Geoffrey Robertson: Using force to stop slaughter is lawful

Will the world stand idly by once Colonel Gaddafi, a man utterly without mercy, starts to deliver on his threat to "fight to the last man and woman" – and, inferentially, to the last child?

The shadow of Iraq invasion illegality has tainted talk of "liberal interventionism" – unfairly, since Bush was no liberal and Blair has wrongly used it as a retrospective excuse. There was no looming humanitarian crisis in Iraq in March 2003, and the aggressor states (the US, UK and Spain) explicitly ruled out this justification: they claimed an entitlement to circumvent the Security Council because of a convoluted reading of an earlier resolution and a bizarre claim to the right of self-defence against Saddam's imaginary weapons for mass destruction. The lesson of Iraq is not that this country should never use force against another, but that never again should it do so in breach of international law.

Which raises the big question, namely the circumstances in which there is a right – or, more importantly, a duty – to use force to relieve a humanitarian nightmare. The UN charter bans "the use of force against territorial integrity or political independence of any state" other than in individual or collective self-defence, or else with Security Council authorisation after the council has determined under Chapter VII of the charter that a threat or a breach of the peace has occurred. This is clear, as far as it goes, but the problem is that the "big five" each have a Security Council veto, and China and Russia generally oppose intervention other than to liberate invaded states (which was the case with Kuwait when it was invaded by Saddam).

In the case of Libya, the council has at least set an important precedent by unanimously endorsing a reference to the International Criminal Court. International justice, however, grinds slowly. So what happens if the unarrested Libyan indictees aggravate their crimes – eg by stringing up or shooting in cold blood their opponents, potential witnesses, civilians, journalists or prisoners of war? If the Security Council in secret session decides to do nothing, does international law permit others – eg the Nato alliance – to pick up the gauntlet?

In my view – contested by some – there is now a narrowly proscribed international law right for states to render assistance to innocent civilians battling for their lives. This right of humanitarian intervention goes back to the "just war" theories of Grotius and Vattel in the 17th century. Examples of such action include the Tanzanian invasion of Uganda to overthrow Idi Amin and India's invasion to stop genocide in Bangladesh. Such actions were, however, justified at the time on very dubious grounds of self-defence and the chief objection to a broadly stated "right of humanitarian intervention" without Security Council approval remains that it is liable to be mistaken for "a right of ideological intervention". Hitler invoked it to justify the use of force to protect German minorities from alleged brutality – in Czechoslovakia and then in Poland.

But more recent examples show that a rule of law is developing to allow "coalitions of the willing" to use appropriate force to prevent a humanitarian catastrophe. The two important precedents are the "safe havens" operation by the US, UK and France, invading northern Iraq without specific Security Council authority to protect Kurds against violent reprisals threatened by Saddam, and the Nato bombing of Kosovo.

But typically when the UN came to embrace a "just war" principle that it called "the responsibility to protect", this was based on state sovereignty in which was optimistically detected an "implied responsibility" to protect its people from suffering serious harm. If a state was unable to protect them, eg because of civil war, its responsibility would devolve on the Security Council. This much-touted "R2P" principle encourages Security Council action when it is "right to fight". That is not much help when the Security Council is blocked by a big power veto. The best the R2P authors can do is to say that operations like Kosovo, while illegal, may nonetheless be "legitimate". This is a silly fudge. It means it would be unlawful for Nato to set up a no-fly zone over Libya, but that the moral "legitimacy" of the action would provide Nato high command with a moving plea in mitigation.

The defects in the Security Council require the acknowledgement of a limited right, without its mandate, for an alliance like Nato to use force to stop the commission of crimes against humanity. That right arises once the council has identified a situation as a threat to world peace (and it has so identified Libya, by referring it unanimously to the ICC prosecutor). To be lawful, the intervention must be at the request of potential victims, for the purpose of stopping crimes against humanity and with no ulterior motive, eg obtaining territory or oil. It must be proportionate – no greater force than necessary. Subject to these preconditions, Nato's intervention in a Libyan emergency would be lawful, unless or until it was denounced by majority vote in the Security Council.

International law is not passed by any parliament: it "emerges" or "crystallises" from state practice, conventions (including human rights conventions), writings of jurists and dictates of collective conscience. The duty to stop the mass murder of innocents, as best we can if they request our help, has crystallised to make the use of force by Nato not merely "legitimate" but lawful. It must be a very last resort, after the Security Council fails to act. The council would be less likely to fail if its deliberations were in public: if the world could watch how the interests of realpolitik and real morality are weighed, the scales may come down more often on the side of justice.

Geoffrey Robertson QC is a member of the UN's justice council. His books include 'Crimes Against Humanity' (Penguin)