In its first weeks in office, the Obama administration has made two major decisions regarding Afghanistan. American combat forces have been increased by 50 percent, and a distinguished ambassador, Richard Holbrooke, has been appointed as presidential representative to what has been designated as the AFPAK region (implying that Afghanistan and Pakistan are being treated as a single geopolitical unit).

But the outcome will depend on the strategy with which we will face the inevitable complexities. The central Islamist challenge has moved to the mountainous Pashtun tribal area along the Pakistan-Afghanistan border, where sanctuaries on the Pakistan side of the border supply and train the assault on Afghanistan and the allied forces assisting it.

No guerrilla war has ever been won in the face of sanctuaries immune to attack. The administration is therefore right in dealing with it as a single problem. But it is also the case that the sanctuaries exist less by the design of the Pakistan government than by its political and military inability to control the territory along the Afghan border, which has never been under civil administration – even during British rule.

The Obama administration faces dilemmas familiar to several of its predecessors. America cannot withdraw now, but neither can it sustain the strategy that brought us to this point.

Heretofore America has pursued traditional anti-insurgency tactics: to create a central government, help it extend its authority over the entire country, and in the process bring about a modern bureaucratic and democratic society.

That strategy cannot succeed in Afghanistan – especially not as an essentially solitary effort. The country is too large for it, the territory too forbidding, the ethnic composition too varied, the population too heavily armed. No foreign conqueror has ever succeeded in occupying Afghanistan.

Even attempts to establish centralized Afghan control have rarely succeeded and then not for long. Afghans seem to define their country in terms of a common dedication to independence but not to unitary or centralized self-government. Once free of foreign forces, the various ethnic and regional groups would resume their autonomies, only reluctantly submitting to central authority and only in a limited way.

Many of President Hamid Karzai's difficulties are structural. I am uneasy about the ostentatious dissociation from a leader in the middle of a civil war and one whom we helped put into office. Without an obvious replacement, our previous such enterprises usually backfired.

The truism that the war is in effect a battle for the hearts and minds of the Afghan population is valid enough in concept. The low standard of living of much of the population has been exacerbated by 30 years of civil war. The economy is on the verge of sustaining itself by the sale of narcotics. There is no significant democratic tradition.

Reform is a moral necessity. But the timescale for reform is out of phase with the imperatives of anti-guerrilla warfare. It will require decades; it should occur as a result of, and even side-by-side with, the attainment of security but cannot be the precondition for it.

The military effort will inevitably unfold at a pace different from the political evolution of the country. But what we are able to accomplish immediately is to make sure that our aid efforts, now diffuse and inefficient, are coherent and relevant to popular needs. And much greater emphasis should be given to local and regional entities.

Military strategy should concentrate on preventing the emergence of a coherent, contiguous state within the state controlled by jihadists. In practice, this means control of Kabul and the Pashtun area. A jihadist base area on both sides of the Afghan-Pakistani border would become a permanent threat to the hopes of a moderate evolution and to all of Afghanistan's neighbors.

General David Petraeus has argued that, reinforced by the numbers of American forces he has recommended, he should be able to control the 10 percent of Afghan territory where, in his words, 80 percent of the military threat originates. This is the region where the clear hold-and-build strategy that had success in Iraq is particularly applicable.

In the rest of the country, our military strategy should be more fluid, aimed at forestalling the emergence of any terrorist strong points. It should be based on close cooperation with local chiefs and coordination with their militias to be trained by U.S. forces – the kind of strategy that proved so successful in Anbar Province, the Sunni stronghold in Iraq.

This is a plausible strategy, though it seems improbable that the 17,000 reinforcements are enough. In the end, the fundamental issue is not so much how the war will be conducted but how it will be ended.

Afghanistan is almost the archetypal international problem requiring a multilateral solution for the emergence of a political framework. In the 19th century, formal neutrality was sometimes negotiated to impose a standstill on interventions in and from strategically located countries. This did not always survive, but it provided a framework for defusing day-to-day international relations. (Belgian neutrality, for example, was not challenged for nearly 100 years.) Is it possible to devise a modern equivalent?

In Afghanistan, such an outcome is achievable only if Afghanistan's principal neighbors agree on a policy of restraint and opposition to terrorism. Their recent conduct would argue against such prospects. Yet history should teach them that unilateral efforts at dominance are likely to fail in the face of countervailing intervention by other outside actors.

To explore such a vision, the United States should propose a working group composed of Afghanistan's neighbors, India and the permanent members of the Security Council. It should be charged with assisting in the reconstruction and reform of Afghanistan and establishing principles for the country's international status and obligations to oppose terrorist activities.

Over time the unilateral military efforts of America can merge with the diplomatic efforts of this group. As the strategy envisaged by Petraeus succeeds, the prospects for a political solution along these lines grow correspondingly.

The precondition for such a policy is cooperation with Russia and Pakistan. With Russia, it requires a clear definition of priorities, especially a choice between partnership or adversarial conduct insofar as it depends on us.

The conduct of Pakistan will be crucial. Pakistan's leaders must face the fact that continued toleration of the sanctuaries – or continued impotence with respect to them – must draw their country ever deeper into an international maelstrom.

If the jihadists were to prevail, Pakistan would surely be the next target – as is observable already along the existing borders and even in the Swat valley close to Islamabad. If that were to happen, the affected countries would need to begin to consult with each other about the implications for them of the nuclear arsenal of a Pakistan in the process of being engulfed or even threatened by jihadists.

Like every country engaged in Afghanistan, Pakistan has decisions to make that will affect its international position for decades.

Other countries face comparable choices – especially our NATO allies. Symbolically the participation of NATO partners is significant. But save for some notable exceptions, public support for military operations is negligible in almost all NATO countries.

It is possible of course that Obama's popularity in Europe can modify these attitudes, but in all likelihood to only a very limited extent. The president will then have to decide how far he will carry the inevitable differences and face the reality that disagreements concern fundamental questions of NATO's future and reach.

Improved consultation will ease the process. It is likely to turn out, however, that the differences are not procedural. We may then conclude that an enhanced NATO contribution to Afghanistan's reconstruction is more useful than a marginal military effort constrained by caveats. But if NATO turns into an alliance à la carte in this manner, a precedent is set that can cut both ways. Those who tempt an American withdrawal by their indifference or irresolution evade the prospect that it will be the prelude to a long series of accelerating and escalating crises.

The new team for Afghanistan faces daunting choices. Whatever strategy they select needs to be pursued with determination. It is not possible to hedge against failure by half-hearted execution.

Henry A. Kissinger served as national security adviser and as secretary of state in the administrations of Presidents Nixon and Ford. Distributed by Tribune Media Services.