In a recent post, I argued that an unexplored and independent legal basis exists for the US practice of targeting “associated forces” of al-Qaeda: the rules governing armed forces “belonging to a party” to a conflict. I contended that this legal ground is stronger than the so-called co-belligerency test which is derived from principles of neutrality and jus ad bellum between states. And I concluded that a narrow reading of “belonging to a party” could undermine humanitarian interests in other contexts.

In a Guest Post, Kevin Jon Heller criticized my analysis. He primarily argued against the methods that I used to apply rules of international armed conflict (article 4 of the Geneva Convention on POWs) to noninternational armed conflict (NIAC). I want to pick up on that thread here, and elaborate on some of the policy implications.

Although a source of the rule of “belonging to a party” is housed in the POW Convention (article 4(a)(2)), it is imported directly into the First Additional Protocol of the Geneva Conventions by reference  in article (50)(1). The latter provides the rules for defining civilians, e.g., for the purpose of distinction in targeting. These rules of Protocol I are widely considered customary international law in NIACs. Indeed, it is routine for practitioners and commentators to apply article 50 of Protocol I to non-international conflict (see Kevin’s own work here). And, indeed, experts routinely refer to Article 4 of the POW Convention – by custom or analogy – to determine the law governing whether particular classes of individuals (such as a cook accompanying armed forces or financiers) are detainable and targetable in a NIAC.

Although unnecessary, there are other forms of interpretation such as analogical reasoning that might be applied here, as they are applied in ordinary cases of interpreting international law and the law of armed conflict. For example, it is commonplace to reason by analogy from international human rights law to the law of war regime with arguments such as: if states are permitted to take X action in times of peace under IHR, they are surely permitted to take X action in times of war under IHL. And, more to the point for our discussion on associated forces, as I have written elsewhere, it is practically an axiom in the laws of war that if states are permitted to take an action in an international armed conflict (e.g., target particular individuals), states are presumably permitted to do so in a NIAC (Goodman, The Detention of Civilians in Armed Conflict, p. 50).

Finally, in my original post, I noted that the ICRC’s Interpretive Guidance on Direct Participation in Hostilities, after it discusses the “belonging to a party” test, includes multiple references to armed forces belonging to a party in a NIAC (e.g., ICRC Guidance: “Organized armed groups belonging to a non-State party to an armed conflict include both dissident armed forces and other organized armed groups.”). Kevin appears to concede that the belonging to test applies in NIACs, but he suggests the term performs different work (without the ICRC explicitly saying so). It stands to reason that whatever entity is considered the “party” in a NIAC (regardless of whether we define it primarily by its core political or administrative unit as Kevin suggests), the test remains the same: determining which organized armed groups are fighting alongside or on behalf of that entity (party). Here, we know that al-Qaeda central is the purported “party” to the conflict with the United States; the question is which organized armed groups are fighting in such a way that they belong to it. Indeed, it would be curious if, in NIACs, armed groups “belonging to” the state were targetable, but organized armed groups belonging in the very same way to the nonstate party were not. As the ICRC Guidance explains in a related context, such a legal asymmetry “would provide members of [nonstate] groups with a significant operational advantage over members of State armed forces.”

Finally, Heller mistakes my argument about the other policy consequences that may result from narrowly defining the “belonging to” concept. I should have been clearer. My point is not that a narrow construction would affect reciprocity in the conflict with al-Qaeda. My point is about the broader impact on the law-of-war regime in other armed conflicts. If we accept a narrow construction of “belonging to” a party today (in the conflict with al-Qaeda), it can have major repercussions for the scope of POW protections that will apply when that same concept arises in other conflicts. A tendentious argument that tries to narrow the concept for targeting al-Qaeda’s associated forces would, as a consequence, throw into doubt the scope of protections that apply to the associated forces of the United States in a conflict with Syria, Libya, etc.

[As an aside, I believe Heller may have meant to say article 43 of the First Additional Protocol (not the Second) when he wrote: “Or (perhaps more likely) the drafters of the Second Additional Protocol could have included the concept [of ‘belonging to’] in Article 43.” His point was to suggest the drafters omitted the concept altogether. However, the drafters of the First Additional Protocol clearly chose to include the concept through Article 50 instead. Also, Heller explains, at some length, why the drafters of the 1949 Convention wrote 4(a)(2) to deal with detention issues. He never addresses the fact that the drafters of the 1977 Protocol chose to directly reference 4(a)(2) to define civilians for the purpose of targeting.]

Update: Heller suggests in his post (and UN Special Rapporteur Christof Heyns does as well in his post) that my argument implies that Al Qaeda would receive POW status. It does not such thing. Al Qaeda would, under any of these circumstances, fail to meet the basic requirements for POW status under Article 4(a)(2) of the POW Convention. See also the 2009 ICRC Interpretive Guidance on Direct Participation in Hostilities (p. 22) (explaining distinction between post-capture status determination and party-to-a-conflict determination).