The starting point is Article 51 of the UN Charter, which says nothing "shall impair the inherent right of individual or collective self-defence if an armed attack occurs…" with "occurs" in the present tense. This has led to two differing views. One says the armed attack must already be occurring, and therefore you have no right of anticipatory self-defence. The other view says, you cannot wait like a sitting duck for the attack to take place, especially in the nuclear era, and therefore you can attack if it can be shown that an attack against you is imminent in a way that cannot be and would not be stopped unless you react.

In any case it is a reaction. You have to show that you are reacting against an imminent attack, that there is an attacker who is about to attack. And here the paradigmatic case was for years the Cuban missile crisis, and there was this famous saying by Myres McDougal that you cannot wait like a sitting duck for an attack to take place, if after the attack you wouldn’t have the capacity to react, and the notion of self-defence would be meaningless. The U.N. charter is not a suicide pact, as it’s been said.

Now there is another argument worth mentioning, which says that Article 51 in fact does not grant a right to self-defence. The right to self-defence, as Article 51 states, is inherent: "Nothing in the present charter shall impair the inherent right…" as if the right exists regardless of Article 51, in a way that antecedes the UN charter. In other words, a right to self-defence exists in customary international law, and Article 51 does not impair it. And if you talk about customary international law, you refer back to the Caroline case, which was in effect a case of anticipatory self-defence.

Now if we think that anticipatory self-defence is not acceptable, then there is no point in us getting further into this discussion. If we think that anticipatory self-defence can under certain circumstances be acceptable, then we have to examine to what extent an army can start the actual fighting, and ask what we have to show in order to be able to claim self-defence.

And here the problem of course is, as it was under the Caroline case, the question of imminence and the question whether there are means other than actual fighting to prevent the attack. This I think ultimately is a question of fact. You have to convince the international community that you acted in self-defence, that you had no other choice, and that is similar to the situation of an individual within a society. I mean everyone has a right of private self-defence, when you are being attacked personally, and you don’t have time to call the police or run away. And then the question whether you acted in self-defence or not will be judged ex post facto.

In international affairs, the same rationale applies. A country feels itself under the threat of an imminent attack and reacts to it, and it cannot be sure that its act will be interpreted as lawful by the international community, say by the UN Security Council, etc. And this gives the correct incentives and precautions for the states involved. They cannot know in advance whether they are acting lawfully or not. They take their chances, they take the risk that if, post facto, their actions are interpreted as a violation of the law, they will be held responsible. They weigh that prospect against the risk of not responding at the time for an attack.

This is exactly the same situation with individuals who are being attacked and they find that they'd better defend themselves and later confront the possibility of being brought to justice.

One specific example we have had was the bombing of the Osirak nuclear reactor in Iraq by Israel in 1981. At that time the reaction was overwhelmingly negative. The UN Security Council denounced Israel for its action, and the United States joined the unanimous condemnation of Israel. Maybe now it might be seen differently. I am not aware of specific arguments nowadays but I recall that during the Gulf crisis in 1991, some said perhaps it wasn't such a bad idea, and perhaps when we deal with leaders who cannot be trusted this kind of action might be justified.

You can look at the law on the use of force and self-defence as a set of norms that are enforced by the principle of reciprocity. All states are repetitive players, and they want the rules applied to them in the same way as others because they can be one day on the receiving end as it were. And therefore the principle of no use of force with the exception of self-defence makes perfect sense under a system of reciprocity, when reciprocity is a genuine possibility.

When reciprocity cannot be ensured, I think we live in a different world. If you take terrorists, there is no reciprocity between the attacked state and the group of terrorists who express no clear interest in their own survival.
The fact that they can be the target of an attack means little to them. So I would venture to say that you cannot use the same logic of no use of force other than in self-defence, when you address actors who are not repetitive ones but have only one strike, if you see what I mean. And therefore when dealing with them, I think you enter into a different way of looking at the prevention of attack – more like a police action.

You know the police don’t have to wait for the criminals to attack. I think that we have a similar concern here, and one government versus unidentifiable terrorists or a terrorist group that exists in different places - there is no reciprocity between the two. If you take say the government of Iraq, a stable government that is sovereign and is a repetitive player. It may not have the same leadership through time, but it's going to be the same country, the same people, and they understand that, they internalise the consequences of an act of war.

Terrorists are not the same in that sense. They seek war and they are not intimidated by international law on the use of force, they are not intimidated by the prohibition and granted they are not inhibited by that, then there's no sense in using a law that applies to reciprocal relationships towards them. Now this I would venture to say is why we should differentiate between the usual question of whether or not to attack a nuclear reactor of a state or a government we don't trust, and whether to attack terrorists who have say a dirty bomb or something, or are about to get a dirty bomb.

In the case of terrorists who have a dirty bomb, I would say we have to apply a different logic, a logic of police action rather than the logic that I said of reciprocal relationships. When we have a stable government on the other hand, say we go back to Iraq in 1981, then I think it would be more difficult to explain why it is legal to strike first well in advance.

In the particular case of Iraq, Israel was technically in a state of war with Iraq. Iraq refused to sign the amnesty with Israel in 1949 and therefore it was arguably legal for Israel to strike. This is a claim that is not accepted by a number of people, but it clouds the picture.

I would say this. If there were a normal relationship between two countries, for example Israel and Pakistan, Pakistan wanting to be the first to construct the Arab nuclear weapon, then I would say you would not be able to justify such an attack under the principle of anticipatory self-defence.

However, I would suggest that even if an anticipatory attack is unlawful under Article 51 and customary international law, it doesn't necessarily mean that if someone used their power in anticipatory self-defence they would be condemned by the UN or brought before the ICC. It could be unlawful but excusable.

You want to provide clear norms that no single country has the capacity to make a judgment over the action of another country, or the intentions of another country. Therefore you say, with regard to anticipatory self-defence, if there is still time to prevent it or still time for diplomacy, then you should wait and see. If you chose to act under those circumstances, it would be a violation of Article 51. Still there might be no sanctions, no condemnation for the act taken – there is room for discretion as to the consequences of an illegal action.

Similarly, let's take what happened in Kosovo. The NATO attack, many people argue, was not a legal act, but it was excusable after the act, ex post facto, because of the humanitarian situation that developed immediately following or immediately before the attack. So what I want to say, there is a third possibility of finessing the consequences of the illegal action, and you actually know it ex post facto. Say the Israeli attack on Osirak, at the time it was considered unlawful but since then in the 1980’s there was the use of chemical weapons against the Kurds and against the Iranians, and there was the 1990 invasion of Kuwait, and you could then realise that the intentions of Saddam Hussein were not pure, and this gives you in retrospect a new sense of what could have happened had Israel not bombed the nuclear reactor.

You know we live in a world of uncertainties. We don't know what will happen if we do this, we don't know what will happen if we don't do this, so there's two ways to look at it - in advance, and ex post facto. Sometimes you know better ex post facto and this gives you an excuse for what initially was a violation of the law, but eventually becomes excusable.

So I would say if the US decides to strike facilities where nuclear weapons are being developed, it takes the risk that it may be deemed a violation of the law and also not seen as justifiable. Of course for the US it's not a real threat, because its position in the world means it can escape sanction. For Israel it's more so – which leads some to say well, this provision doesn't make sense, because it gives the US anyway the right to do whatever it wants. But we live in a world of second-best - this is the least worst situation, where you would say it could be a violation of the law, probably it is not a valid case of anticipatory self-defence, but it is a potentially excusable action.

If we take a related issue, the issue of humanitarian intervention, many people say, I think rightly, that humanitarian intervention has been ultimately accepted in the 1990s, though it was debated earlier. And there was an evolution of the law and pre-emptive self-defence may be another evolution of the law with the same concerns. Again it's a question of the allocation of risks. What do we fear more? Because indeed the UN charter is not a suicide pact, we have to assess what is worse, the situation where you have the ability of one country to strike against another, if there is threat of nuclear weapons, or where you don’t. And ultimately it's a question for the international community to decide and I think we'll see an evolution in that direction.

Back to Top