Detention Under the NDAA and the Limits of Analogy

Detention Under the NDAA and the Limits of Analogy

My thanks to Marty and Steve for their fascinating and insightful posts (here and here) on the NDAA.  I have many thoughts about the Act, but I want to focus here on the idea that U.S. courts can and should analogize to detention in international armed conflict in order to determine what it means for a person to have “substantially supported al-Qaeda, the Taliban, or associated forces that are engaged in hostilities against the United States or its coalition partners.”  I note at the outset that Marty and Steve never mention international human rights law (IHRL), thus accepting — at least implicitly — two very controversial assumptions.  The first is that the United States is engaged in a global non-international armed conflict (NIAC) with al-Qaeda, such that the detention of any member of al-Qaeda anywhere in the world (except U.S. citizens and lawful resident aliens apprehended in the U.S.) is, in fact, governed by the laws of war.  I have addressed the problems with that position ad nauseum (see, e.g., this post) and won’t repeat them here.  The second assumption is that, in situations of actual armed conflict, international humanitarian law (IHL) completely displaces IHRL with regard to detention authority.  That is a very controversial position, particularly in the context of NIAC, given that the only applicable IHL in NIAC is Common Article 3, which regulates detention but does not authorize it.  A strong case can thus be made that it is inappropriate to simply analogize to detention in IAC to determine the limits of detention in NIAC; the applicable rules may well come from IHRL — the concept of security detention, in particular — instead.  (For a discussion of the interplay between IHL and IHRL in NIAC, see this Chatham House report.)

Let’s put those issues aside, however, and focus on what it would mean to determine the limits of “substantial support” by reference to the rules of IHL that apply in IAC.  The first thing that needs to be emphasized is any such analogy is inherently speculative, because there is no concept of “substantial support” in international armed conflict.  There are simply prisoners of war (POWs) and civilians, whose detention is governed by different Geneva Conventions — GC III for POWs, GC IV for civilians.  Thus, as soon as Marty and Steve endorse (I think correctly) the DOJ position “that those who provide unwitting or insignificant support to the organizations identified in the AUMF are not subject to the AUMF detention authority,” we are already pushing the limits of analogy.  The only way to make sense of that statement is to assume (1) that we are talking about civilian detention under GC IV, because anyone who falls under GC III (members of the regular armed forces, volunteer corps, militias, etc.) is detainable at any time for any reason, even if they only thing he or she has ever done to “support” the armed forces is play the trumpet in an army band; and (2) that, under Article 42 of GC IV, unwitting or insignificant support can never make it “absolutely necessary” to detain a person for reasons of state security.  The analogy may still work (I’m inclined to think so), but we can see that even the most basic limit on detention under the “substantial support” requirement is anything but self-evident when analogized to IAC.  And we can see that the appropriate analogy is to GC IV, not GC III — a critical point I’ll return to below.

Marty and Steve’s next two analogies, however, are far more more problematic:

[T]here likely are significant detention limits with respect to persons who provide medical support to enemy forces while “permanently and exclusively engaged as a medic,” see Warafi (discussed above), since such limits traditionally apply to such persons (indeed, even if they are part of an enemy force) in an international armed conflict, under article 24 of the First Geneva Convention and article 33 of the Third Geneva Convention.  On the other hand, perhaps substantial supporters of enemy forces who are apprehended while accompanying such forces can be detained on roughly the same terms as the forces themselves, just as they can be in an international conflict.  Cf. Third Geneva Convention, art. 4(4).

I would be perfectly happy to determine the detainability of medics by analogizing to Article 33 of GC III.  Under Article 33, medics and religious personnel can be detained, but they are entitled to treatment far better than even POWs receive.  In particular they must “be granted all facilities necessary to provide for the medical care of, and religious ministration to prisoners of war.” Does anyone seriously believe that the U.S. would follow the requirements of Article 33 if a medic or a mullah was detained at Guantanamo Bay?  It seems far more likely that the U.S. would simply claim the benefits of Article 33 (detention authority) while disavowing its obligations (continuation of the detainee’s professional function).

And therein lies the fundamental problem with determining detention authority under the NDAA’s “substantial support” prong by analogizing to IAC: the U.S. will always do so selectively, invoking the Geneva Conventions to justify detention but not providing detainees with their rights under the Conventions.  That is even more clear with regard to “persons who accompany the armed forces,” whom Marty and Steve suggest could be detained by analogy to Article 4(4) of GC III.  It is true that such individuals can be detained, but they are entitled to be treated as POWs, just like members of the regular armed forces, volunteer corps, and militias.  Moreover, such individuals cannot be deprived of their POW status for not complying with the conditions of privileged combatancy in Article 4(2) — such as wearing a fixed and distinctive sign — because those conditions do not apply to them.  But again, does anyone believe that the U.S. would treat “persons who accompany the armed forces” as POWs upon capture?

Let me be clear: I am not arguing that the U.S. should recognize POW status in NIAC.  My point is simply that, because POW status does not exist in NIAC, it would be inappropriate for U.S. courts to determine the limits of “substantial support” by mechanically — and selectively — analogizing to GC III’s rules governing POWs.  If the analogy to IAC is at all appropriate, the analogy should be to GC IV, governing the detention and treatment of civilians — a possibility that Marty and Steve acknowledge, citing Ryan Goodman’s excellent article on the subject.  (The one exception might be individuals who assume a continuous combat function in an organized armed group, for whom the appropriate analogy could be GC III.  But such individuals would not be detained for being “part of” al-Qaeda, not for “substantially supporting” it.)  If U.S. courts do analogize to GC IV, however, they must not give the government all of the Convention’s detention authority while not imposing any of its concomitant detention obligations.  On the contrary, if courts are going to use GC IV to ground the detention of individuals who “substantially support” al-Qaeda, the individuals thereby detained must be given all of the rights to which civilians are entitled to under that Convention.

Readers may think that I’m splitting hairs here.  What does it matter if U.S. courts analogize to GC III instead of GC IV?  Actually, it matters a great deal — particularly concerning the possibility of indefinite detention.  Detention “without trial until the end of hostilities,” the centerpiece of the NDAA, is a GC III standard; GC IV is much more restrictive, befitting the detainee’s status as a civilian:

Art. 132. Each interned person shall be released by the Detaining Power as soon as the reasons which necessitated his internment no longer exist.

The Parties to the conflict shall, moreover, endeavour during the course of hostilities, to conclude agreements for the release, the repatriation, the return to places of residence or the accommodation in a neutral country of certain classes of internees, in particular children, pregnant women and mothers with infants and young children, wounded and sick, and internees who have been detained for a long time.

Article 132 is critically important to the debate between David Cole and Marty and Steve, because it means that Marty and Steve are too quick to dismiss David’s argument that the NDAA’s “indefinite detention” provisions are not necessarily consistent with the laws of war.  Even if we grant that the appropriate analogy is to IAC, as opposed to IHRL or the interplay of IHL and IHRL, the rules of detention for those who “substantially support” al-Qaeda or its associated forces must be derived from GC IV, not GC III.  And GC IV does not permit indefinite detention in the manner contemplated by the NDAA.

Print Friendly, PDF & Email
Foreign Relations Law, International Criminal Law, International Human Rights Law, National Security Law
Notify of
Mark Erickson

It is worth repeating arguments ad naseum if they are continually ignored.

“the U.S. will always do so selectively, invoking the Geneva Conventions to justify detention but not providing detainees with their rights under the Conventions”

Creating the new category of “unlawful enemy combatants” is at the heart of the dispute, but this fundamental problem is papered over in all the justifications of indefinite – but not as a POW – detention.

PS was the previous post open for comments for a half hour? So brave.

Patrick S. O'Donnell

[ad nauseam]


Response… I welcome the discussion but particularly given Mr. Lederman’s previous work in the Obama Administration, the lack of comments (cut off after a somewhat critical one, though not saying there is a connection) does send a bad signal.

Back in the day, Mr. Lederman did extended blogging on moves during the Bush Administration over at Balkinization and there were extended comment discussion on them.  This is ideal.

The major criticism, if I might add, of this legislation was that American citizens could be detained indefinitely. My understanding, though it seems to be buried somewhat since only the Gitmo provisions and a matter involving foreigners are included in the “bad” portion of the discussion by Lederman, is that we shouldn’t really worry too much since it is Obama policy not to do something like that.

The concern is for rights, not at will privileges. 

Anyway, I appreciate the discussions, even those w/o comments allowed.

Benjamin Davis
Benjamin Davis

Cutting off comments as a control mechanism is something done at (total control of whom can post) and selectively at  The narrowness of what is apparently considered acceptable criticism is another problematic reality we are confronted with in the Bush and Obama Executive eras.  We should resist that knee jerk reaction of subordination of critics like me who have no interest in positions in government. The National Security law space in the US is SO SO SO narrowly tailored as to who is allowed to participate, who is listened to, and who is considered an “authority.”  It is stultifying for the ordinary citizens out there.  And when we confront them as I did here and as I have done with Meghan Sullivan, Jack Goldsmith, John Bellinger, John Yoo and others along the way, you always come up with this wall of “cone of silence.” In my experience, it is in that silence that all of the troubling things happen.  That is why there is a need to speak out and call out our leaders and challenge challenge challenge every assumption on which their work is based when it tends to decrease our liberty as ordinary American citizens. It… Read more »


Response… Yes, worth repeating (1) that under international law the U.S. cannot be at war with al Qaeda as such, and (2), yes, human rights law applies globally and in all social contexts, especially in time of war.  E.g., 43 Columbia J. Transnat’l L. 811, 820-23 & n.35 (2005), available at ; 43 Valp. U. L. Rev. 1535, 1535-37 (2009), available at It was one of the false claims of the Bush-Cheney “program” and “common plan” regarding attempts to justify “tough” and “coercive” interrogation and secret detention that human rights law did not apply during war — all patent violations of human rights law, the CAT, the laws of war, etc.  Unfortunately, the Obama Administration repeats the nonsense that the laws of war are a lex specialis that can supplant human rights law (what treaty have you ever read that contains the phrase lex specialis?  what limitation of a member-state’s obligation under art. 56 of the U.N. Charter is stated in the Charter regarding the member-state’s duty to promote “universal respect for, and observance of,” human rights?).  See Yet, of course, whether a person has a particular human right might depend on whether the person is in the actual power… Read more »


[…] who wants to understand why Lawfare does not take comments need only take a brief look at this comment thread over at Opinio Juris blasting Lawfare–and others–for not taking comments. As the old […]



You note that POW status does not exist under NIAC.  But, taking the example of Afghanistan, I assume that the initial invasion was an IAC.  The Taliban still didn’t get POW status, even though the US forces in Afghanistan were preparing to hold Article 5 hearings when they were stopped.  Something similar happened at Gitmo, where people were sacked for thinking about Geneva.  This suggests it was only a policy decision that prevented the POW status for the Taliban.

This scenario – an IAC that develops into an NIAC – if that is what happened – may well reoccur.  It will always be the goal of the US to deny that there is a state enemy, in order to forestall POW rights and other Geneva III rules. 

No one seems to be looking at how the NDAA detention provisions may interact with or affect future IAC detentions that the US government characterises as NIAC. 

Is Congress enabling an alteration of the Geneva rules?