Social Boundaries at the EU Border: Engaged Ethnography and Migrant Solidarity in Bihać, Bosnia–Herzegovina

ABSTRACT This paper is a reflexive examination of ethnographic positionality in the Bosnian border town of Bihać as it experienced a bottleneck of migrants and refugees from outside of Europe attempting to reach prosperous EU states by traversing the Balkan Route of irregular migration. Drawing from critical border studies and the principles of engaged ethnography, I approach the relational quality of life on the border as it shapes and also produces social boundaries that must be navigated also by researchers. The paper gives an account of my own active engagement in migrant solidarity activities and chronicles how this positioning came to be seen as my aligning myself with one distinct “side” of the social boundary between those working to support migrants in the community, whether as part of the official migration management response or as autonomous solidarians, on one hand, or those advocating the containment and expulsion of migrants, or “anti-migrant” positions, on the other. I show how this positioning helped to reveal the relational quality of social boundaries created through different ways of relating to the border.

who worked there, and general relations with the local community. I was also hungry and planned to get some take-out food. Unfortunately, my timing was terrible. It was a hot Sunday afternoon and very few people were out. There were no customers, only the owner and a Bosnian woman who was cleaning things in the kitchen.
Suddenly I was keenly aware of how they must see me. To myself, I was an anthropologist familiar with Bosnia and its language, but this was not going to mean anything to the Pakistani owner. I was also someone with a history of working in solidarity with refugees and migrants. I had been talking with migrants along the Balkan Route over several years, visiting official camps and improvised squats in Serbia and BiH and learning about the efforts of local and foreign volunteers to support people on the move with humanitarian aid, hygienic services, and friendly interactions. Much of this activity also brought me viscerally back to my time 25 years previously working in refugee camps in Croatia where the refugees were Bosnians, some of them from this same area of the country, the Una-Sana Canton (USK). 2 In my head, I was somehow still a scruffy young volunteer ready to extend some humanity to people in need -or I hoped this part of me would be recognized in a context of widespread suspicion and outright racist hatred of migrants that was beginning to prevail. What they more likely saw was a strange middle-aged white woman in flat sandals, not groomed much like most local women my age, showing up alone on a dead afternoon in late summer. We settled on English and I tried awkwardly to convey that I was a researcher sympathetic to migrants' predicaments, that I was interested in how the restaurant was doing, what relations they had with locals. The owner looked at me skeptically and suggested I could talk to the woman in the kitchen. She was a bit less wary but still didn't know what to make of me. I let her know she could switch to Bosnian and she told me first about what kind of food they prepared, and then about how rarely any of "our" people come here. After a few more attempts at conversation that didn't quite go anywhere, I left with no food, thinking about how I would be perceived in the year to come.
It was not that I had expected instant rapport or recognition. I had not yet spent the time it takes to build relationships and allow people to form their own understandings of who I was and what I was doing there. As a foreigner from "the west," whether I was perceived as a US-American or someone coming from the EU where I now live, I knew, too, that I might be readily connected to well-resourced international organizations or, more abstractly, to liberal discourses of human rights, democracy, and transparency, as had happened during previous research fieldwork in BiH (e.g. Helms 2006, 2013. The context had shifted, however. Previously, in the decade or two after the end of the 1992-5 war, the actors in the topics I worked on were generally local residents or stranci (foreigners), also referred to as the "international community." This was mostly taken to mean western diplomats and military, aid workers, human rights and democracy practitioners, and other participants in the postwar rebuilding effort (Coles 2007;Gilbert 2020). 3 Now, the focus of the research and one of the most pressing issues in the community centered around a different sort of "foreigner." The migranti (migrants), racialized and othered as a group as non-European in opposition to the locals, were designated like any other outsider by phenotype, dress, language, and patterns of moving through the town, even as these markers were not always reliable or accurate. My appearance classified me outside the category of migranti but not always clearly as either a local or a "western" (read white or European) foreigner, depending on how much I was able to talk with people. And where I was perceived as western, in the current context the most common assumptions were that I could be either a tourist or an aid worker who was here because of the "migrant crisis." Didier Fassin (2011) has pointed to the way in which borders are intertwined with and generate social boundaries "as internal social categorizations" (ibid.: 214), also away from physical state borders, as a crucial aspect to understanding the experience of (im)migration and asylum claiming in host countries. Bihać presents a different sort of relationship between the border and social arrangements, as the migranti spending time in the region were perceived of and often perceived themselves as fundamentally temporary. Bosnia was merely a territory that lay on the route to their goal of more prosperous EU states. Unlike the immigrants aiming to settle in the societies Fassin discusses, here they were assumed to be just passing through. While this was not always an accurate read (there were those who made attempts to stay) and the temporariness sometimes stretched into years, it profoundly shaped the impact of the border on social relations much in the way that Fassin describes. This paper elaborates on this relationship between borders and social relations by focusing on how my positioning as a researcher compelled me to have to confront some of the main social boundaries that had been created or shaped by the presence of migranti, which in turn were present because of the nearby border, on top of the existing social boundaries that influenced everyday interactions in the town. Starting from the principles of engaged research and reflexive ethnography, I describe how I consciously placed myself in different spaces and roles as a participant, coming to inhabit multiple positions and, no doubt, to provoke many more assumptions about what I was doing than I ever became aware of. Some of my activities were read as pure activism or humanitarian work on behalf of migrants, even as my own reasons for engaging in them were also rooted in the ethos of participatory research, but this limited my access to residents who opposed the provision of any form of aid to migrantsthose on the other side of one of the strongest border-created boundaries. This paper is thus an interrogation of the implications of this positioning and on the hurdles and possibilitiesthe borders and boundaries of my own positionalitythat come with doing ethnographic research in this particular border context, just outside the EU. I examine my decisions and how they played out against the relational aspects of social life in Bihać and the "crisis" of migration produced by the border. 4

Research positioning and vulnerabilities
The history of anthropology and ethnographic research is saturated with border crossings and researchers positioning themselvesand being positioned by othersas studying the Other. For several decades now, "native" (Narayan 1993) or "halfie" (Abu-Lughod 1991) ethnographers have offered valuable reflections on the different challenges faced by those perceived as belonging to the community under study, or expectations and perspectives that come with having partial claims to belonging. All ethnographers position themselves in relation to different sorts of borders as well as social boundaries, especially those of class and gender (Narayan 1993;cf. Fassin 2011). As with social relations in general, researchers and their interlocutors have a range of personal and social aspects at their disposal according to which they can evoke closeness or distance of the researcher. The current moment of migration along the Balkan Route to the EU brings border crossing and Otherness into even sharper relief, both through the ways in which migrants and refugees are othered by the local population, and how foreign humanitarians and migrant solidarity activists have been either praised or castigated by association, but also for the ethical and political implications presented to researchers themselves, regardless of the extent to which they are seen to inhabit insider or outsider positions.
Most of the observations, dilemmas, and ethical choices associated with border ethnography are in general terms, therefore, not new or unique. I take direction from longstanding critiques put forth by poststructuralist, decolonial, and especially feminist and queer ethnographers. It is these approaches I find the most insightful in considering the ethics and politics of doing research that is meant to illuminate and even contribute to overcoming inequalities or injustice to marginalized groups, and in dissecting the implications of doing activist or participatory research. One of the most important messages to come out of this scholarship has been about the dangers and traps of taking positions of solidarity with one's research participants in the first place, as Judith Stacey and others pointed out several decades ago (Stacey 1988;Visweswaran 1994). Feminist ethnography and related critical approaches take seriously the ethical principles that should form the basis of any research with human beings: the imperative to do no harm, but even more what Boellstorff et al call an "ethics of care" (2013), the avoidance of orientalist or Eurocentric representations, taking seriously power differentials and the researcher's power to frame "local voices," and critically reflecting on our own and others' positionality in relation to the community being researched (see e.g. Davis and Craven 2016). Despiteor rather becauseof the way we enter the field with these principles in mind, there are dilemmas that cannot be resolved, only explicitly acknowledged and managed.
Driven by these principles, I entered the field, the city of Bihać and its surrounding region of USK, with an interest in how the community was responding to its place in history and geography and to current political realities within the global structure of migration management, EU and regional politics, and the appearance of migrants and refugees in their midst. The ethnographic terrain I moved in ranged from that of everyday life as part of a family (my husband and school-aged son were with me) in the neighborhood, schools, cafes, and shops to spaces of makeshift migrant squats and official camps, to the offices of NGOs and international organizations, to online interactions in social media and mobile phone apps. While not a "native anthropologist," Bosnia-Herzegovina is the place where I am most grounded in terms of language abilities, longterm perspectives, and academic knowledge. In contrast, I did not feel equipped to focus on the trajectories of migrants and refugees, as I don't speak their languages and have no deep knowledge of the contexts from which they come. But I came to interact with many of them, especially those who spoke enough English or Bosnian and with whom I had prolonged contact; they figure in my research, as did other non-Bosnians and non-Bišćani (people from Bihać) I encountered, as part of the community, regardless of their length of stay or degree of integration. The ethical concern of avoiding harm to those in vulnerable or sensitive positions most obviously related to migrants and refugees, depending on the papers they carried, their access to accommodation and food, or other details of their circumstances, but this also applied to local residents whose provisions of food, accommodation, or other forms of material aid to migrants could bring ostracization from neighbors, threats of violence, or fines from the police in a climate of increasing condemnation and criminalization of migrant solidarity (a widespread trend across various borders of the EU: see e.g. Dadusc and Mudu 2020;Fekete 2018;Mainwaring and DeBono 2021;Tazzioli 2018).
These vulnerabilities stemmed from the dramatic extremes created by the border violence and responses on the BiH side of the border to the "migration crisis"violent pushbacks by Croatian police, restricted freedoms of movement and violence from BiH authorities, substandard conditions in internationally run camps, the hardships of living in forests and abandoned structures, and so on. At the risk of engaging in "crisis chasing" (Cabot 2019;Rozakou 2019), these dramatic conditions were also part of what compelled me to engage with this topic and why it is important to understand in context. This, along with my ethnographic proclivities compelling me toward active participation, is also why I sought to engage directly in efforts to alleviate some of the hardships born by both migrants and residents. I simply found it difficult to witness the situation when it was relatively easy to make a difference by buying food or some shoes for these often destitute young men forced to live in squalor on the margins of the town. And I found it awkward not to chip in when I saw acquaintances with limited resources giving what they could to alleviate these hardships. In any case, it was important to balance this with spaces of engagement that might go beyond that of one-way aid provision, to probe the possibilities of more reciprocal interaction, which I found to some extent in my volunteer work in the camps and by joining residents I knew in maintaining family-like bonds with groups of migrants over time.
These modes of positioning, it must be acknowledged, were also practical from a research standpoint, both in terms of access and in the sense that my research was also a means by which I could personally benefit, by maintaining and furthering my academic position. All research serves a similar purpose in the "business of anthropology," as Heath Cabot warns, yet she does not go so far as to call for an abandonment of our endeavors entirely (2019). In the particular context of highly controlled sites of migration management, Katerina Rozakou points to the ways in which researchers can also become part of the power structures they are critiquing as a price we pay for gaining access (2019). Like the dilemmas identified by feminist ethnographers discussed above, the power hierarchies in which we are all embedded and frequently also help to perpetuate are not always possible to dismantle, nor are they necessarily obstacles to relating an authentic account of the relations under studymore often than not power hierarchies in themselves. My aim is therefore not to meaningfully resolve or justify the power differentials in which I am also placed, but to lay bare the specificities and dilemmas presented by these dynamics, especially those shaped by the context of living near a "sensitive" border, one that has become the object of determined efforts both to cross and to prevent this crossing. I return below to the consequences of these choices.

Negotiating borders and boundaries
The dynamics I am concerned with start with the conditions of life at the border, a demarcation that has not changed much in the past few centuries in terms of location but that has carried a variety of meanings and consequences over time. The border in question became solidified as a heavily militarized one separating Ottoman Bosnia and Habsburg Croatia. The population structure of Bihać and nearby towns was fundamentally established because of this positioning along this border, as Slavic Muslims were sent to guard and inhabit its Bosnian side while Serbs and others from other parts of the Habsburg Empire or refugees from Ottoman lands were given land and recruited as soldiers to keep watch over the other side. Bihać and other parts of the Canton concentrated along the northwest border area have thus had a higher concentration of Muslims, now called Bosniaks, than other parts of BiH, a concentration that was reinforced during the ethnicized violence of the Bosnian war in the early 1990s. After the demise of the Hapsburg and Ottoman empires, this border has continued to exist, for a time during socialist Yugoslavia as merely an internal border between republics of a larger state, but since the disintegration of Yugoslavia in the early 1990s, as an international border once again. From the end of the war in 1995 and after the situation stabilized, Bihać residents were able to cross the border with increasing ease, especially since many of them found ways to obtain a Croatian passport. Croatia's capital, Zagreb, continued as the first larger city of reference for many things, the two-plushour drive making it much more convenient than BiH's capital Sarajevo, which takes at least five or six hours to reach by car and more by bus.
When Croatia joined the European Union in 2013, the significance of the border changed again; Croatia was charged with guarding an external border of the EU under the incentive of being allowed into the border-free "Schengen Zone" of the EU were they to show that they could get it right. A BiH passport was still accepted for visafree travel to Croatia, even if visas were sometimes necessary for the rest of the EU, and having a Croatian or other EU passport became even more valuable for BiH residents (see Jansen 2009). But there were soon much bigger changes: already from September 2015, people using the Balkan Route were traversing Croatia, largely because Hungary, a Schengen country, had erected a razor-wire fence and new legal structures backed by violent and harsh measures that posed serious deterrents (Cantat 2020;Kallius 2017). At the same time, an official "corridor" was set up to usher people through Croatia and Slovenia (Bužinkić andHameršak 2018 [2017]). After the closure of the corridor in March 2016 and increasing violent push-backs carried out by the Croatian police across the border with Serbia, the Route once again shifted, now into BiH with wouldbe border crossers concentrated in the western parts of USK especially from 2018 (see Ahmetašević and Mlinarević 2019;. The border now attracted attention for a new set of reasons and by a new set of actors: migrants, local and foreign humanitarians, smugglers, border officials, and police (from Croatia, BiH, and increasingly, the EU's border patrol agency, Frontex) clustered around the border area, entangling local residents into acts of aid provision, surveillance, and bouts of fear fueled by experiences of break-ins, accidental fires, minor thefts, and abandoned debris left by migrantsand the amplification and exaggeration of such incidents by sensationalist media.
In other words, there were multiple distinct groups of people focused on the border in different ways, reflecting not only different social and legal positions but also different and differing ways of conceptualizing the borderwho should be allowed to cross it, where, and how, and what mechanisms should be used to control the flow, as well as ideas about who should be allowed to stay in the community, to open a business or move in certain spaces. Sarah Green (2010) has analyzed a similar border zone, that of the island of Lesvos in Greece, through the ways in which different populations "perform" the border. Different groups of peoplein Green's example, the refugees and stay-ees from the 1923 "population exchanges" that helped produce Greece and Turkey as modern nation-states, or migrants and refugees using this part of the Balkan Route (many of whom would later reach BiH)perform different concepts of border even in the same shared space and in response to changes in geopolitics, legal regulations, migration patterns, or any number of other factors. Key to this approach is recognizing that "diverse theories and people do not simply co-exist; they are also in social relation with one another." This is because, "as a form of classification, or a way of making and marking distinctions borders not only separate, they also imply relations." (ibid., 262).
Coming back to Didier Fassin's attention to the interplay between borders and social boundaries (2011), it is imperative to see such relations at work in a community near the border like Bihać, Here, I encountered longstanding social boundaries that had been played up or down by the "migration crisis" and later the Covid-19 pandemic, as well as new boundaries created by the new condition of the border and clashing ways of conceptualizing it. For instance, when local and state-level officials moved to consolidate and regulate migrant accommodation away from spontaneous (and dangerous) squats to controlled camps, or "Temporary Reception Centers," and management of such centers was entrusted to the International Organization for Migration (IOM), allocation of well-paid positions with IOM and the other organizations assigned to work with migrants was said by residents to depend on existing power networks in the local governmentjobs were handed out to family and friends as favors. I heard laments about the sidelining of "real" Bišćani in favor of those coming from surrounding villages and smaller towns, evoking longstanding social divisions between the civilized and cultured vs. rural-based backward clans that were grabbing all the advantages (Hromadžić 2018;and see Vetters 2014). Similarly, condemnations of how certain entrepreneurs had taken advantage of the war and post-socialist privatization to enrich themselves also informed people's opposition to the establishment of a migrant camp in "Bira," a privately-owned defunct refrigerator factory within the city limits where EU money would be channeled to the (for a time) well-connected owner. Protests against the camp were not only about what might be the best solution for accommodating or deterring migrants but also about who should or should not benefit from their presence. These social boundaries through which many inequities and failures of governance were interpreted thus also infused and were themselves, in turn, shaped by the response to the migrant situation and the EU's border practices.
One of the most relevant newly created social boundaries, and the main focus of my analysis here, was not one marked by social identities like ethnicity, gender, or class (as Fassin mainly addresses) but between those who favored extending humanitarian aid to migrants (whether they approved of their movements or not) and those who advocated various forms of deterrence and expulsion. As I detail below, this was more a wide grey zone than a clearly demarcated line, more a designation of extremes than an accurate measure of bounded, opposing social groups. Those positioned clearly on one side or the other referred to it, yet most of the population remained silent, wavering between worries about safety and hygiene and their discomfort at witnessing the predicament of the migrants. In my research, I was not surprisingly forced to negotiate with the various facets of this division as a way of making sense of people's responses to the influx of migrants, but in the context of increasingly hostile official policies towards migrants and informal humanitarian aid, my own classification in relation to this divide became even more pronounced, as I outline in the following sections.

Positioning
With my interest in local responses, I aimed to position myself at sites of official migration governance, as well as informal sites where local residents encountered migrants, especially where they undertook efforts to influenceto help or to hindermigrants' stay in the community. The most structured activity I had was to volunteer with an Italian NGO called IPSIA 5 that had been working in the community for years. When migrants began appearing in Bihać in 2018, IPSIA was one of the first organizations (along with many private individuals) to provide food and other emergency aid. When official camps were established and IOM was put in charge, IPSIA opened their "Social Café" where camp residents could get a cup of sweet hot tea and socialize with volunteers and each other, finding escape from the tedium of the everyday sameness through board games, crafts, sports, or language lessons. I spent my time with the other local and foreign volunteers mostly in TRC Bira making and serving tea, providing game equipment, enabling the use of an outdoor kitchen, and chatting with residents about their situation, the camp, or life in general. Through this role, over time, I also got to know some of the workers from IOM and other organizations involved in camp management who sometimes stopped by our kiosk for tea, and I chatted with Red Cross workers in the camp's kitchen during our trips to prepare the tea or wash up the pots and utensils. Through the locals I met this way, I got a better sense of how these jobs and voluntary positions were integrated into the community and how people related to migrants when they worked with them daily.
The camps were thus important sites, but they were sites of local incorporation into the larger, official migration management regime itself. Dadusc and Mudu (2020) usefully distinguish between this "humanitarian industrial complex" and what they call "autonomous migration solidarity," the latter of which is part of what threatens the control exercised by official humanitarianism (cf. Ticktin 2011) and is therefore being increasingly targeted with hostile policies and criminalization, accused of attracting and facilitating illegal migration or preying on it for personal gain. Because of the frequency of generalizing comments about how empathetic the local population was or was not towards migrants as human beings, I was keenly interested in the community residents who provided such "autonomous" aid to migrants outside the camps, commonly referred to as volonteri (volunteers) or simply ljudi koji pomažu, "people who help," although a few of them operated through local NGOs or the Red Cross. This was a diverse and scattered set of people who were difficult to get to know, in part, no doubt, because the idea of talking to a foreign researcher did not strike them as a good use of their already stretched time, but also because of the atmosphere of suspicion, negativity, and later de factocriminalization of individual help to migrants that came from the police, local officials, and many in society at large. These volunteers did not want to call attention to themselves even as they depended on outside donations to sustain their work. I approached these people who help through trusted mutual friends from the community of those engaged in solidarity with migranti, but my efforts to participate in their activities did not get off the ground much until I became a volunteer and a (small) donor myself. This was a natural move as I was witnessing so much basic human need that could be easily, albeit temporarily, alleviated by buying a few basic goods at local stores. This is what the autonomous volunteers were doing, some on a larger scale with donations from foreign organizations as well as local groups, but there were also those, like my neighbors, who had befriended small groups of migrants whom they gave food, a warm place to sit, the use of a bathtub, or clean clothes, washed in their machine. An informal collection among my friends, colleagues, and their contacts raised a small pot of money with which I helped my neighbors and another couple I knew help "their" groups of young men and also funded larger purchases for some of the more involved autonomous volunteers, who would distribute packets of food or clothing items to the migrants that contacted them or to whomever they encountered at particular squats. Eventually I also responded directly to appeals from migrants who contacted me on WhatsApp, driving to out-of-the-way pre-arranged spots to hand over bags of groceries or clothing.
These two main activities placed me squarely on one side of the boundary between those who opposed the migrants in all respects and those who supported offering them help. There were many positions along this continuum, taken for different reasons and in response to different inputs, and of course, going back to Green, the marking of these distinctions in fact implied relations among them more than stark divides. Most people in the town lived their lives preoccupied with other things, even as the migranti were a constant topic of conversation and visible point of worry. For some, the worry was for fellow human beings. They worried that these new people in their midst did not have adequate shelter, might be hungry, were far from their families, and, as the border violence became more visible and common knowledge, that they had faced and would continue to face violence and humiliation. Many others worried about the newcomers as a danger. Storiessome credible and others less socirculated by word of mouth, in the media, and especially on social media, about migrants attacking teenagers to steal money or phones, break-ins to property or thefts of clothing hanging out to dry in people's yards, or male migrants as sexual threats to local girls and young women. There were worries about the safety and cleanliness of the City Park and the lush banks of the Una River as it flowed through the town. The first summer with the migrants, a rumor spread that migrants were catching and roasting the beloved wild ducks that lived along the Una River (Hromadžić 2020). Well before the Covid-19 pandemic hit, migrants were feared as carriers of disease, thanks to the contagions like scabies and lice that were spreading too easily in the crowded and unhygienic conditions of camps and squats and to unfounded rumors of more serious diseases they were said to carry. Fear was the main driver of these types of worries. I met people who otherwise expressed great empathy towards the migrants, who may have bought some of them food outside a bakery or collected used clothing for them, who nevertheless avoided public places frequented by migrants or stopped buying fruit and vegetables at a certain market because, they said, they saw migrants shopping there and touching the produce.
Others channeled these fears into much more negative discourse on social media, in everyday talk, or even in public protests against migranti and the camps set up to house them. There was a fine line between concern that migrants had adequate accommodation and a desire to remove them completely from the town, the Canton, the country, or even Europe itself. Calls for discouraging, expelling, or attacking migrants were often couched in dehumanizing, racist language specifically aimed to counter pleas for empathy, humanitarian help, or contextualized understanding of migrants' plights. Crystalizing on social media, people expressing such opinions came to be known as "anti-migrant" groups, after a virulent internet portal of this name, started by a well-known Bosniak nationalist from Tuzla. The portal published wildly false or misleading sensationalizing stories about migrants, often with only a provocative headline and no actual story. 6 Facebook groups emerged with names like "Stop the invasion" (STOP Invazije), "I want to have a saymigrants" (Želim da se pitammigranti), "Stop illegal migrants -Velika Kladuša," and, more neutrally, "migranti -Bihaćrefugees." These and other groups and portals connected to specific towns regularly shared sensationalist and negative media items about migrants, some of which also came from government reports and statements at different levels. Officials tended to address migrants as problems of security or public order, and only through these lenses as objects of humanitarian "solutions." Within this mindset, camps and organizations or individuals who provided food, shelter, legal aid, medical aid, or education to migrants sat firmly on the "pro-migrant" side, not only said to be enabling migrants' stay in Bihać and USK but actually attracting more of them. Commenters in this vein often went so far as to suggest, in line with increasingly common official policies, that anyone helping migrants was personally profiting from it, a conflation of the humanitarian and autonomous solidarity sectors with that of "smugglers" and others who provided a range of services to migrantsfrom receiving money wired to people without papers, to renting accommodation, to providing transportation, to facilitating border crossings in different waysfor profit margins far beyond regular prices. In this world, this concept of the border (Green 2010), there were no selfless humanitarians just as there were no refugees or jadni ljudi ("poor/ wretched people") on the move who were genuinely deserving of concern or help.
It was along these lines that a sharp social boundary was drawn, especially online. As with any discursive dichotomy, in offline everyday practice this boundary was unsurprisingly much more malleable and subject to multiple interpretations. There was much more grey area between the "for" and "against" or between those who actively provided aid to migrants and those who actively worked to make migrants' lives difficult than could be made into a clear boundary. What people were "for" or "against" was likewise far from fixed. There were those who helped migrants as a way to prevent disease and death but did not particularly make friends with them or want them to stay in BiH. The migrants were only passing through in any case; they did not really have a place in Bosnian society. Among those protesting the accommodation of migrants in camps, some said they objected not to the migrants themselves but to the lack of decent facilities available for them, to the enrichment of local tycoons whose received EU money as rent for the housing of migrants in their facilities, or to the non-enrichment of local bus companies when it seemed to be only Sarajevo lines profiting from transporting migranti. While many denunciations of migrants reproduced racist and Othering tropes, there were still those who stayed away from such language, instead emphasizing concerns about public safety, as more break-ins and property damage incidents were happening since migrants had so few options for finding shelter near the border. One man who had been modestly active in a local political party tried, eventually unsuccessfully, to maintain a group on Facebook for reasoned dialogue around the migrant issue. As he put it, people don't love or hate migranti, they pity them. These are scenes they don't want to have to witness in their everyday lives, he told me, but they feel they have little say or influence over the matter.
This distinction went beyond social media or neighborhood talk, however, as local authorities and institutions gradually created and sustained a hostile atmosphere for migrants and those who helped them outside of official camps and organizations registered to provide aid. By the winter of 2019-20, stopping to talk to someone read as a migrant could elicit suspicious or disapproving looks from others; giving them food or clothing might draw "anti" campaigners to take pictures and post them in social media groups, as had happened to several of the autonomous aid givers. Cantonal and city authorities had prohibited providing accommodation to migrants, whether for free or paid; hostel and pension owners were regularly inspected and fined if found to be allowing anyone determined to be a migrant to stay the night. This meant that even inviting someone into your garden was immediately suspect; an "anti" neighbor might confront you about endangering the neighborhood or report you to the police for accommodating migrants. This was a constant fear for my elderly neighbors who regularly helped a group of young Pakistani men, being careful never to allow them to stay the night but hosting them in their garden or house in other ways. Indeed, they were visited a few times quite rudely by the police late at night. The authorities at the same time made possible the establishment of camps where migrants received food, shelter, and necessary clothing, something the "anti" folks resented, despite the fact that local authorities shared much of the "anti" groups' hostile rhetoric towards migrants and a clear determination to close camps within or close to the city limits and to keep migrants out of sight of local residents. The ire expressed at "anti-migrant" protests and on social media was directed as much at IOM and its partner organizations as at other actors they deemed responsible for attracting and keeping migrants in the region.

Consequences: who is this person living among us?
This discursively (re)produced social boundary posed a classic ethnographic dilemma for how I was to proceed with the research. Forging relationships with people doing particular activities, in this case migrant solidarity or humanitarian work, led to insights only possible through long-term consistent participant observation and the building of a common sense of purpose. At the same time, it might have foreclosed similar perspectives on the other side of the divide, in this case those opposed to the migrant presence for various reasons. While the bulk of my time was spent in migrant solidarity activities, I did also interact with ordinary citizens not involved in anything having to do with migrants, who voiced a range of opinions based on empathy or pity but also fear, disgust, resentment, or outright racism towards migrants. I witnessed people's behavior in and around stores where migrants shopped or were prevented from entering, on streets and in cafes, among parents picking up children after school. I went to protests held against the existence of camps for migrants, particularly against the use of the Bira former refrigerator factory as a camp housing between 1000-2000 men and male minors, which was a constant issue during the time I lived in Bihać. These encounters helped me see beyond the more extreme statements being repeated on social media, even as social media narratives surfaced in offline talk, too.
As long as I was seen as a curious outsider who was open to listening, it seemed, I was managing to have some conversations with people who supported the expulsion of migrants. The first wave of Covid-19 from mid-March 2020 put an end to the momentum of these encountersor should I say to the courage I was gathering up to speak to more people with whom I knew I disagreed. But then something happened that changed my positioning: the local leader of the "anti-migrant" protesters found and posted my Facebook profile in his Facebook group of several thousand followers dedicated to "stopping the invasion" of migrants in Bihać. This group was notorious for screening its membership and expelling anyone voicing a dissenting opinion; my request to join several months earlier had been met with silence. A migrant solidarian friend who was a silent member of this group sent me screenshots of the post in question and the comments it had generated. Who was this interesting American living in Bihać, the post asked, and why did she love migrants so much? Commenters speculated that I must be paid by hostile forces, even a spy with the CIA. The police should be sent to check my papers, people like me shouldn't be allowed in the country, and on and on. One person who suggested I may just be naïve but well-intentioned was dismissed asnaïve. Two vaguely threatening messages arrived in my inbox from men I didn't know. A man commenting under his real name whom I recognized as one of my neighbors offered that there were foreigners living on his street and this could be the person in question.
Now I was feeling quite exposed. When would one of these people recognize me and cause problems? Would they come to confront me in my own home? All the carefully thought-out explanations of the goals of my research and the rationale behind ethnography were useless in the face of people with whom I could not communicate to offer those explanations. What would happen the next time I tried to talk with people at a protest? I had planned to try to interview the protest leader who administered the Facebook group, a vocal local politician and public figure, but this turn of events made me reconsider that. Fortunately, perhaps, we were still under Covid lockdown, so there was less chance I would run into members of the group, but I and others were still delivering food to migrants during this time, looking for more and more inconspicuous spots to hand off supplies so as to avoid visibility.
This exposure and labeling as someone working against the local population was in all likelihood a result of my informal call for monetary donations to support migrants and local volunteers. I had used email but also Facebook. This appeal was at first limited to my own circle of Friends but at the request of those who wanted to spread the word, I had reluctantly changed the settings to allow "public" viewing so that others could share it. 7 A Bosnian friend had been among those to share the call; because of this she was later named by the anti-migrant group moderator as one of my local pomagači ("helpers"), as was one of the regular local volunteers I was indeed working with to channel the funds I collected. In fact, the latter and other locally based autonomous solidarians had been directly targeted themselves by the leader of this anti-migrant group, some of them multiple times and with much more explicit threats. For them it was more serious, since they were not just temporary residents like I was. Exposing me in connection to local figures additionally solidified the association of help to migrants with wider foreign forces working against Bosnia, as the "anti" position had it. I was only one in a line of foreigners who were seen helping local solidarians, but an American nevertheless presented a new connection alongside the usual aid from Germany, Austria, and Italy. From a research perspective, this experience of being targeted while witnessing the online harassment of others engaged in similar activities was a form of embodied participation with other solidarians. It was a taste of how it felt to work under the pressure of de facto criminalization and vaguely worded threats, how this entered decision making over whom to respond to with help, how much and what kind to give, where and when it was safe to distribute it. I discussed strategies for not being seen with others doing the same work, even when my deliveries were made alone. Like my work as an NGO volunteer in the camps, being a fully active volunteer outside the camps, not just someone who tagged along on the activities of others, gave me a role from which to interact with local solidarians and humanitarian workers, a practical justification for my presence, precisely the aim of ethnographic participant observation.
One could argue that I might have learned more about those on the "anti" side of the social divide by centering my interactions among them, or that my study would have been more "balanced" had I tried to interact in equal measure with people taking a variety of responses to the migrants. I gravitated towards the solidarians because they aligned with my own sense of what was important in this context but also because these were the people, alongside the migrants themselves, whose perspectives and activities were being made invisible, or paradoxically being made hyper-visible but in a caricatured way, and demonized by the local media and social media spaces. Writing about the place of feminist ethnography in researching the effects of neoliberalism, anthropologists Dána-Ain Davis and Christa Craven make the case for "rais[ing] the volume of subjugated voices" (2011,197). In their research with impoverished women of color in the US, their methodological choices placed them in positions where they could understand and amplify the experiences of those most often denigrated, disciplined, or silenced by the overarching neoliberal power structure. In the case of irregular migrants in BiH, various arms of the "migration governance" machinerylocal and state officials, the EU border regime, international organizations and NGOs, local partner organizations and charities, foreign aid workers and volunteers, and the likehave criminalized and dehumanized irregular migrants and those in solidarity with them or, at best, are part of various mechanisms of control, even while some of these actors also offer aid and humanitarian sustenance to this population. Because of this marginalized positioning, the voices and experiences of ordinary citizens offering solidarity aid, alongside those of migrants themselves were the ones I was interested in amplifying, to the extent that I could, in the spirit of feminist ethnography and other forms of "engaged anthropology" (Mullins 2011).
Having made such choices, I was not in the most favorable position to delve deeply into the lives and motivations of adherents to this anti-migrant stance. The consequences were not, however, that my ability to communicate with and understand such people were cut offexcept, they did exclude me from their Facebook groupbut that they related to me as someone from the "opposing camp." For example, my neighbor who had commented about me in the anti-migrant Facebook group continued to interact with me as a neighbor. He or others on the street had seen that I had greeted migrants or given them packets of food in front of my house and at the house of my next-door neighbors, the elderly couple who often had migrants as visitors. I had heard this neighbor complaining loudly to the elderly couple about the migrant situation, yet they had known each other for many years and continued to visit each other according to the established code of neighborliness (see Helms 2010;Henig 2012;Sorabji 2008). When I later saw this neighbor at an election rally in town, he introduced me to another man he was walking with simply as his neighbor, a foreigner, as if to make clear to the man why he was explaining his position on migrants (the reason he said he had come to the rally) with reference to the time before I had come. He insisted to me that the people of Bihać had offered help in the beginning, but that it had gone on too long. People from countries where there was no war should be sent back home, he said. His explanations were not as harshly worded as those I was used to seeing on social media, even if the conclusions were the same. I sensed that he felt he needed to justify his positions in the face of what he presumed were my arguments, to ensure that he did not sound uncaring or irrational, even though I had not said anything to indicate a position on the matter. It was a moment of connection as neighbors, and at the same time an affirmation of distance and a reminder that I was not fully immersed in his "world" in the ethnographic sense.
On a subsequent visit in May 2022, I had more of a chance to talk to this neighbor, saying I wanted to hear his views. As we talked, he grew impatient and told me that all the information about why migrants were a problem was in the Facebook group. I had not been accepted into the group, I explained, and they later singled me out as suspicious. He considered that for a bit, but then concluded, "that's because you weren't with us." No one actually knew why I was there. When I presented counter arguments to his claims about all migrants being drug addicts and criminals, he responded by saying simply, "it's obvious you're on their side." This encounter confirmed the existence of the social boundary between his group and those who enable the migrants in their view. This opposition did not prevent the usual neighborly relations from being carried outhe was not ostracized by the majority of the neighbors who were less clearly in either "camp," nor did he cut off relations with the elderly couple who favored providing migrants with aidbut it affected those relations. It further confirmed my regret that I had not had much chance to engage my neighbor and his allies earlier, before Covid put an abrupt end to the opportunity. Clearly, I had not been "with them." Yet I suspect that had I known them better, they still would have treated me as a person from the "other side," or at best, a naïve foreigner who didn't get it, not least because of my involvement with different forms of aid to migrants. What my neighbor put into words was the very line of social tension that had been produced by the migrant presence as well as the spaces of connectionimplied relations in Green's terms (2010)revealed through different ways of relating to borders and the people who cross them.

Conclusion: engaged research on the border
The "migration crisis" has introduced new social divisions and tensions among the local population in ways that are directly related to the specific location of Bihać at an external border of the EU and along the Balkan Route of migration. At a time when EU, Bosnian, and international officials project images of security threats, border patrol technology, and the containment of migrants in camps outside the core states of the EU (see e.g. De Genova 2017;Manasijevski 2019; and see Besteman 2020 on the global scale), it is important to investigate the effects of these policies from more human-centered perspectives. This means attending not only to the experience of migrants as full human beings but also to the effects of the migrant influx on social life among local populations. While activist groups do the crucial work of documenting the more dramatic and violent consequences of these policies for people on the move, 8 there is little understanding of how this situation affects communities at the border that, in BiH, are already struggling with political paralysis and corruption, economic stagnation, threats of renewed war, and the out-migration of its own population (see Kurtović 2020;Majstorović 2021).
The social boundary created by discourses that divide people into those who are "for or against" migrants infiltrates relations among people who otherwise interact as neighbors and fellow residents of the town but also reflect differing ways of relating to the border and what it represents, as in Green's framing (2010). If the border should keep out all non-Europeans or any people without (the right) identity papers, then all migrants, whether single men or families traveling together, must be treated as criminals, as people to be contained and away from where locals live. If the border is a marker of hope for a better life where all other avenues of moving across it have been made impossible, a form of rebellion against increasing obstacles to movement by people from the Global South in search of a better life (Besteman 2020; El-Shaarawi and Razsa 2019; Kurnik and Razsa 2020), then people moving through Bihać should be supported in dignified conditions to survive the journey and periods of waiting. These are two possible conceptualizations of the border among others that lead people to differing responses to the migrants in their midst, which in turn shaped the ethnographic landscape I had to negotiate as a researcher.
I have endeavored in this paper to outline the implications of ethnographic positioning as it plays out in a community on the outside of the EU border. I have focused on one of the strongest social fault lines created by the proximity of this border and the related bottleneck of migrants in the community. I have argued for sustained reflexivity in assessing the ethnographer's role in navigating these boundaries, through an examination of my deliberate positionings and their intended and unintended consequences. Chronicling this participation illuminated the drawing of this "for and against" social boundary, attempts to label helpers as enemies of the community, attached to migrants as a destructive and alien force. "Anti-migrant" advocates clearly attempted to place local people who helped migrants into the category of izdajnici (traitors) to their community, while my western association was used to reinforce their sense that broader forces and foreign governments were working against the interests of the community. For the "anti" camp, this strengthened the argument about local "enemies"; they had powerful foreign backers. "Anti-migrant" activists could thus see themselves as part of a global struggle against illegalized border crossings and non-Europeans attempting to move to the EU, of white Europe against the less civilized. Connections to purported western and European values were also at times important to solidarians; they tended to stress associations with human rights, antiracism, or humanitarianism, even as ideals that connected them to movements in the EU that opposed the violence of the EU border regime. Making these links to different aspects of the EU's contradictory messages and practices was in any case all the more important for a population that was regularly represented as peripheral, even outside of Christian European culture as Muslims and/or inhabitants of the backward Balkans, a long-standing association in relation to BiH (e.g. Helms 2008;Rexhepi 2015;Žarkov 2007). 9 Ethnography presupposes participant observation. It was not enough to simply live in the town and move around the region of USK. My insights into the realm of informal, autonomous and volunteer efforts to support migrants by members of the community were only possible as a participant in these efforts, even when participation meant also experiencing the unpleasantness of being singled out by people who saw our activities as directly opposed to their agenda of deterring and expelling migrants from the region. In a context of increasing criminalization of (autonomous) solidarity, these are more and more experiences that illuminate growing inequalities. From a perspective critical of the existing European and global militarized border regimes, these voices are worth amplifying. International activist and aid networks often profile individuals supporting migrants locally as heroic and dedicated humanitarians, but this obscures the full extent of these people's positions and relationships in the community, and their struggles with local social divisions. The model of engaged ethnography of the border that I have tried to follow instead takes seriously the embeddedness of such activities in an existing and evolving social field profoundly affected by the presence of borders. Reflexive engagement through an awareness of the relations created and affected by border dynamics helps illuminate and contextualize the profound effects on local communities of broader movements of migrants and border practices by state and transnational actors.
Notes interrupted by the first wave of the Covid-19 pandemic and is thus still ongoing at the time of writing. 5. IPSIA, or Instituto Pace Sviluppo Innovazione ACLI (Associazione Cristiani Lavoratori Italiani), is based in Milan and is funded largely by Caritas Italy. 6. This portal was eventually shut down for a time under accusations of spreading hate speech but it was allowed to resume posting when a judge threw out the case against the owner: https://depo.ba/clanak/224113/fatmir-alispahic-osloboden-optuzbe-za-izazivanje-vjerske-irasne-mrznje-evo-sta-je-sud-bih-utvrdio?fbclid=IwAR0Hs2srqaM12-rWApRTPykyzWwo 5sxZ6Bpa-D0Y70claJExzpHVngKN3DM (accessed Nov. 19, 2021). 7. I capitalize Friend here to indicate this particular relationship on Facebook as those who have mutually agreed to become Friends and therefore see what the other posts. For readers unfamiliar with this platform, a post can normally be seen only by Friends unless one allows it to be shared further (to the Friend circles of one's Friends) by changing the settings to "public." 8. See documentation and reports by the Border Violence Monitoring Network at https:// www.borderviolence.eu 9. As others have documented elsewhere along the Balkan Route, being Muslim was only sometimes seen as a point of commonality with migrants passing through (not all of whom were Muslim in any case) but could also be taken as an imperative to demonstrate one's Europeanness in opposition to migrants, lest Balkan versions of Islam become a reason for rejection by the EU, western governments, or even local governments striving to join the EU (see Kurnik and Razsa 2020;Rexhepi 2018).

Disclosure statement
No potential conflict of interest was reported by the author(s).