Devices of suspicion. An analysis of Frontex screening materials at the registration and identification center in Moria

ABSTRACT

The identification of migrants and the creation of data identities lies at the core of datafied forms of migration and border control. In recent years, Frontex has made identification to one of its key tasks and conducted so-called screenings in many EU member states. Yet only little is known about the screening materials in use. Based on an ethnographic inquiry of Frontex’ data practices, this paper analyses Frontex booklets, dossiers, questionnaires, images, and forms and studies how they structure the situation of identification. Making use of research in science and technology studies and recent research on suspicion and credibility assessment, it argues that those materials not only compile information but work as socio-material devices of suspicion that render migrants into fraudsters and translate peoples’ actions, stories, and performances into accounts of truthfulness or deceit. As devices, they frame cases, script interactions, code statements, create stigmata of belonging, and produce purified accounts, and thus enact multiple forms of suspicion. The paper concludes with a critical reflection about the credibility of those materials and speculates about how devices could be designed otherwise.

KEYWORDS:

• Credibility assessment
• device
• Frontex
• migration and border control
• suspicion

Introduction

Until 2020, the registration and identification center Moria was in a former military camp in the south-east of Lesvos Island.¹ There were several residential areas, a central road with stores, a central place, and some closed facilities. In the center of the camp, the Registration and Identification Center (RIC) was installed, fenced and guarded by Hellenic Police. One container of this arrangement was the screening center, where Frontex officers identified the arrivals and created data identities. In this process, officers seek to determine the name, age, gender, country of origin, and other details. All too often passports, documents, and entries in birth registers are not available, were not issued in home countries or got lost during the journey, or data identities have not (yet) been uploaded or failed to be distributed across information systems. In these cases, credibility assessments are conducted, that is interviews involving border guards, translators, and migrants. For this process, border guards relied on a variety of materials, such as dossiers about so-called countries of origin, questionnaires, compilation of images, and forms.

Created, stored, and distributed data identities are the basic elements for surveilling and controlling people on the move. Frontex and Hellenic police officers upload them to several national and EU databases, which multiplies the bordering of people (Pollozek and Passoth 2019). By this, depending on the data entries, migrants might face unpleasant forms of interrogations, waiting times, the refusal of entry, or even incarceration at other sites and by other authorities across the EU. Those crafted identities are key for issuing temporary documents – ‘Restriction of liberty cards’ – and for making people containable in camps and on islands (Tazzioli 2018). Since 2016 and in the context of the EU-Turkey deal, Greece has implemented fast-track border procedures on the Aegean islands that put people into separate asylum tracks. While in an admissibility procedure, Syrian nationals needed to prove why Turkey is not a ‘safe third country’ to them for applying for asylum, in eligibility-procedures people with low recognition rates were deprived of applying for asylum (RSA and Pro Asyl 2021). In many cases, non-Syrian nationals were rendered ineligible for applying for asylum and Syrian nationals’ claims inadmissible, with the consequence of detention and repatriation back to Turkey or elsewhere, or of being caught in a state of limbo on the Greek islands (Greek Council for Refugees 2021, 86).

The identities created throughout the procedure of identification and registration affect peoples’ stay in Greece in many ways. Details of age, vulnerability, or country of origin determine access to accommodation and social services and prepare the sorting of people into different administrational tracks with different state authorities being responsible. At least during the time of our research, identities indicating that people are not willing to apply for asylum were forwarded to the Hellenic Police, identities indicating the willingness of applying for Asylum were distributed to the Hellenic asylum service, and those classified as vulnerable were taken care of by doctors of the world or E.K.K.A. (Pollozek and Passoth 2019). By this, people could face detainment and repatriation, even before speaking with the respective asylum administrations (MIAR 2017). While a large range of research has been published on the biometric control regimes of re-identification (e.g. Muller 2010; Scheel 2019), surprisingly little work has focused on the creation of such identities in the first place. This is even more surprising when considering the manifold utilizations of such identities on a regional, national and EU level, and the consequences they have for people on the move.

This paper contributes to this special issue by asking how migrants are enacted as knowable and governable subjects and how control knowledge in border encounters is produced through low-tech devices (Amelung, Scheel, and van Reekum 2024). By deploying an STS-inspired analytical vocabulary, it argues first that devices have ‘agential effects’ (Amelung, Scheel, and van Reekum 2024) on the creation, stabilization, and authorization of control knowledge. They not only compile information but make information operational for Frontex officers and the practice of screening. Our analysis will reconstruct the work and contribution of such devices to the production of control knowledge in five different ways: they frame the cases, script the interactions, code the statements, create stigmata of belonging, and produce coherence through purification. By this, the paper not only contributes to recent research by providing an in-depth analysis of the knowledge and materials Frontex has fabricated for identification and screening but also develops an analytical lens that is attentive to the performativity of such knowledge devices. Second, the paper argues that such knowledge devices render migrants into fraudsters, and translate peoples’ actions, stories, and performances into accounts of truthfulness or deceit. Thus, they are not merely innocent instruments for coordination and cooperation of diverse actors at the border, but themselves effective elements of a culture of suspicion enacted in local screening and reporting practices as well as through and with the devices used in such practices. In the following, we will show how devices of suspicion enable and encourage questionable interrogation techniques that explicitly seek to provoke contradictions and that are entwined with violence and coercion.

The paper is based on ethnographic inquiry into Frontex’ data practices conducted between 2016 and 2018. One of the authors visited multi sites, such as Lesvos, Athens, Piraeus, and Warsaw, and collected empirical data on the so-called Moria hotspot, as well as on various regional, national, and international coordination centers and headquarters. Throughout this process many hands-on documents have been collected and analyzed ranging from operational plans, standard operations procedures, dossiers, guidelines, manuals, administrational forms, classification sets and other templates. More than forty interviews were conducted with personnel from Frontex, police and coast guard authorities, non-state institutions and activist groups working at, for and against the identification and registration of migrants at Moria.

In the following, this paper will develop a framework for analyzing devices of suspicion that considers work from science and technology studies and recent research on suspicion in credibility assessment. Then, the paper will turn to the empirical data and analyze how the Frontex materials work as devices of suspicion for the practice of screening. The paper will close with some critical reflections about the credibility of the knowledge Frontex has compiled and speculate about how devices could be designed in another way that could strengthen migrants’ position and bind migration and border control actors to legal standards and procedures of accountability. Even though these low-tech devices play a decisive role in the larger political context of migration governance and the role of Frontex as a powerful coordinating entity of such a regime of migration governance, this paper will explicitly not study the role of devices as part of a broader political analysis of migration governance. This was done elsewhere (Pollozek and Passoth 2023a; 2023b), where we examined how low-tech devices coordinate heterogeneous and fragmented forms of European migration governance in quite provisional ways – which produce quite severe issues of accountability. Moreover, an analysis of the broader political circumstances of such devices in migration management would, at least from the material and situated approach we utilize in this paper, have to follow the construction, circulation, deployment, use, maintenance and contestation of devices of suspicion, that is a multi-sited analysis that would go far beyond the scope of this paper.

Devices of suspicion

In Frontex joint operations, border guards from EU member states come and work together. Even though they are experienced, they come with different skills, backgrounds, routines, and languages. During a joint operation, border guards may also be interchanged (Pollozek and Passoth 2019). Frontex problem as a coordinating agency is turning disparate practices into institutionalized ‘mechanisms of repetition’ (Singleton and Law 2013, 260) that are kept similar throughout time and despite changing actors (Kasparek 2021). When one of the authors of this paper conducted ethnographic fieldwork, we learned that Frontex officers receive an USB-stick when they are deployed to a Frontex operation. The stick contains the Frontex Operational Plan, materials on fundamental rights, forms, and reporting templates, as well as a folder with material for screening. The latter includes the Frontex Screening booklet (2015), a compilation of files with background information, country dossiers, questionnaires, lists, maps, and images. The screening booklet is organized by a folder structure with three main folders: one for identification forms in different languages, one that accommodates country specific materials, and another for background information.

Interviewees repeatedly referred to those materials when describing how to solve problems of identification as well as of coordinating with colleagues from different national bureaucracies. These documents played a decisive role in our interviewees accounts of how to organize and perform the screening process and they were, at least in these accounts, used for the practice of screening. They were instrumental in producing questions, raising concerns, and justifying mistrust. They were, as we will call them throughout this paper: devices of suspicion.

Suspicion in migration and border control

Extensive research has shown how migration and border control regimes work as filters and seek to produce the increase and restriction of mobility at the same time, letting pass those who are allowed to, and stopping and refusing all the others (e.g. Mau and Barfoot 2022; Transit Migration 2007). Such a sorting function enacting and differentiating between ‘political refugees’, ‘wanted’ or ‘unwanted’, ‘economic’, ‘illegal’ and ‘irregular migrants’ is not only established by migration law and policy, but also by the migration and border apparatus with all its paperwork, visa requirements, credibility assessments, as well as border technologies ranging from databases, biometrics, risk calculations, to gray – and blacklists. This does not only lead to unequal conditions of mobility but turns suspicion – assuming in advance ‘that someone is guilty of an illegal, dishonest, or unpleasant action’ (Oxford Dictionary, cited in Borrelli 2020, 4) – into a central element of migration and border control.

As the governance of such filters is based upon spotting, containing, detaining, illegalizing, and repatriating those who are considered ‘unwanted’, a culture of suspicion is embedded within everyday interpretations and assessments, institutionalized in bureaucratic practice, and migration law and policy (Borrelli, Lindberg and Wyss 2021; Jubany 2016; Pratt 2010). Such a culture of suspicion is also affected by political and public discourse that normalizes and reaffirms suspicion (Affolter 2022; Alpes and Spire 2014; Dahlvik 2018; Noll 2006). By this, suspicion becomes the default mode of migration and border control linking migrants with tropes of the ‘criminal’, ‘terrorist’, ‘trickster’, and the ‘undeserving’ (Borrelli 2020). Hence, people are not presumed trustworthy but need to proof that they have not committed a crime, that their reasons for applying for a visa are not dubious, or that their reasons for applying for asylum are legit.

This implies that suspicion is more than an attitude of (some) individual case workers that ill-interpret ‘rule-bound, neutral and non-discriminatory migration laws’ (Borrelli, Lindberg and Wyss 2021: 3). Suspicion is also not a result of instances of fraud and criminal activities, but a performative effect of the European migration and border regime and its dynamics of securitization (Borrelli 2020; Scheel 2022). Through ever new products, methods, practices and technologies of security and risk assessment, and ever new articulations of threats, risks, and dangers seemingly coming from migration, further demands for and practices of sensing, detecting, listing, and assessing migrants are (re)produced.

Practices of suspicion in credibility assessment

Recent research has studied this active (re)production of suspicion in the context of identification and credibility assessment by focusing on various socio-material practices within institutional cultures (Borrelli, Lindberg and Wyss 2021). Some of this work is also part of this special issue, see for example van der Kist’s analysis on tinkering practices with country-of-origin information or Scheel’s research on mobile phone data analysis in asylum procedures. Work on the practice of credibility assessment has focused on discretion, that is how laws and formal forms of institutional knowledge are applied when doing credibility assessment (Evans and Hupe 2020; Gill and Good 2019). Street-level bureaucrats transform subjective interpretations into ‘objective’ legal criteria by selectively referring to rules or laws (Zampagni 2016). Scheel has highlighted the discretionary power of such practices, as they render the fabrication of institutional facts opaque and produce uncertainty on the side of migrants (Scheel 2022). Furthermore, work on the practice of credibility assessment points to the common-sense assumptions within an institutional culture passed along peers and shaping everyday forms of suspicion. In teaching novices how-to knowledge – how to tell when a story is ‘off’, which alert indicators to keep in mind, what is counted as ‘‘logical’ and ‘illogical’ behavior’ (Affolter 2022, 1078), or when and how to dig deeper – also professional norms are invoked that are at the same time linked with suspicion (Affolter 2021).

Practices of credibility assessment also interrelate words with bodies, and thus interweave language, the corporeal and the visual (Kelly 2012). Fassin and others have argued that the body not only becomes an inscription of power through persecution and suffering, but also an inscription of truth and evidence (Fassin and D’Halluin 2005). In this context, research has started to investigate technologies of suspicion in credibility assessment and their targeting of the body. DNA tests in family reunification cases reducing complex family constellations to genetics, medical age assessments for unaccompanied minors through X-ray based on western models of childhood, or automated language tests to validate claims of origin of asylum applicants show vividly how the practice of credibility assessment has been extended to medical professions, health examinations, and laboratories (Bohmer and Shuman 2018; Netz 2022; Netz, Lempp, and Krause 2019).

Analyzing devices of suspicion

This focus on materiality and technicality of practice provides a fruitful background for our analysis of materials and how they structure the practices of Frontex screenings. By analyzing the Frontex materials as ‘devices of suspicion’, we make use of an approach developed in different empirical fields of Science & Technology Studies (STS). The notion of the device was introduced in Bruno Latour and Steve Woolgar’s ‘Laboratory life’ (1986) to make sense of the role of the equipment and material setup of research laboratories and office spaces. Inscription devices – ‘any item of apparatus or particular configuration of such items which can transform a material substance into a figure or diagram which is directly usable by one of the members of the office space’ (Latour and Woolgar 1986, 51) – are used not simply to automatize tasks or to carry out activities otherwise performed by staff. Such devices, as Rheinberger (1997) later argued, are ‘epistemic things’ producing inscriptions – serialized measurements, numbers, diagrams – that in turn become as important elements of scientific practices and outputs as journal articles, conference proceedings, and gray reports.

From this starting point, the notion of the device traveled to multiple approaches and fields of application, later called (post-)actor-network theories and practice and object focused methodologies. Work has focused on algorithmic and automatized calculative devices, but also on low-tech devices such as forms and calculation tables in financial markets (Callon and Muniesa 2005), publics and participatory setups (Marres and Lezaun 2011), governance and public institutions (Ruppert 2012), and social research (Law and Ruppert 2013). Device-oriented approaches also have traveled to research on security practices, where it has been used to ‘(1) sidestep developmental and epochal renderings of security that can be encouraged by discussions of technology, (2) take the debate beyond the confrontation between understandings of technology as pure instrumentality for social ends or as an autonomous driving factor for social practice, and (3) recast recent discussions on ‘materiality’ in international and security studies by extending the attention to ‘things’, to the arrays of equipment and instrumentation through which security is performed’ (Amicelle, Aradau, and Jeandesboz 2015, 295). In recent years, research has studied how maps, security scanners, credit scoring, or walls work as devices, and worked out how the design process of a security device is shaped by assumptions, expectations, and visions of scientists and security actors alike (see the special issue on security devices by Amicelle, Aradau, and Jeandesboz 2015; Pallister-Wilkins 2016).

By focusing on materials as devices means to start with hybrid collectives (Callon and Law 1995) instead of human actors or autonomous technologies alone. Frontex screeners are not ‘unaided human individuals’ (MacKenzie 2006, 268) but equipped with materials that require as much attention as what they practically say and do. In contrast to elaborated technological objects – which carry and blackbox inscribed programs of action, visions of users, or use-case scenarios –, the Frontex devices explicate them visually, verbally, and procedurally. They describe and prescribe (Akrich and Latour 1992) how to conduct screening, which actors to involve, which roles to play, what typical procedure to follow, and which different methods of assessments to apply. Paraphrasing Law and Ruppert (2013) on methods as devices, the Frontex materials ‘teach us about the social’ as they make ‘assumptions about the character of the social as a collectivity, the components of that collectivity, the attributes of those components, and how best these might be researched’ (Law and Ruppert 2013, 234).

Moreover, Law and Ruppert (2013) point out that devices are materially heterogeneous and carry teleologies. They are neither high-tech nor low-tech, nor are they tools that instrumentally produce a single, causal, or even predictable effect. Their composition might not be internally coherent and still have effects on practice. Some devices support, some force a certain practice. Devices can also be an obligation or obstacle to overcome and work-around. They play an active role, they sometimes act or make others act, not always according to an inscribed plan or pattern but by materially and socially altering the situation in which a practice is enacted. In any case, to treat a thing as a device means to analyze it as something that is related to and tailored towards a certain practice – designed for and used as something that seeks to alter and shape this very practice.

Treating devices as materially heterogenous demands to consider the particular history, the particular arrangements, their inscribed teleologies and envisioned scenarios, as well as forms of use. This is especially challenging in the realm of migration and border control, where secrecy cuts off access to the design and construction of devices, such as the Frontex screening material. But it is also challenging because devices are ‘either forgotten or taken for granted as being merely technical matters’ (Latour and Woolgar 1986, 63). This requires workarounds and inventive forms of research. In our empirical inquiry, we could not trace the design process of the Frontex documents and materials, even though many of the materials were not framed explicitly in terms of secrecy. Furthermore, we were not allowed to attend actual screening situations, so we could not observe how exactly the materials have been used in practice. But we did speak to border guards and made the materials a topic of our interviews, using the accounts of screening situations given by our interviewees and the individual reflections about the use of respective materials as material for collective analysis. We collected, asked for and inquired as many details as possible regarding the practice of screening and the materials in use. Some of those materials were shown, read out aloud, or described to us, and we were particularly sensitive to comments, reflections, and assessments our interlocutors articulated.

Devices of suspicion in credibility assessment of Frontex screenings

In the remainder of the paper, we will analyze the Frontex screening materials as devices of suspicion by unpacking their teleologies and methods. The latter teach screeners about the social of the screening situation, they are also performative in framing and organizing the circumstances of screening practices. By this, they shape the respective processes and enable and persuade involved actors to act in a certain way. As we will see, devices shape the situation of screening in several ways: (1) they script the process of identification and distribute roles, (2) frame credibility assessment in terms of countries of origin, (3) code utterances in terms of correct and incorrect, (4) assess a person’s body in terms of belonging, and (5) purify the very process by reporting templates. With this, they enact multiple forms of suspicion: suspicion is generalized and associates migrants with the figure of the fraudster. By containing people within social groups and within nation states, those outside of the expected framework are denormalized. Coding methods and chains of coding provoke inconsistencies, and suspicion can only be mitigated gradually throughout the course of questioning and answering. Migrants’ claims are put into context with racist stigmata of belonging constructed through images, while in the end purification devices obfuscate the process of identification and its different forms of suspicion. In the following five sections we will focus on each of these practices and the role devices of suspicion play in framing, shaping, and enacting them.

1.

Scripting

Given the heterogeneity and different background of border guards and the fluctuation of ever only temporally deployed personnel at Frontex joint border operations, the Frontex screening materials are a device for keeping screening practices similar throughout time and despite changing actors. They not only provide general information for screening but also specify and explicate how the situation of screening should look like. They do this by providing scripts defining how the practice of screening and credibility assessment should look like, imagine users with specific competences, motives, aspirations, and problems, and speculate on further actors, entities, and factors that could become relevant (Latour 1990).

The script reads like this: The Frontex materials explicitly articulate a standardized procedure and specify the screeners’ role in it. The screener should get prepared, that is, check if colleagues are around who can provide safety and perform a body and luggage search, check if the fitting screening materials for claimed nationalities are compiled, and check if interpreters are approachable. After calling a migrant into the container, screeners are asked to scan the appearance from the top to the ground and assess size, age, color of skin, scars, and clothes. Then screeners should offer the migrant a seat, hold eye contact, try to find out which language the migrant speaks, and then call for an interpreter. When the screened person has claimed to be from a particular country, screeners should consult the screening booklet and open the corresponding country folder. Then several questions about this very country should be posed. Step by step the identification form should be filled out and at the end screeners and migrants should sign the identification form (Frontex 2015, Frontex 2017).

Moreover, the script articulates different roles and actors. While the screeners are positioned as experienced and skilled ‘experts’ (Frontex 2015), migrants are associated with the roles of fraudsters or tricksters. Throughout the materials and interviews, migrants are positioned as suspicious, sometimes even as dangerous (I20, I23). Also, the Frontex screening booklet assumes migrants as suspicious differentiating between ‘claimed’ and ‘presumed’ nationality, estimating chances and success rates about ‘finding out if the claimed nationality is right or wrong’, introducing techniques of taking someone by surprise, and giving advice for precaution (Frontex 2015). In this regard, Frontex screening is not an exception. Well-prepared fraudsters or tricksters are common tropes within credibility assessment (Scheel 2019). It is assumed that tricksters know how to behave, to lay low and to remain inconspicuous. They know how to hide their true self and to act convincingly as someone else.

In our interviews, tricksters were positioned as the antagonists of the screeners and become the primary target of border guards (I15, I20). By this, the role of tricksters was assumed by default and suspicion was generalized. This produces an ambivalent relation between screener and migrant. The Screening booklet states that on the one side, the screener is supposed to gain trust by introducing herself properly, describing the screening procedure, informing migrants about their rights, speaking with a calm but clear voice, and encouraging migrants to start talking (Frontex 2015). This is also deemed important in case migrants need medical treatment, turn out to be vulnerable, or are willing to act as witnesses (Frontex 2017). On the other side, screeners are asked to keep distance and are warned of not getting too close or affected to a migrant – ‘never become a friend’ (Frontex 2015) – as the scripted process of screening is all about questioning and disputing migrants’ claims.

To unpack the assumed hidden intentions of migrants, the screening booklet refers to what we call the authenticity of the body that can only be tested when implementing the script in concrete screening situations. The screening booklet states that a screener should ‘focus on the body language and on the impression management’ (Frontex 2015). This assumes that the body might lack full control and thus reveal if someone lies. Behavior, such as getting nervous, avoiding eye contact, starting to sweat, starting to play with hands, and getting noisy or even aggressive are taken as indicators for hiding something (I15, I20). In our interviews, we heard various confession stories that are not explicit parts of the script, but account of its performance in practice. Screeners would try to mirror that they are not buying the claims of migrants and then start confronting them with alternative assumptions (I15). They try to keep the pressure high while at the same time continuing the communication until the screened persons would give up their claims and confess who they ‘really were’ (I15). In some of the stories, this constant push towards confession was parallelized with an increasing failing and finally a collapse of impression management, for instance when the screened person would start crying (I20, I23).

This creates an ambivalent positioning for migrants. The script expects them to act as competent, accountable, and intelligible while at the same time doubting their intentions and motivations in a substantial way. Throughout its steps and in instances of its performance, the script serves as a device producing and reproducing suspicion. Impression management for migrants means to keep face even if screeners explicitly and loudly question their statements and to remain polite, and to keep going even if their accountability, intelligibility, and selves are disputed and denied with tremendous consequences. Furthermore, the screening often takes place right after people have crossed the Aegean Sea and people might be traumatized, exhausted, and overwhelmed by the new situation in the camp, worried about their loved ones, and being concerned about what might happen next in Moria. However, when they step out of the role defined and imposed by the screeners by not managing to remain calm, patient, and polite, all those circumstances are crossed out and replaced by the preformatted role of the competent and sovereign trickster. With this, the envisioned script of the Frontex screening is tightly related to questionable interrogation techniques that aim to provoke contradictions, ignore the conditions of migrants and the screening situation, and force confession through means of violence and coercion.

2.

Framing

In contrast to asylum case workers, Frontex screeners are no experts on countries of origin. Often, the identification of people is only one activity out of many border guards do. During our research, we learned how screeners use different devices for framing and assessing statements. One of them were dossiers. The screening materials contain at least twenty country specific subfolders that contain collections of materials about a particular country. Most of the dossiers are about countries from the Middle East, Northern, West, and East Africa, and South Asia. In the following, we will show how those materials work as framing devices, creating ‘boundaries within which interactions […] take place more or less independently of their surrounding context’ (Callon 1998, 249), shaping the situation and its interactions by defining the problems at stake, ascribing causes, and attributes, and qualifying relevant issues. The outputs of such framing devices are often precarious because putting ‘the outside world in brackets’ (Callon 1998, 249) pushes actors quite violently ‘onto a clearly demarcated ‘stage’, which has been specially prepared and fitted out’ (Callon 1998, 253). They are loaded with power and imperfect producing processes of framings and overflows.

The dossiers of the Frontex screening booklet serve as framing devices to turn an uncertain and contingent situation into an ordered situation of credibility assessment. They construct an ordered world with ever similar and typical features of territories, borders, nation states, and national cultures. Almost all dossiers define countries based on a geography of states, detailing out geographical borders and neighboring countries. The capital city and its geographical location as well as flag, anthem, size of the population, official language(s), national holidays, and currency are listed next to geographical landmarks, such as mountains, rivers, lakes, forests, or deserts. The dossiers also construct an administrational geography of a country. Several maps and lists detail out major cities, major roads, and postal and phone codes. They also mention provinces or districts by name, and the size of their population. We also find entries of landmarks, administrative or public buildings, parks, squares, markets, and other historical structures. Additionally, brief overviews sketch out the political history of a state and names of political leaders and presidents are collected in form of lists (Frontex 2015).

Moreover, the dossiers link individuals with typified cultural and social groups. In this context, food plays a crucial role. There are lists, descriptions, translations, and pictures of and questions about ‘typical’ Afghan, Egypt, Iranian, or Eritrean food. While some paragraphs provide ingredients of dishes and how they are prepared, others elaborate on food in everyday life, such as baking bread in rural areas, or street food in cities. Further entries are portraits of famous singers and other public figures, as well as sports including popular teams and stadiums. In a few dossiers, a typical culture of a country is constructed out of references to literature, graphic and performance arts, styles of writing, and music creation. In most cases, overviews of languages and dialects are provided – however with huge discrepancies regarding details and length. Some dossiers only state the official state language (e.g. Arabic), while others contain maps displaying complex and overlapping geographies of language families as well as lists and explanations of regional dialects (Frontex 2015).

These geographical and cultural features are constructed as fixed and static (Johannesson 2012) building the basis for investigation and intervention (Law and Ruppert 2013). After screeners have asked migrants about their country of origin, they open the corresponding dossier and develop questions along the lines of information outlined above. The dossiers as framing devices translate the interactions with all their contingencies into an asymmetrical question-answer game and turns the screeners into ‘receivers and testers of information that others present’ (Smith-Khan 2017a, 519). Responses can be assessed not only in terms of correct or incorrect but also as far-fetched, implausible, or simply irrelevant. By this, migrants are put into the ambivalent position of having to guess which answers the screener might have in mind, and are confronted with the fact that their answers, believes, stories and choices are directly linked to a typified cultural background (Smith-Khan 2017b). As migrants are asked to claim to be a citizen of a particular country, they are expected to have access to knowledge about institutions, insignia, figures, histories, monuments, events, infrastructure, and high and popular culture. Screeners get suspicious when people do not meet these expectations of what should be known (I9, I15). However, the framing devices are biased towards a Eurocentric notion of a middle-class citizen (Amicelle, Aradau, and Jeandesboz 2015) which ‘presents obvious challenges for people who come from socially stigmatized groups that fall outside such institutional or stereotypical framings’ (Smith-Khan 2017a, 520). To our knowledge, no minorities or translocal or transnational communities are recognized, and people on the move, or people who are not, cannot or do not want to be part of a nation state are largely ignored. Multi-religious, multi-ethnic and multi-cultural constellations are as absent as details on regional dynamics and current conflicts. Structural violence, and different forms of discrimination regarding particular social groups and regions are widely ignored.

It is striking that the framing devices for credibility assessment itself lack credibility, and thus does the related mode of suspicion. An inquiry into the sources and references of the dossiers suggest a non-coherent patchwork of heterogenous materials, stitched together without any standards of source validation or methodology. The subfolders of the screening folder differ at great lengths. For instance, according to informants, the dossier about Syria contains even sub-dossiers about regions and districts while other dossiers are empty. Furthermore, the dossiers seem to differ very much in the ways of how materials are produced. While some documents are structured with headers, sections, keywords, and links, others show running text without layout. Drawing on insights of our informants, it seems that heterogenous sources have been used to produce the screening booklet. Some of the materials being described or shown to us looked like online encyclopedias or educational learning websites. We also encountered global lists of postal codes, telephone codes, and other communication standards which reminded us of compilations on Wikipedia. Other materials looked like travel guides and overviews from sources such as wikivoyage suggesting UNESCO World Heritage Sites, national parks, and other touristic sites (Frontex 2015). Materials about languages and dialects were like a blogpost we found on a website hosted by a company that provides translation services. This suggests that the screening booklet is not a completed publication but a patchwork of different materials and a work in progress with a folder structure suitable for adding further documents. Taking the pragmatics and politics of dossier creation into account, the dossiers produce a geography with fringes and blind spots and a variety in terms of detailedness that mirrors dynamics of border control.

3.

Coding

In addition to the broad background information of a country, Frontex has developed devices that are even more hands-on. So-called screening forms are also part of the country folders, and they can be understood as a condensed version of the dossiers. They are much shorter, more formalized, and they serve as coding devices that turn the complexity and contingency of an interaction into a binary pattern with a positive and a negative value. They structure the ongoing interaction into a sequence of questions posed by the screeners and answers provided be the screened persons. By this, the complexity of assessment is reduced drastically. While in open questions all kind and parts of responses from migrants could have informative value und become relevant for assessment, close questions imagine only a tiny realm of answers considered correct. All other answers do not matter regarding the content but are simply regarded as incorrect. The implication of such coding devices is that they request facts as answers and exclude other modes of speaking. They also do not account for and disqualify other forms of answers. This may produce severe uncertainty and disturbances for people who are not experienced with such forms of interrogation and who are used to respond in other formats, such as stories, or who relate things to their own lives and subjective experiences. Moreover, the authoritative form of the coding devices makes it difficult to articulate and claim alternative responses.

In a screening process, several coding devices are at work that help turning a ‘large proportion of questions [into] closed and fact-checking questions’ (van Veldhuizen, Maas, and Horselenberg 2018). One part of the screening forms are short profiles of a country. Structured as brief fact sheets, they are like the key fact boxes of Wikipedia articles. Those short profiles entail images of the country’s flag, maps of territories and districts, lists of items such as regional phone codes, and several details about a country collected in the dossiers. By providing a structured account in form of a list, they allow that questions can easily be formulated by the screener.

Moreover, the screening forms contain prefabricated questionnaires including correct answers. The screeners may also make use of images and photographs of famous buildings, mosques or monuments, singers, and other popular people. Those images are flanked with questions and answers, such as ‘how is this building called’, ‘in which city do you find this mosque’, or ‘who is this person’ (Frontex Screening Booklet). Furthermore, Frontex collected images of national flags, car plates, telephone numbers, banknotes, politicians, pop singers, and many other things and people. Each compilation brings together items from different countries. Such coding devices could be coined as select-the-right-items (Figure 1). For instance, after a person has stated that she is from Syria, she would need to pick the item from this very country. Another coding device in use could be called what-is-nearby. After migrants have indicated her hometown or the region, the screener uses Google maps to ask questions concerning local daily life: ‘What is the nearby bigger city, where is the nearby market, or what is the near-by bigger square?’ (I15) This test is supported by the dossiers offering additional collections of places, buildings, and parks of major cities.

Figure 1. Reconstruction of an image questionnaire.

Through coding devices, chains of intertextual inconsistency can be constructed. The screening booklet points out that screeners should ask migrants not only a few but many questions. Screeners also may repeat similar questions several times throughout the interview. For this, the Frontex screening booklet gives advice for often changing the subject and going back and forth to confuse migrants (Frontex 2015). From screeners point of view those techniques are designed to unsettle people and provoke stammering, nervousness, and lack of focus to reduce the chances of potential fraudsters to get their series of prepared answers straight (I15, I20). Based on general suspicion towards migrants that assumes fraudsters by default and on the cumulative process of coding, inconsistencies are provoked and made visible along the way. Hence, becoming unsuspicious is an unlikely accomplishment of answering all the questions in a way considered correct and of getting through the whole interview without contradictions.

4.	Stigmatizing

Credibility assessment not only includes the assessment of claims of migrants but also the articulation of alternative claims by screeners. When clashing claims about the identity of migrants are articulated, the screeners try to harden their assumptions and to refute the claims of the other party. In this process, screeners relate statements of migrants not only to the framing and coding devices analyzed before but also to migrants’ bodies. It is assumed that a country-of-origin leaves traces, signs, and marks on and in the body that can be crosschecked with the statements people give.

For this process, the screening booklet offers materials, which work as devices of stigmatization binding an individual to certain categorizations. They turn actions or the display of the body into signs and attributes that speak for and are considered characteristic of an individual. Stigmata define the status or position of an individual and link her to a social group. At the border, this can be done by sophisticated biometric technologies, but also by rather low-tech devices. By fingerprinting someone, a mark left literally on the skin of a person is ‘read by the appropriated equipment wherever these bodies go’ (Van der Ploeg 1999, 301). In the case of Frontex screening, border guards look for bodily signs that can be connected to a particular region and cultural group (I15). The Frontex screening booklet offers material and suggests methods for doing so.

Inconsistencies between someone’s claim about her country of origin and her dialect is almost considered as a knock-out criteria (I9). Another method of connecting statements and body is to read the body as a display of belonging. Screeners are advised to read the bodies of migrants for signs that might reveal the truth from where they come from. The screening booklet suggests scanning migrants from the top to the ground and to check on size, age, color of the hair, haircut, hands, clothes, ‘color of the skin’, ‘shape of the head’, ‘tribe marks’, and ‘traditional jewelry and decoration’ (Frontex 2015), such as rings, necklaces, or tattoos. There is an own folder with a collection of images with ‘tribal marks’ and ‘traditional scars’ including comments that link those marks to gender categorizations like man, girl, and woman, as well as to countries, regions, and tribal identities (Frontex 2015). In combination with the assessment of migrants’ hands, which may be informative about the labor of someone, and the assessment of the clothes of someone, which may provide hints to a milieu, a person is linked to a cultural group, position, and status.

By this, a person’s identity is reduced to a taxonomy of belonging and suspicion is constructed when migrants’ claims differ from these stigmata. The devices of stigmatization not only ignore the colonial and racist practice of recognizing individuals as part of ethnic and cultural groups, but they also misrecognize the complex processes of signification. Body signs, such as scarification, henna, nose, or earrings, have become global (pop) culture phenomena that circulate in and through many different genres. Cultural studies have elaborated in full length about the complexities of signification of body displays and the intertextuality, transmutation, and appropriation of postcolonial identities. The Frontex devices of stigmatization gloss over such complexities and instead construct stigmata that bind migrants to a logic of static, racist, and naturalized belonging.

5.

Purifying

After the screening interview, the screener fills out a so-called identification form, a template with several items that is the basis for all subsequent administrational steps and procedures. Next to ‘date’ and ‘presumed nationality’, several items need to be filled out, such as language, family and given name, date and place of birth, sex, marital status, or job. The form additionally asks for signatures of migrant and screener (I11).

With reference to Bruno Latour, purification is a process that turns a messy, complex, and heterogeneous practice involving humans and non-humans into a drained version with clear cut divides between subjectivity and objectivity, and nature and society (Latour 1993). Purification devices are crucial for administrational procedures, as they produce consistent, comparable, and interchangeable objects, define clear cut roles, and ascribe responsibilities. In our case, the identification form obfuscates the heterogeneous and dynamic nature of the screening process, silences alternative or oppositional claims from migrants, and renders both screener and migrant liable.

First, the identification form constructs a consistent identity. By structuring the identity with a set of fixed categories, it creates discrete and unambiguous information comparable with other identities and applicable to database uploads. The required dates and signatures not only mark the starting point of an administrational case but also turn the outcome of the screening process into an official document both parties formally agree upon. By this, the process of information creation is immunized while details about the interaction process, the devices, and the different forms of suspicion at work with all their problematic effects are obfuscated.

The identification form ignores objections and alternative claims from migrants and establishes an authoritative position of screeners, as they are the ones filling out the form and defining what is the case, leaving out objections, or reframing claims. But even if oppositional claims are documented, they are usually resolved at a later stage of the procedure and in absence of the person in question – usually in favor of the screeners’ assessment (I11, Magalhães 2016). Next to the dislocation of objections, the identification form allows for even another way of silencing objections, that is through residual categories (Star and Bowker 2007). Frontex guidelines inform border guards that they may make use of two residual categories, both proxies for no country of origin: While the category ‘unknown’ implies that migrants were unwilling to cooperate, the category ‘not identified’ states that migrants were cooperative but still no country of origin could be assessed (I23). By this, the identification form renders migrant subjects processable, traceable, and sortable, even in cases where someone's identity remains incomplete or oppositional claims have been articulated. It makes, however, a decisive difference for the interrogated migrants as it can have serious consequences for their subsequent procedure. Some may be approved to apply for asylum or to be tested for the vulnerability track, while others, such as ‘unknown’ or ‘not identified’ people, may be channeled into the repatriation track (MIAR 2017, I11).

Conclusion

This paper reconstructed how suspicion towards migrants is enacted through Frontex screening materials used in screening arrangements at Europe's external border. We analyzed booklets, guidelines, dossiers, questionnaires, images, and forms as devices of suspicion. As such, they not only compile some information but make knowledge operational by providing methods that give hands-on guidance on how screening should be conducted. By this, devices translate suspicion embedded within institutional culture into the everyday practice of identification and credibility assessment.

We identified five devices of suspicion that do different things for the practice of Frontex screening and credibility assessment. They script a typical scenario of the screening situation by specifying its components, actors, and procedures, and they render migrants into fraudsters by default and thus generalize suspicion. They frame interactions between screeners and migrants by creating a diagram of a perfectly ordered world consisting of bordered nation states and hegemonic cultures that produce both a realm of intelligible answers and suspicion sensing deviations from this realm. They code answers as either correct or incorrect and provoke inconsistencies through chains of questions and answers. Starting with suspicious by default, suspicion can only be mitigated gradually throughout the course of questioning and answering. Devices of stigmatization relate utterances of migrants to their bodies. Signs of the body are linked with a region, cultural group, position, and status, which reduce migrants’ identities to a static, racialized, and naturalized taxonomy of belonging. Suspicion is constructed as the deviation from this stereotypical taxonomy. Finally, devices of purification produce consistent, comparable, and interchangeable identities and turn unknown arrivals into processable subjects. These devices obfuscate the violent screening process, silence objections, and render both screener and migrant liable. They also allow the administrational processing of incomplete identities through residual categories.

Our paper suggested that the Frontex materials themselves lack credibility. We can only speculate about the crafting process of information, and we are uncertain (but pessimistic) if Frontex has developed a peer-review process, a country-of-origin information methodology, or standards for source selection and assessment. This is the more surprising considering that the identification of migrants at Europe’s external borders has become a key activity for Frontex and its recently founded standing corps (Frontex 2022). In the context of rapid interventions, joint border operations, and the hotspot approach, Frontex has conducted screenings not only at so-called hotspots in Greece and Italy, but also in Bulgaria, Latvia, Lithuania and Poland (EU COM 2021). As screening devices also play a role in harmonizing disparate practices throughout time and changing actors across the EU, and as Frontex understands itself as a vanguard for evidence-based migration and border management, Frontex should be more attentive to its own practices and procedures of knowledge production.

Through our in-depth analysis, we have shown how devices render migrants suspicious, silence migrants’ objections, and encourage questionable interrogation techniques that seek to provoke contradictions and confessions through means of intimidation, violence, and coercion. This is objectionable considering the manifold and tremendous consequences those dubious methods of knowledge production have for people on the move determining the access to accommodation, social services, and the sorting into different administrational tracks. We can imagine devices that could play another role within Frontex screenings strengthening migrants’ position and binding migration and border control actors to legal standards and procedures of accountability. A procedure including contact points and information materials could advise migrants and encourage systematic objections. Or information about the identification process, purposes, about sites and limits of data storage, and about restrictions of responsible authorities could sensitive both migrants and border control officers in favor of migrants’ rights. But even though such imagined devices might change the situation for migrants gradually, they would still be situated within a framework of migration and border control based upon securitization and suspicion and caught in the inhumane and dehumanizing set-up of the EU-Turkey deal.

Disclosure statement

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

Additional information

Funding

This article was supported by the Deutsche Forschungsgemeinschaft and Exzellenzinitiative des Bundes und der Laender.

Notes

1 After a fire destroyed the Moria camp in September 2020, a new camp has been established on the island, which lies directly at the sea and closer to Mytilini, the island’s capital. Regarding the conditions and constellations regarding this new campsite, see for instance the report by GCR and Oxfam (2020).