Turkey has been part of an expanding European border regime through the construct of ‘transit’. Against the essentializing use of this term, this article aims to draw attention to the varied nature of ‘transit’ migration based on ethnographic research conducted in two cities of Turkey: Edirne and Kayseri. While both cities are subject to the EU-ization of migration and asylum management, we argue that their geographical positions cause variations in how they experience EU-ization and how they receive border-crossers and refugees. Variations are further intensified by different configurations of il/legality and il/licitness in each city. We claim that neither the state nor the European border regime, as actors and producers of il/legality, can predetermine the outcome of such configurations. ‘Transit’-ing is rendered il/licit, depending on the visibility and duration of the stay of border-crossers and refugees, their impact on local economies and the attitudes of local state actors.

Introduction

Turkey became a part of the European migration regime1 under the notion of ‘transit’, starting in mid-1990s.2 This notion emerged in conjunction with policy concerns to capture irregular movements of migrants and refugees on the peripheries of destination states, especially those of the European Union (EU). As the notion came into favour with European policy-makers, it enabled member states to convince their neighbours to roll out migration and border-control policies in line with the EU interests—a process commonly known as externalization (Walker 2000; Walters 2002; van Munster and Sterkx 2006; Duvell et al. 2014). Yet scholars criticize the notion not only for its role in the spread of European migration and border-control policies, but also for its lack of explanatory power. The notion implies fixed starting and ending points, and essentializes ‘transit’ space as a space of moving through, disregarding the multiplicity of irregular and regular migration experiences in such spaces (Collyer 2010; Collyer and de Haas 2012). Building upon such critique and rejecting an essentialized understanding of ‘transit’ that leaves both border-crosser experiences and the spaces they move in undifferentiated, we compare two ‘transit’ cities in Turkey, namely Edirne and Kayseri. As an actual border city, Edirne is subject to irregular border crossings and serves as a short-term transit city, whereas, in Kayseri, we find the regular and longer stays of asylum-seekers3 who are sent there to wait out the results of their asylum and resettlement claims.

The comparison does not aim to demonstrate difference across two bounded locales. Rather, we seek to situate these places in their connections to the European migration regime and argue that variations are conditioned by the roles they play within this regime, which are assigned according to their geographical proximity to the EU borders. In so doing, we respond to the call of Glick-Schiller and Caglar (2009) for ‘locating migration’, for a double act of location: locating border-crossers and refugees in cities they use or are forced to use as entry points, stopovers and exits, while locating cities in larger, transnational forces and restructurings. Such an approach enables us to make three interventions to the literature on transit migration. We first bring out the different experiences border-crossers have in Kayseri and Edirne and thereby reiterate the importance of the city scale in accounting for the multiplicity of transit experiences. We then identify public perceptions of legitimacy (i.e. licitness) regarding border-crossers as a major driver of variation of transit experiences at the city scale. Last but not least, we argue that, by connecting transit cities to transnational regimes of migration, border-crossers rescale such cities. In the case of Edirne, this rescaling moves Edirne to a more central position within the European migration regime and allows trans-local cooperation. In the case of Kayseri, it reinforces the sense of being peripheral, both in the eyes of the EU and in the eyes of refugees. The demand made on Kayseri to contain refugees produces no perceived benefit in return. This causes scepticism towards the EU migration regime as well as refugees.

Our work is based on year-long (2013–14) ethnographic research conducted separately in two cities of transit in Turkey—Edirne and Kayseri—where heterogeneous experiences take place. In Edirne, as part of larger PhD research that took place on both Greek and Turkish sides of the Thracian border, Zeynep Kaşli collected local reports on migration, did archival research on local newspaper Hudut, conducted 25 semi-structured in-depth interviews, mainly with local state agents, representatives of civil society organizations, smugglers, lawyers and journalists, as well as participant observations at specific sites, such as the coach station, town centre and two specific neighbourhoods where ‘transit migrants’ are sporadically present (Kaleiçi and Yeniimaret). In Kayseri, Asli Ikizoğlu Erensü, again as part of larger PhD research, conducted semi-structured interviews with different groups of asylum seekers and refugees and followed them in their daily lives in the city, while also meeting local state officials, non-governmental organization (NGO) and charity workers, municipal employees and neighbourhood-level administrators. In order to better grasp refugees’ reception in the city, she also engaged in informal conversations with a variety of social and occupational groups, including real estate agents, shop-owners, nurses, journalists and urban planners.

In what follows, we begin with a short review of the literature on transit migration. We then discuss recent legal changes in Turkey’s migration management as a result of EU-ization, which are complicated by what we refer to as factors of licitness (i.e. public perceptions of the legitimacy of border-crossers). Before we analyse factors of licitness under headings such as local administrative cultures, temporalities of stay, visibilities in public space and impact on local economy, we examine how the geographical positions of the two transit cities lead to differential roles in European migration regime, which then shape the licitness of border-crossers.

Unpacking Transit

There exists a lively literature that discusses from multiple perspectives how the notion of transit fails to capture what it claims to describe. For one, many migrants and refugees end up remaining in transit for indefinite periods of time. Second, it is common for migrants and refugees to change their plans during transit, postponing further movement or revising its direction. Hence, Collyer and de Haas suggest we use the term ‘fragmented journey’ instead of transit migration (2012: 476–477). Papadopoulou-Kourkoula (2008) likewise argues that the experience of transit cannot be examined independently of the broader migration process. Sampson et al. (2016), on the other hand, contend that the representation of transit as a break and suspension in migrants’ lives obscures the variety of experiences with long-term effects that migrants might have during transit, such as marriage, child-bearing, community-building and other political struggles.

If the notion of transit is insufficient to account for the experiences of border-crossers, it is also deficient in illuminating the behaviour of those countries labelled as ‘transit countries’. Scholars studying the shifts in the migration policies of transit countries illustrate how such shifts cannot be understood solely from within the framework of transit migration, since governments often use the label of transit to link together differing areas of policy. Kimball (2007), for instance, when developing the term ‘transit state’, argued that such states actively sought to negotiate other interests with northern countries in return for compliance in the area of migration and border control. Yet, conditions of transit are not shaped exclusively on the plane of inter-state politics. On the contrary, transit migration constitutes a dynamic phenomenon that evolves with the interactions between various actors and their multiple interests at different levels. Andersson’s (2014) recent study on the ‘illegality industry’ along the clandestine route from Senegal and Mali to Ceuta and Melilla vividly portrays the cooperation among various sectors involved in controlling the Euro-African borderlands where a ‘gift economy’ between European paymasters, African border authorities, local youth and potential ‘smugglers’ keeps would-be migrants away from certain routes and opens other.

That the phenomenon of transit produces increasingly expanding borderlands is indicative of its trans-scale character. Transit cannot be isolated at any one scale. While policy-makers tend to break transit routes into composing nation states for purposes of international politicking and policy transfer, the experience of transit is hardly spread homogeneously across a country. Not only border-crossers concentrate in particular cities (and, within those, in particular neighbourhoods) for specific reasons, but also the experience that each city presents may not be directly inferred from national policies and regulations. Marconi (2008) therefore invites scholars to scale down to the ‘transit city’ from the ‘transit country’. While we take this invitation to heart, we also take care not to treat transit cities as containers. Rather, we follow Glick-Schiller and Caglar (2009) in questioning how border-crossers contribute to the ways in which cities are rescaled. In our specific case, we examine how border-crossers impact how transit cities are positioned vis-à-vis transnational regimes of migration.

National Legality and Local Licitness

Turkey’s inclusion into the European migration regime is usually discussed at the national scale by examining changes in the legal infrastructure. Indeed, Turkey has come a long way in terms of reforming its laws and regulations. Up until recently, Turkey’s migration and asylum governance was fragmented across various old laws, secondary regulations, circulars and by-laws. Their contemporary ineffectiveness, internal incoherence and inability to prevent human rights violations were subject to criticism by migrant and asylum rights advocates as well as the EU and the European Convention on Human Rights (ECHR) (Helsinki Citizens’ Assembly 2007; TOHAV 2014). However, recent changes that took place in this area are attributed more to EU-ization, namely adoption of EU measures and acquis as part of membership candidacy, than to civil society pressure (Içduygu et al. 2014). 2001 and 2003-dated Accession Partnership documents committed Turkey to harmonize with the EU requirements and priorities in all matters of migration: visa systems, border control, readmission, fight against human trafficking and ‘illegal’ migration, refugee status determination and reception conditions. 2005-dated national action plans further set the framework for legal and institutional transformation and culminated in the enactment of a new and much more comprehensive law, entitled the Foreigners and International Protection Law, in 2013.

In the area of border management, in addition to the 2001-dated bilateral readmission agreement between Greece and Turkey, the cooperation agreement between Frontex and Turkey was signed in May 2012 and an EU-Turkey readmission agreement was initiated on 21 June 2012 and ratified in the Turkish parliament in 2014. With this agreement, the EU has opened up the way for the current EU-Turkey Statement that came into existence on 18 March 2016.4 As part of existing Turkey-EU cooperation, the construction of a new detention centre in Edirne was already funded by the EU, whereas reception centres for asylum seekers and refugees are under construction again through EU IPA funds in seven cities, including Kayseri and Kırklareli, neighbour of Edirne. While the detention centre in Edirne is not designed as one of those seven reception and accommodation centres, it is still possible to ask for legal aid to seek asylum.

In the area of asylum specifically, while the recent Foreigners and International Protection Law grants more rights, for example with regard to access to health care, it also keeps two significant practices in place. On the one hand, it retains the defining feature of Turkish asylum governance that is the geographical limitation on the Geneva Convention of 1951, not granting asylum to refugees from outside of Europe, but only allowing them to remain within Turkey as the United Nations High Commissioner for Refugees (UNHCR) Turkey office tries to find a durable solution for them. Most refugees hope this solution will be resettlement to a third country, options being the United States, Canada, Australia and certain European countries. On the other hand, Turkey continues the mandatory dispersal of asylum seekers within the country. This policy was first specified in a by-law in 1994, although mandatory dispersal of particular population groups has deeper roots in imperial and Republican history, enacted in the 1934 Settlement Law. It aimed to keep asylum seekers and refugees away from borders and metropolitan centres. While, at first, only a dozen cities around the capital were used for this purpose, the policy has gradually extended to 60 cities out of a total of 80.

While Turkey’s legal restructuring aimed to streamline migration and asylum administration across the country, such streamlining cannot be assumed to automatically take place. Looking from the city scale, we find other factors that inform how laws may be implemented and how border-crossers and refugees may experience ‘transit’. If national legal restructuring cannot fully determine its manifestations at the local scale, then we need conceptual tools that address the different contextual factors and historical–social frictions that complicate this process. We thus turn to the notion of ‘licitness’ that coexists with and at times rubs against legal structures. Building on Abraham and van Schendel’s (2005) distinction between what states consider to be legitimate (‘legal’) and what people involved in transnational networks consider to be legitimate (‘licit’) in the current public discourses about terrorism, security and all sorts of ‘illegal’ border-crossing activity, an edited volume by Kalir et al. (2012) disentangles further ‘how dynamic “regimes of permissiveness” condition and sustain informal migration and how transnational flows generate zones of licitness’ (Kalir et al. 2012: 19). They argue that such zones are created through negotiations among many actors, including state agents, who are part of networks of informal transnational brokerage. In our case, licitness includes not only perceptions regarding border-crossers and refugees, but also state and non-state actors’ local relations with and understandings of the EU. Below, we will address a multiplicity of factors that feed into the licitness of border-crossers and refugees in the eyes of the ‘locals’. These factors do not necessarily overlap with criteria defined by state law or international authorities. But we first need to set these against the background of relations with the EU migration regime, which are in turn determined by the geographical locations of the two cities.

Geographical Positions and Roles

Although, in different ways, both Edirne and Kayseri function as the EU’s external borders, Edirne, historically considered as the gate to the Balkans, is now a border town neighbouring two EU member states: Greece to the west and Bulgaria to the north. Kayseri enables extraterritorial processing of asylum seekers within the bounds of the international asylum regime. However, the consequences of serving as external borders have not been the same in two cities. In what follows, we illustrate how cooperation with regard to border control has drawn Edirne closer to Europe, while the presence of asylum seekers and refugees has not altered Kayseri’s distance.

As a border town, Edirne is not a city of settlement for asylum seekers and irregular migrants, whereas most of its population is composed of people of Turkish descent from Bulgaria, Greece and other Balkan countries who, at different periods since the demise of the Ottoman Empire, migrated to Edirne as well as other Western cities in Turkey. Edirne has acquired a new role for European border protection especially in the aftermath of the 2001 readmission agreement between Greece and Turkey. Although not strictly implemented by either party, cooperation between Greek and Turkish authorities in the area of border control became a part of a broader neighbourly rapprochement of 1999. This was also a time when the Turkish state was relaxing its visa regime for citizens of all neighbouring countries, thus easing local cross-border interactions. Consequently, the people and the state authorities in Edirne were able to come into closer contact with Greece as well as Bulgaria.

Edirne has been the top province, where almost half of the total apprehensions at Turkey’s border cities were done, namely 155,628 out of 392,545 between 1995 and 2010 (Karci Korfalı et al. 2014). While the unauthorized border crossings from various countries to the east and south of Turkey were directed, in 2005–09, through the sea routes of the eastern Aegean islands, it later shifted to the Evros-Edirne border in Thrace since 2010, with the daily number of border-crossers reaching up to 250-300 (ProAsyl 2012) which constituted the majority (estimated at more than 80%) of migratory flows into the EU (Hatziprokopiou and Triandafyllidou 2013). According to the report of the Turkish Parliament Human Rights Commission (2012), in 2010, one-third of the unauthorized border-crossers (11,384) and, in 2011, half of them (22,664) were apprehended in Edirne. As a deterrence against this ‘flow’, the Greek–Turkish border has gradually become highly securitized through the involvement of the intelligence-driven EU-agency Frontex monitoring the Evros river and the land crossings; the construction of new detention centres on both sides of the border; the local Operation Aspida (Shield) and deployment of 1,800 extra officers along the 206-km-long border in Evros region and the completion of the fence along the 12.5-km-long land border; and increasing intelligence sharing between Greek and Turkish security forces (Migreurop 2009; ProAsyl 2012, 2013). Following the measures taken by Frontex and implementation of stricter and joint border controls as a result of regular monthly meetings between Greek and Turkish local authorities since 2013, the migratory routes have shifted back towards the Aegean Sea in the south (Amnesty International 2013) and more recently to Bulgaria in the north, where new fence construction immediately started. More interestingly, as the local state agents’ accounts show, these monthly meetings, recently including the Bulgarian counterparts, worked as socialization channels that took on a new life outside work, through family visits in the weekends and cross-border sports tournaments organized among the personnel.

These new practices of cooperation stand in stark contrast to the historically conflictual relations that go back to the Greek independence/secession of the Greek millet in early nineteenth century, reoccur during the Turkish independence war/Great Catastrophe in the early twentieth century and endure with the Cyprus and Aegean Conflicts. It is already stated that Greek–Turkish relations were always shaped at the periphery of Europe and in its ‘long shadow’ (Anastasakis et al. 2009). Yet the recent process of Greek–Turkish rapprochement and the externalization of the EU border regime seem to have a deeper societal impact as manifested in the changing role of Edirne from protecting the national border against Greece into controlling the EU’s external borders. In this process, Greeks have gradually ceased to be the Other, while it was now the irregular border-crossers that constituted an Other, common to both Greeks and Turks on each side of the border. As a top-rank officer on border management in Edirne says, during our informal talk:

Of course it is always delicate. But through these regular meetings we are certainly building a common understanding, we have a common aim, we have to cooperate to solve it together and now have a personal contract in such a way which is harder to break.

In Kayseri, on the other hand, there is a major geographical difference: Kayseri is in the interior of the country and shares no physical borders with the EU. Yet, it mimics a border that separates Turkey from the EU as well as several other resettlement countries. In August 2014, there were 2,200 Iranians, 1,150 Iraqis and 550 Afghans in the city, sent there by the Ministry of Interior to wait either for their refugee status determination by the UNHCR or for the results of their resettlement applications to various European countries, the United States, Canada and Australia. While these asylum-seeker groups have been assigned to Kayseri since 1994, when Turkey systematically began its mandatory dispersal policy, Syrians are relative newcomers. Around 15,000 Syrians under temporary protection status spontaneously arrived in Kayseri between 2013 and 2014, without the direction of governing authorities. However, certain religious charities played the role of a magnet by providing shelter and aid to incoming Syrians.

While Kayseri acts as an external border, it cannot engage in border-zone interactions like Edirne. On the level of bureaucratic interaction, instead of face-to-face meetings or joint trainings with personnel across the border, state personnel in charge of refugee presence are on the receiving end of occasional visits by UNHCR and EU representatives, without being able to control their timing or content. This structuring of relations situates local officials in a defensive position from the beginning. In her interviews with officials across different departments of the local state apparatus, Asli Ikizoğlu Erensü typically encountered two lines of defence that officials used when pressed for what rights were accorded to refugees in practice and what changes EU-ization brought. The first line of defence amounted to the argument that ‘refugees do not want to be here anyway, so why should we care about what they can access’. This stance was most clearly taken in the field of education. An assistant to the provincial head of K-12 education defended the department’s lack of interest in refugee pupils’ access to school on the grounds that refugees did not want their children to learn Turkish (the language of instruction) and that they would only enrol to qualify for certain UNHCR funds, after which they would simply drop out. The second line of defence criticized international agencies for demanding and investing in the welfare of refugees without mirroring the same for the welfare of locals.

These lines of defence, more importantly, were not unique to local officials, but were shared by other segments of the urban society. From the doctor at the hospital to the taxi driver, from the housewife to the factory worker, every informal chat that Asli Ikizoğlu Erensü had supported the same cynicism regarding refugees and the international agencies that were involved on their behalf. Add to this the tendency to explain asylum as a geopolitical conspiracy, such as by believing that all asylum seekers are paid by the United Nations to become an asylum-seeker or that the United States accepts Iranian refugees only to turn them into spies against Iran. To be sure, these reactions cannot be understood independently of the fact that the city is predominantly right-leaning, and that conservatism in Turkey comes with its particular forms of anti-Europeanism and anti-Americanism. This, however, does not mean an outright rejection of the ‘West’, but a more ambivalent relationship simultaneously including feelings of desire and resentment. A sizable portion of Kayseri’s population lives in Europe, particularly Germany, as migrants visit Kayseri over the summer. Their visits are enthusiastically marked by local companies, which go as far as putting up ads on public buses welcoming them home. In such welcomings by those left behind, not only the notion of home becomes contested, but also an underlying tension is visible with regard to the question of which home is better.

Kayseri’s border mimicking vis-à-vis refugees does not shift the perception and experience of Europe and the ‘West’ as Other. Rather, the border-mimicking experience builds on existing frames of reference. Europe and the ‘West’ continue to be perceived both as distant objects of desire (as seen through the eyes of refugees) and as dangerous subjects (manipulating in real politik and corrupting with regard to one’s sense of identity). Refugees who want to move on to and become a part of this Other are then doubly-othered. In the eyes of locals, they become those who ‘sell out’ to foreign institutions and states, and betray their countries of origin. The ambivalence between desire and distrust gives way to a peculiar sense of patriotic pride that vilifies both the refugee and her hoped-for destination. While most refugees in Kayseri are there legally, this is not sufficient to make their presence legitimate in the eyes of the locals. In Kayseri, the desire to cross the border is much more stigmatized than in Edirne.

In sum, while in Edirne, externalization of the EU borders takes place in tandem with rapprochement with Greece, in Kayseri, it does not bring or follow any such rapprochement with those on the other side. As a result, in Edirne, otherness shifts from its historical or usual subjects, Greece and Europe, to irregular border-crossers. In Kayseri, otherness expands from those countries whose borders now crisscross the city without effectuating a shared border zone, and includes refugees caught and defined by the same borders. Although border-crossers and refugees are subject to othering in both Edirne and Kayseri, in the former city, this othering serves to bring Europe closer, whereas, in the latter, it reproduces the existing distance to Europe.

Factors of Licitness

Local Administrative Cultures

While national legal reforms strive to standardize and EU-ize migration and asylum governance, local administrative cultures may lead to variations in the implementation of those reforms. Our comparison suggests that perceptions regarding the EU and ideological distance to the ruling party condition how local administrative cultures receive border-crossers and refugees. In both Edirne and Kayseri, the differing levels of suspicion towards the EU-ization process impacts the compliance of local state actors with new rules and regulations. The two cities are also differentiated by their response to the arrival of Syrian refugees, which is in turn structured by ideological distance to the politics of the ruling party.

Local alignments with EU-ization: In Edirne, in 2014, the detention centre was still managed by police together with the civilian officers.5 As just mentioned, local migration authorities in Edirne recognize the externalization of European borders and comply with the EU and national authorities’ measures othering transit border-crossers. Yet, this othering does not automatically lend itself to criminalization and in fact leaves room for an implicit understanding of border-crossers’ victim status. Local state actors of the new migration regime, both the police and civilian officers, do sympathize with the desire to cross the border and it is in this light that they generally view ‘illegality’ as victimizing rather than criminalizing. As the then provisional civilian head of the detention centre says in May 2014:

Unless the root causes of migration are not solved, people will continue to try to cross these borders. But we do what we have to do. We cannot let everyone pass the border. After all we must also show everyone that we do our job right.

While officers comply with the policies passed down to them, they remain suspicious towards the new role assigned to Edirne as part of the new EU migration regime, since they foresee the ways in which border-crossers’ presence would be experienced in Edirne. Sharing similar concerns with the civilian officer regarding the financial burden that this new role brings to the local government in Edirne, the police head of the detention centre in May 2014 also shows his resentment towards the Greek and the EU authorities for not sufficiently supporting them financially for their operations which, according to the rules, starts with apprehension on the border, continues with processing asylum claims and, in cases of rejection, deporting them to ‘safe third countries’. Local state actors complain that the costs of running a detention centre, namely providing everything from diapers for babies to sanitary pads for women, is being disregarded not only by the EU authorities, but also by the bureaucrats in Ankara who, as the provisional head of the detention centre says, can only grasp what running a detention centre takes when they make field visits.

Such friction between central authorities who pass down orders and local officers with their own opinions about those orders exists also in Kayseri and, for that matter, in all cities of mandatory dispersal. Rights granted to refugees on paper by central authorities stop short of materialization unless their local practitioners are trained to facilitate them or separate budget lines are identified for their realization. Refugee rights advocates not only in Kayseri, but also in other cities of mandatory dispersal that Asli Ikizoğlu Erensü visited recount numerous stories of wrestling with local personnel who act as gate-keepers especially with regard to access to education and health care. Local budget concerns play an important role in such gate-keeping. For example, while Social Assistance and Solidarity Foundations under the governor’s office have the authority to distribute aid to asylum seekers and refugees, the amounts distributed and the frequency of distribution vary from city to city. When probed for the reasons of such variance, local officials tend to compare the numbers of citizen aid applicants and refugee applicants. In cities where citizen aid applicants are many in number, the notion of helping ‘our’ poor first prevents refugees from receiving much aid, whereas, in cities where citizen applicants are lower in number, refugees can get a bigger share. Thus, in the eyes of local officials, the licitness of refugees is dynamically measured against the impact that their welfare might make on the perceived welfare of citizens. If reforms under EU-ization necessitate recognition of certain rights for refugees, local officials filter that recognition through a sense of duty of protecting citizens.

Local alignments with party politics: Kayseri is generally known as a conservative city with a large Justice and Development Party (JDP) voter base and home to various religious factions which may act in tandem with JDP to boost their reach and power but are not necessarily organically tied to it. The arrival of Syrian refugees in the city on the one hand provided an opportunity to grow closer to JDP. On the other hand, JDP’s portrayal of the Syrian civil war as Alawite brutality towards Sunnis resonated in the discursive universe of many such groups. Therefore, eager to embrace the ruling party’s position on the Syrian civil war, or perhaps concerned that this is what is expected of them, the governor’s office, municipal heads and particular charities joined forces and established a coordination to provide for Syrian refugees. It was for Syrian refugees that the municipal soup kitchens doubled the number of their servings, distributed daily bread to each registered Syrian family and continued to do so even when it became a known secret that municipal budgets were struggling with huge deficits. Such coordinated efforts were distinctly lacking when it came to other groups of refugees in the city.

In contrast to Kayseri, Edirne is one of these cities labelled as the ‘city of white Turks’ with a large Republican Peoples Party (RPP) voter base with unconditional loyalty to the principles set by M. Kemal, the founding father of the republic, and with suspicion towards all religious and ethnic movements which, in their view, are threats to the founding principles. Therefore, the arrival of Syrian refugees deepens the tension not within the local, but rather between the local residents and the JDP government which, in their view, is acting against the people and founding principles of the state: ‘I think it is Erdogan’s [then prime minister] fault who let them travel freely and settle wherever they want,’ says a member of the City Council, an independent civilian initiative yet legally operating under the municipality. It is seen as a matter of bad management at best, as the government let Syrians go anywhere, even all the way to Edirne, instead of containing them in the border towns. It is bad intentions at worst, as the government let them settle in Turkey to increase Arab influence in Turkey while cutting Turkey’s ties with the West, away from ‘modern way of living’ with which everyone in Edirne proudly identifies themselves and their city. Consequently, while in Kayseri, the proximity of local authorities and certain sections of the civil society to the ruling party’s politics endowed Syrian refugees with an initial licitness not granted to other refugee groups; in Edirne, non-state actors kept their distance from Syrians, whom they automatically identified with the conservative Islamist incumbent government that they were critical of.

Temporalities of Stay

In both Edirne and Kayseri, border-crossers and refugees are tolerated based on the understanding that they are here temporarily. While most locals of Edirne sympathize with the desire to cross the border, when probed further, they openly declare that border-crossers are not welcome to remain and settle in Edirne. Border-crossers’ temporary presence right before they cross the border is considered as a licit act, since this is quite the opposite of the claim for settlement, even for a temporary period. Their presence becomes less licit if they are caught by legal authorities before they cross the border. From the moment border-crossers are caught on the border and brought into the detention centre until they are released with an administrative paper to leave the country or for official deportation, their presence is marked with a new label: a label that itself has been changing over the years, in the way the local newspaper Hudut uses, from ‘illegals’ (kaçaklar) and to finally now ‘illegal refugees’. In other words, while the objective to leave is seen as licit despite its illegality, right after the crossing attempt takes place and ends with capture, not only the failed attempt is then considered illegal due to violation of legal entry and exit rules, but also the ex post presence of the border-crosser becomes a criminal act. In that sense, the representation of the ‘smuggled border-crosser’ also follows international humanitarian discourse that victimizes and criminalizes the border-crosser at the same time (Kaşli, 2011). However, with the arrival of Syrians, the invisible yet longer presence of border-crossers in a few neighbourhoods began to stand out as a new temporal form. Opaque in terms of length and meaning, it shifts from the usual practice of licit temporary presence in Edirne as the last stop before the border crossing into an illicit place of refuge for being the furthest point in Turkey away from the Syrian border and closest to the EU border, which thus puts them in a suspect position with possibilities of engaging in illegal acts of border crossing.

In Kayseri, temporary is measured in years during which refugees rent apartments, find employment, consume and use public spaces. More than 40 per cent of the Afghan refugees and asylum seekers in Kayseri, for example, had been there for longer than three years in 2013.6 Yet, while interviewed refugees shared individual stories of discrimination, exploitation and harassment, it was not until the arrival of Syrian refugees in the city that a mass disgruntlement surfaced. Between 2013 and 2014, the number of Syrians in the city surpassed the 10,000 mark (and was even estimated by charities to be upwards of 15,000) and a mass-lynching attempt erupted in summer 2014 (Ikizoglu Erensü 2014). The number of Syrians at this point was at least three times as much as the total of the other refugee groups in the city; however, local reaction could not be explained only by numbers. Small talk with local shop-owners regarding refugees in the city registered another, more significant difference between Syrian and non-Syrian refugees. In 2013, when asked what they thought about refugees in the city, most would remark that refugees wanted to and were trying to leave Turkey. In 2014, the same probes provoked anxious comments about Syrian refugees, who could not easily be dismissed as temporary, although they had been in Kayseri for a shorter duration than other refugee groups. Perceptions regarding the possibility of permanence were further aggravated by the perceived difference in the approach of local state actors to Syrians. While local state actors were not perceived to be providing for other groups of refugees, their active involvement with regards to Syrians incited antagonism, fuelling the sense that the stay of Syrians will impact local communities more extensively than other refugee groups do.

Visibility in Public Space

Transit migration as a phenomenon that is not so much visible in the public space in Edirne as border-crossers try to leave Edirne with no trace. For example, as early as the first half of 1999, in Hudut, there are on average three news pieces every month on the apprehension of ‘kaçak’/’kaçak mülteci’ in Edirne; this goes up to six in the summer months. These were almost standard pieces, using a template to report daily legal and illegal border crossings only with changes in numbers and countries of origin. Besides this standard ‘News from the Border’ column, border-crossers’ temporary presence is worth recording only when there is a case of drowning or dead bodies found by the river or a relative searching for someone who got lost on her/his way to Greece. In other words, unlike the positive representation of ‘visitors’ from Western countries who stop in Edirne during their road trips or who come as invitees of some international or regional events organized in Edirne, irregular border-crossers are seen either only as numbers or as voiceless subjects involved in illegality, even when they are indeed victims of the illegal acts of others, being left alone to die or being kidnapped.

It is not wrong to claim that the only time at which border-crossers’ personal stories, their traumas and aspirations were heard and maybe circulated a little in Edirne seems to be during a social responsibility project conducted at a public high school in 2013. Apart from that, the Migrant Solidarity Network, a non-hierarchical activist group from Istanbul, together with the City Council of Edirne organized a day-long event on irregular migration in April 2011 (Hudut, 16 April 2011). However, the fact that the teacher who mentored the social responsibility project cannot recall this event, despite being active in several NGOs, shows that the migration issue was and still is off the radar of even these socially engaged people in Edirne. The experience of high-school students, who volunteered in preparing a playroom for kids detained with their families in the newly constructed detention centre, remained the only example of locally initiated activity in town. None of the local officials or the members of civil society organizations involved in this project has tried to continue similar projects or to endure relations that were set up with the detention centre during this project, whereas the invisible presence of border-crossers in the new detention centre is now known to many people thanks to the statistical and procedural information in the local newspapers.7

Moreover, just like the issue of temporality, border-crossers’ visibility has taken a new though not necessarily positive meaning as Syrians with temporary protection status have begun to come to Edirne. Since the summer of 2014, people have begun to mention Syrians renting houses mainly in two neighbourhoods. One of these neighbourhoods is Kaleiçi, one of the oldest neighbourhoods and the old non-Muslim quarter of the city, lately inhabited by the working class and the urban poor, namely migrants from different Anatolian cities as well as Romas from the outskirts of Edirne. The other one is Yeniimaret, an old neighbourhood at the outskirts of the city and where the new detention centre is located. These neighbourhoods are places where one can find cheap, old and abandoned houses, as opposed to new apartment blocks, where newcomers are easily recognized and visible each time they enter and exit the building. Even though many people mentioned seeing Syrians every now and then in the neighbourhood markets for their immediate needs or the old market place in Edirne centre, they are not so much seen in the parks, cafes, weekly bazaars or the new malls of the city. Two other public spaces in which Syrians as well as other irregular border-crossers are seen transiting are the coach station, where they come from Istanbul mostly early in the morning and wait to be picked up for their next destination, and the yard of the Selimiye mosque, where people whose entire belongings can easily fit into small plastic bags spend the night.

While the legal status of Syrians under temporary protection is not fully known in public, they blur the picture for local authorities and the people in Edirne. During one meeting with the civilian head of the detention centre and a member of the city council, they revealed their suspicion in the following way: ‘You ask them where are you from and he just makes up something. They do not tell the truth.’ The survival strategy, in which irregular border-crossers have to make the right confession at the right moment to the right authority, becomes a reason for mistrust for the locals and questions the licitness of their presence in the city. While the reasons for migration are well founded, being seen with a border-crosser in public or even renting your house to border-crossers is not considered licit. Thus, the moment of interaction with any local becomes the moment when the boundary of the licit–illicit is redefined. This line of il/licit does not always coincide with the line of il/legal act, especially in the case of interactions with Syrian border-crossers who have the papers to stay and travel in Turkey.

In Kayseri, refugees cannot escape visibility as they live in the city. Locals pick up less on phenotypical difference than on other aspects of appearance, such as make-up and clothing. These aspects prove to provide easy material for moral panics, which then inform the licitness of refugees. One such moral panic, for example, ensued around the traditional clothing of Syrian men that includes a long, white dress at the height of the tension between communities in the summer of 2014. As a local journalist remarked in an interview with Asli Ikizoğlu Erensü, some locals became worried that men’s underwear would suddenly be visible to local women and some even began speculating that Syrian men did not wear underwear, revealing their genitals. Local women then needed to be protected from men in dresses and on such moral panic rose a coalition of local men.

While Syrian men were just trying to wear the same clothing that they wore in Syria, there are also refugees who want to look different to how they looked back in their countries of origin (especially true in the case of Iranian refugees). Those refugees actually find themselves in a dilemma: on the one hand, they wish to be finally free with regard to the way they dress and look. On the other hand, they soon realize that such freedom remains to be socially costly and feel the need to protect themselves. Female refugees become subject to cat-calling and LGBTI refugees become subject to outright harassment. Yet many continue to fight for the right to physically express themselves. In the words of a frustrated LGBTI refugee:

I began growing my nails as soon as I crossed the border, because this was something I could never do in Iran. But here [Kayseri] is no different than Iran… I will continue to grow them [nails] because I need to be myself finally.

By challenging the heteronormative urban order, such acts of self-expression also feed into perceptions of an illicit presence.

Yet, cultural purity is sought with regard to not only gender roles, but also religious identity. The location of the church, where Christian refugees attend, has to be kept a secret from the locals, with no markings or signs on the building. In contrast, Bahai refugees can afford to be more openly public. Their house of worship is clearly identified in Turkish, and they report no incidences of harassment. Interviewed Bahai refugees half-seriously half-jokingly claimed that this was because many locals either could not understand what Bahaiism was or thought Bahaiism was a sect within Islam. As such, Bahaiism is not perceived to threaten religious normativity in the city.

Local perceptions of licitness are also shaped by the extension of visibility. Most refugee groups in Kayseri cluster in the old housing stocks of the city centre where locals have vacated. Locals visit the city centre for business and administrative affairs but prefer not to take residence there. Such residential separation limits interactions between locals and refugees especially in the evenings, conferring to refugees a tacit licitness. Interviewed refugees explain their clustering pattern by referring to their weekly reporting requirements at the Foreigners’ Police station (to verify through signature or fingerprints that they have not illegally left the city). Syrian refugees, on the other hand, have sought shelter in a variety of neighbourhoods no matter how far away these are from the city centre because their temporary protection status frees them from reporting responsibilities to the police. Looking for even cheaper accommodation or following the lead of the religious charities that aid them, Syrians thus spread to neighbourhoods where there was no previous familiarity with the refugee presence in the city. The uncontained settlement pattern of Syrians was not only perceived to be threatening, but also put a sudden, increased demand on public space resources in unexpected neighbourhoods. In such neighbourhoods, this combination pitted locals against refugees to the extent that the apparent reason behind the mass-lynching attempt in the summer of 2014 was a brawl over the use of a public park.

Impact on Local Economy

In both Edirne and Kayseri, migrants are valuable to the economy exactly because their irregular/informal/unregulated status enables extra exploitation. The value extracted out of them in this way makes them tolerable yet by no means accepted. Again, in both Edirne and Kayseri, material assistance given to migrants by state and local authorities or charities raises eyebrows, as locals are resistant to sharing resources.

Smuggling does not encounter much social stigma in Edirne while nobody would even think to demand it to be legalized. People openly say that everybody knows who is involved in smuggling circles in their neighbourhood or village, and also point out that the act of smuggling of people and goods has been considered a legitimate practice and a source of income for border citizens for a long time. It was on that Thracian border that the illegal crossings of the Turkish Muslim minority of Western Thrace into Turkey had continued until the democratization of minority rights in Greece in the 1990s and Kurdish and leftist asylum seekers were also smuggled into Greece. The same route has also been in use for smuggling of goods—mostly electronics to Turkey and textile to Greece. Yet, like all border-zone activities, it takes two sides to smuggle: according to local smugglers in Edirne, smuggling people is increasingly becoming difficult on the Turkish side, not necessarily because of Frontex intervention or the hardening of Turkish police, but because of the increasing penalties on the Greek side which make the smuggling occupation less attractive for their local counterparts. While few houses rented in Kaleiçi and Yeniimaret points out that housing might be a niche for economic returns, the fact is that both locals in Edirne, politically active or otherwise, and local state actors involved in migration management are wary of ‘defending European borders with our tax money’ and ‘feeding migrants with our tax money, as if we already took care of our own people’.

In Kayseri, local economy benefits from the presence of refugees not through smuggling, but through abuse of refugees in legal sectors such as real estate and factory production. Because there is no central body that regulates rents, refugees complain that they are asked to pay either higher rents or the entire year’s rent in full in advance. The opposite happens when it comes to labour: because refugees are informally employed, they are paid lower wages or sometimes not paid at all and made to work longer hours. Refugees in the city also have to shop for food, home appliances and utensils, fuel and clothing, not only benefiting the economy by increasing volume, but also by paying taxes that never return to them: tax is pre-included in items sold in Turkey. While there is no local initiative against the economical exploitation and unfair treatment of refugees in the city, there have been sporadic mobilizations against the aid that Syrian refugees receive. One of the local coordinators of a religious charity moaned in an informal chat how they had to conceal aid distribution to Syrian refugees in a particular neighbourhood because local Roma protested that they should also be receiving aid. While contributions of refugees to consumption and even production feed into their acceptability as long-term tourists, any use of aid and benefits that indicates possible permanence and inclusion draws objections.

Conclusion

This comparison of ‘transit’ experiences in two cities of Turkey not only reaffirms Collyer and de Haas’s (2012) emphasis on the multiplicity of irregular and regular migration experiences in what are labelled as ‘transit’ spaces, but also highlights the importance of the city scale in accounting for this multiplicity. While we thus situate border-crossers and refugees in cities, we simultaneously situate cities in larger, transnational regimes of migration—specifically the European migration regime. Each city’s geographical position assigns them particular bordering duties within the European migration regime. We then argue that this has implications for the encounters between the ‘locals’ and the border-crossers, informing the public perceptions of their legitimacy or, in other words, licitness. We claim that different experiences of bordering produce dissimilar disjunctions between national legal reforms triggered by EU-ization and their implementation, as these are tampered with by local senses of licitness.

The fewer the city’s direct ties with the EU border, the more violent is the othering of border-crossers and refugees. In Kayseri, the legality of border-crossers and refugees that arises from conforming to the international migration regime ironically becomes a source of illegitimacy. In contrast to Edirne, where both the international migration regime and those it ir-regularizes are deemed legitimate in their own way, in Kayseri, the perceived illegitimacy of the international asylum regime and refugees mirror each other. In other words, while the presence of refugees in Kayseri showcases the illicit yet legal, their border crossing through Edirne sits at the intersection of the licit and illegal.

1.
Our use of the notion of regime is informed by Sabine Hess and her colleagues’ multi-sited ethnographic project entitled ‘Transit Migration’ (2002–04), conducted in the south-east border region of the EU (including the Former Yugoslavia, Greece and Turkey). For the team of researchers involved in this project, the notion of regime makes it possible to include a multitude of actors whose practices relate to each other in a space of conflict and negotiation and are not ordered in the form of a central logic or rationality (Tsianos et al. 2009; Tsianos and Karakayali 2010; Hess 2012).
2.
Turkey was first captured as a transit country in 1995 in an IOM report, which was consequently quoted as scientific evidence in the EU Accession Partnership Document (APD) of 2000. Ever since, more reports and studies have continued to portray Turkey as a transit country (IOM 1995; Içduygu 2003) and, more significantly, government policies also evolved to embrace the notion, with bureaucrats increasingly internalizing the label with the institutional help and expertise of IOM field mission as well as the EU authorities (Hess 2012).
3.
We use the term ‘refugee’ for everyone who self-identifies as such. We use the term ‘asylum-seeker’ when it is legally relevant, namely when we specifically refer to those who have not yet received refugee status. Both asylum seekers and refugees can be found in Kayseri. Syrians in Turkey categorically belong to neither grouping; they are under temporary protection.
4.
According to this EU-Turkey Statement, ‘all new irregular migrants’ arriving in Greece after 20 March were going to be returned to Turkey, in exchange for resettlement of Syrian refugees residing in Turkey to EU member states, accelerating visa liberalization for Turkish nationals and improving existing financial support for Turkey’s refugee population. In total, 1,546 irregular migrants have been returned from Greece to Turkey in the course of 2016. According to the European Commission report in June 2016, 462 people were returned from the Greek islands to the Turkish coasts by Frontex boats based on this deal, including 31 Syrians who returned voluntarily, whereas 511 Syrians resettled in various EU countries, which exceeds the number returned under the EU-Turkey Statement. The rest of the returns were done under the framework of the Bilateral Readmission Protocol between Turkey and Greece, which was signed in 2001. Information retrieved from http://multeci.org.tr/haberdetay.aspx?Id=140.
5.
The Directorate General of Migration Management under the Ministry of Interior became the main authority in this area in April 2014; however, in provinces, Foreigners’ Police were still kept active to assist with the transition period until May 2015. We have witnessed the messiness of this transition period during our field research in both Edirne and Kayseri.
6.
This percentage is calculated based on a combination of UNHCR statistics and interviews with refugee community leaders.
7.
Finally, in September 2015, the issue received a lot of local as well as national and international attention when thousands of people, mostly Syrians, who were mobilized through the ‘Crossing No More’ Facebook page and, in reaction to the deaths in the Aegean Sea and smugglers’ exploitation of their fragile situation, decided to march to Edirne to demand ‘safe passage’ to Greece. During her short visit to Edirne at that time, Zeynep Kaşli observed that the majority of the public in Edirne were intimidated by the appearance of this crowd in the town centre, whereas there were also some local initiatives eager to help with their basic needs. Yet this relationship between locals and the border-crossers were cut when the people were transferred to a stadium that is away from the centre and guarded by the police officers.
ABRAHAM
I.
VAN SCHENDEL
W.
(
2005
) ‘
Introduction: The Making of Illicitness
’. In
Van Schendel
W.
Abraham
I.
(eds)
Illicit Flows and Criminal Things: States, Borders and the Other Side of Globalization
 .
Bloomington, IN
:
Indiana University Press
, pp.
1
38
.
AMNESTY INTERNATIONAL
(
2013
) ‘Greece: Frontier Europe: Human Rights Abuses on Greece's Border with Turkey.' Available at <https://www.amnesty.org/en/documents/EUR25/008/2013/en/> (Accessed September 2016).
ANASTASAKIS
O.
NICOLAIDIS
K. A.
OKTEM
K.
(eds) (
2009
)
Under the Long Shadow of Europe—Greeks and Turks in the Era of Postnationalism
 .
The Hague
:
Brill
.
ANDERSSON
R.
(
2014
)
Illegality, Inc. Clandestine Migration and the Business of Bordering Europe
 .
Oakland
:
University of California Press
.
COLLYER
M.
(
2010
) ‘
Stranded Migrants and the Fragmented Journey
’.
Journal of Refugee Studies
 
23
:
273
293
.
COLLYER
M.
DE HAAS
H.
(
2012
) ‘
Developing Dynamic Categorisations of Transit Migration
’.
Population, Space and Place
 
18
:
468
481
.
DUVELL
F.
MOLODIKOVA
I.
COLLYER
M.
(
2014
)
Transit Migration in Europe
 .
Amsterdam
:
Amsterdam University Press
.
GLICK-SCHILLER
N.
CAGLAR
A.
(
2009
) ‘
Towards a Comparative Theory of Locality in Migration Studies: Migrant Incorporation and City Scale
’.
Journal of Ethnic and Migration Studies
 
35
(
2
):
177
202
.
HATZIPROKOPIOU
P.
TRIANDAFYLLIDOU
A.
(
2013
)
‘Governing Irregular Migration: States, Migrants and Intermediaries at the Age of Globalisation.’ Deliverable 1.1.: Concept Paper. Hellenic Foundation for European and Foreign Policy (ELIAMEP)
.
HELSINKI CITIZENS’ ASSEMBLY, REFUGEE ADVOCAY AND SUPPORT PROGRAM
(
2007
)
Unwelcome Guests: The Detention of Refugees in Turkey’s ‘Foreigners’ Guesthouses
 .
Istanbul
:
Helsinki Citizens’ Assembly
.
HESS
S.
(
2012
) ‘
De-Naturalising Transit Migration: Theory and Methods of an Ethnographic Regime Analysis
’.
Population, Space and Place
 
18
:
428
440
.
IÇDUYGU
A.
(
2003
)
Irregular Migration in Turkey
 ,
IOM Migration Research Series No. 12. Geneva: IOM Publications
.
IÇDUYGU
A.
ERDER
S.
GENCKAYA
O. F.
(
2014
)
Türkiye’nin Uluslararası Göç Politikaları, 1923–2023: Ulus-devlet Oluşumundan Ulus-ötesi Dönüşümlere [Turkey’s International Migration Policies, 1923–2023: From the Formation of the Nation-State to Transnational Transformations]
 .
Istanbul
:
Mirekoc
.
IKIZOGLU ERENSU
A.
(
2014
)
‘Suriyeli Linçlerine Karşı Misafirin Adını Koymak’ [Countering Syrian Lynchings By Naming the Guest].
.
INTERNATIONAL ORGANISATION FOR MIGRATION (IOM)
(
1995
)
Transit Migration in Turkey
 .
Budapest
:
IOM Migration Information Programme
.
KALIR
B.
SUR
M.
VAN SCHENDEL
W.
(
2012
) ‘
Mobile Practices and Regimes of Permissiveness
’. In
Kalir
B.
Sur
M.
(eds)
Transnational Flows and Permissive Mobilities: Ethnographies of Human Mobility in Asia
 .
Amsterdam
:
Amsterdam University Press
, pp.
11
27
.
KARCI KORFALI
D.
SENOL SERT
D.
ACAR
T.
(
2014
) ‘
Mapping and Analysis of Types of Migration from CEE countries: Country Report Turkey’. Imagination Working Paper No. 4
. .
KAŞLI
Z.
(
2011
)
‘Criminalising and Victimising the Migrant: Reflections on the UN Protocol and UNHCR's Position against Smuggling
’.
Oxford Monitor of Forced Migration
 
1
(
2
):
64
68
.
KIMBALL
A.
(
2007
)
The Transit State: A Comparative Analysis of Mexican and Moroccan Immigration Policies
 ,
Working Paper 150. San Diego: Center for Comparative Immigration Studies, University of California
.
MARCONI
G.
(
2008
) ‘
Transit Cities in Transit Countries: Steering the Consequences of US and EU Closed Doors Policies
’,
paper presented at the workshop ‘Narratives of Migration Management and Cooperation with Countries of Origin and Transit’, Sussex, 18–19 September 2008
.
MIGREUROP.
(
2009
) ‘
Europe's Murderous Borders
’. .
PAPADOPOULOU-KOURKOULA
A.
(
2008
)
Transit Migration: The Missing Link between Emigration and Settlement
 .
New York
:
Palgrave Macmillan
.
PROASYL.
(
2012
). ‘‘
Walls of Shame’ Accounts from the Inside: The Detention Centers in Evros
.’ .
PROASYL.
(
2013
) ‘
Pushed back – Systematic human rights violations against refugees in the Aegean sea and the Greek-Turkish land border
.’ .
SAMPSON
R. C.
GIFFORD
S. M.
TAYLOR
S.
(
2016
) ‘
The Myth of Transit: The Making of a Life by Asylum-Seekers and Refugees in Indonesia
’.
Journal of Ethnic and Migration Studies
 
42
:
1135
1152
.
TOPLUM VE HUKUK ARAŞTIRMALARI VAKFI (TOHAV)
(
2014
)
Göç Analiz Raporu [Migration Analysis Report]
 .
Istanbul
:
Toplum ve Hukuk Araştırrmaları Vakfı Yayinlari
.
TSIANOS
V.
HESS
S.
KARAKAYALI
S.
(
2009
) ‘
Transnational Migration: Theory and Method of an Ethnographic Analysis of Border Regimes’. Sussex Center for Migration Research Working Paper No: 55
.
Available at <http://www.sussex.ac.uk/migration/documents/mwp55.pdf> (Accessed September 2016)
.
TSIANOS
V.
KARAKAYALI
S.
(
2010
) ‘
Transnational Migration and the Emergence of the European Border Regime
’.
European Journal of Social Theory
 ,
13
(3):
373
387
.
TURKISH PARLIAMENT HUMAN RIGHTS COMMISSION.
(
2012
)
‘Report on the Deportation Centres in Edirne, Istanbul and Kirklareli' [Türkiye Büyük Millet Meclisi İnsan Haklarını İnceleme Komisyonu “Edirne, İstanbul ve Kırklareli İllerinde Bulunan Geri Gönderme Merkezleri Hakkinda İnceleme Raporu.” 24. Dönem 3. Yasama Yılı 2012
. .
VAN MUNSTER
R.
STERKX
S.
(
2006
) ‘
Governing Mobility: The Externalization of European Migration Policy and the Boundaries of the European Union
’. In
Holzhacker
R.
Haverland
M.
(eds)
European Research Reloaded: Cooperation and Integration among Europeanized States
 .
Dordrecht
:
Springer
, pp.
229
250
.
WALKER
R. B. J.
(
2000
) ‘
Europe Is Not Where It Is Supposed to Be
’. In
Kelstrup
M.
Williams
M.C.
(eds)
International Relations Theory and the Politics of European Integration
 .
London
:
Routledge
, pp.
14
32
.
WALTERS
W.
(
2002
) ‘
Mapping Schengenland: Denaturalizing the Border
.
Society and Space
 
20
(
5
):
561
580
.