Vulnerability-Making at Europe’s Edge: How Policies, Documents, and Spatiality Intra-Act in the Context of Young Refugees

When young refugees who arrive in the European Union (EU) are categorized as “unaccompanied minors,” an implicit assumption and ascription coincides with their categorization, namely that they are vulnerable. Being underage and being vulnerable are inextricably linked and often equated. Thus, vulnerability is understood as a bodily fact linked to a person’s age, and discourse consequently overlooks how vulnerability is enacted in the EU’s border regime. To demonstrate how vulnerability is produced and stabilized through entangled practices of human and nonhuman agencies, I examine vulnerability-making in the context of young refugees who were classified as “unaccompanied minors” in Malta. I pay specific attention to policies, documents, and spatiality and ask: How is vulnerability of young refugees in the EU’s border regime produced and stabilized? And, how are nonhuman agencies implicated in these dynamics? The article, on one hand, makes an important empirical contribution to understanding the construction of vulnerability in respect to unaccompanied minors. The theoretical contribution of the article, on the other hand, lies in offering a new way of conceptualizing vulnerability in the border regime and within bordering practices in the EU by examining it through the epistemic lens of intra-action.

the move (Belloni, 2016a(Belloni, , 2016bLems et al. 2020;Menjivar and Perreira 2018;Raghallaigh and Gilligan 2010;Rallaghaigh and Thornton 2017), and makes an important empirical contribution to understanding the construction of vulnerability in relation to "UAMs." Despite the wider literature on "UAMs," little is known about their classification along the EU's external border. The fundamental assumption on vulnerability employed here does not suggest that bodily vulnerability does not exist, but that the sole focus on vulnerability as a bodily feature overlooks important processes of contextualized vulnerability-making. To address the lacuna, I propose engaging with vulnerability-making through the lens of understanding spatialities and policies as intra-acting (Barad 2007) in producing and stabilizing vulnerability. The theoretical contribution lies in offering a new way of conceptualizing vulnerability within bordering practices in the EU.
The article adds to a body of literature that offers complex accounts of refugees' agency, and how they are made vulnerable by institutions, border regimes, and powerful systems, placing them at the nexus of "care and coercion" (Andersson 2017, 69). The tension between the parallel demand and denial of acting agentively in refugee management and reception (Ong 2003) highlights the interwovenness of refugees' experiences, humanitarian systems, and border management (Ramsay 2019), and the politics of vulnerability in particular. Malkki's work (1996, 384) is fundamental in particular, as she showed the distribution of humanitarian aid follows logics of verifiable vulnerability, thereby erasing refugees' individual histories and individualities and, ultimately, silencing them. In a similar vein, Fassin and D'Halluin (2005) raise attention to the issue that in times of rising doubt within asylum procedures, refugees' bodies increasingly count as "proof" to be eligible for protection. Ticktin (2014) argues that access to humanitarian aid is more likely the more a body is wounded, through which refugees become "casualties of care" (Ticktin 2011). Vulnerability is often used to "shock" Western societies (Le Espiritu and Duong 2018) or to produce refugees as enduring victims (Hyndman 2000). Accounts which deconstruct vulnerability and address the term's "stickiness" (Bragg 2021) are needed.
In what follows, I introduce debates about vulnerability in the context of young refugees, demonstrating the shift of viewing vulnerability as embodied toward viewing vulnerability and agency as two sides of the same coin. This shift plays a crucial role in perceiving young refugees as active agents rather than passive victims, but the debate on vulnerability in the context of the border regime still predominantly focuses on the role human actors play in creating vulnerability. The discourse, I argue, overlooks nonhuman agencies, which are crucial to consider when asking how vulnerability takes form. Here, I propose viewing vulnerability as co-constituted and stabilized by referring to the literature from within Science and Technology Studies and new materialism. Following the development of my theoretical approach toward vulnerability, I engage with the emergence of the Maltese border regime, taking into account historical and contemporary developments which highlight situational understandings and exactments of vulnerability. An ethnographic vignette illustrates how young refugees become "UAMs" in the island state, and I reflect on how I approached them and their vulnerability-making ethnographically. The subsequent empirical part focuses on how (a lack of) documents and policies which govern "UAMs" and Malta's insularity intra-act in the production and stabilization of vulnerability among human and nonhuman actors. To conclude, I argue that vulnerability is-as other categories and classificationstemporally and historically situated in the EU's border regime in which spatiality and regulations play decisive roles in its enactment. This perspective does not contribute to assessing embodied vulnerability but offers the prospect of recognizing the complex entanglements of human and nonhuman agencies in the production and stabilization of vulnerability.

Vulnerability and Young Refugees: Definitions, Policies, and Academic Discourse
Vulnerability is an overarching term. Cohn (2014) points out that anyone or anything-people, places, and societies-can be vulnerable. What, then, is vulnerability and what does it mean to be vulnerable? There is a tension in the literature as to whether vulnerability is innate or if it is situationally produced (Gilodi et al. 2022).
The innate understanding of vulnerability can, for example, be found in policy documents. The Reception Conditions Directive of the Common European Asylum System (EASO 2016) understands "minors, UAMs, disabled people, elderly people, pregnant women, single parents with minor children, victims of human trafficking, persons with serious illnesses, persons with mental disorders and persons who have been subjected to torture, rape or other serious forms of psychological, physical or sexual violence, such as victims of female genital mutilation" to be vulnerable.
Critical scholars engage with policy documents which understand vulnerability as a static criterion, which can be assessed and then applied, thereby granting access, entitlements, and protection to some, and denying it to others (Ticktin 2011). The situational approach-inspired by feminist scholarshipargues against groupings. It instead highlights that all persons can be vulnerable, depending on the situations and contexts they find themselves in and which create vulnerability. As such, people can become "vulnerabilized" (Lind 2019). The situational approach toward vulnerability takes seriously social, political, economic, and environmental circumstances, indicating vulnerability's "intersectional" (Mendola and Pera 2022) and "temporal" (Gilodi et al. 2022) character.
Several scholars have carved out differentiated understandings of the creation of vulnerability and the political force they have, which play a crucial role in migration management, governance, and access to humanitarian aid (Hruschka et al. 2019;Janmyr and Mourad 2018;Pearce and Lee 2018). The "discretionary nature" of how refugees are dealt with allows states not only to add additional criteria to what constitutes vulnerability, but it also leaves space for translating these criteria into practice through "border work" (Reeves 2014). Research on UAMs is especially fruitful to analyze border work with and through vulnerability, and researchers have begun paying increasing attention to their narratives and situations (Belloni 2016a(Belloni , 2016bCrawley 2011;Lems et al. 2020;Raghallaigh and Gilligan 2010;Terrio 2008;Wernesjö 2020).

From Inscribed Vulnerability to Vulnerability as a Social Construct
The approaches above conceptualizing vulnerability as either inherent to certain groups of society or as created are important in research on young refugees: Some studies identify them as vulnerable per se (Bean et al. 2007;Hodes et al. 2008), thereby not only overlooking the circumstances which make them vulnerable but also underestimating their resilience, agency and their competences (Adefehinti and Arts 2018;Rallaghaigh and Gilligan 2010).
A debate ensued which offered more nuanced accounts, embedding both vulnerability and agency within concrete situations (Rallaghaigh and Gilligan 2010). In addition, scholars have criticized the overreliance of chronological age as being a "neutral" indicator of biological and social development (Clark-Kazak 2013). In this vein, research has moved toward a relational understanding of vulnerability, resilience, and agency (O'Higgins 2012; Tisdall and Punch 2012), arguing that vulnerability is not merely a category but also an experience (Clark 2007). Research is thus concerned with trying to better understand young migrants' relational and contingent positions in the context of forced migration (Huijsmans 2012). This has led to approaches which challenge the idea of equating (ascribed) underage with vulnerability, instead showing that young refugees can still find themselves in vulnerable situations after having reached adulthood in legal terms (Otto 2020a(Otto , 2020bRaghallaigh and Thornton 2017). A contingent and constructivist perspective acknowledges that vulnerability is not merely embodied but also relates to legal and social circumstances (Crawley 2011;Eastmond andAscher 2011, 1195;Orgocka 2012, 3;Raghallaigh and Thornton 2017, 389;Silverman 2016). It has shown thereby that agency and vulnerability are not mutually exclusive, but two sides of the same coin (Orgocka 2012, 3). The central idea is to demystify "UAMs" as inherently vulnerable and as victims of their circumstances-examples such as recruitment as child soldiers, upbringing in hostile environments or growing up as orphans spring to mind-instead highlighting young refugees' agential practices under these conditions (Adefehinti and Arts 2018, 428).
The abovementioned studies criticize the unreflected reproduction of vulnerability and argue against its overemphasis. Still, they predominantly focus on the roles human agencies play in processes of vulnerabilization. But what role do nonhuman agencies play?

Powerful Classifications and Vulnerability as Intra-Action
The discussion of the nonhuman agency has unfolded in recent approaches of new materialism on a variety of topics (Barad 2007;Braidotti 2013;Haraway 2007). In its line of thought, actors-be they human or nonhuman-do not have or possess agency but gain agency through particular intra-actions and agential cuts (Barad 2007;Haraway 2007). Barad suggests replacing the term "interaction"-which implies that preexisting and preestablished bodies or entities participate in action with each other-with "intra-action," a term which emphasizes that agency is not an inherent, determinate property of an actor but a "dynamism of forces" (Barad 2007, 141). The idea of intra-actions is not that the discursive and the material simply "meet" and interact, but rather that they are entangled and permanently co-constitute and coproduce each other (ibid.). Barad (2007, 14) stresses the aspects of human and nonhuman agency and understands agency itself as "enactment." While new materialism perspectives have been adopted in research on children and young adults-which has led to rethinking and deconstructing childhood and its traditionally associated categories (Murris 2016;Taylor 2013)-I propose analyzing vulnerability through an approach inspired by intra-action among human and nonhuman actors by examining how documents, policies, and spatiality are entangled with human actors in the production and stabilization of vulnerability. This requires an understanding of vulnerability as a phenomenon in which intra-actions take place and give form to vulnerability in the first place. In addition to being inspired by concepts and perspectives summarized under new materialism, the perspective I develop is also informed by Bowker and Star's (1996) analysis of classifications and categories. They demonstrate how classification systems can shape social interactions and worldviews, carving out the role classifications and categories play in ordering the modern world. When bringing together their understanding of categories with perspectives from new materialism, the "UAM"-category does not simply exist independently of time and space. The often-associated notion of vulnerability is, then, not an essential quality of young refugees but is given form (Haraway 2007) within these intra-actions. Vulnerability, as such, is not simply a "matter of fact," but a "matter of concern" (Latour 2004), that is, vulnerability is not just something measured and quantified but is closely interwoven with the lives and daily routines of young refugees. It, therefore, promotes a relational understanding of vulnerability that does not reproduce naturalized, oversimplistic, and biologistic conceptions thereof. In addition, it furthers the debate by addressing the anthropocentrism dominant in debates around vulnerability.

Refugees and Vulnerability-Making in Malta: Historical Context, Contemporary Dynamics, and Ethnographic Fieldwork
The theoretical contours discussed here call for situating the interpretation and emergence of refugees' vulnerability in the context of Malta, an islandstate, which is characterized by a long history of emigration (Grigg 1980, 25) and immigration (Ciappara 2013;Okolski 2012). At times, Maltese were colonized and became subalterns-the island-state was a British colony between 1814 and 1964-at other times, Maltese became colonizers, for example, when they settled in Algeria to stabilize French colonial power (Smith 2006).
These ambivalent dynamics have played a role with regard to refugee reception since Malta's independence in the mid-1960s, and Baldacchino (2010) emphasizes that conditions of refugee reception mesh with its colonial past and postcolonial present. Today, Malta is the only country among all Western Mediterranean states that receives refugees which was itself at one point colonized, and it now deals with refugees who have left a postcolonial African continent (ibid.). Since the 1970s, the trend toward immigration to Malta has become apparent, with more people arriving than leaving. The first case of refugee arrival for the independent Maltese government occurred when they agreed to house Asian refugees who were expelled from Uganda in the early 1970s. In the 1990s, Malta offered refuge to people who fled the Gulf War, as well as offering protection to victims of the Kosovo War later in the decade. These people were considered to be victims of the war and therefore vulnerable and in need of help. During each of these episodes, Maltese society understood itself to be hospitable, and refugees encountered a welcoming and supportive environment, which had to do with the fact that the number of refugees was relatively small and agreements were in place administrating their temporary refuge (McAdam and Otto 2020).
Since the mid-1990s, people have been arriving by boat from Africa, and the island-state has experienced what is termed "boat migration." Current forms of "boat migration" differ in various ways from previous migration: People who arrive from Africa do so in much more spontaneous and less controlled ways compared to refugees from the Kosovo or Uganda, and they are less likely to return to their countries of origin. In addition, Malta became a member state of the EU in 2004; ever since it has played an important role in "fortifying" its external border (Mainwaring 2019). The Mediterranean islandstate became part of a broader enforcement archipelago of refugee detention (Mountz 2011). With its accession to the EU, Malta has been involved in supranational agreements such as the Dublin Regulation or the EURODAC system. The latter is the EU's database for identifying asylum seekers and persons who cross borders without the necessary documents. Ratified in 2003, it is closely related to the former, the Dublin Regulation, which determines the member state responsible for examining asylum applications. Typically, the state through which the asylum applicant first enters the EU-where the person's fingerprints were taken and are saved in the EURODAC systembecomes the state responsible for processing the asylum application.
"Boat migration" is often viewed to be a new form of colonization, and I repeatedly heard during fieldwork that African refugees were now taking possession of Malta, replacing Maltese culture with their own, and destroying the island's identity. Recent opinion polls are revealing: 63% identified migration as a problem, less than one-third of respondents viewed it as culturally enriching, 79% of survey respondents believed that migrants worsen the crime problem, and Maltese society is the least willing of all EU societies to interact with them (Debono 2018). Discourse analysis (Falzon 2012) illustrates that the general view of refugees renders them as transient, as not "fitting" due to cultural differences, and of contributing to overpopulating the island (Author and Author Year). In contrast to the 1970s and the 1990s, refugees who arrived by boat more recently have been considered to be much less vulnerable and deserving of support (Author and Author Year), which also applies to those considered to be underage. But who are these young people, what have they experienced before arriving in Malta, how have they encountered the border regime, and how have they become (vulnerabilized) "UAMs"?

When Elais Arrived: An Ethnographic Vignette on Becoming "UAM" in Malta
In 1994, Elais was born in Mogadishu, Somalia's capital, to a single mother. As a young boy, Elais began helping her sell donuts-Kac Kac in Somali-in the streets and at the local market. He passed the money he earned to his pet monkey, Danyeer, who hid the income in his mouth so no one could steal it. One day in 2012, as Elais was walking around his neighborhood and was enjoying the evening chill after a day of hard work in the burning sun, he witnessed a life-changing event: The rape of a young woman. He recognized her tormentor: a respected, yet feared man from the neighborhood. The perpetrator saw Elais before he could run away. Elais was afraid-afraid that something would happen to him because of what he had witnessed. He did not want his mother to be dragged into the story and so he decided to run away. Elais made it to Ethiopia, where he kept his head above water with different jobs. When he had earned enough money, he moved on and arrived in Libya after an exhausting Sahara crossing. There he was imprisoned but managed to escape. He learned that he could not return to Somalia and that crossing the Mediterranean to Europe was the only way to improve one's life. For $800, Elais heard, he could make it to Italy. He managed to earn the money and decided that he would attempt the boat crossing. But Elais landed in Malta, a place he had never heard of, where the Immigration Police conducted short interviews upon his arrival. Elais was asked about his place of origin, his health, and his age. After having provided police with that information-place of origin: Mogadishu; health: in good condition; age: 15-he was brought to a detention center, which was set up for the management of newly arrived migrants. Even though Malta adopted EU Directives in 2005 which held vulnerable persons, including "UAMs," should be released from detention quickly (Pace et al. 2009), Elais, like most young refugees I met, was detained for several weeks. While detained, the authorities conducted an age assessment. In conversation with me, Elais remembered that he was brought to a hospital for an X-ray of his wrist and psychosocial interviews. Police drove him to his appointment, his hands handcuffed and his feet in nonmatching shoes. People looked at him as if he was a criminal, he said, and he felt sad and disgraced. Elais claimed to be 15 years old, but-based on an age assessment procedure-the Maltese authorities concluded that he was 16 years old. Following age assessment, he was considered to be an "UAM," and accordingly housed in a state care facility, where we met in 2013.

Approaching Vulnerability-Making Ethnographically: Methods and Research Project
EU policies often follow innate understandings of vulnerability and render young refugees like Elais to be among society's most vulnerable members (European Parliamentary Research Service 2021). But "vulnerability" is not translated and practiced in uniform ways in everyday life (Belloni 2016a(Belloni , 2016bLems et al. 2020;Otto 2020a;Punch 2002;Terrio 2008). Young refugees' arrivals reveal tensions and controversies: on one hand, they are viewed as vulnerable, wounded persons who deserve and need special protection and care services; on the other hand, it is often assumed that they are not inherently vulnerable children or youths Otto 2020b). Ethnographic research enables one to trace how vulnerability emerges situationally. But what does it mean to study vulnerability ethnographically in a border regime?
Before answering that question, a detour into the ethnography of border work is worthwhile. Border studies in general-and ethnographic approaches in particular-increasingly recognize the importance of everyday practices in the management and control of human mobility and borders (Isleyen 2018, 851;El-Qadim 2014;Mountz 2004). Nick Gill and Anthony Good (2018) make clear that ethnography is an especially fruitful method to study the often-messy character of border work, as it enables researchers to grasp the contingent, inconsistent, and unjust processes within asylum regimes, with Mountz (2011, 58) concluding that the state is a "series of performances." Scholars have taken seriously the production of borders and mundane border work (Côte-Boucher 2014) among humans and nonhumans, highlighting how borders are made real, and thereby contributing to its understanding in socio-material and socio-technical terms (Sontowski 2018). In addition, research on the EU border regime and processes of borderization demonstrates that "everyday actors," such as volunteers, social and care workers, or migrants, can shift how bureaucratic and legal categories or classificationssuch as "UAM" or vulnerability-are defined and practiced in everyday life (Coutin 1998). With that in mind, what better place could exist than a state care facility to house "UAMs" to ethnographically approach their vulnerability-making?
Until 2015, the Maltese government housed refugees classified as "UAMs" in such facilities in which the residents were supervised by care workers around the clock. Regarding Malta, the European Migration Network states that these homes should "provide a safe residential setting, education, preparation for employment, cultural orientation and leisure activities" (Pace et al. 2009). My initial fieldwork spanning six months commenced in February 2013 when I was granted access to one of these state care facilities. Revisits in 2015 (four weeks), 2016 (three weeks), and 2018 (one week) followed my initial research. My project was guided by Marcus' (1998) idea of following, which led me to conduct fieldwork not only in the facility but also in nongovernmental organizations (NGOs), ministries, refugees' homes, in bars, buses, and sports clubs. In total, 48 young refugees categorized as "UAMs" and 17 young refugees who were not considered to be underage participated in my study, as well as 22 actors with institutionalized positions, such as ministry employees, politicians, NGO officials, or care workers. My material consists of narrative interviews with non-refugees, such as NGO volunteers or ministry employees, informal talks as well as participant observation. To better contextualize the collected ethnographic material, legislative texts, NGO reports as well as newspaper articles complement the corpus of material.
The young refugees who lived in the state care facility when I began fieldwork predominantly originated, like Elais, from Somalia, while a few of them fled Ethiopia, Eritrea, Gambia, and Sudan. Most of the young people self-identified as men (85%), the others as women, and one person as nonbinary. Some of them originated from higher income families and participated in formalized education in their countries of origin. Others lived on the streets of Mogadishu before they left, earning money for their flight along the route, for example, as day workers in Sudan or Libya.
In the facility, these young people met Maltese care workers and facility management, which I refer to as the "Local Migrant Aid Organization" (LMAO); both made my presence as a researcher in this environment possible. The staff at the center saw me primarily as a relief in their daily work, as the care and administration of my research partners was characterized by chronic understaffing. Employees often called in sick or were fired. Refugee work is challenging in Malta, as it is not very respected; employees frequently faced hostility from friends and family for the work they conducted. I was given the role to act as a supervisor in the facility's computer room and was charged with organizing outside activities, teaching residents how to cook, and assisting them with bureaucratic work. During fieldwork, I was-especially when I tried to organize outside events but also when I talked to staff members-repeatedly confronted with anti-immigrant sentiments, which is also reflected in the surveys on attitudes toward refugees cited above, which the following entry from my diary, written in May 2013, illustrates: The day after tomorrow marks an important day in European soccer: Dortmund and Bayern play against each other in the Champions League final. The residents would like to watch the game and asked me if I could accompany them. I offered to find a suitable bar and looked for one in Hamrun, which is within walking distance from the facility. I entered a bar at Church Square and asked if I could reserve a table for 10 people for the day of the game. The bartender asked me with whom I would come, and I said with residents from the accommodation for underage refugees. He informed me that it was not possible; I related his scepticism to their being minors and promised him that we would not order alcohol. That was not the problem. He could not guarantee that it would be a safe place for them because they were black. My attempt to reserve a table went nowhere.
In the end, we went to another bar where many non-Maltese people were present. That I organized such activities led staff, at least early on in my research, to view me as a colleague. My already powerful position in society as a white European woman was bolstered with me having the keys to the computer room, knowing the Wi-Fi password, which was to be kept secret, and being allowed to look at the files generated about the residents. It thus comes as no surprise that the young refugees first approached me with scepticism: Would I be someone they could trust? Would I reveal intimate information to care staff? Would I really be willing to listen to their stories? It took several weeks of my presence before I was approached with more trust; later, I was often called "sister," only being a few years older than some of the residents. The young refugees hoped that I would report in public on the situation in Malta and generate attention for their discontents. In addition, I listened and often tried to help them find solutions for their problems, for example, when homelessness was imminent once they turned 18. And even though one manages to gain trust in such circumstances, research with young refugees poses specific questions and challenges.
Is it responsible to conduct research with those classified as minors? What must one consider ethically? As argued elsewhere (Chase et al. 2020), while I have adhered to standard ethical codes in my work-all research participants were informed about my research and consented to participating, either verbally or in a written form, and all actors were given pseudonyms-existing ethical guidelines are overly formalistic and should not only be viewed as a tick-the-box process to receive research permission. Bragg (2021) argues that research ethics boards tend to perpetuate an oversimplified image of refugees as being vulnerable per se, which leads to science becoming unable to uncover "real life [. . .] vulnerabilities" (ibid.). The "vulnerability doctrine" (ibid.) ultimately leads to the exclusion from research of people designated as "vulnerable" and contributes to silencing refugees in its attempt to "protect" them. Instead, "ethics needs to become more integral to the life of the whole research endeavor, forcing us all to constantly think about the consequences of how we conduct ourselves in the pursuit of knowledge" (Chase et al. 2020). There is no perfect solution in handling power imbalances and young people's needs, but Rodgers' (2004) concept of deep hanging out proved to be suitable to build relationships, to listen to young peoples' stories carefully, and not to confront them with the formalized situation of an interview again (ibid.). My aspiration was to use ethnographic methods in ways which do not harm my interlocutors, and I tried to accommodate their preferences wherever possible. To do so, I explained the ethnographic toolkit and let them choose how to proceed. Informal conversations and observation were the preferred methods among refugees. I was always aware of the danger of possibly retraumatizing them and made sure to let them speak to me whenever they wanted to, instead of asking questions related to their flight. These informal conversations usually came about when we spent time together in the computer room, on the beach, or in my apartment.
When I began my research, I aimed for a balance between the actors-it was not my goal to take a position in the field. This intention turned out to be naïve and unfeasible. Ethnographic research in these powerful and often hostile environments also implies becoming a witness of what happens on the ground-in my case that included incidents of police violence against residents inside the facility, of kicking out residents without money or food because their beds were needed for younger arrivals, of using racist and patronizing language toward residents, etc. Over the course of my research, I departed my intended neutral position and began criticizing the management, the conditions under which the young people were housed, and filed reports at different EU institutions in charge of child welfare. As time proceeded, it is no surprise that I was viewed by staff less as a colleague but was understood as a troublemaker instead. The residents, however, valued my attempts. My behavior led, in the end, to what Becker (2001) calls "exclusion from within" (my translation).
The exclusion from within led by no means to the end of my research. I had established enough contacts and built strong relationships to enable me to continue my research for years to come. However, the state care facility was essential as a starting point for an ethnographic study on vulnerability, as my access enabled me to observe and learn about situational, more-thanhuman vulnerability-making. It was my experience in the facility that irritated my own understanding of vulnerability and underage. I observed how vulnerability becomes enacted situationally and that an innate understanding is not sufficient when we aim at developing differentiated understandings of border work.

Young Refugees' Enacted Vulnerability: The Intra-Action of Regulations and Spatialities
Given the tendency to study vulnerability in static terms and to underemphasize processual approaches to studying vulnerability, the perspective adopted here explores its enactment. To carve out the proposed relational understanding of vulnerability and how it is produced and stabilized, it is worthwhile examining how intra-actions make vulnerability more pronounced. In the Maltese case, this implied that both EU and domestic policies as well as the country's insularity played significant roles. The social and the material are inseparable, and vulnerabilization is, much like other border work (Reeves 2014), socio-material (Sontowski 2018) and relational.

Dublin Regulation and (the Lack of) Travel Documents
The Dublin Regulation and the obligations it entails for the Maltese state were recurring topics during research. The impression was often that Malta was left alone with the arrivals of boat refugees. Pelizza (2021) discusses that the Dublin Regulation contributes to and has given rise to criticism of "geographical unbalances." John Mitchell, who worked for a ministry in Malta, explained during an interview to me in the summer of 2016: "Asylum seekers and migrants do not have free movement within the EU. Their movement is limited by the Dublin Regulation [. . .]. Even once you are granted some sort of protection by a member state, in this case Malta, you are only entitled to the rights of this state and not elsewhere. That is something migrants often misunderstand. Yes, Malta is part of the EU, but you only have certain rights here, Maltese rights so to speak." The Dublin Regulation "fixates" refugees geographically and legally. The states along the EU's external border do not become transit countries as many refugees and Maltese citizens had hoped, but rather places where they must remain. The Dublin Regulation restricts refugees' mobility significantly. Asylum seekers cannot leave Malta in legal ways if they have applied for asylum and the application is still being processed-which can take up to several months. After arrival, it is not only the Dublin Regulation that comes into play, but refugees' fingerprints are also taken and saved in the EURODAC system. Several scholars have shown how these databases "do borders" (Kuster and Tsianos 2016). Pollozek and Passoth (2019) argue that technologies and databases, like EURODAC, not only realize the data transfer among different member states and make information about refugees available across the EU, but that they also assign a legal status, assign the state's responsibility, and assign refugees to specific places in which they can stay (temporarily). These databases and regulations help, as Pelizza (2021) argues, to channel, control, and manage refugees.
Both the Dublin Regulation and EURODAC exert power over refugees' mobilities, and so do the documents refugees (do not) receive. Cabot (2012) emphasizes that documents not only exert agency on the state's regulation of refugees but also on their own conditions of movement and freedom. Once refugees are granted a status of protection in Malta, their possibilities to move change: Recognized refugees receive a convention travel document. People eligible for subsidiary protection or whose asylum application was rejected, receive an alien's passport. With that document, they can travel to countries that recognize this document. They can move more freely and can also stay, up to 90 days, in other EU countries; still, as mentioned above, certain legal rights are only accessible for them in Malta, but these documents transcend the Dublin Regulation temporarily.
While adult migrants who were granted a status can travel within the EU, the situation for "UAMs" is fundamentally different. Here, their age intersects with their (im-)mobility and a lack of documents. These refugees do not receive an alien's passport or a convention travel document as long as they are not of legal age. Instead, the young refugees I conducted research with received an identity card with their name, gender, and country of origin. The ID card stated in capital letters: "Not eligible to travel." I discovered their inability to travel through coincidence: In April 2013, I had to register with the Immigration Office in Malta, because I had been in the country for over 90 days. At the authority, I met Warsame. Together, we went through our papers and forms and I saw on his ID card that he was not allowed to travel. This forced immobility is primarily because the Maltese state is responsible for these refugees who are considered vulnerable minors. Disappeared minors have repeatedly led to debates in the EU, and state authorities are concerned with preventing their disappearance and possible abuse. For the young refugees, however, it meant that they were forced to remain in a country considered the least open toward refugees in the EU (Debono 2018): they were not allowed to seek out family members in other EU countries and had no access to EU-wide migrant support networks. "UAMs" (ascribed) underage manifests itself on their ID cards, and when they interact with border guards, they are easily identified as "UAMs" and required to stay in Malta. Their ID cards, like other documents, bring together space and place (Pollozek and Passoth 2019). The materiality of the ID card ensures that underaged refugees who try to leave Malta stay put, and border guards and other actors from within migration management do not need to engage with their stories or reasons why they wish to leave Malta. Instead, the ID card reduces them to their ascribed vulnerability, renders other information invisible, and inscribes the idea that "children" should not travel by themselves. These ID cards play a crucial role in border work and the stabilization of vulnerability: Moving elsewhere would have meant for many of my interlocutors that they could have left circumstances of vulnerabilization (Raghallaigh and Thornton 2017) in Malta, as they, for example, had family members in other EU countries who were willing to house and support them.
Absimil, a young refugee from Somalia whom I first met in March 2013, after he was released from detention, was granted one of these ID cards. He mentioned to me in a conversation in 2015: "That document is a big problem. Because everybody can see that I cannot leave Malta. So what chances do I have? Because staying is no option, and I don't want to wait my legal age. I want to finish Malta now." Absimil did not necessarily consider himself to be "vulnerable"; instead, he often told me that he views himself as a person who still has to learn a lot. He argued that he cannot understand why he cannot leave Malta legally, as the state would not support him in his personal growth and development. His main goal was to participate in education, which was denied because he was considered too old (in Malta, compulsory education ends at the age of 16), and he was cognizant that he would be allowed to go to school in other EU countries.
ID cards, which "tell" others about one's age and travel restrictions play decisive roles in vulnerability-making of young refugees. Their ID cards are a good example which illustrates how ascribed vulnerability of being underage materializes in these documents, as well as highlighting the card's agentive role. They not only "tell" others about a person's legal age, but they also fixate young refugees in Malta. Vulnerability is thus not only a bodily feature or an ascription of being underage but is produced and stabilized by documents (not) granted to young refugees. It suggests that vulnerability does not have a definable ontology. If we think of vulnerability as intra-action, then we can see how the agencies of documents, border officials, and rights enact a specific understanding of vulnerability of young refugees in the EU's border regime. It is the interplay of the ID cards' materiality and border guards' response to the document which forms their interaction with "UAMs." Social norms and interactions in the border regime cannot be thought of as being separate from the materiality of nonhuman actors.
Despite efforts of keeping young refugees in place, some of my research participants found ways to leave Malta while still classified as "UAMs." One was Yahya, who left Somalia in early 2012 and arrived by boat in Malta during November of the same year. Malta served as his entry point into the EU, and with its logistical setup including EURODAC and age determination, it became the place in which data was produced and stored about Yahya. These technologies and data locate, sort, and monitor refugees, and help to move control away from territorial borders alone to refugees themselves who are assessed, registered, and become "traceable objects," as Pollozek and Passoth (2019) show. Yahya, as I show below, became one of these.
He was, like others, very frustrated with his situation in Malta, as he was not allowed to go to school or find work. He heard from others that Germany offered a better life: that one could attend school, and that one was treated better. Yahya succeeded in travelling to Germany; during the days before his departure, I observed him being particularly friendly to the facility's staff: he always greeted them, asked them about their well-being, and emphasized how much he liked Malta. I already knew that he was planning to leave-he had secretly moved his belongings out of the home and organized his departure. After his arrival in Germany, he registered with the authorities and was placed in a shared apartment for underage refugees. Upon arrival, German police found his fingerprints in the EURODAC system, but they said that he did not need to worry about it for the time being. Yahya, who was considered to be 16 years old, started going to school, gradually made friends and told me that he never thought about going back to Malta. Shortly after his official 18th birthday, however, the authorities contacted him: Germany was not responsible for him and since he was now of legal age, he had to return to Malta. He was brought back in 2016 and was then considered a so-called "Dublin case." Yahya's story is a prime example of how databases and technologies enact multiple spaces on the local, national, and European levels. His story not only reveals "the power of infrastructures" (Lin et al. 2017) but also shows how border control and refugee management actors gain agency through intra-actions with nonhumans (Barad 2007): EURODAC helped to remind German authorities of who Yahya is, where he should (not) be, and who is (not) responsible for him (Pollozek and Passoth 2019). Beyond demonstrating how border work "works"-of illustrating how vulnerability is created rather than inherent (Grigg 1980)-Yahya's story also reveals what these border enactments do to young refugees: "In Germany, my life was better. The housing situation was better, and I found a school. I really felt that I can start my new life there." I knew Yahya as a cheerful and positive-thinking young person: he cracked jokes on outings, he liked to lead a team when bowling, and he often made everyone laugh. After the forced return, things changed: our meetings were marked by Yahya's noticeable lack of zest for life, he appeared distracted, and sometimes he cried.
Yahya is only one example of many. Being viewed as a minor in the EU's border regime both includes and excludes, makes one mobile and immobile at the same time. To a certain extent, Yahya had taken his ascribed vulnerability with him to Germany, and it allowed him to stay for slightly more than two years, as authorities did not want to return a minor. Here, vulnerability was stabilized even across different member states: While minors were officially not allowed to leave Malta, those who managed to reach another member state were, however, protected by their ascribed vulnerability. But when his ascribed vulnerability no longer applied in official terms, he was forced to return to Malta.
Yahya's and others' experiences illustrate that being categorized as a minor leads to different spatial experiences compared to those of adults: Where he was (not) tolerated in the EU depended significantly on his age. The Dublin Regulation is, as Yahya's case shows, not only entangled with space and mobility but also with age. While he was tolerated in Germany as long as he was considered a minor and others ascribed vulnerability toward him, he was returned to Malta once he turned 18; a situation, in which vulnerability was not further stabilized. Both the implementation and the suspension of the Dublin Regulation create a specific border space, which significantly contributes to vulnerability-making. Spatialities become efficacious in various ways and are closely related to (nonexistent) identity cards, supranational agreements, and attributed age. Vulnerability-making is therefore inextricably linked to other agencies and goes beyond individuals' bodily features. As other standards and mechanisms apply to the case of adults, it is fair to say that classification regimes and regulatory mechanisms play crucial roles in the (de-)stabilization of vulnerability. Reflecting on the "behind-thescenes machinery" (Bowker and Star 1996, 158) of classifications and categories highlights that vulnerability is a phenomenon in which EU policies, the "UAM"-category, and notions of vulnerability intra-act and thus give form to what is often understood as the inherent vulnerability of young refugees.

Insularity and Immobility
Young refugees' lack of travel documents means that they have little chance of leaving the island. However, there is another central actor which ensures that minors "stay put": The physical borders of the island-state itself play a central role and are part of enacting young refugees' vulnerability.
As an island-state, Malta has no mainland to which refugees could move as the coasts of the island constitute the state border. These coasts take on agentive roles in stabilizing vulnerability among young refugees along the EU's external border. In comparison to other countries, it is more difficult to leave Malta if one does not have any or the wrong documents. Mountz (2011) notes that EU islands play a crucial role in channeling refugees, arguing that islands have become part of a broader "enforcement archipelago of detention." Islands' geographical features and boundaries, such as cliffs, water, and beaches, help move border control away from their fringes to their centers, where multiple border sites emerge. These features help humans avoid having to control all physical borders (Pollozek and Passoth 2019) and can instead focus on making refugee tracking much more efficient.
Two of these border sites in Malta are the seaport and the airport. Both are subject to intensive controls; black people, in particular, are repeatedly asked to show their documents, displaying how intensely the EU's borders are racialized (Hess et al. 2017;Mainwaring 2019). While Absimil was affected by the racialization of the border, other factors also played a role. He told me that he tried to leave Malta twice. His first attempt consisted of trying to leave the island-state by boat to Italy, but when he arrived at the port early in the morning on a summer day in 2014, he realized that police controls were present along the pier and by the ferry. As he was too afraid of getting in trouble, he decided not to attempt to enter the boat to Sicily. Believing he did not have bright prospects in Malta, he attempted to leave again and booked a flight to Germany. In conversation with me in 2015, Absimil shared his experience at the airport in Malta: "They let everybody pass in the security queue. So I thought I am lucky today. But to me, the security lady she said: 'Wait.' She check again and again my passport. So then the police came. And all my luck was gone. You know, I was not lucky. That day my lucky [sic] was dead. And I could not run away." A court case for Absimil followed as he had tried to leave the country with the passport of another person. At his court appointment, it was made clear to him that if he tried again, far-reaching sanctions would ensue, and he was charged with community work. The physical conditions of the island-state direct agency toward human actors-in this case a border guard. Refugees in Malta must always make contact with border guards to leave the country, which exemplifies how a "dynamism of forces" (Barad 2007) unfolds among human and nonhuman actors in border work.
In 2018, Absimil still lived in a shared apartment in the town of Hamrun. He never attempted to leave the country again. He said: "You know, you can't just disappear here. Elsewhere you can walk across borders or drive yourself, but I can't swim away from here. All that remains is the harbor or airport, and I'm too scared now". Absimil's experience illustrates that his immobility does not only relate to insularity but also to the fact that Malta, as an islandstate, does not itself have mainland. It contributes to the stabilization of vulnerability, which is inextricably linked with immobility. Spatiality and how it contributes to vulnerability-making is therefore highly contingent. While it is clear that the security guard played a decisive role in Absimil's (im-)mobility, the stabilization of his ascribed vulnerability is also linked to the materiality of the island-state. The geographic features of the island and sea lead to circumstances in which he cannot easily escape.
It is not the coastal state borders alone which help to ensure that the young refugees remain where they are supposed to be legally; in addition, Malta's relatively small island-state spanning only 316 square kilometers leads to a situation of close proximity and an atmosphere characterized by an everyoneknows-everyone sentiment. When young refugees left the care facility, they were asked to tell the care staff where they were going, with whom they had an appointment, and what their plans were. The young refugees' plans were, as I observed, noted in a green report book. If their plans changed spontaneously and they were seen on the island by off-duty caregivers, they often checked with those on shift to see if the refugee in question was permitted to be outside. Often when I spent time at the facility, staff frequently went over the book's content with me to show me who had misbehaved in the past.
It illustrates that the entanglement of care staff, the green report book, and Malta's geographical features stabilizes young refugees' vulnerability. These dynamics led to young refugees feeling permanently observed. They, therefore, tried to stay in places they thought were safe and certainly found places they were able to visit, such as the streets around the port of Marsa, but they also knew moving across the island could lead to being observed. Several times during research I was told that no surveillance cameras were needed in Malta: The approximately 400,000 pairs of islander eyes obviated the need for official observation and surveillance. Malta's smallness meant that the minors were made observable and controllable, thereby stabilizing their ascribed vulnerabilities. Malta's geographical smallness and the way it was used played a central role in vulnerability-making. Elements in the making of vulnerability thus included the island itself, report books, registration and deregistration practices, and a neighborhood-is-watching-you atmosphere.

Conclusion
I have demonstrated that vulnerability is viewed as an integral part of the category of the "UAM," and it is closely associated with its official classifications within the EU's border regime. Young refugees under the age of 18 are often considered vulnerable, whereas those above 18 are not considered to be a particularly vulnerable group of asylum seekers. The categorization thus constructs young refugees to be vulnerable because they are minors. While I have been agnostic about the question of whether they are inherently vulnerable, my intention has been to ask how vulnerability is enacted in the border regime. To that end, I analyzed the intra-actions of documents, policies, and spatialities which produce and stabilize vulnerability. To do so, I have employed an epistemic lens rooted within the theoretical context of new materialism, illustrating that vulnerability is a phenomenon in which intra-actions take place and give it its form in the first place. I have demonstrated that vulnerability is not a pre-social characteristic of young people but is (co-)constituted in entangled human and nonhuman relations. While the concept of vulnerability often remains abstract or simplistic, these empirical explorations provide a more complex understanding thereof. My aim has been to write against the "black boxing" (Bowker and Star 1996, 158) of classifying young refugees and to look at "behind-the-scenes machineries" (ibid.).
By recognizing nonhuman agencies, it is possible to develop even more complex understandings of "border work" in the EU and move further away from essentializing representational approaches. It enables the development of differentiated and differentiating perspectives instead. To contribute to the debate, I situated enactments of vulnerability ethnographically and illustrated how different agencies stabilize vulnerability. While the Maltese state has a duty of care and tries to ensure that young people remain "in place," the laws, measures, and (nonexistent) documents lead to a particularly high degree of "staying put" of young refugees. It was not only immobilizations brought about by human and nonhuman agencies that stabilized vulnerability, but the intense proximity of different actors in Malta played an important role in stabilizing young refugees' ascribed vulnerability, too. A certain controllability of their spatial navigations has been coproduced through different agencies in play. Young refugees' vulnerabilities are entangled with domestic and international frameworks, as well as with geographic entities. Thinking with Aradau, the enactments of vulnerability are "an open-ended practice involving dynamic entanglements of humans and non-humans" (Aradau 2010, 498), and they do the "ever-local, ever-partial work of making [classifications] appear" (Bowker and Star 1996, 161). Examining the ever-local and ever-partial classification work through the empirical material revealed that materialities and artifacts-such as the island's geography or information displayed on young refugees' ID cards-often served as elements that had an impact on human agencies in either concealing or evoking them. As such, there cannot be an adequate verdict on the question of inherent vulnerability among young refugees. Yet approaches like the one employed here offer the prospect of recognizing the complex entanglements and processual nature of human and nonhuman agencies in the production and stabilization of vulnerability.

Declaration of Conflicting Interests
The author declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.