Boundaries and Belonging Among Settled Minorities and Refugees in Bulgaria

Abstract The importance of settled minorities for facilitating refugee belonging is seldom discussed in research on refugee integration. Drawing on scholarship on belonging, boundary-making, and bordering, this study investigates how boundaries are drawn between settled minorities and refugees in Bulgaria. Based on interviews with integration workers and organizations of settled minorities in a state with the largest historically present Muslim minority in the EU, an Arabic-speaking diaspora settled decades ago, and with minimal state involvement in refugee integration, the study shows how spatial, linguistic, and religious boundaries separate settled minorities from newly arrived refugees. Arabic-speaking diasporas are nevertheless witnessed to overcome the boundaries through geographical proximity, a shared language, and shared countries of origin, whereby they have functioned as facilitators of refugee belonging and inclusion. Furthermore, Muslim institutions led by Bulgarian Turks have functioned as spaces for refugee belonging. The study finds that settled minority communities have, despite multiple boundaries and some assimilatory discourses, contributed to refugee belonging in ways that in part has compensated for the state absence. The study calls for further research investigating the role of settled minorities in inclusionary processes in society.


Introduction
Imagined borders between East and West, Europe and the "Orient," and Christianity and Islam strongly inform ideas on nationhood and belonging in Bulgaria. Constituting half of the EU's outer border with Turkey, and with around 9 percent of the population belonging to the EU's largest historically present Muslim, Turkish-speaking minority, Bulgarian national narratives are shaped in relation to complex boundary-making processes involving the Ottoman past. The EU's failure to provide safe pathways for refugees 1 fleeing the numerous conflicts in Turkey's neighboring countries made Bulgaria part of the politicized game to hinder intra-EU refugee mobility during the 2010s. Since the Dublin III Regulation 2 makes the first EU country of arrival responsible for the asylum process, many, mostly Muslim, refugees found themselves in involuntary transit in Bulgaria after crossing the Turkey-EU border.
As well as being the poorest EU state 3 with hostile narratives on the "Orient" shaped by a process of post-Ottoman nation-building, Bulgaria also has a historical, institutionalized presence of Islam in the form of state mosques and muftis' offices. 4 In addition, it is host to an Arabic-speaking diaspora 5 of different faiths, originating mostly in Lebanon and Syria, that settled in Bulgaria following participation in study programs of communist fraternity between 1946 and 1989 (Zhelyazkova 2004). The predominantly Muslim refugees who arrived during the past decade have provided yet another layer to existing struggles over nationhood in Bulgaria, further politicizing national narratives on Islam, the Turkish border, and Europeanness.
By drawing from scholarship on belonging, boundary-making, and bordering, this study deepens our understandings on the role of settled minorities in the politics of belonging that relates to refugees. Based on interviews with integration workers and Muslim and Turkish organizations in Bulgaria, the study aims to explore how settled minorities were part of shaping the belonging of refugees who arrived in Bulgaria between 2013 and 2019. It asks: How are spatial, linguistic, and religious boundaries drawn between settled minorities and new refugees? The study unpacks how practices of refugee reception and integration relate to belonging in a country where, on the one hand, the state's involvement in refugee integration policy is deemed insufficient (Nancheva 2015), and, on the other hand, state Muslim institutions led by a Turkish-speaking minority (Troeva and Mancheva 2011, 156) have contributed to integration practices. By analytically combining the position of "old" minorities in Bulgaria with policies and practices of contemporary refugee integration, this article sheds light on how contestations of nation-building, boundary-making, and belonging related to contemporary migration are rooted in the past.
The study first presents an overview of scholarly debates on belonging, boundary-making, and bordering in contexts of migration, secondly, discusses boundary-making in the Bulgarian context from a historical and contemporary perspective, and thirdly, outlines its methodology. Fourthly, the empirical analysis shows how spatial, linguistic, and religious boundaries related to majority nationbuilding keep settled minorities separated from recent migrants, but also how marginalized minorities do take part in important inclusionary processes, shaping belonging in an era where minorities and migrants face increasingly restrictive discourses. Finally, its findings are concluded, showing how contemporary migrations in Europe closely connect with past and present nationbuilding processes and how integration policy is both shaped by and shapes existing politics of belonging.

Settled Minorities and the Politics of Belonging
The awarding of national belonging to majorities, minorities, and migrants is shaped by processes of boundary-making. How boundaries are drawnincluding those between -Europeans and "others," national and non-national groups, can be captured by investigating the dynamic relationality and processual nature of group and identity construction (Lamont and Molnár 2002;Barth 1969). What has been characterized as "the politics of belonging" describes how the inclusion and exclusion of certain people is justified and resisted by those who hold power over granting belonging (Antonsisch 2010;Yuval-Davis 2011, 18). These boundary processes, where parts of the population determine who has the right to belong, often either exclude or differentially include groups such as Black, Roma, or Muslim minorities in subordinate ways (Anthias 2021, 146). The exclusions, which can be connected to racial ideologies promoting whiteness (Imre 2005), are typically based on national myths of common descent, culture, religion, or language that are to be protected against undesirable minority cultures (Yuval-Davis 2011, 20;Anthias 2021, 158). Indeed, nation-states have employed a number of tactics in order to protect the hegemonic culture, ranging from assimilation, deportation, and displacement to conditional accommodation of minorities (Mylonas 2013, 21-22). In current politics of migration control, of which integration forms a key part, such tactics are widely used to justify who is awarded belonging and who is not. Yuval-Davis (2011, 21) identifies racialized requisites of belonging that relate to origin and race as less permeable than those that are more open for assimilatory identification such as language, culture, and, in part, religion. Whereas the attributes are closely interrelated, most discourses on migrant and minority (non)belonging in Europe explicitly circulate around religion, language, and culture.
The construction of linguistic, religious, and cultural belongingness is multifaceted and imbued with complexity. In the context of migration, scholarship has shown how specific groups are awarded different degrees of belonging, and that boundaries can shift depending on perceptions of linguistic, cultural, and religious proximity (Elchinova 2005;Castles 2002;Rottmann and Kaya 2021). Russian authorities portrayed the Ukrainian forced migrants who arrived between 2014 and 2016 as compatriots, which partly facilitated their settlement process (Myhre 2018). In Turkey, cultural affinity related to perceived religious similarities and shared Ottoman heritage have been shown to enhance the belonging of Syrian refugees (Kaya 2017). Such expressions of belonging can be tied to "space-belongingness," entailing a personal, intimate, and homely feeling of belonging connected to a specific place (Antonsisch 2010). In addition, scholarship has demonstrated how settled minorities perform important roles for facilitating refugee belonging. In the USA, Muslim and Catholic institutions have been argued to function for bonding and bridging purposes, strengthening connections not only between refugees and previous migrant groups but also with wider society (Allen 2010). The concept of integration as presupposing settlement into whitemajority communities has furthermore been criticized from the perspective of superdiversity (Grzymala-Kaslowska and Phillimore 2018), showing how friendships with co-ethnics constitute a social capital that promotes the integration and belonging of recently arrived migrants in the UK (Wessendorf and Phillimore 2019, 124). Hence, during initial stages of settlement, perceived cultural proximity or contact with previously settled co-ethnics may facilitate the sense of belonging of refugees and erase boundaries between settled groups and newly arrived refugees.
The softer boundary-making processes that take place between communities are crucial for the politics of drawing physical borders (Eder 2006, 269;Van Houtum 2005, 674). Especially in the context of migration, spatial dimensions of bordering are determined by questions such as who has the power to control whose mobility and how and why specific borders are constructed and controlled (Yuval-Davis, Wemyss, and Cassidy 2019, 4). In such processes, racial, linguistic, and religious boundaries are at the forefront. The complex post-imperial nation-building of Eastern Europe make questions of bordering particularly poignant due to its imagined position constituting the border between the Orient and the Occident, while being characterized by "several layers of imperial, national, regional, and linguistic borderlines" (Kulawik 2019). In order to assert the Europeanness of the region, an unspoken and invisibilized insistence on whiteness (Imre 2005, 82) coexists with Islam and the Ottoman Empire as boundary markers for determining what Europe is and what it is not, where its borders pass, and who belongs within them (Boatcă 2015). Contemporary European discourses on multiculturalism and integration have found vigor in the historical boundary, widely portraying Islam as incompatible with Europeanness (Lentin and Titley 2011). In Western Europe, Muslim communities are generally understood as newcomers whose belonging in the perceivedly Christian nations is made conditional and is used for legitimizing stricter border controls. Rather than seen as facilitating belonging, "Islam is analyzed as a barrier or a challenge to integration and a source of conflict with mainstream institutions and practices" (Foner and Alba 2008, 368). Such discourses, which upon closer investigation conceal more than merely religious boundaries, take contextual expressions that are rooted in deep historical contestations in post-Ottoman, post-communist Bulgaria.

Boundary Processes Toward Minorities in Bulgaria
The role of migration in the politics of belonging in Bulgaria is shaped by the nation's charged perceptions of Muslimness and the Middle East. Whereas Muslims were a privileged group during the Ottoman era, which lasted between 1396 and 1878, they were excluded from the modernizing nation-building that followed Bulgarian independence (Neuburger 2004, 3). The reconstruction of the religious boundary between Christians and Muslims along linguistic lines during post-Ottoman nationalization (Todorova 1998, 476;Köksal, 2006, 507) produced a Muslim, Turkish-speaking "other" in the then young nation-state, where Bulgarian was the official language and Orthodox Christianity the state religion. While the new focus on language did not erase the religious boundary between Christians and Muslims (Todorova 1997, 177), it added a central identity marker that came to be targeted politically alongside religion. During the Bulgarian Communist Party (BKP) rule between 1946 and 1989, Muslim and Turkish-speaking minorities were targeted by assimilatory policies. The five-year "Revival Process" at the end of BKP rule resulted in one million Muslim names being Bulgarianized by force over the course of a few weeks in 1984, a ban of Islamic practices, and the 1989 forced expulsion of more than 300,000 Bulgarian Turks, which has been characterized as ethnic cleansing (Kamusella 2018;Dimitrov 2000;Mahon 1999). Following the fall of the BKP regime, around half of the expulsed Bulgarian citizens were able to return, whereas others settled permanently in Turkey or other countries. The expulsion is a clear example of when "some citizens are treated as 'internal others,' and may be subject to processes whereby they cease being citizens" (Anthias 2021, 150) through concrete acts of violent bordering.
Today, the heterogeneous Muslim national minority includes Bulgarian Turks, (Turkish-speaking) Roma, and Bulgarian-speaking Pomaks. Bulgarian Turks form the largest linguistic minority in Bulgaria, greatly overlapping with the Muslim minority. Many Roma are both Turkish speakers and Muslims, and even though many Turkish-speaking Roma self-identify as Bulgarian Turks, the identity is rarely externally accepted by Bulgarian Turkish communities or majority Bulgarians who view them as Roma (Ladányi and Szelényi 2001, 86). Hence, linguistic, religious, and racial processes interact in the boundary constructions between minorities. With more than six hundred years passed since the arrival of the first Ottoman settlers, it has been claimed that Bulgarian Muslims should be considered "native" populations (Zhelyazkova 2001, 285). The statement is motivated by the uninterrupted presence of Turkish-speakers and Muslims that precedes the establishment of the modern Bulgarian nation-state by centuries. The position of minorities is nevertheless still marginal: following Bulgaria's post-1989 market transition, Bulgarian Turks and Roma have for instance been disproportionately living under what Emigh, Fodor, and Szelényi (2001, 11) call racialized poverty. Bulgarians from "pockets of extreme poverty," who often are Turks or Roma, have been identified as overrepresented among the seasonal labor emigrants enabled by the EU membership of 2007 (Markova 2010, 11).
Bulgaria's EU membership made the country part of the core European institutional space, reshaping its mobility regimes. While its citizens within a few years gained access to free movement within the European Single Market, Bulgaria also became part of the Common European Asylum System (CEAS) 6 and began to control the EU's outer border. Following political unrest, which has been especially acute in a number of Turkey's neighboring countries, Bulgaria has gone from being mainly a country of emigration to a country of immigration too. Largely due to the Dublin Regime that hinders intra-EU refugee mobility, 91,214 persons applied for asylum in Bulgaria between 2013 and 2020, out of whom 26,458 were granted protection, 16,589 had their applications rejected, and more than half of the asylum cases, 46,446, were terminated (SAR, n.d.). In the absence of official numbers, interviewees in this study estimate that at most 1,000-2,000 of the refugees who arrived between 2013 and 2019 have stayed in Bulgaria. Indeed, partly due to its economical precarity, 7 Bulgaria can be described as a transit country where migrants are "expected and encouraged to migrate again" (Erolova 2017, 362). Among those who stayed, we find persons settling in Bulgaria after having received asylum and persons stuck in "grey zones" (Yuval-Davis, Wemyss, and Cassidy 2019) who are awaiting status, pending deportation, or have been deported back to Bulgaria from other EU countries through the Dublin system. Current restrictive policies towards refugees stand in contrast to earlier Bulgarian state policy that actively welcomed students from the very same countries whose citizens are today seeking refuge. An Arabic-speaking diaspora of different faiths originating mostly in Lebanon and Syria settled in Bulgaria between 1946 and 1989, having participated in student programs with roots in the policy of "socialist solidarity with the fraternal peoples" (Zhelyazkova 2004, 29). The programs were not seen as immigration in the eyes of the state (Zhelyazkova 2015). The construction of specific borders varies based on time and in relation to whom, how, and why inclusions or exclusions take place (Yuval-Davis, Wemyss, and Cassidy 2019, 4). Hence, persons from the same background, class, and country of origin may have highly different paths of migration during different periods of time. At the same time, old and new migrations, settled and new minorities are discursively connected in multiple ways. Krasteva (2020, 686) calls the expulsion of Bulgarian Turks in 1989 and the refugee "crisis" of 2013, which followed the upsurge of Syrians fleeing the civil war to Europe, the two major instances of rebordering in Bulgaria, stating that they are used in similar, nationalist ways in politics even though their connections are silenced. As Anthias (2021, 150) states, "we cannot draw sharp boundaries between migrants and racialized groups, since they are intertwined, and often experience similar regulatory regimes and ordering." Indeed, newly arrived refugees have, according to Erolova (2017), been targeted by variants of long-present negative racializing discourses where Turks are blamed for being "carriers of Islam" and Roma are associated with criminality and bad manners. Prejudices facing Arab and Black immigrants are similar to the levels of prejudice directed toward Roma (Zhelyazkova 2015, 586), indicating how racial logics directly inform the discrimination and othering of minorities. Besides being targeted by discourses usually directed at "old" minorities, Muslim refugees have also faced new negative stereotypes that can be linked to forms of Islamophobia that spread following 9/11, which frame them as security threats and terrorists (Erolova 2017, 375).
The relations between new and old minorities also take more inclusionary dimensions. Zhelyazkova (2004, 11) has found that Arabic-speaking immigrants have previously tended to settle in Turkish-speaking Roma Muslim neighborhoods due to a perceived absence of prejudice and an ease in recruiting workers, rather than in Turkish or Muslim-majority towns. Many mosques built during the Ottoman era, led by Bulgarian Turkish imams, have furthermore become meeting places for Muslims of different origins for religious discussions and Friday prayer (Zhelyazkova 2004, 12). Minorities then appear to play a part in coproducing belonging for newly arrived persons in Bulgariaa state that has erased and externalized Islam and its Ottoman heritage from understandings of Bulgarianness and that provides scarce integration support for refugees.

Methodological Notes
In an endeavor to investigate connections and separations between recent asylum migration, historically present minorities, and nationhood in Bulgaria, a total of two months of field research was carried out between March and July 2019 in different parts of Bulgaria: in Sofia, where most integration organizations and refugees are located, in Northeastern Bulgaria, which has a significant Bulgarian Turkish presence, and in Southeastern Bulgaria, with its Bulgarian Turkish-dominated regions and multiple refugee camps. As the main concern of the study regards official, organized integration practices, perceptions, and testimonies, interviews were carried out with state, NGO, and Muslim representatives working with issues related to refugee integration or the linguistic or religious maintenance of Turkish communities. They made it possible to "pursue questions that are difficult to locate in documentary sources or everyday interactions and to explore such questions in intricate detail" (Soss 2006, 141). Fifteen semi-structured interviews were conducted, of which eight were with organizations working with refugee integration (including one group interview with four persons representing the State Agency for Refugees), and seven with Turkish and/or Muslim organizationsof which three have a cultural-linguistic focus (in Northeastern and Southeastern Bulgaria) and four are religious organizations operating under the state, namely the Grand Mufti's Office in Sofia and three Regional Mufti's Offices in Southeastern Bulgaria. Hence, the knowledge claims of the study are not derived from first-hand experiences described by refugees, but stem from testimonies by official organizations as actors working with the settled minorities or as actors shaping refugee belonging through integration practices.
All interviewees were promised confidentiality, which is why smaller cities are not named, and the ethnic background of non-Turkish interviewees is not disclosed, even though we find both Arabic-speakers and majority Bulgarians among the integration workers. Questions were asked with the aim of acquiring knowledge of the situation of minorities, understanding the challenges of integration in Bulgaria, the consequences of transit migration and emigration for integration and belonging, how the Ottoman past is perceived in contemporary Bulgaria, and what connections exist between Turkish/Muslim communities and refugees in relation to official, state-funded integration and unofficial integration taking place between private persons. The interviews, which were carried out in Turkish (with Bulgarian Turkish and Muslim authorities) or English (with integration workers) lasted between 30 minutes and 2 hours, the most common duration being 1 hour. They were fully transcribed, and quotes in Turkish have been translated to English by the author. Furthermore, a few selected newspaper articles 8 featuring connections between refugee reception and nation-building or minorities in Bulgaria are used, for "providing contemporaneous accounts of key actors and their views along with more general sentiment at the time, especially for periods when the researcher was not or could not be present" (Yanow 2007, 114).
The material has been analyzed using an interpretive framework (Yanow 2007) with the constant aim of exploring how boundaries are drawn between settled minorities and recently arrived refugees. Concretely, it was done by identifying how the state, (international) non-governmental organizations [(I)NGOs] working with integration, and minority institutions narrated boundaries and (non)belonging between different groups within the nation-state, with a focus on political identity markers such as language, culture, and religion, but also conditions created by the asylum and border system in the EU.
In the following, the empirical material is analyzed in four sections, manifesting firstly how spatial, linguistic, and religious boundaries come into expression between settled minorities and refugees, and, finally, how minorities have taken part in crucial inclusionary processes for refugee belonging.

"Most People Leave Bulgaria": Spatial Boundaries Between New and Settled Minorities
The connections between settled minorities and new migrants identified in prior research were described by interviewees in this study as scarce. The lack of physical contact between Turkishspeakers and refugees was reported as one main reason for the separations. A rigid spatial boundary was identified between refugees, who are mainly located in bigger cities, and Turkish-speaking minorities, who mostly reside in rural areas. Even the refugee camps in the vicinity of Turkishspeaking regions were not said to contribute to connections between settled minorities and refugees, due to their spatial isolation from the surrounding society and the way the asylum system prevents long-term contact and fuels mobility. The testimony of an interviewee located in Sofia, 300 km from the Turkish border, illustrates how peripheral Turkish-dominant regions are perceived from the capital, as the center where most activities for refugee integration are carried out: These [Turkish] communities are normally very concentrated in some regions of the country. The old Muslim communities, they are high up in the mountains and they're pretty isolated. The Turkish community is mostly on the border with Turkey … But I don't know if they're engaged in some way like a relationship there because most of the inhabitants of the camp, which is on the Southeastern border, are Afghans. So maybe that's also one reason. I mean even though they're Muslims as well. (Representative of social enterprise, Sofia) The interviewee further described how most (Sunni) Syrians had left the camp and the country as they received asylum quickly, whereas those left in the camps (Shi'ite Afghans) would not have as much in common with the local Muslims as Syrians would. Transit migration makes the presence in the regions where Turkish is spoken short-term: whereas many of the Afghans, who are rarely awarded asylum and thereby will not be allowed to settle in Bulgaria, are still present in the camps, most of those who received asylum and the right to settle in the country had left the region and usually Bulgaria, too.
Transit migration, as something preventing contact, integration, and long-term connections, was connected by many interviewees to the meager support provided to refugees in Bulgaria. Some even claimed that the seeming passivity of the state in providing integration support was actually a conscious policy to encourage transit migration. Syrians in particular were identified as a group who were granted asylum with the purpose of enabling intra-EU mobility, and thereby making it possible for refugees to leave Bulgaria within legalized frameworks of mobility: And I'm sure that it was by intention that everyone from Syria received refugee status because it was clear that they would not stay, and they would not be interested in integrating. They were two or three consecutive years, the Syrians didn't allow the kids to go to school. They said we are here for a few months and we leave. (Representative of INGO, Sofia) The transit that precedes the emigration of refugees creates a state of limbo, of temporariness, and prevents settlement, in ways that largely characterize contemporary border politics (Yuval-Davis, Wemyss, and Cassidy 2019, 4). The emigration that follows the period of limbo was placed by interviewees on the same continuum as the emigration of young Bulgarians, especially Bulgarian minorities from less prosperous regions, to other EU countries. Interviewees saw this emigration as something with a deep societal impact and as a factor that weakens the Turkish and Muslim communities: So there are many who leave of course; actually there is decline in the number of young people. Many go to work in Europe. Out of the visitors of our mosque, many left for Europe and the size of the mosque congregation has thereby diminished. (Representative of Regional Mufti's office, Southeastern Bulgaria) Rather than a question of specifically refugee preferences, the mobility toward more prosperous EU countries can thus be seen as a structure that permeates Bulgarian society, adding a layer of temporariness and transit to it. As a phenomenon, this mobility is multifaceted, provoking emotional reactions that put pressure not only on minorities struggling against a decline in their numbers but also on majority-minority relations: Now the Bulgarian population is also decreasing. Bulgarians are afraid of it. They say that if we get fewer and the Turks get more, the border is also close to here; there is a fear that because of that some people would claim [the land]. (Representative of Bulgarian Turkish cultural organization, Southeastern Bulgaria) This fear, which stands in contrast to the observation of a weakening Muslim congregation, is directed at the integrity of the physical state border between Bulgaria and Turkey itself. The feeling of threat associated with majority emigration can be connected to instances where internal "others" become threatening through "periods of instability and crisis, when the territorial and symbolic boundaries of the ingroup are unstable and/or unclear" (Triandafyllidou 1998, 603). Fears associated with emigration-related phenomena such as depopulation and brain drain (Markova 2010) are, in this case, connected to separatism and imagined challenges with maintaining the physical national border. Such reactions can be connected to nation-building in post-communist Eastern Europe, where national core groups highly coincide with statehood and national sovereignty (Pogonyi 2017). As a consequence, Bulgarian Turks, who were excluded from the core nation, have been constructed as a national enemy (Elchinova 2005, 95).
The demographic question, where a boundary is drawn between "belonging" majority Bulgarians and settled minorities, with loyalties questioned by some, also has a dimension connected to refugee migration. The Bulgarian Orthodox Church, accused of passivity in the wake of refugee arrivals, has managed to connect such a discourse to refugees. Its Holy Synod issued a statement in 2015, asking the government not to admit any more refugees, since it would threaten Christianity and the "ethnic balance" of the Bulgarian state: Also, [it poses] the question what spiritual context, what spiritual environment will the Orthodox Bulgarian people live in if this influx continues to the extent that it shifts the existing ethnic balance in the territory of our Fatherland Bulgaria that God determined for our Orthodox people to inhabit. (Sofia News Agency, September 25, 2015) The remarkable wording used by the state church not only draws a near-impermeable boundary between the Bulgarian Christians and Muslims but also falls directly in line with violent acts of bordering that Krasteva (2020) has analyzed as connecting the expulsion of 1989 and current refugee politics. They can also be connected to political discourses in Europe that give preference to Christian migrants over Muslims, who are instead viewed as a threat (Hafez 2015).
As the findings demonstrate, spatial separations fueled by emigration shape boundaries between refugees, minority Bulgarians, and majority Bulgarians. Rather than viewing emigration as an issue that deeply impacts all communities in Bulgaria and immigration as something that balances the effects of emigration, majority discourses have mobilized demographic fears against the Bulgarian Turkish national minority and refugees, portraying Muslims as a threat to Bulgaria's national unity and territorial integrity.
"Our Turks Do Not Know Their Language Well": Linguistic Boundaries Between Bulgarian Turks and Refugees Language has been identified as an important factor for belonging (Antonsisch 2010, 648) and, in post-Ottoman nation-building, it also became a main identity marker alongside religion (Todorova 1997, 177). The Turkish language has held a politically controversial position in Bulgaria over the years, culminating during the "Revival Process" when speaking Turkish was policed even inside homes. Today, Bulgarian is the only official language, public schooling is conducted exclusively in Bulgarian, and teachers can be fined for speaking in Turkish to pupils. The monoethnic and monolingual nature of Bulgarian nation-building in the post-communist period is reflected in the lack of constitutional recognition of the Turkish minority and a political consensus among the mainstream Bulgarian parties to exclude the Turkish minority from nation-building (Aktürk and Lika 2020). Even though the official policies of assimilation have ended in Bulgaria, interviewees connected the present linguistic decline to the idea of monolingualism. One interviewee calls the present situation a "secret" assimilation: So there is no ban but there is also no encouragement for the [Turkish] mother tongue. In order for a pupil to study their mother tongue, one has to request it each year; there are no teachers, there is no book. The Ministry [of Education] does not issue books. They study from the old books from 1993. So Turks from here did everything for mother tongue studies. We have the right, yet it is not taught. Let me give just an example: in 1994, 114,000 pupils studied their mother tongue in schools. Now it has gone down to 5,000. (Representative of cultural organization, Southeastern Bulgaria) The impression of a weakening position for the Turkish language was widely shared by interviewees working with the Turkish-speaking community and connected to the past policies of assimilation. Another interviewee nevertheless had some hope for an improvement: Unfortunately, our Turks do not know their language well. They don't make an effort to read. From what I can see. But there is of course an improvement too. İnşallah, it will be better. We exited communism 30 years ago. Recently, Muslims started to recover. (Representative of regional Mufti's office in Southeastern Bulgaria) Despite the recovery perceived by the interviewee, they further stated that a majority Bulgarian studying Turkish would be a rare exception. Rather, the Turkish language is confined to the Turkish speakers and is separated from the majority, which does not learn or engage with the language. For minorities, knowledge of Bulgarian was nevertheless seen as self-evident even in Turkish-dominated areas, as described by another interviewee: Of course it is important to know the language of the homeland one lives in. We know that. We know Bulgarian. Besides that, it is a good thing to learn Western languages. But our Bulgarian minorities have a good sideeach citizen knows at least two languages: Turkish and Bulgarian. Bulgarian is compulsory. (Representative of cultural organization, Southeastern Bulgaria) The act of emphasizing the mandatory knowledge of Bulgarian while pointing out advantages of multilingualism may be interpreted as an expression of how narrow the space provided for minority cultural claims is (see Aktürk and Lika 2020), as well as a signal of what it takes to belong in Bulgaria namely mastering the Bulgarian language.
The weak position of Turkish and the strong position of Bulgarian means that refugees too face an officially Bulgarian-speaking society. Posing the question of whether Bulgaria does not also have Turkish speakers to an interviewee who works with integration in an INGO led to a further confirmation of the marginal position of the language: Yes, it has, but still the one official language in Bulgaria is Bulgarian. And the main language that you can take your exams for a degree in is Bulgarian. So there is no doubt about that. But you are right, some of the schools in Bulgaria have Turkish language courses as a subject like Dutch or French. (Representative of INGO, Sofia) Rather than being seen as a domestic language, Turkish is placed by the interviewee in the category of foreign languages. In a similar manner to the representative of the Turkish cultural organization, the interviewee mentions Turkish in relation to Western languages.
The Bulgarian language also has a strong position in the Banya Bashi Mosque, which, as the only functioning mosque in Sofia, has become a meeting space between the Turkish-speaking minority, Muslim refugees, and other Muslims who reside in the capital. The position of Turkish was stated to be relatively marginal in the mosque too: Generally, the main official language is Bulgarian, so generally everyone starts to speak Bulgarian; then, if we feel that, for example, we could speak in Turkish, we speak the one that is easier for us. (Representative at Grand Mufti's Office, Sofia) Even though virtually all refugees in recent years arrived in Bulgaria through Turkey, where they had spent varying amounts of time, the interviewee stated that very few had learned Turkish: Some of them knew some things from the time they had spent a few months in Turkey, but they knew very little. Generally, they spoke Arabic. (Representative at Grand Mufti's Office, Sofia) The Turkish language is then, at best, marginal in the lives of refugees, which can be expected given the weak societal position of Turkish in Bulgariabut also given the struggles many Syrian refugees experienced with learning Turkish in Turkey (Kaya 2017, 341), where the language is in dominant majority position.
The interviewees then paint a picture where Turkish is a marginal language within the state, from which refugees are separated. Efforts for linguistic integration that refugees encounter are thereby clearly centered around the official language, Bulgarian. According to one interviewee, INGOs provide Bulgarian language courses when they have funding, to an insufficient extent and on an irregular basis: No Bulgarian language courses [are given by] the government; there are some classes beginning this year in the reception center but there is only one teacher or two, I don't know. Twice a week, which is not enough. (Representative of INGO, Sofia) As stated in the quote above, the linguistic dimension of nation-building achieved through integration is mostly left to non-state actors. The definition itself of what the target nation of integration should be is nevertheless formulated by the state. The scarce, state-provided courses offered by the State Agency for Refugees (SAR) at the Council of Ministers, commissioned to provide initial facilities for reception and registration to refugees, are described as "adaptation" rather than "integration." However, the courses clarify that the aim is to find a belonging in majority Bulgarianness: Adaptation is the first, the primary step of the integration process and during this period of adaptation the asylum seekers actually get aware of the [Bulgarian] language, the [Christian] religion here, the Bulgarian culture, the customs. And the integration phase is the [learning about] different government systems, the government in Bulgaria and how it functions, to understand it, who the President is, who the Prime Minister is … And when [refugees] start buying tickets for buses, this is integration. Because they know what they are supposed to do and what they are expected to do, like all the Bulgarian citizens. They know what they are expected to do. This is the main difference. (Representatives of SAR, Sofia) As can be expected for state institutions working with reception, the phase of adaptation is used for formulating "the" national linguistic, religious, and cultural identity markers of the state, namely Bulgarian and Christianity. Fourteen days after an asylum seeker gets their decision, they must leave the reception centers and independently enter the stage of "integration," as the state-provided courses thereby end.
Since Bulgarian language tuition is insufficient in relation to the need, refugees are confined to learning the majority language independently, which not everybody manages to do. One interviewee nevertheless shared an example of how, according to them, successful integration and belonging did not necessarily require a common language: I know refugees who are integrated perfectly in Bulgaria without Bulgarian language. They speak with neighbors, I don't know how. And sometimes they come [to the NGO and] … say "I have a Bulgarian neighbor, she is very poor, she has four kids, do you have something to help her with?" … [One Afghan woman] is about 80 years old, she cannot speak the Bulgarian language but … she has more friends than anybody where [they] live. … All the old [Bulgarian] women from the village they live in, they say "every morning we go to the Afghan family." She used to make them Afghan tea, black tea with milk, and they speak, we don't know how but they were all laughing and all. (Representative of refugee NGO, Sofia) The testimony turns around the preconception of migrants needing help from the settled populationindeed, many majority Bulgarians suffer from poverty, which here prompted a refugee to seek material aid for their neighbor. In addition, it shows an example of how language is only one factor in processes of belonging. Elchinova has shown how even a shared mother tongue does not erase "otherness"expulsed Bulgarian Turks were perceived in Turkey to be more "alien" by majority Turks than they had been by Christian Bulgarians in their places of origin (2005,107). Hence, in contrast to the testimony above on how boundaries can be erased despite the lack of a common language, we find indications of boundaries being drawn despite a shared mother tongue. Belonging is indeed more complex than mastering individual identity markers.
Between the Turkish language and refugees, interviewees identify a rigid boundary fueled by past and present majority nation-building. Even though the Bulgarian language is clearly stated as the refugees' target language, the state barely supports language acquisition. Rather than an endorsement to live in another language than Bulgarian, which may be one of the consequences of the policy, it can be connected to the prior testimonies in which the low level of state support was seen as deliberately encouraging transit migration, thereby aiming at excluding refugees from the nation.
"We Drink Our Alcohol": Religious Boundaries Between Old and New Muslim Communities In contrast to the strict linguistic and spatial separations identified thus far, the production of boundaries in relation to religion take, in part, more multifaceted and complex expressions that vary between religious institutions, minority NGOs, and individuals. As with language, assimilation was nevertheless widely perceived to impact how Islam is practiced in Bulgaria: Also, those who we call Muslim in fact even eat pork, drink alcohol, don't go to the mosque; even if you tell them to, they still don't go, but for example he has a Muslim name and identifies as a Muslim. Judging by that, for example, there is no Muslimness in him … Assimilation lasted for 45-50 years so there is an effect, of course. (Representative at Grand Mufti's Office, Sofia) The Islam in Bulgaria, of course it has been affected. There are 1.5 million Muslims … but those taking that path, I mean those who follow Islam, the number is very low due to that effect. But 30 years have passed since the exit from communism. A development has started, slowly, slowly, an improvement has started. (Representative at Regional Mufti's office, Southeastern Bulgaria) The interviewees separated self-identified Bulgarian Muslims from those they perceived as true followers of the religion. The interviewee at the Regional Mufti's office described the increasing number of those they viewed as real Muslims as an improvement. Another interviewee, representing a Turkish cultural organization, was however content with current religious practices that they described as "democratic," as contrasted with perceived religious practices outside of Bulgaria: Generally, Turkey's Muslims, Bulgaria's Muslims, Europe's Muslims, perhaps there is 3-5 [extremists] among them, but generally we are democratic, there is no extremism in our Muslimness. We, for example, don't have headscarves, of course the elderly may wear a headscarf but those wearing black ones 9 like in Turkey we don't have … We drink our alcohol; when required, we naturally go to the mosque. When we go to the mosque, they say according to Islam it is like this, don't do like that. We don't have that; we are democratic. Here, girls walk around with uncovered hair in summer days. What will happen? Our men also don't go and look at them. Everyone is used to it; this is how it gets about. (Representative of cultural organization, Southeastern Bulgaria) Most Muslims in the Balkans and in Turkey follow the Hanafi school of Islam, often described as more flexible than the schools predominant in the Arab world (Öktem 2010, 15), which may in part explain the testimony above. Largely focusing on gendered aspects of clothing, the interviewee points out differences in the use of the veil in Turkey and in Bulgaria. The differences in clothing have historical, context-bound roots, which are exemplified, for instance, through the Western gaze of Bulgarian nation-building where the veil was seen as a symbol of oppression and framed as non-European (Neuburger 2004, 116-118). Regulating the use of the veil among women is indeed part of the crucial role women and gender relations play in nation-building (Yuval-Davis 2011, 125). Troeva and Mancheva describe how some Muslim migrant women in Bulgaria dressed the same way as in their country of origin while previously living in Turkey, but stopped wearing the veil in Bulgaria due to the negative attention it attracted from the surrounding society (Troeva and Mancheva 2011, 181-183). Indeed, one interviewee shared how a Muslim refugee woman was requested to remove her veil in order to obtain employment with a countryman who settled in Bulgaria decades ago: She was looking for a job and she wore a headscarf; he said to her "if you want to work with me in my shop you have to uncover that scarf" …. "I will give you a good job, good salary but without the scarf." She said, Bulgarians, they didn't ask me for this. (Representative of refugee NGO, Sofia) The interviewee explained that the same requirement of removing the headscarf had occurred in a medical practice run by a settled Arab Muslim, out of fear that Bulgarian customers would stop coming were they to receive veiled customers. Fears of being perceived as less Bulgarian, less integrated, and of becoming less accepted by majority Bulgarian society when associated with persons bearing visible religious markers such as the veil, seem to have led the Arab Muslim employer to reproduce discrimination against, in this case, a veiled Muslim woman. The conditional belonging of previously settled migrants extended to sending a signal to veiled women questioning their belonging in Bulgaria, in ways that majority Bulgarians would not, according to the testimony. Hence, settled minorities do take part in acts and discourses that can be deemed assimilatory.
Even though Bulgaria has an established, state-recognized Muslim community, the space of religious belonging is diminished by the legacy of assimilation, the minoritized position of Islam, anti-Muslim discourses, and the specific history and local habits of Balkan Islam. Öktem (2010, 17) states that "'[b]eing Muslim' is often a very local affair and only one of many identities, i.e. national, ethno-linguistic." The boundary drawn between religious habits perceived as different was indeed seen as one factor contributing to the limited interaction between Bulgarian Turks and refugee communitiesthe shared religious belonging being only one factor in the complex interplay of identities. As an example, Bulgarian Turks who settled in Turkey following the expulsion found that considerable differences existed in linguistic and religious practices between Turks in Bulgaria and those in Turkey (Elchinova 2005, 105), despite seemingly having a language and religion that coincide with the national core.
Notwithstanding the perceived differences in religious practices, connections were identified between historical Muslim institutions and newly arrived refugees. The Mufti's Office, led by Bulgarian Turks, was described by several interviewees as an important charity actor providing for refugee needs. In the wake of increased Syrian arrivals in 2013, media reported how the Mufti's Office provided iftar meals for breaking the fast during Ramadan, as well as material aid (Kırcaalihaber 2013). One of the initial campaigns launched by the Grand Mufti's Office named "Elini Uzat, Ensar Ol [Reach out your Hand, Be an Ansar]" used the religious rhetoric of ansar (helper), a call for solidarity between Muslims that was also used by Turkish state authorities in relation to Syrians 10 (Kaya 2017, 340). Beyond charitable actions, religious organizations were nevertheless perceived as having had only a small role in refugee reception and integration, from which the Orthodox state church was stated to be fully absent: We had in the crisis time, …the Protestant church, these strange religious formations like the Jehovah's witnesses and others who showed interest in supporting migrants and refugees, but then we heard that they attract them to convert. And of course, the mosque and the Muslims they helped of course. They helped the people, but we don't have any special relation. I mean they help individuals; we haven't heard they made big donations. They have made donations on occasions like Kurban Bayram [Feast of the sacrifice] something in the past, some food or something like that. But it is more an exception than a rule. (Representative of INGO, Sofia) In line with the Christian communities with a missionary agenda described by the interviewee, a refugee NGO representative had also observed missionary objectives when approaching a Muslim NGO for donations: They said to me: we are supporting the Bulgarian Muslims. I said that, out of all refugees, 80% are Muslims. They said we are supporting the Bulgarian Muslims … they want to gain more Bulgarian people to be Muslims. So because the refugees are all Muslims, they just don't need them to convert. (Representative of refugee NGO, Sofia) Warning about missionary charities following Wahhabism, the interviewee also points out that Roma communities living under poverty are particularly vulnerable to being targeted by missionary Salafists. Severe discrimination and poverty have been identified by researchers as factors increasing acceptance for Salafi mission (Öktem 2010, 18;Zhelyazkova 2015, 573). Just as in the case of the organization mentioned by the interviewee, which was later reported to authorities and banned, Öktem (2010, 43) has shown that such charities have not managed to impose their versions of Islam in the regions where "Muslimness" is deeply embedded with linguistic, national, and historical markers. Indeed, the encounters with non-state Muslim and Christian organizations with missionary agendas were marginal. It was rather the landmark Banya Bashi Mosque that became a visible location that many refugees turned to for help when first arriving in the capital.
Of course, they came, they came a lot; many Muslims also helped them. Some of them they could not help. Some stayed in the streets. It depended. But now there is not that view [of homeless refugees] like back in the days, thankfully. No, not now. Everyone disappeared somewhere, of course to Germany, Sweden … So we also did some things, from the part of the Mufti's Office, as an institution, food aid, aid with clothes, such things. (Representative of Grand Mufti's Office, Sofia) The mosque was seen to function as a space people in need could turn to for charity, with zakat, or alms-giving, indeed being one of the mandatory five pillars of Islam. While its role in distributing charity was widely acknowledged, one interviewee nevertheless questioned whether it provided a space for more long-term integration, claiming that the Turkish employees at the Grand Mufti's Office "hate Arabs": For this reason, there is a new mosque, new masjid; it is built by the Arab community in Sofia in Lyulin. The Arabs said that they are discriminated in these mosques; they discriminated the Arabs. (Representative of refugee NGO, Sofia) Boundaries drawn between groups practicing religion in the same institutional spaces have then, according to the quote above, taken a spatial, physical form. Anti-Arabic discourses are prevalent in Turkey (Kaya 2017, 351)discourses that can be expected to reach Bulgarian minorities, among other things, through Turkish media consumption.
Notwithstanding the exclusionary and seemingly rigid nature of some boundary-making processes, identities are fluid, situational, and contextually shaped. Previous research has described testimonies of Muslim migrants who, after settling in Bulgaria, began to visit churches and incorporated Christian elements into religious practices as part of local adaptation (Troeva and Mancheva 2011, 189). Christians have also been found to attend Sufi tekkes that function as places of worship for non-Sunni Bulgarian Muslims, such as Alevis (Zhelyazkova 2015, 592). Religious boundaries between minorities and majorities are then shown to be fluid, shaped by social situatedness, and renegotiable through migration.
The sometimes rigid boundaries identified between religious institutions and refugees are demonstrated to be surpassed on an individual level. Testimonies were shared of Bulgarian Muslims going beyond occasional charity by providing employment to refugees staying at camps who were in need of an income (Interview with representative of INGO, Southeastern Bulgaria). Furthermore, a newspaper describes how a Bulgarian Turkish businessman welcomed a Syrian family to his Turkish-majority village of birth. The family relocated and got employment and a home after having faced racism and exclusion in the Christian village they initially settled in (Kırcaalihaber 2017). The businessman shares his cosmopolitan philosophy of helping in the newspaper: I also assist at both church and mosque construction sites. Before anything else, we should not forget that we are human. When we see someone in need, we should not forget that it is our primary duty to help. If Allah gave to us, we should also give [to others]. (Kırcaalihaber, October 15, 2017) The businessman names a general desire to help people in need as his driving force rather than a special interest to help those who share his religion. The assumption that the Muslim family may be better received in his Muslim-majority home village than in the Christian village, where migration was politicized by local politicians, however, directed his actions. Other accounts of Bulgarian Turks, and also Pomaks helping refugees, were also shared. An interviewee working in proximity with refugee camps described for example how small Bulgarian Turkish organizations discretely donated clothes and utilities to refugee camps without seeking attention (Representative of INGO, Southeastern Bulgaria). A refugee NGO in Sofia, located far from the areas where most Bulgarian Muslims live, had contact with Pomaks living on the other side of the country who sent generous, good quality donations by post (Representative of refugee NGO, Sofia).
The religious boundaries identified between local religious practices and forms of Islam practiced elsewhere, and within local institutions, were perceived to contribute to the scarce contacts between old and new minorities. However, despite the linguistic, spatial, and religious boundaries that sometimes prevent contact between national minorities and newly arrived refugees, the communities have also been engaged in important actions facilitating refugee belonging, taking place through both religious organizations and individual efforts. The more profound dimensions of facilitating refugee belonging were nevertheless connected above all to previously settled, Arabicspeaking diasporas.
"They Know the Situation": Settled Arab Diasporas as Facilitating Belonging The connections between Arabic-speaking diasporas and contemporary refugee migration were perceived to be stronger than those between migrants and Turkish-speaking national minorities. The bond can be explained in terms of the experience of prior migration and a shared language but also by geographic proximity, as both diasporas and refugees mainly reside in the capital or other bigger cities.
The fraternal country immigration during the communist era that took place mainly from the Near and Middle East can also be traced to institutional pathways that shape contemporary refugee integration. The Bulgarian Red Cross started their first program of integration in 1992-1993 upon the arrival of Afghan refugees who had previously studied in Bulgaria. While the former students were familiar with local society, the aim of the program was to familiarize their family members with Bulgaria. Mintchev has observed how still in 1999, most of the refugees and asylum-seekers are from Afghanistan, Iran, Iraq, former Yugoslavia, Armenia and Ethiopia. Many are former students or graduates in Bulgaria who have not been able to return because of political changes in their home countries. (Mintchev 1999, 143) Almost two decades later, most of the former students who aspired to seek refuge in Bulgaria had already settled there. Now, most asylum seekers instead lack prior connections to Bulgaria and originally aimed to settle in other EU countries, where they possibly had prior connections through family and friends who had taken up residence in these countries earlier.
After the first programs that were adapted to the former students' families were completed, other integration programs were put in place. In the wake of the increase of refugees in need of support in 2013, the state nevertheless pulled out of the field of integration, ending the funded national integration program (Interview with INGO representative, Sofia). However, the established Arabicspeaking diasporas were deemed by several interviewees to be highly valuable for refugee reception: So the others are businessmen and, when we faced the crisis situation in 2013-2014-2015, those people were really helpful. They wanted to help a lot. They donated a lot. The doctors were the first to offer medical assistance in the reception facilities where there was nothing in place etcetera. (Representative of INGO, Sofia) Even though a range of volunteers mobilized all over Europe during the years preceding the increased border restrictions, the significance of such efforts is accentuated in states where the humanitarian needs are extensive due to scarce state support.
Another interviewee described how, after the initial humanitarian crisis support, diasporas turned to long-term assistance by, for example, providing employment. The interviewee found the experiences and insights gathered over a long presence in Bulgaria, often living with majority Bulgarian spouses, to be important in facilitating integration: There are Arab communities who are living in Bulgaria since more than 50 years, they are integrated so well with the Bulgarian. So they are advising the newcomers to get integrated with the locals. They ask them not to be the same way as in their countries. No. They are open people, they want you to know, because they have children and are living in Bulgaria so they know the situation. (Representative of refugee NGO, Sofia) The interviewee had witnessed members of the diasporic communities encouraging newcomers to start following habits deemed as (majority) Bulgarian and relinquishing habits from their home countries. Such acts reflect societal dynamics that, to an extent, could be connected to previously discussed assimilatory ideas. At the same time, minority belongings are awarded space in the public in Bulgaria. Historical Muslim institutions, especially the Banya Bashi Mosque, were observed to contribute to the spatial belonging of Arabic-speaking communities in a formerly Jewish neighborhood in central Sofia: This [neighborhood] is where Muslims normally concentrate and Arabs as well because it's like the Arab street with a lot of shops and restaurants and the mosque and so on and so refugees are definitely integrated in this area just because there are also a lot of Arab populations from before … Maybe some Bulgarians settle their families here at home but it's traditionally the Arab quarter and a lot of refugees have come there. (Representative of social enterprise, Sofia) The mosque, designed by chief Ottoman architect Mimar Sinan, is located near the Women's Market Zhenski pazar with multiple Arab stores. It brings the Muslim presence in Sofia from the periphery to the center, and, as the interviewee describes the process, provides a space for integration. According to Venkov (2019), the neighborhood is nevertheless politically contested by some majority Bulgarians, who link it to othering discourses: Because of the pre-existing Middle Eastern presence, most of those who reached Sofia [as refugees in 2013] sought sanctuary in the area of the Women's Market. The zone added a second stigmatising name, "Little Beirut," to an earlier one, "the Gypsy Market," and the gentrification ambitions of the redevelopment project wilted. (Venkov 2019, 11) These discourses stigmatize both the Roma national minority and newly arrived persons from the Middle East, othering them through narratives where the newer minority is added as a layer on the historical target of discrimination, namely the Roma. Indeed, Anthias (2021, 150) has noted the intertwined nature of boundaries between racialized groups and migrants who experience similar regulation and ordering.
Even prior to the refugee arrival, however, the Arabic language had already been institutionalized through three private schools in Sofiaone Lebanese, one Palestinian, and one Iraqiwhere pupils can study in Arabic. Studying in Arabic-language schools is, however, discouraged by some NGOs, who view it as an obstacle to integration: At the beginning, [the refugees] all said they want to send their kids to Arab schools, while we were explaining to them, counseling them, telling them [if they attend a Bulgarian school] we will support [through an INGO]. They have a good program for children; after school they have teachers who are helping them with homework. So there is good support from NGOs for kids in Bulgarian schools and Bulgarian schools are free of charge. (Representative of refugee NGO, Sofia) Holding the national, Bulgarian-language exams, which determine later study opportunities as a key reason to discourage people from schooling in Arabic, the NGOs aimed to prevent the option of "integrating" in Arabic. Only teenagers with few remaining school years were encouraged to pursue the path of schooling in Arabic.
In contrast to the NGO representative who witnessed a willingness to attend Arabic schools, representatives of the SAR perceived the opposite: We have an Arab community here in Bulgaria and most of the asylum seekers they don't want to benefit, to be part of, to get involved with, to integrate with them, to stay here [with] them.
Religion is not what they want, it is not a factor that will make them stay here … We have Arabic schools, three schools, only in Sofia. We have mosques here also in Sofia, but they don't want to be part of this community, they don't want to integrate here with them … The material assistance is not attractive here, is not enough for them and they don't want to stay here. Most of the Syrian refugees have relatives abroad and they want to reunite with them and not to stay here alone. (SAR representatives, Sofia) The state representatives frame the Arabic-speaking communities as an option for facilitating integration and belonging in Bulgaria. Economic precarity and separation from family members in other EU countries were nevertheless seen to trump the opportunities that exist there to follow one's religion and use one's language of origin in established religious institutions and schools. Even though the state representatives fashion a negative discourse on migrants' unwillingness to integrate with the help of existing minorities, the statement reflects a different narrative than those prevalent in Western countries, which construct Muslim communities as a barrier to integration (Foner and Alba 2008). Rottmann and Kaya (2021) show how some Syrians living in majority Muslim Turkey feel a religious connection and cultural proximity to Turks that, despite dire economic conditions, contributed to an unwillingness to move to Europe. In Bulgaria, amid similarly challenging economic conditions, the small size of the diasporic and Muslim communities may have contributed to the willingness to move on to countries with larger diasporas, as well as friends and relatives, as suggested by the interviewees. Indeed, just as many former students applied for asylum in Bulgaria due to their familiarity with the country and contacts there, many persons that have family or networks in other EU countries wish to settle in those countries instead.
Hence, the minority position of the communities, transnational family ties, but also prior ties to Bulgaria through studies closely shape how refugee belonging is constituted in Bulgaria. Precarities faced by refugees in Bulgaria are, to a certain degree, mitigated by both organized and individual assistance in which settled minorities play a role. Diasporic groups, historical state Muslim institutions, and national minorities thereby form a marginal, but in many ways important, piece in the complex picture of European asylum politics in Bulgaria. They shape the dynamics of nationhood that refugees face in Bulgaria, and by extension shed light on ongoing contestations of belonging, Europeanness, and inclusion in the EU.

Conclusion
This article has shed light on nation-building, boundary-making, and belonging in Bulgaria by analytically combining the position of "old" minorities with contemporary refugee immigration. By analyzing testimonies gathered from integration and minority workers, the study investigated how boundaries are drawn between settled minorities and newly arrived refugees in Bulgaria. Drawing on scholarship on belonging, boundary-making, and bordering, the study has shown how spatial, linguistic, and religious boundaries keep settled national minorities separated from recently arrived refugees. Bulgarian Turks were identified as being spatially separated from refugees, a boundary that is further fueled by the emigration of young Bulgarians and refugees. The legacies of assimilation that weaken the position of the Turkish language and Islam within Bulgaria were identified as a further reason for boundaries being drawn between the groups. The study nevertheless shows how settled Arabic-speaking diasporas in particular have overcome the boundaries due to spatial proximity, a shared language, and experience of migrating from the same countries of origin. They have thereby participated in important practices contributing to refugee reception, integration, and belonging. However, some acts were witnessed to be conditioned by assimilatory undertones such as demands on women to remove the veil and requirements of adjusting behaviors connected with the country of origin. Hence, the strained position of Islam in Bulgarian nationbuilding, impacted by ideas of assimilation, shapes how minority communities also relate to visible expressions of religious belonging or behavior deemed as unsuitable for Bulgaria or "Europeanness".
Despite the boundaries identified between the established Bulgarian Turkish minority and newly arrived refugees, the historical presence of the Turkish minority contributes to a variety of institutional spaces such as mosques that function as spaces for interactions, meetings, and charity activities between refugees and established Muslims. While profound contacts exist between some individuals, ranging from material support to long-term integration assistance such as support with housing and employment, most of the organized help for refugees has consisted of emergency humanitarian assistance. The lack of state support was stated to lead to the emigration of refugees to EU member states with larger diasporas and a greater degree of support. At the same time, the Muslim and Arabic-speaking communities were viewed as facilitators of belonging rather than obstacles to integration.
Since the present study is concerned with minority perspectives that are often neglected in the study of refugee integration, it has not put a focus on the many majority Bulgarians working in refugee NGOs or helping on an individual basis. Hence, it makes no valorized claims about which groups have assisted more or less. Furthermore, it is limited to the perspectives of organizations working with producing refugee belonging through integration activities and, as such, does not provide firsthand accounts of the actual sense of belonging of refugees. Instead, the study has brought attention to, and described, how politics of refugee integration temporally and relationally interconnect with past and present policies of nationhood and minorities, constantly (re)producing boundaries and categorizations with shifting functions. The study calls for further investigations on the role of othered minorities in processes of integration and belonging in Europe, particularly analyses that raise the voices of those affected by such politics and that acknowledge Islam as a religion with long-lasting roots in Europe rather than a new phenomenon. seminars; Çağla Demirel for valuable input with regard to translations; the editors of Nationalities Papers for their work during the publication process; and finally, the two anonymous reviewers for all their crucial comments that helped rework the article.
Financial support. This research was supported by a grant awarded by the Swedish Research Institute in Istanbul.

Notes
1 The present study designates all prospective asylum seekers as refugees, while also sharing the critiques of the arbitrary dichotomy of "refugees" and "economic migrants." Instead, different types of migration are viewed as being part of the same, long continuum (Yuval-Davis 2011, 37). 2 The Dublin system is formed by the EURODAC Regulation establishing a centralized asylum fingerprint database, and Regulation (EU) No. 604/20133, determining that the first country of entry is primarily responsible for processing an asylum application. 3 Per capita, see World Bank 2020; https://data.worldbank.org/indicator/NY.GDP.PCAP.CD? locations=EU 4 The Grand Mufti's Office is responsible, among other things, for politically representing Bulgarian Muslims, for Islamic education, and for issuing non-binding opinions (fatwas). It was founded in 1909 through an agreement between the Ottomans and King Ferdinand of Bulgaria. 5 Statistics on the size of the diaspora are unavailable, but estimates have varied over time between 17,000 in 2004 to a few thousand persons in recent years. Afghans, Iranians, and several other nationalities also took part in the programs and have smaller diasporas in Bulgaria. Other members of the diaspora settled through engaging in business following the fall of communism (Zhelyazkova 2004). 6 CEAS sets out asylum related minimum standards for EU member states, regulating, for example, the conditions surrounding asylum decisions, reception conditions, and assigning responsible EU states for each asylum seeker (Dublin system). 7 Unlike many other EU countries, asylum seekers in Bulgaria do have the right to work already after three months, yet the labor market and social support is mostly precarious. 8 The purpose of using media material was, rather than to provide an exhaustive analysis of all media coverage on the topic, to add further perspectives to the interview material. The articles (listed in the reference list) were selected from Turkish-and English-language Bulgarian media outlets (using keywords such as "refug*", Syria*, etc.). 9 The interviewee refers to a çarşaf, a loose robe similar to the abaya or chador, worn by a small minority of Turkish women. 10 The author wishes to sincerely thank Professor Ayhan Kaya for pointing out this connection.