Jonathan Sumption

The new Machiavelli

Text settings

Preemption: A Knife That Cuts Both Ways

Alan M. Dershowitz

W.W. Norton, pp. 348, £

Should the state take action against people who have done nothing wrong, if there are plausible grounds for thinking that they are about to? Suppose, says Alan Dershowitz, that reliable intelligence shows that a large-scale terrorist attack is about to happen. Should the law allow the police to round up whole categories of potential perpetrators in advance, in the hope of stopping the conspiracy in its tracks? Should it authorise the use of torture to obtain information that will prevent the attack? What if the suspected mastermind is identified in a foreign country where the authorities are too weak, wicked or incompetent to arrest him? Might one assassinate him instead? Or invade the foreign country? Or bomb it from a great height?

These are ancient dilemmas, but it has taken the events of the last five years to bring them to the forefront of political debate. Even now the public, although largely in favour of pre-emptive action, is reluctant to contemplate its implications. Politicians have been in no hurry to point them out, preferring to retreat behind a fog of emotion and fear. Alan Dershowitz will have none of this. He believes that the issues need to be confronted. What follows, after the initial statement of the problem, is an extended justification of pre-emptive action against the enemies of society, not just as a pragmatic response to periodic crises, but in principle. In extreme cases, Dershowitz would not stop short of torture or assassination, provided that there were proper legal safeguards against abuse.

The reasoning is subtle, moderately rigorous and illustrated with much historical learning. Any short summary is therefore bound to sound cruder than the real thing. But Dershowitz is basically making three points. First, he argues that terrorists, instead of being treated as particularly heinous criminals, should be reclassified as combatants. Their methods are after all essentially military, and the damage they can do is more characteristic of war than of large-scale crime. So if there is no other way of stopping them, let them be subjected to the kind of pre-emptive violence which has hitherto been reserved for the enemy’s armed forces in wartime.

The second argument is a variant of the Benthamite calculus. You have to weigh up, says Dershowitz, the strength of the evidence, the imminence of the threat, the gravity of the consequences if the attack were to succeed, the damage that pre-emptive action would do to the intended victim, and the risk of collateral damage to innocent third parties. What you end up with is certainly a departure from traditional legal and moral principles. But it may nevertheless be right if it brings the minimum acceptable level of security to the greatest possible number, at the smallest inconvenience to the innocent or possibly innocent. Of course, it is hard to be precise about these things. Abraham discovered that when he tried to haggle with God about the exact number of false positives that would be acceptable in the effort to destroy the sinners of Sodom. But judges and tribunals are trained to deal with such difficult conundrums.

Dershowitz’s third main point is a surprising one. The authorities, he says, will engage in pre-emptive detention, torture, targeted assassinations and the like anyway. So we might as well make it legal. Then at least it can be regulated, to ensure that it is limited to what is necessary and proportionate. It is not clear how far Dershowitz would carry this principle. Would lesser crimes than terrorism warrant lesser responses than precautionary torture or preventive assassination? Policemen, for example, sometimes beat up suspected criminals in the cells for want of any better method of social control. This is inconvenient for the criminals, but perhaps not excessively so. Such incidents will no doubt continue. Should they be legalised, under appropriate judicial controls, if it can be shown that they are an effective way of reducing crime?

The problem with Dershowitz’s argument, as he fairly acknowledges, is that it means that there are no moral absolutes to restrain the coercive action of the state. It all comes down to weighing the ends against the means. The fact that Dershowitz would have it done with the authority of the courts is hardly reassuring. At an international level, there would not even be this safeguard. As he points out, if pre-emptive action succeeds, one may never know whether the end justified the means. If Britain and France had invaded Germany in 1936 and removed Hitler from power, we would not know what horrors he would have perpetrated in the next nine years. In hindsight the invasion would have saved the world a great deal of agony. In fact it would have gone down in history as an unprovoked aggression against the elected government of Germany.

If Dershowitz’s views were widely accepted, it seems most unlikely that their application could be confined to extreme cases or subjected to effective legal control. War, murder and torture would be justified by little more than the intensity with which the victim was feared, and the awfulness of the consequences if the fear should hypothetically turn out to be justified. Dershowitz is a professor of law at Harvard, but his prescription for dealing with terrorism is the negation of the rule of law. It would undermine the moral integrity of the state far more completely than the official cheating which admittedly goes on at the moment.

Preemption has been much abused in the United States since its first publication there, partly because its author’s views have been quoted without referring to his reservations. It is in fact a book of singular intellectual honesty, which argues a difficult case with skill and verve. It does not cheat, and it gives proper attention to most of the arguments pointing the other way. Traditional liberals need to have their convictions tested in this way. His views deserve to be widely read and universally rejected.