In the media celebration of our "victory" over the Taliban in the Helmand Valley, little attention has been given to the nature of insurgency: the proper tactic of guerrillas is to fade away before overwhelming power, leaving behind only enough fighters to force the invaders to harm civilians and damage property. This is exactly what happened in the recent fighting in Marja. Faced with odds of perhaps 20 to 1, helicopters, tanks and bombers, the guerrillas wisely dispersed. Victory may not be quite the right description.

That battle will probably be repeated in Kandahar, which, unlike the agricultural area known as Marja, is a large and densely populated city. Other operations are planned, so the Marja "victory" has set a pattern that accentuates military action. This is not conducive to an exit strategy–it will not lead out of Afghanistan but deeper into the country. Indeed, there is already evidence that this is happening. As the Washington Post reported shortly after the Marja battle ended, not far away "the Marines are constructing a vast base on the outskirts of town that will have two airstrips, an advanced combat hospital, a post office, a large convenience store and rows of housing trailers stretching as far as the eye can see."

Since the Helmand Valley is the focal point of the military strategy, it is important to understand its role in Afghan affairs. The Helmand irrigation project, begun in the Eisenhower administration as a distant echo of the TVA, was supposed to become a prosperous island of democracy and progress. As a member of the Policy Planning Council in the Kennedy administration, I visited it in 1962. What I found was deeply disturbing: no studies had been made of the land to be developed, which proved to have a sheet of impermeable rock just below the surface that caused the soil to turn saline when irrigated; the land was not sufficiently leveled, so irrigation was inefficient; nothing was done to teach the nomad settlers how to farm; plots were too small to foster the social engineering aim of creating a middle class; and since there were no credit facilities to buy seed, settlers were paying 100 percent interest to moneylenders. In short, after the buildup of great expectations, disappointment was palpable.

Was it a portent? It seems likely. At the least, it’s striking that precisely where we carried out our first civic action program is where the Taliban became most powerful.

So what should that experience have taught us? That we should learn about the Afghans, their country and their objectives before determining our policy toward them. There is much to be learned, but I will here highlight what I believe are the three crucial issues that will make or break our relationship.

The first issue critical to evaluating US policy is the way the Afghans govern themselves. About four in five Afghans live in the country’s 20,000-plus villages. During a 2,000-mile trip around the country by jeep, horseback and plane half a century ago, as well as in later trips, it became clear to me that Afghanistan is really thousands of villages, and each of them, although culturally related to its neighbors, is more or less politically independent and economically autarkic.

This lack of national cohesion thwarted the Russians during their occupation: they won many military victories, and through their civic action programs they actually won over many of the villages, but they could never find or create an organization with which to make peace. Baldly put, no one could surrender the rest. Thus, over the decade of their involvement, the Russians won almost every battle and occupied at one time or another virtually every inch of the country, but they lost about 15,000 soldiers–and the war. When they gave up and left, the Afghans resumed their traditional way of life.

That way of life is embedded in a social code (known in the Pashtun areas as Pashtunwali) that shapes the particular form of Islam they have practiced for centuries and, indeed, that existed long before the coming of Islam. While there are, of course, notable differences in the Pashtun, Hazara, Uzbek and Tajik areas, shared tradition determines how all Afghans govern themselves and react to foreigners.

Among the shared cultural and political forms are town councils (known in the Pashtun areas as jirgas and in the Hazara area as ulus or shuras). The members are not elected but are accorded their status by consensus. These town councils are not, in our sense of the word, institutions; rather, they are "occasions." They come together when pressing issues cannot be resolved by the local headman or respected religious figure. Town councils are the Afghan version of participatory democracy, and when they act they are seen to embody the "way" of their communities.

Pashtunwali demands protection (melmastia) of visitors. Not to protect a guest is so grievous a sin and so blatant a sign of humiliation that a man would rather die than fail. This, of course, has prevented the Afghans from surrendering Osama bin Laden. Inability to reconcile our demands with their customs has been at the heart of our struggle for the past eight years.

As put forth in both the Bush and Obama administrations, our objective is to prevent Al Qaeda from using Afghanistan as a base for attacks on us. We sharpened this objective to the capturing or killing of bin Laden. That is popular with US voters, but even if we could force the Afghans to surrender him, it would alienate the dominant Pashtun community. Thus it would probably increase the danger to us. But it is unnecessary, since a resolution of this dilemma in our favor has been available for years. While Pashtunwali does not permit a protected guest to be surrendered, it allows the host, with honor, to prevent the guest from engaging in actions that endanger the host. In the past, the Taliban virtually imprisoned bin Laden, and they have repeatedly offered–provided we agree to leave their country–to meet our demand that Al Qaeda not be allowed to use Afghanistan as a base. Although setting a withdrawal date would enable us to meet our objective, we have turned down their offers.

The second crucial issue in evaluating our policy is the way the people react to our civic action programs.

Afghanistan is a barren, landlocked country with few resources, and its people have suffered through virtually continuous war for thirty years. Many are wounded or sick, with some even on the brink of starvation. The statistics are appalling: more than one in three subsists on the equivalent of less than 45 cents a day, almost one in two lives below the poverty line and more than one in two preschool children is stunted because of malnutrition. They are the lucky ones; one in five dies before the age of 5. Obviously, the Afghans need help, so we think they should welcome our efforts to aid them. But independent observers have found that they do not. Based on some 400 interviews, a team of Tufts University researchers found that "Afghan perceptions of aid and aid actors are overwhelmingly negative." We must ask why this is.

The reason, I think, is that the Taliban understand from our pronouncements that civic action is a form of warfare. The Russians taught them about civic action long ago, and Gen. David Petraeus specifically proclaimed in his Iraq days, "Money is my most important ammunition in this war." Thus many ordinary citizens see our programs as Petraeus described them–as a method of control or conquest–and so support or at least tolerate the Taliban when they destroy our projects or prevent our aid distribution.

To get perspective on this, it is useful to look at Vietnam. There too we found that the people resented our efforts and often sided with our enemies, the local equivalent of the Taliban: the Vietminh, or, as we called them, the Vietcong. The Vietminh killed officials, teachers and doctors, and destroyed even beneficial works. Foreigners thought their violence was bound to make the people hate them. It didn’t. Like the Kabul government, the South Vietnamese regime was so corrupt and predatory that few supported it even to get aid. When we "inherited" the war in Vietnam, we thought we should sideline the corrupt regime, so we used our own officials to deliver aid directly to the villagers. It got through, but our delivering it further weakened the South Vietnamese government’s rapport with its people.

Is this relevant to Afghanistan? Reflect on the term used by Gen. Stanley McChrystal when his troops moved into Helmand: he said he was bringing the inhabitants a "government in a box, ready to roll in." That government is a mix of Americans and American-selected Afghans, neither sent by the nominal national government in Kabul nor sanctioned by local authorities.

How will the Afghans react to McChrystal’s government? President Karzai was at least initially opposed, seeing the move as undercutting the authority of his government. We don’t yet know what the inhabitants thought. But we do know that when we tried similar counterinsurgency tactics in Vietnam, as the editor of the massive collection of our official reports, the Pentagon Papers, commented, "all failed dismally."

If we aim to create and leave behind a reasonably secure society in Afghanistan, we must abandon this failed policy and set a firm and reasonably prompt date for withdrawal. Only thus can we dissociate humanitarian aid from counterinsurgency warfare. This is because once a timetable is clearly announced, a fundamental transformation will begin in the political psychology of our relationship. The Afghans will have no reason (or progressively less reason, as withdrawal begins to be carried out) to regard our aid as a counterinsurgency tactic. At that point, beneficial projects will become acceptable to the local jirgas, whose members naturally focus on their own and their neighbors’ prosperity and health. They will then eagerly seek and protect what they now allow the Taliban to destroy.

If under this different circumstance the Taliban try to destroy what the town councils have come to see as beneficial, the councils will cease to provide the active or passive support, sanctuary and information that make the Taliban effective. Without that cooperation, as Mao Zedong long ago told us, they will be like fish with no water in which to swim. Thus, setting a firm and clear date for withdrawal is essential.

This leaves us with the third issue, the central government. We chose it and we pay for it. But as our ambassador, Gen. Karl Eikenberry, has pointed out in leaked reports, it is so dishonest it cannot be a strategic partner. It is hopelessly corrupt, and its election last year was fraudulent; General Petraeus even told President Obama that it is a "crime syndicate." It is important to understand why it lacks legitimacy in the eyes of its people.

For us, the answer seemed simple: a government must legitimize itself the way we legitimize ours, with a reasonably fair election. But our way is not the Afghan way. Their way is through a process of achieving consensus that ultimately must be approved by the supreme council of state, the loya jirga. The apex of a pyramid of village, tribal and provincial assemblies, the loya jirga, according to the Constitution, is "the highest manifestation of the will of the people of Afghanistan."

Like the Russians, we have opposed moves to allow Afghanistan to bring about a national consensus. In 2002 nearly two-thirds of the delegates to a loya jirga signed a petition to make the exiled king, Zahir Shah, president of an interim government to give time for Afghans to work out their future. But we had already decided that Hamid Karzai was "our man in Kabul." So, as research professor Thomas Johnson and former foreign service officer in Afghanistan Chris Mason wrote last year, "massive US interference behind the scenes in the form of bribes, secret deals, and arm twisting got the US-backed candidate for the job, Hamid Karzai, installed instead…. This was the Afghan equivalent of the 1964 Diem Coup in Vietnam: afterward, there was no possibility of creating a stable secular government." An interim Afghan government certified by the loya jirga would have allowed the traditional way to achieve consensus; but, as Selig Harrison reported, our ambassador at the time, Zalmay Khalilzad, "had a bitter 40-minute showdown with the king, who then withdrew his candidacy." We have suffered with the results ever since.

Could we reverse this downward trend? If we remove our opposition to a loya jirga, will the Kabul government respond? Probably not so long as America is willing to pay its officials and protect them. But if we set a clear timetable for withdrawal, members of the government will have a strong self-interest in espousing what they will see as the national cause, and they will call for a loya jirga. Indeed, President Karzai already has.

Would such a move turn Afghanistan over to the Taliban? Realistically, we must anticipate that many, perhaps even a majority, of the delegates, particularly in the Pashtun area, will be at least passive supporters of the Taliban. I do not see any way this can be avoided. Our attempts to win over the "moderates" while fighting the "hardliners" is an echo of what we tried in Vietnam. It did not work there and did not work for the Russians in Afghanistan. It shows no sign of working for us now. As a 2009 Carnegie Endowment study of our occupation and the Taliban reaction to it laid out, even after their bloody defeat in 2001, "there have been no splinter groups since its emergence, except locally with no strategic consequences."

Nor, as I have shown in my history of two centuries of insurgencies, Violent Politics, are we likely to defeat the insurgents. Natives eventually wear down foreigners. The Obama administration apparently accepts this prediction. As the Washington Post reported this past fall, it admits that "the Taliban cannot be eliminated as a political or military movement, regardless of how many combat forces are sent into battle."

A loya jirga held soon is the best hope to create a reasonably balanced national government. This is partly because in the run-up to the national loya jirga, local groups will struggle to enhance or protect local interests. Their action will constitute a brake on the Taliban, who will be impelled to compromise. Today the Taliban enjoy the aura of national defenders against us; once we are no longer a target, that aura will fade.

If we are smart enough to allow the Afghans to solve their problems in their own way rather than try to force them to adopt ours, we can begin a sustainable move toward peace and security. Withdrawal is the essential first step. Further fighting will only multiply the cost to us and lead to failure.