Peter Berkowitz’s Unconvincing Justification of “Preventive” Self-Defense

In a previous post, I noted that Peter Berkowitz defended the legality of a U.S. or Israeli attack on Iran’s non-existent nuclear weapons program by invoking U.S. practice — and only U.S. practice.  He now has offered another defense of such an attack in response to an editorial by Bruce Ackerman in the L.A Times.  Unfortunately, Berkowitz’s second article is no more convincing than his first.

First, Berkowitz claims that “Article 51 of the U.N. Charter, which recognizes states’ ‘inherent’ right of self-defense, does not exclude preemption… Article 51 has always presumed a right of anticipatory self-defense that is significantly broader than the Webster test.” Berkowitz offers precisely one piece of evidence in support of that claim — a 1914 statement by Elihu Root, then a U.S. senator, at an ASIL conference.  That’s it.  Berkowitz does not even explain why, if the U.S. believed that the customary scope of self-defense extended well beyond Caroline‘s “imminence” standard, Webster would have specifically relied on that standard to criticize the British attack — which the British Foreign Minister later agreed was the correct one (see Ruys, 256).

Second, Berkowitz argues that “[w]hen the U.N. Charter was ratified in 1945, Article 51 did not change things. As esteemed Yale Law School scholar of international law, Myres McDougal, wrote in 1963 about the Cuban Missile Crisis, ‘There is not the slightest evidence that the framers of the United Nations Charter, by inserting one provision which expressly reserves a right of self-defense, had the intent of imposing by the provision new limitations upon the traditional rights of states’.” Not the slightest evidence?  Ruys convincingly rebuts Berkowitz’s argument in the second chapter of his book, painstakingly reviewing the drafting history of Article 51 and concluding (p. 66) that “[n]othing in the preparatory works suggests that Article 51 was merely perceived as an example of a broader, pre-existing and unimpaired customary right of self-defence. No explicit reference was made to pre-existing custom, nor to any alleged right of self-defence in circumstances not involving an armed attack.”  Moreover, here is what he says (p. 259) specifically in the context of preventive self-defense (and note the U.S.’s own position during negotiations):

[M]ore convincing, the ‘restrictionists’ argue that, even if anticipatory self-defence was permitted in the years prior to the adoption of the Charter, pre-existing custom was nonetheless modified by Article 51 UN Charter. This argument, which the present author subscribes to, was extensively discussed in previous chapters and will therefore not be restated in full. In essence, it can be broken down into two parts. First, in light of the equal normative position of customary and conventional law, and in accordance with the lex posterior principle, the introduction of the Charter rules on the use of force has removed incompatible pre-existing custom. Second, each of the primary elements of interpretation supports the view that the occurrence of an ‘armed attack’ is a sine qua non under Article 51 UN Charter. In accordance with the principle of effectiveness it is hard to give a different twist to the clear wording of the phrase ‘if an armed attack occurs’. If the drafters did not intend this phrase to be regulatory, one might have expected them to use a different wording (‘for example, if an armed attack occurs’), or to have omitted the phrase altogether (expressio unius est exclusio alterius). McDougal’s objection that ‘if A, then B’ does not equal ‘if, and only if, A, then B’ cannot be upheld if one looks at the Charter context. Article 51 constitutes an exception to (or better, qualification of) the comprehensive prohibition on the use of force, and must accordingly be interpreted restrictively. The fact that ‘the threat of force’ and ‘threats to the peace’ are mentioned in Articles 2(4) and 39 proves that the drafters were aware of such situations and chose to submit incidents concerned to the Security Council. The ‘objects and purpose’ of the UN Charter only confirm this reading, since its goal was precisely to limit the unilateral use of force as much as possible and to subject it to the control of the Security Council.

While the present author believes that the primary elements of interpretation settle the matter conclusively, one might, for argument’s sake, object that a literal reading of Article 51 leads to manifestly absurd results and therefore warrants a recourse to the Charter’s travaux by way of supplementary means of interpretation. Again, however, nothing in the records of the San Francisco Conference suggests that the phrase ‘if an armed attack occurs’ was intended to be declaratory instead of regulatory. The word ‘inherent’ was included without any debate taking place as to its meaning. US Governor Stassen emphasized that ‘we did not want exercised the right of self-defence before an armed attack had occurred’. When asked what action could be undertaken against a fleet that ‘had started from abroad against an American republic but had not yet attacked,’ Stassen responded that ‘we could not under this provision attack the fleet but we could send a fleet of our own and be ready in case an attack came’.

When taken together, these arguments make clear that the ‘pre-existing custom’ paradigm is not only artificial, but fundamentally flawed. This view is shared by a considerable majority of legal doctrine. Even among proponents of anticipatory self-defence, many concede that legal support must be sought elsewhere, primarily in post-Charter customary practice. Franck, for instance, claiming that a purely reactive posture has become ‘logically indefensible by the advent of a new age of nuclear warheads and long-range rocketry’, nonetheless admits that ‘it is beyond dispute that the negotiations [at San Francisco] deliberately closed the door on any claim of “anticipatory self-defence”’.

Third, Berkowitz dismisses the Reagan administration’s support for SC Res. 487, which condemned Israel’s 1981 attack on Iraq’s Osirak nuclear reactor in the face of Israel’s explicit insistence that such a preventive attack was permitted by Article 51 of the UN Charter.  As he points out, the U.S. did not support the resolution because it necessarily agreed that such attacks violated Article 51.  With his myopic focus on U.S. practice, though, Berkowitz fails to discuss why the resolution is not relevant to the customary status of preventive self-defense — after all, SC Res. 487 was adopted unanimously, and soon thereafter the General Assembly adopted a resolution that mirrored SC Res. 487 and condemned the attack as “aggression” by a vote of 109-2 (the two “no” votes being Israel and the U.S.) with 34 abstentions.  Indeed, “dozens” of states condemned the idea of preventive self-defense on the floor of the Security Council and General Assembly, including the UK, Spain, Germany, Austria, India, Mexico, and the USSR (see Ruys, pp. 280-86).

Fourth, and finally, Berkowitz cites the Cuban Missile Crisis as evidence — quoting McDougal again — that “[t]he understanding is now widespread that a test formulated in the previous century for a controversy between two friendly states is hardly relevant to contemporary controversies, involving high expectations of violence, between nuclear-armed protagonists.”  As Ruys points out (pp. 270-71), however, the U.S. “provided little in way of legal justification” for its blockade of Cuba and — more importantly — made “no attempt… to justify the operation as an exercise of self-defence.”  Indeed, “[t]his was no coincidence… according to US Legal Adviser Chayes, the operation did not fall within the purview of Article 51, but ‘within a third category: action by regional organizations to preserve the peace’, pursuant to Article 52 of the Charter.”  Chayes was even more explicit a few years later, writing the following:

No doubt the phrase ‘armed attack’ must be construed broadly enough to permit some anticipatory response. But it is a very different matter to expand it to include threatening deployments or demonstrations that do not have imminent attack as their purpose or probable outcome. To accept that reading is to make the occasion for forceful response essentially a question for unilateral national decision that would not only be formally unreviewable, but not subject to intelligent criticism either… In this sense, I believe an Article 51 defence would have signalled that the United States did not take the legal issues very seriously, that in its view the situation was to be governed by national discretion, not international law.

Berkowitz’s response to Ackerman is entitled “A Misreading of Law and History on Preemptive Strikes.”  As the above demonstrates, it is Berkowitz, not Ackerman, who misreads the relevant law and history.