Al Hassan Symposium – Towards the Acceptance of the Aggregated Violence Approach?
Editors’ note: This post is part of a joint symposium hosted by the Armed Groups and International Law and Articles of War blogs. The symposium addresses the ICC’s judgment in the Al Hassan case. The introductory post is available here.
On June 26, 2024, Trial Chamber X of the International Criminal Court (ICC) delivered its judgment in the case against Al Hassan, a former member of the Islamist group Ansar Dine and de facto chief of the Islamic Police in Timbuktu (Mali). The Chamber convicted Al Hassan of several war crimes and crimes against humanity committed between April 2012 and January 2013, during the occupation of Timbuktu by Ansar Dine and Al-Qaeda in the Islamic Maghreb (AQIM).
The significance of the Al Hassan case has been addressed by several authors, notably with regard to questions on rebel governance (here, here, and here), gendered crime, complicity in torture, and the crime of sentencing or execution without due process, among others. Crucially, this case may also mark a new chapter in the classification of armed conflicts saga. In a cryptic sentence, integrated by a convoluted and relatively vague footnote, the judgment seems to have abandoned the traditional fragmented approach to classify the situation of violence in Mali in 2012-2013.
The fragmented approach represents the traditional method to classify non-international armed conflicts (NIACs). This method focuses on evaluating whether the required thresholds of intensity and organization are met in each distinct bilateral relationship between armed groups and/or State forces. Accordingly, conflicts involving multiple parties are disaggregated into individual relationships (e.g., State A vs. armed group X, armed group X vs. armed group Y) and each relationship is assessed separately for the intensity of violence and the organizational capacity of the parties involved.
After rejecting the Defence argument that individual confrontations taken in isolation “were not ‘intense’ enough by themselves” (para. 1266), Trial Chamber X added the following footnote:
The Chamber finds unconvincing the Defence’s argument that it is necessary in the present case to assess each bilateral situation on its own (the intensity between each non-state armed group and the Malian army). … In this regard, the Chamber observes that the reference cited by the Defence envisages an assessment of the aggregated military actions carried out between all the non-state armed groups fighting together and their common enemy, ‘rather than requiring that each bilateral relationship of violence meets the criterion on its own’. … In any case, the evidence makes it clear in the present case that Ansar Dine, AQIM, the MUJAO (and the MNLA to a certain extent and at a certain point in time) undertook military actions in a coordinated manner (n. 4064).
While this footnote alludes to abandoning the traditional fragmented approach to classification, the Court does not seem to provide any clear indications as to the legal basis for this decision. This post explores the scholarly debate addressing classification of armed violence in international humanitarian law (IHL) in light of the judgment in the Al Hassan case.
The Fragmented Approach
The fragmented approach has received vast scholarly support (see e.g. here, here, and here) and has attained the status of a foundational framework for conflict classification. Nevertheless, it presents some limitations in addressing the realities of modern, complex conflicts. Any practitioner or scholar tasked with classifying conflicts in situations of overlapping and decentralized violence has had to contend with the inadequacy of this approach in capturing the interconnected dynamics of contemporary hostilities. For instance, applying the fragmented approach to the situation in Myanmar would be particularly challenging due to the sheer number of armed groups involved, and the challenge of attributing specific acts of violence to particular armed non-State actors (ANSA).
Moreover, many of these groups frequently form and dissolve coalitions, further complicating efforts to identify stable, organized parties to the conflict. This dynamic and fluid landscape makes it challenging to apply the fragmented approach, potentially leading to the conclusion that there is no armed conflict in certain areas which are clearly affected by NIACs.
These challenges have led to proposed alternative methodologies, which have emerged from those involved in the practical day-to-day task of conflict classification. Under the fragmented approach, many complex situations would fail to meet the threshold for a NIAC due to the disaggregation of violence among multiple actors. Accordingly, a number of authors have proposed to aggregate violence in complex scenarios, as outlined below. These authors posit that aggregation better reflects the cumulative impact of violence across loosely coordinated groups, offering a more accurate and operationally relevant framework for conflict classification.
The ICRC Position: Aggregating Violence of Coalitions
In its 2019 Report on International Humanitarian Law and the Challenges of Contemporary Armed Conflicts, the International Committee of the Red Cross (ICRC) affirmed that:
When several organized armed groups display a form of coordination and cooperation, it might be more realistic to examine the intensity criterion collectively by considering the sum of the military actions carried out by all of them fighting together (p. 51, emphasis added).
In 2021, ICRC legal advisers Jelena Nikolic, Thomas de Saint Maurice, and Tristan Ferraro published a post in which they suggested that, when organized armed groups “are essentially operating in some sort of a coalition, objectively and effectively demonstrating a collective approach of the fighting,” the intensity of violence should be assessed cumulatively.
More recently, the ICRC opinion paper titled “How is the Term ‘Armed Conflict’ Defined in International Humanitarian Law?” addresses the question of aggregation of violence distinguishing between two different scenarios:
1. If there is a preexisting NIAC, and an ANSA (or a State or an international organization) supports one of the parties to the conflict in a substantial way, such that “the support has a direct impact on the opposing party’s ability to carry out its military operations” (p. 16), then the classification exercise should rely on the support-based approach (for a definition of the support-based approach, see here).
2. If there is no preexisting NIAC, and multiple organized armed groups act in coordination as part of a coalition, violence should be assessed for the coalition, not bilaterally for every single group. Indeed, it would be impractical to expect the opposing party to address the resulting high-intensity violence solely through law enforcement measures in case the intensity criterion is not met applying the fragmented approach. Since the support-based approach is outside the scope of our analysis, the following will focus only on this second scenario.
Relying on the existence of coalitions to aggregate violence seems challenging, as it struggles to address the dynamic and fragmented realities of coalitions in contemporary conflicts. The following examples suggest how overlapping coalitions can create fluid dynamics that challenge classification efforts:
– Coalitions of coalitions.In May 2021, the Coordination of Azawad Movements (CMA) and the Platform coalition established the Permanent Strategic Framework for Peace, Security, and Development (CSP-PSD) to unify their efforts in northern Mali. By September 2023, the Platform withdrew from the CSP-PSD due to escalating hostilities between the CMA and the Malian government. Subsequently, the CSP-PSD was renamed the Strategic Framework for the Defense of the People of Azawad (CSP-DPA), with the CMA continuing its operations independently.
– Dissolution of coalitions. Formed in November 2021, the United Front of Ethiopian Federalist and Confederalist Forces (UFEFCF) included groups like the Tigray People’s Liberation Front (TPLF) and the Oromo Liberation Army (OLA). Over time, factions exited: the Afar Revolutionary Democratic Unity Front left in January 2022, the Benishangul People’s Liberation Movement signed a peace agreement and exited in October 2022, and the TPLF ceased supporting subversive activities by November 2022.
– Proliferation of unstable coalitions. In Myanmar, over the past decade there has been a proliferation of coalitions, such as the Northern Alliance, the Brotherhood Alliance, and the Federal Political Negotiation and Consultative Committee (FPNCC), which have formed and dissolved or fragmented over time, with groups frequently shifting allegiances or pursuing independent agendas.
This constant realignment of coalitions and the movement of groups between alliances make it challenging to rely on coalitions for conflict classification, as their instability undermines consistent assessment of organization and intensity thresholds. Accordingly, other scholarship suggests alternative approaches to aggregating violence.
Scholarly Approaches to Aggregating Violence
Coordination or Cooperation
Professor Marten Zwanenburg draws upon the ICRC’s 2019 Report and posits that coordination or cooperation among armed groups can suffice for conflict classification because it reflects the operational realities of many non-international armed conflicts. He contends that where groups share a common purpose, engage in coordinated activities, or cooperate to a degree that creates a unified threat, their actions can collectively meet the intensity threshold required for a NIAC.
Temporal and Geographical Continuum
Professor Jann Kleffner argues that in complex conflict scenarios, such as those in Libya and the Democratic Republic of Congo (DRC), there are strong reasons to adopt a cumulative approach to assessing whether the intensity threshold for a NIAC has been met. This is particularly relevant when multiple organized armed groups are engaged in hostilities within the same geographical area and timeframe, as their combined actions may reflect the required level of intensity, even if the violence between individual groups and the opposing party does not.
Armed Groups Fighting Against a Common Enemy
I have previously argued that “the intensity of violence between organized armed groups and the government or between such groups should be considered in an aggregated fashion when three cumulative conditions are met: namely when a number of organized ANSAs are fighting (1) in the same geographical region; (2) during the same period of time; and (3) against a common enemy.”
The geographical and temporal continuum criteria are crucial to aggregate acts of violence that occur within a specific location and timeframe, distinguishing them from unrelated acts spread across wide areas. While membership in the same coalition could replace the geographical criterion, this raises challenges. Aggregating intensity based solely on coalition membership risks complicating the classification exercise in situations when ANSAs join and leave different coalitions, forming and dissolving alliances.
On the other hand, relying only on a temporal and geographical continuum would risk overclassifying situations of violence. Accordingly, demanding ANSAs fight against a common enemy could serve as a more effective criterion to limit cumulative intensity. This would capture situations where splinter groups or independent actors share a common adversary, ensuring IHL applies consistently to all parties meeting the organizational threshold.
Sliding Scale of Complex Conflict Situations
Lastly, Professor Martha Bradley has put forward a grading scale of complex conflict situations, where the three aforementioned approaches could be applied depending on the context. The first category Professor Bradley defines involves coalitions of organized armed groups engaging in hostilities. Here, an aggregation approach combines the ICRC’s coalition-based model with Professor Kleffner’s geographical and temporal continuum. Coalition members must meet IHL’s organizational threshold, be actively engaged in fighting, and operate within a shared temporal and geographic continuum. For example, in a coalition, only the actions of groups within the same operational area would qualify for cumulative assessment, while remote groups would be assessed bilaterally.
The second category focuses on groups fighting a common enemy without clear coalition membership due to insufficient intelligence. In this case, Professor Bradley argues that my common enemy approach would be more appropriate, inasmuch as it aggregates intensity based on shared adversaries within a geographic and temporal continuum, even without coordination or cooperation. For instance, groups fighting in the same region against the same enemy could qualify for aggregation, while distant or isolated groups would not.
The third category addresses extreme cases like the hostilities in the DRC, where the number of actors, shifting alliances, and widespread violence make individual attribution impossible. In this case, Professor Kleffner’s approach should be preferred, which considers the cumulative impact of hostilities within a defined area and timeframe, ensuring IHL applies without risking overclassification in highly fragmented conflict settings.
Conclusions
Proponents of the fragmented approach even in complex scenarios put forward two main critiques of aggregating violence. First, they posit that aggregating violence is a concept rooted in lex ferenda rather than lex lata, reflecting a normative aspiration rather than a legal reality. These approaches would therefore lack grounding in established legal standards, which currently require each ANSA involved to meet specific criteria for intensity of fighting and degree of organization independently. Second, they claim that aggregating violence for conflict classification purposes would raise concerns about overclassification, potentially resulting in the unwarranted application of IHL to situations that do not meet the required thresholds. This could blur the distinction between armed conflicts and lower-intensity violence, leading to the inappropriate invocation of IHL in situations where IHRL should apply.
Nevertheless, these remarks fail to consider that the reason why aggregated violence approaches have been suggested is to address scenarios where the overall intensity of violence is very high, however, applying the fragmented approach would lead to the paradoxical conclusion that IHL is not applicable. In other words, aggregated violence approaches have been suggested to avoid the risk of under-classification, whereby civilians would be left without the protection granted by IHL.
Furthermore, IHL has evolved through State practice and jurisprudence, with its principles adapting to contemporary realities of conflict. Courts and tribunals have occasionally expanded traditional interpretations of NIAC to address emerging issues (e.g., the International Criminal Tribunal for the former Yugoslavia’s articulation of “protracted armed violence” in the Tadić case). Similarly, aggregating violence could reflect a contextual interpretation responsive to complex realities.
In sum, aggregating violence addresses gaps in the fragmented approach, aligning IHL with the evolving nature of conflicts where decentralization and multi-group dynamics increasingly prevail. This approach ensures a more coherent application of IHL, reducing ambiguities in determining the threshold and classification of armed conflicts. While the Al Hassan case embraced the aggregated violence approach in a footnote, it appears to reflect a broader movement towards the recognition and use of aggregated violence in complex conflict scenarios. This case will likely play a pivotal role in fostering the crystallization of this approach into customary international law, reinforcing its applicability in future conflicts.
***
Chiara Redaelli is a Senior Legal Adviser at Diakonia IHL Centre and visiting professor at the Catholic University of Lille and the University of La Sabana in Bogotà.
Photo credit: Magharebia
