The affective border: Missing migrants and the governance of migrant bodies at the EU’s southern frontier

While living migrants are the object of enormous attention, counted and screened on arrival and throughout their presence in European states, those who die crossing the Mediterranean remain uncounted and largely unidentified, as a result of a failure of European states to collect data from both bodies and from families looking for missing loved ones. Living migrants are seen as objects of interest to and control of biopolitical regimes, yet, once migrants are dead, they become marginal to their concerns. The body of the migrant is neither fully inside nor outside the social and legal order, while family members waiting for news of a loved one are entirely invisible to it. The result of death at the frontier is that the border gains an affective dimension, with the capacity to communicate trauma over large distances, plunging into ambiguity families in states of migrant origin who are unable to learn if loved ones are alive or dead.


Introduction
Affect [. . .] is not a discipline of study of which the politics of affect would be a subdiscipline. It is a dimension of life [. . .] which directly carries a political violence. (Massumi, 2015: Preface) The so-called 'crisis' of migrant arrivals at the Europe's southern border in the last decade has seen thousands of migrants and refugees have drowned while seeking to cross the Mediterranean and enter the European Union (EU) (Last and Spijkerboer 2014). 2 Sovereign power in contemporary Europe lets migrants die at the border by framing their deaths as accidents, unrelated to the machinery of militarization and securitization that accompanies those deaths (Albahari 2006). That no comprehensive data around such migrant deaths are available reflects a commitment of the EU and its member states to refuse to enumerate such migrant deaths. The International Organization for Migration (IOM) has estimated that between 2014 and 2017 15,348 deaths occurred at the EU's borders (IOM 2018). In 2016, when the data presented here were collected, more than 5,000 are known to have died crossing the Mediterranean, with an additional untold number of unrecorded deaths. Kovras and Robins have noted the impact on families of deaths on this scale, noting that: For every body that is washed ashore at the EU border and not identified, there is a family living with ambiguity, not knowing if their loved one is dead or alive. For such families their loved ones are missing, having left home and not having been heard from since. (2016: 41) Without information about the fate of relatives, families live with the trauma of ambiguous loss. This article seeks to use a novel and unique dataset to further explore the challenge presented in the paper by Kovras and Robins. More than this, it aims to show how the border gains an affective dimension as it kills migrants whilst simultaneously denying the information to the families of the dead that would address their ambiguity and the trauma that results. 3 The problem is one of governance, and the impact is trauma: I seek to both describe how the migrant dead are governed, as well as the impact on the families of those who die at the border and are never identified, which is a direct result of that failure of governance.
Data from an empirical project is presented that has made interviews with both families of missing migrants from Tunisia, Syria and Iraq, and authorities and others concerned with migrant bodes in Greece and Italy, in 2015-16. It is seen that there are multiple failures around the collection and management of post-mortem data from the bodies of migrants, and insufficient effort to reach out to the families of the missing to collect ante-mortem data that could be used to identify them. The result is a population far from the EU -in countries of migrant origin -who see the trauma of deaths at the border transmitted into their daily lives, due to the absence of any information about the fate of loved ones who left for Europe. The border both kills migrants and acts as a barrier to the transmission of information that could deliver answers to families. The border becomes a highly extended object with the capacity to traumatise at a great distance.
For the dead governance ultimately has no impact, so here I will discuss the impact that governance of the dead has on the living, with a focus on the family members who remain unaware of the fate of dead loved ones. Whilst it is a given of secular modern ontology that the dead have no agency, anthropology shows that in most of human history the community of the dead has been seen as able to nourish, or haunt, the living. The migrant dead continue to haunt the living in the sense that the uncertainty over their fate has dramatic impacts on their family members, possessed of an agency that can traumatise.
The families of missing migrants are the victims of this humanitarian disaster and yet the impacts on them are unknown. They remain unrepresented in discussions about the management of bodies and the broader crisis, and largely unable to engage with the authorities who can identify their loved ones. This report uses data from the Mediterranean Missing research project (Ben Attia et al. 2016a;Ben Attia et al. 2016b), which collected data in interviews with 84 families who are missing loved ones, from Tunisia, Syria, Iraq and elsewhere. The study also engaged with authorities in both Greece and Italy, interviewing authorities, civil society and others to understand policy approaches to migrant bodies Mirto et al. 2016). 4 Governance and the migrant dead While death is usually understood in terms of emotional impacts on those close to the deceased, an emergent literature exists that examines the intimate links between death -and in particular the dead body -and sovereignty. In all political and cultural communities discourses and institutions have developed intended precisely to manage the ''object of horror and dread'' (Hertz 1960) that is the corpse, demonstrating that death and the rituals that surround it are highly social phenomenon that reflect society's investment in the person of the deceased. Conversely, the bodies of those unknown to the community inspire little or no emotion or ritual (Ibid.). As such, we see that governance, through both law and broader social norms, drives both attitudes and obligations to the dead: authorities govern dead bodies in ways that directly reflect how sovereignty is claimed. Bataille sees the fear of death and the powers of the dead body as having a key role in his conceptualisation of sovereignty (1991): the body as a symbol of transgression, most notably of the taboo against killing, sees sovereignty embedded in the corpse, emphasising the role of the authorities in protecting populations from death, echoing the biopolitics of Foucault in which power manifests itself in control over every element of life (Foucault 1991). Most concretely, this sovereignty is articulated through laws and procedures, as well as technologies such as forensic medicine, which are key to maintaining the state's claim to a monopoly of violence and the elimination of killing outside the law. Governance through death can be understood in terms of Mbembe's necropolitics (2003) in which the capacity to permit migrants to die at the border is a demonstration of sovereign power. Mbembe draws on Schmitt's definition of sovereignty as the power to decide on the state of exception, ''where the controls and guarantees of judicial order can be suspended'' (Ibid.: 24), as they are for migrant bodies where the legally prescribed processes of investigation are curtailed. This resonates with how Agamben perceives politics as ''an ongoing tension between inclusion and exclusion, between forms of life that the sovereign will protect and represent and those it will not (Agamben 1998)'' (Kovras and Robins 2016: 43): both the fact and the treatment of the migrant dead is an articulation of sovereign power.
The rituals around death are a way of confirming and affirming the existing social order: they are public events that communicate the departure of a member of a society from it. Socially, there is a clear divide between the living and the dead and a funeral or other death ritual serves to mark the The affective border 3 traversing of that boundary. Where someone's fate is unknown of course this clear divide is blurred, not just psychologically and emotionally, but socially. Some however have seen the rituals of death as not only concerned with symbolic order, but also with political order, since the contingency of events manifested pre-eminently in death poses a challenge to authorities that claim to be eternal and continuous, and so death must be negated through the mechanism of the funeral (Bloch and Parry 1982). This ''sacralisation of authority'' (Stepputat 2014: 20) is most clearly seen in the enthusiasm of authorities, including in secular Western states, to retrieve the bodies of their citizens, particularly where they are perceived to have died serving the state, such as in war. The seminal work of Verdery (1999), on the political lives of dead bodies in post-socialist Europe, has served as the basis of a literature that demonstrates how governance and sovereignty are articulated and contested through the corpse (Fontein 2014). Human remains have huge potentialities and are asked to carry the weight of a range of political meanings and reconstructed as specific kinds of political subjects.
The migrant body however sits in contrast with the citizen body, becoming a part of the legal order precisely through its constitutive exclusion, with migrant deaths depoliticised and framed as accidents, unrelated to the securitization of the border. The invisibility of migrant bodies at Europe's southern borders and the denial of their political subjectivity, can be understood in terms of Butler's concept of grievability (Butler 2004;. Such an analytical lens demands that we 'critically evaluate. . .the conditions under which certain humans are more grievable than others' (Ibid: 30). Mountz makes the link between the political invisibility of migrant bodies with the lack of value migrants are given in life (Mountz 2015:188): those not valued in life, are by extension not grievable after death. This serves to emphasise the contrast in political value in Europe between the body of a dead citizen and that of a migrant, and indeed between the body of a living and dead migrant. The dead bodies of would-be border crossers are framed as non-grievable by state authorities, and subject to almost no attention, while living migrants are perceived as a potential security threat and constantly surveyed: ''dead migrants are ignored and the management of their bodies circumscribed by legal and bureaucratic ambiguity'' (Kovras and Robins 2016: 4), defaulting to local authorities.
Death has been hypothesised as the limit of biopolitics. Whilst we understand living migrants as objects of interest to and control of biopolitical regimes and conceptualize their deaths as vital to how sovereign power and the nation state operate, once migrants are dead they become marginal to the concerns of the biopolitical. The dead are subjected to a distancing from the living for reasons of purported public health and aesthetics, driven by the need to control leaking and putrefying bodies, but beyond this the body is often an object of disdain wherever it is found. 5 This is of particular relevance for the body of the migrant that occupies an indeterminate space that is neither fully inside nor outside the social and legal order. The techniques of governance applied to migrant and refugee bodies in contemporary European Union states resonate with totalitarian histories in which those who died in systematic exterminations -notably in colonies of European powers and the Holocaust -were disposed of without dignity. Depriving the dead of proper disposal denies them the most basic sign of being human, emphasising the exclusion that characterises attitudes and policies toward living migrants.
The political impact of the transgressive treatment of the bodies of the ungrievable migrant dead is precisely that they are unknown and neglected. Only rarely, and temporarily, are migrant bodies made politically visible, as for example when the washed up corpse of the three year old Syrian boy, Alan Kurdi, appeared on newspaper front pages globally. However, even this shocking image has not been permitted to change either dominant narratives or policy in the EU, and migrant bodies have continued to arrive largely unseen on European beaches. The articulation of the Mediterranean border that emerges from a study of such deaths supports an understanding of a performative model of the border (Salter 2011). Migrants and agents of the state seeking to control their movement ''enact and resist the dominant geopolitical narratives of statecraft as they cross, or are prevented from crossing, borders'' (Johnson et al. 2011: 66). The border is thus created by three elements: its formal territorial demarcation, the politics of how movement across it is controlled and popular performances of the border which include the ''overtly public and political contestation over the meaning of the border'' (Salter 2011). Border deaths represent a particular manifestation of all these elements, demonstrating borders as enacted and performed as ''emotional landscapes of social power'' (Johnson et al.: 62).

The EU's southern borders and the management of migrant bodies
The issue of the migrant dead is one of governance, and whilst it might be imagined that governance concerns -at least primarily -the living and not the dead, here it will be argued that governance and sovereignty are articulated and contested through the corpse and that the treatment of the dead can have dramatic implications for the living. For families to address their need to know what has happened to loved ones demands both that authorities make efforts to identify bodies and that they families are aware of what action they need to take. Here the response of the Greek and Italian authorities to migrant bodies is discussed, in terms of how the dead and the data collected from them are managed and what efforts are made to identify them. It demonstrates a policy vacuum at state and EU levels, with no concerted response to the problem. When a body is found in an EU state, both national and international law create obligations for states to investigate, including but not limited to taking testimonies from witnesses, preparing forensic reports, medical examination and collection and storage of DNA profiles (Romano, 2016;Kerasiotis and Spiliotakara, 2016). It is the absence of such an active The affective border 5 investigation that most demarcates the difference between how a migrant body and that of a presumed citizen are treated.

Mapping Policy Responses
The two key entry points for migrants crossing the Mediterranean into the EU are Lesbos, which received 490,000 migrants between 2015 and mid-2016, more than six times the total population of the island (UNHCR 2017), and the Central Mediterranean island of Lampedusa. While Lampedusa received only a quarter of almost 1.5 million people who arrived since 2014 on all routes, it accounted for 85% of all migrant deaths in the Mediterranean (ibid). Since 2014 an estimated 11 per day have lost their life in shipwrecks in the Mediterranean Sea (IOM 2018), and an unknown number have died but their bodies never been recovered. This sets the scene for a transnational challenge that affects the lives of thousands of families in countries of migrant origin who live with no knowledge of the fate of loved ones.
Successful identification of bodies retrieved requires the systematic collection of different types of data and their management, bringing together postmortem data from the body (distinguishing features, material effects such as SIM cards and wallets, DNA data, witness testimony) with ante-mortem data from the families of the missing (a description of the missing, DNA data etc) (Ben Attia et al. 2016a). Both the international and national legal frameworks impose a duty on national authorities to investigate 'non-natural deaths' (Grant 2016), yet: in practice there is little active investigation, with even easily accessible personal effects (e.g. SIM cards, notebooks, credit cards) often left on the beach [. . .] Collecting ante-mortem data from families is one of the most challenging tasks, as it demands outreach to families of the missing, including potentially through cooperation with relevant state authorities in countries of migrant origin. (Ben Attia et al. 2016a: 6).
Experiments have been made in using social media, such as Facebook pages, to successfully reach out to missing relatives (Reidy 2017), but these raise concerns about data protection, confidentiality and potential security threats to those involved. In the absence of any data from bodies found, social media spaces dedicated to the issue are a tumult of photographs and descriptions of the missing from desperate families, accompanied by rumour and the presence of those prepared to extort money in exchange for fake information.

Identification
The vast majority of identifications currently made in both Italy and Greece are made visually. This is driven by the need only for a relative to be present who can identify either a body directly or from photographs and demands no international outreach. It constrains such identification however to those cases where a relative was either traveling with the deceased and survived 6 Simon Robins or can quickly come to the scene. (Ben Attia et al. 2016a) In both Lesbos and Sicily, photographs are taken of the body and an examination carried out by coroners or general practitioners, to collect information on height, weight, presumed age, and distinguishing marks such as tattoos or scars. Identification can then be made either visually, through relatives examining photographs or bodies, or using primary identifiers, such as DNA sampling, odontology or fingerprints. In many cases an autopsy is not made, since the cause of death is usually assumed to be drowning. The situation in Greece is complicated by the fact that the 'new' Muslim cemetery in Lesbos, where most bodies are buried, is managed entirely by an Egyptian volunteer, with no state engagement. ) It remains a concern that there is no route to identify a given set of post-mortem data with a particular gravesite, and thus even where a DNA-based identification is made no way to return human remains to a family. Identification based on DNA data is anyway rare in both contexts due to the difficulties for families to access or even to be aware of the relevant institutions with which to share antemortem data. (Ben Attia et al. 2016a) In Greece DNA data from all unidentified bodies are centralised at the Forensic Science Division (FSD) laboratory in Athens, a branch of the Greek police which is responsible for storing all genetic samples and for carrying out any future identification . Although in theory this means that relatives can send samples from their countries of origin to the FSD through an official agency, or through the Athens embassy of their country of origin, there is no outreach and so families in countries of migrant origin are unaware of the possibility to share tissue samples for DNA analysis.
In Italy DNA data remain dispersed, with tissue samples routinely taken from all bodies and stored in different locations depending on the institution carrying out the examination. As such, the data are scattered but this is not necessarily perceived as problematic as long as the coordinating office of the Commissioner knows where the data are. In practice, the lack of ante-mortem data with which to make comparisons, means this system is rarely tested. A Special Commissioner for Missing Persons is responsible for three high-profile cases of shipwrecks with hundreds of deaths, which benefit from the forensic capacity of a coalition of academic institutions. The capacity to make identifications, even with such state of art forensic capacity has however also been limited by a lack of access to ante-mortem data, with only 8.5% of some 365 recovered bodies having been identified (Oliveri et al. 2018). As long as families are not considered actors who can contribute, identifications will remain few.

The impact of having a missing relative
Death ritual serves two purposes relevant to this discussion. It serves to separate the living and the dead, transforming persons into non-persons, into corpses, ancestors and spirits, but also entails a restatement of social The affective border 7 order with the intensity of mourning correlated to the social status of the dead. Hertz portrays death ritual as a process of parallel transformation of the dead body, the mourners and society (Hertz 1960). The lack of mourning for the migrant dead, the absence of emotion, occasion and ritual, demonstrates the position of the migrant in the society in which he dies, with his or her passing having no impact. Secondly, death ritual has an affective role for those close to the dead person, serving to delineate the boundary between the living and the dead and creating both collective spaces for mourning and social meanings for family and community. Funeral rites serve to support the coming to terms with someone's passing, and redefine the family and society as now excluding that individual. As such, the rituals around death have implications both for governance and for the everyday lives of surviving family members. However, families of those missing in migration are not simply denied access to a body and the rituals that permits, but to knowledge of death.
For families of the missing, due to the lack of clarity over their fate, the death of their loved one is something almost impossible to admit to themselves. This can disrupt the normal grieving process and may lead to arrested grief or atypical reactions, known as 'complicated grief' (Blaauw and Lahteenmaki 2002). The ambiguous loss model has begun to be used as a basis for understanding the impacts of disappearance on families and in steering interventions. Ambiguous loss is ''a situation of unclear loss resulting from not knowing whether a loved one is dead or alive, absent or present'' (Boss 2004: 554), and occurs when a family member is psychologically present, but physically absent. It will be used here as one way of understanding the affective impacts of having a relative missing in migration.

Disappearance and its meanings
The way families of those missing in migration lose contact with relatives takes two fundamental forms that can have dramatic impacts for the meanings that family members give to the absence of their relative. For most, those now missing left home and at some point on their journey all contact ceased. For some however, the crossing of the Mediterranean was one undertaken together, as a family, and surviving family members were a part of the shipwreck in which their relative went missing. In such circumstances, there is a cumulative trauma of being rescued from a sinking vessel and then learning that family members are dead or missing: We travelled on the 13 th of September from Bodrum in Turkey. Our boat carried nearly 152 passengers, which was a very big number. . .. It was stormy and the huge waves hit and pushed our boat backwards till it hit a rock and we started to sink before reaching the island. My husband, I and other people next to us threw ourselves from the boat's windows into the sea, then climbed up the boat again when a monster wave hit it and pushed us far from it. My legs were stuck in water till my husband released them, then we decided to swim away from the 8 Simon Robins boat before it capsized over us. We kept swimming in the dark for 4 hours, but the coastguards did not come to rescue us. After that, my husband got tired and decided to swim on his back. He was behind me, but we could not see each other in the darkness despite the beacon's flashing lights. . . Suddenly, I heard him yelling for me, then he disappeared in the middle of the sea and I have not seen him since. (Wife of missing man, quoted in Ben Attia et al. 2016b) Those who have survived shipwrecks -and often seen the bodies of relatives who died in them -are thus exposed to two very different types of trauma: that of a single, traumatising event that can lead to post-traumatic stress disorder (PTSD), and the traumatic uncertainty of an ambiguous loss, as loved ones have simply disappeared. In contrast, the missing from Tunisia were mostly younger men who left on their own, or with friends, and in many cases without informing their families. All carried mobile phones that allowed those travelling to keep families updated with their progress, often calling right up until the point that they left the Tunisian network range and were not heard from again. (Ben Attia et al. 2016b) Whilst in most cases no official information has been given, many families have been told things that impact upon the meaning they give to their relative's disappearance, by smugglers, those travelling with the missing, and others who claim they have information. This, combined with opinions in the community and among other relatives of the missing, will ultimately determine how relatives interpret disappearance.
In some cases, the sinking of boats was reported in the media and families could draw conclusions depending on what they knew to link that shipwreck to their relative's journey. Informal sources are however often much more important to families, with relatives of the dead and missing, and survivors of shipwrecks often trying to inform the families of those they were travelling with. (Ben Attia et al. 2016b: 13) In most cases however, no trustworthy information at all is received, leaving families subject to rumour and speculation.
One coping mechanism for families with no knowledge of the fate of loved ones is to search for both evidence that they are alive, and a reason to explain why they have not been in touch, and rumours can feed this need. Family members who had travelled with those now missing and who had been a part of the shipwreck have seen exactly its consequences, including in most cases the deaths that resulted. As such, they are more likely to understand that there is a possibility that the missing have drowned, but still have a human reflex to want to believe they could have survived. We don't know anything at all about him since that day. My heart is broken [crying]. He is my eldest son and I swear my heart is broken. I wish to see him or hear his voice and then I will die happy. I don't want anything from life but to hear his voice. [. . .] The problem is I don't know where he is. I would feel better knowing his location. The problem is he is missing. I lost him. It's absolutely hard. I would feel better knowing where he is. My heart is broken and there are thousands who are missing. We don't know where they are. If they are alive or dead. The data from families was unequivocal about the presence of ambiguity:

Ambiguous loss
[F]amilies demonstrated this ambiguity during the interview by saying both that they await the return of the missing and that they expect he is dead, demonstrating how they are trapped between hope and despair, seeking an answer but fearing it will be the worst answer. Families made many statements showing how they were constantly moving between these two contradictory understandings (Ben Attia et al. 2016b: 13): We would like his body, I'd like my brother. Even a part of him, it would relieve us to bury him here in his country among people who know him and love him. But I'm convinced that he's alive, and I need an unalienable proof so that I admit his death. I don't want to say that he's dead, my brother is alive. (Brother of missing man, quoted in Ben Attia et al. 2016b) A significant number of family members stated unequivocally that their missing relative was alive, even with no evidence to support this assertion, articulating the incomprehensibility of their death. In many cases, mothers express a visceral, physical awareness of their belief that their son is alive, demonstrating the primacy of the affective and sensory over the rational in families' responses to their experience. The most intriguing stories however concern those who have been shown photographs or videos of their loved ones, apparently in Europe This can also be seen as families articulating their right to know the truth, to believe that their loved ones are alive until they receive evidence otherwise. It is however also interesting to compare the certainty of families that they have recognised their family members with the known inadequacy of visual identification of the dead (ICRC 2009a). This emphasises that the issue of missing migrants concerns not only those who have died, but all those who are unable to contact family at home, including the thousands of potentially vulnerable unaccompanied minors. (Ben Attia et al. 2016b) The impact of having a missing relative We live in mystery. My parents died, but I forgot them. However, my son's leaving is unforgettable, I can't get over it. . . (Mother of missing man, quoted in Ben Attia et al. 2016b) The difference between the impact of a death on an individual and family, and that of a disappearance, is profound. A death that can be mourned is always a tragedy but one that can be comprehended, both personally and socially. Funeral ritual after death can construct meaning, formalise the departure of the deceased from the community, and create a space for mourning that allows both a processing of the new situation and a shared understanding of a family's pain. A disappearance however must be given a meaning, precisely because it is something that cannot be understood in the light of everyday experience. The trauma of ambiguity creates a challenge for families because they cannot make sense of a situation defined by uncertainty. Disappearance is of a chronic nature that has emotional, psychological, economic, and social consequences (Blaauw and Lahteenmaki 2002): the net impact on families and individuals of having a missing relative will be the sum of these effects, subject to the resources of individuals and communities to cope.
Families of the missing are mostly seen to suffer from normal emotional distress after extremely upsetting events, rather than psychiatric disorders. It is also key to understand that families and communities have resources to address some of their needs, and during the empirical research of the Mediterranean Missing project an effort was made to identify these, notably positive coping mechanisms that offered emotional support. The Mediterranean Missing data revealed the impact of having a missing relative: The families demonstrated a range of symptoms associated both with the impact of trauma and of ambiguous loss, including sadness, sleep disturbance and dreams of the missing, anxiety and hypervigilance. One of the most commonly articulated impacts of having a missing relative was the very present absence of the missing in their life. Families reported that they were constantly thinking of the missing, involuntarily dwelling on the life and disappearance of their loved one. (Ben Attia et al. 2016b: 20) [My mother] wakes up every day asking about them. For example; you called inquiring about them now, she was alarmed that maybe you have any news about them. . . she is attached to the ringing of the phone, is depressed, has heart and pressure diseases which are aggravating even more. . . The affective border 11 The pain families demonstrate can be understood in the light of ambiguous loss: This generalised anxiety disorder was the most common symptom encountered, together with expression of extreme pain. This anxiety, and its prevalence, appears to allow the problems facing most families to be discriminated from PTSD. Whereas PTSD is linked to a specific event of trauma, the anxiety expressed by families was about the missing person, rather than the event of his disappearance: it concerned an ongoing situation that has no closure (Boss 2006) (Ben Attia et al. 2016b At its most extreme, this takes the form of neglecting all aspects of life to dwell on the missing person: ''I don't want to eat, I don't want to wear clothes and I don't want to go out anymore'' (Mother of missing man, quoted in Ben Attia et al. 2016b) and this was seen to be associated with making an extra effort to include the missing as a part of the family: Feasts have no sense anymore, neither Ramadan nor Eid. I place the pan on the stove to prepare something to eat, and then I start crying. Even while preparing the A common theme of family members was sleep disturbance, and problems of insomnia. Others report the feeling of stasis, the impression that life has stopped or is somehow suspended pending resolution of the uncertainty around the missing: Until today, I've felt that the world has stopped. Nothing moves. I do not feel anything moving in this life. Since my son is gone, everything has been frozen! The air, the sun, the wind, the rain, the night and day. . . everything has stopped until now and since the day of his departure as I still expect a call from my son. Some family members met reported psychiatric problems beyond the distress described above and that may require medical intervention. (Ben Attia et al. 2016b) Such impacts can be truly debilitating, as well as serving to increase problems with self-isolation and societal stigma of the families of the missing.
There are also social impacts, with some family members who are confronted with the antagonism of family and community choosing to retreat into isolation. This is linked to those, notably mothers, who appear to be living their lives for the missing -or in searching for them -rather than for the living. Women whose husbands are missing face some of the most difficult social environments, as the ambiguous status of their husbands -neither dead not alive -has implications for their own identities. The result is that they are neither wives nor widows but have a unique status that a largely conservative Muslim society often fails to understand. A women's identity is largely defined through her relationship to a man, either as daughter, wife or widow and women existing between these categories challenge social norms. This is partially driven by women themselves, who seek to deny at all costs that their husbands are dead and valorise the suffering of waiting for their return.
The need for information about the missing was the overwhelming priority of all families met, seen as the route to addressing the problems they faced. Most obviously they sought closure -to know if their relative was dead or alive: [We need] [T]o know where they are. If they are alive or dead. I am staying here just for this. Its fundamental I know. We are Muslims and we like to recite the Quran for our dead and if they are alive we want to know where they are. This may be particularly impacting in a Muslim society, where ''there is a preference for dying at home [. . .] and to die away from home is generally considered a calamity'' (Petersen 2012: 2). Beyond an answer it was also clear that families wanted an explanation as to what had happened, information not just about the fate but something that communicates and demonstrates the how and why of their fate, and this links to the need to give meaning to their situation: ''I want the truth about my son. How did he die? Was he killed? Has he drowned in the sea? I want to know!'' (Mother of missing man, quoted in Ben Attia et al. 2016b) If the missing person is alive then families obviously want to know where he or she is; if they are dead they want a body so that mourning can begin.
The Mediterranean Missing data includes attitudes to bodies of loved ones: In some cases, families' belief that their relative was alive prevented a discussion about the bodies of the missing. However, in many cases families acknowledged that a lack of information indicated that the missing could be dead, and were prepared to discuss their need to retrieve the body of their loved one. (Ben Attia et al. 2016b: 17) The body represented proof of death to families. Receiving the body would permit families to 'honour' the dead, and to bury them locally so that they can again be a part of family and community, and tradition can be respected: The presence of the body of a loved one close to the family is considered a comfort. Despite the stigmatization of exhumation in Islam, none of those The affective border 13 interviewed opposed exhumation and return of the body in the event their loved one was dead (Ben Attia et al. 2016b).

Ambiguous loss as a lens on impacts on families
[A]mbiguous loss is the most stressful kind of loss. It defies resolution and creates long-term confusion about who is in or out of a particular couple or family. With death, there is official certification of loss, and mourning rituals allow one to say goodbye. With ambiguous loss, none of these markers exists.
The persisting ambiguity blocks cognition, coping, and meaning-making and freezes the grief process. (Boss, 2006: xvii) The lens of ambiguous loss is one way to interpret impacts on families, as well as offering a route to therapeutic approaches, and here is presented as aiding understanding of the impact of a relative missing in migration.
In contrast to individualized trauma approaches, such as that of PTSD, the lens of ambiguous loss is explicitly relational, characterizing the stress as external and ongoing. An empirical evidence base shows that ambiguous loss can lead to symptoms of depression, anxiety, and family conflict. This has been used to develop a model in which the stress of ambiguity impacts mental health and well-being, subject to individual and collective resilience. While the data show the need for closure for families of the missing to end the ambiguity of loss, this is denied to the families of those missing in the Mediterranean. The ambiguous loss model confronts this perceived need for closure: The goal is to find meaning in the situation despite the absence of information and persisting ambiguity. Here, resiliency means being able to live with unanswered questions. Instead of the usual epistemological question about truth, we ask, ''How do people manage to live well despite not knowing?'' (Boss 2007: 106) The ambiguous loss model can be used to both justify the need of families of the missing to receive answers about the fate of their loved ones and as a framework for the delivery of support in the absence of such an answer. The qualitative data collected from families of missing migrants by the Mediterranean Missing project demonstrate the relevance of ambiguous loss to families of those missing in migration. In Boss' work with ambiguous loss, the need of families to give their experience meaning is prominent -most obviously in terms of the fate of the missing -and this is visible in the data addressed here. Identity is also crucial as disappearance provokes anxiety about the role of those left behind, as mothers without sons and wives without husbands seek to reconstruct an identity for themselves despite the uncertainty. (Ben Attia et al 2016b) There is always a danger that defining a population as traumatised invites external intervention to address their problems, and diminishes their own agency to do so. As Pupavac has said in characterising therapeutic governance: ''injustice is recast as psychological injury and exclusion, and rights become reconceived therapeutically as addressing psychological needs.'' (Pupavac 2004: 379) However, for the families of missing migrants, the challenge is not that they are pathologized by Western states but that they and their needs and rights are entirely invisible to them, despite those states being in a position to address their trauma through identifying the bodies of the dead. The priority now is to see their suffering acknowledged as a platform to ensuring their visibility and agency in addressing it.

Conclusion: The affective border
The performative border defines itself by its ability to separate citizen and alien, and today the EU's Mediterranean border does this in part by killing a significant fraction of those who attempt to cross it. Sovereignty, following Foucault and Butler, has no essence, and must continually be articulated and rearticulated in terms of stylised repetition of acts of sovereignty, and as such migrant bodies as well as the trauma they generate are one product of the assertion of the border that incarnates sovereignty. Such a performative understanding of the construction of the border sees the border being built by and upon the bodies of migrants.
The way the migrant dead at Europe's southern borders are governed has dramatic impacts on those who are waiting desperately to hear news of their migrating relative. As such, the treatment of migrant bodies in Europe has traumatic implications for the families of the dead and constructs certain relationships between those families and states of the EU. Understanding governance as the production of subjectivities, this then is the production of trauma in a population through the explicit action or inaction of a state authority: governance through trauma. The explicit form of this trauma is ambiguous loss caused by the border, which spatially demarcates politically sovereign lives -denying to migrants the closure that is considered a right for the relatives of citizen dead. The border ultimately both kills migrants and prevents families knowing the truth of the fate of missing loved ones, insinuating itself between the dead and their families, creating ambiguity and trauma.
The empirical work of this study has focused on how the bodies of the migrant dead are addressed by the authorities in Italy and Greece. In particular, it notes that national authorities largely avoid taking responsibility, and this sees obligations default to local authorities in Lesbos and Sicily. While in Italy a large-scale and state of the art forensic operation is underway, this addresses only a minority of migrant deaths and in both states efforts remain handicapped by an apparent inability to access the families of the dead. While post-mortem data collection and storage from migrant bodies found in southern Europe has been poor , and remains imperfect, it does offer the prospect of identification through a comparison with ante-mortem data from families. The principle obstacles are The affective border 15 the difficulties in accessing families far away separated by multiple barriers of culture, resources, technology and political will. As such, the geographical border becomes a highly extended object dividing not just Europe from its southern neighbours, but families from their loved ones and -through the social and familial impacts of ambiguous loss -families from their communities, and surviving family members from each other.
The Mediterranean border is only one of many contexts where migrants are dying while seeking security or a better life. In the Sinai, the Sahara, in Central America, at the US-Mexican border, and on the Andaman sea, as well as in in other areas, migrant deaths are occurring on a large scale, with often extreme challenges in families gaining information about them, or indeed any data at all emerging. There remains a need to focus global attention on both the issue of migrant deaths and on identification of those who die. Specifically, there is an urgent need for a transnational architecture to manage data around missing migrants, both for European states holding post-mortem data centralize such data, and for them to have access to ante-mortem data from a large range of other sources, including states of migrant origin, other European states, and directly from families. Data to facilitate identification must be collected and shared with an awareness of data protection standards, of the need to keep them separate from those accessed by those policing the border, and of potential security threats to those fleeing regimes that threaten them.
Deaths at the border have transnational political, psychological and social effects on families in migrant countries of origin. Hence, ''the border is defined not only in spatial, geographical, or political terms'' (Robins and Kovras, 2016: 41), but ''also has a strong emotional component: its presence has an affective impact far away'' (Ibid.). The struggle of families to cope with the lack of clarity of the fate of their loved ones becomes a permanent and dominant feature of their daily lives. The tendency to police bodies, rather than merely spatial borders that has become a principal doctrine in recent decades, impacts populations of which European authorities remain unaware. This demands that we shift our attention to both the corporeal and affective dimensions of deaths at the border, building on a growing trend in the literature (Coleman and Stuesse 2014;Andrijasevic 2010;Pugliese 2009;, and seeking to make the families of missing migrants visible. The dead who have been laid to rest, for whom rituals have been made and who have made the social and political transition to non-persons, make few affective or other demands of the living. Particularly in Western societies where ghosts, spirits and ancestors no longer have power over the living, the buried and mourned dead have no agency and laws dictate the modest obligations of the living to the dead (Stepputat 2014). However, where the dead cannot be mourned, or -even more so -where their absence is an uncertain and ambiguous one, the dead continue to haunt the living. In this case, and particularly in the affective domain, the dead have something close to agency, a subjectivity that emerges from deep emotional links to family members that only closure can sever. The very existence of the body and its presence at the border as a product of politics shows the corpse itself as a political subject. ''The dead migrant body appears to have agency; such bodies can both nourish and haunt the living, animating the social and political processes around death and challenging the body as purely an object of politicisation.'' (Kovras and Robins 2016: 43) The affective impacts of the dead body give them power over the living (Borneman 2014).
Methodologically, this work suggests a richer ethnography of migration and its impacts can be found by focussing not on the categories that law and policy produce, but rather to engage in an embodied scholarship that emerges from the persons of migrants being studied. Geographies of emotion and affect can challenge the distancing of the emotional from scholarship, in favour of 'embodied thinking' that emphasises an engagement with everyday emotional subjectivities. Scholars of affect believe that: ''placing emotion in the context of our always intersubjective relations offers more promise for politically relevant, emphatically human, geographies.'' (Thien 2005) This seems particularly relevant in the field of migration where the movement and interaction of people will always be governed by spatialities of emotions.

Note on funding
The project that collected the data upon which this paper is based was funded by the Economic and Social Research Council (ESRC) grant no. ES/ N01345X/1.