The International Court of Justice (ICJ) will hold public hearings in the case brought by the Gambia against Myanmar starting next week. The case involves serious allegations of genocide committed against the Rohingya group. This article concerns some technical issues which the Court may wish to clarify in its final judgment, regarding inference of genocidal intent from a “pattern of conduct.”
The Court applies a high standard of proof in genocide cases. As first explained in Bosnia v. Serbia, “[t]he Court requires that it be fully convinced that allegations made in the proceedings, that the crime of genocide or the other acts enumerated in [the Genocide Convention] have been committed, have been clearly established.” Allegations of genocide must be proven by evidence that is “fully conclusive.”
Genocidal intent may be proven by direct evidence of a genocidal plan or by evidence of a pattern of conduct from which the only reasonable inference is that the conduct was committed with genocidal intent. As the Court later explained in Croatia v. Serbia, its “only reasonable inference” test is “in substance identical” with the standard applied by international criminal tribunals, namely that:
it is permissible to infer the existence of genocidal intent based on ‘all of the evidence taken together’, as long as this inference is ‘the only reasonable [inference] available on the evidence’. (Prosecutor v. Tolimir, para. 745)
So understood, the Court’s standards of proof and inference fit together. Evidence of genocidal intent is fully convincing if, and only if, genocidal intent is the only reasonable inference based on “all of the evidence taken together.”
Unfortunately, some States appear uncertain and apprehensive of the Court’s approach to pattern-of-conduct evidence. Some worry that only the similarity of acts will be considered while their scope and severity will be ignored, or that dissimilar acts such as killing and sexual violence will be considered separately, compartmentalizing evidence that should be viewed together. Others fear that pattern-of-conduct evidence excludes other relevant, probative evidence of genocidal intent, such as statements other than official plans. This was never the Court’s intent, and it has not been the Court’s practice. Gambia v. Myanmar presents an opportunity to dispel any lingering misunderstandings.
Briefly, a pattern of similar conduct by different individuals, at different times, in different locations, may provide important evidence that the conduct was directed by higher authorities, or was otherwise coordinated, and reflects a common intention. Pattern-of-conduct evidence is unnecessary when a single incident raises an inference of genocidal intent on its own, or when other evidence exists that multiple incidents were directed or coordinated. Once a common intent is established, other evidence may be introduced to prove genocidal intent, including the scope and severity of the conduct, overlapping patterns of conduct, and statements other than official plans. So long as genocidal intent is the only reasonable inference from all the evidence taken together, the Court should be fully convinced.
The Caselaw
In Bosnia v. Serbia, the Court found that the Army of the Republika Srpska (the VRS) committed genocide in Srebrenica, specifically intending to destroy a substantial part of the Bosnian Muslim group through a combination of killing the military-aged men and expelling the women, the children, and the elderly. The Court did not find an official statement of genocidal intent, nor did it infer genocidal intent from a pattern of conduct. The Court simply evaluated the evidence placed before it and was fully convinced that the acts were committed with genocidal intent.
With respect to atrocities committed outside of Srebrenica, the Court found that genocidal intent had not been “conclusively shown in relation to each specific incident.” Anticipating this finding, Bosnia attempted to link these specific incidents with each other and with Serb leaders. Bosnia proposed that the Court should infer genocidal intent from “the pattern of [potential] acts of genocide committed throughout the territory, against persons identified everywhere and in each case on the basis of their belonging to a specified group.” Bosnia alleged that the genocidal intent of the higher authorities “directing the course of events is clear from the consistency of practices, particularly in the camps, showing that the pattern was of acts committed ‘within an organized institutional framework.’” According to Bosnia, the consistent pattern established that higher authorities must have directed the conduct. Given the nature and scale of the conduct, these Serb leaders must have intended to destroy the group in whole or in part.
The Court noted that Bosnia’s argument “moves from the intent of the individual perpetrators of the alleged acts of genocide complained of to the intent of higher authority.” The Court examined whether genocidal intent could be deduced “from the pattern of actions against the Bosnian Muslims taken as a whole.” It was in this context that the Court famously said that:
The dolus specialis, the specific intent to destroy the group in whole or in part, has to be convincingly shown by reference to particular circumstances, unless a general plan to that end can be convincingly demonstrated to exist; and for a pattern of conduct to be accepted as evidence of its existence, it would have to be such that it could only point to the existence of such intent.
As the Court says, genocidal intent may be convincingly shown by reference to “particular circumstances,” a broad notion that covers any circumstantial evidence of genocidal intent (as opposed to direct evidence of a genocidal plan). A pattern of conduct is one particular circumstance, among others, that may arise in a specific case and point to genocidal intent. In the case at hand, the particular circumstance invoked by Bosnia was a pattern of similar conduct, and the Court was not fully convinced. But the Court never suggested that a pattern of similar conduct is the only “particular circumstance” that can ever convincingly show genocidal intent across multiple incidents.
In Croatia v. Serbia, the Court “examined and assessed the totality of the evidence advanced by Croatia.” For its part, Croatia “advanced a single claim alleging a pattern of conduct increasing in intensity” over time, involving acts of violence in towns and villages across multiple regions, from which “the only reasonable conclusion to be drawn is an intent on the part of the Serbian authorities to destroy in part the Croat group.” As before, a pattern of conduct was invoked to link together specific incidents and establish that “Serbian authorities” or “Serb leaders” must have directed the conduct which, given its large scale and targeted nature, must have been intended to destroy a substantial part of the group.
The Court loosely defined a pattern of conduct as “a consistent series of acts carried out over a specific period of time.” It found “a pattern of conduct that consisted … in widespread attacks by the JNA [Yugoslav People’s Army] and Serb forces on localities with Croat populations in various regions of Croatia, according to a generally similar modus operandi.” Turning to whether this pattern of conduct revealed genocidal intent, the Court recognized the importance of:
the scale and allegedly systematic nature of the attacks, the fact that those attacks are said to have caused casualties and damage far in excess of what was justified by military necessity, the specific targeting of Croats and the nature, extent and degree of the injuries caused to the Croat population.
The Court focused on “the context in which those acts were committed and the opportunity which the JNA and Serb forces had of destroying the Croat population.” Ultimately, the Court found that all the evidence taken together was consistent with an intent to displace or collectively punish the group without destroying it.
In its counter-claim, Serbia alleged that “the acts and statements of the Croatian authorities taken as a whole, before, during and immediately after Operation Storm manifest a consistent pattern of conduct which can only show that those authorities were animated by a desire to destroy, in whole or in part, the group of Serbs living in Croatia.” Here, too, a consistent pattern of conduct was alleged to show that different incidents were motivated by a common intention of higher authorities. Ultimately, the Court found that the alleged acts which were proven “were not committed on a scale such that they could only point to the existence of a genocidal intent.”
Both cases reflect the same core idea: patterns of similar conduct may indicate that the conduct was directed by higher authorities or otherwise coordinated, and therefore reflects a common intention. Once a common intention is found, an evaluation of genocidal intent may consider other evidence, including the scope and severity of the conduct as well as statements other than official plans. The ultimate question remains whether genocidal intent has been convincingly shown by reference to the “particular circumstances” of the case.
The Court never suggested that direction, coordination, or common intention must be established by a pattern of similar conduct, rather than by witness testimony, authenticated records, or public statements other than official plans. Such evidence may be sufficient with respect to specific incidents, like the Srebrenica massacre, and they may be sufficient with respect to multiple incidents as well. In the case of an advanced military, with hierarchical command and control, direction or coordination reflecting a common intention may be more readily inferred (if not presumed). If the conduct involves large-scale atrocities systematically targeting members of a group, then a strong inference of genocidal intent will naturally arise, and the space for other reasonable inferences will naturally diminish.
Recall the Court’s statement that its standard of inference is “in substance identical” to the standard applied by the International Criminal Tribunal for the former Yugoslavia (ICTY) in the Tolimir case. As the ICTY explained in the very next sentence, genocidal intent may be inferred from a number of factors, including
the general context, the perpetration of other culpable acts systematically directed against the same group, the scale of atrocities, the systematic targeting of victims on account of their membership in a particular group, or the repetition of destructive and discriminatory acts. The existence of a plan or policy, a perpetrator’s display of his intent through public speeches or meetings with others may also support an inference that the perpetrator had formed the requisite specific intent.
Only one of these illustrative factors involves a pattern of similar conduct, namely “the repetition of destructive and discriminatory acts.” Pattern-of-conduct evidence is an important factor, but it is one factor among others, with a specific but limited function.
As an aside, the Court’s treatment of pattern-of-conduct evidence should not be confused with the substantive requirements of the International Criminal Court (ICC)’s Elements of Crimes. The Elements provide that an individual commits genocide only if they commit a predicate act (such as killing a group member) “in the context of a manifest pattern of similar conduct directed against [a] group” or their conduct “could itself effect” the total or partial destruction of a group. The ICJ did not refer to the Elements in Bosnia v. Serbia. In Croatia v. Serbia, the ICJ mentioned in passing that the Elements are concerned with a related but distinct problem, namely when genocide can be committed by “a single individual or a small number of individuals.” In contrast, the ICJ looks to pattern-of-conduct evidence as one way (though not the only way) to infer that conduct was directed by higher authorities, or otherwise coordinated. Once direction or coordination is proven (and however it is proven), the ICJ looks to the scale of the conduct, potentially including whether it could itself effect the total or partial destruction of a group, as one way (though, again, not the only way) to infer genocidal intent.
The Confusion
In their joint declaration of intervention, Canada, Denmark, France, Germany, the Netherlands, and the United Kingdom observed that
the Court stated that the “only reasonable inference” test should be used with respect to drawing an inference of specific intent from a “pattern” of conduct only. This cannot be the threshold of the test when other methods of inference are also present, such as when examining the scope and severity of a perpetrator’s conduct to evidence specific intent.
These States appear concerned that the Court will ignore the scope and severity of a perpetrator’s conduct when inferring their intent. This concern is misplaced. In its prior cases, the Court identified a pattern of conduct only to infer direction by higher authorities reflecting a common intent. To infer genocidal intent, the Court also examined the scope and severity of the conduct.
Alternatively, these States may be concerned that the Court will treat a pattern of similar conduct as a necessary prerequisite for a finding of genocidal intent, when the scope and severity of the conduct may suffice in a given case. For example, if an advanced military engages in large-scale, systematic violence targeting an ethnic group, then the scope and severity of the conduct alone may raise a strong inference that the violence must have been directed by higher authorities, or otherwise coordinated, and must reflect a common intent to destroy a substantial part of the group. While such large-scale violence may include patterns of similar conduct, it is the scope and severity of the conduct, rather than its similarity as such, that may raise a strong inference of common and genocidal intent. All of this is consistent with the Court’s basic point that genocidal intent must be “convincingly shown by reference to particular circumstances,” of which a pattern of similar conduct is only one.
In its declaration of intervention, the Democratic Republic of Congo (DRC) writes that the Court’s “only reasonable inference” test “applies when the intent is inferred from particular circumstances or a pattern of conduct and does not concern cases where express statements provide evidence of genocidal intent.” In fact, the Court has examined express statements alongside patterns of conduct when applying the “only reasonable inference” test (Croatia v. Serbia, paras. 422, 438). While a pattern of similar conduct may establish the existence of a common intent behind multiple incidents, express statements may establish the (genocidal or non-genocidal) content of that common intent. In principle, express statements could establish genocidal intent when viewed alongside the scope and severity of conduct, whether or not the conduct falls into patterns of similar acts.
Finally, in its declaration of intervention, Ireland interprets patterns of conduct very broadly, to include “a pattern of widespread and systematic violence against the protected group” as well as inciting statements and discriminatory practices. According to Ireland:
Instances of the acts enumerated in Article II (a)-(e) [of the Genocide Convention], together with circumstantial evidence such as statements or incitement by State organs, or persons or entities empowered to exercise governmental authority, may form a pattern of conduct from which the existence of a general policy, plan or campaign may reasonably be inferred.
Regular public statements made on behalf of State organs, or by persons or entities empowered to exercise governmental authority, that denigrate the protected group, as such, or that incite hatred or fear of it should, in Ireland’s view, be considered as forming a pattern of conduct from which reasonable inferences can be drawn for the purposes of the Convention.
Likewise, measures that systematically discriminate against or persecute the group should also form part of a pattern of conduct to be assessed.
As a practical matter, Ireland’s position gets to the right results by the shortest route. The ultimate question before the Court is whether it is fully convinced by all the evidence taken together or, in other words, whether genocidal intent has been convincingly shown by reference to the particular circumstances of the case. By treating actions, statements, and context as part and parcel of a “pattern of conduct,” Ireland’s position arguably just cuts to the chase and minimizes the risk that the Court will be distracted from a full consideration of the evidence. At the same time, Ireland’s position risks losing sight of the original point that a pattern of similar conduct committed across multiple incidents is one basis, though not the only basis, from which to infer direction or coordination reflecting a common intent. The better approach remains to give pattern-of-conduct evidence a specific, significant, but limited role in the evaluation of all the evidence before the Court.
Conclusion
Genocidal intent may be inferred from a pattern of similar conduct, or from other evidence of direction or coordination reflecting a common intent, whose genocidal character may be inferred from the scope and severity of conduct, from statements other than official plans, or from overlapping patterns of conduct (such as killing and sexual violence). The Court’s standard of proof is high but it can be met in a variety of ways. Whatever the Court decides in Gambia v. Myanmar, it should decide based on all the evidence taken together and the particular circumstances of the case before it.
