Data, detection, and the redistribution of the sensible in international law.

Author:Johns, Fleur

    One dusty day in 2002, at Takhta Baig Voluntary Repatriation Centre near Peshawar in northwestern Pakistan, an Afghan woman--let us call her Amena--entered a nondescript room and sat down in front of a camera. A brief conversation took place with a woman sitting nearby at a computer terminal. Amena placed her chin where she was directed to do so, swept back a few strands of hair creeping out of her veil, and stared straight ahead for a few seconds while a series of photographs of one of her eyes was taken. Almost immediately, a small alarm sounded on the computer terminal of the woman seated alongside her. Amena was gently ushered toward the other side of the room for discussions with other officials. Some short time later, she was advised that her request to the Office of the United Nations High Commissioner for Refugees (UNHCR) for a modest cash grant and some supplies to aid her and her family's repatriation to Afghanistan had been denied. This was because, according to output of the UNHCR's iris verification program, she had already received assistance earlier the same year. When asked, Amena admitted that she had indeed sought UNHCR repatriation assistance multiple times, under pressure from family members. She walked away. Soon, she could soon no longer be seen amid the press of trucks, cars, bicycles, and people that stretched to the suburbs in the distance.

    Let us now imagine Amena in the same scenario--seeking UNHCR repatriation assistance at a refugee processing center in Pakistan near its border with Afghanistan--but envisage how it might have occurred in 1992. In that year, between April and December, an estimated 900,000 Afghans returned to Afghanistan with UNHCR support in what has been described as the "largest and fastest repatriation prografm] ever assisted by the UNHCR" up to that point. (1) The UNHCR's role and responsibilities vis-a-vis Amena at that time would have been the same, but its enactment of them would have been quite different. Most notably, iris scanning would not have featured in the process. Amena and members of her family would have been required to attend an interview with UNHCR staff, to present documents (including ration books) and photographs for inspection attesting to their identity, and would likely have waited a long time for a determination of their eligibility for repatriation assistance. This would have been made on the basis of staff observations, including potentially the staffs sense of whether Amena and her family "looked tired and sick." (2)

    Amena's story--in both these versions--is a fictional one, but it depicts a change in UNHCR practice that did in fact take place. The UNHCR did introduce biometric registration (specifically, iris scanning), alongside other identification methods, in its support for Afghan repatriation from Takhta Baig in September 2002 and it has since introduced this at other locations. (3)

    Faced with this two-part story, there are a number of questions and concerns that one might expect international lawyers to raise, even while acknowledging that the changes described may have beneficial effects. Was Amena's "confession" of having sought multiple rounds of assistance truthful, or was she in fact wrongfully denied assistance on the basis of some error, human or mechanized? Even if that admission was accurate, were her rights respected in the course of eliciting it, in both the 1992 and the 2002 modes? Was her consent to the process in each case informed and freely given? Were the data drawn from her body during iris scanning, or disclosed during interview, kept secure, anonymized, and confidential? Were her religious freedoms respected during these interactions? What of the pressure from her family to which she referred--could she be a victim of sexual or gender-based violence? Adopting a wider-angle lens might elicit the following questions: Under what lawful authority was the UNHCR engaged in repatriation from Pakistan in 1992 and again in 2002? And in view of the security conditions in Afghanistan at these times, and obligations owed to refugees under international law, were these programs sound and well-advised?

    These are all important questions. They are not, however, the sorts of questions on which this article dwells. What this combination of questions makes of this scene, in both time periods--a site of principled, rights-based engagement, forensic inquiry, privacy, and protective conundrums--is also not, in the main, what this article aims to make of it. What this article invites us to see in both versions of Amena's story is the constitution and exercise of jurisdiction on the part of an international institution--in this instance, the UNHCR--and also on the part of the discipline of international law. It also invites us to take note of significant changes in how this takes place, especially through the mediation of technology.

    The changes to which this two-part story draws attention could signal expansion in international law's and institutions' capacity, enabling more precise and rapid responsiveness to need. It may be that automation allows international institutions to anticipate and temper violence and suffering like never before. Yet the very same changes may threaten to undermine international law's and institutions' operations, or narrow their effective jurisdiction, insofar as they heighten inequalities, magnify distrust, and impede communication and connection. Exercises of international legal authority that cannot be understood, represented, or justified in recognizable terms may be prone to rejection. The changes in international legal and institutional work to which this story draws attention are also important because of their role in shaping allocations of power, competence, and capital. It is these distributive implications on which this article will focus.

    Sensing practice--or the work of trying to detect and verify certain worldly phenomena-that is carried out by international lawyers and international institutions creates uneven distributions of capacity and resources. It contributes to the creation and allocation of divergent ability to generate shareable sense-information. That distribution helps to ensure that particular configurations and aggregations of sense-information come to be experienced in common (however contentiously) as the factual condition of the world, while others do not: configurations that, in turn, shape prospects for lawful relation and action. This is not necessarily a bad thing; there are good reasons for vesting certain detection and verification responsibilities in those with particular expertise. Nonetheless, that which international law takes as its factual background, or the context in which it operates, are made out of relations of inequality to which international legal work contributes. This is important to highlight because it contradicts international law's implicit claim to be about all, by all, and for all. Far from enclosing "all" in an equalizing embrace, and responding to naturally occurring conditions of common humanity, international law generates prospects for experiencing commonality out of its profoundly unequal distribution of eligibility to sense, and to be sensed.

    With the advent of new, technologically advanced modes of data gathering and analysis, those relations of inequality are taking on new configurations. These are reconfigurations with which international legal thought, doctrine, and practice are, in the main, poorly equipped to deal. One reason for this deficiency is that the introduction of machine learning (a form of artificial intelligence) to international institutional work takes those operations outside the scope of preexisting legal analyses of, and methods for grappling with, quantification and statistics. The very experience of characterizing the world's conditions, and of exercising power to govern, make legally significant decisions, and conduct juridical relationships on the global plane are under revision in the face of automation, especially with the growing prevalence of machine learning. Also shifting is the global distribution of capacity to engage critically with law and policy, and to attest to or verify phenomena to that end. These changes are occurring in ways with which international lawyers, and the publics of which they are a part, are, and should be, concerned--more so than most in the international legal field have been to date. These changes are explored here by recourse to two illustrative snapshots of changing technical practice: those surrounding the global movement of weapons (through the work of the International Atomic Energy Agency (IAEA)) and the mass movement of refugees (through the work of the UNHCR). As intimated by the reference to "snapshots," the accounts of these institutions' work put forward in this article are designed to be suggestive, not definitive.

    Part II of this article explains some key terms and how the contribution made here both builds upon, and departs from, other scholarship in international law, historical and contemporary, including scholarship from the pages of this Journal. Part III describes how sensing practice has occurred in two areas of international legal and policy work: in the tracing and handling of certain categories of nuclear material; and in the detection of, and response to, mass movement on the part of refugees. Part IV discusses some developments reshaping, or carrying potential to reshape, these two areas of international law and policy work--nuclear nonproliferation and the verification of refugee numbers and identities--and examines how these developments may affect the common sensorium of international law's crafting. The last section is a short conclusion, reflecting on the possibilities occasioned and questions raised by studying distributions and redistributions of the sensible in international law and policy.


To continue reading