Can a country ever justify military intervention in another country's affairs? Only in extreme cases such as Rwanda, argues Tzvetan Todorov, when there is evidence of genocide.

The military intervention in Kosovo has often been presented as arising from a new doctrine, summarised in the phrase "the right to intervene". This is taken to mean that a group of states, such as Nato, has the right to undertake armed intervention anywhere in the world where massive and systematic human-rights violations are taking place.

The right to intervene had no sooner been formulated as a doctrine than objections were raised. One of the most common was that it required us to abandon the principle of national sovereignty, which had previously underpinned international relations.

Resistance to the right to intervene has been particularly lively in countries outside Western Europe and North America. But the theory of the right to intervene is not new. On at least two occasions in history, European states have justified actions beyond their borders by reference to "values and principles" rather than national interest - either for Christian reasons as in the Crusades and the conquest of America, or to advance European civilisation as in the new wave of colonisation of the 19th and 20th centuries. On both occasions, the European states proceeded with the conviction that right was on their side and that other countries in other continents were allowing wrong to triumph.

ÖйúAƬ

It might be objected that human-rights values are accepted in theory almost everywhere in the world, even in those countries where they are daily trampled underfoot. But the universal love preached by Christianity and the reign of reason asserted by the European powers in the 19th century were universal values also worth defending. Suspicion in non-European countries is, therefore, understandable. They have not forgotten previous interventions made "for their own good". Declared good intentions offer no guarantees and are too often camouflage.

Moreover, imposing right by force is never an unmixed benefit. Before we inflict our universal values on others, it might be proper to ask their opinion: if we acknowledge that they are no less human than we are, their opinions weigh no less in the balance than ours. However much their governments incur our censure, should we, where the will of an electorate has been freely expressed, simply discount it? In the past, we have frequently mistaken for universal values our own traditions and desires.

ÖйúAƬ

This does not mean we should give up our attempts to identify universal values. The question being debated is not the universality of rights and values, but that of their practical implementation in societies.

There is another reason why we might prefer national sovereignty to the right to intervene. Sovereignty takes the form of state institutions; intervention destroys the nation-state. The inhabitants of any country have many more rights as citizens of a state than they have as members of the human race. Anarchy is worse than tyranny because anarchy, the absence of a state apparatus, replaces tyranny by the tyranny of the multitude. Even unjust laws have the merit of predictability. It seems unlikely that an order imposed vi et armis from without will ever appear legitimate to the people of the country concerned. Negotiation, indirect pressure and charm offensives may well prove more effective than war.

Too often, the right to intervene is clearly exercised in a highly selective manner on those who are weaker or less strategically important or where the country intervening has a strong interest. What kind of justice is it that is not the same for all?

And even supposing that the right to intervene could be exercised to perfection, should it be generalised? Human-rights violations are far too widespread. If all violations were to be prevented or punished, war would never end.

Attempts to eradicate injustice from the earth resemble totalitarian utopias in their efforts to perfect mankind and establish a paradise on earth. They also require us to be convinced that right is embodied solely and exclusively in ourselves, an attitude reminiscent of the wars of religion. Put simply, the promotion of universal justice implies the scientistic construction of a universal state, with the same laws, institutions and police. Would this be the perfect state? By no means - its drawbacks would be substantially greater than its advantages.

ÖйúAƬ

Such terms are not much used. Yet evidence of the tendency they represent is not lacking. For example, the return of the kind of medical metaphors applied to the body politic once used in totalitarian regimes. We speak of surgical interventions, for instance, and argue that prevention is better than cure - as if social defects could be analysed in terms of illness. Imagery of the body is meaningful only if we conceive of humanity as a single entity, with a brain and a heart, and its agents, the inevitable arms of the law; as an entity in which there may be zones of corruption and malady from which the body must be protected, if necessary by excision. Such a view could justify preventive strikes - even where it subsequently emerges that danger was illusory.

Why is a pluralist ideal better than a unitary one? Because human knowledge can never be complete and because it cannot in any case realise its ideals. A science of politics will therefore never exist.

In its absence, balances of powers, mutual tolerance and a plurality of centres of decision are more valuable than unity. Plurality ensures liberty and the possibility of inquiry. Translated into international relations, this means that mutual deterrence among several groups of countries or several superpowers may be more advantageous than the exclusive domination of a single power. Military intervention in a foreign country, to the detriment of its national sovereignty, is justifiable in one extreme case - on condition that the intervention is not likely to cause more victims than it saves. For the past few decades, this extreme case has had a name, that of genocide.

ÖйúAƬ

Kosovo was not genocide. Since the second world war, there have been only two genocides - Cambodia and Rwanda. Military action in such cases presents no fewer dangers than other forms of intervention, but they must be accepted, given what is at stake. We cannot know if we are acting too late or how many will die as a result of our intervention. Nor can we distinguish accurately between collective massacres and embryonic genocide.

Error cannot be eliminated; but if we take as our principle that genocide alone justifies military intervention, we are permitted to hope that interventions - perhaps by neighbouring countries with better reasons and more empathy with those affected - will be few and far between.

This is no panacea. It could favour regional powers to the detriment of universal principles, but it is preferable to genocide. Rejecting intervention does not, however, mean doing nothing. Those who suffer have a right to assistance, whether legal, political, humanitarian or economic. It will not bring miracles, but it will avoid the imposition of force and a multiplication of the number of victims.

Balance of power is an integral factor in international relations. But we are not obliged to accept that power relations be tricked out as the magnanimous bestowal of some transcendent or practical benefit, as they were in the past. Nor should we confuse defence of the national interest, a legitimate aim of any government, with the struggle for universal justice. We must champion right over might, but when faced with two powers, if one frankly avows its objectives and the other sports a mask of virtue, we should know which to choose.

ÖйúAƬ

This is an abridged version of Tzvetan Todorov's lecture for the Oxford Amnesty Lecture 2001 series, co-sponsored by The THES , which continues until February 23. Todorov is director of research at the Centre National de Recherches in Paris, France.