Why Preventive Self-Defense Violates the UN Charter

by Kevin Jon Heller

I am teaching IHL in Jericho this week, so I don’t have as much time as I’d like to weigh in on the increasingly surreal debate over whether the right of self-defense in Article 51 of the UN Charter permits the U.S. or Israel to attack a country that does not have nuclear weapons, could not build a nuclear weapon anytime soon, and is not — according to the U.S. itself — even trying to build a nuclear weapon.  Readers who are interested in taking the issue seriously, however, should check out Tom Ruys’ magisterial recent book “Armed Attack” and Article 51 of the UN Charter: Evolutions in Customary Law and Practice.  Ruys carefully analyzes state practice and opinio juris to determine whether customary international law permits states to engage in preventive self-defense — the use of military force in response to the threat of an armed attack that is not imminent.  Here is a long snippet from the book, discussing the international response to the report issued in 2004 by the UN’s “High-Level Panel on Threats, Challenges and Security,” which specifically rejected preventive self-defense but cautiously embraced pre-emptive self-defense — the use of military force in response to the threat of an imminent armed attack (the Caroline standard):

[T]here can be no doubt that even among States adhering to the ‘counter-restrictionist’ view, support for self-defence against non-imminent threats is virtually non-existent. Apart from the fact that the sponsors of Operation ‘Iraqi Freedom’ avoided this justification, it may be observed that many States, such as Germany, Japan, Switzerland, Uganda, Singapore or Liechtenstein, which professed support for anticipatory self-defence after 2002, nonetheless placed great weight on the imminence requirement. Germany, for instance, expressly denounced an erosion of the Charter framework and State practice via the notion of ‘preventive self-defence’. Likewise, the French politique de defense unequivocally ‘rejects . . . the notion of preventive self-defence’.

What is more, even the ‘traditional’ adherents of the counter-restrictionist interpretation of Article 51 generally appear to uphold the imminence requirement. Despite bold statements by its Prime Minister on the need to adapt the UN Charter, Australia’s response to ‘In Larger Freedom’ was rather cautious: it simply ‘[supported] reaffirmation by the Secretary-General that Article 51 of the Charter adequately covers the inherent right to self-defence against actual and imminent attack’. Israel called for an explicit recognition in the World Summit Outcome that States may use force in self-defence ‘in the event of both actual and imminent attacks’. As far as the British position is concerned, Attorney- General Lord Goldsmith in 2004 declared before the House of Lords that: ‘It is . . . the Government’s view that international law permits the use of force in self-defence against an imminent attack but does not authorize the use of force to mount a pre-emptive strike against a threat that is more remote’.


This brings us to the third and last question, regarding the balance of majority and minority opinion among States. Interestingly, Reisman points to a ‘significant lack of comment’ on the interpretation of self-defence in the Reports of the High-Level Panel and the UN Secretary- General, which, nonetheless constituted ‘an attempt at adjustment of the Charter to meet part of the U.S. claim’. In light of this alleged tacit approval, one might be inclined to qualify the extended support for pre-emptive self-defence as a decisive shift in the opinio iuris of the international community. An analysis of State responses, however, learns that this conclusion is rather precipitate. Indeed, notwithstanding a number of positive (cf. supra) and rather vague reactions by several States, numerous UN Members opposed the acceptance of pre-emptive self-defence. Turkey, Argentina and Mexico, for instance, observed that the admissibility of self-defence in response to imminent threats was very controversial and warned that recognition thereof might lead to complications and abuse. Many States went even further and combined a rejection of the interpretation of Article 51 in the two Reports with an explicit affirmation that self-defence could only be exercised in response to an actual armed attack. In the words of Pakistan: ‘We certainly do not agree that Article 51 of the Charter provides for the pre-emptive, preventive or protective use of force. In our view, the language of Article 51 is quite explicit and highly limitative. It provides for the use of force in self-defence only in case of an actual attack against a Member State’.

Comparable statements were submitted by the representatives of Vietnam, Belarus, Bangladesh, Algeria, Iran, Cuba, Costa Rica and Egypt. Other States that explicitly or implicitly voiced opposition to pre-emptive self-defence include China, India, Syria, Malaysia and Indonesia. China, for example, argued that Article 51 needed neither rewriting nor reinterpretation: ‘[E]xcept in case of self-defence against armed attacks, any use of force must have the authorization of the Security Council. Any “imminent threat” should be carefully judged and handled by the Security Council . . .’. India noted that the framers of the Charter never intended Article 51 to cover anything beyond its text. Significantly, the Non-Aligned Movement in 2005 issued a position paper emphasizing that: ‘Article 51 . . . is restrictive and recognizes ‘the inherent right of individual or collective self-defence if an armed attack occurs against a Member of the United Nations.’ This Article should not be re-written or re-interpreted. This is supported by the practice of the UN . . .’.

In the end, because of the resistance of a considerable number of UN Members to the acceptance of pre-emptive self-defence, the relevant recommendations of the High-Level Panel were not included in the Outcome Document of the 2005 World Summit of the UN General Assembly. No trace of Article 51 can be found in the resolution. Instead, paragraph 79 simply reaffirms ‘that the relevant provisions of the Charter are sufficient to address the full range of threats to international peace and security.’

In light of the available evidence, it can be concluded that there has indeed been a shift in States’ opinio iuris insofar as support for pre-emptive self-defence, fairly rare and muted prior to 2001, has become more widespread and explicit in recent years. At the same time, it seems a bridge too far to claim that there exists today widespread acceptance of the legality of self-defence against so-called ‘imminent’ threats. Such assertion tends to forego the opposition of a considerable group of mainly Latin-American, north-African and Asian States. In the present author’s view, it would therefore be more appropriate to argue that the crack in opinio iuris among States has widened, without, however, identifying one approach or the other as the majority view. The implication is that, taking account of the Charter ‘baseline’ and the absence of a concrete precedent in State practice which convincingly demonstrates the international community’s support for some form of anticipatory self-defence, it is impossible to identify de lege lata a general right of pre-emptive – and a fortiori preventive – self-defence.

This is what careful analysis of customary international law looks like — a refreshing contrast to the myopic discussion provided by someone like Peter Berkowitz, who defends the legality of a pre-emptive attack on Iran solely on the basis of U.S. practice.  (And it’s worth noting that Berkowitz’s reliance on the invasion of Iraq is itself flawed; as Ruys points out, “[t]hose States supporting the operation carefully avoided any justification based on a broad reading of Article 51 UN Charter. Furthermore, a majority of States apparently held the opinion that the operation violated the UN Charter.”)  As much as it likes to pretend otherwise, the U.S. does not get to unilaterally dictate the content of customary international law.  And that is nowhere more clear than in the context of Article 51 and “preventive” self-defense, where the U.S. position is opposed not simply by “minor” states whose views the U.S. has no problem disregarding, but also by most of the world’s major powers: Germany, France, the UK, Australia, Japan, India, China.

So no, international law does not permit the U.S. or Israel to attack a country that does not have nuclear weapons, could not build a nuclear weapon anytime soon, and is not even trying to build a nuclear weapon.  Such an attack would be a flagrant violation of the UN Charter.


13 Responses

  1. This is only correct to the extent that your three propositions regarding Iran – and specifically the second and the third – are correct. I think this is where the argument raises difficult dilemmas.

  2. Does Iran have nuclear weapons?  If not, the threat cannot — by definition — be imminent.  The Caroline standard is clear: “a necessity of self-defense, instant, overwhelming, leaving no choice of means, and no moment for deliberation.”

  3. @Kevin
    People who devised the ban of use of force had the following in mind: how do we design rules that avoid abuse as much as possible? This is because they knew from experience that once you give a state a possible loophole, states will use that to wage wars for other reasons – for imperial ambitions and free oil.
    Therefore the rule of thumb is very simple: Does a legal argument give the charlatans a cheap excuse to use military force? If yes, than it is a bad argument.

  4. But then the problem is – and I don’t have an easy solution for this – that once Iran does in fact have a nuclear weapon, the threat becomes imminent but also uncounterable. Not to mention the collateral effects of an attack at that stage. That’s why it’s a bit problematic to analyze the immanency test regarding nuclear weapons in the exact same manner that we would discuss conventional threats.
    I am not writing this to defend a legal stance according to which an attack at this stage is legal. I’m just pointing out that it’s not an easy case.

  5. Eliav,

    You’re right, it is a difficult case.  I think most states would respond that, until the state has a nuclear weapon, countering the threat of them building one is entrusted to the international community acting through the Security Council — not through unilateral military action.

    I would add, though, that there is a difference between a state possessing a nuclear weapon and a state threatening to use one.  I’ll quote Ruys again:

    [E]ven if the Security Council has labelled the proliferation of nuclear, chemical and biological weapons a threat to international peace and security, there is no rule in general international law which prohibits a State from developing and/or possessing nuclear weapons per se. Article II of the Non-Proliferation Treaty inter alia requires non-nuclear weapon States Parties to the Treaty ‘not to manufacture or otherwise acquire nuclear weapons or nuclear explosive devices’ and ‘not to seek or receive any assistance in the manufacture of nuclear weapons or other nuclear explosive devices’. Article VI moreover requires nuclear weapon States to pursue diplomatic negotiations in order to achieve nuclear disarmament. On the other hand, Article X affirms that States Parties have the right to withdraw from the treaty by giving three-months’ notice when they are faced with ‘extraordinary events’ that ‘jeopardize their supreme interests’. States that have properly withdrawn from the Treaty or that have not joined the regime in the first place (e.g., India, Pakistan and Israel) are probably not bound by any international legal duty not to develop or acquire nuclear weapons.

    Third, there is some difficulty in equating the development of a nuclear weapons programme by an enemy State to an imminent threat of an armed attack. Indeed, from a military–strategic perspective, nuclear weapons are in principle not offensive weapons, but rather ‘safety devices’ intended to deter large-scale attacks against the possessing State. Put differently, States presumably do not seek to acquire nuclear weapons to carry out concrete attacks, but rather to shift the strategic and geopolitical balance to their advantage. In Charter terms, this falls within the frame- work of collective security rather than that of self-defence. No State has ever deployed nuclear weapons in combat since the notorious bombings of Hiroshima and Nagasaki on 6 and 9 August 1945. Every Head of State is aware that a first use of nuclear weapons would not only make it a pariah in the international community, but would most likely entail a massive military response.

  6. Thanks a lot, Kevin – this is very useful.

    I would add one thing. The rule of thumb could actually be the following: is State A prepared to accept its own arguments for preventive self-defence if propounded by State B against State A? For instance, there might be a good argument for Israel to bomb Iran in preventive self-defence due to perceived threats by Iran if Israel agrees that Iran can do the same against Israel on the basis of (similar) perceived threats. If Israel agrees that this would be in Iran’s right, then they do have a slightly better argument…

  7. Response…
    Kevin: The Caroline incident is NOT justification for anticipatory self-defense, what you term preemptive self-defense (which is confusing because others consider preemptive self-defense to encompass what you term preventive self-defense — neither one of which is justified under Article 51 of the Charter, which expressly requires the start of an “armed attack”).  There were ongoing armed attacks by insurgent non-state actors against Canada prior to, during, and for months after the British attack on The Caroline.  The U.S. and the U.K. argued about “self-defense,” and the U.S. viewpoint that is often quoted was with respect to the method and means of response in context — e.g., whether the actual attack was necessary when the U.K. could have waited until The Caroline went back into British waters.  Please see http://ssrn.com/abstract=1520717 re: The Caroline incident and other patterns of practice and expectation regarding self-defense, and please see http://ssrn.com/abstract=1798582 regarding overwhelming textwriter opposition to preemptive (vs. anticipatory) self-defense.
    With respect to the start of an “armed attack,” one can recognize that a process of armed attack has started, for example, when the enriched uranium is placed on missiles aimed at Israel and the leaders of Iran declare that they are going to wipe Israel off the map.

  8. Response…
    p.s.  I agree that that is apparently not the present circumstance.

  9. Response…
    obviously my comment before the “p.s.” was lost or is “awaiting moderation”

  10. Taking together Holder’s speech, War with Iran and Intervention in Syria mutterings, it does feel like hysteria is being built – once again.

  11. Bravo, Kevin. Even the threat of illegitimate force is against the UN Charter.

  12. I wrote about a similar issue long ago.  If you look at arguments made by states before the GA and the SC about uses of force, it is generally impossible to discern any opinio juris in favor of changing the meaning of either Article 2(4) or 51.  States sometimes do make broader claims in political contexts, but when legally justifying their acts they tend to rely largely on traditional arguments about 2(4) and 51.  (At least this was the case when I wrote the original article in the late 1990s.)  Their actions may not be consistent with the legal arguments they make, but they still lack the opinio juris necessary for the formation of new CIL.

    Therefore, I proposed treating them as offers to interpret the text of 2(4) and 51.  Seen this way, they are subject to a requirement of good faith (pacta sunt servanda). The best way to evaluate whether they are good faith interpretations is to look at how they are received by the other members of the UN.  Largely, these “interpretations” are rejected.  Sometimes even by states that had previously made similar “offers” to interpret the Charter.  This suggests quite strongly that most states view them as not being made in good faith.

    If anyone is interested, the article is available at http://papers.ssrn.com/sol3/papers.cfm?abstract_id=1995608

  13. “In the present author’s view, it would therefore be more appropriate to argue that the crack in opinio iuris among States has widened, ******without, however, identifying one approach or the other as the majority view******. The implication is that, taking account of the Charter ‘baseline’ and the absence of a concrete precedent in State practice which convincingly demonstrates the international community’s support for some form of anticipatory self-defence, it is impossible to identify de lege lata a general right of pre-emptive – and a fortiori preventive – self-defence.”

    I have a probably naive (and definitely Neanderthal American positivist) question about the way Ruys formulates his conclusion here. It sounds like he’s saying that there’s insufficient agreement in the international community to generate a CIL *right* of pre-emptive self-defense (or VCLT subsequent agreement to the same effect). But isn’t that getting it backwards? Lotus and all that? (There’s that American Neanderthalism!!)

    The best response that comes to mind is: I’m thinking about it backwards — there’s a clear ban in 2(4), and it’s that unequivocal *baseline* w/r/t which clear understandings would have to coalesce to create a *new exception*. But that seems not just functionally weird but conceptually incoherent — actually too faux naive/legalistic in its own way. I.e., both conceptually and in terms of the negotiated agreement, you can’t understand 2(4) without 51, and vice versa — they are inherently and necessarily embedded with one another such that understandings (and as relevant here, LACK of settled understandings!) about the one are embedded in what the other “means.”

    Please note — I’m not actually sure what I think the right answer is here. Am trying to understand the argument from an internal perspective, which at the moment I’m not fully following . . . 

Trackbacks and Pingbacks

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