‘The whole history of Palestinian struggle has to do with the desire to be visible.’

Edward Said

On the evening of March 3rd, 1991, Rodney King, a 25-year-old Black American, was pulled over by LAPD officers after a high-speed chase. What followed was recorded by George Holliday on a handheld camera. The grainy footage showed four officers—Laurence Powell, Timothy Wind, Theodore Briseno, and Stacey Koon—brutally beating King with punches, kicks, tasers, and batons. In 85 seconds, King was struck 56 times. The video, aired by CNN, went viral, triggering global outrage and leading to a criminal case against the officers. This was the first time a court had access to such clear and visual evidence of police brutality against a young Black man. The video became the trial’s central piece of evidence.

Having unprecedented confidence in this piece of evidence, the prosecutor opened his case by saying: ‘What more could you ask for? You have the videotape that shows objectively, without bias, impartially what happened that night. During the closing argument, he played the video once again and asked the jury: ‘Now, who do you believe, the defendants or your own eyes?’ Yet, the jury acquitted the officers.

During the trial, the defense, instead of discrediting the footage, decided to use the footage and instead discredit what people saw in it. To begin with, they claimed that the technological footage cannot speak for itself. It needs to be talked for. Then, they presented the short video not as a continuous event, but as slow-motioned, cut-outs of 56 single blows. Isolating each kick, punch from the preceding and the ensuing one.  Then, they placed each cut-out in a professional discourse associating escalation or de-escalation of force (i.e., beating or pausing) to corresponding signs of aggression or submission (i.e., different movements in Rodney King’s body). A cramping leg, a moving hip, or a raised hand was tagged as signs of aggression and therefore legitimating the escalation of violence by police. Lying flat, motionless, almost lifeless under the blows of batons and kicks was read as submission and explaining a pause in beating. Through a mixture of technological reordering of the events and professional discourses, the defense team convinced the jury that what the whole world saw as objective evidence of brutal police violence was, in fact, a measured, textbook application of the law.

This case illustrates that although sight is a physiological and objective act, seeing, as a basic act necessary for making sense of the evidence and meaningfully making the distinction between categories, here, of (un)lawful use of force, is in fact socially situated and technologically mediated. As such, technological seeing, that is, slow-motioned, replayed, re-ordered, separated vision, enables another mode of representation, where new meanings are produced in places where such meanings otherwise did not exist. In Rodney King’s case, what seemed to anyone as agony, cry, and struggle of a body under attack, was presented as signs of aggression, which justifies more attack. 

This reconfiguration of meanings through visual mediation opens the law to contestation and uncertainty over the truth of an event and hence the distribution of its protective categories. This social permeability of law allows for its categories to be continuously exposed to reconfiguration in ways otherwise unimaginable in exposure to new technologies of vision. This case tells us something about the law too, that is, the sources of law are not just to be searched in the law books, cases, norms, or rules of interpretation, but that law finds its sources in the relationships and social formations. Formations that are sometimes made through technological operations. 

This dynamic extends to another form of state violence: War. On 3 April 2024, the Palestinian-Israeli magazine +972 and The Local Call revealed the Israeli military’s use of an AI target-generating system called Lavender in its ongoing military campaign in Gaza.

Put simply, Lavender processes data on Palestinians to algorithmically generate human targets. Accordingly, ‘During the first weeks of the war … Lavender … clocked as many as 37,000 Palestinians as suspected militants — and their homes — for possible air strikes.’ In observation of the outrageous death toll brought by Lavender, we may see the problem to be one of a lack of regulation and oversight, a remedy to which would be more human supervision and regulatory control. But lavender, like its American predecessor, the Disposition matrix – used in Iraq and Afghanistan – signals a deeper problem if we unpack the laws of war’s relationship with technologies of visualization.

The key to understanding the impact of advanced visual technology of targeting, I propose, requires acknowledging how the laws of war themselves produce an ‘order of visibility’ in order to legitimize lethal violence. The fundamental principle of the laws of war that draws the difference between lawful and unlawful modes of wartime killing is the principle of distinction. When broken down to its operative elements, this principle is simply an order of visibility, a modality of visuality if you will. That is the organization of a field of vision through strict regulation of modes of appearance for people and places, as well as the sense that is to be made of them, and finally authorizing the use of force against some modes of wartime visual appearances and prohibiting it against others.

It is for this reason that the Geneva Conventions contain numerous rules about appearances, dressing, perfidy, camouflage, or the use of color, and patterns on buildings, signs, flags, vehicles, etc. Soldiers must signal hostility through the wear of a distinct military uniform or signs visible from a distance, protected vehicles must be painted white, signs signifying civilian status must be painted in a specific size on specifically measured flags with material visible from a distance and in the dark, and so on and so forth.  The battlefield in the eyes of the Geneva Conventions is a visually distinct and ordered field in which every sign, color, and pattern has a stable meaning.

Problems arise in asymmetric warfare. Non-state or insurgent fighters forgo the visual markers. Such illicit invisibilitythen explains, but does not justify, the reliance on technologies that generate targets based on digital visualizations.

Militaries like the US and Israel have adopted technological solutions that combine hyper-visualisation and surveillance of the battlefield with algorithmic analytics referred to as pattern-of-life analysis. The US army defines pattern-of-life analysis as ‘an extensive, intelligence-oriented activity that establishes a physical, visualised, and tangible infrastructure of relations in the environment of operation by ‘connecting the relationships between [observed] places and people.’ The ultimate function of such analysis is to make the invisible enemy networks ‘visible and vulnerable’, thus negating ‘the enemy’s asymmetric advantage of denying a target.’ Everything that is recorded by surveillance technologies, including everyday activities, movements, patterns of purchases, regular visits to different locations, or regular meetings or online/ phone contacts with other individuals, is analysed to produce a visualised network of relations between people and places, and determines their degree of targetability.

These tools visualize the battlefield as an array of networks in which nodes are identifiers (IP address, phone number) and links are their interactions. A node’s importance, then, is determined by the intensity or frequency of the interaction it has with other nodes. These factors, in addition to the overall character of the observed network, in turn determine the value of each node (individual) as a probable target. 

In the Israeli context, these links are referred to as features. Lavender processes surveillance data to identify Palestinians with similar behavioural patterns to previously defined features of an enemy target. The result is assigning a risk value from 1–100 to the observed population. The allocated risk determines the individual targetability. Lavender and Disposition Matrix are essentially tools for providing alternative modes of visibility, in the form of patterns and networks that compensate for the insurgents’ intentional invisibility.

The stakes extend beyond questions of legality or temporal and spatial expansion of wartime violence. In reenacting the law’s requirement of visibility for legitimation of violence, these technologies are rewriting the very idea of legitimate violence in ways otherwise unimaginable. What is happening is a great perversion of the laws of armed conflict in producing enemy targets in terms of resemblance, approximation, and association, as opposed to the hostile actions of politically motivated agents. As such, rendering as targets whoever is in proximity of whatever that resembles an enemy feature. In this sense, the digital technologies of targeting are tools for recasting as targets an otherwise civilian population. Thus, the new modes of representation (i.e., creating visibilities and meanings associated with those visibilities) facilitated by these technologies are attempts to produce new forms of wartime legality.

Industrial Scale of Killing

Target-generating technologies are often described as factories of targets. This is so because the politics of representation enacted by these machines has a concrete impact not only on the people who are observed and targeted, but also on those who survive. These machines do not operate in a vacuum. The lives that are constantly monitored, those that are taken by missiles and bombs, and those who survive these attacks share an ecosystem – a platform of affecting and being affected by – with the very technologies that monitor and target them. As Lavender targets patterns of lives and features, the targeted population in turn adapts their lives to patterns that, they come to understand, do not trigger the lethal algorithms of these machines. This is to say that as bombs strike certain patterns of life, they eventually produce, favour, or impose other features as normal and unthreatening ones. Then deviations from these newly imposed patterns, in turn, signify new anomalies and, as such, create new possibilities for lethal targeting. It is in this sense that these automated machines become self-perpetuating, industrial-scale killing machines that can produce new targets out of patterns of life that they initially did not consider bearing that meaning.

Can You Be Palestinian and Look at This?

Technologies of vision entangled with state violence often involve and provoke another politics of vision, that is, politics of witnessing and evidencing. In an essay titled: ‘Can you be BLACK and Look at This?’ the poet Elizabeth Alexander reflects on the political, social, and interpersonal consequences of the viral circulation of images such as that of Rodney King. She asks how the constant circulation of footage of Black bodies in pain consolidates the category of Black? She explores the formation of self and collective identity in the midst of the constant circulation of images of Black people as victims of extreme violence for public consumption. And asks, ‘How can this photograph exist?’ It is perhaps fair to ask similar questions in the context of Palestine and the ongoing genocide. Can you be Palestinian and look at this?

Reflecting on a similar set of questions but in the context of Palestinian occupation and its representation in films and documentaries made by Palestinians as well as Israeli filmmakers, Gil Z. Hochberg raises different questions, one that engages us, the onlookers, as well.

In her analysis, Hochberg argues that the Israeli-Palestinian conflict ‘cannot be simply recognized only in terms of colonial land grabbing, but that classification of distinct ethnonational and religious identities is dependent upon particular visual practices and distribution of visibility.’ After all, the mantra of ‘a land without a people for a people without a land’ is an illustration of a politics of visibility that, from time immemorial, renders Palestinian political subjectivity and being invisible. A politics of visibility in which Palestinians keep appearing as a shapeless mass, not a people or a community with roots in the lands they inhabit, just masses of entities that happen to be there, or in the eyes of the killing machine as entities lacking political agency lacking an intentionality, a formless mass of mendaciousness and malevolence by virtue of proximity and association. No need, ever, for their individual knowability.

In her book ‘Visual Occupations: Violence and Visibility in a Conflict Zone’, Hochberg appears to be skeptical that the way to undo the injustices of occupation involves providing witness accounts through alternative modes of visibility or evidencing through showing, recording, and circulating images of and footage of violence. She notes that the presupposition underlying attempts to provide an alternative vision of Israeli violence is that ‘if only the world had the opportunity to see, then it could react’. But the problem, according to Hochberg, is ‘not that the world must be given an opportunity to see, but rather that the world spends too much time seeing and that seeing secures no political intervention.’ Hochberg continues: ‘the problem we are facing, then, consists less of making an otherwise invisible catastrophic reality visible and more so of challenging the dominant modes of representations through which the very visibility of others’ suffering remains nothing but a spectacle, providing at best a momentary source of ethical speculation and at worst a source of voyeuristic pleasure.’

The solution to all this is not so much to speak truth to power but to enact a politics of opacity, one that involves not inaction, passivity, and invisibility, but one that entails ‘turning our backs on power’. That is rejecting the dominant structures of representation, negation of visibility, and refusal to cooperate. For some, this may mean politics of public occupation, protest, and appearance as willful subjects; for others, it may be in the form of conscientious objection, refusal to participate, and boycotting.