The Gambia v. Myanmar and the Nature of Genocidal “Intent to Destroy” before the ICJ (As Opposed to Criminal Courts)

[Zsuzsanna Deen-Racsmány holds a Ph.D. in public international law from Leiden University. She is an independent researcher and has worked, inter alia, at Leiden University and the University of Amsterdam, and been a rapporteur for Oxford International Organizations.]

In January 2026, the International Court of Justice (ICJ or Court) held public hearings on the merits in Application of the Convention on the Prevention and Punishment of the Crime of Genocide (The Gambia v. Myanmar). In their oral pleadings, both Parties accepted the dominant interpretation of “intent to destroy” in Article II of the Genocide Convention (Convention)—developed in the jurisprudence of the ad hoc criminal tribunals for Yugoslavia (ICTY) and Rwanda (ICTR), confirmed by the ICJ in Bosnia v. Serbia (para. 344) and, more categorically, in Croatia v. Serbia (para. 136)—that genocidal intent must seek the physical-biological destruction of a protected group (e.g., here, p. 32, para. 26; and here, pp. 44-45, paras. 15-16). The declarations of the intervening States, not heard during the oral pleadings, have not challenged this position either. Consequently, the Court may not need to address this matter in its forthcoming judgment, beyond restating its previously reached conclusions. 

In this post, I argue that the Judges still can and ought to seize this opportunity to revisit the ICJ’s position regarding this point. To underpin the need to do so, I briefly address a main flaw of the Court’s prevailing interpretation of genocidal intent, one related to a point raised by Myanmar during the oral pleadings: the ICJ’s mechanical reliance on the jurisprudence of the ICTY and ICTR and its corresponding failure to distinguish between interpreting Article II of the Convention as a treaty rule and as a criminal law provision.

Why Must the Court Still Consider This Matter in Detail?

The interpretation of “intent to destroy” is clearly relevant to the case. Already a 2018 report of the Independent International Fact-Finding Mission on Myanmar (FFM) concluded that large-scale sexual violence (i.e. genocidal actus reus) had been committed with intent “at least in part, to weaken the social cohesion of the Rohingya community and contribute to the destruction of the Rohingya as a group and the breakdown of the Rohingya way of life” (para. 941, cited in the Gambia’s Application, para. 98), i.e. with intent to destroy the group other than physically or biologically. Moreover, in its 2026 oral pleadings, the Gambia frequently referred—or quoted texts referring—to destruction of the group’s social cohesion, social fabric, the “architecture of Rohingya social life”, “Rohingya way of life” or culture, and to forced displacement (here p. 20, para. 3; p. 21, para. 4; p. 25, paras. 14-15; p. 26, para. 18; p. 29, para. 10; p. 66, para. 29; p. 71, para. 46 and here pp. 18-19, para. 31; p. 30, para. 25). Though the relevant arguments may have been presented in support of a pattern of conduct from which genocidal intent may be inferred the invoked concepts are commonly associated with groups’ social destruction and thus with cultural genocide, a legal category which the ICJ declared not to fall under the Convention’s prohibition. 

However, mounting volumes of critical studies (e.g., Novic, Mundorff, Abtahi (Chapter 5), Berster (‘Article II’) and Greenawalt) published since Croatia v. Serbia convincingly argue that intent to destroy a group in a non-physical-biological sense also constitutes genocidal mens rea under the Convention. Accordingly, and considering the previously long-prevailing disagreement regarding this matter (for dissent in the jurisprudence see, e.g., Krstić Appeal, Partial Dissenting Opinion of Judge Shahabuddeen, paras. 48-54, Blagojević and Jokić, paras. 657-666, Krajišnik, para. 854; and Jorgić, para. (III)(4)(a)(aa), margin note 22), it is understandable that Myanmar pre-emptively emphasized already in 2019 the “important distinction between physical and cultural genocide, one to which this Court has previously attached considerable significance” (p. 34, para. 40), pointing to a statement in the above-quoted FFM Report (para. 1438).

As the Gambia, too, accepted the position that actus reus committed with intent to destroy the group in a non-physical-biological sense does not constitute genocide, the Court strictly speaking does not need to address that matter beyond reciting its established position. However, it is still entitled do so. Whereas the principle of party presentation may limit the Judges’ freedom to step beyond the Parties’ submissions in terms of facts and inferences drawn from those, the interpretation of the Convention (i.e. of the law) is a different matter: jura novit curia. The ICJ’s competence to interpret the law extends beyond considering the arguments raised by the Parties, and it may reject legal arguments presented by them even if those are not contested by the other Party.

It may even depart from its own jurisprudence when “there is cause not to follow the reasoning and conclusions of earlier cases” (para. 28). Accordingly, it is not bound by its position arrived at in Bosnia v. Serbia and confirmed in Croatia v. Serbia regarding the nature of genocidal “intent to destroy”. It may reopen this question, should the Judges deem it necessary to do so. 

Importantly, in its intervention in South Africa v. Israel, Mexico (para. 35) rejected the Court’s established position on this question. Though—and, indeed, as—the Court’s judgments are final and may not be subjected to review to correct the underlying legal reasoning (ICJ Statute, Art. 61), it would appear prudent for the Court to address this issue in detail in The Gambia v. Myanmar in light of this already known challenge. 

Revisiting the matter in The Gambia v. Myanmar appears called for also due to Myanmar’s submissions concerning an apparently unrelated matter: the distinction between individual criminal responsibility and State responsibility. This was—in part due to questions posed to the Parties by Judges Charlesworth and Hmoud (pp. 44 and 47-48, respectively)—a recurring issue in the oral pleadings, predominantly regarding the required standard of proof. However, Myanmar’s original argument challenging the Gambia’s submissions (para. 3.5)—which clearly inspired at least Judge Hmoud’s question—was that the distinction between individual criminal responsibility and States responsibility had to be observed in relation to the interpretation and application of the criteria relevant for proving genocide in general, including genocidal intent (p. 41, paras. 5-6). “Par conséquent, la prise en compte de la jurisprudence des juridictions pénales internationales doit se faire avec discernement et non automatiquement” (here p. 41, para. 6, see also p. 46, para. 21; p. 51, para. 41 for similar statements). While the Gambia subsequently did not address this argument except regarding the required standard of proof (pp. 24-28), it continued to cite ICTY and ICTR judgments (e.g., here p. 29, para. 31; p. 41-42, paras. 31, 33) in support of its arguments in other contexts without explanation. The Parties thus appear to disagree on this point. Moreover, as discussed below, Myanmar’s submission concerns a fundamental issue which, left unaddressed, could impair the validity of the Court’s conclusions concerning several elements of its judgment, including the treatment of the nature of genocidal intent. 

The ICJ’s Interpretation of Genocidal “Intent to Destroy” Replicates the Jurisprudence of the Ad Hoc Criminal Tribunals

The Convention does not specify that genocidal intent must seek to destroy a group physically or biologically. The interpretation of “to destroy … a group” in the chapeau of Article II consistently with the law of treaties (see VCLT, Arts. 31-32, reflecting customary international law) does not support such a requirement either. 

Significantly, even Serbia’s counsel (p. 14, para. 12.) acknowledged in Croatia v. Serbia that

[d]epending upon how canons of interpretation were applied to the problem, it was possible, in the past, for reasonable people to differ about the scope of the words “with intent to destroy”, and more specifically as to whether the word “physically” should be added as an adverbial modifier of the verb “to destroy”. 

However, as he promptly added, the question was conclusively settled by the ICTY—starting with Krstić (para. 580)—and by the ICJ in Bosnia v. Serbia (para. 344), in favour of reading that qualification into the text of Article II. Significantly, the ICTY and the ICJ—like the International Law Commission (‘ILC’) on whose position expressed in the commentary of the Draft Code of Crimes Against the Peace and Security of Mankind (Draft Code) (pp. 45-46, para. 12) both heavily relied in this context—based this interpretation on the Convention’s travaux préparatoires, without even briefly referring to the primary means of interpretation prescribed by the law of treaties. In Croatia v. Serbia (para. 136), the Court endorsed this approach. 

Differences between the Interpretation of Treaties and Criminal Statutes

Here, I do not wish to challenge the accuracy of the reading of the Genocide Convention’s travaux préparatoires that underlies the dominant interpretation of the required nature of genocidal intent (for recent assessments concluding that that interpretation of the preparatory work is incorrect, see, e.g., Mundorff, Abtahi (Chapter 5), Berster (‘Article II’) and Greenawalt). However, it is respectfully submitted that the ICJ in any event erred in giving preponderant weight to the preparatory work when interpreting genocidal intent under the Convention: it should not have unreservedly followed the ILC’s and ICTY’s approach, adopting the same “canons of interpretation”. 

Admittedly, the ILC (see Draft Code, pp. 45-46, para. 12), the ICTY (see Krstić, paras. 574-580) and the ICJ faced the task of interpreting literally the same definition. It thus seems sensible that they reached the same conclusion. However, while the ILC (see Draft Code, Arts. 2 and 4) and the ICTY (see Krstić, para. 97) addressed individual criminal responsibility for genocide and interpreted Article II or a corresponding provision in that context, the ICJ dealt with it from the perspective of State responsibility (see Bosnia v. Serbia, paras. 155-179 and Croatia v. Serbia, paras. 128-129). 

Significantly, as noted by Judge Gaja in his Separate Opinion (pp. 394-395, para. 2) in Croatia v. Serbia, “[w]hile it would seem logical to give to the definition of genocide the same meaning with regard to State responsibility and the criminal responsibility of individuals, there are reasons for the international criminal tribunals to adopt a restrictive approach to the definition which are not applicable when one considers State responsibility”. He identified the overlap between genocide and other crimes within the jurisdiction of international criminal courts and tribunals which are easier to prove as a pertinent practical consideration to interpret the definition of genocide narrowly. In turn, the criminal law principle of legality and its corollary requirement of strict construction offer criminal courts and tribunals was a compelling legal reason to do so. Though Judge Gaja’s observations addressed the definition of genocide in general, there are no obvious reasons why they should not apply to the here addressed aspect of that definition.

Let us zoom in on the second (legal) consideration cited by Judge Gaja, which in essence concerns the applicable rules of interpretation. The criminal law framework (of the Draft Code and Krstić)—and in particular the principle of legality in that context (see Krstić, para. 580 on the decisive role of this principle)—places considerable emphasis on “legislative intent”. As noted by the ICTY Trial Chamber in its brief yet enlightening discussion of the “exercise […] of interpretation generally and of the [domestic an international] criminal law in particular” in Čelebići (paras. 402-413), in criminal law, “[t]he rule of strict construction requires that the language of a particular provision shall be construed such that no cases shall be held to fall within it which do not fall both within the reasonable meaning of its terms and within the spirit and scope of the enactment” (para. 410, emphases added). 

Moreover, Grover (pp. 377-378, emphasis added) observed in relation to interpreting the crime of aggression—a crime subject to a similar interpretative dichotomy—in criminal law that 

the general rule of interpretation in article 31 of the Vienna Convention remains applicable, but ‘object and purpose’ and the idea of effective interpretation are to be read down to exclude considerations of the treaty regime’s perceived goals (e.g. ending impunity). Without reading down these rules, they could be applied at such a general or abstract level that the rule of strict construction would itself be rendered ineffective. […] The rule of strict construction does not […] have a fixed place in interpretative reasoning—all canons of interpretation may be considered in any order—but must be considered at some stage and, in doing so, regarded as an essential device for safeguarding the principle of legality.

Conversely, the law of treaties does not have a principle corresponding to that of strict construction in the context of interpretation (or a principle of legality). In addition, it does not attribute the same weight to specific means of interpretation as criminal law, or even require consulting all of them in all circumstances. In particular, under the law of treaties, recourse to the travaux—a subsidiary means of interpretation—is explicitly foreseen (apart from the possibility to confirm the interpretation arrived at through the application of the customary rule identified in Article 31 VCLT) only to determine the meaning of a provision “when the interpretation according to article 31: (a) leaves the meaning ambiguous or obscure; or (b) leads to a result which is manifestly absurd or unreasonable.” Although the phrase “to confirm the meaning” may be interpreted liberally (see, e.g., Gardiner and le Bouthillier), this does not change the fact that—contrary to criminal law—the law of treaties attributes a limited role to legislative intent. 

Furthermore, as noted in the Krstić Appeal Judgment (para. 37), the Convention itself already “guard[s] against a danger that convictions for this crime will be imposed lightly”, by “demanding proof of specific intent and the showing that the group was targeted for destruction in its entirety or in substantial part.” These observations appear to negate the need for reading the definition of genocide more strictly than is required by the Convention as interpreted according to the law of treaties, in any event before the ICJ. 

To conclude, the difference between the two legal disciplines is too substantial, rendering it all the more plausible that the interpretation of the Convention—and identically worded penal statutes—may yield different results in a criminal law context than its interpretation according to the law of treaties from a civil law perspective. After all, the Genocide Convention has a dual nature (pertaining to individual criminal responsibility and State responsibility), and its contents do not negate the possibility of two different interpretations in the two contexts wherein it accordingly operates. Significantly, the ICJ too confirmed that “State responsibility and individual criminal responsibility are governed by different legal regimes and pursue different aims”, and that its task was limited to the first framework (Croatia v. Serbia, para. 129). Judge Gaja took that argument even further, positing that the underlying definition may be interpreted differently in the two contexts. Recognition of this possibility would enhance the effectiveness of the Convention’s primary goal to suppress genocide, allowing its optimization within the specific confines of each context. Accordingly, and for the above reasons, it is respectfully submitted that the Court should reconsider its reliance on the (criminal law) approach adopted by the ILC and ICTY—in particular the weight they attributed to the travaux—and on their resulting interpretation of the “intent to destroy”.