A Brief History of Torture in Israel

The working of Mavtan we review in this paper is to be situated within a longer history of the use of torture in Israel (or more accurately, a history of the relations between GSS’s torture practices and the Israeli judiciary) for which we cannot do justice here. The first chapter in this history would begin in the early state years and end roughly in 1987. During this period, the executive and judiciary branches preferred to turn a blind eye to how the GSS conducted itself, allowing the GSS to effectively operate with full impunity (Al Haq Reference Al Haq1990; Cohen and Golan Reference Cohen and Golan1992; Langer and Bishara Reference Langer and Bishara1984).Footnote ⁵

In 1987, following a series of high-profile scandals which exposed the GSS’s unruliness, the Landau Commission was established and marked the beginning of a new era. Formally, this was the era in which Israel introduced judicial oversight over the GSS interrogation methods, also in light of the coming into force of the UN Convention Against Torture in that same year (marking the potential problematics torture would pose for Israel in the international arena henceforth). Effectively, however, this era marked the introduction of torture into the Israeli legal system; torture became an ordered and sanctioned form of state violence (Hajjar Reference Hajjar2004). Indeed, the Landau Commission justified the resort to what it euphemistically called “moderate measures of physical pressure” during interrogations as long as these measures were not too “excessive”, and most famously introduced a classified list of authorized torture methods (Landau Commission 1987, 72).Footnote ⁶ In so doing, it constituted a seminal change in the relationship between torture and the law that would later be adopted by the United States (Hajjar Reference Hajjar2006; Khalili Reference Khalili2013; Yoo Reference Yoo2012). For the sake of our argument here, it should be noted that this commission was mainly troubled by the fact that the GSS evaded judicial oversight, that whenever allegations of torture were raised in court, the interrogators involved would simply lie under oath,Footnote ⁷ and that these jeopardized “the image of the State as a law-abiding polity” (Landau Commission 1987, 49, emphasis added). The Commission therefore took upon itself the task of mending the relationship between the GSS and the Israeli judicial system, including, among other measures, recommending the establishment of Mavtan (Kremnitzer Reference Kremnitzer1989).

The next significant milestone in that history was the 1999 landmark ruling of the Israeli High Court of Justice (HCJ) in Public Committee Against Torture in Israel vs. State of Israel. The ruling, which was interpreted at the time as “a bold prohibition” of the use of TIDT, introduced, in effect, “a torture policy” (Mann and Shatz Reference Mann and Shatz2010, 60-62). What has been “truly new” following this decision, Mann and Shatz argue, “is the administrative structure in which torture is managed” (63). While it outlawed torture and made way for holding interrogators criminally responsible for using TIDT, the ruling also introduced the “necessity defense.” Adopted from the French-Algerian context (Macmaster Reference Macmaster2004), the “necessity defense” could shield interrogators from criminal liability, post-factum, if during an interrogation they believe that they had to use illegal interrogation methods to thwart an imminent threat. Shafir (Reference Shafir, Brysk and Shafir2007) thus claims that rather than forbidding torture altogether, this was a doctrine of restraint. The assumption was that if interrogators knew that they might be indited for using TIDT after the fact, they would be deterred from resorting to such methods in all but the very rare exceptions. And yet, as Smadar Ben Natan (2019) demonstrates, the ruling allowed the creation of a mechanism—that was soon thereafter formalized and widened by the Attorney General—for a-priori authorization, enabling a de-facto procedural licensing of torture (see also Mann and Shatz Reference Mann and Shatz2010). The doctrine of restraint nevertheless remained the system’s implicit rationale. Mavtan, which is tasked with determining whether a criminal investigation should be conducted against GSS agents who allegedly used illicit interrogation methods, is hence a key node in this doctrine. Our intervention belongs here, in what can be seen as the latter days of the second chapter of the longer history of torture in Israel.

At the time of writing this article, in the winter of 2023–2024, there are strong indications that following Hamas’s attack on October 7, torture by the GSS has entered a new era. In response to what was the deadliest attack on Israeli civilians in the history of the conflict, Israel significantly relaxed, if not completely abandoned, its rules of engagement in the war it conducts on Gaza and, differently so, against Palestinians more generally. In many respects, its adherence to International Law has all but disappeared in both practice and rhetoric (Abraham Reference Abraham2023). At the same time, Palestinians in Israeli prisons and in a newly established IDF detention camp for detainees from Gaza, have been subjected to torture, with little to no restraint. Testimonies tell of unprecedented levels of wardens’ and soldiers’ violence, including sexual violence (Abraham Reference Abraham2024). These were met by a clear unwillingness of the Israeli authorities to intervene, let alone investigate and hold responsible those involved (PCATI 2024). While testimonies from people interrogated by the GSS during that period have yet to emerge, it would come as little surprise if such interrogations have also intensified, and that after October 7, new levels of violence have been practiced also by the GSS.

Nevertheless, as the Israeli defense in the ICJ clearly shows, the rationalities which brought Mavtan into being, and which are at the heart of our analysis here, are still at play even in this post-October 7 era.

Democracy, Legitimacy, Violence: The Argument

In her analysis of torture in India and the United States, Jinee Lokaneeta (Reference Lokaneeta2011, 5) shows that torture makes “a particularly paradoxical proposition” for liberal states. This is because “the standard narratives on the history of torture have created a discourse of impermissibility of torture in modern liberal democracies” (8; see also Luban Reference Luban2005). Thus, to sustain its image as a liberal democracy, Israel needs to contend with its systematic deployment of torture. This, we argue, is what Mavtan does. Mavtan (as an emblem for many other judicio-bureucratic systems) bestows democratic semblance on Israel as it continues to torture Palestinians. Our argument, then, joins a large body of work that has shown how various mechanisms and discourses designed to restrain state violence have not merely failed to protect disenfranchised individuals and groups, but, further, often serve to enable the propagation of violence (Gordon and Perugini Reference Gordon and Perugini2020; Kolsky Reference Kolsky2019; Sanders Reference Sanders2018). Within such mechanisms, we focus on the routine practices of judicial bureaucracy, thus contributing also to the analysis of violence and bureaucracy, particularly in (post-)colonial settings (Berda Reference Berda2017 Reference Berda2022; Gupta Reference Gupta2012; Mathur Reference Mathur2016).

In the case at hand, the mechanics of legal bureaucracy produce what we call, as a shorthand and as a paraphrase of Timothy Mitchell (Reference Mitchell and Steinmetz1999), “the legitimacy effect”. “Legitimacy” operates here both internationally and internally (as an investment in a certain self-image) and corresponds to the notion of a liberal democracy. In this context it is enacted through a cluster of principles, including the rule of law, legal accountability, and a specific concept of good governance, to legitimize a state of affairs that is essentially non-democratic. If for Mitchell “the State arises from techniques that enable mundane material practices to take on the appearance of an abstract, nonmaterial form,” (77; see also Gupta Reference Gupta1995, 378) here we argue such practices can sometimes make way not just for the state as such to emerge, but further allow the state to assume a particular form or identity. Drawing on Butler’s (Reference Butler1993) notion of performativity, Cynthia Weber (Reference Weber1998) has already argued that much like individuals, states should be seen as subjects-in-process whose identities are shaped through reiterated material practices (see also Edensor Reference Edensor2020; Visoka Reference Visoka2018).

At play here is what Elif Babül (Reference Babül2017, 14) identifies in her research on Turkey as an “identity crisis,” in which “the desire to belong” (to the EU, and to a certain notion of Europe or the family of liberal democracies more generally) drives introducing principles of good governance and a language committed to human rights, while other national interests and dynamics undermine the actualization of these principles. Eventually, bureaucratic attentiveness to the protection of individual rights and liberties (demonstrated through increasing “markers of accountability and openness to scrutiny,” “improving bureaucratic performance,” a language of professionalism) works in tandem with (and even facilitates) the violation of those very principles (Babül Reference Babül2017, 63, 9).Footnote ⁸

Much like Turkey, Israel has both internal and external motivations to identify itself as a democracy (emerging both from a similar “desire to belong” to the category of liberal democracies and from its susceptibility to international pressure), yet its ongoing control over the occupied Palestinian territory and over the occupied Palestinian population means it cannot align itself with the dictations of international laws and norms in full. And similarly to Babül’s findings, attempts to bridge this unbridgeable gap often take place through “democratic performativity” that emphasizes what Vivien Schmidt (Reference Schmidt2013) termed “throughputs”: processes pertaining to procedural competence which allow states to gain their legitimacy.

What we show in our analysis of Mavtan should accordingly be seen as a dual act. First, much like Mitchell, Bigo (Reference Bigo2012) and others, we claim that abstract and normative concepts must be understood as emerging from the tangible, from the operations of concrete institutions, agencies and organizations, their structures, procedures and regulations, and from the workings of bureaucracy. Interlinking the abstract and the empirical, we can see how specific conceptual clusters emerge through the work of Mavtan and other judicial bureaus. Second, this cluster amounts to an act of substitution in which procedural competence come to stand for that which they presumably deliver, such as, most evidently in this case, the rule of law. That is, by going through the motions of proper governance, professionalism, efficacy, accountability, or transparency—which, in turn, underwrite a more abstract language: “the rule of law,” “democracy,” and potentially even “justice”—Mavtan puts into operation an apparatus of legitimacy, whilst allowing torture to persist. In this process, what we identify below as “lapses” (Grinberg Reference Grinberg2018; Stein Reference Stein2021), performative failures (Butler Reference Butler1990, Kotef and Amir Reference Kotef and Amir2007) or “glitches” (Lisle Reference Lisle2021), have a structural role: they mitigate, if not fuse, a notion of democracy (constricted to its procedural meaning) with Israel’s nondemocratic rule.

The effort to bridge this gap—between a language of democracy and a prolonged military occupation—appears in many of the discussions pertaining to Mavtan. While it is never articulated as such, and is always coded through talking about Israel’s “security challenges” or “unique circumstances”, these “circumstances” are, for the most part, Israel’s colonial control over the Palestinians. For instance, when asked to rule on the legality of Mavtan following an appeal by PCATI and others who argued Mavtan’s very existence obstructs justice, Supreme Court Justice Elyakim Rubinstein quoted himself declaring that “Israel is an abnormal normal country.” He continues by explaining:

Responding to the June 2024 ICJ ruling in the SA vs Israel case, Dina Zilber (Reference Zilber2024), a former Deputy to the Attorney General, explicitly articulates what is at stake, and why it is so crucial to insist on presenting Israel as a democratic state, “abnormal” as it may be. According to her, the hearing has shown that “the professionalism, independence and autonomy of the public legal counsel and of the judiciary system,” which attest to Israel’s “democratic character,” protect Israel from extensive international intervention. As such, she argues, this system is “part of the national security of the state,” (emphasis added) and is even “necessary to secure its survival”. Note how Israel’s “democratic character,” which amounts to the “legal standards” Rubinstein required, appears as protecting the state rather than protecting human rights. By producing what we termed above “the legitimacy effect,” it works to shield Israel from interventions, interventions which are perceived as a threat to its security.Footnote ⁹

The role of this democracy, or the image thereof, became most evident as the sixth Netanyahu government attempted to curb the independence of the Israeli judicial system as of the winter of 2023. Many of the Israelis who objected to the government’s plans explicitly vocalized what has long been a latent perception: the independence of the judiciary—which was explicitly identified as the core of Israel’s democracy—serves as a “body armour” that protects the state’s security interests (Kotef and Amir Reference Kotef and Amir2023). The former Attorney General, Avichai Mendelblit, for instance, warned that “the moment that the justice system in Israel isn’t perceived as [independent], Israel will lose international legitimacy for its military operations and will no longer be shielded from accusations of war crimes.” (again, note the importance of “perceptions”) (Weitz Reference Weitz2023; see also AGI 2023). Many in the streets and on countless written interventions vocalized a similar understanding: that democracy is equitable to the independence of the judiciary system, and that this independence is key to cordoning off persecutions in international courts or other courts internationally.

Our analysis of the link between “democracy” and torture is thus tangential to Darius Rejali’s (Reference Rejali2007) work on the particularities of torture techniques in democratic contexts. Both projects show how a particular adherence to some form of democratic principles or language allows torture to persist, even if it requires some adjustments. As Rejali shows, in the case of Israel, as in many other contexts, adhering to the principles of democracy had not ended torture, but had rather led to the modification of torture techniques (Ben-Natan Reference Ben-Natan2019). We, however, focus on how the problem that torture poses for the perception of Israel as a democracy (in and of itself, but also as an example of colonial violence more generally) is allegedly resolved by introducing bureaucratic solutions. It should go without saying that such solutions are conditioned on, and reproduce, the complete side-lining of the question of torture itself and the rights of its victims. In this sense too, we look here at torture not only as one element within the problematics at the core of Israel’s concept of democracy, but also as an emblem of this problematics as a whole: it is only by discounting its human rights violations, by completely disregarding the rights of Palestinians, and only by utterly side-lining the political question of Palestine, that Israel can call itself a democracy at all.

Mavtan: Structure and History

Individuals who wish to take legal action for being subjected to torture or ill-treatment by the GSS are required to first file a complaint with Mavtan. Mavtan then conducts a preliminary examination of their complaint and submits its findings and recommendations to the Mavtan Comptroller.Footnote ¹⁰ The Mavtan Comptroller, who is a deputy of Israel’s Attorney General, decides whether or not to refer the case to the police for a criminal investigation. The first conclusion can be stated already here: Mavtan is predicated on a contradiction. It is simultaneously the conduit for providing legal remedy for victims of torture, and an obstacle to their ability to see justice. The requirement that complaints must first go through Mavtan, and that victims do not have direct access to criminal proceedings, is what constitutes it as an obstacle (HCJ 1265/11 2012).

The very existence of Mavtan can be seen as an example of what Nasser Hussain (Reference Hussain2007) referred to as “hyperlegality” which, he argued, characterizes the aftermath of British colonial rule—a mode of governance that creates complex, fragmented bureaucratic units and authorities. The fragmentation of the bureaucratic system then “fractures the process of action” into semi-judicial bodies, that accomplish the role of doing very little while reassuring the public that a responsible course of action has taken place (523; see also Raman Reference Raman, Balachandran, Pant and Raman2017). Yael Berda (Reference Berda2017) similarly shows that far from the Weberian principle of rationality, (post)colonial bureaucracies are convoluted, self-contradictory, and often nonsensical systems, which ultimately perpetuate the racially differential logic of colonial rule, forming a system the colonized can never successfully navigate.

Yet Mavtan does more than mounting additional hurdles and delays or fragmenting the complaint process. The preliminary examination necessarily impairs the integrity of the future investigations by delaying them, tampering with evidence and allowing the accused to better prepare, if not coordinate their testimonies. The institutional requirement for conducting a preliminary examination therefore means that any criminal investigation would already be compromised well before it is launched (if ever). It should therefore come as little surprise that Mavtan has proven to be an insurmountable institutional obstacle for torture victims.

Mavtan was established in 1992 as an internal unit within the GSS. As to be expected, and as highlighted repeatedly in both internal and external reviews, during this period Mavtan’s ability to effectively examine GSS’s conduct was curtailed by its positioning within GSS, and the unit was largely ineffectual (HCJ 5722/12 2017; Turkel 2010; UNCAT 2009). Moreover, the employee assigned to run Mavtan did not have the competence to fulfil this role. A 2007 MoJ review found that he “is very limited in his skills as an investigator” and “does not know how to confront GSS interrogators with diverse findings and conflicting testimonies” (quoted in Turkel 2010, 415). Mudzgurishvilly (who would later head Mavtan), insisted that the unit was completely useless at that time, if it could be said that it existed at all.Footnote ¹¹ Indeed, despite reviewing close to 800 complaints between 2001 and the end of 2013 Mavtan failed to find even a single case warranting a criminal investigation (PCATI and FIDH 2022).

This impotence, however, served a purpose: it resolved the inherent tension between the requirement to fully investigate torture and the need to prevent the exposure of GSS’s interrogation methods to judicial review. Mavtan allowed Israel to exhibit that its judiciary has the capacity to scrutinize the security agency, while shielding the bench from contending with GSS interrogators who were evidently violating not only international law, but also Israel’s own legislation.

While Mavtan’s ineptitude allowed Israel to cover up its use of torture for many years, this solution did not withstand increasing external pressures. With Israel facing growing criticism for its (mis-)handling of allegations of torture,Footnote ¹² Mavtan’s evident inactivity and, more critically, its lack of independence, became more difficult to justify, and in 2014 the unit was transferred to the MoJ.

Bureaucratic Industriousness

The transfer of Mavtan to the MoJ was accompanied by several operational changes. Most visibly, the unit has expanded significantly: between 2014 and 2022, it grew from a single person to a staff of ten, including four investigators.Footnote ¹³ Its volume of activity similarly increased: slowly but surely, it started to conduct lengthier examinations and to issue more detailed reports. It also digitized its data collection and filing systems, and began to operate with a greater level of transparency: information is now shared more freely with human rights organizations, and Mavtan’s examination reports (albeit redacted) are made available to plaintiffs and their lawyers. Yet the final outcome changed very little: the absolute majority of all complaints are still dismissed. Out of close to 600 complaints filed from 2014 to 2023, Mavtan Comptroller recommended launching a criminal investigation in three cases alone, and these, too, were eventually closed.

In its present incarnation Mavtan hence became a highly industrious unit; an industrious unit whose work still amounts to nothing. Mavtan’s industriousness manifests through elaborate examinations that come up empty handed; modes of narration that evade blunt deception but do not reveal the truth; ways for producing accountability that lead away from disclosure; and institutional mechanisms for sustaining the object of knowledge in an indeterminate realm—not fully concealed, yet never out in the open. Despite ample evidence that illegal interrogation measures are prevalent, Mavtan’s examinations repeatedly fail to lead to criminal investigations in all but the extremely rare occasions. Yet even though it does not produce the outcomes it is formally tasked with delivering, Mavtan’s work does have a significant effect: It is precisely through Mavtan’s heightened activity, its diligence, its productiveness, that Israel can demonstrate that allegations are examined, thus deflecting accusations of collusion. The unit’s operational intensity thus lends itself to the claim that Israel’s juridical system offers justice to GSS interrogees, despite all evidence to the contrary.

In our interview, Asher, the current head of Mavtan, kept referring to himself as “a man of numbers” and returning to “the numbers” which—he insisted—“speak for themselves.” The sheer weight of numbers, he seemed to imply, proved that Mavtan was performing its work efficiently. While he predominantly referred to the volume of processed complaints, he also referred to other numbers: of investigators employed, the number of days it took to launch an examination, months spent processing each complaint. But when we tried to ask about one number, the number of criminal investigations that had been opened thus far, and whether this number—three—was too low, he got agitated and reprimanded us for insisting on asking the wrong questions (Asher 9/22). Mudzgurishvilly, the first head of Mavtan following its move to the MoJ, similarly stated that judging Mavtan through this figure is simply wrong. When it comes to legal cases, she said, statistics are irrelevant, and one has to look at each case individually (Mudzgurishvilly 10/22; see also Grinberg Reference Grinberg2016). This number, arguably the most important number since it could have proven that Mavtan secures access to justice for torture victims, is the one number they both discounted. Those other numbers, that “speak for themselves,” that testify for nothing but Mavtan’s vigorousness, have thus come to stand for the meaningful indicators of Mavtan’s effectiveness. Devoid of content, these numbers are what Mavtan is there to produce: a mode of enumeration that is used, according to Ken MacLean, as a practice establishing credibility and authority by creating “evidentiary weight through the repetition of numbered sources” (Reference MacLean2022, 198). Indeed numbers, Diane Nelson noted, can do “magical things”, such as creating something “out of, quite literally, nothing” (Reference Nelson2015, 27).

Mavtan’s effectiveness, we should note, is not simply produced through vigorousness. As Asher stressed, it is also demonstrated by Mavtan’s compliance with the principles of good governance and a well-operating bureaucracy, which, he said, are key to any democratic society, such as accountability, transparency, and maintaining the appropriate relations between state institutions. In the next two sections we look more closely at how accountability and transparency are manufactured.

Accountability: Meticulous Examinations

As noted earlier, Mavtan’s transfer to the MoJ in 2014 revamped its operation—and above all, its examinations. Since then, the examination team has been expanded to include more staff; more examinations are opened each year; the examinations are more extensive; more of the involved parties are interviewed in each examination; these interviews are more detailed; and the protocols of these examinations are much more elaborate.

To better understand how this dynamism fails to translate into meaningful results, let us look briefly at one example. In a recent complaint, a Palestinian woman stated that on the thirty-fifth day of her interrogation, exhausted from incessant sleep deprivation and in a deteriorated physical state, she was taken to a room which the interrogators called the “VIP room” and the wardens called “Room 220”. The room was pitch dark. One may begin to imagine what it means to be placed in a dark room amid a violent interrogation, particularly when a woman is interrogated by men; not being able to tell who was in the room or where they were; being shouted at by people one cannot see, not knowing how close the interrogators are; suddenly feeling someone up close; constantly being scared of being attacked. Being in that room was so intimidating that the female soldier who was there to accompany the interrogationFootnote ¹⁴ felt too distressed herself, and very quickly left the room.

The decision letter notifying that her complaint had been dismissed revealed that as part of the preliminary examination, an investigator was sent to find whether the interrogation facility indeed had such a room. After verifying that Room 220 exists, the investigator reported that “The examination I conducted revealed that the room has two lighting options: dimmed lights and full lights.” “When using the dimmed light option,” the report continued, “I was still able to see all objects in the room. In fact, from each position within the room I was able to see the entire room … without any difficulties.”Footnote ¹⁵ He then called his supervisor on a video call, placed the phone in several areas of the room, and made sure the supervisor could clearly see him. These “findings” then served to disqualify the complaint.

The absurdity of this “finding” is obvious. Light switches can be used to switch the light off. One may further question why a “dim light” option is installed in an interrogation room. But what is significant for us here is the lengths to which Mavtan went in order to repudiate the complaint: an investigator travelled to a remote interrogation facility (roughly a 90-minute drive in each direction); searched for the room; checked the different options of the light switches; examined the room from different angles; repeated this examination on a video call with his supervisor, and then composed a report detailing these findings. It is through this kind of exuberant activity that Mavtan can argue that its examinations are thorough and demonstrate the seriousness with which it treats complaints. Yet such an examination can only be understood as a parody; a mimicry of an “exhaustive” enquiry, a show of industriousness tactically deployed to fabricate “accountability.” And much like Butler’s (Reference Butler1990) understanding of parody, through this mimicking of an “examination,” or perhaps a “rigorous examination” (alongside other procedures, some of which we reviewed earlier and some we identify later), clusters of bureaucratic principles emerge: “oversight”, “due process” and, by extension, also “the rule of law” or even “justice-making”, while rendering them all void of substance—a façade. Yet since this parody of an examination supplants an earnest one, it undermines the very possibility of the latter, and hence has a very material effect: it shields torture from justice-making.

The issue of the examination’s content (or lack thereof) becomes particularly pertinent in Mavtan’s questioning of GSS agents. Mavtan takes great care in these questionings, making sure that all the interrogators involved are summoned and that every single detail of the complaint is addressed. This seriousness, however, only serves to obscure the fact that nothing of any significance ever emerges in these conversations.

In many respects, Mavtan’s questioning of GSS interrogators adheres to a well-rehearsed structure. Drawing on Foucault (Reference Foucault and Hurley1978, 18), it can be seen as an “institutional incitement to discourse”, in which the subject is probed and solicited to provide the minutest of details. But it is not the truth that GSS agents are prompted to disclose; neither—technically speaking—are they asked to lie. Imitating the confessional act, what we have here is the art of profession that says nothing; the institutionalization, in a way, of the mechanism of a meaningless disclosure when silence is no longer an option.

Still, these questionings provide an abundance of irrelevant information. Asked about a particular interrogation, interrogators can recall the finest details. One remembered bringing a particular type of chocolate to an interrogee or recalled that the latter asked for a newspaper in Arabic, which the interrogator was able to find at a nearby corner shop. Another remembered that his fellow interrogator brought to an interrogee fresh clothes and falafel.Footnote ¹⁶ Interrogators seemed to have accurate recollection of the exact chair on which the interrogee sat (of a normal height, in a comfortable position), or that the interrogee was always allowed to go to the toilet whenever they needed. They could recall the number of times the interrogee had asked to pray and that they were always allowed to do so, or how many hours of sleep they got (again, always enough hours). This level of detail serves both GSS and Mavtan. It allows the GSS interrogators to demonstrate that they are fully cooperative while substantiating their account as both accurate and full. Concurrently, by extracting a wealth of information (albeit completely irrelevant), Mavtan’s investigators can show that an exhaustive examination had taken place and that due process had been followed. No less importantly, this wealth of detail is geared to divert attention from the fact of torture. Both Mavtan and GSS thus seem to share the same desire: that Mavtan incites the GSS interrogator to talk and yet say very little, together performatively producing a mirage of accountability.

Importantly, while some (irrelevant) information is provided in detail, questions which pertain to criminal conduct are evaded. When asked about the use of physical or verbal violence, or about other acts which could implicate them in torture or illicit interrogation methods, interrogators respond laconically and often claim that they remember nothing.

The intensive work that goes into the lengthy questioning, and then to the composition of lengthy reports (demonstrating the lengthy questioning), is accompanied by a third layer of bureaucratic labor: censorship. Significant sections of the reports are redacted by the GSS before complainants’ lawyers get to see them. Importantly, the laborious censorship is no less integral to the performativity of bureaucratic industriousness. The long black lines serve to substantiate Mavtan’s competence, testifying to its diligence, implying that investigators did not shy away from asking difficult questions and were exposed to highly classified information (sensitive enough to merit censorship), while conferring an air of esteem on Mavtan for being privy to state secrets (Bigo Reference Bigo2006; Weber Reference Weber1958). The censored information, we must presume, still failed to reach a criminal threshold (otherwise—theoretically—it would have warranted a criminal investigation). Moreover, the practice of redaction works to counteract accusations that the GSS covers-up illicit acts since all information, including highly sensitive details, was ostensibly shared with Mavtan. This semiotics of classified revelation is, in a sense, saying: trust us, we have left no stone unturned, and still, we did not find a shred of evidence that torture took place. By the time it reaches the complainants’ lawyers, Mavtan’s examination reports often entail formulations such as:

Note the somewhat dazzling dance around the fact of torture in this formula. It starts with a lapse of memory (“I do not remember”); then a turn to the hypothetical form (“had it happened, I would have…”) that turns a fact into a hypothesis; then an admission that something indeed took place (“when…”), which is almost fully redacted; concluding with a return to the hypothetical, to undermine the status of the previous statements.

The legalistic conditionality which appears in the responses of GSS interrogators may come as no surprise and adheres to the structure of incitement into meaningless discourse with which we started this discussion. However, these evasions are never contested by Mavtan’s investigators. The inexplicable and all too convenient memory gaps, the evident implausibility in which violence is explained away, the obvious inconsistencies in interrogators’ recollection, and any discrepancies, no matter how glaring, between versions of the events including in a single testimonial: none are ever challenged.

Indeed, as explained to us by A.R., a former GSS interrogator, even though having to answer to Mavtan may be unpleasant for GSS agents, these exchanges should not really be seen as “questionings” but as “interviews”. Going in, he further stated, it is clear to the Mavtan investigator who is trustworthy and who is not: “the Mavtan investigator knows that they need to take the Palestinian’s accusations with a grain of salt.” And whereas Mavtan has indeed become more thorough in its examinations, “when I enter the room, I start with a score of one hundred and the Palestinian has a score of zero.”Footnote ¹⁸ This score may shift a little, but A.R. insisted that the likelihood of Mavtan investigators completely changing their mind is close to nil. The dynamic of the examination thus serves to justify the presupposition that GSS interrogators are in the right and that the Palestinian victims are in the wrong. And yet, even though—or, more accurately, precisely because—the outcome is preordained from the onset, the act of simulation is crucial.

In the setting of an interrogation, Elaine Scarry (Reference Scarry1987) argues, the dynamic between causing pain and generating information is essential. Whereas pain is not necessarily inflicted to reveal the truth (Barela et al. Reference Barela, Fallon, Gaggioli and Ohlin2020), the interrogation serves to both structure and justify torture. The information the tortured discloses is less important in this context, Scarry suggests; what is crucial is that an answer is provided. Extracting an answer from the tortured converts pain into power: the mere fact of an answer not only reflects the torturers’ ascendency, it is also the very form of a political betrayal, a disavowing of resistance. It is hence the victory of the torturer and the breaking point of the tortured (Scarry Reference Scarry1987, 28-9). What we see in Mavtan’s questionings is an inversion of this dynamic: at stake is the mere existence of an examination, the questioning and the answering, and not the content of the answer. Yet instead of a forced betrayal, these reciprocal dynamics are structured to reaffirm existing political relations. This laboured incitement into (meaningless) speech harbors the coverup of torture in an intermediate discursive realm, which is neither a blatant lie nor the truth, thus protecting the integrity of the Israeli legal-security system, as an apparatus in which torture and legality coexist.

Transparency: The Non-Recording Cameras

In 2011, a petition challenging the existence of Mavtan was filed on behalf of ten Palestinian plaintiffs by six human rights organizations. The HCJ ruling, delivered by Justice Elyakim Rubinstein, dismissed the petition, expressing full confidence in Mavtan despite all evidence to the contrary. Rubinstein’s professed confidence in Mavtan notwithstanding, his ruling deemed the moving of Mavtan to the MoJ essential. Mavtan’s “new institutional position,” he reasoned, would allow “such monitoring [of GSS to] be as transparent as possible,” (HCJ 5722/12 2017, para.21) reflecting Israel’s security agencies’ exposure to “the sunlight and judicial review in a new era of transparency” (para.19). Transparency, from then on, becomes the leitmotif of the ruling, echoing the prevailing supposition that increased visibility serves to eradicate political violence (cf. Linfield Reference Linfield2012, 151-62).Footnote ¹⁹ This emphasis on transparency was carried into the ethos of the reformed Mavtan: when we met with Asher (the head of Mavtan), he took great pride in the increased transparency introduced under his management, and repeatedly equated it with the very essence of democracy and justice-making.

The story of Mavtan and the changes it has undergone is accordingly also a story of this enhanced transparency. And much like in the story concerning accountability, transparency was from the onset very much a question of “appearances.” Indeed, as long as Israel appears to seriously treat allegations of illicit interrogation techniques, it can fend off international pressure. In a way at least, the story of the ICC and ICJ decisions and preliminary decisions is a story of these systems’ inability or lack of interest to continue even playing this performative role. Rubinstein accordingly argued, somewhat frankly, that Mavtan’s relocation to the MoJ was necessary not only for good governance, but also “for the sake of appearances” (HCJ 5722/12 2017, para.21). The cameras installed in GSS interrogation rooms best illustrate the unique form this transparency has taken. We turn to them now.

GSS’s refusal to allow recordings of its interrogations has been a consistent issue of contention. Israel’s State Comptroller determined that GSS interrogations should be recorded already in 1995 (State Comptroller 2000), and the UN Committee Against Torture (UNCAT) has raised the lack of such recordings as a grave concern in all its reports since 2009 (UNCAT 2009). In 2010, this necessity was reiterated by the Turkel Commission which was tasked with determining whether Israel’s investigation mechanism complied with international law (Turkel 2010). The implementation of the Turkel Report was then relegated to the Ciechanover Team. The latter concurred with the Turkel Commission’s findings, yet raised several objections to the full recording of interrogation. Echoing what has long been GSS’s justification for protecting its interrogations from external scrutiny,Footnote ²⁰ the Turkel Report claimed that exposing GSS interrogation techniques might impede their efficacy, deter suspects from cooperating with the authorities, and “substantially [impair] the ability to frustrate terrorist threats” (Ciechanover Report Reference Report2015, para.135). It hence devised an “alternative” solution: installing cameras in all interrogation rooms—but insisting that what they capture would not be recorded. Instead, the cameras would broadcast in real-time to a separate location which would “be accessible and available to a supervising entity on behalf of the Ministry of Justice at any time without giving prior notice”, thus ensuring that “the interrogators will have no indication of when the supervising entity is watching them.” The supervising entity will be tasked with composing “a concise memorandum on what he [sic] saw”; and if he “believes that illegal means have been used during the interrogation, an immediate obligation to report the matter to the Mavtan will arise” (para.139).

Installed in 2018, the CCTV cameras have subjected GSS interrogations to an unprecedented level of transparency. And yet, much like the lengthy examinations which fail to deliver any meaningful information—and part of the same structure—what we have here is a mechanism for producing transparency that does not allow for torture to be seen. Since what the cameras capture is never recorded and stored, and is only transmitted in real-time to the control room, this system replaces solid evidence (the audio-visual recordings) with selective and subjective reports. Moreover, these CCTV cameras also constitute, quite literally, an institutionally closed-circuit system which is hermetically sealed: the inspectors, selected and assigned by Mavtan, report only to Mavtan, and only on what Mavtan instructs them to report. And all of this is assuming that anyone is watching at all: according to Israel’s own admission, only a small fraction of the interrogation hours is monitored.Footnote ²¹

Analyzing the efforts of human rights organizations to visually document human rights violations, Rebecca Stein (Reference Stein2021) identifies the material conditions leading to partial, unclear, and sometimes unusable visual products. Whereas Stein talks about different circumstances altogether (dropping one’s phone camera while running away from soldiers or settlers, for example), similar “glitches and lapses” (16) occur here: blind zones in interrogation rooms, or, more evidently, a video stream broadcasting to an empty monitoring room. In this respect, if Stein shows how visions of transparency and accountability in cases of human rights violations were met with “broken camera hopes and dreams” (4), in our case, the system’s breakdown is, in some sense at least, embedded into its design. Our analysis thus draws closer to Omri Grinberg’s (Reference Grinberg2018, 267) suggestion that we should not assume that such lapses “are (micro) failures.” Rather, “we can also view them as an instigated, or at least uninterrupted” lacunas, that are “constitutive elements in colonial domination” (see also Kotef and Amir Reference Kotef and Amir2007). At any rate, as Stein shows, and as Hedi Viterbo (Reference Hedi2014) warns us following the cases of Abu Ghraib and the Syrian civil war, even when cameras provide clear visual evidence, it often makes little difference, retaining, in the words of Chatterjee (Reference Chatterjee2023, 8), “violence visible yet unaccountable”.

Perhaps trying to prove the cameras’ efficacy, and contra their official purpose of capturing violations as they occur, Asher emphasized the power of the cameras to prevent future violations, regardless of the actual number of monitoring hours. In a proto-Panoptical fashion (Foucault Reference Foucault and Sheridan1979), he argued that the knowledge that someone may be watching deters GSS agents from resorting to illegal interrogation methods. However, in our interview, one interrogator disclosed that it is not only that GSS interrogators know that the inspectors are rarely there; crucially, and despite Ciechanover’s recommendations, cameras are not installed in all interrogation rooms. Interrogators hence know how to evade this presumed panoptical effect.

Moreover, the very activity of watching is veiled under additional layers of opacity since Mavtan releases very little information regarding the inspectors and what they do. Even more so, the little that it provides is diluted to the extent that nothing of significance can be concluded from it. For instance, when asked for the number of inspectors’ reports submitted in a given year (that is, how many “abnormal events” the inspectors witnessed), Mavtan offers a number, but immediately adds a disclaimer: “abnormal events” that merit a report, Mavtan’s standard reply states, are based on “a wide definition. Reports may pertain to a wide range of instances, including instances wherein very minimal deviations from interrogation procedures are suspected of having taken place.” Mavtan responses further stressed that “there may be cases wherein after an examination it is found that there was no inappropriate conduct or any deviation from interrogation procedures.” Thus, whereas Mavtan provides a number indicating what seems to be an objective measure of the fact some violations have taken place, the meaning of the number and what may be concluded from it is immediately undermined. Curiously, as if to further diminish any conclusion that could be drawn from the figures provided, and in a stark deviation from the MoJ policy concerning what may count as an “abnormal event”, Mavtan also states that in some instances, reports indicate not a violation by interrogators, but rather of an “unusual behavior of an interrogatee towards his interrogators” (SoI 2019). In our conversation, Asher gave another example of an “abnormal event”: Palestinians who hurt themselves in order to complain against the GSS. We will return to this example in the last section, but for now, we can note that this disclaimer pulls the grounds from under what would have been an indicative number: the number of times the inspectors observed the use of unlawful methods. Thus, the transparency Asher highlighted as an indicator of adherence to democratic principles, as evidencing Mavtan’s accountability to the public, reveals itself as a mimicry of compliance. Accordingly, the cameras, the very devices of transparency, do little to expose the interrogations, and themselves remain completely opaque. In this sense, the CCTV system is both an example and an emblem of a much wider tendency: a constant increase in transparency that in fact reveals very little, if anything at all.

Partly, what is crucial here is that this system flounders on its own terms, a point that becomes clear if one considers the logic guiding these terms. As we show later, the rationale for the necessity of Mavtan is predicated on the presupposition that violations by GSS agents rarely occur and that when they do occur, they are the aberration. As described on the MoJ website: “The purpose of the Department is to examine exceptional incidents or complaints of misconduct by members of the Israel Security Agency [GSS]” (MoJ 2024, emphasis added). By this logic, the CCTV monitoring is not fit for purpose: Since the inspectors only sample a small percentage of the interrogations, they are statistically unlikely to capture such (presumably) rare transgressions. But in its failure to provide actual transparency, this mechanism nevertheless produces at least three important effects. First, despite it being inadequate to expose the interrogations to external scrutiny, the CCTV system is the very institutionalization of transparency: cameras are installed, inspectors are hired, training is conducted, monitoring hours are logged, reports are issued and reviewed. All this amounts to an apparatus of transparency: an idea that is so often linked to the ethos of due process and even justice-making, that is aligned with their form, or, as we saw in Rubinstein’s quote above, their image. The mere existence of this system can then serve to demonstrate that Israel properly monitors its own security forces, that international bodies should therefore not intervene, that the entire legal apparatus is independent and well-functioning, and the sovereignty attached to it should therefore be respected. “Transparency”, then, shifts from being a means to secure good governance, to what “good governance” means; accordingly, it no longer matters what such transparency reveals, or whether it reveals anything at all.

Second, the failure to expose transgressions is itself deployed to undermine the possibility of delivering justice: the failure to see serves to prove that there is nothing to see. Indeed, since 2018, Mavtan’s justifications for rejecting complaints increasingly refer to inspectors’ reports that allegedly do not support the complainant’s version of events. A complaint filed by a Palestinian man who was arrested in 2019 for suspected involvement in terror attacks illustrates this point. His interrogation by the GSS was conducted over more than 600 hours in total, with many sessions lasting up to 48 consecutive hours. In his complaint, he described that he was shackled to a chair the entire time and that the inability to move for so many hours caused him severe pain. He further complained of being subjected to threats, curses, severe sleep deprivation, humiliation, and psychological torture. Mavtan Comptroller dismissed the complaint and turned to the inspectors’ reports to justify his decision. He acknowledges that “the inspection did not cover the complainant’s entire interrogation, and some [elements of] his complaint may have occurred at times which were not monitored”. Indeed, the dismissal letter states that the inspectors only watched two of the 14 interrogation sessions, and that they only monitored a few hours of each. The Comptroller nevertheless drew on the monitoring of this fraction of the interrogation to conclude that the complaint was unreliable. “The inspectors’ reports,” the letter states, “do not appear to support the complainant’s claims,” since “the inspectors did not report any abnormal events.” He adds that according to one inspector’s account, the interrogation was “conducted in an amiable atmosphere.” These few hours of monitoring hence sufficed for the Comptroller to conclude that the reports “confirm that the GSS interrogators operated legally” (ibid). Demands for increased transparency can thus prove to be double-edged. In a state-controlled “representational economy” of torture (Viterbo Reference Hedi2014), the assumed superiority of visual evidence means that such evidence is often used by state authorities to refute the reliability of victim testimony. We return to this point in the next section.

Third and finally, the CCTV system serves another purpose beyond demonstrating transparency and justifying the dismissal of complaints (even if it does so inadvertently): it helps Israel substantiate its position that torture, if it happens at all, is an aberration rather than the rule. For had torture been systematically used against Palestinian detainees, one could argue, even a selective sampling of interrogations would have detected it. Buying into this argument, however, requires considerable leaps of faith: we would need to assume that the inspectors monitor a sufficient sample of interrogations, that this sample is representative of all interrogations, that the inspectors perform their duties diligently, that they are proficient with the letter of the law, that they would know to identify an infringement when seeing it, that had they witnessed such incidents, they would have reported them to Mavtan, that Mavtan would then recommend a criminal investigation, and that the Mavtan Comptroller will indeed follow up on this recommendation in his decision. The fact that this system fails to identify torture is hence what renders it effective—not in combatting torture, but in contradicting allegations that torture is prevalent in GSS interrogations.

Final Analysis: What Is Justice?

Thus far we proposed that Mavtan’s heightened activity should be understood primarily as addressing the international legal system. Since appeals to international courts are conditioned on the inability to attain legal remedy domestically (“the complementary principle”), this performativity of justice-making works to protect Israel from appeals to external judiciaries. The unit’s mechanisms and practices can thus be seen as merely meant to create a mirage, empty gestures of justice-making. Indeed, some of the HCJ reviews we quote earlier allude to such an interpretation (cf. HCJ 5722/12 2017, HCJ 9018/17 2018).

While this is, no doubt, part of the story, we proposed that there is more at stake. Performativity, as Butler (Reference Butler1993) shows, is never simply an empty gesture, and always does something or produces something in the world, even when it merely affirms power structures. Shielding Israel from international interventions—what we termed “the legitimacy effect”—is a crucial aspect of this performativity that must be understood as emerging through the bureaucratic performativity of accountability, transparency, and due process we reviewed here. Through its thorough examinations, the expansive reports, the detailed decisions issued by the Mavtan Comptroller, its elaborate responses to FOI requests, Mavtan creates a material reality that, as Mitchell (Reference Mitchell and Steinmetz1999, 77) identified in relation to the state, is indistinguishable from the appearance of “the abstract or ideal.” The work of governmental departments, the activities of bureaucrats, the paperwork, the tactics of management, and above all, the appropriate relations between state institutions or different organs of government (institutional independence, appropriate subordinance, oversight mechanisms, legal compliance, and demonstrable transparency and accountability), all these end up producing, we argued, the shielding effect of legitimacy.

Thus, documents are composed, data are generated and accumulated, reports are extracted, new positions are created, expertise are established, training is provided, data-protection mechanisms are introduced, and so on. This bureaucratic laboriousness builds a world, and a way of seeing the world—or, at least, a material image of the world. And this material image comes to stand for “transparency”, “accountability”, or “good governance” more generally, which then become the ground through which Israel substantiates itself as a legitimate player when facing international jurisprudence. But this performativity of justice-making has another effect: it is geared to reshape the discourse of justice.

As we briefly showed earlier, this apparatus serves to provide “proof” that Israel does not torture, and that Palestinian complaints are groundless. In our interviews with MoJ officials and several former GSS interrogators, we were repeatedly reminded that despite having a dedicated unit which laboriously scrutinizes the many thousands of complaints filed by Palestinians, no evidence of torture had ever been found. For our interviewees, this fact did not attest to Mavtan’s failures; rather, they argued, it persuasively showed that the allegations are all false, or wild exaggerations at best. Having such an industrious system conducting lengthy examinations, monitoring interrogations in real-time through a CCTV system, compiling detailed reports and rationalizing data, while being fully transparent and sharing this information with human rights lawyers and advocates, seems to give credence to the claim that Mavtan finds nothing because there is nothing to be found.

But it is not merely the complaints that are deemed fraudulent here; the Palestinian plaintiffs—and even more so, the Palestinians as a collective—are similarly portrayed as deceitful. One GSS interrogator even went as far as pinning the volume of discarded complaints on Palestinian “culture”, prone—so he said—to fabrication and hyperbole. The presumption that Palestinians are untrustworthy, or, at the very least, have ulterior motives for claiming that they have been tortured is so thoroughly embedded in Mavtan’s structure that it can be found already in the documents that led to its establishment. The very reasoning for conducting preliminary examinations was premised, inter alia, on the notion of Palestinian deceptiveness. The guiding thread from the onset presupposed that Palestinians would wrongfully accuse Israel of torturing, not only as individuals but as a collective; as part of an organized campaign against Israel. In the words of the 1987 Landau Report:

The assumption that almost all such allegations are fictitious has led Landau to conclude that it is paramount that prior to launching a criminal investigation, “the examination of complaints should first distinguish between those which are false, and those which are genuine” (para.4.18). And it is this mission with which Mavtan was tasked.

Rubinstein’s 2012 HCJ ruling similarly explained Mavtan’s “necessity” by the need to protect GSS from idle complaints: “Let us not feign innocence”, he argued; there are valid “concerns regarding politically and ideologically motivated false complaints given the nature of the subject at hand” (HCJ 5722/12 2017, para.21). In their justification of Mavtan’s structure, Landau and Rubinstein therefore propose that the prevalence of complaints of torture does not reflect the pervasiveness of illicit interrogation methods by GSS, but rather the dubious motivations of complainants. Such a presumption adheres to, and indeed is part of the production of a wider logic we find in colonial and post-colonial settings: the classification of racialized groups along what Berda (Reference Berda2022) terms “the axis of suspicion.” This classification is “founded on the colonial assumption that native witnesses and their statements were not to be believed” since natives, generally so, “could not distinguish fact from fiction” (Kolsky Reference Kolsky2019, 24). As Chatterjee (Reference Chatterjee2023, 21) argues in his work on anti-Muslim violence in India, in such settings, the law has a role in creating particular subjects “whose very existence … precludes them from witnessing,” and whose testimonies thus can never provide creditable legal evidence.

In reality, however, the volume of complaints underrepresents the actual number of torture cases: most Palestinians do not trust the Israeli legal system enough to file complaints, do not want to validate the system by filing complaints, or drop their complaints due to the hurdles that this system mounts. This is, of course, a self-reinforcing feedback loop: because plaintiffs are presumed to be liars, the elaborated mechanism of examinations which dismisses complaints ends up “proving” that Palestinians are indeed deceitful. Mavtan’s diligent production of what seems to be nothing therefore works to change the meaning of justice itself: justice stands for protecting Israel from Palestinian deceitfulness.

The effects of this heightened performativity range beyond obstructing torture victims’ access to justice. By articulating Palestinians’ search for justice as fraudulence, Mavtan’s work feeds Israel’s broader narration of the Palestinian search for justice as “warfare”: a politically motivated attack by legal means on the Israeli state (cf. Gilboa Reference Gilboa2021). In other words, it is not merely that the system both presupposes and then “proves” that each Palestinian complaint is fraudulent; using the power of state institutions, it can further argue that collectively, these complaints are part of an orchestrated Palestinian campaign to undermine Israel’s very existence, and is hence terror by other means.

And once the Palestinian search for justice is re-defined as terror, the concept of justice itself shifts its meaning. As one reads through the documents involving Mavtan’s establishment and operation, what is striking is the degree to which justice has come to note the vindication of the GSS. The Ciechanover Report, for instance, deemed the introduction of the CCTV cameras necessary since ultimately it would “assist in combating false claims regarding the use of improper measures during interrogations” (Ciechanover Report Reference Report2015, para.138, emphasis added). In our interview, Asher made a similar claim, which was illustrated by providing the conjectural example we mentioned earlier—of an inspector seeing Palestinians injuring themselves and then complaining that they were beaten up by an interrogator. Such a scenario, not only serves to draw a picture—speculative as it may be—of Palestinians as unreliable and of violence in the interrogations as a complete fabrication, it further has the power of taking the most tangible proof that violence had occurred (injury) and rendering it meaningless.Footnote ²² In this inverted system, the transparency Mavtan assumes to provide is geared towards demonstrating the absence of torture.

The current Mavtan’s Comptroller, Shlomo (Shlomi) Abramzon, summarized this attitude best: in some of the cases, he told the PCATI legal team, Mavtan dismisses complaints after finding what he referred to as indications that help expose the truth, or at least, he added, expose that the version provided by the complainant is unreliable.Footnote ²³ By equating the truth with disqualifying the complaint, Abramzon once again reveals, perhaps inadvertently, that ultimately, disproving torture allegations is the justice that Mavtan is striving to produce. Mavtan then does not merely fail to provide justice to torture victims; rather, it labors to rearticulate what justice stands for: the vindication of Israel from Palestinian “conspiracies”.