First Strikes & NIAC: Thoughts on the Haque/Horowitz Debate

I have been following with great interest the debate at Just Security between Adil Haque and Jonathan Horowitz over whether the existence of a non-international conflict (NIAC) exists the moment a state launches a “first strike” at an organized armed group or whether hostilities of a certain intensity between the two are required. Adil takes the former position (see here, here, and here); Jonathan takes the latter one (see here and here).

Though Adil’s posts exhibit his typical brilliance, my sympathies lie with Jonathan. To begin with, as a matter of the lex lata, I don’t think the argument is even close: the Tadic test, which requires both organization on the part of the armed group and adequately intense hostilities, has overwhelming support from states. After all, the test is based squarely on Art. 1(2) of Additional Protocol II, ratified by 168 states, which provides that the “Protocol shall not apply to situations of internal disturbances and tensions, such as riots, isolated and sporadic acts of violence and other acts of a similar nature, as not being armed conflicts.” And, of course, as the ICRC notes in its new commentary on the First Geneva Convention, the AP II standard is used by a number of more recent conventions that apply to all NIACs — Common Article 3 or AP II — such as the Rome Statute (1998), the Second Protocol to the Hague Convention for the Protection of Cultural Property (1999), and the Convention on Certain Conventional Weapons (2001).

Adil, it is worth noting, has a different interpretation of AP II, one that does not require intensity:

In my view, if an organized armed group has the capacity to sustain military operations then any military operation by or against that group should be constrained by the law of armed conflict. The organization and capacity of the group is sufficient to distinguish military operations by or against the group from “internal disturbances and tensions, such as riots, isolated and sporadic acts of violence and other acts of a similar nature.”

This is a difficult position to defend. The text of Art. 1(2) of AP II clearly contemplates actual hostilities, not a single act by government forces. What could be a more “sporadic act of violence” than a single act that does not meet with a response from the targeted group and may never be repeated by the government? More importantly, despite some stray practice cited by Michael J. Adams and Ryan Goodman in this post, states have simply never interpreted the AP II standard to require only organization.

Even more problematic, though, is Adil’s argument that the “object and purpose” of IHL counsels against conditioning the application of IHL on adequately intense hostilities:

In my view, we should interpret both the substantive rules of IHL and the conditions for the application of IHL in light of the object and purpose of IHL. The primary object and purpose of IHL is to protect human beings against dangers arising from military operations. Accordingly, IHL should apply to all such military operations. To postpone the application of IHL until a first strike triggers an armed response, or until military operations reach a high level of intensity, would be inconsistent with the object and purpose of IHL.

As regular readers know, whenever I see arguments based on the supposed “object and purpose” of a treaty, I reach for my pen. All too often, such arguments simply use object and purpose to justify interpreting a treaty in a manner that specifically contradicts the intention of the states that drafted and concluded it. And unfortunately I think that is what Adil does here. He defends applying IHL to first strikes by claiming that the “object and purpose” of IHL is “to protect human beings against dangers arising from military operations” (emphasis mine). But that is misleading: the object and purpose of IHL is to protect human beings against dangers arising from military operations in armed conflict. If there is no armed conflict, IHL has nothing to say about the danger of military operations — because IHL doesn’t apply. And as discussed above, states have always insisted that a first-strike military operation is not enough to create an armed conflict — IHL applies only once there are adequately intense hostilities between government forces and the organized armed group.

Adil is free, of course, to normatively argue that IHL should apply to first strikes in NIAC because doing so would better protect human beings. I would disagree, but the claim is coherent and deserving of discussion. What he can’t do is base that claim on the object and purpose of IHL, because that would be to use an object and purpose that only applies within armed conflict to justify changing the definition of armed conflict itself. The definition of when IHL applies cannot be determined by reference to what the goals of IHL are once it applies. That definition has to be sought outside of the IHL system — and again, it is clear that states do not want IHL to apply to first-strike military operations against organized armed groups.

There is, however, an even deeper problem with Adil’s argument that the need to protect human beings from military operations counsels a definition of NIAC that does not require adequately intense hostilities: if that is true, there is also no reason why the application of IHL should require armed groups to be organized. All of Adil’s arguments against the intensity requirement apply equally to the organization requirement. If we need to protect human beings from the dangers of first-strike military operations by states against organized armed groups, surely we also need to protect them from the dangers of first-strike military operations by states against unorganized armed groups. After all, Adil’s central argument is that the inherent danger of military operations means that IHL should apply to a first-strike regardless of whether that military operation leads to any kind of hostilities.

I see no convincing response to this criticism. It is tempting to argue that the organization requirement is important because a first-strike military operation against an organized armed group is much more likely to lead to actual hostilities than a first-strike military operation against an unorganized armed group. But Adil rejects the idea that hostilities are relevant to the application of IHL. He believes IHL should apply even if a first-strike military operation meets with no response whatsoever.

Another potential response would be to argue that first-strike military operations against organized armed groups pose greater dangers for innocent civilians than first-strike military operations against unorganized armed groups. But that would be a difference of degree, not of kind — and thus far from a convincing basis for applying IHL to the former and not the latter. I’m also not sure whether the claim is even empirically sound. It is at least equally plausible to assume that states are more willing to use military force against unorganized individuals whom they can assume will not fight back (or will not fight back effectively) than against an organized armed group with the capacity to respond to a first strike with military operations of its own.

Adil’s desire to protect human beings from the dangers of military operations is laudable, but his claim that IHL should apply to first strikes against organized armed groups cannot be sustained. Not only have states insisted that IHL applies only to hostilities that reach a certain level of intensity, the idea that protecting individuals from the danger of military operations requires eliminating the intensity requirement is underinclusive. Those dangers exist for all military operations, even those against unorganized armed groups. So the only consistent — if still objectionable — position is that IHL applies to any military operation launched by a state, regardless of its object. I’m curious whether Adil would be willing to take that position.