The Responsibility of Syria under the Convention Against Torture before the ICJ

With the oral proceedings and marathon round of interventions over the past two weeks on the preliminary objections in the Allegations of Genocide (Ukraine v Russia) case (on which, see the excellent summary here), one might forget that the International Court of Justice will be in session again from 10 October to hear pleadings in yet another case concerning an ongoing armed conflict. This time, it will be to hear the pleadings of the parties in relation to Canada and the Netherlands’ request for the indication of provisional measures following their joint Application in the summer against Syria, alleging violations of the Convention Against Torture. The applicants assert that, since the start of the civil war in 2011, Syria has committed, and continues to commit, a range of violations of the CAT, including through acts of torture and inhuman treatment attributable to the State itself, as well as through its failures to meet the various positive obligations of prevention and investigation under the CAT.

The case, including the first phase of proceedings on provisional measures, is an interesting and important one for a number of reasons, a few of which I will briefly consider here: first, the broader context of armed conflict; second, the scope of obligations under the CAT that is asserted by the applicants; and, third, the particular relief sought.

The context of armed conflict

First, like the Allegations of Genocide case, it is one of five current contentious cases pending before the Court that relate to alleged violations of international law in the context of (ongoing or recurring) armed conflicts (the others being Application of the ICSFT and CERD (Ukraine v Russia), Armenia v Azerbaijan, and Azerbaijan v Armenia). The Court has certainly become an established front in the progress of these conflicts. Yet, unlike a number of the Court’s earlier cases involving claims in the context of armed conflicts that were brought under the optional clause, all of these pending claims are brought on the basis of compromissory clauses that appear to prevent the explicit airing of the broader ad bellum and in bello disputes that lurk in the background. This has not stopped allegations of aggression or international humanitarian law violations from being made by the disputing parties, however, though to different degrees of relevance to the precise causes of action.

In some of the cases, references are merely made in passing to responsibility under these other rules of international law, e.g. in Armenia’s frequent references to Azerbaijani ‘aggression’ in its Application. At other times, these other rules of international law are hinted at by the invocation of alleged facts that potentially raise issues under those rules as well as the particular treaty on the basis of which jurisdiction is established, e.g. in Azerbaijan’s claims in its own Application that the alleged use and non-removal of landmines by Armenia engage Armenia’s responsibility under the Convention on the Elimination of Racial Discrimination. Indeed, this potential overlap was a point of contention during the oral proceedings on Azerbaijan’s request for the indication of provisional measures (e.g. CR 2021/26, p 16). In the end, the Court agreed with Armenia that CERD could not plausibly be read as imposing ‘any obligation on Armenia to take measures to enable Azerbaijan to undertake demining or to cease and desist from planting landmines’ (2021 Order, para 53).

Finally, as seen in Ukraine’s submissions in the last few weeks in Allegations of Genocide, there has also been a more explicit renvoi to those extraneous rules in order to interpret the applicable law. In this case, Ukraine argues that Article I of the Genocide Convention itself obligates States to comply with other rules of international law (such that violations of those other rules in the course of preventing genocide, e.g. the UN Charter rules on the use of force, necessarily also violate Article I) (Ukraine, Written Statement on Preliminary Objections, paras 129-40; a more reasoned articulation of this same idea, grounded in Art 31(3)(c) of the Vienna Convention, was also advanced, e.g., in Latvia’s intervention, para 52).

This is an altogether different context to those earlier ICJ cases dealing with armed conflicts, such as DRC v Uganda and Nicaragua v US, where the Court’s broad jurisdiction extended to consider the parties’ ad bellum and in bello responsibility. Instead, it seems likely that we will see the same interpretive battles in some of these cases as we saw in Oil Platforms, where again an apparently narrow compromissory clause became the door through which the respondents’ responsibility under the law on the use of force was litigated (see Judge Higgins’ Separate Opinion, paras 45-54).

Relatedly, and as has become a common feature of its defence in cases brought against it, Russia argued at length that Ukraine’s application under the Genocide Convention represented a mischaracterisation of the ‘real’ dispute between the parties (i.e. that concerning Russian recognition of the two so-called republics and responsibility under the UN Charter) (Russia, Preliminary Objections, paras 122-37): ‘What this case is really about is the legality of the special military operation and the recognition of the DPR and LPR as States. The legality of these actions not under the Genocide Convention, but under the UN Charter and customary international law. The entire case before you hinges upon this’ (CR 2023/13, p 60).

As I have shown elsewhere, the ICJ has tended to reject such arguments concerning an alleged ‘mischaracterisation’ of a dispute where it otherwise appears on the facts that a legal dispute concerning the particular treaty invoked by the applicant exists (see here; see also the excellent post by Matina Papadaki here). Nonetheless, one can see some disquiet by judges over the tendency to bring cases under compromissory clauses within treaties that somehow seem far removed from what is alleged to be the ‘real’ dispute. Thus, in his declaration attached to the February 2023 Order on provisional measures in Armenia v Azerbaijan, Judge Yusuf made the following point:

‘My objection relates to the continued misuse of the compromissory clause of CERD as a basis of jurisdiction of the Court with respect to alleged acts and omissions which do not fall within the provisions of that Convention. A regrettable tendency seems to have developed, whereby any State that fails to find a valid basis of jurisdiction of the Court for its claims, but still wishes to bring a case before it, tries to stuff those claims into the framework of CERD’ (para 3).

For Canada and the Netherlands, their particular action against Syria should comfortably avoid any such criticism about the ‘real’ dispute. First, the facts alleged directly fall within the provisions of the CAT without appearing to misrepresent an alternative claim, ‘stuffed’ into the framework of the CAT. Second, the preconditions to ICJ jurisdiction under the CAT’s compromissory clause in Article 30 (attempted negotiations and arbitration) mean that, once those preconditions are met (as seems to be the case here), it is difficult to imagine that the Court will not consider there to be a dispute between the parties (of which all were aware) specifically in relation to the CAT (cf Belgium v Senegal, para 54) (though it is possible that, on the facts, an especially granular assessment by the Court of the existence of a dispute might reach different conclusions in relation to particular CAT provisions). In any case, at this early stage, establishing prima facie jurisdiction and the plausibility of rights (and claims) should not be too burdensome.

The scope of obligations under the CAT

The second point of interest I wish to flag relates to the scope of the obligations under the CAT that are asserted by Canada and the Netherlands. In their Application, Canada and the Netherlands assert violations of the CAT both on the basis of alleged acts of torture and other cruel treatment by Syria itself (e.g. at para 59(a) and (c)) and on the basis of its failure to fulfil the various prevention, investigation and other procedural obligations under the CAT (e.g. para 59(b), (d)-(l)). It is the former claims that are especially interesting, as nowhere in the CAT is there an explicit prohibition of torture; indeed, the prohibition of torture as a matter of international law was rather taken for granted by the drafters of the CAT, which itself was designed to make that prohibition effective (see Zach, Commentary to Art 1, para 133). The CAT thus focuses not on the negative duty (a prohibition of torture) but rather on various positive duties (of prevention and investigation).

In their request for provisional measures, Canada and the Netherlands assert that the CAT contains a negative duty prohibiting torture and other cruel treatment, without citing any authority for that proposition (PM Request, para 16). The UN Committee Against Torture does appear to treat torture by the State as itself a violation of the CAT (specifically of Articles 1 and/or 2(1)) (see, e.g., Ali Ben Salem v Tunisia, paras 16.4-16.5). Indeed, the preambular reference to the prohibition of torture under Article 7 ICCPR and the overriding aim of the CAT as enhancing the ‘struggle’ against torture (see also Belgium v Senegal, para 68) certainly offer a strong basis for arguing either that the prohibition itself is implicit in the CAT (most persuasively in the Article 2(1) duty of prevention), or that the conventional and customary prohibitions of torture are ‘relevant’ rules in the Article 31(3)(c) VCLT sense when interpreting the substantive CAT provisions. The latter approach brings us back to the Allegations of Genocide analogy and the question of extraneous rules.

The former approach, on the other hand, invites an analogy either to the Court’s (expansive) approach in Bosnia v Serbia or its more restrictive approach in Application of the ICSFT and CERD (Ukraine v Russia), both of which raised similar issues. In the former case, the Court relied on the text and object and purpose of the Genocide Convention to find that the positive duty of prevention in Article 1 a fortiori contains a negative duty prohibiting States parties themselves from committing genocide and the other covered acts (paras 166-9). In the latter case, the Court was faced with the similar argument that the International Convention on the Suppression of the Financing of Terrorism, which contains various positive duties concerning the domestic criminalisation of terrorist financing by individuals, also implicitly contains a negative duty prohibiting States parties themselves from financing terrorism. On the basis of the text and travaux, the Court rejected this argument, without explicit consideration of its reasoning in Bosnia v Serbia (notwithstanding that the parties themselves raised this analogy in their pleadings) (para 59).

At the provisional measures stage, the Court need not concern itself too much with this question. And when it does arise in the subsequent stages of the proceedings, there would seem to be a strong argument that the approach of the Court in Bosnia v Serbia is the better analogy here, given the similar scope and placement of the obligation of prevention under the CAT and the Genocide Convention, the peremptory nature of the rules they seek to enforce, and the absence of any evidence in the travaux that direct State responsibility under the CAT for torture was rejected (Zach, Commentary to Article 2, paras 6-10). This would seem an even stronger argument here given that the definition of torture in Article 1 requires official involvement. It should also be noted that the unusually worded compromissory clause in Article IX of the Genocide Convention, that specifically envisages claims ‘relating to the responsibility of the State for genocide’, was explicitly stated by the Court in Bosnia v Serbia not to be dispositive of the scope of obligations, given the non-substantive nature of the clause (para 166); the absence of such wording from the compromissory clause in the CAT thus does not exclude this reading, which had been one of Russia’s arguments regarding the ICSFT (CR 2019/9, p 38).

Invoking the aggravated regime of responsibility

The third and final point of note that I wish to raise relates to the relief sought by Canada and the Netherlands in their Application, in particular the request that the Court ‘declare that Syria has committed a serious breach of a peremptory norm … and determine the legal consequences thereof’. This is a reference to the aggravated regime of responsibility codified in Articles 40 and 41 of the ILC Articles on State Responsibility. Notwithstanding the rather uneasy birth of these provisions and the criticisms that have been levelled at them (see, e.g., Tams, pp 770-75; Cassese), the notion of certain additional consequences for all States attaching to violations of obligations of a particular character does find some (varied) support in ICJ jurisprudence (e.g. in Israeli Wall, paras 154-60; Chagos, para 180; Germany v Italy, para 93 (albeit obiter)).

As is well known, however, the Court in Israeli Wall and Chagos did not rely on a general international law obligation arising from violations of peremptory rules, but rather on other legal bases, such as common Article 1 of the Geneva Conventions or the erga omnes nature of the obligations. Indeed, the Court has tended to shy away from relying explicitly on the more controversial (or ‘progressive’) provisions of the ILC Articles (Paddeu, p 424). Nonetheless, the ILC reiterated, and partly strengthened, Articles 40 and 41 in its Draft Conclusions on the identification and legal consequences of peremptory norms, adopted last year (see Conclusion 19) (not without controversy in the Sixth Committee: see Aust).

This particular request by Canada and the Netherlands asks the Court, in the context of contentious proceedings, explicitly to engage with this question of the scope of additional legal consequences flowing from a serious breach of peremptory norms under existing international law. This would be a very welcome opportunity for the Court to clarify its views here, should the case reach the remedies stage. On the particular facts of the case, the request could even possibly raise (if only by implication) the responsibility of States that have supported the Assad regime, such as Russia. Of course, any attempt to take the pleadings that far would risk the Court avoiding (some or all of) these questions on the basis that they would too strongly implicate the legal interests of a State that is not a party to the proceedings (Monetary Gold, p 32).