Liberal Violence and the Racial Borders of the European Union

This paper examines how racial violence underpins the European Union’s border regime. Drawing on two case studies, in northern France and the Balkans, we explore how border violence manifests in divergent ways: from the direct physical violence which is routine in Croatia, to more subtle forms of violence evident in the governance of migrants and refugees living informally in Calais, closer to Europe’s geopolitical centre. The use of violence against people on the move sits uncomfortably with the liberal, post-racial self-image of the European Union. Drawing upon the work of postcolonial scholars and theories of violence, we argue that the various violent technologies used by EU states against migrants embodies the inherent logics of liberal governance, whilst also reproducing liberalism’s tendency to overlook its racial limitations. By interrogating how and why border violence manifests we draw critical attention to the racialised ideologies within which it is predicated. This paper characterises the EU border regime as a form of “liberal violence” that seeks to elide both its violent nature and its racial underpinnings.


Introduction and Methods
I have a question for you: If the European Union stands for liberty and equality and all that, how can it be taken from us? They say they do so much humanitarian work, but they don't want even 2000 people from this camp? And they beat us, aggressively. (Interview with Afghan victim of Croatian police violence, July 2019, Biha c, Bosnia-Herzegovina) This paper brings together evidence from two research projects on border violence in Europe-in northern France and the Croatia-Bosnia and Herzegovina (henceforth Bosnia) border. Each of these case studies demonstrates that within the European Union, violence is routinely used to control immigration. Between these case studies, we see how European border violence is obscured by the concealment and displacement of violence to spatial "peripheries" where they are less likely to be detected. These empirical findings of border violence stand in stark contrast with Europe and the European Union's liberal, post-racial self-image, which positions human dignity, human rights and the rule of law as fundamental tenets of "European values" (European Union 2019). Rather than accepting this hegemonic understanding of liberal bordering practices, we argue that the technologies of concealment and displacement evidenced within these case studies amount to a form of "liberal violence", marked by the obscuring of violent governance and the racial logics underpinning it.
The first of these case studies presents fieldwork conducted in France, ongoing between 2015 and 2019, with the town of Calais providing a focal point. Northern France constitutes the north-western edge of the EU's Schengen border zone, and the port town of Calais is a transit-point out of the Schengen area to the UK. Calais has seen significant numbers of migrants arrive and reside informally over the last 20 years, en-route to the UK via the English Channel (see Rygiel 2011). Research here began in 2015 when thousands of migrants who had been living in informal settlements in the port-town were corralled into a space on the edge of the town's limits, a site which became the infamous "Calais jungle" (Mould 2017; Van Isacker 2019). Our research in Calais has involved Environmental Health Surveys (see Dhesi et al. 2015Dhesi et al. , 2018 and ethnographic research alongside volunteers and camp inhabitants, and has been used to explore the subtle and everyday acts of state-enforced structural violence, through inaction and abandonment (see Davies et al. 2017).
The second case study presents research from Bosnia, ongoing between 2016 and 2019. The wider region known to the EU as "the Western Balkans" is one setting to which EU border security has been "outsourced" (Trakilovi c 2020) and is thus a productive site through which to examine the contestation of Europe's exclusionary practices. Since the re-imposition of Dublin 3 border controls in 2016, migrants and refugees have become stranded on the borders between Croatia inside the EU, and Serbia and Bosnia on its periphery (Umek et al. 2019). Those migrants and refugees who have attempted to make journeys into Croatia and Slovenia (i.e. into the EU), have routinely faced physical abuse from border guards within EU territory. This violence, documented by activists (Border Violence Monitoring Network 2020), humanitarian organisations (Amnesty International 2019; MSF 2017), and our own long-term research, is more direct and crude than that which affects migrants and refugees in Calais. It is characterised by theft and physical assaults: fists and batons against flesh and bone. These abuses are catalogued by border violence monitors with whom we are collaborating on a scholar-activist research project. Our data here consist of the long-term observations of scholar-activist Karol ına Augustov a who has worked with grassroots volunteers in collating violence reports over an 11-month period between January 2018 and December 2018, and a month of qualitative fieldwork in Bosnia by the other three authors, specifically focusing on assembling testimony about experiences of violent and illegal pushbacks from the EU. The data are supplemented with observations and interviews from fieldwork in Serbia in 2019.
The contrast between these two modalities of violence at different external borders of the European Union ( Figure 1) have prompted us to consider more closely the function of border control within European liberal democracies. The European border's "integrity" is enforced through various forms of violence, both direct and indirect, that can lead to wounding and death to those who try to enter irregularly (Jeandesboz 2014). As critical migration scholars have shown, often the more indirect and geographically "distant" forms of violence are the most lethal: the removal of rescue ships in the Mediterranean for example, leaving thousands to drown as they make perilous journeys to Europe (Stierl 2018). Whilst there exists much insightful scholarship on the technologies of border governance (see Jones 2016), which have answered vital questions about how irregular bodies are immobilised and rendered precarious, there remains a critical need to also interrogate its foundational principles: its rationales, its logics. In other words, questions of why these violent technologies of bordering are allowed and justified.
It is in this space that our paper makes its contribution. This paper looks at the two aforementioned case studies of border violence used to securitise EU borders and suggests that although such acts of violence differ in manifestation, they share a dialectical unity. Each form of violence works to obscure its brutality, often through concealment and displacement-and thus mask the racial logics which underpin the EU's border work. This obscuring of violence helps sustain notions of European liberalism by camouflaging liberalism's contradictions with respect to race. The paper begins by reviewing literature on violent migration governance in a liberal age, making the distinction between violent governing technologies and its rationales, using Galtung's (1990) notion of "cultural violence". Drawing on the work of Goldberg (2002) in particular, it then examines the racial logics underlying the modern European nation-state, by tracing the long-standing contradictions between Europe's self-avowed universal liberalism, and its spatial and ideological limits. This postcolonial perspective is crucial to understanding how the boundaries of Europe and European liberalism have been ideologically constructed (El Tayeb 2011;Losurdo 2011). We then examine the two case studies described above, detailing the types of violent technologies used, but also excavating the ways in which racialisation can be detected in that violence in subtle and more overt ways. We conclude by framing contemporary Europe's border governance as a form of "liberal violence" that works to obscure its racial underpinnings.

Violence in Liberal Times
Whilst the notion of a "European" continent predates liberalism, contemporary Europe and Western European values are largely accepted to be couched in liberal ideology. Individual freedoms and rights to speech, expression, political organisation and movement, together with the architecture of democratic political states are all considered hallmarks of liberal European society. Furthermore, liberal rights are considered inherent and universal, rather than predicated upon inherited traits, for instance of race or ethnicity (Parekh 1992). The nation-state itself with its structured bureaucracy is designed to balance popular democratic will with individual rights, and becomes the primary vehicle through which the liberal rule of law is sustained. On the eve of the Iraq war, Habermas and Derrida (2003:295) summarised in their overly optimistic call for European unity and peace: "Europeans have rather a lot of faith in the organizational and governing capacities of the state ... they possess a keen sense of the dialectic of enlightenment ... Their level of tolerance towards violence against persons is comparatively low". A sentimental equivalent of this notion is also evoked in the self-proclaimed values of the EU's border and coastguard agency Frontex, which is charged with the "protection" of liberal European states: "We are professional. We are respectful. We seek cooperation. We are accountable. We care" (Frontex 2019).
As this paper will show, that notion of "caring" is very far from the reality of policing the European border, which is unquestionably violent. But the words of Habermas and Derrida (2003) on European society's aversion to violence echoes the orthodox tale of modern European states' relationship to violent governance. This tale sees European states as having overturned the autocratic sovereign power of former European kingdoms and empires. It sees Europe-and its settler colonial offshoots-as having introduced universal rights and eventually, democratic accountability. Societal violence is then reduced, or at the very least made "just" (Neu 2018), by being delivered through legal infrastructure that is premised on those liberal, democratic rights. This is one reason why violence is often under-theorised, by accepting the notion that lingering acts of violence in liberal settings are "mere collateral casualties produced along a steady path to enlightened modernity" (Hutta 2019:65).
On the contrary, as geographers and allied scholars have demonstrated, violence is still very much part of liberal society (Neu 2018), and not merely an outcome of unequal social relations (Galtung 1969), but also as a reflection of society itself: its priorities, its anxieties, and its vulnerabilities (Davies 2019). In other words, violence holds a "mimetic" relationship to the fundamental value system of society (Springer and Le Billon 2016:1).
The work of Galtung (1969Galtung ( , 1990 is key to understanding the mechanisms through which violence operates in liberal settings, highlighting how violence is either hidden or legitimised. If liberal societies would like to see themselves as peaceful and fair, then the presence of violence must be disguised, displaced or rendered just. Galtung (1969Galtung ( , 1990 offers important insights into the ways in which violence operates, from its socialisation and structuration, down to the actual acts of violence themselves. For Galtung (1969) the very point at which violence is inflicted on a body is not the most sociologically significant moment. Instead, he positions structural and cultural violence as precursors and preconditions to acts of direct violence, arguing that structural and cultural forces both legitimise and embed violence into the routine patterns of everyday life (Galtung 1990). Both structural and cultural violence are marked by a reduced visibility next to the very act of violence itself (Rodruiguez et al. 2014). For Galtung (1990:294), structural violence constitutes the processes through which violence is sustained, and this very much corresponds to the orders through which violence may be concealed, displaced or denied.
Those documenting the structural violence of bordering practices have evidenced some of the ways in which it can be shrouded, and these can vary in their levels of subtlety (Davies et al. 2017;Jones 2016). The consequences of banning rescue missions in the Mediterranean (Garelli et al. 2018), or the illegalisation of providing water for migrants in the Sonora desert both have deadly implications. Similarly, the attempts to illegalise food distribution to asylum seekers which may keep them in states of permanent hunger (Tazzioli 2019; Tyerman 2019); or "domicidal" practices such as the systematic soaking of belongings and the routine destruction of improvised migrant shelters in Calais and elsewhere (Hagan 2019;Van Isacker 2019). Other techniques of bordering take place away from the physical borderzone in institutional settings, where for example, healthcare for refugees can be denied or limited, which can allow preventable illnesses to debilitate those who might suffer from them (Dhesi et al. 2018;Ilcan et al. 2018). Within these examples, states can use a legal or extra-legal architecture of public policy to enact harm through the violence of inaction and denial (Davies et al. 2017).
Migrant deaths can thus occur out of sight, displaced and concealed along desolate southern borderlands of Arizona or New Mexico, thousands of miles from Liberal Violence and the Racial Borders of the EU policymakers in Washington DC, or beneath the waves of the Mediterranean Sea. The suffering deliberately enacted through the subtle technologies above, also allows such violence to appear self-afflicting, where the "violent conditions" that refugees are forced to endure are framed as of their own making (Laurie and Shaw 2018). Just as governmentality subverts power to look and feel like the outcome of individual free choice, so too does the politics of inaction and the withholding of basic human rights re-centre blame to that of the migrant "Other".
Unlike the direct and structural forms of violence discussed above, cultural violence takes us beyond technologies and processes of violence towards its underlying rationales. Though post-structuralist writers may contend that processes of violence and their logics are overlapping, for Galtung (1990:294) cultural violence is the permanent substrata beneath which the process of structural violence, and the event of direct violence takes place: the patterns of prejudice, repression and exploitation that construct the very subjects to whom violence may then be apportioned. Galtung (1990:296) specifically refers to ideologies such as nationalism or religion as the basis upon which popular and political consent are often drawn to normalise repressive violence. Here, by centring race in our analysis of border violence, we begin to isolate the rationales of governance that can often be drowned out by the mundane rhetoric of everyday politics, or overshadowed by the blunt immediacy of violent acts themselves.

Europe and Race: Liberal Contradictions
Let us then return to the aforementioned liberal self-image of Europe and its attendant values of universal liberty: values that provided an intellectual basis for the democratic revolutions and evolutions of political orders in Europe between the 17 th and 20 th centuries. Notions of "freedom" and "justice" capture the turn away from autocratic sovereign rule towards a social contract which amplifies the rights of domestic citizenry, or the supposed domination of Nature by Reason. However, the cracks and contradictions of this self-image are apparent from the very inception of European liberal thought. Morefield (2014:4) contends that it is imperialism that "presents perhaps the ultimate challenge to the founding narratives of liberal equality, individual freedom and sovereign authority". This is because the foundational tale of European liberalism readily excludes narratives of women, non-freemen, and crucially, black and brown people who had been colonised by ostensibly "liberal" states (Mehta 1999). This section explores how echoes of this liberal contradiction, borne in the context of colonialism, is sustained in the contemporary racial border logics of the European Union.
Goldberg (2002) throws the dominant narrative of liberalism into doubt with his exegesis of liberal philosophy, by re-evaluating its tenets through the lens of race and colonialism. Goldberg (2002) dissects essays by John Locke, Thomas Carlyle and even John Stuart Mill, whose liberalism scarcely extended to bestowing equal rights to colonised subjects. At its most egregious, the works of Carlyle manifests an explicit "negrophobia" (Goldberg 2002:58-60) justified through biological determinism. Even at its least inflammatory, the liberal racism of John Stuart Mill and others demarcate African, Eastern and Asian entities as insufficiently developed, socially or politically to be considered deserving of the liberal self-rule that Europe was inherently thought to merit. This logic, as Parekh (1994) astutely observed, allowed for distant colonies to be violently governed as though existing in supposed Hobbesian "states of nature"-with little regard for the freedoms under which Europeans had discovered themselves. This contradiction-a wilful myopia-is also extended to the very construction of that other invention of the enlightenment: the nation-state. The nation-state, bound as it is by geography, serves to protect citizens from the supposed alternative "state of nature". As Goldberg (2002:240) observes, it does so by mobilising the spatially bound configurations of race, in order to comfort and control those within, and compel those without. A violent "racial rule", as Goldberg (2002) puts it, becomes a natural extension of this rationality. We therefore come to expect liberal values and rights to be lent and upheld in Europe-but no further.
This colonial "original sin" of European liberal states is significant because it builds the racial configurations that become the raison d'être of violent techniques of border violence in contemporary European and settler colonial regimes. Aim e C esaire (2001), among many others, referred to the centuries of European rule of much of the colonised world as civilisational. As Cedric Robinson (1983: xxxi) concludes on European domination in the colonial age, "Race was its epistemology, its ordering principles, its organising structure, its moral authority, its economy of justice, commerce and power".
Western colonial powers routinely used the notion of civilising processes as justification for imperial conquest (Fabiani 2011)-a racist ideology that was also replicated within the discipline of geography. Colonialism did not merely reshape the colonies but constructed the notions of "Western Civilisation", as opposed to non-Western barbarism. It is for this reason that Frantz Fanon (1961:53) famously quipped, "Europe is literally the creation of the Third World". The very definition of Western civilisation and the liberalism it purports to exemplify is defined in relation to this uncivilised Other (Said 1978). And yet, as C esaire (2001:32) emphasises, and as this paper will reiterate, it is the very dehumanisation of racialised subjects in this way that allows for them to be violently, and illiberally treated.
The 20 th century post-war order might ostensibly be seen to provide a postcolonial break from what Goldberg (2002) described as "racial rule", yet we see a continuation of these debates today through European migration policy (El Tayeb 2008;Turner 2015). Even when the 1951 Refugee Convention was signed by liberal nation-states in the wake of European civilian displacement after the second world war, Britain and other colonial powers harnessed the technologies of bureaucracy to restrict the right to asylum to non-Europeans (Mayblin 2017). Asylum, in other words, was never designed for colonial subjects.
In contemporary settings we can see the distinctive echoes of imperialism in the violations of rights of migrants and refugees (Walia 2013). Bhambra (2017) notes that contrary to prevailing narratives of Europe as a refugee and asylum destination, the reality is that Europe, as the wealthiest continent in the world, hosts fewer refugees per population than any other (Bhambra 2017:396). Developing countries-not their former colonial masters-host 80% of the world's refugees (Bhambra 2017:396). In this century, EU countries including the United Kingdom, France, Germany, Denmark, the Netherlands, Austria, Italy and Hungry have all seen a great deal of anxiety being expended over the place of immigrants in those countries by mainstream political parties (Betz 2016). Moreover, the European Union has presided over a migration regime which has eased restrictions to movement internally, only to strengthen the barriers to entry for those from outside the EU. As El Tayeb (2011:2-3) explains, the construction of "fortress Europe" in this way is routinely bolstered by discourses that see Europe as a beleaguered "vanguard" in an otherwise dangerous world. Similarly, on the eve of European political expansion after the fall of the Berlin wall, Stuart Hall (1991:18) explicitly identified how the emerging discourse around liberal European cosmopolitanism could easily pivot to countervailing discursive markers around refugees, fundamentalism and "illegal migration". It is this very logic that M'charek et al. (2014) identify when detailing the ways in which the governance of European borders hinges upon the technology of race and the processes of racialisation.
As scholars routinely remind us, race is not to be essentialised, but considered a shifting category (Gilroy 1987;Sivanandan 1990), driven by social, political and cultural discourse and practice. We will see in the second case study in this paper, how refugees racialised through their perceived Muslim-ness, have their Muslim identity used as justification for violence. The unsettling "otherness" of migrants is then bolstered through discursive frames that draw migrants as, for example, barbarians at the gates of civilisation (Todorov 2010), as too illiberal to find a place in liberal society, as dirty (Zimring 2017), as disgusting, and even animalistic (Vaughn-Williams 2015). Each of these tropes forms the cultural violence (Galtung 1990), which then allows harm to be perpetrated to the very bodies of migrants and refugees who are racialised in this way. The racialisation of non-European Others provides the ideological and epistemic base on which the structures of anti-migrant oppression are built. We need to explicitly contextualise the violence evident at the world's borders-and in both case studies presented here-as racial violence.
There are risks of not confronting the racial underpinnings of border violence. De-racialised readings of migration governance upholds the near-mythical version of European and settler colonial history which simply draws violence as aberrations or policy-failure in otherwise egalitarian and progressive societies (Mondon and Winter 2020). Such readings re-inscribe what Lentin (2008) describes as the myth of a "post-racial Europe", alongside its intellectual handmaiden which simply takes anti-migrant violence to be a technocratic result of the state's governing practices with respect to citizenship, rather than structural exclusions with respect to race; a trope so often repeated within "migration studies". As this literature review has argued, this very elision reproduces the race-blind logics of liberal violence. Building on this literature review, in what remains of the paper we will explore two interconnected case studies that demonstrate racial violence at the European border. Drawing on long-term research in Calais and the Balkans, we will show how "liberal violence" is a dominant form of EU border governance, which operates through racialised logics and is sustained through the concealment, displacement and denial of racial violence.

Violence against Refugees and Migrants in Calais
Gentlemen, we must speak more loudly and more honestly! We must say openly that indeed the higher races have a right over the lower races. I repeat that superior races have a right, because they have a duty. They have the duty to civilise inferior races. (Jules Ferry, 28 March 1884) Jules Ferry was a liberal statesman and two-time Prime Minister of France. As his speech above suggests, he was also "one of the main theoreticians of the colonising process" (Fabiani 2011:8) and an ardent white supremacist during the Third Republic (1870Republic ( -1940. It is not without irony that some 130 years later, descendants of the colonial subjects that Jules Ferry had lobbied to subjugate, would be stood in line, waiting in the rain to charge their phones at a refugee shelter named "Le Centre Jules Ferry", situated on the edge of the largest displacement camp in mainland Europe, on the outskirts of Calais. Let us explore how this came to be the case.
For over 20 years, Calais in northern France has been a transit site for refugees travelling to the UK to seek asylum. Throughout this period, the port town became a reluctant host to thousands of displaced people, who lived-and continue to live-informally in numerous squats and disused sites on the town's periphery. In April 2015, the shifting population of refugees were corralled onto one peripheral site: a former toxic dumping area alongside Le Centre Jules Ferry, two miles from the streets of Calais, out of sight for the town's residents and tourists. At its peak, the site hosted up to 8000 residents, and was the only location in Calais where displaced people were allowed to sleep and reside. Not since the wars in former Yugoslavia in the 1990s had the continent witnessed the political use of encampment on such a scale.
We posit that the designation of the Calais camp in 2015 was an act of both concealing and displacing the racial violence of refugee governance. Furthermore, we suggest this violence was made possible by the racialisation of the camp's inhabitants. Rather than working with the UK to allow asylum claims to be handled either side of the English Channel, or to allow for processes of family reunification, refugees had been abandoned by state agencies. Their right to family lives and resettlement would not be innate, but dependent on making dangerous border crossings. However, their very presence in the centre of Calais was a problem for authorities; their visibility in civic spaces was for them, egregious (Migration Observatory 2014). Accordingly, in an act of concealed violence, the camp's residents were deliberately fed only one meal per day from Le Centre Jules Ferry, which only covered less than half of the camp's population (Davies et al. 2017). Many residents reported hunger; they also had no safe storage for food, which led to gastrointestinal illnesses (Dhesi et al. 2015). Shelters consisted of donated tents, or rudimentary tarpaulin over improvised wooden structures. During our own fieldwork in Calais in April 2015, we noted the feeling of sudden rupture and shock when walking from the suburban streets of Calais into the space of the camp where the suffering of refugees was laid bare (Davies and Isakjee 2019). In the early days of this new camp, merely a road and a hedgerow concealed from view the living conditions of migrants, who would otherwise be living directly alongside French citizens and residents of Calais.
Liberal Violence and the Racial Borders of the EU After French authorities had forcefully displaced refugees to the site of the new camp in 2015, the so called "jungle" became a highly concentrated site of refugees living informally-a space where "global racialized inequalities [were] suddenly writ large on the European landscape" (Davies and Isakjee 2019:215). The displacement of refugees to this site however ultimately failed to hide the "problem" of migrants, the injustice of their habitation, or the inability of state violence to act as a deterrent. From the smell of burning rubbish due to the lack of waste collection, to the acrid taste of chemicals in the air from the adjacent industrial park (Figure 2), the camp became a state-induced public health disaster. Indeed, an environmental health survey of the camp conducted in 2015 provided a systematic account of the public health conditions in the camp, concluding that the French government failed to meet basic humanitarian standards set out by the UNHCR (see Dhesi et al. 2015Dhesi et al. , 2018. Simply put, abandoning the refugee population to live in the so-called "jungle" without sufficient state provision was leading to extensive bodily harms. Together with a dearth of shower facilities, and with no safe way to wash clothes or bedding, an estimated fifth of the camp's population contracted scabies. What's more, there was an acute shortage of toilets, which left refugees with no option but to defecate on open ground. Of course, the French government and local prefecture did not ascribe a racial motive to these acts of violence. Nor too, would the status of "violence" be ascribed to the systematic denial of sanitation, water, and food to subaltern people, the vast majority of whom originated from former European colonies, or "the colonial outside" (Goldberg 2006:332). Yet the spectre of race was apparent as a logic under which these policies operated-from the camp's creation to its eventual demolition in October 2016. As Galtung (1990:292) explained, cultural violence may sometimes operate "by making reality opaque, so that we do not see the violent act or fact". One such concealment is the way the creation of the camp involved the discriminatory forcible eviction of refugees from more visible parts of Calais. Concurrent with all forms of racialised policing, racial profiling was key to this enforcement: during our fieldwork in 2016 we observed how refugees, racialised and identified as unwelcome, would routinely be rounded up by patrolling police, and transferred from central Calais back to the squalor of the campback to the degrading conditions to which they would be deemed to belong. As a result of this eviction and encampment of the so-called "jungle", the racialised divisions between migrant and citizen entered the realm of spectacle. The camp became "a concentrated visible symbol of the 'apartheid' of migrant Others from the Global South" (Davies et al. 2017(Davies et al. :1268. Before long, the stark divide between the condition of a majority-white Calaisian population, and the encamped residents became perceptible, uncomfortable and even horrifying. Within a week of the camp being established, we came across local farmers who were offering photojournalists flights over the campscape for a small fee. Journalists began to frequent the camp to write about the squalor and monetise the unfolding scandal. For journalists the Calais camp was a "third-world slum" (Sinibaldi 2015), a "shanty town" (Lichfield 2015) jutting out of an otherwise orderly, "civilised" (Spence 2016) European scene. For geographers, it was "a slum of London's making" (Mould 2017), and the camp became politically intolerable because it made the violent consequences of anti-migrant racism spatially concentrated, visible, and unmistakably there.
For this reason too, as the racial violence became apparent, solidarity organisations and charities began to protest, resist and deliver humanitarian supplies (Mould 2017). One sign next to an improvised shelter on the edge of the camp read, "2 B Black is not a crime" (Figure 3), and refugee residents themselves protested against the racism which made their violent treatment "acceptable". The concentration of racial othering turned the camp into an avatar of global inequality (Davies and Isakjee 2019). Scarcely since the second world war had such a distinct postcolonial spectre of racial difference been etched so unmistakably onto a Western European landscape. Like the racialised trope of "gypsy camps" in Europe, the so-called "jungle" produced a "perfect juxtaposition of a racially connoted marginalized population with a secluded urban location, eventually crystallizing racist perceptions" (Picker et al. 2015:742). As the camp became an icon for racial exclusion, the uneasy covenant of liberal violence-which relies upon the ability to conceal and displace violence-had failed: the camp had to go. The bulldozers rolled over the smouldering camp in October 2016 and its inhabitants were bussed to distant locations across France, allowing once more for racial border violence to be displaced, concealed, and denied.
Today, the violent governance of the Calais border continues to be obscured, and the racial logics that presuppose it remain camouflaged. As we write this paper, around 1500 people still reside informally in Calais and Dunkirk in a shifting constellation of smaller camps. Le Centre Jules Ferry has been demolished, and the so-called "jungle" is long-gone, but the politics of concealment and displacement has arguably become more sophisticated. The violence against refugees today often takes the form of "domicide" (Mould 2017; Van Isacker 2019), where police routinely demolish shelters, and soak sleeping bags in flooded ditches; state-sanctioned cruelty rendered innocuous by its repetition (Obradovi c-Wochnik 2018) (Figure 4). It also takes the form of attempts by local authorities to outlaw the distribution of food in public (Hagan 2019). In these acts-systemic, subtle, and often out of sight -we see border violence continue, underpinned by the ever-present cultural violence of racism. Even within the politics of destitution and exhaustion, a racialised hierarchy persists; during our fieldwork in 2019 we noticed how small groups of homeless white people were permitted to sleep rough in the centre of Calais, while black and brown homeless migrants would regularly be apprehended and evicted from these same sites. Though the overt racial violence promoted by the likes of Jules Ferry and his liberal peers may be gone, the racial logics that underpin the borders of Europe remain intransigent. This violence lives on in subtle and-as the next section will demonstrate-not so subtle ways.

Violent "Pushbacks" at the Croatia-Bosnia Border
If the violence of the camp in Calais can be stealthy, along Croatia's borders with Bosnia and Serbia, a seemingly different violence operates, the hallmark of which is open aggression and assault rather than insidious subtlety. Displaced people attempting to pass through Croatia to seek asylum in the EU are routinely violently "pushed back" and illegally expelled into Bosnia and Serbia, without having their asylum claims processed (Amnesty International 2019; European Council on Refugees and Exiles 2019). As both the EU and the Schengen Area expands, the countries on the EU's external borders increasingly become responsible for this border work. Whilst the Mediterranean Sea (Garelli et al. 2018) is "allowed" to passively deter and even drown migrants journeying to Europe by boat, the methods being used in Croatia take an active turn, and include direct, bodily violence. This empirical section compiles evidence of systematic physical assault on migrants crossing into the EU through Croatia, with details of how that violence is overt, at times explicitly racial in character, and often displaced to the EU borderzone, with the southern and eastern borders of Croatia providing the backdrop for this violent deterrence.
The "Balkan Route" or the "Western Balkan Route"-as described by Frontex (El-Sharaawi and Razsa 2019)-came to prominence as immigration to Europe peaked during 2015. Our fieldwork indicates that the Balkan Route was initially an alternative to the more dangerous sea routes. Typically, refugees travelled from Greece, through Serbia, and then either through Hungary into Austria and then Germany, or through Croatia and Slovenia. Countries of the route themselves experienced outward and forced migration, particularly during the 1990s onwards, and this was felt acutely in Bosnia where the Muslim population was subjected to genocide perpetrated by Serbian and Bosnian Serb forces; as well as forced displacement within the country and abroad. However, the relationship between that displacement and the current migration through the region is complex and, as Hromad zi c (2019:128) highlights, contradictory: in Bosnia, for instance, local populations both "embraced" the new migrants and "wished them gone"; a point that resonated in our own fieldwork encounters. More broadly, as Rexhepi (2018Rexhepi ( :2218 notes, the region came to be seen as a "zone of vulnerability requiring constant surveillance", which necessitated the increased securitisation of borders at multiple geopolitical scales. Border security is sometimes explicitly supported through racialised discourses and practices (see e.g. Rexhepi 2018), as well as implicitly through the racialised segregation of refugees into isolated and enclosed spaces (Bird et al. 2020). For example, the "closure" of the Hungarian border with barbed wire fences and increased violencethe original exit point of the early iteration of the Balkan Route-was supported by Viktor Orban's narratives about "defending Christianity" from the "virus of terrorism" and "mixed-race nations" (Fekete 2018). Each border closure along the Balkan Route pushed, pulled and diverted displaced people around the region in circulatory movements into increasingly precarious border crossings (Stoji c Mitrovi c and Vilenica 2019). When apprehended in Croatia and Slovenia, migrants are often expelled back to Bosnia in what has come to be termed as a "pushback".
The pushbacks and the associated violence are so endemic that during fieldwork we recorded that on any given day, even a short walk in the small border town of Velika Kladu sa would result in encounters with groups of migrants who would have been pushed back on that very day, mostly violently. When working with the activist group No Name Kitchen to provide showers and clothes for migrants living in informal squats, the sight of migrants with bruises, wounds or bandages from border violence in Croatia was an everyday sight. The effects of border violence are not just visible in the surrounding border towns; aid workers and volunteers reported seeing injured survivors of pushbacks as far away as the Eastern Bosnian town of Tuzla, and the Serbian capital Belgrade after they were forced to turn backwards on their journeys.
Violence and racism are key features of the pushbacks away from the EU, as practiced, particularly by Croatian border forces (Border Violence Monitoring Network 2020). Author Karol ına Augustov a has worked as part of a border violence monitoring network in the area to collect evidence of the violent and systematic nature of these expulsions. Since January 2018, 680 testimonies of violent pushbacks have been reported, based on detailed interviews with victims, by only a handful of activists working principally in Western Bosnia and Serbia (Border Violence Monitoring Network 2020). Of these, 270 involved the forced removal of minors. However, the number of people being pushed back are likely far higher than the number of cases that can be collated by the small numbers of volunteers working in the region. Our fieldwork conducted in a camp and informal settlements in Western Bosnia which receive refugees returning from pushbacks, as well as interviews with border violence monitors, indicates that between 50 and 200 people are forcefully removed from Croatia and pushed back into Bosnia each day during the summer months, without having asylum claims processed. Those being pushed back routinely return with visible physical injuries, and those subsequently interviewed reported being subject to an array of violent abuses whilst detained in Croatia.
It is important to detail some of the typical types of violence that asylum seekers have reported, specifically when being expelled from Croatia. Almost universally, migrants and refugees have mobile phones confiscated and destroyed, which are crucial for both navigation and contact with family members and smugglers. With visibly smashed mobile phones quickly becoming evidence of violent pushbacks (Figure 5), refugees showed us how their phones were being destroyed in less obvious ways-with Croatian police destroying the circuitry inside the phone, thereby concealing the criminal damage. Interviews with refugees and border violence monitors indicated that other possessions including money are also routinely taken by Croatian authorities, being a hallmark of this type of violence.
In contrast with the situation in northern France, hundreds of cases of physical beatings have been recorded in various forms: refugees commonly report being assaulted with punches, kicks or being beaten with batons by Croatian border agents, as well as electric tasers. For instance, a report made of an incident on 7 August 2019 recorded in Croatia near the Bosnian border at Maljevac, details how 23 persons between the ages of 25 and 32 were beaten with batons and fists, as well as being kicked. On 5 August 2019, a group of ten individuals aged 18-45 from Afghanistan were similarly beaten with hands and batons. From the extensively detailed testimony in this case is the following excerpt: They beat us one by one and then they told us to go to Bosnia. There was a river to cross to enter Bosnia. So each of us crossed the river after being beaten up. Some people got beaten up hard, some people got beaten up less. Me, the last one, they beat a lot. The first and the last ones get beaten up the most. One grabbed my shirt The systematic violence meted out to people crossing the borders into Croatia is corroborated too by reports from international aid organisations and NGOs (Amnesty International 2019; MSF 2017). For example, NGOs and aid organisations have reported the use of violence including electric shocks, beatings and sexual violence (Border Violence Monitoring Network 2020), as well as the use of razors, knives, and food depravation (MSF 2017). During fieldwork in 2019, we frequently met refugees with serious cuts, bruises and bandaged limbs which had resulted from violent pushbacks. In addition to the systematic recording of border violence, our research team spoke to over 50 migrants who had been stripped and returned from Croatia to Bosnia with little clothing, and with shoes confiscated, so that the long walks back to camps inside Bosnia would result in severe injuries to their feet. As in Calais, there is a concealment of this violence, notably in the fact that the physical beatings take place under the cover of night and often far from populated areas, amidst the forested hills and mountains along Croatia's border with Bosnia. It is important to stress that the racial nature of border violence cannot be contingent on the explicit avowals on behalf of the oppressor. However, in the violence reports and in our interviews with participants, a number of people testified to how race was explicitly evoked during their violent pushbacks. A report on 21 September 2018, for example, describes how a refugee from Afghanistan was beaten once he confirmed that he was Muslim: [the Croatian border police] started beating us ... I said why do you do that because this place is Europe ... The police officer said to me that I am an animal. He used the electricity and also baton. (Interview conducted by Karol ına Augustov a, 21 September 2018; extracts published by Border Violence Monitoring Network 2020) As Vaughn-Williams (2015:2) noted, the "animalisation" alluded to here, structures both "irregular" migrants' testimonies and is also revealing of the racialised nature of the abuses migrants suffer. Similarly, a report on 1 October 2018 described how a border official who administered physical violence made reference to Muslims killing the official's father during the Yugoslav wars as justification for not helping the refugee, and before beating them (Interview conducted by Karol ına Augustov a, 1 October 2015; extracts published by Border Violence Monitoring Network 2020). A report of a pushback in Slovenia from 5 August 2018 describes a Muslim woman having her headscarf refused and being told "This is the last time you [will wear] your scarf ... here is not Afghanistan, here is Slovenia, here is no Islam" (Interview conducted by Karol ına Augustov a, 5 August 2018; extracts published by Border Violence Monitoring Network 2020).
The racialising of Muslim migrants as unwanted "Others" within a white Europe (Fekete 2018; Rexhepi 2018) is explicit in these testimonies-but as with Calais, the racial violence is elided, and obscured through both displacement and denial. It would be inaccurate to see this violence as grounded in some form of perceived "Balkan exceptionalism and the perceived "criminality" of the region and the "Balkan Route" (El-Sharaawi and Razsa 2019), outside the sphere of "liberal Europe". Neu (2018:22) writes about the violence of "unexamined liberalism" and urges analysis to go beyond the atomised acts of violence, to understand instead the construction of violence through the spatial decision-making and policy networks. In order to make sense of this racial border violence, it is therefore vital to comprehend the wider geopolitical context. The proliferation of violent pushbacks can be directly linked to Croatia's aspirations for Schengen membership. For instance, a fact-finding report by the Council of Europe has linked the heightened focus on border security and strengthening as part of "preparations to access the borderfree Schengen area" (Council of Europe 2019:3; also European Council on Refugees and Exiles 2019). According to whistle-blower reports, violence against migrants at the border is systematic and institutionalised (Ombudswoman for Human Rights 2019). As such, countries including Germany, France and Italy are "protected" from the very presence of racialised outsiders; whilst Croatia is "rewarded" with Schengen Accession (Human Rights Watch 2019).
Our interviews with EU Commission and Frontex staff also indicate that there is widespread awareness of Croatian border violence-and so too its political role in supporting Schengen Area expansion. Moreover, Croatia has repeatedly been praised by EU officials specifically for its handling of "migration issues" (Vladisavljevi c 2018). In 2018, German chancellor Angela Merkel delivered public praise for its border governance, claiming Croatia was doing "an outstanding job with its security forces" (Merkel 2018). As M'charek et al. (2014) have evidenced, racial bordering practices of the European Union are enmeshed in and depend upon an array of security and technological infrastructure. The EU has provided crucial funding and support to the bordering operations along this borderzone (European Commission 2018a): 108 million Euros has been spent on border management in Croatia since 2014, with an additional 23.3 million Euros on emergency border security infrastructure since 2016, and a further 6.2 million euros in 2019 (Council of Europe 2019:26). Since 18 July 2018, Frontex has been using airplanes and visual data to stream to the over 1000-strong Croatian border police, to assist them with "interception operations" (Council of Europe 2019:34). Furthermore, the European Commission (2018b) has reported that technologies including watchtowers, thermal vision cameras, drones, helicopters, and barbed wire acquired through EU funds have been used in these operations to secure the Bosnia-Croatia border. Just as the EU has been shown to outsource border security to countries outside the union (Bialasiewicz 2012;Borg 2014), as countries join the EU and prepare to join the Schengen zone, they too become responsibilised for the EU's border security. In short, brutal physical violence is peripheralised to the EU's external borders.
There is much scope for further focused exploration of the array of financial, technological, institutional and infrastructural architecture that allows for violence to take place against migrants in Croatia and in former Yugoslav space. However, we return to the overarching point drawn from the empirical data; evidence clearly shows that EU policies support border violence through funding of material border security, and EU bodies to date have largely evaded confronting its existence, let alone its illegality. The acts of distancing the violence from Northern Europe and centres of power in Brussels, Berlin and Paris allow for a level of plausible deniability, or as Doty (2011) describes it, the space of moral "alibi", away from centres of geopolitical power, whose resources are crucial to maintaining the profound racial violence evidenced in this section.

Liberal Violence and the Racial Border
Human dignity is inviolable. It must be respected, protected and constitutes the real basis of fundamental rights. (European Union 2019) European border policies routinely deliver violence to racialised migrant groups, in abdication of both the liberal values and legal rights that the EU purports to uphold. This ethical and legal failure is no aberration, but is systemic to European border governance, as exemplified in this paper. At first glance, the racial violence of EU bordering is hard to detect-border violence is routinely displaced, concealed or denied where possible. In France on the north-western edge of the Schengen Area, the obscuring of violence takes on more subtle forms; those seeking asylum have restricted rights to even the basics of food, shelter and security through the pernicious violence of inaction (Davies et al. 2017). Yet border policy on the south-eastern edge of the Schengen Area features much more direct violence: systematic beatings of racialised migrants that seem more akin to the torture and corporal punishment of a pre-liberal age of governance.
We contend however that it is useful to conceive of both these different types of violence in our contrasting case studies as forms of liberal violence. Liberal violence involves the structural obscuring of violence as is de rigueur in liberal societies-but also reproduces race-blind logics, which have long been part of liberalism's own history (Losurdo 2011;Morefield 2009). The manifestation of liberal violence ostensibly contradicts the liberal values with respect to human rights, dignity, and asylum in the case of this paper, but this contradiction is sustained through the concealment of and displacement of violence. Liberal violence can also be obscured through other mechanisms-as Neu (2018) wrote of "just liberal violence", liberal violence is often conceived of as being part of a humanitarian project in and of itself.
The literature review demonstrated how European liberal thought has routinely been guilty of deep inconsistency in relation to applying liberal rights to racialised groups, an inconsistency which finds its roots in European colonialism, and can also be witnessed across other settler colonial spaces. In these postcolonial times, that inconsistency has metamorphosed into violence directed at migrants as racialised outsiders to whom, as the empirics show, the theoretical rights to asylum and freedom from physical harm can be withheld-and so too grossly violatedby the very states who espouse to uphold them. European and settler colonial states can articulate their liberalism as a political mechanism with which to put aside their colonial past and overlook their postcolonial present-but it simultaneously re-inscribes racial divisions through restrictive refugee and immigration policies and violent bordering practices. Crucially, the concealment and displacement of racial violence towards migrants does political work in maintaining the fac ßade of European liberalism over its compromised edifice.
Scholars of race and post-colonialism rightly ask us to consider how the racial logics underscoring racial violence are constructed, and similarly we must also uncover the ways in which that very racial violence is obscured. The multiple forms of social resistance, activism, and advocacy for migrant rights (Stierl 2018) and against racial violence in Europe suggest the potential to challenge what Losurdo (2011:344) described as liberalisms' exclusionary clauses. However, the strength of these resistances is at least partly dependent on the visibility of those injustices and their racial components. By eliding the violent and racial nature of EU border governance, liberal violence is allowed to continue undisturbed, easing itself into a position of social embeddedness. We can find liberal violence elsewhere too: mediated in the vast distance between the controllers of military drones in the US, and its victims in Pakistan and Afghanistan (Singh 2017); or within the toxic geographies of industrial sacrifice zones, where petrochemical companies are permitted to expose racialised communities to the liberal violence of pollution (Davies 2019). Liberal violence unleashes severe racial repression to indefinitely detained denizens on Nauru Island, or in the sweatshops and workhouses of the "global South" (Neu 2018) which create products cheaply for wealthy consumers, to whom the reality of that labour-and the racialisation of capital-is conveniently out of sight.
Within much migration scholarship on the so-called "European refugee crisis", the analysis of race and racism has remained curiously absent. At times this is a function of the fact that liberal society makes a virtue of its liberal image, which is why even contemporary racist practices often come pre-packaged with the epitaph "I'm not racist but" (Augoustinos and Every 2007). At other times this omission is accompanied by the under-theorised notion that "citizenship" rather than "race" is the target of discriminatory practice; as if, somehow, the historical workings of citizenship, biopolitics and borders have remained hermetically sealed from the systemic and epochal racism that has touched everything else. Occasionally, there is an implicit understanding amongst scholars that race plays a key role in the governance of migration. However, without explicit recourse to this notion there are pitfalls for scholars of migration. In simply unpicking the mechanics, techniques, and modalities of violent governance alone, there is a danger of missing the structural forces of race, racism and racialisation that shape border governance and border imperialism. Without an analysis that considers race centrally, the various technologies of violence may appear to be an almost unintentional, spontaneous, unpredictable result of shifts in the modes of liberal governance; an innocent side effect of advanced liberal bureaucracies, governmentalities, and technologies of bordering. The racial violence of the border outlined in this paper is no eccentricity or mistake, it is the foundation upon which border regimes are built.
We close this paper by re-affirming the importance of acknowledging the role of race and racialisation in all forms of border violence and paying attention to the ways that the racial violence of the state is routinely obscured. By calling this violence "liberal" we hope to focus the lens of critique back within liberal societies as a systemic issue, rather than as an aberration. Moreover, by considering these acts as forms of liberal violence, we foreground how the obscuring of violence against migrants, through concealment, displacement-or even denial or ignorance-helps sustain the pleasant and uncritical notions of "liberal Europe" and its civilisational pretentions, against which the uncivilised Other is drawn, and then violated, in Europe's supposed defence.