How Bosnia and Herzegovina Was Bordered: The Supervised Making of a Border/Mobility Assemblage in the European Semiperiphery

Border Security > State-Building

In 2010, as part of a collaborative study of borders of Bosnia and Herzegovina with Čarna Brković and Vanja Čelebičić, I conducted interviews with a range of professionals involved in the bordering of BiH. Here I turn to two of those people, both working in the capital Sarajevo.^[2] Mr Mehić was a geodesist, a member of successive commissions to identify and mark the country’s outer border through negotiation with colleagues from neighbouring states. Mr Popović was a middle-ranking officer in the BiH Border Police. I will now identify two unsurprising but important patterns that marked both conversations.

First, both interviewees framed their professional engagement with the border in terms of non-political, technical expertise. For Mr Mehić, this primarily concerned identification of the precise topographical coordinates of the border. Mr Popović focused on undocumented border crossings, which he presented entirely as a security problem. Both men left unspoken that neither of those two issues was considered a priority by most people in BiH at the time. Topography was important to small local constituencies only (e.g. Jansen and Brković 2019) and, in 2010, undocumented crossings were hardly ever a topic of conversation. All three of us knew that these issues stood for another question that was of great public concern: BiH state sovereignty. Right since the dying days of Yugoslavia this had been contested from within and from outside the country. After the US-brokered and -enforced 1995 Dayton Peace Agreement, which ended a three-and-a-half-year war, reaffirmed BiH sovereignty, at least in theory, many disputes shifted to the forms this sovereignty should take. Mr Mehić and Mr Popović might well disagree on this question. This too was left unspoken, but it could be read between the lines and detected from word choice. For example, in references to contestations of the topography of the border, Mr Mehić posited the legitimacy of the “BiH” side, using the register of sovereignty that prevailed amongst those who favour the strengthening of BiH statehood on the grounds of the 1992 declaration of independence. Mr Popović, on the other hand, had until the late 1990s worked in the service of a competing sovereignty project: in the police force of Republika Srpska (RS), one of BiH’s war-produced “Entities”. Now he commuted weekly from a town in RS to the BiH Border Police headquarters in Sarajevo. Mr Popović did not make any mention of legitimacy.

What is much more important here than any such potential disagreements between these two officials is an obvious difference in the prioritisation of the actual work of institutionalisation in BiH. In 2010 (as today), the sovereign capacity of the country’s “central” organs remained minimal,^[3] but border security received massive investment. Consider the contrast between the surroundings where we met. Mr Mehić’s office was in the late nineteenth century building that initially served as the headquarters of the Austro-Hungarian occupation of BiH. Today it is the seat of what is formally the highest political office in the country, the “central” BiH Presidency. Reflecting its de facto status as a mostly dysfunctional, toothless body, Mr Mehić’s room was in a state of disrepair, with peeling paint and threadbare carpets. Spare furniture and paperwork were stacked in a corner. Mr Popović, on the other hand, received me in the shiny headquarters of the Border Police, built in 2007. Costing a reported 3.65 million euros, of which 2.8 million were provided by the EU, the architecture, the furniture, the equipment, the uniformed personnel, the logos—everything exuded slick professionalism. If nothing else, then, funding priorities were obvious: first border security, then sovereignty/state-building.

This leads us to the second pattern in those interviews. Both men framed their explanations directly in the terms of conditionality for EU accession, which featured as a largely unquestioned end goal. Exasperated with politicians’ regular announcements that border agreements with neighbouring states were about to be signed, Mr Mehić lamented that the border had not even been marked anywhere. This required pillars and pyramids for the three-way border points, he explained, “as is the standard between other states”. But then he added: “… perhaps less than that, if we move towards EU accession”. Later, when Mr Popović spoke about security, with examples of drugs, terrorism and, most of all, undocumented migration, I recalled Mr Mehić’s concerns and asked him if the state border was marked. “No”, he said, “not at all”. “In principle”, he said, “this could be a problem, but that’s not a security issue … And, anyway, I somehow hope we will join the EU before that and then marking would become much less relevant.” Later in the conversation, Mr Popović picked up on this theme to steer our interaction in a peculiar direction. Addressing me as a citizen of a Western European state, he said:

We’re really doing everything that can be expected from us, and perhaps more. We accepted, without any conditions, a readmission agreement, we take back anyone who illegally entered the EU through BiH; we established an asylum centre and a reception centre for people illegally residing in BiH and awaiting deportation. […] We’re really doing everything so as not to hinder or slow down that process in any way, and we have often been singled out for praise. In terms of the control of migration, the situation has really drastically improved. There are EU reports to confirm this.

Mr Popović closed our conversation by stating politely, almost pleadingly:

I would like, hm, diplomatically, to express a feeling that many people have, that the EU is not fair towards us. In border management, we in BiH have accepted almost anything that the EU has demanded from us. Much more than other states! The best example is that we have implemented—okay, this is still ongoing but it’s ninety per cent done—an Integrated Border Management system. Few states even in the EU have that! And Serbia, which is not even in the phase of design in that regard, was given White Schengen …^[4]

Indeed, visa requirements for citizens of Serbia to visit the Schengen Zone had been lifted in late 2009. This was referred to as “getting White Schengen” and in BiH, at the time of our interview in 2010, it remained a key topic of everyday speculation (Jansen 2009).

Let us link this back to the contrasting priorities detected above. As Mr Popović stated, EU reports confirmed the achievements of the Border Police. Six years earlier a security scholar had already judged that BiH had “the most sophisticated system of border management in Southeast Europe” (Hills 2004, 42). She attributed this overwhelmingly to effective foreign involvement. Tellingly, more than a decade before the movement of people from Asia and Africa through BiH made any serious media headlines, Hills’ short list of “problems” faced by BiH started with it being “a transit route for illegal migration and trafficking” (2004, 41; see also 2006). Of the two “fundamental themes shaping the development” of the Border Police, security and state-building, she continued, “the first theme, security, is crucial” (2004, 45). This was borne out, some years later, in a broader analysis of the EU’s role as the “primary state-builder” in BiH, in which a political scientist argued that EU “conditionality” had been “largely ineffective with regard to state building” (Bieber 2011, 1783). Operating from competing angles, a vast literature on the role of the so-called “international community” in BiH tends to agree on this latter point. And while they may use different vocabularies, and cherish conflicting aspirations regarding which state should be built, most inhabitants of BiH also feel, even today, that theirs is not a “normal state” (Jansen 2015). To add insult to injury, they are reminded of its inadequacy in the continuous evaluation that marks the country’s projected “Road into Europe”. So how did BiH, for once, come out top of the class and why precisely in the domain of securitised border management?

Upgrading Borders

Latour states that Actor Network Theory—the “empirical sister-in-arms of the more philosophical assemblage thinking” (Müller 2015, 29)—works best when focusing on “what is not—not yet—a sort of social realm” rather than on “what has already been assembled” (Latour 2005, 12). The bordering of the newly proclaimed state of BiH would then seem to be an ideal topic for an assemblage approach to work its magic. At the time of the interviews in 2010, it is not just that the precise topography and policing of this border was in the process of being assembled, but that neither the border nor the state itself could be considered any stable “sort of social realm”. Heterogeneity and provisionality abounded here and an assemblage approach allows us to account for this.

At pains to impress the achievements of the Border Police on me, Mr Popović emphasised throughout our interview that, apart from Sarajevo Airport, BiH had not been able to rely on previous experience with the management of inter-state borders. Prior to 1991, it had been surrounded by other republics of the Yugoslav federation. Such inter-republican borders were relevant territorial delineations for administration but they were unmarked and unguarded. In 1991, the EU’s Badinter Commission had indicated that new states could be recognised within those existing intra-Yugoslav borders. After a majority vote in a 1992 referendum an independent Republic of BiH was proclaimed and, as Mr Mehić emphasised, sought to upgrade its 1500 km long outer border right from the day of international recognition.^[5] Yet the outbreak of war rendered this impossible. Almost all road and rail crossings into neighbouring states were controlled by Serbian nationalist or Croatian nationalist armed formations, both with heavy involvement from across that border: the governments of Serbia and Croatia were key to their functioning in terms of arms, personnel, logistics and diplomacy. To spell it out: police units of polities that challenged the legitimacy of BiH statehood were presumably regulating movement to and from what they considered ethnic “mother-states” with which they wished to have no border at all. Even after some tightening under pressure of Western governments, checks remained discretionary. In practice, in conditions of war, it was not the internationally recognised outer border that demarcated any sovereignty of the state subject called Republic of BiH, but the trenches at the far end of the much smaller territories controlled by its army. Clearly, any account of the production of this highly provisional border/mobility assemblage must account for the heterogeneity of its production, involving operations by competing (para)military formations, the laying of landmines, ceasefire negotiations, but also displacements, smuggling, (non)interventions by UN soldiers, and so on (Jansen 2013).

The 1995 Dayton Agreement confirmed the geopolitical recognition of BiH sovereignty and its territorial and institutional division into Entities and Cantons, under a wide-ranging mandate for in-country foreign supervision.^[6] In practice, internal borders remained much more effective than the external one. For over five years after the war, almost all crossing points into neighbouring states were still manned by Entity and Cantonal police units controlled by Serbian and Croatian nationalist forces, with separate command structures, legislations, logistics, insignia, flags and, of course, political agendas. As we saw, Mr Popović and Mr Mehić had occupied different positions in this 1995–2000 constellation. The former had worked in the police force of RS, which continued to reinforce separation. The latter had been engaged in a section of the “central” state organs of BiH that sought further integration. At the time undocumented border-crossings by non-Europeans did not feature much in local disputes. Instead, the dominant register was that of sovereignty.^[7] Yet by the late 1990s there was considerable foreign pressure to establish unified security organs in BiH. Initially, this pressure came particularly from US officials. They too evoked BiH sovereignty, but were most focused on trans-border crime and transit migration, especially from certain Asian states such as Iran, for which they identified Sarajevo Airport as a hub (Andreas 2008, 127). The integration of police structures was rejected by domestic politicians, especially from RS, and the precise shape of unified security was contested by all. In a particularly interventionist phase around 2000, the High Representative in BiH, formally the top in-country official of foreign supervision, overruled part of this and imposed the establishment of a unified State Border Service under the aegis of the BiH Presidency.^[8] By the time of our meeting in 2010, Mr Popović displayed much pride in his institution, but he made no bones about the fact that it had come about through foreign imposition. In his words: “Some high-up politicians went to some meeting or something and they were told they had to sort this issue out and create a unified border service.”^[9]

In order for this new institution to establish an effective border/mobility assemblage in the name of BiH statehood, a vast array of measures was introduced from 2000 onwards, involving legislation, infrastructure, equipment, organisational structures, recruitment and training. Most applicants to the State Border Service, like Mr Popović, were already working in one of the existing Cantonal or Entity police forces. They were drawn by significantly higher salaries and opportunities for career advancement. The in-country UN International Police Task Force, which was replaced in 2003 by the EU Police Mission, organised “transition training”, including human rights education. It also vetted applicants, particularly concerning war-time activities and postwar occupation of other people’s accommodation (property restitution was another priority of BiH’s supervisors at that time) (Collantes Celador 2005; Padurariu 2014). The first operational unit was deployed in Sarajevo Airport in 2000, and over the following two years the State Border Service gradually took over all 89 official border crossing points, of which 55 were for international traffic. They now flew the BiH state flag,^[10] the design of which was also imposed by the High Representative. Uniforms bore the new logo that included this flag, a sword, the country’s contours and the service’s name in Latinic and Cyrillic script.

Supervision and the Road into Europe

In 2007 the State Border Service was renamed as “Border Police”. Its precise status changed over time, but, in a situation where few public institutions functioned effectively across the internal divisions of BiH, it retained its reputation as a truly “central” BiH service. In official communication this was often referred to as “autonomy”, evoking a contrast with the presumably far greater capture of other organs by competing domestic political forces. Yet even to the degree that such “autonomy” really existed on the domestic scale, it obviously came with (and was produced through) alternative strings. Namely, once they had forced the establishment of this new agency through the political institutions of BiH, functionaries of the USA, of the EU and of individual member states closely steered its work. Their governments also provided the bulk of the funding. Millions of euros came in through the EU’s Community Assistance for Reconstruction, Development and Stabilisation (CARDS) programme. CARDS, which aimed to implement the Stabilisation and Association Process, commenced in 2005 and was envisaged as a step on the Road into Europe. Security concerns were always high on the CARDS agenda, and “border management” was a key priority, with investment in the set-up of the Border Police, its headquarters, its training and equipment, and later its collaboration with the Customs Service as part of the Indirect Taxation Authority (which was also imposed by the High Representative). Particularly from 2005 onwards, the EU paid the lion’s share of the brand new, high-tech border crossing points that were gradually replacing offices improvised out of shipping containers.^[11] This included funding for vehicles, parking space and access roads, and, in some cases, renovation of bridges. Less visibly, the EU funded IT equipment, software, and ongoing specialist training on border security, often in seminars with EU and US border agencies. Coordination and collaboration (such as joint patrols) with colleagues from neighbouring states, and integration with customs and sanitary inspection agencies were also on the agenda. Much assistance was organised in packages geared at “Integrated Border Management” (IBM), referred to by Mr Popović above and proclaimed “a priority in the European Partnership” by the EU Delegation to BiH.^[12]

Public legitimisations of this ongoing consolidation of the comparatively well-funded Border Police, and of the border/mobility assemblage it was at the heart of, were invariably coated in the language of BiH sovereignty and its projected EU accession. Opening ceremonies of new border posts featured foreign functionaries, usually from Western Europe, alongside those from BiH and neighbouring states. Sometimes tensions arose as a result of the key role of the EU in funding them and because of the frequent rhetorical mobilisation of the “central” Border Police’s achievements to showcase the BiH state’s increasing sovereignty over Entities and Cantons. For example, in 2005 RS government representatives refused to attend the opening of a new border crossing post with Serbia because they rejected the “central” protocol of the ceremony (i.e. they wanted a more prominent place for RS officials and symbols). The chief of the (also “central”) Customs Service remarked: “It’s a little awkward to me too, but it’s known who donated the resources for this border crossing point and who participated in the creation of the protocol, and that is the [Office of the] European Commission in BiH.”^[13]

In speeches by local and foreign functionaries, prickly questions of sovereignty were therefore mostly secondary to references to borders as bridges that do not separate but bring people together and to evocations of “European standards”. But most prominently, they framed every new border crossing point as a significant step on BiH’s Road into Europe. This was part of a much broader discursive register of conditionality that revolved around progress and reward. The work of bilateral border commissions in which Mr Mehić participated to solve disputes on border topography was also monitored as a measure of such progress,^[14] as were numerous other “reforms”. As we saw, a prominent manifestation of this mechanism of conditionality concerned the promise of “White Schengen”. In 2008 the European Commission had presented BiH and neighbouring states with a “Road Map” for visa liberalisation, which stipulated measures organised in four blocks (document security; illegal migration, including readmission agreements; public order and security; and external relations and fundamental rights). Implementation was monitored by policy analysts in BiH and beyond, spawning regular progress reports and analysis (usually with EU or US funding).^[15] In all this, while EU rhetoric did favour increased BiH sovereignty (or at least functionality of its “central” organs), mobility-restricting border security featured as a touchstone for potential rewards on the Road into Europe across the entire first decade of the new millennium. Clearly, it was key to the establishment of the State Border Service in 2000. As active opposition to this imposed institution decreased, massive investment in training, equipment and border infrastructure continued, culminating in the implementation of the 2008 Road Map and the award of “White Schengen” in 2010. So this is how BiH, for once, came out top of the class: when border security was assessed.

It is important to realise that in that decade EU demands regarding border security predominantly concerned the cross-border movement of citizens of Balkan states themselves. Controlling that unruly mobility—through measures including increased surveillance, biometric passports and readmission agreements—was rewarded with visa-free travel to the Schengen zone (Hameršak et al. 2020, 18). In early 2016, the German government’s “Western Balkan Regulation” came into force, making it possible for citizens of BiH, Albania, Kosovo, Macedonia, Montenegro and Serbia to be legally employed in Germany. This resulted in a vast increase of documented emigration from BiH. Domestically, the notion of irregular migration thus came to be associated increasingly with non-white, non-European persons rather than with locals or citizens of neighbouring states. The topic came to prominence with the mass movement along the formalised corridor through Serbia and Croatia in 2015–2016 but by early 2017 even the IOM still presented transit migration in BiH as a small-scale problem and foresaw no redirection of the Balkan Route through its territory (Stojić Mitrović et al. 2020, 74). Nevertheless, it is clear that securitised concerns with such mobility by non-Europeans were always a factor in the supervised bordering of BiH. As we saw, US government demands regarding transit migration by Iranians are said to have prompted the imposition of the new unified State Border Service in 2000. And it seems fair to say that the very establishment of that institution was already an integral part of the EU policy of “externalisation” of mobility regulation (Stojić Mitrović et al. 2020). As early as 2002, a top British UN-policeman in BiH put it plainly to a UK newspaper: “A lot of [BiH Border Police] officers think that if people are transiting into Western Europe that’s not a problem for Bosnia. We have to educate them that it’s creating lots of problems and that Bosnia wants to be part of Europe.”^[16]

So, for their citizens to enjoy future freedom of movement in a presumably borderless union, BiH authorities should prevent undesired movements towards the northwest of their own people and of others. The tensions inherent in this conditionality are perhaps visible especially sharply in BiH due to the fact that the formation of an effective border/mobility assemblage in the name of this state was initiated as late as 2000. Both domestic authorities and EU supervisors considered it to be provisional right from its inception. The prime minister of the Federation of BiH (one of the country’s Entities) used the occasion of two opening ceremony speeches in 2005 to politely and, hm, diplomatically, express his “regret that significant resources were invested in the building [of the border crossing points], while they will have to be destroyed in five-six years. However, building them is unavoidable if we want to enter Europe.”^[17] Later, with Croatia’s EU accession (which occurred in 2013) approaching, anxieties intensified, and politicians in BiH and in Croatia reassured people that no new “Berlin Brussels Wall” would be built because BiH was “on the same path [to the EU]”.^[18]

Today this may sound wildly optimistic, but it does reflect how the BiH border/mobility assemblage was initiated and provisionally stabilised as part and parcel of EU enlargement policies. This process, starting in earnest in 2000, was marked by recursive disciplinary dynamics. At the border posts the BiH Border Police combined surveillance and pedagogy: it monitored crossings and taught (its own and other) people how to orderly engage in (or refrain from) movement across borders into another space. Sanctions were at hand. On another scale EU agencies monitored the work of the Border Police and, with sanctions at hand, taught BiH how to engage in orderly temporal movement towards another timespace, that of EU membership. As Mr Mehić, Mr Popović and the prime minister all pointed out, since neighbouring states—Croatia, Montenegro and Serbia—were (and are) continually “ahead” on this Road into Europe, this produced a paradox: the degree of success of the entirely new BiH border/mobility assemblage could be measured in terms of the speed by which it reduced its own relevance. It is this configuration that a multitude of actors encountered as the Balkan Route swerved through BiH in 2017.

Border/Mobility Assemblages and Hierarchy in the European Semiperiphery

Let us now turn to the uses of assemblage theory for providing this historical baseline. Much of my reconstruction of the making of the BiH border/mobility assemblage until 2017 has revolved around what are effectively practices of “implementation”. This term has a bad name in assemblage theory (Beznec and Kurnik 2020, 39; Clarke et al. 2015, 59) because it is perceived to overemphasise hierarchical, top-down determination at the expense of heterogenous and provisional becoming. I hope it is clear that my approach is not the result of a reluctance to refrain from unpacking “black boxes” or to register alternatives; and that it constitutes neither approval of the actions I document, nor legitimation of the dominance of certain actors. But it is true that my account does emphasise hierarchical dynamics throughout. Therefore it seems important to specify in what ways exactly hierarchy is central to my analysis of the bordering of BiH and to do this in light of the key contribution of assemblage theory: its attentiveness to heterogeneity and provisionality.

First, heterogeneity. A reconstruction of the formation of the BiH border/mobility assemblage from 1992 to 2017 must acknowledge how it was assembled, for example, through war-time military operations by competing armies; through interventions of domestic and foreign political functionaries; through state-centred loyalties, fears and hopes of people in BiH and in neighbouring states; through EU enlargement policies and geopolitical interests; through labour, security and wealth disparities between the European centre, its semiperiphery and other continents; through electoral concerns in Western Europe regarding immigration from the Balkans and from beyond; through the travels, subjectivities and aspirations of people on the move, from the Balkans or from further East and South; through attempts to regulate mobility by various police forces in BiH, by the US government and EU agencies; and so on.

But let us be frank: the attentiveness to heterogeneity that assemblage theory encourages would be pointless if it merely led us to a list like the one in the paragraph above. Such lists are never complete. Nor, I think, are they particularly interesting. To go beyond mere listing, our analysis must “cut the network” somewhere and develop a rationale for one particular focus rather than another. As we saw, some recent studies of the assembling of the Balkan Route have chosen to emphasise the role of people on the move themselves, and sometimes also of solidarity activists. Their rationale is political: they focus on these subjectivities to counteract previous neglect and to foreground possible alternatives. I find this legitimate and important. In that sense another valuable contribution consists of recent anthropological accounts of the hitherto neglected role of non-activist (and not necessarily supportive) residents of the border region in BiH (Helms 2022; Hromadžić 2020) and in Croatia (Leutloff-Grandits 2022). Of course, like these other studies, my reconstruction of the pre-2017 making of the BiH border/mobility assemblage has been selective too. So, in all this heterogeneous becoming, what were my grounds for emphasising certain constituents over others?

I too sought to offer more comprehensive understanding by foregrounding less documented dimensions (here: the history of the BiH border/mobility assemblage prior to 2017). Yet I focused on the role of actors that cannot be said to be subaltern or marginalised. I believe it is important to acknowledge that not all constituents of this assemblage were equal and I wanted to account for the relative weight of the concerns and actions of different actors. Of all the constituents of the heterogeneous becoming of this assemblage, I foregrounded the concerns and interventions of foreign, and particularly Western, governments and their representatives because their capacity to steer the actions of others was particularly notable. This was so, for example, in the EU’s 1991 Badinter Commission’s position on the legitimacy of upgrading inter-republican Yugoslav borders, which preceded BiH’s 1992 declaration of independence; in the 1995 Dayton Peace Agreement, which ratified the division of a nominally sovereign and supervised BiH state, whose constitution was its annex; in the 2000 imposition of a unified State Border Service by the High Representative; and, during the following years, in the prioritisation of border security in EU conditionality. Assemblage theory, then, usefully draws our attention to the relevance of a multitude of actors, but it does not a priori favour any particular ones. It is to heterogeneity that it makes us attentive, not to any specific subaltern constituents of it. And when it comes to the making of the border/mobility assemblage of BiH prior to 2017, I believe that any adequate account must, as mine has done, primarily frame it in the hierarchical workings of EU conditionality regarding mobility regulation.

But of course, this does not mean that that hierarchy is necessary, or fixed forever. And that brings us to the issue of provisionality. Assemblage theory rightly urges us to acknowledge that institutional arrangements that may appear to us as stable orders are actually only provisionally stabilised assemblages. Their stability is never complete nor unassailable. So, if at any given moment a hierarchy is at work, assemblage theory encourages us to trace the contingent historical processes of its formation and functioning. Here it is important to acknowledge that the violent disintegration of Yugoslavia and the 1995 Dayton establishment of BiH as a divided semiprotectorate occurred in the early post-Cold War period, when the authority of Western players on the European continent met with little competition. As I am writing this, recent geopolitical developments, as exemplified in the war in Ukraine, serve as a reminder for many in BiH that such things may and do change. Yet for the first few decades after BiH’s declaration of independence, and until further notice, the default position for almost all actors involved in this country’s bordering was to conceive of themselves as entering an already-assembled hierarchy. In that hierarchy, they knew, they occupied a lowly position, and Western institutions held both carrots and sticks in a peculiar relation where this hierarchy was itself explicitly presented as provisional, with a normative way forward framed as the gradual fulfilment of conditions on the Road into Europe.

Now, of course, hierarchy does not imply consensus on either end of it. There are disputes about migration and enlargement policies within EU organs, and actors in BiH disagree about the bordering of BiH, as about much else. Top-down demands do not always go unchallenged either. We saw that, in BiH, EU conditionality has been ineffective in many realms. Yet when it came to border security, EU agencies used a broad arsenal of particularly forceful intervention and conditionality to incorporate certain BiH institutions as implementers of its policies.^[19] BiH even became a model pupil for a while. Sure, much foot-dragging characterised the processes I have described and, clearly, the unruly cross-border movement of BiH citizens and others was never fully suppressed. Contestation, deviation and insurgent agency existed all along, and assemblage theory makes us particularly attentive to all that. But after the direct imposition of the unified State Border Service and ancillary institutions, such friction rarely made a dent in the making of the BiH border/mobility assemblage along EU priorities, and I therefore did not turn to them as central explanatory and interpretative tropes. Domestic actors in that assemblage encountered the machinery of EU conditionality and it is within those already-established hierarchical relations that they could use EU resources, EU-compliant legal and institutional reforms and/or the very image of their own (respectful, cunning, defiant) subordination to “Brussels” in order to pursue their interests and shore up their domestic positions.

So when in 2017 tens of thousands of people from Asia and Africa rerouted their attempts to reach Western Europe through BiH, the ensuing reconfiguration of the BiH border/mobility assemblage appeared as another phase in a process that had operated according to an already-established hierarchy for decades. Amongst the heterogeneous multitude of people on the move, solidarity activists, non-activist local residents, BiH domestic authorities and many others, EU institutions again operated as master-constituents. It was EU states that these people wished to reach; it was EU priorities that informed the closing of the 2015–2016 formalised corridor that circumvented BiH; it was EU police forces, including members of its Frontex agency, that multiplied violent “pushbacks” from Croatia and other member-states. Clearly, this is why large numbers of people from Asia and Africa got stuck in that particular place (northeast BiH) at that particular time (in 2017). Domestic BiH authorities were unable and/or unwilling to exercise sovereign powers to prevent their entry into the EU or to provide them with humanitarian relief on their territory. Divided and often largely dysfunctional anyway, most saw little reason and received little encouragement to forge a sense of responsibility for issues that appeared to them as concerning other people either way: poor Asians and Africans and rich Western Europeans. So Western intervention intensified again, but this time not so much through the procedures of in-country supervision specified in Dayton. Instead, as in many other states, the EU contracted the IOM as the key actor for the externalisation of its mobility regulation. In BiH, this intergovernmental organisation had previously been active predominantly in the forced repatriation of hundreds of thousands of refugees from Germany in the late 1990s and in the roll-out of the IBM system in the late 2000s, mentioned by Mr Popović. Unaccountable to any democratic public, now it became the dispenser of multi-million budgets for the domestic implementation of EU priorities. It enhanced the capacity of the Border Police and ancillary services, which had already been enrolled for EU border security purposes since their establishment anyway, and also of the police force of the northwestern Unsko-Sanski Canton. As others have documented, due to the labyrinthine set-up of Dayton BiH, at times this situation fed into a further weakening of BiH statehood due to periodical and racially-profiled upgrading of internal borders (Ahmetašević and Mlinarević 2019). So, again, the hierarchical functioning of the BiH border/mobility assemblage crystallised the tensions between the prioritisation of border security and the rhetoric of BiH sovereignty. Over time, and as in earlier phases, EU conditionality—with its all-important funding streams—re-entrenched the hierarchy as most domestic authorities sought to display, enthusiastically or grudgingly, cooperation in a chain of command in which they could not meaningfully negotiate “either the terms of their inclusion or the migration policy they [are] to implement” (Stojić Mitrović et al. 2020, 100; see also Hameršak et al. 2020, 19).

Balkanism and the Supervised State

While hierarchical relations may be particularly visible in the supervised semiprotectorate of BiH, such deeply unequal processes of assembling in the realm of borders and mobility are neither rare, nor new. Jane Cowan’s excellent analysis of the “supervised state” demonstrates the historical persistence of such deeply “asymmetrical supranational governance” (2007, 547). Focusing on the question of “minority policies” in Southeastern Europe, she reveals sharp parallels between the post-1990 period and the post-World-War-I period. In the Committee on New States during the Versailles negotiations, she explains,

rather than making minority protection a universal treaty obligation for all states implicated in the Versailles arrangements, including Britain with its “Irish problem”, France with its Bretons and Basques and Italy with its multinational Tyrol, the Principal Allied and Associated Powers imposed such treaties only upon particular states, largely in east and southeast Europe, […] as the price of their international recognition (Cowan 2007, 549).

Representatives of “neutral”, established states monitored, reported, advised, and, ultimately, decided on the sovereignty claims of such “treaty-bound states”—here mainly through supervision of the field of minority protection policies.

Developing a broader critique of understandings of states as discrete entities, Cowan then urges us to acknowledge a long-standing, global “spectrum of sovereignties”, with colonial rule at one end and fully independent European statehood at the other, and with many other forms in between. Turning to 1990s BiH, she clearly shares much of assemblage theory’s concerns: for example, she rejects the register of “empire” as too blunt (553) and she acknowledges that the terms of EU conditionality are contested and provisional at any given moment in time (545–55). However, commenting on EU integration, she also writes: “We would be wise to keep an eye on the reproduction of hierarchical relations between states within this process, and especially in current processes of accession” (571–2, italics in original). Cowan identifies BiH and Kosovo as the sharpest examples of “supervision” in Europe, but warns that “virtually all states now installed in former socialist territories are currently subjected to scrutiny by supranational organisations and often, too, by NGOs closely tied to them” (553). In the decade and a half since the publication of her article, we have witnessed the direct involvement of EU officials in the making of austerity policies in southern Europe. This is but one example that suggests we may now want to leave out the “former socialist” clause in Cowan’s statement. Perhaps it is time to consider most states as to some degree “supervised”.

If that is so, we need modes of attentiveness that allow us to understand such hierarchical relations. Assemblage theory helpfully guides us away from any assumptions of mechanistic top-down command and from the invocation and legitimation of “black boxes”. Yet turning to this approach raises some specific concerns when addressing the specific spatiotemporal configuration of the European semiperiphery (see e.g. Blagojević 2009; see Jansen et al. 2016; Tošić 2017). Namely, assemblage theory makes us attentive to heterogeneity and provisionality. In other words, to difference, friction, incompleteness, fragmentation, deviation, unpredictability, disjointedness, divergent meanings, uncertainty, plurality, fluidity, instability, unintended consequences, failure of top-down legal-rational intentional causality, complexity, and so on. Notably, most of these terms are also central to the register of Balkanism: a discourse that constructs the southeast of Europe as insufficiently European (Todorova 1997). In Balkanist representations it is precisely heterogeneity and provisionality—and excessive investment in it by “locals”—that are considered to be key problematic characteristics of “The Balkans”: never stable, always in chaotic flux; never homogeneous, always violently fragmented by diversity.

The attunement to heterogeneity and provisionality that assemblage theory entails, then, must acknowledge its own positioning with regard to already-assembled hierarchy. In this article, I was wary of presenting an account that did more to rhyme with Balkanist representations of the European semiperipheral condition than it would do to add to our understanding of it. This is why I awarded priority to tracing how the remarkable persistence and effectiveness of the hierarchy that constitutes that condition (as well as that of the “centre”) is encountered as already-assembled by different actors.