Controlling borders: A note on ordinary technologies of power

When scholars of migration and borders investigate technologies, it is often very contemporary technologies that capture their attention. This is understandable: there are some very good ethical, legal, political and economic reasons why we should carefully scrutinise emergent technologies, asking how the phenomena known as “artificial intelligence” (see Aradau, this issue), “drones” (see Topak, this issue) and “digital borders” (see Frowd et al. and Molnar, this issue) shape and differentiate the experience of people on the move. But in this commentary, I take a somewhat different path: I argue we should also think critically about a theme I want to call ordinary technologies of power. Here I mean not necessarily the world of “low-tech” (Bonelli & Ragazzi, 2014), though this matters too. What I have in mind are technologies that have been around for some time, that are settled, and well-integrated into societal and governmental domains—whether in the migration field or elsewhere. It is precisely because they are settled, embedded and not spectacular that we easily take them for granted, overlooking the subtle ways in which they mediate relations of power and domination in the governance of borders and migration. Certainly, these technologies were once “new,” and as with “AI hype” (Aradau, this issue) at the moment, the object of all sorts of fantasies, fears and projections. But their lustre has faded. This can be opportune for the scholar of technologies and migration. Studying old technologies can be a useful exercise if it gets us to approach new ones with a deflationary eye and if it allows for a degree of caution and circumspection about the extravagant claims that are sometimes made for new technologies. In the interest of space but also in the spirit of pragmatism, I will focus this discussion around one particular family of ordinary technologies relevant to questions of migration: the aeroplane. And to be even more specific, I will confine myself to aeroplanes inasmuch as they feature in operations of deportation. This is a useful exercise not just to build an argument for opening up the blackbox of old and seemingly well-established technologies. It is also useful as a way to partially redress what I have elsewhere called the ingression bias. This is the tendency in scholarship to examine the role technologies play in policing mobilities towards the territories of the global North while overlooking the role technologies play in the expulsion of people (Walters, 2018: 2799). Deportation, I argue, is no less technical than other modalities of policing and enacting borders, including the new technologies called “AI.” But it is, of course, not merely technical, as we will see. Let us start with an early attempt at air deportation. On 31 March 1939, the UK newspaper Daily Mail published a large photograph of a man, Oskar Goldberg, being carried by four police officers against his will to a waiting aeroplane at Croydon airport (Frankl, 2017).1 It was actually the same Danish tri-motor plane that had been chartered the day before to fly him, together with 12 other Jewish refugees, from Warsaw via Copenhagen to this aerodrome on the outskirts of London. In fact, hundreds of refugees had recently landed there, all fleeing the murderous advance of Nazism. Britain was not prepared to offer asylum to Goldberg, and his group and the authorities were now seeking to deport them. The publication of this and many similar photographs in British newspapers was no doubt a factor that helped these refugees gain public support and official permission to remain in Britain. But if the generation of positive publicity about their plight was important, so too in a much more immediate way were certain affordances of the aeroplane itself. It is reported that after they were forced to board the plane, the refugees pounded on the windows and threatened to jump out of the plane in flight. The Danish pilot was so alarmed by the prospect of disturbance in the air that he refused to take off. This refusal bought the refugees vital time. Air deportation is today a very routinised and widespread practice.2 Yet, the possibility of obstructing a given deportation by leveraging the relative fragility of mechanical flight is something that can still disrupt expulsions to this day (Independent, 2017). Goldberg may be long dead but it seems his ghost still haunts the deportation machine. Aeroplanes were already a relatively settled technology by 1939. Less than four decades had passed since the Wright brothers had famously demonstrated the possibility of motorised flight, taking off from fields in North Carolina and then Ohio. At that time, the aeroplane had already been pressed into service for commercial, diplomatic, touristic and military purposes. Indeed, <2 years before Goldberg's photograph was published, Picasso's Guernica had been unveiled at the Paris International Exposition. Guernica was certainly not the first settlement to be bombed from aeroplanes, as Libyans, Iraqis and Moroccans among others will attest (Lindqvist, 2001). But like no other work of art, this mural would come to monumentalise the aeroplane's capacity for destruction, heralding the dawn of the age of air war. We can only surmise that as tricky as it must have been to turn aeroplanes into bombing machines, or make them capable of speeding wealthy tourists, movie stars and businesspersons across vast oceans, British authorities had by 1939 not yet worked out how to transform the aeroplane into a reliable technology of deportation. It is one thing to carry people who want to fly, indeed for whom flight was a badge of wealth and distinction. It is quite another to transport people against their will. How do you pacify and control recalcitrant subjects herded into a narrow metal tube flying at great speed above the clouds? For the aeroplane to become a viable and routine technology of deportation, authorities would need practical answers to these questions. These answers did not appear overnight. The history of what we have called air deportation remains largely unwritten.3 But any attempt at writing it will surely include an account of the different practices, strategies, regulations and technologies that have been used to adapt the aeroplane to the ends of the deportation state, along with the kinds of questions authorities, experts, campaigners and others have asked about these procedures—the field of questions Foucault (1997) called problematisation. Can we use drugs like Haldol to sedate deportees or is that illegal (Weissman et al., 2020: 479)? Is it acceptable to use pain to elicit compliance and if so, how (Walters, 2017)? If we use straps and cuffs to immobilise the deportee in transit, what form should they take and what training should officials receive in their use (Bosworth & Singler, 2023)? How many security personnel do we need during the operation to ensure a “superiority of numbers” (IAPNCM, 2014: 68)? Why not charter planes so there are only deportees and security personnel on board, distancing the whole operation from the gaze and interference of regular passengers (Fekete, 2011)? Might we scale up assisted return programmes so that we will not need to use force at all: offer people cash incentives and other supports so that more of them will fly of their own accord (Cleton & Chauvin, 2020). Or make everyday life for unwanted migrants so “hostile,” so debilitating that they will move on themselves (Tazzioli, 2022). Reflection on these varied practices and devices matters. It matters conceptually because it complicates the image of the aeroplane as a self-standing, bounded technology of deportation. Rather than fixating on objects in isolation, we have to understand how technologies operate within networks; indeed, that what we call technologies are networks. A plane does not fly without an extensive infrastructure of supports such as radar systems, fuel supplies, aviation regulations, seat markets, ground crews and, increasingly, AI (Xiang & Lindquist, 2014: S124). Likewise, planes do not become the backbone in the deportation programmes of the powers of the global North without a whole network of experts, devices and technologies of the kind just mentioned. It is these that allow deportation to flow through the veins of the world's aviation system, becoming global where once deportation, like convict transportation, was largely confined to geographies of land and sea. But this reflection also matters politically. And here, there is a wider lesson for anyone interested in technologies. When it is written, the story of air deportation will not be a linear one in which authorities and experts devise ever more efficient methods and technologies of removal. Rather, it will be a history that shows that this entire domain is shot through with struggles, shifts and sometimes reversals on every scale and along multiple axes. If many European authorities have come to favour a deportation enforcement model that relies on the superior numbers of security officials coupled with some very old technologies like the waist restraint belt, it is because various struggles have succeeded in rendering other methods as unethical or impractical. If there are empty seats on deportation charter planes, or indeed numerous cancelled charter flights (Bosworth & Singler, 2023), it is because detainees and their allies manage, whether through legal appeals, political campaigns or physical resistance—like Goldberg on the runway in Croydon—to thwart or at least delay the state's most concerted attempt to remove them. In short, deportation is an eminently technological domain even if many of these technologies are not especially “new.” But it is also a political domain and it is this that both frustrates these technologies of power while inciting attempts to fashion new ones. The opinions expressed in this Commentary are those of the author and do not necessarily reflect the views of the Editors, Editorial Board, International Organization for Migration nor John Wiley & Sons.

Récupéré en direct depuis OpenAlex et désinversé. Les résumés ne sont pas conservés dans cette base de données : les index inversés représentent 8,6 Go des 9,3 Go de texte de la base, et le serveur dispose de 13 Go libres.