A “Broad Consensus” — of Between Two and Four States

by Kevin Jon Heller

Yes, the “unwilling or unable” test marches on. The latest step forward is a Just Security blog post by Kate Martin, the Director of the Center for National Security Studies, that cites absolutely nothing in defense of the test other than another scholar who cites almost nothing in defense of the test. Here is what Martin says in the context of the UK’s recent drone strikes in Syria (emphasis mine):

Some issues raised by the UK Article 51 legal theory are less controversial than others. The US and other states understand customary international law to include the right to use military force in self-defense against armed attacks, and claim the right to use military force under Article 51 outside of an armed conflict. As Lubell has noted, there is support for reading Article 51 as justifying the use of military force against non-state actors. There is broad consensus that there is a right to use military force in self-defense when the host country is unable or unwilling to stop the attack.

And what does Martin offer in support of this “broad consensus”? A link to a blog post at Lawfare by Ashley Deeks, in which Deeks (1) correctly points out that the US and UK both support “unwilling or unable,” (2) claims that “France appears to be prepared to invoke the ‘unwilling or unable’ concept in the Syria context,” and (3) states that Australia is “apparently relying on a collective self-defense of Iraq/unwilling and unable theory.”

So at most there is a “broad consensus” of four states in support of “unwilling or unable.” And perhaps there are only two. That’s quite a consensus.

This isn’t even instant custom. This is custom by scholarly fiat.


24 Responses

  1. Kevin, I wonder, do you realize the meaning of ” consensus ” in customary international law??

    Consensus is not only the spread and numbers of states (valid by itself) but also (and among others):

    The ” jus cogens ” of it , means :

    It is a law of nature, fundamental principal , and indeed, if you would read the article, carefully, professionally, instead of solely preaching on it , you would discover there the words ” inherent right ” here the ” chapeaux ” of the article:

    ” Article 51
    Nothing in the present Charter shall impair the inherent right of individual or collective self-defense …… “

  2. “Consensus”: “majority of opinion,” “general agreement.”

    There are 194 states in the world. Agreement between 2 and 4 of those 194 is not “consensus” that unwilling/unable represents customary international law.

  3. How many countries though explicitly repudiate such a test?? If the answer is none or an equally small group I would think that would be just as important as how many directly endorse unwilling or unable. If international law is defined at the boundaries by things states do not reject I don’t see silence by the majority as any sort of knock against the US/UK position.

  4. I have blogged extensively about states that have rejected the unwilling/unable test. Besides, I know of no scholar — much less state — that agues the customary rule is established by acquiescence. If someone want to make that argument, they should. I would disagree, but at least we could have a discussion. Arguing that four states at most equals “broad consensus” makes discussion pointless.

  5. Kevin , you don’t seem to distinguish , between , legal definition , and linguistic one . consensus or not , can be attributed to subjects or issues , out of the scope of ” jus cogens ” like :

    Drugs , homosexuality , death penalty , but :

    Never for ” jus cogens ” !! Sheer numbers, won’t do here!! Because , this is the very heart of the CIL : natural law , acting due to sense of obligation . And so, no state for example, in its law:

    Would permit , genocide for example . The same for self defense :

    Every state in the world, is granting acquittance ( or alike ) holding a perpetrator, not criminally responsible, when he was acting in pure self defense, this is a: Natural law!! This is the very corp of CIL, notwithstanding, validity of sheer numbers ( In other domains , sheer numbers or acceptance and objections can prevail ) .

    Anyway, the UN charter, prescribe it deliberately, as mentioned in my first comment , by itself a source for CIL .

  6. I suppose two can form a consensus. Ultimately the conflict is arguably a war, in fact, whether or not both sides are nation states. If combatants are harboured anywhere they may be attacked where they are. It isn’t a game.
    As to weapons, I cannot see a difference between a drone and an assegai, except that the first is sent from further away. Yes, control is even further away than the launch site, but what of it?

  7. Fantastic! Regardless of Republican opposition, the Democrats should immediately pass legislation by consensus creating a single-payer national health care system, raising taxes on the rich, funding schools and highways, guaranteeing a living wage, and bailing out underwater mortgages.

  8. We’ve been over this before, in great detail; and so I’ll simply refer interested readers to the comment section of this post:


    As that discussion reveals, the dispute is not really about the “unable or unwilling” test–which is a *constraint* on the use of force (an application of the requirement of necessity under the jus ad bellum, i.e., a condition precedent to the use of force)–but instead on the antecedent question whether there is a right of self-defense against armed attacks by nonstate actors who have taken sanctuary in another state (such as in the Caroline case itself). My view, reflected in that comment section, is that of course the notion of self-defense, incorporated in Article 51, does encompass such cases, and has done so since at least the Caroline–and that virtually no states have ever rejected it. (The principal dispute in that discussion was this was so from approximately 1947-2001; there was much less disagreement about the years before and after that period.)

  9. I, too, won’t go over all this again. Suffice it to say that Marty is (still) wrong to describe unwilling/unable as a constraint on the use of force; the actual issue is whether unattributable attacks by NSAs can be considered an “armed attack” within the meaning of Art. 2(4), the condition precedent to the use of force in self-defense. He is also (still) wrong to categorically assert that no state has ever rejected the idea that armed attacks include attacks by NSAs that are not attributable to a state — the rejection is Art. 2(4) itself, as the drafting history of the UN Charter makes clear. He also (still) simply ignores all of the state practice and ICJ decisions prior to 9/11 reaffirming the attribution requirement, none of which would have made any sense at all if attribution was, as Marty claims, never required in the first place.

  10. I will leave it to readers of those two previous posts to assess whether I’ve ignored anything important. For now, suffice it to say:

    1. Most importantly, Kevin and I now agree that “the actual issue is whether unattributable attacks by NSAs can be considered an “armed attack” within the meaning of Art. 2(4), the condition precedent to the use of force in self-defense.”

    2. I believe there’s also widespread agreement that the answer to this question was “yes” prior to the Charter — see The Caroline.

    3. As the current events in Syria demonstrate–with even the UN Secretary General apparently blessing the attacks on ISIL there–there are many nations and others that agree that nonstate actors’ attacks can justify self-defense (and, as far as I’m aware, virtually no states that have said otherwise).

    4. There was some dispute *among commentators* about this question (but not so much among states) in the years between 1947-2001; but as I believe even Tom Ruys agrees, that dispute appears now to be largely a thing of the past.

    5. I am not “wrong to describe unwilling/unable as a constraint on the use of force.” The rule of necessity is a constraint–a precondition–on the use of force in self-defense; and if the host state is willing and able to stop the threat from the nonstate actor, there is no necessity for the threatened state to use force.

    6. For evidence that states have “rejected the idea that armed attacks include attacks by NSAs that are not attributable to a state” Kevin points to “Art. 2(4) itself.” Well, Article 2(4) speaks for itself–and it includes no such exception.

  11. Sorry, meant to say in Point 6, of course, that Article 2(4) does not speak to the question of the scope of permissible self-defense; that Article 51 is an acknowledged exception to the prohibition in 2(4); and that *Article 51* includes no such distinction for armed attacks by nonstate actors: it recognizes “the inherent right to individual or collective self-defense if an armed attack occurs,” full stop. And that “inherent right” had long included self-defense against nonstate actors who engage in “armed attacks.”

  12. Three issues keep confusing me:
    – the (ab)use of Caroline to support any position on self-defence against non-State actors; Caroline was not about self-defence (se ILC commentaries on necessity), and it is rather doubtful whether one can use a pre-Charter incident as element for the alleged customary character of rule, when IL at that time (pre Kellog-Briand, pre League of Nations) was far more permissive when it comes to using force.
    – the use of the language “unable OR unwilling”; if the territorial State is willing but not able, the victim State should seek to obtain consent. Only when consent is not given the territorial State becomes unwilling. If there is any test at all, it should be unwillingness, not inability a combination of both used as alternatives.
    – Can’t persistent unwillingness be regarded as “substantial involvement” as per Art. 3(g) Def. of Agression, thereby circumventing the more arduous task of positive attribution of the act itself?

  13. Remy,

    Your first point is exactly correct. Marty simply ignores that — and ignores the entire drafting history of Art. 2(4), preferring instead to adopt a naive textualism. As I have pointed out again and again, all post-Charter, pre-9/11 practice makes sense only if one acknowledges — contra Marty — that the drafters of Art. 2(4) understood the concept of “armed attack” to encompass only attacks launched by states or by NSAs whose acts were attributable to a state.

  14. Note also what Marty’s textualism implies: that states must wait until they are actually attacked before acting in self-defense. After all, Art. 51 specifically states that “Nothing in the present Charter shall impair the inherent right of individual or collective self-defence if an armed attack occurs against a Member of the United Nations.” I’m pretty sure that Marty doesn’t believe that Art. 51 doesn’t permit any form of preventative self-defense.

  15. “it is rather doubtful whether one can use a pre-Charter incident as element for the alleged customary character of rule”? really? Art. 51 preserves the “inherent” right to individual or collective self-defense if an armed attack occurs. The scope of that “inherent” right is defined by pre-Charter custom; and I had thought virtually everyone agreed that the Caroline incident was a seminal and important precedent in that respect. (People cite it repeatedly, for example, for its gloss on “necessity”–“instant, overwhelming, and leaving no choice of means,” etc. And when they do, does anyone respond: “Who cares? That’s an nineteenth century case”?) Is there *any* support for the idea that, when they wrote and ratified the Charter, states intended to deny themselves the use of self-defense that the British employed w/r/t the Caroline? That would have been a very significant relinquishment, indeed — surely it would have occasioned some comment and debate.

  16. Ruys says it all in his book: “The ‘expansionists’ for their part would do better to abandon the ‘pre-existing custom’ discourse altogether, including the far-fetched assertion that the 1837 Caroline incident authoritatively explains the contours of a twenty-first century right of self-defence.”

  17. Here’s a fun quote from the 1949 Report of the US Senate Committee on Foreign Relations concerning the North Atlantic Treaty:

    [T]he words ‘armed attack’ clearly do not mean an incident created by irresponsible groups or individuals, but rather an attack by one State upon another. However, if a revolution were aided and abetted by an outside power such assistance might possibly be considered an armed attack.

  18. I should add that one could take the position that the adoption of the UN Charter did not affirmatively resolve the attribution issue. That would still not help Marty’s case, though, because there is no disagreement in the literature — literally none — that post-Charter, pre-9/11 state practice established that attribution was required as a matter of customary international law. (As the ICJ found on multiple occasions.) That customary rule is itself subject to change, of course, and that is the point of paying attention to post-9/11 practice, which has indeed (as I’ve pointed out numerous times) put pressure on the attribution requirement. But one cannot simply disregard the attribution requirement, as Marty does, on the ground that pre-Charter practice did not require attribution. Any such argument fundamentally misunderstands how customary international law is established and changes over time.

  19. Marty: yes, there is no unwilling or unable limitation of the inherent right of self-defense if an armed attack occurs, but there is no requirement that article 2(4) be violated before a state can engage in lawful self-defense measures against a NSA armed attack emanating from another state — especially because article 2(4) only applies re: state actors. The Caroline incident did not involve what would be an article 2(4) violation by the U.S. and there was no expectation evidenced that the UK had no right to use proper measures of self-defense against NSA armed attacks if the U.S. did not attack Canada or there was no imputation of the NSA armed attacks to the U.S., etc. Re: Caroline, etc., see, e.g., http://ssrn.com/abstract=2402414 and http://ssrn.com/abstract=1520717 and regarding Operationalizing Self-Defense, http://ssrn.com/abstract=2459649 These also demonstrate why there is no “unwilling or unable” limitation on the inherent right of self-defense, which you seem to accept.
    Remy: re-read the exchanges during the Caroline incident. It was all about slef-defense against NSA armed attacks (esp. http://ssrn.com/abstract=2402414 ).

  20. Of course the other problem of unable or unwilling is that the side that wants to use force is asserting the authority to make both the decision that they are under armed attack and that the host country is unable or unwilling. Way too easy to start kinetic action. Reminds me of Poland in 1939. Godwin’s law.

  21. Prof. Keller wrote in relevant part” “I know of no scholar — much less state — that agues the customary rule is established by acquiescence. If someone want to make that argument, they should. I would disagree, but at least we could have a discussion.”

    Let’s have that discussion. A sovereign’s consent to a customary international legal norm can be implied. See The Schooner Exchange v. McFaddon, 11 U.S. 116, 136 (1812). A state’s silence regarding its acceptance of a customary rule may preclude a state from challenging another state’s claim. See Thomas Franck & Dennis Sughrue, The International Role of Equity-as-Fairness, 81 Geo. L.J. 563, 568 (1993). This Acquiescence Principle is a general principle of law recognized by civilized nations and operates in conjunction with the Persistent Objector Rule (which, of course, has been recognized by the ICJ, Inter-American Commission on Human Rights, and Vattel). The application of the Acquiescence Rule and Persistent Objector Rule is important to ensure that silent states are not unjustly enriched at the expense of states that have expressly consented to a customary international legal norm. (Of course, the Unjust Enrichment Prohibition Rule also is recognized as a general principle of law.) For a fuller discussion, See Martin, The Constitution as Treaty 68-92 (Cambridge Univ. Press 2007).

    Francisco Forrest Martin

  22. But Francisco, customary international law does not rest on “consent,” express or implied, but on general patterns of opinio juris, which could be expressed or possibly implied in a given context (your comment reminds one of the principle of estoppel). Further the minority claim of immunity from C.I.L. created by the international community runs counter to the nature od C.I.L. and smacks of an unacceptable “consent” theory, at least in part.

  23. The question is not whether there is some magical definition of “armed attack” that somehow makes only states capable of launching one. The question is whether state X is entitled to defend itself against hostile acts undertaken by Y, be it a state or a non-state actor. Hostile acts cause death and destruction, it doesn’t matter who stands behind them. Whether they merit a forcible response is a matter of their organization, length, scope, nature, intensity etc; and whether the response is legal is a matter of necessity and proportionality.

    I think we are too caught up in the intricacies of the “war on terror” to deal with the real question. We reject the unwilling/unable test because it seems tainted with Bush-era reasoning, while in principle, there are definitely situations that would justify transnational self-defense actions.

    Let’s assume that state A shares a border with failed state B. Organized Armed Group C – which is hostile to A – controls the territory between A and B, since B is failing and can’t do anything about it (nor does it wish to, since it is perfectly fine with C harassing A and keeping it in check). C has managed to amass substantial armed forces, and has invaded A. Is there anyone that would seriously claim that A cannot attack C’s bases in B’s territory (assuming peaceful means were exhausted) because C is not a state?

    It’s not about preexisting customs, it’s about common sense. If we don’t want the self-defense debate to become “transcendental nonsense” we have to discuss the effects of the non-state attacks on the alleged victim, and these really do not depend on the political status of the attacker.

Trackbacks and Pingbacks

  1. There are no trackbacks or pingbacks associated with this post at this time.