Security Prisoner
Security Detainee
Administrative Detainee
The Court Presumes: HCJ 5553/09 Public Committee Against Torture in Israel v. Prime Minister of Israel (Judgment of April 26, 2010)
Court Watch | 5553/09 | 1.11.2012 | Dr. Na’ama Carmi

In a High Court petition filed in 2009, the Committee Against Torture in Israel (PCATI), sought a complete ban on the use of the use of restraints as an interrogation method during Israel security Agency (ISA) interrogations; a ban on restraining in positions that cause humiliation, pain or suffering – with specific reference to several such positions, and publication of clear directives for restraining measures that do not cause pain or suffering.[1]

The rights sought to be protected in this petition are the rights to be free of torture and other forms of cruel and/or inhuman and/or degrading treatment; the right to health and physical and mental integrity and the right to dignity. The right to be free of torture and the likes forms part of customary international law, and has attained the status of jus cogens, meaning it cannot be derogated from. Within international treaty law, torture is prohibited under Art. 7 of the International Covenant on Civil and Political Rights (detailed in General Comment No. 20 of the overseeing UN Human Rights Committee); the Convention against Torture and Other Cruel, Inhuman or Degrading Treatment or Punishment (hereinafter: CAT); and in Art. 3 that is shared across the Geneva Conventions as well as elsewhere.[2] Israel is party to all the listed conventions. Torture is considered a grave breach of the Fourth Geneva Convention, and it carries international criminal liability. As noted, the ban on torture is absolute – a matter that is reiterated in the relevant UN committees’ comments, together with the fact that there are no circumstances that could justify torture (e.g., General Comment No. 2 of the UN Committee against Torture).[3] Reference to this is also made in the UN committee recommendations regarding Israel specifically.[4] Torture also contravenes Israeli law, as ruled in HCJ 5100/94 (hereinafter: the torture case).[5]

The petitioner provided a neurological expert opinion that stated restraints could cause interrogees temporary or permanent neurological damage, and argued that even if we assume that the restraining is meant for the safety of the interrogators and to keep the interrogees from fleeing, the same result could be achieved with alternative measures which do not harm the interrogees and cause them pain. In other words, the argument was that restraining was disproportionate under the second proportionality subtest used to assess constitutionality – the less injurious means test – which checks whether the same result could be achieved with less injurious means.[6]

The petitioner argued that these positions, whether in and of themselves, or through a combination of the position and the duration of the binding, cause varying degrees of pain and suffering according to the circumstances and duration of the binding. Two positions of “routine binding” cause interrogees severe pain and suffering “especially through pressure and excessive tightening of the binds on the upper arm or the forearm. Use of this method may result in irreversible neurological damage amounting to partial paralysis of the arm”.[7] The petitioner argued that another position of “elevated binding” could amount to torture.

The State’s Response

In its response, the state rejected the petitioner's claim that restraints are used as an interrogation measure, which partly relied on the fact that many interrogees stated that their restraints were taken off when the interrogators were under the impression that they were willing to cooperate and provide information. This last argument was vehemently denied by the ISA. The response stated that the release of restraints, inasmuch as it occurred, serves as no indication that interrogees provided information. According to the response, interrogees often become less dangerous as they begin to cooperate with their interrogators. ISA protocols give interrogators discretionary power to decide when the threat posed by an interrogee has diminished to the point where the restraints could be taken off. In other words, the state argued that even when interrogees are released from their restraints around the time they start providing information, there is no causal relationship between the two, unless there is a correlation between the interrogees’ cooperation and their diminished threat. However, inasmuch as this occurred in any specific interrogation, the interrogators’ actions are to be examined according to the circumstances.

According to the state, the petitioner's claim regarding use of binding as an interrogation measure should be dismissed for lack of cause, since the ISA does not have the power to use force during interrogations: “There is no protocol or permit allowing to use pressure during interrogation”, and if an ISA interrogator does use such measures, the issue of criminal liability must be explored retroactively, at which point an assessment is made as to whether the circumstances at the time support the necessity defense provided for in the law, in retrospect.[8] This conforms to the ruling made in the torture case (wherein the court ruled that interrogators may avoid criminal prosecution if the attorney general has been persuaded that they acted out of necessity in the specific case, or acquitted on these grounds, if prosecuted and the court accepts this line of defense – N.C). The state pointed to the procedure in place for inquiries in cases where “ISA interrogees who believe improper means were used during their interrogation” and stressed that the cases in which the ISA considered the necessity clause to apply in recent years, and proceeded to act based on this assessment (i.e., used forced during interrogation – N.C.) were “Few and exceptional, amounting to a miniscule proportion of all cases in which suspects were interrogated on terrorism related suspicions”.[9] The state’s contention that this particular argument should be dismissed on lack of cause is odd. The fact that the ISA lacks the legal authority to use force during interrogations, in this instance through painful binding, does not mean such use of force does not take place in reality. The petitioner make factual claims, and the respondent denies them on the factual plain. Dismissal for lack of cause is irrelevant.

As for restraining as a safety measure, the state argued the petition must be dismissed for two reasons: failure to provide the complete factual basis, which amounts to lack of good faith with regards to the description of the manner of restraining (the hands are not tied behind the back, as the petitioner claimed, but on the sides) and on the merits, since such restraining is used “solely for the safety of the interrogators, and, on occasion, of the interrogees as well, and to prevent escape”.[10] Restraining for safety reasons is within the purview of the ISA, as determined in the torture case.

The distinction between measures required for the interrogation, that are instrumental to it, and measures used as an end in itself was made by Supreme Court President Barak in the torture case. Barak distinguished, for instance, between sleep deprivation that is “a ‘side effect’ inherent to the interrogation” and sleep deprivation as a goal in itself, in which case it constitutes prohibited torture. The same holds true for restraining interrogees for the safety of the interrogators, which is legally permitted and is “included in the general power to investigate”. However, painful binding is prohibited. “There are other ways of preventing the suspect from fleeing from legal custody which do not involve causing the suspect pain and suffering”.[11] The UN Committee Against Torture has also expressed concern over the use of restraining measures that are liable to cause unnecessary pain and degradation, as noted in the petition, and held that use of restraints must be constantly supervised and recorded.[12]

The Judgment and the Role of the Court: Critical Review

On the petitioner's arguments regarding use of restraints as a safety measure, the court ruled: “following intervention by the legal advisor’s office and the request of the Petitioner, instructions which improve the posture and the manner in which chaining is carried out have been introduced”.[13]

As noted by the state in its response, the improvements relate to the length of the chain (connecting the handcuffs), that are meant to alleviate the discomfort of the interrogee and were put in place following communications from the petitioner prior to the petition. However, the ruling is interesting as it points to a sort of “negotiation” with the authorities with legal mediation. This opens the door to two more generalized, principled questions. One concerns the need to take to the court for the authority to acknowledge its obligations – the other is the court as a mediator. While the classic role of the court is to decide disputes, and, sitting as the High Court of Justice, provide remedy directly to victims of rights violations by the authorities, we are increasingly seeing the court act as mediator rather than arbitrator, sometimes sending the parties off to reach an agreement or compromise – at times during deliberations, or even when fighting is going on – instead of ruling according the applicable legal norms.[14] In the case considered here, the High Court cannot be said to act as a mediator, but, as is often the case “the shadow of the Court” has a mitigating effect on the arrangements reached.[15] However, in reality, by introducing the improvements, the state implicitly acknowledged that the practice challenged in the petition was improper, or at least, required improvement.

One issue on which the court remained silent was the petitioner's question why binding as a safety measure (to protect the interrogator, to prevent escape from custody and to prevent the interrogees from self-harming – N.C.) was required in ISA interrogations, while interrogees were not restrained during police interrogations, sometimes held in the same facility. In its detailed response, the state addressed the distinctive features of investigations into terrorism or security offenses in contrast to ordinary criminal investigations: the former are forward looking. These types of offenses are committed for ideological-nationalistic reasons and are backed by a supportive, ideologically motivated environment as well as secrecy within the terrorist network. This means that interrogations in these cases are longer and more continuous and that interrogees are more hostile and dangerous. The threat they pose is ongoing and their interrogators are a preferred and attractive target. The state’s response listed multiple examples of attacks or attempted attacks on interrogators, and included a document indicating that according to the Israel Police, interrogees that are restrained during ISA interrogations are restrained during police interrogations as well. This was contrary to the petitioner's claims, which were based on a letter from the Ministry of Public Safety ombudsman, which stated his review indicated that interrogees were not restrained during police interrogations. The court remained silent on these matters, without stating if the response was satisfactory. Did the court agree with the state that there is a crucial different between ISA and police interrogations that could justify restraints during the former? Or did it accept the state’s position that no such difference exists? What was the court’s view on the discrepancies in the description of the facts provided by two official state bodies?

Torture is a classic example of a violation whose legal remedy is obvious. The remedy the court was asked to give was the immediate cessation of the measure. True, when the respondent announced it was changing its conduct – the remedy was provided, whether it was given through a court decision or just thanks to its involvement as a mediator matters little. Still, the petition referred to improvements the respondent purportedly planned to implement even before the petition was filed:

[…] Despite the reported plans by Respondent 1 to increase the length of the chain that connects the handcuffs to alleviate the interrogees’ discomfort by allowing them to place their hands on the sides of their bodies more comfortably, this change has turned out to be cosmetic only, and has hardly affected the pain and suffering of the interrogees, partly because the position and the duration of the binding continue to cause varying degrees of suffering according to the specific circumstances.[16]

The next matter in the state’s response the court addressed is even more interesting: “The Respondents further noted that procedures regarding chaining are in place but that they, by nature, could not be published”.

The distinction between the content of secret discussions and their very existence is addressed by Sissela Bok in her book, Secrets: On the Ethics of Concealment and Revelation,[17] where she also discusses what issues should be considered secret. According to Bok, the unavoidable tension between publication and secrecy could be mitigated by publicizing the moral choice with respect to secrecy itself. The moral arguments regarding any secret practice must be publicly debated. They cannot remain a secret themselves, nor can the fact that these practices exist be kept secret. In other words, the fact that there are secret protocols can be acceptable if their existence is public knowledge and if public debate about the fact that they are a secret can take place. The protocols in question do, at the least, meet the requirement for knowledge that they exist. As evidence, this fact is made public in the state’s response, and is quoted in a public judgment. Everything seems to be in order, until one takes a closer look.

Not publishing protocols because of their “nature” is nothing new. But, what is the issue here? If restraining is not an interrogation measure, as the state claims, whey can it not be publicized? What is it about its “nature” that precludes publication? While there is no expectation for the ISA to expose its interrogation methods, if the restrains are used solely for the safety of the interrogators, it is unclear why the protocols governing their use cannot be disclosed.

In its response, the state references Section 19 of the Israel Security Agency Law 5762-2002 which stipulates that “The rules, directives and protocols of the ISA, the identity of its employees and those working on its behalf, in the past or in the present, as well as other information regarding the ISA set forth in regulations are confidential and their publication is prohibited”. The fact that there is a legal basis for non-publication is, naturally, relevant, but it is a formalistic response that fails to provide a pertinent reason for not publicizing the rules. Given the criticism laid out above, it appears that the balance between the exposure of protocols that, according to the state itself, do not constitute methods of interrogation, and the need to ensure force is not used during interrogations and that the state meets its obligation to constantly monitor the use of restraints – tips in favor of publication. This would enhance the state’s ability to meet the absolute ban on torture and ensure it does not take place.

The court concludes the dismissal of the petition:

In view of the general nature of the petition and with attention to the explanations provided by the State, we have not seen cause to intervene and provide a concrete remedy in the petition herein. It has been confirmed before us that each individual complaint is investigated, vouched for by the State Attorney’s Office. We do not find, given the information before us, that we are able to conduct a different individual examination. It has also been made clear to us that the legal counsel department conducts its own review of the lawfulness of employed measures and that procedures are in place with respect to this issue. We presume that the procedures take into account the type of handcuffs in use, that care is taken that they are not used for exerting pressure, that they do not cause pain and that the manner in which the chained interrogees are seated will not harm their health [emphasis added – N.C.].


Legal discourse, and the court in general, served as a mediator of sorts, through whose mediation, or under whose shadow, improvements were made to restraining practices. In other words, there was a tacit admission that the practice that was in use was improper – an admission the court remained silent on. The court put its trust in the fact that “interrogee complaints are examined individually in order to make sure that chaining is not misused” and that “the Petitioner may bring allegations in individual cases”. It also accepted the state’s assertion that if it came to light that an interrogator had broken protocol and used restraints as an interrogation method “the case will be judged according to the circumstances”. It is worth noting that the state addressed the individual cases included in the petition in detail, stating that is some of them, the Inspector of ISA Interrogee Complaints had never contacted, a point that blunts the force of the petition. Other cases, the state argued, were “irrelevant” because they concerned interrogations carried out before the improvements were introduced. In still others, the state charged false claims were made as to the length of the interrogation. Complaints made by three interrogees had been transferred to the State Attorney’s Office and were still under review. On the other hand, the report issued by the UN Committee against Torture indicates that complaints of torture by the ISA are reviewed internally by the Inspector of ISA Interrogee Complaints and that recommendations made by this office to end probes without launching criminal investigations are routinely upheld by the Ministry of Justice and the attorney general. According to the UN Committee this indicates Israel does not have a true system for investigating torture complaints.[18]

The court accepted the fact that protocols on the use of restraints as a safety measure did exist, and that there was no need to publicize them, even though it accepted the state’s assertion that these were not interrogation methods that, by nature, could not be disclosed. The court itself was not made privy to the protocols but did make presumptions regarding what they took into account and what they took care to do, and, partly based on these assumptions, dismissed the petition.

In this case, as in others, the question that needs to be asked is – what is the basis for these presumptions regarding the rules and how rigidly they are followed? What is the basis for the court’s assumption that the authorities obey the law and actually practice what they claim to practice? Did the court not put too much faith in the authorities? Is it really a presumption that “care is taken”? Does past experience justify such faith in the arguments the ISA presents to the court? Or – should the court have been more cautious with its presumptions, especially with regards to protocols it never sees.

The petition in question was directed against a restraining method. The court was not prepared to grant remedy or explore the issue as the petition was too general. Still, could it have not “presumed” but rather affirmed that clearly the ISA must ensure that “the procedures take into account the type of handcuffs in use, that care is taken that they are not used for exerting pressure, that they do not cause pain and that the manner in which the chained interrogees are seated will not harm their health”? Such an assertion would have carried more weight. Then again, the court still would have had no way of ensuring its instructions were followed, since, as stated, not only did it stop short of ordering the publication of the protocols, which would have allowed for monitoring and supervision, it was not privy to them itself.

Dr. Na’ama Carmi

The author has a PhD in philosophy and has served as chair of the Association for Civil rights in Israel. She is the author of the blog Reading and Writing.

HCJ 5553/09 Public Committee Against Torture in Israel v. Prime Minister (2010), petition of July 5, 2009.
See UN Committee Against Torture (CAT), General Comment No. 2: Implementation of Article 2 by States Parties, (2008).
See UN Committee Against Torture, Concluding observations of the Committee against Torture: Israel, June 23, 2009, para. 10; and UN Human Rights Committee, UN Human Rights Committee: Concluding Observations: Israel, August 21, 2003, para. 18.
HCJ 5100/94 Public Committee Against Torture in Israel v. State of Israel (1999), judgment of September 6, 1999; see commentary on this judgment.
The other two proportionality tests are the rational means test – whether there is a rational connection between the means chosen and the purpose pursued; and the test of proportionality in the narrow sense – whether the damage does not outweigh the benefit. Any action by the authorities must meet all three subtests, and failing to meet just one suffices to rule the action disproportionate. This proportionality analysis has been used by the courts in many scores of judgments; see, e.g., HCJ 2056/04 Beit Sourik Village Council v. Government of Israel (2004), judgment of June 30, 2004, para. 41.
See above note 1, sect. 7 of the petition.
The necessity defense is provided for in Sec. 34(K) of the Penal Code: “No person shall bear criminal responsibility for an act that was immediately necessary in order to save his own or another person's life, freedom, bodily welfare or property from a real danger of severe injury, due to the conditions prevalent when the act was committed, there being no alternative but to commit the act”.
See above note 1, Preliminary Response on behalf of the Respondents, September 15, 2009; emphasis in the original.
See above note 5, paras. 31 and 26 (respectively) of the judgment.
UN Committee Against Torture, Concluding observations, Australia, UN Doc. A/56/44, paras. 47-53, November 21, 2000, para. 52(b).
See above note 1, judgment of April 26, 2010 (in Hebrew).
See David Kretzmer, “The Supreme Court of Israel: Judicial Review during Armed Conflict”, German Yearbook of International Law, Vol. 47, 2004.
See David Kretzmer, The Occupation of Justice: The Supreme Court of Israel and the Occupied Territories, 2002.
See above note 1, sect. 52 of the petition.
Sissela Bok, Secrets: On the Ethics of Concealment and Revelation, 1989.
The Public Committee Against Torture in Israel, Accountability Denied: The Absence of Investigation and Punishment of Torture in Israel, (2009). PCATI petitioned the HCJ together with other human rights organizations in a bid to change the attorney general’s policy and have him exercise his authority to order a criminal investigation in every case of suspected torture. The court gave a partial judgment in this petition (on the principled portion of the petition), ruling that the operation of the Inspector under the Ministry of Justice was not improper. See HCJ 1265/11 Public Committee Against Torture in Israel v. Attorney General (2012), petition of February 14, 2011; (02) 627 1698   (02) 627 6317

red-id | רד אינטראקטיב