This essay adopts a critical perspective of the idea of humanizing refugee research. It argues that much social scientific research is intrinsically dehumanizing, as it simplifies and reduces human experience to categories and models that are amenable to analysis. Attempts to humanize research may productively challenge and unsettle powerful and dominant hegemonic structures that frame policy and research on forced migration. However, it may replace them with new research frameworks, now imbued authority as representing more authentic or real-life experiences. Rather than claiming the moral high ground of humanizing research, the more limited, and perhaps more honest, ambition should be to recognize the inevitable dehumanization embedded in refugee research and seek to dehumanize differently.

Cet essai adopte une perspective critique à l’égard de l’idée d’humaniser la recherche sur les réfugiés. Il soutient qu’une grande partie de la recherche en sciences sociales est intrinsèquement déshumanisante en ce qu’elle simplifie et réduit l’expérience humaine à des catégories et des modèles se prêtant à l’analyse. Les tentatives d’humaniser la recherche peuvent efficacement remettre en question et déstabiliser les structures hégémoniques puissantes et dominantes qui encadrent les politiques publiques et la recherche sur la migration forcée. Par contre, elle pourrait les remplacer par de nouveaux cadres de recherche, désormais imprégnés d’autorité comme représentant des expériences plus authentiques ou plus proches de réalités vécues. Plutôt que de revendiquer une position de supériorité morale en humanisant la recherche, une ambition plus restreinte, et peut-être plus honnête, serait de reconnaître que la déshumanisation est inévitable dans la recherche sur les réfugiés et de chercher à déshumaniser autrement.

In this forum, we are focusing on the theme of humanizing refugee research. To my mind this suggests making our analysis more nuanced, more responsive to the reality of people’s experiences, and contributing to more humane outcomes. The notion is rather comforting and positive. Who can object to it? This short essay offers some provocation that counters this cozy perspective. I start by looking at the dehumanization of refugees in the process of bureaucratic labelling. I then look at how this relates to the similar process of categorization that is fundamental to much social scientific research. My proposition is simply that the work of social science can be seen by its very nature to be dehumanizing because it is impossible to analyze the full human experience. Instead, we work with models and simplifications that amplify some aspects and play down others. I suggest that our challenge is to recognize this and ask how we can dehumanize differently, for example, by looking at alternative ways of formulating these models and more critical analysis.

It is clear that there are competing narratives shaping states’ policies on refugees and their practical responses to them. On the one hand, there is the story of ideals and a moral imperative. This emphasizes state concerns with upholding rights—in particular, basic human rights and the right to seek asylum—and duties—the international obligation to provide protection. On the other hand, there is the narrative of refugees as a burden and threat for the state and the society that receives them. Here, the noise about human rights is muted. Instead, the first response of states is more concerned with controlling access to refugee status and the resources associated with it, such as humanitarian aid or state welfare. An important technique for this control is the establishment of rigorous criteria for the recognition of refugees. These criteria are used as the basis for challenging the status of the people making claims for asylum with a presumption that they do not meet them.

In crude terms, we can see the former narrative paraded more strongly when it comes to claims about the obligations of others—for example, the international concern about Kenya’s repeated threats to close Somali refugee camps in 2016 and 2021 (Muiruri, 2021). While it was prominent within the boundaries of western Europe during the Cold War, this rights narrative has lost ground to that of burden and threat since the late 1980s, particularly in recent years, as the number of asylum seekers arriving in Europe from poorer regions of the world has grown dramatically. One of the impacts of this shift towards seeking grounds for rejecting asylum claims is the development of ever finer-grained categories. For example, between 2013 and 2014, Germany created three new forms of humanitarian status that admitted Syrians to the country but with fewer rights than those available to convention refugees (Tometten, 2018). By such means, people are diverted to a secondary status, and the criteria for eligibility to full refugee status can be narrowed. This limits the numbers of people with entitlement, whether to protection, humanitarian aid, or other support.

To some extent, we can argue that such categorization and bureaucratic labelling are always somewhat dehumanizing. They demand that we focus on particular qualities of people to group them together as people who will exhibit particular behaviour, have particular needs, be exposed to particular risks, and so forth. AsWood (1985) and laterZetter (1991) argued, referring particularly to refugees, labelling replaces the individual with a stereotyped identity; the complex human story is reduced to the bare facts of the case. It is on the basis of this bureaucratic process that critical decisions are made about refugees’ lives. Subsequently,Zetter (2007) observed how new labels emerged to tease out distinctions between people and determine their level of entitlement. He cited the way the United Kingdom’s Home Office used the label asylum seeker, breaking it down to seven subcategories, or segments, including minors (under 18 years of age), those making late or opportunistic applications, or those who had passed through a third country, where they could have claimed asylum. Once applicants were assigned to a segment on the basis of their fit with assessment criteria, the outcome of their claim was largely determined.

While it is easy to decry the dehumanizing aspects of the labelling process, it serves an important role: “Bureaucracies need labels to identify categories of clients in order to implement and manage policies designed for them” (Zetter, 2007, p. 184). It is a necessary part of a bureaucracy to ensure that the action of the state can be based on “objective,” “standardised,” or “technical” criteria rather than a result: either of individual negotiation that reflects a person’s power and access to resources, whether in the form of wealth, knowledge, social networks, or other forms of privilege; or of states privileging particular groups of people over others on the basis of race, religion, political expediency, or other grounds. In this sense, formal, public criteria for determining action introduces some element of accountability for states and prevents the most egregious bias. At minimum, it makes it possible for deviations from the criteria to be noticed and challenged.

However, such bureaucratic labelling may only create a superficial veneer of neutrality and justice as the criteria underpinning the labels may be carefully designed around political and social discrimination. This was evident in the late 1940s and 1950s, when European governments created an amazing array of categories to separate out those they saw as desirable from the undesirable migrants, refugees, and displaced persons in the aftermath of the Second World War (Gatrell, 2019, pp. 35–68). These are far from being objective or merely technical bureaucratic processes. Therefore, there needs to be ongoing and important critical discussion about the nature of these labels and the criteria that assign them. They will embody and entrench structures of privilege, reinforcing and establishing norms of identity and behaviour. They must be contingent on evaluations and critiques of their (im)partiality and subject to revision. Nonetheless, the labels at play at any particular time perform a critical role in enabling states and other actors, such as non-governmental organizations, to operate and provide protection, however flawed.

In the face of such bureaucratic systems, we are all reduced to the categories that we are deemed to fit—student, employee, graduate, pensioner, citizen, and so forth. In this sense, this practice of labelling is dehumanizing as it cuts away a large part of our lives and does not acknowledge us as unique, rounded, and complex individuals. However, it is an open and contentious question whether it would be better to have a state bureaucracy to which all the different aspects of our lives are legible and known (assuming that this were even possible). Perhaps this idea of disclosure takes us close to the vision being opened by the recent developments in data capture by the likes of Google, Facebook, or Sesame Credit in China (Lee, 2019). This makes many of us very nervous. I am not sure we want to be seen completely in the round, particularly by states.

Reflecting on a comment from Elena Fiddian-Qasmiyeh (at the workshop on Humanising Studies of Refuge and Displacement, 2018), I find it is important to recognize the performativity of people’s presentation of themselves. At times, they may consciously play up particular aspects of their lives, depending on the audience and their ideas of its expectations. This is a common experience for someone undergoing an interview, whether for a job or an asylum application. It is not just on such occasions that we perform. All of our social interactions constitute a performance that varies with the audience and the context. Our behaviour changes unconsciously as we shift between settings. As a result, there is no consistent and objective identity to be observed; our identities are produced by our constantly varying performance. In the face of this fluidity, bureaucrats will inevitably struggle to work out who is the “genuine” asylum seeker. Equally, those objecting to the dehumanizing processes of states may be demanding that they recognize a person’s “real” identity, which does not exist.

Hence, rather than harbour a growing expectation of people showing authenticity and openness, perhaps the best that should be asked is that they are not subject to the crude system of labelling and that stereotypes be challenged. These reflections make me wonder if the creation of ever finer categories that take more account of the complexity of people’s lives (as we often see in refugee or asylum systems) perhaps comes closer to recognizing them as more human. If this is correct, it brings up the question of whether our complaint should be about the dehumanization of the policies so much as the politicization of bureaucracies. This is not to suggest that bureaucracies have ever been apolitical; but the ever-sharper delineation of categories, often couched in terms of technical refinement, serves to exclude more people from recognition as refugees or other categories that secure their protection and access to rights. It can be seen as an important element in the political project of dismantling of the architecture of rights.

With this in mind, I turn to the issue of humanizing refugee research in a turbulent world. What are we concerned with? As social scientific researchers, I think we are seeking to make some generalizations: we aspire to take our knowledge from one setting to build up our understanding of another setting. We use concepts and categories to build up “models” or ideas of how things connect to each other. Like bureaucracies, social scientists (and here I would include even anthropologists) are reductive in a somewhat ordered way. They may carefully rationalize their decisions to include some aspects of life and exclude others to create this order. Of course, given the comments above, we could say that all human interactions involve some element of reductionism, as we can see only what is performed by others, but it is not systematic in the same way. For researchers, this reductionism results in a set of analytical categories that can be shared between social scientists and used to try to make sense of the world and to generate new insights.

Some years ago, I wrote of the importance of distinguishing between such analytical categories and those used in policy (Bakewell, 2008). I argued that research should not be bound by policy categories as it creates three critical constraints. First, it limits the questions that we ask as it sets up a taken-for-granted view of the world that compares those within the category (say, refugees) with those outside (say, hosts) and prioritizes this above other intersecting areas of comparison (e.g., class, education, or sexual orientation). Second, it confines the object of study to particular categories of people seen as in need or otherwise problematic (e.g., displaced people). Finally, it constrains methodology and analysis as we frame discussion around the issues that are legible to policy-makers. Even when we challenge policy categories, we come back to them to make our work comprehensible to others.

Reflecting on this over 10 years later, I am not sure that the issues have changed so much. Many others have written on similar themes in the intervening years, and in this essay, I make no claim to do justice to all this inspiring literature (e.g.,Brubaker, 2013; Cole, 2021; Crawley & Skleparis, 2018; Dahinden, Fischer, & Menet, 2021; Benson, Temprosa, & Shlebah, 2020; Nguyen, 2019; Sözer, 2019; Will, 2018). This work makes it clear that there remains the great challenge of navigating the priorities and frameworks that are shaped by powerful interests—at times including those of academics. Of course, states, international organizations, businesses, and a whole host of other powerful actors not only are implicated in creating the problems confronting refugees and other forced migrants but also play a crucial role in addressing them. Along with any others who aspire to ameliorate the situation, scholars need to work constructively with these parties and manage the multiple compromises this entails, which includes engaging with the categories they impose. While we may question the way that people are assigned to categories such as refugee or harbour doubts about the use of the term at all, it seems inevitable that we must use them if we are to communicate and have any impact in the world. This does not mean that we accept these categories as given, nor do we need to base our analysis on them. The puzzle we face is how to challenge and possibly subvert these categories to provide alternative perspectives.

Much of my research is on migration and refugee issues, which means that the fact that people have moved in some way is often at the forefront of my mind. I tend to assume its salience to understanding their position. I explain the behaviour of someone with reference to their status as a refugee or other forced migrant, and it is easy to assume a set of vulnerabilities to particular hazards that are associated with this status. For example, in a study on returns to Somalia conducted in 2018, it was clear that there were very high levels of displacement, but we found that for many the category of internally displaced person (IDP) is seen as a trap. Those identified as IDPs are subject to discrimination—accused of bringing disease, unemployment, and insecurity—and are more vulnerable to rights violations, including violence and evictions with no recourse. However, their living conditions and economic position may be little different from rural–urban migrants, many of whom have been compelled to move for similar reasons. These migrants are not recognized as IDPs, and they are able to avoid the stigma of the label. Others have the resources to move into better living conditions, and their economic position helps ensure they avoid being labelled as IDPs, despite having been forced to leave their homes under the threat of violence (Sturridge, Bakewell, & Hammond, 2018). This suggests that the (policy) category of IDP may say much more about issues of class and clan than it does about people’s experiences of displacement. As an academic, it is possible to explore how these different axes of inequality intersect both with the way that the category of IDP is applied in policy and experience of displacement. Nguyen (2019) takes this much further in a reconstruction of the refugee category using the concept of refugeetude.

My discomfort with the policy categories does encourage me to question them in this way, along with many others (e.g., Cole, 2021; Crawley & Skleparis, 2018; Sözer, 2019). Nonetheless, even when I challenge the meaning of the IDP label, I am implicitly replacing it with other categories based on other criteria: income, livelihood sources, type of housing, clan, place of birth, and so forth. This may result in a much more nuanced intersectional account that draws attention to the problems with the term IDP, challenges some of the assumptions about those categorized as IDPs, and accounts more for people’s lived experiences. Nonetheless, as a research team, we still impose our interests on the analysis. We focus on some aspect of people’s lives and inevitably neglect others. The process is still reductionist and to some extent dehumanizing. Perhaps this is a legitimate part of our role as social scientists, and it can serve to generate some useful understandings. I struggle to see a way to escape from some form of categorization to make some sense of the world.

This then makes me skeptical about the feasibility of humanizing refugee research, in the sense of enabling it to take account of the full human experience. As I have argued above, the process of much social scientific research requires the researcher to work with abstract concepts and models that cannot possibly capture the complexity of human life. Moreover, there are profound questions about the nature of personhood or identity that can be represented in research. Our attempts to deconstruct the narrow, reductive, policy-related concepts and categories may result in the formulation of much more nuanced and grounded concepts and categories, emerging from empirical research and dialogue with a wide range of actors, in particular, the voices of those often marginalized or silenced. These are the product of a different political process, but it is not clear that they are necessarily any more of a “natural,” “humanized” set of categories than those we started with.

As we work more closely with those who are marginalized, attempting to humanize refugee research may inflate our sense of worthiness. It may change profoundly the way situations are analyzed and understood. However, it may also create a new hegemony of concepts and categories that are imbued with a sense of authenticity and moral superiority but no less dehumanizing. I should stress that I am not for a moment suggesting that this process of changing the frames for debate and analysis is not worthwhile. I have not given up on my call for policy-irrelevant research into forced migration (Bakewell, 2008). My skepticism is about aspiring to humanize refugee research. It may be more productive and usefully subversive to acknowledge the dehumanization inherent in our work as social scientists. We may not want to celebrate it, but we need to politicize it, dehumanize differently, and provide alternative perspectives so that we can resist the standard scripts, such as those embodied in the categories that frame so much research into forced migration.