this ontology was weaponised by the hegemonic powers of the Global North to sustain and perpetuate the territorial hierarchies that followed from geographical racism¹³ and colonialism. When IR came of age in the post–World War II era, it inherited the dominant geographical meta-narratives of the Western world; IR scholars’ reluctance to explicitly define and scrutinise territory and territoriality then rendered TTT rather resilient, or more precisely, almost invisible.

The remainder of this article is organised into four main sections. The first briefly elaborates on TTT. In the second section, I offer a discussion of the notions of territorial hierarchies and geographical racism. The third section explores the impacts of the selective nationalistic ontology that dominated IR especially in the first half of the 20th century. Finally, I briefly weigh in on the implications of the arguments and potential venues for future research.

Making sense of the territorial trap

During the 1990s, just like today, IR as a whole did not profess explicit geographical assumptions. Apart from exceptions such as Friedrich Kratochwil, Richard Ashley, and John Ruggie,¹⁴ most leading IR scholars of the time took territory and territoriality for granted, as if they were concepts so self-obvious that they required no attention whatsoever.¹⁵ Agnew’s main contribution was to identify the implicit geographical assumptions in IR theory, which he arranged in three categories. The first entails the reification of ‘state territories as set or fixed units of sovereign space’, which in turn serves to ‘dehistoricize and decontextualize processes of state formation and disintegration’.¹⁶ Second, modern IR theory draws upon the polarity of domestic and international politics, which then obfuscates the ways in which these two dimensions interact. The third component of the territorial trap, in turn, involves viewing ‘the territorial state as existing prior to and as a container of society’, which artificially normalises the complicated and multifaceted nature of space–society–politics interaction simply by assuming away its complexity.

Overall, Agnew’s main point was straightforward: in their research, IR scholars were making some very strong geographical assumptions without either recognising or acknowledging that they were making geographical assumptions in the first place. IR scholarship’s failure to engage its own geographical assumptions, in turn, deludes our collective understanding of territory’s role and place in world politics, erasing the extant socio-spatial complexities across time and space. This tendency then ossifies a [territorial] version of what R. B. J. Walker dubbed the ‘discourse of eternity’ that dominates mainstream IR theories,¹⁷ where the interactions between different kinds of states, different kinds of territorial arrangements across time and space are all reduced to the interactions between similar (territorial) units. Conversely, territorial ideas and arrangements that do not ‘fit’ the dominant Eurocentric geographical imagineries are regarded as (extra-systemic) abnormalities or insignificant footnotes that did not matter as much as the relations between the Western states.

Take the example of mainstream IR’s approach to the so-called peaceful long 19th century between 1815 and 1914. IR scholars have carefully assessed the existing datasets and established that this specific ‘international’ peace was ‘remarkably’ robust.¹⁸ If we are students of the truly ‘international’, and not merely students of the Global North, a major conceptual – and not methodological – problem with this approach instantly reveals itself. In IR, ‘war’ implies a specific type of organised violence, between specific type of political-territorial entities, and it is not a coincidence that the baseline, or the norm, for such a conceptualisation is essentially Eurocentric.¹⁹ The relevant datasets (and the subsequent analyses based on them) then follow this Eurocentric conceptualisation and erase colonial violence from the long nineteenth century.²⁰ In this reading, the territories in the Global South, and the hundreds of millions of people living in these territories, have not always been seen as part of the ‘international’ geopolitical space. Once we moved beyond TTT, the peaceful long nineteenth century reveals itself to be an essentially European peace, which was packaged as the international.

With hindsight, Agnew’s TTT intervention can be seen as both a success and a failure. On the one hand, it is a success story insofar as it has influenced how political geographers – and, to a much lesser extent, some IR scholars – approached territory and territoriality in the context of global politics. On the other hand, in terms of its target audiences, that is, IR scholars, TTT intervention was also a partial failure. If we are to define ‘mainstream IR’ vis-à-vis scholars’ responses to surveys conducted by the College of William and Mary’s Teaching, Research, and International Policy (TRIP) Research Lab in terms of ‘top/relevant journals’,²¹ outlets where careers are literally made or broken, in mainstream IR, TTT remains almost invisible.²² This is hardly a hyperbole. While Agnew’s 1994 article reached canonical status in political geography (while also attracting more than 3,500 citations as of 2023), a simple Google Scholar search suggests that the term the ‘territorial trap’ has not appeared even once in leading IR and/or political science journals such as the American Political Science Review, the American Journal of Political Science, International Security, and the Journal of Conflict Resolution.²³ World Politics, in turn, appears to have published a single article where TTT is invoked.²⁴ International Organization and International Studies Quarterly fare slightly better, as each has published six articles citing (but, not necessarily exploring) TTT.

On the surface, it is rather easy to explain mainstream IR’s indifference towards TTT: either Agnew’s intervention did not add up to much, or IR has already moved beyond TTT. Both arguments would be incomplete, for two reasons. First, as highlighted above, IR research beyond critical approaches²⁵ has never fully engaged TTT. Accordingly, mainstream IR neither analysed and then decided to dismiss TTT nor consciously moved beyond it; TTT was simply ignored. Second, more importantly, engaging or moving beyond TTT requires a necessary condition: defining territory and territoriality explicitly. However, IR scholars rarely define territory, even when their research entails a key territorial component.²⁶ In this sense, moving beyond TTT becomes possible only by defining the territorial. Conversely, invoking territory and territoriality without explicitly defining these terms empowers the idea that TTT does not exist or that it has already been transcended. More precisely, while Agnew’s specific criticisms themselves are open to debate, his core argument retains its validity: IR scholarship would be better off if it further conceptualised and theorised territory and territoriality.

As political geographers have long recognised, territory is not a synonym for physical space.²⁷ ‘Territories are not frozen frameworks where social life occurs’; instead, they are made (and remade) through social and individual action.²⁸ Territory is about people as much as it is about physical space.²⁹ For territories to emerge, human groups need to develop institutionalised ideas about the space–society–politics nexus and then act upon these ideas. In other words, territories are what states and societies make of them.³⁰ If we adhere to the simplest definition of territory, territories emerge when (1) physical space is demarcated for social and political purposes; and (2) demarcated space is organised in order to manage and regulate the association between space and society.³¹

Of course, since 1994, IR research became far more reflective in numerous ways, and partially moved beyond some of the individual assumptions that collectively constitute TTT; most notably, numerous IR scholars have transcended the ‘domestic–international’ (or, IR–Comparative Politics) divide in recent decades.³² In many ways, these developments suggest that IR is now more amenable to the task of moving beyond TTT. More specifically, the ‘constructivist turn’ in the study of territory renders TTT not only (even) more relevant, but also far easier to engage and transcend. Of course, not all IR scholars need to engage and move beyond TTT. However, for IR research that builds on the concepts of territory and territoriality, engaging TTT in particular, and the relevant research in political geography in general, offers numerous opportunities in terms of novel research questions or new ways to examine existing ones. It is at this juncture where ‘explaining TTT’ becomes important: especially for IR scholars, moving beyond TTT is possible (or feasible) only by scrutinising why it existed in the first place.

Territorial hierarchies and geographical racism

At first glance, TTT is best seen as a by-product of what can be referred to as the nationalistic ontology.³³ However, how IR ‘interpreted’ this ontology was shaped by, and filtered through, geographical racism and territorial hierarchies. Temporally speaking, geographical racism and the notion of territorial hierarchies preceded the rise of the nationalistic ontology, making it essential to elaborate on these topics before delving into the aforementioned ontology.

IR was born into and out of the ideological currents of the late 19th and early 20th centuries in the Global North.³⁴ In that, IR emerged as an independent field of study partially (or mostly) to serve and inform colonial administrations.³⁵ This foundational episode was never truly about the ‘international’; instead, it was founded on the assertion that the ‘Western’ could be a substitute for the international. By implication, non-Western geographies did not (yet) deserve to be a part of the international geopolitical space.³⁶ IR’s very own Western origins and its infatuation with European history, or the history of the Global North, partially mask this fact, but they also prove the point: from its early days, IR was construed as an ‘international’ enterprise, but it was produced and consumed mainly in the Global North, reflected the biases of the Global North, and, for a long time, took pride in serving the interests of the Global North.

For generations, IR scholars started from the assumption that the secrets to understanding the ‘international’ lie with understanding the ‘European’, treating the Western experiences as ‘the only true subjects of history’.³⁷ In such a setting, dominant IR approaches imagined the birth and evolution of the modern state system and colonisation/imperialism as almost two distinct topics.³⁸ However, colonial [territorial] arrangements survived well into the second half of the 20th century.³⁹ By 1900, more than 60 per cent of the globe was governed by imperial rule or some form of colonial dependency; the same ratio was roughly 40 per cent in 1920, and more than 20 per cent in 1960.⁴⁰ In such a setting, the ‘West’ did not emerge and evolve in a political, economic, cultural, and geographical vacuum; on the contrary, it evolved into what it is today because of its interaction with the Global South.⁴¹

At the heart of this ‘interaction’ lay two interrelated dynamics: territorial hierarchies and geographical racism. Political geographers have long studied the notion of territorial hierarchies: territorial arrangements not only in the past but also in the present clearly suggest that some territories have been more equal than others.⁴² From such a vantage point, the so-called modern state system is better defined as a territorial caste, where the Global North has always sat at the top. Roughly from the mid-17th century onwards, European states began constructing a ‘European [territorial] club’, where members of the club eventually came to respect each other’s borders and territorial sovereignty during peacetime. When the Europeans turned to the [sic] ‘New World’ (new, according to whom?), Africa, Asia, and beyond, they saw inferior, open territories, which were deemed ripe for the taking. Even ‘international’ law was instrumentally deployed to ‘reorder colonized spaces and bodies’.⁴³ One notable example is the concept of terra nullius, or ‘territory belonging to no one’.⁴⁴ Most notably, in European eyes, in contrast to Europe, there were no real borders and boundaries to respect in the aforementioned open spaces.⁴⁵

In fact, European powers first experimented with modern cartography and the concept of linear borders not in Europe, but in the Western Hemisphere:⁴⁶ the European landscape was ‘crowded’, in terms of not only (or primarily) population, but also a multitude of political actors and overlapping (spatial) jurisdictions. The ‘New World’, by contrast, was perceived as an unpopulated, empty land. Of course, there existed millions of human beings and a wide variety of political and socio-spatial arrangements in these lands. They just didn’t matter that much – that is, in the eyes of the Western colonisers.⁴⁷

For example, while the Western states conceived of the relationship among themselves in binary terms, that is, war and peace, their approach to actual and potential colonies entailed the notion of ‘permanent war’.⁴⁸ The omnipresent nature of this colonial (organised) violence, however, was essentially asymmetrical. It was the European/Western colonial powers that were directly or indirectly responsible for this perennial war, but the same powers rarely thought themselves to be waging ‘real’ wars; real wars, different from colonial military campaigns and ‘emergencies’, were to take place in Western lands or between Western powers. In the end, these targeted geographies were merely ‘unruly’ spaces, waiting and deserving to be tamed and transformed by the European powers. This was the territorial logic of colonisation.⁴⁹

So, where do these territorial hierarchies come from? IR scholarship has long paid attention to the so-called anarchy–hierarchy dichotomy in world politics. Typically, IR researchers scrutinise the power (a)symmetries and/or institutional arrangements at the interstate level to make their case, regardless of the position they take.⁵⁰ However, neither institutional arrangements nor power asymmetries can account for the aforementioned territorial hierarchies all by themselves.⁵¹ One crucial dimension to consider is the question of race and racism in world politics and IR.

As numerous IR scholars have recently emphasised, even the concept of (territorial) sovereignty cannot be thought to be independent of the question of race.⁵² As Kelebogile Zvobgo and Meredith Loken aptly put it, ‘race is not a perspective on international relations, it is a central organizing feature of world politics’.⁵³ In the words of Freeman, Kim, and Lake, ‘the invisibility of race in IR was not accidental but designed’.⁵⁴ Especially during its formative decades, IR openly reflected, fed off of, and fed the rampant racist biases of the time.⁵⁵ In world politics, in turn, racism is not only about a certain group of people; it is also about geography.⁵⁶ Put differently, in world politics, racism is geographical. Since the dawn of the age of colonialism, European powers portrayed other geographies as empty, backward, barbaric, feminine, and dangerous lands.⁵⁷ The reason they did so – or the reason that the practice of portraying ‘other’ geographies in such demeaning and inferior ways seemed ‘natural’ to most spectators – followed from the inherent geographical racism embedded within colonialism.

Of course, the dominant ideational currents in the Western world were neither limited to, nor can be reduced to (geographical) racism. For example, capitalism and liberalism also played important roles in the ways in which the West approached the Rest. Still, during the early decades of IR, both of these ideational currents (and associated practices) dovetailed very closely with racism, far more explicitly than they do in the present. More precisely, capitalism and liberalism played significant roles in the formulation of the ‘Western gaze’,⁵⁸ though usually through the prism of geographical racism. As Rosa Luxemburg highlighted, by its very nature, the capitalist system needed to expand and to do so almost indefinitely.⁵⁹ For capitalists, ‘open, inferior’ geographies meant few, if any, legal and [sic] moral constraints in terms of exploitation. For instance, King Leopold of Belgium unleashed one of the most vicious exploitation-for-profit campaigns in modern history in Congo; capitalism might have driven Leopold’s calculus, but it was geographical racism that allowed him to violently exploit the Indigenous populations, at the cost of millions of human lives. Similarly, while it was the ‘cotton capitalists’⁶⁰ that helped institutionalise slavery in the pre-Civil War United States and beyond, the slave trade was a viable and acceptable option precisely because sub-Saharan Africa, which was conceived as an open, inferior collection of territories, could be seen as a source of forced labour. The ‘jointness’ of racism, liberalism, and colonialism, in turn, is well documented.⁶¹ In many ways, racism and liberalism collectively ‘legitimized expansion into [non-Western] spaces’, leading to ‘the re-articulation of imperialism as a progressive practice’.⁶² This (liberal) geopolitical vision was propelled by the self-adopted European mission to ‘civilise’ uncivilised populations and geographies under the pretext of universalising principles.⁶³

On both accounts, geographical racism acted as the meta-narrative which other ideational currents worked through and with. To cut a long story short, when IR was ‘born’ as a scholarly field, geographical racism and the associated territorial hierarchies constituted the widely accepted, broadly defended, and internalised characteristics of ‘international’ politics. In such a setting, (geographical) racism drove not only the policies pursued by the hegemonic powers of the Global North, but also how IR scholarship thought of itself and the world it was studying.

For instance, the very first American Political Science Association (APSA) Meeting in 1904 designated ‘Colonial Administration’ as one of the five core branches of ‘politics’.⁶⁴ From such a vantage point, it is no surprise that Foreign Affairs, the first IR journal in the United States, was initially established as the Journal of Race Development (JRD) in 1910, just as the United States was flexing its colonial muscles in the Pacific. The JRD’s introductory essay specified the journal’s foundational purpose in terms of paternalistic racism, that is, ‘to discover, not how weaker races may best be exploited, but how they may best be helped by the stronger’.⁶⁵ In an article Foreign Affairs published in 1992, on the journal’s 70th anniversary (which takes 1922, when the journal was renamed as Foreign Affairs, and not 1910, as the journal’s birthdate), the journal’s origin story does not even mention the JRD episode, instead offering a rather convenient narrative: disillusioned with the post–World War I settlements and the US response, a group of intellectual trailblazers conjured Foreign Affairs almost out of thin air, with the primary purpose of informing the American public about foreign affairs.⁶⁶ Sympathy for Wilsonian internationalism can perhaps explain the new, race-blind name for the journal, but it also points towards an irony. As Erez Manela has shown, Wilsonian internationalism came with a robust geographical bias in the context of the principle of self-determination: almost without exception, societies in non-European geographies were categorically and systematically denied the right to determine their own destinies.⁶⁷

This is hardly surprising. The late 19th and early 20th centuries were in fact the apex of what can be referred to as (explicit) dual territoriality.⁶⁸ Western countries professed imperial territorial ideas in their colonies but were also increasingly privileging and reifying the nation-state form within their borders.⁶⁹ In this sense, at least in the context of the aforementioned period, it is utterly misleading to discuss whether the ‘international order’ resembled anarchy or hierarchy; the international order was not really international (it was largely a Western-dominated order), and from a territorial perspective, it was both anarchical and hierarchical, with a spatial twist.

In the Global North, the leading states of the time increasingly emphasised the ideal of clearly demarcated and compartmentalised pieces of sovereign territorial units that act like ‘culture and social containers’,⁷⁰ a (territorial) hallmark of the nation-state ideal, for their own core territories. The same powers’ territorial ideas as well as practices in (or for) the Global South were a far cry from what they upheld at home; Western powers recognised no Indigenous boundaries and ruled their colonial possessions through overlapping, contingent, and ever-shifting jurisdictions. Within the Westphalian club, the rules pertaining to territorial sovereignty were not always upheld⁷¹ but often respected. Outside the geographical scope of the club of the sovereigns, those rules did not apply, and whenever and wherever rules existed, they were mostly about managing colonial powers’ relations among themselves in the ‘open’, lesser territories in the Global South. As Adelman and Aron highlighted in the context of the Rush-Bagot convention of 1817 in North America, the (Western) sovereign states ‘fixed the lines separating political communities’ in colonised geographies, but ‘no one consulted the [Indigenous populations]’.⁷²

As will be discussed in detail in the next section, TTT is more of a product of the nationalistic ontology. However, the hegemonic powers in the Global North (and, by extension, IR theories) interpreted this ontology primarily through the prism of geographical racism and territorial hierarchies.

TTT and the nationalistic ontology

Political geographers have long recognised that TTT entails a nation-state-centric reading of world politics.⁷³ As an ontological benchmark,⁷⁴ nationalism provides purpose and meaning for one’s life, lays out a convenient shared past, offers the promise of a common future, and does so by collapsing the collective of individuals under a singular and exceptionally robust identity. At its core, the nationalistic ontology is as much about ‘territory’ as it is about the ‘people’: almost all nationalistic ideologies involve a close association between society (the putative nation), the state (politics), and space (the homeland). In fact, what separates nationalism from its ideational rivals or predecessors is how explicitly territorial it is.⁷⁵

The nationalistic ontology is both divisive and remarkably singular (or ‘modular’).⁷⁶ It is divisive in the sense that, say, French nationalism is different from American nationalism, which is different from Indian nationalism, and so on. Furthermore, particular nationalistic discourses are bound to be in disagreement from time to time, especially over the topic of territory, usually over the question of to whom a certain territory belongs. Alternatively, some nationalistic discourses deny the status of nationhood to minorities. However, even under such circumstances, the relevant minority groups rarely challenge the nationalistic ontology per se: on the contrary, they make the case that they constitute a nation too, with a legitimate claim over what they perceive to be their national homeland.⁷⁷

In sum, while it is true that there are many different nationalisms, almost all nationalistic discourses point towards a singular nationalistic ontology where ‘nations’ are not only real and pre-existing socio-spatial entities, but where they are (or should be) also the primary actors in world politics, each with its own homeland. The homeland, in this setting, is depicted as a discrete, bounded territorial unit with a unique and unifying socio-spatial identity. The linchpin that connects the nation and the homeland is the state or the desire to establish a state. In some nationalist narratives, the states are almost perennially attached to their respective nations and homelands (e.g. France); in other cases, nations ‘awaken’ to establish their state after hundreds of years of dormancy, in their homeland (e.g. Germany). Borders may change, ‘new’ members may eventually join the club of ‘nations’ (or the United Nations [UN]), but all these changes take place within and through the nationalistic ontology.

As far as territory is concerned, IR scholarship, to a large extent, also operates through this ontology. One crucial implication of this tendency is usually referred to as methodological nationalism.⁷⁸ The concept refers to the (usually implicit) ‘assumption that the nation/state/society is the natural social and political form of the modern world’.⁷⁹ In other words, ‘methodological nationalism is found when the nation-state is treated as the natural and necessary representation of modern society’.⁸⁰ Overall, methodological nationalism can lead to ‘explanatory reductionism’:⁸¹ the practice entails taking ‘nationally bounded societies’ as ‘naturally given entities to study’⁸² and does so without ‘[problematising] national discourses, agendas, loyalties and histories’.⁸³ Wimmer and Schiller also point towards the territorial consequences of methodological nationalism, which paves the way for a rather limited and (potentially) misleading ‘territorialization of social science imaginary and the reduction of the analytical focus to the boundaries of the nation-state’.⁸⁴

To be precise, IR scholars’ inclination to succumb to methodological nationalism is not specific to IR; sociologists and historians, among others, can and do fall into the same analytical trap.⁸⁵ What renders IR rather unique is a paradox: as far as territory and territoriality are concerned, IR scholarship (implicitly or explicitly) usually leans towards methodological nationalism, but many IR scholars still imagine the territorial underpinnings of the modern state system to follow from the Peace of Westphalia, which legal scholar Leo Gross famously (and inaccurately)⁸⁶ christened ‘the majestic portal which leads from the old world into the new world’ in 1948.⁸⁷

Recall that, in its simplest definition, a territory emerges when (1) physical space is demarcated for social and political purposes; and (2) demarcated space is organised in order to manage and regulate the association between space and society.⁸⁸ Both the Westphalian ideal and the nationalistic ontology emphasise the importance of clearly demarcated, rigid borders (as opposed to, say, fluid frontiers).⁸⁹ In other words, both the Westphalian state and the nation-state, as aspirational models, are built on similar demarcation principles. However, these two territorial models differ from each other in terms of socio-spatial organisation. The nationalistic ontology explicitly stipulates that the space–society association should be defined in terms of socio-spatial homogeneity.⁹⁰ This homogeneity is always aspirational and can be based on language, ethnicity, culture, or civic values. Almost all nationalistic discourses emphasise the primacy of socio-spatial singularity within a discrete, inviolable homeland.

In the original Westphalian arrangement, neither the rulers nor the ruled demanded or sought a comparable sense of socio-spatial homogeneity. The territorial politics of the 17th and 18th centuries clearly supports this interpretation: state territories were deemed divisible, simply because neither the ruler nor the ruled considered their attachment to ‘state space’⁹¹ in terms of national, indivisible homelands.⁹² As sociologist Anthony Smith pointed out, ‘pre-nationalist outlooks tend[ed] to treat territory simply as “land” to be worked and settled’.⁹³ Similarly, Clausewitz wrote that the typical state of the time ‘behaved as though it owned and managed a great estate that it constantly sought to enlarge – an effort in which the inhabitants were not expected to show any particular interest’.⁹⁴

In Europe, the rise of nationalism eventually transformed what territory meant for leaders and societies in fundamental ways.⁹⁵ At least in theory, state spaces evolved into national spaces. This gradual transformation also had significant implications for the ways in which borders were construed. Most notably, the main distinctive characteristic of the nationalistic understanding of borders is exactly what borders (should) delineate: while borders in pre-nationalistic Westphalian states focused on compartmentalising the extent of sovereign jurisdiction, for a nation-state, borders also act as ‘cultural containers’.⁹⁶ Put differently, national borders, in theory, are supposed to distinguish and isolate one cultural group (or even the culture itself) from other cultural groups. In order to accomplish this goal, nation-states tend to sacralise the state space through symbols and (often invented) traditions.⁹⁷

To cut a long story short, IR’s geographical assumptions and the very territorial underpinnings of the present-day state system are a function more of the nationalistic ontology, rather than the original Westphalian arrangement. The dominance of the nationalistic ontology, in turn, is rather new and ‘modern’ by world-historical standards.⁹⁸ Obviously, nationalism as an ‘ism’ had not permeated an overwhelming majority of the global (or even European) landscape until some two centuries ago.⁹⁹ However, especially during the course of the 20th century, nation-state form was rendered into an ideological, legal, and normative hegemon, due to both international and domestic political dynamics. Internationally, the emergence of the League of Nations and then the United Nations has promoted and privileged the nation-state form. In domestic politics, territorial nationalism (which is sometimes branded as ‘patriotism’ due to the connotations associated with the term nationalism especially in the Western world) remains an exceptionally robust reference point, that is, in terms of the dominant territorial ideas in the present day.¹⁰⁰

According to the associated geopolitical imagery, international politics emerges from the interaction between sovereign (national) territorial units, which are usually taken as pre-existing entities. From a world-historical point of view, countless nation-states, especially nation-states that emerged out of colonial arrangements, can hardly be taken as pre-existing territorial units or polities. Regardless, during the course of the first half of the 20th century, the Western powers interpreted the nationalist ontology and its territorial implications through the prism of geographical racism and imperialism. The notion of ‘nationhood’ was reified by the Western powers, but it was also used as a discursive weapon to deny ‘equal’ status to countless geographies and societies, with lasting consequences.

This would hardly come as a surprise to students of nationalism. Nationalism is – by definition – particularistic and ‘exclusionary’;¹⁰¹ some rights, especially ‘territorial rights’,¹⁰² are reserved for some social groups while being categorically denied to others.¹⁰³ The spread of nationalism and the rising salience of the nation-state form during the course of the 20th century also followed this exclusionary logic with an openly Eurocentric perspective and did so on a global scale. In many former colonial geographies, it wasn’t the ‘nations’ that somehow ‘woke up’ and established their own nation-states and homelands. In many cases, sovereignty was ‘granted’ to existing colonial administrative-territorial units, not to the ‘peoples’ per se. In many parts of the non-Western and/or colonised geographies, it was the decisions of Western powers that determined which territorial units would be considered as real nation-states.

The relevant processes usually followed three stages.¹⁰⁴ In the first, especially during the course of the 19th century, the Western powers quite literally sliced and diced territories across the globe for their colonial ambitions. The second stage entailed dual territoriality on the part of the Western powers: these powers promoted the nation-state ideal in their core territories while also professing imperial (territorial) idea(l)s and policies in their colonies and dependencies. The apex of this episode involved the interwar period. The immediate aftermath of World War I represented ‘a world of sharp contradictions’, where ‘empires both disintegrated and expanded’.¹⁰⁵ In contrast with the idealistic pretensions at the discursive level, victorious European great powers aimed to establish an ‘imperial peace’.¹⁰⁶

Most notably, post–World War I arrangements simultaneously exalted the so-called self-determination principle and categorically denied it to non-European societies and geographies.¹⁰⁷ In this sense, the colonised geographies were still being portrayed as territorially distinct from the colonising geographies.¹⁰⁸ The relevant arguments were based on a certain assumption or assertion: societies in the ‘backward territories’¹⁰⁹ had not yet matured into full-blown nationhood; therefore, they needed the guidance and assistance of the so-called advanced nations. This assumption was also engrained in the Article 22 of the League of Nations Covenant.¹¹⁰ In other words, while ‘nationhood’ was conceived as the master criterion for becoming a sovereign member of the so-called society (or league) of states, Western states exploited the same criterion to exclude countless societies and geographies from the same society, or (eventually) to decide which geographies and societies would deserve the status of nationhood (not to mention, when exactly they would qualify). In this context, the mandate system that emerged after World War I (or its extension, the institution of trusteeship) is best understood as a slight modification of existing colonial arrangements.¹¹¹

The third stage culminated in decolonisation, a process whereby colonial and quasi-colonial geographies were finally admitted to the modern state system. Most notably, the architects of the post–World War II settlements aimed to ‘freeze’ borders by establishing the so-called border fixity norm, while also emphasising the inviolability of the self-determination principle.¹¹² In the words of political geographer Liam O’Dowd, the settlements of 1945 ‘[replaced] the imperial ideal with the ideological hegemony of the nation-state ideal on a global basis’.¹¹³ Similar to the interwar period, the answer to the question of which geographies would qualify as proper nation-states had a lot to do with the existing territorial hierarchies.

Perhaps the most robust reflection of this trend was the principle of uti possidetis.¹¹⁴ The principle was inspired by a maxim of Roman law, uti possidetis ita possidetis (as you possess, so you possess). According to legal scholar Joshua Castellino, uti possidetis is ‘perhaps the single greatest influence on the shaping of the map of the world today’ and ‘the most important factor in the creation and maintenance of modern postcolonial identities’.¹¹⁵ Historically speaking, the application of the doctrine to interstate borders can be traced to the early 19th century, or the decolonisation process in Latin America. The principle ‘treats the acquisition and possession of a state’s territory as given, with no territorial adjustments allowable without the consent of the currently occupying parties’.¹¹⁶ The doctrine of uti possidetis, more specifically, does not differentiate between de facto and de jure possession and favours ‘actual possession irrespective of how it was achieved’.¹¹⁷ The doctrine was eventually canonised in international law in 1960, with UN Assembly Resolution 1514, which also finally formalised colonial geographies’ right to self-governance.

As applied to colonial geographies, the doctrine stipulated that ‘boundaries left behind by colonial rulers, whether sanctified by treaties or not, remain untouched by the succession of the state by independent non-colonial rulers’.¹¹⁸ In this sense, uti possidetis ‘upgraded’ colonial administrative boundaries to the status of interstate borders,¹¹⁹ inadvertently emboldening the ‘dogma’ of the intangibility of the colonial boundaries.¹²⁰ In practice, uti possidetis not only upgraded the legal status of the existing colonial borders, but also ‘froze’ them – presumably, for posterity – by way of foreclosing debates (and claims) over their location.¹²¹

On the one hand, the application of the uti possidetis principle to international boundaries could be seen as an attempt on the part of the UN to ‘maximize the viability of the new states, rather than ethnic or tribal ties’¹²² and prevent, or at least limit, the extent and scope of territorial conflict in the post-colonial environment.¹²³ On the other hand, the uti possidetis doctrine inverted the territorial logic of the original self-determination principle. In theory, the principle of self-determination, as it was conceived (or imagined) after World War I, had a simple, if impractical, logic: the global (or more specifically, European) landscape was imagined as a territorial ‘puzzle’ waiting to be solved, and the solution would entail ‘figuring out’ the correct socio-spatial distribution of the ‘peoples’ and matching them with appropriate borders.¹²⁴ Put differently, this vision aimed to take the identities of numerous ‘peoples’ (that is, nations) as givens and define borders to establish socio-spatial homogeneity (or national uniformity) in terms of such identities by drawing/redrawing borders. By contrast, the doctrine of uti possidetis took the existing (colonial) borders as givens and defined the socio-spatial identities in terms of the same borders. Simply put, the doctrine neglected ‘all the other criteria of identity’ in favour of pre-existing colonial boundaries,¹²⁵ further ossifying what legal scholar Gerry Simpson referred to as ‘legalized hegemony’,¹²⁶ while also perpetuating the ‘spatial history of colonialism’.¹²⁷ In that, the application of the uti possidetis principle to colonial geographies can be seen as one of the most robust and ‘visible’ implications of TTT: by buying into IR’s geographical assumptions without questioning them, we, as students of global politics and history, run the risk of deluding ourselves about the true nature of global politics, while also erasing the very complex and hegemonic origins of the modern state system.

Conclusion

Paradoxically, what paved the way for TTT is also what TTT erases and whitewashes: IR’s racist-colonial origins and the selective nationalistic ontology that has long sustained the illusion of ‘sovereign equality’ among not only states, but also ‘peoples’ and territories. In many ways, TTT functions as a carnival mirror, with a twist. Carnival mirrors distort the image of an object, so that it appears unpleasant (which we may find funny or scary). TTT distorts the exceptionally unpleasant and overly complicated origins of the modern state system (and IR), projecting a palatable and convenient ‘image’ instead.

In the past decade, in more ways than we can count, ‘mainstream IR’ has been challenged from within numerous corners of the discipline, not to mention the globe. In this context, this study joins countless others in trying to unpack core aspects of the dominant approaches in IR. That being said, the intent here is not to make a case for simultaneously reifying ‘mainstream IR’ as a monolithic enterprise while also ‘cancelling’ it. On the contrary, the aim is to take advantage of the salience of the rising intellectual diversity and associated critical perspectives to advance our collective understanding of the international. In other words, while this study challenges some tendencies of the so-called mainstream IR, it does not seek to reduce it to a singular, static, and irredeemable enterprise to be undermined. Instead, it offers a constructive criticism (pun intended) with the purpose of ‘recovering IR’, as Daniel Levine put it.¹²⁸

For instance, it is possible to deduce two conclusions from the preceding argument. First, TTT is perhaps more of a problem for ‘American IR’ (which some may use as a synonym for mainstream IR) than for others. Second, IR scholars’ reluctance to engage TTT (or, similar critical perspectives) can potentially be explained by a tendency to support the existing hegemonic tropes. However, placing an overwhelming emphasis on such claims would not help us, as IR scholars, to recover IR. For example, associating TTT exclusively with American IR detracts from two facts: some segments of non-American IR (however defined) also tended to ignore TTT,¹²⁹ and a good portion of IR scholars who are participating in the ‘constructivist turn’ in the study of territory hail from (North) American institutions. As the past decade has shown, American IR, however defined, is also amenable to critical (and self-critical) perspectives. In other words, if our intention is to move forward as a discipline (or move beyond TTT), it is more constructive as a practice to treat American (or, mainstream) IR as a moving target that is open to criticism and change, rather than essentialising and vilifying it en masse.

Similarly, the argument that it was geographical racism and a penchant to serve the hegemonic powers that paved the way for the racist underpinnings of IR theory in its early decades does not necessarily mean that present-day ‘mainstream IR’ scholarship as a whole shares similar inclinations and motives. As is the case for almost all epistemological communities, in IR, some practices may turn out to be ‘sticky’ and be passed on across generations even when the original motives no longer apply. Especially in the context of the recent ‘self-reflective moment in IR’, where the ‘mainstream’ is becoming more critical of itself, a far more productive path forward is to try and figure out how IR scholars, mainstream or otherwise, can move forward together.

In this context, the arguments provided above have two broad implications for future research. The first involves the concept of geographical racism. The concept of racial hierarchy has been gutted out of IR,¹³⁰ but, as John Hobson put it, it was then replaced by an ‘equally distorted conception of “cultural hierarchy”’.¹³¹ This ‘subliminal Eurocentrism’, according to Hobson, merely recycles some of the old themes of the colonial-racist past but does so without ever invoking direct references to race. In this new-but-old narrative, ‘decolonization becomes reimagined as a result of the “triumph of the moral ideas of the West”, specifically the principles of national self-determination and social justice’.¹³² This new master narrative is part of a ‘happy story’, ‘in which the West diffused its “rational” civilizational institutions and practices outwards so that the East too could come to enjoy the benefits of residing within civilized international society’.¹³³ The arguments offered above suggests that TTT is closely associated with subliminal Eurocentrism.

In such a setting, the notion of geographical racism deserves further attention, for instance, in the context of the recent debate involving the Sovereign Territorial Order and the Liberal International Order.¹³⁴ The debate is most certainly a step forward in terms of improving our understanding of ‘international orders’. However, unless students of global politics integrate (or at least account for) the impacts of geographical racism and the associated territorial hierarchies in(to) their analyses, the aforementioned debate runs the risk of either being essentially incomplete (at best) or recycling subliminal Eurocentrism in different forms and formats (at worst).

Lastly, this study makes the case for further interdisciplinary dialogues between political geographers and IR scholars.¹³⁵ One potentially fruitful area for interdisciplinary research involves nationalism. In a global landscape where nationalisms are being simultaneously challenged (for example, by immigration patterns and refugee inflow) and reinforced (say, due to the rise of anti-immigrant/refugee sentiments), further inquiries into the roles that territoriality and the notion of nationalism play in world politics is not merely a matter of academic curiosity; it is an essential task. Both IR scholars and political geographers, with their own disciplinary ‘comparative advantages’,¹³⁶ can benefit from speaking to one another about the relationship between the national and the international in the context of territory and territoriality.

Another potentially fruitful area of research is to compare and contrast Western and non-Western conceptions of (geopolitical) ‘space’.¹³⁷ Such a research agenda would be beneficial for IR research in two ways. First, as Erik Ringmar highlighted, a fuller understanding of the Western spatial and political arrangements is possible not by studying the Western ad nauseam, but by scrutinising Western and non-Western experiences in tandem.¹³⁸ Second, as Amitav Acharya forcefully argued, a truly global IR requires a further and deeper engagement with non-Western ideas, traditions, and institutions.¹³⁹ In this context, further inquiries into the non-Western spatial practices and imagineries, which at minimum requires moving beyond TTT, can help IR scholarship to move towards a more global perspective. In that, the relevant research in political geography offers numerous insights.¹⁴⁰

Some 30 years ago, John Agnew made one of the most significant interdisciplinary contributions in the context of territory and territoriality. Especially in the last decade, IR scholarship has developed a rather cumulative interest in the socially constructed nature of territories, which has created potential interdisciplinary ‘intellectual beachheads’ that can help IR scholars and political geographers to engage each other’s ideas and approaches in far more constructive ways. In this sense, revisiting TTT and subsequent research in the light of contemporary scholarship in both political geography and IR from a truly interdisciplinary perspective can contribute to our understanding of the central role that territories and territoriality play in world politics, past and present.