The Folly of Comparing Al-Awlaki to Admiral Yamamoto (Updated)

by Kevin Jon Heller

It appears the right-wing has settled on a shiny new historical comparison to justify the targeted killing of Anwar al-Awlaki.  Here is Jack Goldsmith in the New York Times:

An attack on an enemy soldier during war is not an assassination. During World War II, the United States targeted and killed Adm. Isoroku Yamamoto, the architect of the Japanese attack on Pearl Harbor.

And here is John Tobin in the American Spectator:

Anwar al-Awlaki was actively recruiting terrorists to attack the U.S. He was, in effect, a battlefield commander, and the operation to kill him was in that sense well-grounded in the laws of war — little different from Operation Vengeance, in which the US military targeted and killed Admiral Isoroku Yamamoto during World War II. If Yamamoto had been an American-born traitor, Operation Vengeance would have been no less legitimate.

It genuinely amazes me that anyone could compare Al-Awlaki to Yamamoto with a straight face.  World War II was an international armed conflict (IAC), while the U.S. war on al-Qaeda is at most a non-international armed conflict (NIAC).  (The correct position is that it is not an armed conflict at all.)  That is a critical distinction, because the targeting rules in IAC and NIAC are completely different.  Admiral Yamamoto was the Commander in Chief of Japan’s Combined Fleet, the prototypical combatant who was targetable at any time by the U.S.  Al-Awlaki was a radical cleric whose targetability depended on whether he assumed a “continuous combat function” in AQAP (in which case he was, like Yamamoto, targetable at any time), or whether he was a civilian who directly participated in hostilities on various occasions (in which case he was targetable only for the duration of his direct participation).  Which is it?  I frankly don’t know — but I do know that determining Al-Awlaki’s targetability is vastly more legally and factually complicated than determining whether it was legal to kill an enemy Admiral in a formally-declared war.

Not so, of course, for Goldsmith and Tobin.  For them, that pesky international-law distinction between international and non-international armed conflict is irrelevant.  (Except, of course, when it comes to things like combatant’s privilege and POW status; the rules of IAC and NIAC are interchangeable only when interchangeability works in the United States’ favor.)  The U.S. once killed a bad guy during World War II, so of course it can kill a bad guy during the war on terror.  What could be more obvious?

UPDATE: Goldsmith responds to my post here, although he doesn’t bother to address the substance of what I wrote — concerning his elision of the distinction between IAC and NIAC and his failure to grapple with the distinction (critical for purposes of targeting in NIAC) between members of organized armed groups and civilians who directly participate in hostilities.  Instead, he simply cites other Americans who believe the analogy is justified (why that’s relevant he never explains) and claims that I somehow admit that the analogy is justified, because I say that Al-Awlaki could be targeted in much the same way as Yamamoto if he had assumed a continuous combat function in AQAP.  Of course, the entire thrust of the post was to point out that Al-Awlaki might also have been a civilian who directly participated in hostilities, in which case he could not be targeted in the same way as Yamamoto, making the analogy between the two deeply misleading.

Notice, also, how Goldsmith selectively quotes my post, ending the block quote with “I frankly don’t know.”  The entire sentence reads, of course, “I frankly don’t know — but I do know that determining Al-Awlaki’s targetability is vastly more legally and factually complicated than determining whether it was legal to kill an enemy Admiral in a formally-declared war.”  Goldsmith thus conveniently cuts off the thesis of my post in order to claim that I somehow agree with the Yamamoto analogy.  As is obvious from the omitted clause, I acknowledged my uncertainty concerning Al-Awlaki’s status precisely to make the point that determining targetability in NIAC is far more difficult than determining targetability in IAC — the primary reason why the Yamamoto analogy is so flawed.  Goldsmith ignores that aspect of my post; I guess, for him, the targeting rules in NIAC are so self-evident and straightforward, and the facts of Al-Awlaki’s case so obvious, that the only possible conclusion is that Al-Awlaki is no different than Yamamoto.  I envy his certainty.

22 Responses

  1. What could be more obvious?
    That if the President does it, it’s not illegal. Duh.

  2. I don’t think many people are terribly concerned about international law. The legal questions that most people are concerned with are those of domestic law.

  3. Jack Goldsmith does say: “The Obama administration has tried to explain the basis for its actions under international law . . . . But its international law arguments are more controversial.”  It seems that for you, that pesky sentence is irrelevant in evaluating JG’s argument, and it does not stop you from claiming that JG misses the obvious.

  4. Whilst I agree that too many commentators have failed to make the IAC/NIAC distinction, I think Al-Awlaki probably did fit into the NIAC criteria for targeting. He hadn’t ‘retired’ from his business of planning and inciting attacks against both civilian and military targets; whether that counts as a continuous combat role or ongoing direct participation, it should surely be one or the other.

  5. It is worth pointing out that SOME who have opposed drone strikes in Yemen have even contended that the Yamamoto shootdown would be illegal under today’s understanding of IHL (see Mary Ellen O’Connell), so there is some value in making it clear that that strike was legal.  More to your point, there is something fundamentally wrong with using the same standard to define the scope of IHL in all NIAC’s.  Tadic factors are fine for internal NIAC’a, but their application to transnational NIAC’s has the perverse effect of creating sanctuaries for groups that are otherwise IHL’s most disfavored (terrorist organizations) because of the way they conduct themselves (targeting civilians, hiding amongst civilians).  Get to a place where there is no internal armed conflict going on and where there is no effective law enforcement (e.g. Somalia, the Sudan, Yemen before the uprising, etc.) and you are effectively immune from attack while you regroup and plan your next operation.  IHL should not be read to allow that.    

  6. Michael,

    Your argument assumes that IHL should govern all attacks on terrorists anywhere.  But why should that be the case?  With very few exceptions, states are perfectly comfortable determining the legality of the use of force against terrorists according to IHRL, not IHL, and it is far from clear whether the restrictions that IHRL places on things like targeted killing would significantly undermine the ability of the U.S. to conduct drone strikes.  (Relatedly, i fail to see how terrorists are IHL’s most disfavored groups, when international sources are generally in complete agreement that sporadic terrorist attacks do not fall within IHL; see Tadic.)  The US opposition to that paradigm, properly understood, is not that IHRL is too limiting, but that human-rights obligations should not apply extraterritorially.  And that is an indefensible position.

    IHL is also not as limiting as you suggest.  To be sure, attacks on terrorists who have no relationship to terrorist groups that are involved in an actual NIAC ( such as Afghanistan, perhaps Somalia)  are subject to IHRL, not IHL.  But terrorists who are members of organized armed groups that are involved in such NIACs are targetable anywhere at any time under IHL, and terrorists who directly participate in those NIACs are targetable anywhere for the duration of the direct participation.  So “safe havens”  exist only for terrorists who engage in sporadic attacks — precisely those attacks that the international community believes should not be addressed through the military model.

  7. In a paper of 500 words or less ompare and contrast the imagined killing of Jack Goldsmith to the actual killing of Al-Awlaki. You may use the Goldsmith-Gonzales-Yoo-Feith analytical style. The paper is due before the use of drones comes to the U.S., or yesterday.

  8. Steve,

    It won’t take 500 words.  Drones will not come to the US.  So many commentators conceive of drones as this unstoppable fantasy weapon that can “rain death from above”.  Factually, any nation with an even moderately capable air defense network will not have to worry about drones.  They can be detected and once detected easily defeated.  They require a constant telemetry link with their controllers, a link that is easily jammed, interrupted or even intercepted.  So if you want to imagine killing Jack Goldsmith I’m afraid you will have to do it with a suicide bomber or perhaps a sniper, but not with a drone.  Although I would be interested in the basis for your apparent belief that Goldsmith and Awlaki occupied the same status under IHL/IHRL. 

  9. Steve,

    And that is precisely the point.  The reason that the U.S. adopts such a permissive legal standard for drone strikes, one that finds little support in existing international law, is that in practice only powerful Western states will ever be able to use them successfully.  If the U.S. could be the victim of drone strikes, the U.S. would be the first in line to proclaim their illegality.

    (The same logic, of course, explains why the U.S. invents war crimes like “murder through the use of perfidy” — such acts are within the capacity of weaker states and terrorist groups, so they must be deemed illegal.)

  10. Kevin,

    Actually no, my argument is not premised on a belief that all terrorists are subject to IHL.  Only those involved in a NIAC.  I believe that members of al Qaeda (and we can argue about whether AQAP qualify or not at some other time, I concede that is not a certainty) are involved in a NIAC with the United States.  I believe that members of Hezbollah were involved in a NIAC with Israel in 2006.  THOSE terrorists, as you say, can be struck anywhere, anytime.  But what is the legal basis for targeting an al Qaeda terrorist that has achieved the status of a CCF “anywhere, anytime”?  It cannot be self-defense because that depends on “imminence” which may or may not be present.  So it must be based upon the application of IHL to them in a geography well removed from “hot” battlefields (like Yemen).  How do you justify that?  My answer is neutrality law.  Yemen must either permit our actions on their territory (as an ally) refuse such actions and expel al Qaeda (maintain neutrality) or refuse strikes and harbor al Qaeda (acting as an enemy, as Afghanistan did in 2001).  This makes far more sense than applying Tadic to Yemen and saying “no war in Yemen, therefore no IHL in Yemen against al Qaeda operating from Yemen” (essentially MEO’s argument). 

    So what legal framework do you use to conclude that terrorists involved in a NIAC may be targeted “anytime, anywhere”?

    P.S.  I say “terrorists” are IHL’s most disfavored group within the context of an established NIAC.  I agree with you that many (perhaps even most) terrorists are not involved in NIAC’s.  But those that are, who routinely target civilians and intentionally blend in with them for cover are strongly disfavored by IHL.  No combatants’ privilege AND forfeited civilian immunity, you can’t get more disfavored than that.  Yet those that claim Tadic determines the scope of IHL in ALL NIAC’s ARE providing such terrorists with safe havens in states where there is no NIAC and no functioning law enforcement. 

  11. Prof. Lewis: Please do “argue about whether AQAP qualify or not” right now, because that is the pertinent question with the Awlaki and Khan killings.
    Also, I left a comment on Lawfare’s FB page, on the update for “A senior administration lawyer involved in national security issues writes in” in regard to your Lawfare guest post, because it didn’t get a FB update.

  12. Followup question: Identify each and every weapons technology and/or technique for which migration from so-called “first world militaries” to so-called “non-first world” militaries is factually impossible. Assume the would-be possessors dream of eternal rewards for his/her/its efforts, have sufficient funds, boundless enthusiasm, fervently believe “God is on his/her/its side” and adamantly refuse to take “No!” for an answer. I will take the answers in a fallout shelter.

  13. Michael,

    The law of neutrality applies only in international armed conflict, whether between states or between a state and an insurgent group that has been recognized as a lawful belligerent.  Does that mean you are prepared to recognize al-Qaeda as a lawful belligerent with the same rights and privileges as the U.S., including combatant’s privilege and POW status?

  14. Michael,

    As for CCF, see p. 206 of Gary Solis’s book, where he explains that individuals who assume a CCF in an organized armed group are targetable anywhere and anytime for as long as they continue in that role.

  15. Kevin,

    That is precisely my point.  You say that the only way that neutrality law applies (and al Qaeda CCF are targetable 24/7) is if somehow al Qaeda are “recognized as lawful belligerents”.  If al Qaeda followed and enforced the laws of war they COULD (in my view) be regarded as lawful belligerents.  But because their leadership does not in any way follow or enforce the laws of war, then members of al Qaeda CANNOT be considered lawful belligerents entitled to the combatants’ privilege.  If you agree with Solis on that point, as I do, that al Qaeda CCF are targetable 24/7, what is your basis for applying IHL to them? 

    If not neutrality law then what is the basis for targeting AQ 24/7?

  16. Fascinating stuff.
    The crux for me is not whether a drone – or more accurately, Remotely Piloted Vehicle (RPV) – delivered the missile, (not least because some reports have manned aircraft ready to bomb the cars if there was a problem with the drone), but the legal status of the engagement.
    It’s most interesting that we can juxtapose the jury selection for the trial of Umar Farouk Abdulmutallab starts tomorrow as a criminal matter, whilst his purported leadership in AQAP are killed in a drone attack. Now, either AQAP is in either an IAC or a NIAC with the US, in which case the drone strike is probably legal, but there is a legal knock on to Abdulmutabllab’s case, or the US would have to show that under IHRL there was no other option to remove the imminent danger to civilians than the use of lethal force against al-Awlaki et al.
    At some level either argument can be sustained – even, in my view, that there is a continuing IAC. But what cannot be sustained is treating members of the same group engaged in, or planning/inciting, violent acts to different legal regimes simply because it suits us to do so. Consistency is critical to both rule of law here and to the broader political case for the use of force; reciprocity against western interests based on this sort of double standard would (rightly) be completely unacceptable.

  17. I took care of the rapid response to Goldsmith on FB – see second trackback below.

  18. Kevin,
    Have you ever seen or read Giraudoux’s Electra?

Trackbacks and Pingbacks

  1. […] that President Obama did not have the Constitutional authority to kill al-Aulaqi or that the strike violated international law, others argue that the strike was both legal and […]

  2. […] on a shiny new historical comparison to justify the targeted killing of Anwar al-Awlaki,” says Kevin Jon Heller at Opinio Juris, in response to my invocation of the Yamamoto precedent, and a […]

  3. […] legal case for killing Awlaki (here, here, and here). Case against (here, here, and here). Update: For background on AQAP and its ties to Yemeni tribes, there’s a brand new report […]

  4. […] Anyone who imagines the very thorny issues presented in the current case are somehow illuminated by analogies from World War II is just kidding themselves: if this conflict were not so plainly unlike World War II and other […]