The Ntaganda appeal judgment and the meaning of ‘attack’ in conduct of hostilities war crimes

One of the many interesting issues raised by the ICC’s recent Ntaganda appeal judgment, and the subject of this post, is whether war crimes referring to ‘attacks’ require a nexus to hostilities.

IHL has traditionally distinguished between protecting persons from the conduct of hostilities (Hague law) and protecting persons in the power of a belligerent (Geneva law). This is the difference, for instance, between attacking civilians and killing civilians in occupied territory, or between declaring that quarter will be denied and killing captured prisoners. Since war crimes are generally interpreted by reference to the underlying IHL rules, the distinction between conduct of hostilities and other armed conflict conduct is also reflected in the corresponding war crimes. Thus, for war crimes involving attacks, ‘attack’ has been defined in accordance with its meaning in IHL (e.g., Katanga trial judgment ¶ 798), where it is defined in AP I Art. 49(1) as ‘acts of violence against the adversary whether in offence or in defence’. The restrictive IHL interpretation of ‘attack’ as combat actions (ICRC ¶ 1880), as opposed to the ordinary meaning of violent acts generally, means that attack-related war crimes (like the corresponding IHL rules) can only be committed during the conduct of hostilities. This requirement of nexus to hostilities for attack-related war crimes operates in addition to the general requirement of nexus to the armed conflict for all war crimes.

This restrictive scope of attack-related war crimes was challenged in the Prosecution appeal in the Ntaganda case. The Prosecutor argued that in the context of the Rome Statute Art. 8(2)(e)(iv) war crime of attacking cultural and medical property, ‘attack’ should have a broader meaning, encompassing all acts of violence in an armed conflict regardless of nexus to hostilities. The appeals judgment seems to have endorsed the Prosecutor’s argument, but in doing so, also seems to have extended the broad interpretation of ‘attack’ to war crimes generally.

This broadening of the scope of attack-related war crimes in the Rome Statute raises significant concerns regarding the relationship between ICC war crimes and corresponding IHL rules, and between Hague and Geneva law war crimes in the Rome Statute. Below, I summarise relevant aspects of the appeal and appeal judgment, highlight the broad scope of the judgment, and conclude by discussing my concerns.

The appeal and the judgment

One of the charges against Ntaganda related to ‘ratissage’ operations. Following operations intended to seize towns and displace the Lendu population, UPC/FPLC soldiers under Mr. Ntaganda’s command looted a hospital in Mongbwalu and destroyed a church in Sayo (Ntaganda trial judgment, ¶ 1138). This conduct was charged under Rome Statute Art. 8(2)(e)(iv). The trial chamber dismissed the charge, referring to the narrow definition of ‘attack’ in IHL (¶ 916) and the requirement of nexus to hostilities (¶ 1136), and concluding that ratissage operations could not constitute attacks (¶ 1141-2).

This finding was challenged in the Prosecution appeal on the grounds that ‘attack’ has a special meaning in relation to Art. 8(2)(e)(iv) (and Art. 8(2)(b)(ix)) and is not restricted to the conduct of hostilities (¶ 9). This argument was based on the special protection of cultural property and medical establishments under international law, which extends beyond the conduct of hostilities to the armed conflict more generally. In pursuing this argument, the Prosecutor was building on similar arguments in the Al Mahdi trial judgment (¶ 15) and the Al Hassan confirmation of charges decision (¶ 522).

The appeal judgment rejected the Prosecutor’s appeal by a majority (Judge Ibáñez Carranza dissenting) (¶ 1163), but for a range of reasons which together seem to amount to an endorsement of the underlying argument. Two judges (Judges Morrison and Hofmański) upheld the trial chamber’s interpretation of ‘attack’ and endorsed its applicability to the war crime set out in Art. 8(2)(e)(iv). The other three judges, however, disagreed on this point.

Judge Ibáñez Carranza (her dissenting opinion on this point is set out at ¶ 1165-8 of the appeal judgment) held that ‘attack’ in Art. 8(2)(e)(iv) should be interpreted in accordance with its ordinary meaning rather than its meaning under IHL (¶ 1166). She based this conclusion on the objective of Art. 8(2)(e)(iv) (preventing harm to protected objects), the impunity-ending objective of the Rome Statute, and the reference to the ‘established framework of international law’ in Art. 8(2)(e) (the broad scope of protection of cultural property in armed conflict) (para 1167). She also argued that the impugned acts of destruction were substantively and temporally linked to the preceding combat actions, and should be construed together as a method of warfare directed against the Lendu populations of the attacked towns (¶ 1166, 1168).  

Judge Balungi Bossa agreed with Judge Ibáñez Carranza on the importance of interpreting the Rome Statute so as to avoid impunity gaps (¶ 16), and on the absence of a substantive or temporal distinction between the combat actions and the ensuing violence (¶ 3-10). She rejected the prosecutor’s appeal because she considered that the impugned conduct should more appropriately have been charged under Art. 8(2)(e)(xii) (destroying or seizing the acts of the adversary in the absence of military necessity) (¶ 12-5). But she expressly rejected a narrow interpretation of ‘attack’ (¶ 18).

Judge Eboe-Osuji joined the other two judges in preferring the ordinary meaning of ‘attack’ unencumbered by a requirement of nexus to hostilities (¶ 110-1), and in denying a substantive or temporal distinction between the combat actions and subsequent destruction (¶ 127-32). Additionally, he saw no reason for the term ‘attack’ to have different meanings in relation to crimes against humanity under Art. 7 and war crimes under Art. 8 (¶ 112-21). Like Judge Balungi Bossa, he agreed that this conduct should more appropriately have been charged under Art. 8(2)(e)(xii), and on this basis rejected the Prosecutor’s appeal (¶ 136). 

The broad scope of the appeal judgment

In effect, three out of five judges accepted the Prosecution’s argument to reject the narrow IHL interpretation of ‘attack’ in favour of a broader, ordinary meaning. There is much that could be said about this conclusion and the reasoning that produced it. For now, I want to focus on the astonishing breadth of the opinions issued by Judges Ibáñez Carranza, Balungi Bossa and Eboe-Osuji.

The Prosecutor’s appeal was limited to Art. 8(2)(e)(iv), as were the earlier pronouncements on this issue in the Al Mahdi and Al Hassan cases. The argument was based explicitly on the scope of protection of medical and cultural property and argued for an exceptional, special meaning for ‘attack’ in Art. 8(2)(e)(iv).

Setting aside the merits of this argument (but see, e.g., O’Keefe), the reasoning and conclusions of Judges Ibáñez Carranza, Balungi Bossa and Eboe-Osuji seem to go beyond the narrow scope of the appeal and to extend to all attack-related war crimes. The opinions are framed by reference to Art. 8(2)(e)(iv), but there is little to suggest that the analyses are restricted by that frame. Their reliance on plain text interpretations of ‘attack’ and on substantive and temporal connections between combat actions and ratissage operations is potentially applicable to all war crimes involving ‘attacks’. A restriction of scope to Art. 8(2)(e)(iv) may be inferred from the footnotes of Judge Ibáñez Carranza’s opinion, but not from the substantive text. A close reading of the opinions of Judges Balungi Bossa (¶ 7-9) and Eboe-Osuji (e.g., ¶ 112, 121-2, 130-1) suggests a deliberate effort to expand the meaning of ‘attack’ more generally.

There are two common threads across the expansive interpretations offered in these opinions.

One is the emphasis on the humanitarian and impunity-ending objectives of the ICC. This is used to conceptualise the ICC’s war crime jurisdiction not as a penal enforcement mechanism for selected IHL rules, but instead as an independent source of regulation of armed conflicts.

It is not unheard of that international criminal courts use their war crimes jurisdiction to develop IHL ‘progressively’. The ICTY undeniably did so, and the Ntaganda trial chamber’s conclusion that war crimes could be committed against the defendant’s own forces clearly departed from ‘traditional’ understandings of IHL. But the symbiotic and intertwined nature of IHL and war crimes law demands care in doing so. There is little suggestion of any such caution in the meagre engagement of these opinions with the longstanding jurisprudence supporting a narrow interpretation of ‘attack’.

The other common thread across the opinions of Judges Ibáñez Carranza, Balungi Bossa and Eboe-Osuji is that all three construed the looting of the hospital as part of the attack on Mongbwalu and the destruction of the church as part of the attack on Sayo. They construed the combat actions and the subsequent violence as part of the same method of warfare and denied the possibility of establishing a temporal or substantive line between them.

But it is not clear why this was not possible. Admittedly the formal line between active hostilities and other armed conflict conduct can be difficult to draw, particularly in inter-ethnic conflict. But ICC trial chambers have done so before. The Ntaganda trial chamber examined the same military operation and was able to distinguish between deaths that were relevant to the war crime of murder and those relevant to the war crime of attacking civilians (§ V(C)(4)(a)-(b)), as did the Katanga trial chamber (§ VIII(B)).

There may be cases where the formal distinction between conduct of hostilities and other armed conflict conduct makes it difficult to regulate violent acts by armed groups in territory under their control. But this does not seem to be such a case, particularly given that two of these three judges note not only the possibility, but the preferability, of charging the impugned conduct under Art. 8(2)(e)(xii) (destroying enemy property) rather than as an ‘attack’.

‘Attack’ in Rome Statute war crimes going forward

The broadening of the scope of ‘attack’ across Art. 8 raises significant concerns. It effectively severs the link between Rome Statute war crimes and the underlying IHL rules for attack-related war crimes and creates overlaps between Hague and Geneva law war crimes within the Rome Statute.

Consider the elements of the war crimes of murdering civilians (Arts. 8(2)(a)/(c)(i)) and attacking civilians (Arts. 8(2)(b)/(e)(i)). The former requires a result (causing death) and knowledge of the civilian status of the victim; the latter requires conduct (directing an attack) and intent to attack civilians. If ‘attack’ is not interpreted narrowly in accordance with its definition under IHL, an attack that complies with the rules of distinction and proportionality (it was not directed at civilians but caused proportionate civilian deaths) can still be prosecuted as murder.

Thus, a general broadening of the scope of ‘attack’ across war crimes in the Rome Statute significantly expands international legal regulation of battlefield actions without state consent, while simultaneously adversely impacting defendants’ rights. It can only be hoped that the Prosecutor does not follow Judges Ibáñez Carranza, Balungi Bossa and Eboe-Osuji in this conclusion.

But that still leaves the possibility of a broad interpretation of ‘attack’ at least in relation to Art. 8(2)(e)(iv). A few days before the judgment, the Office of the Prosecutor released a draft policy on cultural heritage which suggested a preference for charging the destruction of cultural property as ‘attacks’ against cultural property (¶ 45), in recognition of the special status of cultural heritage (¶ 42). That preference seems to find endorsement in the 3-2 split of the appeals chamber on this point and seems likely to continue. Thus, if nothing else then it is likely that this judgment effects a partial broadening of the scope of ‘attack’ limited to Art. 8(2)(e)(iv), and its international armed conflict counterpart – Art. 8(2)(b)(ix). 

But that then raises a further question regarding a schism between the protection of cultural and medical property in international armed conflicts under Art. 8(2)(b)(ix) and in non-international armed conflicts under Art. 8(2)(e)(iv). The Prosecutor’s argument drew on the special protection of these objects in international law and applied across international and non-international armed conflicts. But the appeal judgment relies on the difficulty of distinguishing between combat actions and ensuing violence, and on the need to end impunity. That argument may find purchase in the context of non-international armed conflicts where there are gaps in the regulation of armed groups exercising territorial control, but it has no basis in international armed conflicts where there is an extensive law of occupation, supplemented by IHRL. The same humanitarian impulse and teleological revisionism which motivate expansion of the scope of active hostilities in non-international armed conflict, militate against similar expansion in international armed conflict.