For lasting peace, the colonial powers must leave the warring nations to find their own solutions: most conflicts are about internal failure, not simple border quarrels.

The precarious system of states bequeathed to Africa by the colonial powers is disintegrating fast, with domestic instability increasingly leading to inter-state conflicts. War in the Democratic Republic of the Congo has sucked six other countries into its maelstrom. Civil conflicts in Sudan, Uganda, Rwanda, Congo-Brazzaville, Sierra Leone, Guinea-Bissau and Lesotho have also attracted the military intervention of neighbouring states. As alarm mounts over these new African wars, it is time for the international community to step back, recognise that 40 years of post-colonial intervention have often done more harm than good, and for once do little - not out of indecision, but because it is the most helpful thing to do.

Africans are no longer playing by the rules they established for themselves when they formed the Organisation of African Unity in 1963: inviolability of colonial borders and non-interference in each other's internal affairs. This shift in behaviour is reviving old fears of chaos on the part of the international community. Not that anybody has anything good to say about colonial borders: drawn on a map in Berlin in 1885, they are artificial lines dividing ethnic groups and old kingdoms, and generally playing havoc with African societies. But they are the only borders Africa has. Start questioning them, common wisdom assumes, and a Pandora's box of conflicts will fly open.

But Africans are not fighting over boundaries. Wars are raging because many states have become hollow entities. Governments cannot exercise basic control over their territories, let alone carry out other functions of a modern state.  There are some 15 active conflicts in Africa today, but only that between Ethiopia and Eritrea can be properly described as a border conflict. The problem is not boundaries but state failure.

Take the Democratic Republic of the Congo. It has not had an effective government for 30 years. By the end of the long reign of Mobutu Sese Seko, the government could not maintain security, provide services, or pay civil servants and soldiers; Mobutu was overthrown in 1997 by a weak military movement with even weaker political roots, led by an incompetent leader, Laurent Kabila. History is repeating itself: once again, a weak rebel movement has been able seize control over one-third of the country in a few months.

Or take Sierra Leone, a small diamond-rich country where a feckless elected government, propped up by Nigerian troops and international support, is battling a brutal rebel movement fuelled by diamonds and Liberian military aid. These are not states, but vacuums in which conflicts fester.

Does it matter why these wars are raging? The humanitarian consequences are the same, no matter what the cause: people die horribly and in large numbers. The economic consequences are the same: devastation for the modern economy the subsistence economy of the peasants and the informal economy of the urban poor.

But causes matter because they affect possible solutions. The international community harbours the dangerous illusion that conflicts resulting from state failure can be handled like any other. Border conflicts between states, or even civil wars between strong states and strong movements, are relatively simple. They are the continuation of politics and diplomacy by other means; mediators can help lead these conflicts back into those realms. Conflicts in imploded states are a different matter.

Such conflicts are more diffuse, the sides less well defined, the goals unclear. The choices open to the international community are stark and unpalatable.

One is to continue the present course: promotion of negotiations, small-scale interventions of short duration and no clear goals. The problem is not only that such policies are ineffective when state failure is at the root of the problem; worse, they can actually do much harm. Mediation attempts among groups with no goal but their own power prolong conflicts as internationally negotiated cease-fires give the two sides time to regroup, rearm and start fighting again. Angola is a case in point. Ten years of internationally sponsored negotiations and agreements have merely led to more conflict. Even humanitarian assistance can prove dangerous at times, fuelling conflict, as it did in Somalia and is probably doing in Sudan. Continuing on this course is not a viable choice.

The second option is to impose order from the outside. Realistically, this means intervening militarily, not persuading states to abide by the OAU rules, because these states are beyond abiding by rules. But the conflicts are too numerous for intervention, and some involve vast multi-ethnic countries (such as the Congo, Sudan and, if elections fail, potentially Nigeria). Even if peace could be imposed, the international community would have to remain for a long time to rebuild a new system. The colonial powers called this pacification. It is not a route outsiders should travel again.

The third option is to do nothing except seek to limit the supply of arms to all combatants in the hope that either one side will prevail sufficiently to reconstruct a state, or that the opponents will reach a stalemate forcing them to seek an accommodation in good faith. Intervention should be limited to the most extreme situations such as Rwanda in 1994, where the necessity of stopping crimes against humanity superseded questions about the long-term outcome of intervention.

The third option is not particularly attractive, but it is the best. Time has come to accept the limitations of what the international community can do, take stock of the damage that intervention can cause, and sit on the sidelines as the old order crumbles. It was not a particularly good order, it never worked without outside intervention, and it is not worth restoring. The only sustainable order in the long run is one Africans establish and maintain themselves. And if some countries break up in the process, if borders change, if new entities appear, that is simply the march of history, not a catastrophe to be prevented at all costs.