First published in The New York Review of Books, 17 December 2009.

Cool poker-players, we are tempted to believe, only raise or fold: they only increase their bet or leave the game. Calling, making the minimum bet to stay, suggests that you can’t calculate the odds or face losing the pot, and that the other players are intimidating you. Calling is for children. Real men and women don’t want to call in Afghanistan: they want to dramatically increase troops and expenditure, defeat the Taliban, and leave. Or they just want to leave. Both sides—the disciples of the surge and the apostles of withdrawal—therefore found some satisfaction in one passage in President Obama’s speech at West Point on December 1:

I have determined that it is in our vital national interest to send an additional 30,000 US troops to Afghanistan. After 18 months, our troops will begin to come home.

But the rest left them uneasy. This was not, as they might have imagined, because he was lurching between two contradictory doctrines of increase and withdrawal, but because the rest of his speech argued for a radically different strategy—a call strategy—which is about neither surge nor exit but about a much-reduced and longer-term presence in the country. The President did not make this explicit. But this will almost certainly be the long-term strategy of the US and its allies. And he has with remarkable courage and scrupulousness articulated the premises that lead to this conclusion. First, however, it is necessary to summarize the history of our involvement and the conventional policies that have long favored surge and exit.

A legion of arguments almost drove Obama away from this new moderate position over the last ten weeks of discussion. There was our general fear in Afghanistan and Pakistan of the modern demons, which policy experts dub “insurgency, terrorism, civil war, human rights–abusing warlords, narcotics, weapons of mass destruction, and global jihad” and the spawn of “safe havens, rogue, fragile, and failed states.” There was our developing sense, over the last eight years, that the status quo was unacceptable.

From 2001, sections of the international community attempted to assist the Afghan government in the construction of a state. The British Department for International Development put 80 percent of its funds into direct budgetary support for the Afghan government and NGOs implemented health, education, and rural development projects as contractors for the Afghan government. Such efforts were described by NATO as a “comprehensive approach to security, governance and economic development” in which the UN, an apparently benevolent Karzai government, NATO, and the NGOs would all play their part—largely in concert because there was no perceived conflict between their aims and values.

Challenges from warlords, druglords, lack of funds, and lack of government authority were to be met through cen- tralization, disarmament of opposition groups, crop eradication, coordination, and closer partnership. It was assumed that it would be possible within a reasonable time (some documents claimed within seven years) to build a stable centralized state, largely independent of foreign support, arranged around the rule of law and a technocratic administration, with a vibrant economy based on lawful commerce and trade. Few expected the Taliban to reemerge. Comparisons were drawn with the development of Korea or Singapore.

Eight years later this seems a tragic fantasy. Frustrated by lack of progress, the US and its allies have oscillated giddily between contradictory policies. The British government that once championed more generous budgetary support for the Kabul government now portrays it as corrupt, semi-criminal, ineffective, and illegitimate. “Warlords” such as Gul Agha Shirzai, who we once demonized, are now tolerated or even praised, and are almost certain to be given good positions in the new Karzai government. We armed militias in 2001, disarmed them through a demobilization program in 2003, and rearmed them again in 2006 as community defense forces. We allowed local autonomy in 2001, pushed for a strong central government in 2003, and returned to decentralization in 2006. First we tolerated opium crops; then we proposed to eradicate them through aerial spraying; now we expect to live with opium production for decades.

Meanwhile, the Karzai government and the nations involved in Afghanistan have fallen into a cruel and dysfunctional arranged marriage that seems too often to lack common values, common projects, trust, and even patience. Each undermines the other’s legitimacy. NATO is blamed for being associated with a corrupt and illegitimate administration; the Karzai government is blamed by Afghans for bombarding civilians and for accepting the support of foreign infidels. And each has sought to shift blame to the other side.

Many of these tensions were illustrated in the first week of November: five British soldiers were killed by the Afghan policeman they were training; nine Afghan policemen, trying to come to the rescue of lost American servicemen, were killed by a coalition bomb; five UN election observers were killed by the Taliban in their Kabul guesthouse, causing the UN to begin to withdraw its staff. A PBS journalist interviewed President Karzai:

Margaret Warner: “The UN did reluctantly withdraw about two thirds of its foreign staff…. What impact is that likely to have?”

Hamid Karzai: No impact. No impact.

Margaret Warner: So you don’t care if they return?

Hamid Karzai: They may or may not return. Afghanistan won’t notice it. We wish them well wherever they are.

Even an optimist would now describe Afghanistan as a poor, dangerous country, struggling to survive in the face of jihadist ideology, insecurity, and poor governance. It is now hoped that good development in Afghanistan might allow it over decades to draw level with Pakistan. The Taliban have a growing presence even outside their traditional heartland in the south and east of Afghanistan and they mount attacks on previously safe areas and communities. Civil war is now seen as very likely. Comparisons are drawn with Somalia.

Through all these bewildering years, a subtle and refined edifice of justification for troop increases has emerged, in which arguments are categorized by type and family and reinforced with analogies and precedents, in a structure in which each claim supports another. The tone, history, and arguments in this liturgy are not only the product of soldiers, spies, explorers, journalists, administrators, writers, aid workers, professors, think-tank directors, and politicians. They have been developed by the great alliances of NATO and the UN and have drawn on World Bank economists, veterans of Iraq and the frontier, linguists with decades of experience in rural Afghanistan, and even, occasionally, Afghans. The creed, hammered out in the great international councils of Washington, Bonn, and Paris, runs as follows:

Afghanistan is an existential threat. It is the epicenter of international terrorism and the epitome of a failed state. We must fight in Afghanistan for six reasons: (1) to protect the United States and the rest of NATO from terrorist attack; (2) to protect Pakistan and the region; (3) to protect the credibility of the United States and NATO; (4) to protect the Afghan people; (5) to defeat the Taliban; and (6)to create an effective, legitimate, stable state.

Our enemies include corruption, drugs, poverty, and insecurity and we will address them through governance and capacity- building, alternative livelihoods, a regional solution, a comprehensive approach, and an exit strategy. The surge worked in Iraq. We have a moral obligation to the Afghan people. By abandoning them in 1989, we created the conditions that led to September 11. We must, therefore, implement counter-insurgency operations across the spectrum.

Just as Buddha’s fourth noble truth can be divided into an eightfold path, so each justification, need, ethical claim, doctrine, precedent, and analogy of this modern metaphysics can be further subdivided. Thus the article of faith that our operations in Afghanistan are crucial to the stability of Pakistan can conventionally be defended by reference to the need for a two-sided pincer movement against the Taliban on the border; worries about safe havens, failed states, and global jihad; the support for drone attacks in Pakistan conveyed in one opinion poll on the frontier and by one Pakistani general; the appearance of the Taliban “only sixty miles from Islamabad.” And the possibility that mad mullahs will seize the nukes.

Each argument echoes much deeper assumptions about the world: a belief in the moral imperative of humanitarian intervention, backed by our failures in Rwanda and our success in the Balkans; a maximal vision in which no one good (“security,” for example) can be achieved without the achievement of every other good (such as “development” or “the rule of law”); a rhetorical tradition in which all goods are seen as consistent and mutually reinforcing; and an Enlightenment faith that there is nothing intrinsically intractable about Afghan culture and society and that all men can be perfected (to a Western ideal) through the application of reason and the laws of social science.

But perhaps more importantly there are our more recent theories about the global order. There is the credit we take for the success of postwar Germany, democracy in Eastern Europe post-1989, and economic growth in South and East Asia. There are our apparent mistakes with Mossadeq in Iran in the 1950s; fighting in Vietnam in the 1960s, Latin America in the 1980s, and Somalia in the 1990s; the September 11 attacks; North Korea today; and the different lessons we have chosen to take about working against the popular will, supporting dictators, leaving, or failing to act. All of this experience is reflected in our division of the world into friendly, puppet, rogue, fragile, and failed states and our anxieties about instability, insurgency, terrorism, or weapons of mass destruction.

All these fears, frustrations, and doctrines contributed to the relentless logic that drove Obama to state, last year, “We must win in Afghanistan”; and to claim that Bush failed in Afghanistan because he did not invest enough resources. Even Obama’s latest speech began with the story of how Afghanistan fell and September 11 occurred because “the attention of America and our friends had turned elsewhere,” and the speech reminded us of “a nuclear-armed Pakistan,…NATO’s credibility,…failed states.”

Such arguments explain why he sent an extra 17,000 troops last March, insisting that “there is an uncompromising core of the Taliban. They must be met with force, and they must be defeated,” and he committed the US to “promote a more capable and accountable Afghan government” and “advance security, opportunity and justice.” This is also why he announced a more maximalist counterinsurgency strategy in the March White Paper and appointed a new commanding general, Stanley McChrystal, to implement it.

By agreeing to a counterinsurgency strategy, Obama implicitly committed to all the doctrine contained in a two-hundred-page field manual, derived from the analysis of seventy-three previous insurgencies. “Full-spectrum counter-insurgency,” or COIN, the President was informed in the manual, “is all-encompassing.” It is expressed in aphorisms such as “the center of gravity is the population” and “we are not being out-fought but out-governed”; and mottoes like “Clear, Hold, Build.” It includes economic development, infantry tactics, political negotiation, building capacity for governance, and eliminating “high-value” targets using predator drones. The soldiers, according to the COIN doctrine, need to have considerable cultural sensitivity, knowledge, and good fortune. They must work in close and constructive concert with a credible local government. They need to be able to control the borders and protect communities during the lengthy process of reconstruction.

It is almost impossible to say what counterinsurgency does not include. But it almost always requires more troops. I first heard almost a year ago that General Petraeus was pressing for another 40,000 troops. When I finally saw McChrystal in Kabul in October, he had completed his report and formally requested another 40,000 troops. Obama could not refuse the bulk of the general’s requests without being personally blamed for the future of Afghanistan.

Little wonder that some called (in the President’s words) “for a more dramatic and open-ended escalation of our war effort—one that would commit us to a nation-building project of up to a decade.” How could they ask for any other course when they argued from within a conceptual prison, founded on fears, boxed in by domestic political calculations, restricted by misleading definitions, buttressed by syllogisms, endorsed by generals, and crowned with historical analogies? Yet this is what the President said about full-scale escalation:

I reject this course because it sets goals that are beyond what can be achieved at a reasonable cost, and what we need to achieve to secure our interests. As President, I refuse to set goals that go beyond our responsibility, our means, or our interests. And I must weigh all of the challenges that our nation faces. I don’t have the luxury of committing to just one. Indeed, I’m mindful of the words of President Eisenhower, who—in discussing our national security—said, “Each proposal must be weighed in the light of a broader consideration: the need to maintain balance in and among national programs.”

I felt as though I had come to hear a fifteenth-century scholastic and found myself suddenly encountering Erasmus: someone not quite free of the peculiarities of the old way, and therefore haunted by its elisions, omissions, and contradictions; but already anticipating a reformation. Obama’s central—and revolutionary—claim is that our responsibility, our means, and our interests are finite in Afghanistan. As he says, “we can’t simply afford to ignore the price of these wars.” Instead of pursuing an Afghan policy for existential reasons—doing “whatever it takes” and “whatever it costs”—we should accept that there is a limit on what we can do. And we don’t have a moral obligation to do what we cannot do.

The US must husband its resources to meet other strategic challenges. Obama’s description of these is still narrowly focused on failed states and terrorism: it does not include the threats posed by states such as China or Russia, still less Iraq, Egypt, Lebanon, or Kashmir, and it does not attempt to compare the conflict in Afghanistan to the risks posed by climate change or threats to the supply of food in poor nations. But he names Pakistan, Yemen, and Somalia as posing challenges. The US responsibility to the Afghan people is only one responsibility among many and “the nation that I’m most interested in building is our own.” He emphasizes the competing demand of domestic priorities and costs:

Over the past several years, we have lost that balance. We’ve failed to appreciate the connection between our national security and our economy.

Or to return to poker, he argues that we have limited chips and the amount we stake in Afghanistan should reflect the amount we stand to gain and the likelihood of winning.

This may imply that Obama has given up and is in favor of a rapid exit. (I, for one, have rarely managed to convince anyone during the last four years that I can be both against troop increases and against withdrawal.) But Obama opposes precipitate withdrawal. He acknowledges that although “our responsibility, our means, or our interests” are limited, they exist in Afghanistan. We have a certain responsibility to the Afghan people who would suffer a civil war if we withdrew. This would initially be between the Taliban and the Karzai government, but it could expand (as it did in the 1990s) into more fragmented local conflicts, fueled by neighboring countries, in which no faction is strong enough to win or weak enough to give up the fight, and in which Afghans are plunged back into anarchy, cruel conflict, and poverty. We have the means, however, to make a positive contribution and we have an interest in preventing a defeat that would wreck our hopes, humiliate the United States and NATO, embolden our enemies, and weaken our allies (and not only in Pakistan). He implies that just because we cannot do everything does not mean we can do nothing.

Obama’s objectives in remaining in Afghanistan are as follows:

We must deny al Qaeda a safe haven. We must reverse the Taliban’s momentum and deny it the ability to overthrow the government. And we must strengthen the capacity of Afghanistan’s security forces and government so that they can take lead responsibility for Afghanistan’s future…. And we will also focus our assistance in areas—such as agriculture—that can make an immediate impact in the lives of the Afghan people.

In other words, he would continue to use intelligence and special forces to keep the pressure on Osama bin Laden. He would continue to deliver humanitarian assistance and economic development aid particularly to the many poor and neglected communities who want to work with us in the north and center of Afghanistan. In addition (which differentiates this model from the strictly counterterrorism approach), he would retain a sufficiently robust presence to prevent the Taliban from ever gathering an army or mounting a conventional threat or rolling artillery and tanks up the highway to take an Afghan city like Kabul. And combine US military presence with political action and incentives to keep tribal leaders and other regional power brokers on our side and away from the Taliban. And ultimately, through all these techniques, decrease the likelihood of civil war, increase the likelihood of a political settlement with the Taliban, and leave Afghanistan in twenty years’ time a more stable and prosperous country than it is today.

This strategy assumes that the Afghan Taliban are less of a threat to Pakistani stability and NATO than they appear. It also assumes that a counterinsurgency strategy and Iraq-style surge will not—on their own—succeed and a state-building strategy will not work. Obama still needs to find the language to express these insights without falling into the trap of withdrawal.

There are, in reality, no inescapable connections between Afghanistan and Pakistan, al-Qaeda and the Taliban. There are positive and negative effects of our Afghan operations on Pakistan, (positive, through increasing pressure on the Taliban; negative, through inflaming Islamist anti-US sentiment in Pakistan and driving “bad guys” over the border into Afghanistan). But the future of Pakistan will be determined predominantly by factors internal to Pakistan, such as the military, the feudal system, and the relationship between the institutions of Islam and the Pakistani state. Similarly, although al-Qaeda and the Taliban cooperate and share funding, they are still largely divided between a non-Afghan group focused on international terrorism and Afghan–Pakistani groups whose primary aim is to drive foreign troops from Afghanistan and spread Islamist rule in Pakistan. You could at least in theory defeat the Taliban without eliminating al-Qaeda, and the Taliban could return to power in Afghanistan without bringing back al-Qaeda.

The counterinsurgency strategy and surge in Iraq led to a drop in violence (against predictions), but the same will not happen in Afghanistan. The Iraq insurgency was the movement of a minority sectarian group, the Sunnis, whose supporters have been driven from most of the neighborhoods in the capital city and whose leaders were tribal figures with a long-standing relationship to the central government. The Shia-dominated Baghdad government was a powerful, credible force, from the majority ethnic and sectarian group, and was supported by mass political parties, with their own militias. The challenge for Petraeus and his predecessors in Iraq was to grasp this political opportunity; provide support, money, and status to the losing Sunni groups to separate them from al- Qaeda; and convince Nouri al-Maliki to disengage from some of the Shia militias and endorse the settlement. In Afghanistan, neither the Karzai government nor the Taliban have the history, the structure, or the incentives to foster such a deal.

Afghanistan contains a diffuse rural insurgency spread among a population of 30 million people, 80 percent of whom are scattered among 20,000 remote, often mountainous villages. It is different from Iraq, where the insurgency was largely centered around the flat urban areas surrounding Baghdad. Nor is it like the much smaller Malaya of the 1950s, where the British in their antiguerrilla operations were able to move villagers to walled and guarded camps. At least half of Afghanistan (a country almost the size of Texas) is now threatened by insurgency, and the COIN doctrine requires sufficient troops to secure and protect the population areas.

This is why the architects of the COIN doctrine are calling for a ratio of one “trained counterinsurgent” (a category that includes Afghans, if they have been given the necessary skills) for every fifty members of the population or a combined total that would amount in Afghanistan to 600,000 troops, if they intended to cover the country (though most theorists believe it is only necessary to cover half). The effective, legitimate Afghan government, on which the entire counterinsurgency strategy depends, shows little sign of emerging, in part because the international community lacks the skills, the knowledge, the legitimacy, or the patience to build a new nation. In short, COIN won’t work on its own terms because of the lack of numbers and a credible Afghan partner and in absolute terms because of the difficulties of the country and its political structures.

But equally history does not doom the allies to absolute failure. The situation may not be that of Iraq in 2006 or Afghanistan in 1988, but neither is it Afghanistan in 1842, still less in 330 BC (even if we actually understood the victories of the Victorians or Alexander). Pakistan may not be a failed state and mullahs may not be a hand’s breadth from its nukes; but Pakistan is facing serious instability and a moderate, constructive policy in Afghanistan could at least prevent Afghanistan from con- tributing further to its instability. The US and its NATO allies would be able to survive withdrawal from Afghanistan but it would be damaging to their reputations. While we cannot write a blank check to Afghans, we would like to prevent their country from falling into civil war, which would probably result in tens of thousands of deaths. It makes sense to stay, if we can maintain a realistic, affordable, and legitimate presence in Afghanistan and do some good.

It is difficult to find the appropriate language to express such insights. A moderate, light policy runs against a natural tendency to invest extravagantly in defending against even minor threats to our national security (the reverse of our systematic tendency to “lowball,” i.e., to undercompensate for, or underprice, risk in our banking system or the environment). This partly reflects a general, ancient view of the “night watchman” state, involved not in internal regulation but in security. It is partly because terrorism seems a much more immediate and horrifying prospect than financial collapse, climate change, or threats to food security and is more directly linked to loss of life (even if the other issues ultimately may kill many more people). And our culture puts a very high value on life (though a higher value on the lives of our own citizens than on those of other nationals).

We would prefer, therefore, to believe that any war in which we engage is a vital threat to our very existence—in which case the odds of victory are irrelevant and any sacrifice is justified. And there must be a defined end. It would be difficult for a president to argue that we should sacrifice lives without winning in order to prevent something worse (although we build dams when we can’t control the flow of water and employ a police force when we can’t end crime).

We would be revolted by someone who tried to calculate how many lives the objectives in Afghanistan were worth (fifty? a thousand?). And these are all healthy intuitions: we would not want to be in a world where lives were treated simply as units, to which we assigned a definite and explicit expendable value in a grand cost-benefit analysis. But these intuitions still reinforce an all-or-nothing approach to foreign policy.

The simple process of naming our past and present strategies already generates and restricts our response. Thus by naming operations in Afghanistan a counterinsurgency, we may feel compelled to deploy one trained counterinsurgent for every fifty members of the population; by labeling our approach “an Afghanistan–Pakistan strategy,” we imply that our actions in Afghanistan are vital to the security of Pakistan; by putting the Taliban in the category of those pursuing a global jihad, we conclude that we cannot negotiate with them; by naming Afghanistan a terrorist safe haven or a failed state, we conclude that failure (or even a light “footprint”) is not an option.

Obama deftly avoided all these words and traps in his speech, perhaps because he has become aware of their extreme implications. There was no talk of victory. His aim was no longer to defeat but to contain the Taliban: to “deny it the ability to overthrow the government.” He explicitly rejected a long “nation-building project.” He talked not of eliminating but of keeping the pressure on al-Qaeda. He did not speak of a moral obligation to the Afghan people. He did not specify any necessary logical connections between the Taliban, al-Qaeda, Afghanistan, and Pakistan. He asserted that “there’s no imminent threat of the [Afghan] government being overthrown.” He emphasized that “we will support efforts by the Afghan government to open the door to those Taliban who abandon violence and respect the human rights of their fellow citizens.” He did not draw parallels with the surge in Iraq. And most strikingly of all, whereas he had referred four times in March to insurgency, now he stated that “unlike Vietnam, we are not facing a broad-based popular insurgency.”

Such moderate analysis disappointed those who wanted a call to arms. The West Point cadets in the audience yawned, stared at the floor, and clapped only halfheartedly. Bush’s surge in Iraq was a troop increase of only 20 percent; Obama’s contributions to Afghanistan since he took office will more than double US troop presence on the ground. Bush spoke at a time of overwhelming public opposition to the war and with one of the lowest popularity ratings ever recorded; but it was Bush, not Obama, who spoke about determination, commitment, victory, and doing whatever it takes. Obama sounded like those he criticized for wanting to “simply maintain a status quo in which we muddle through.”

But this moderate tone gains Obama the leverage that Bush lacked. As long as the US asserted that Afghanistan was an existential threat, the front line in the war on terror, and that, therefore, failure was not an option, the US had no leverage over Karzai. The worse Afghanistan behaved—the more drugs it grew and terrorists it fostered—the more money it received. If it sorted out its act, it risked being relegated to a minor charitable recipient like Tajikistan. A senior Afghan official warned me this year “to stop referring to us as a humanitarian crisis: we must be the number one terrorist threat in the world, because if we are not we won’t get any money.” By asserting convincingly that Afghanistan is not the be-all and end-all and that the US could always ultimately withdraw, Obama escapes this codependent trap and regains some leverage over the Afghan government. In his politer words:

It must be clear that Afghans will have to take responsibility for their security, and that America has no interest in fighting an endless war in Afghanistan.

But perhaps even more importantly, defining a more moderate and limited strategy gives him leverage over his own generals. By refusing to endorse or use the language of counterinsurgency in the speech, he escapes their doctrinal logic. By no longer committing the US to defeating the Taliban or state-building, he dramatically reduces the objectives and the costs of the mission. By talking about costs, the fragility of public support, and other priorities, he reminds the generals why this surge must be the last. All of this serves to “cap” the troop increases at current levels and provide the justification for beginning to reduce numbers in 2011.

But the brilliance of its moderate arguments cannot overcome that statement about withdrawal. With seven words, “our troops will begin to come home,” he loses leverage over the Taliban, as well as leverage he had gained over Karzai and the generals. It is a cautious, lawyerly statement, expressed again as “[we will] begin the transfer of our forces out of Afghanistan in July of 2011.” It sets no final exit date or numbers. But the Afghan students who were watching the speech with me ignored these nuances and saw it only as departure.

This may be fatal for Obama’s ambition to “open the door” to the Taliban. The lighter, more political, and less but still robust militarized presence that his argument implies could facilitate a deal with the Taliban, if it appeared semi-permanent. As the President asserted, the Taliban are not that strong. They have nothing like the strength or appeal that they had in 1995. They cannot take the capital, let alone recapture the country. There is strong opposition to their presence, particularly in the center and the north of the country. Their only hope is to negotiate. But the Taliban need to acknowledge this. And the only way they will is if they believe that we are not going to allow the Kabul government to collapse.

Afghanistan has been above all a project not of force but of patience. It would take decades before Afghanistan achieved the political cohesion, stability, wealth, government structures, or even basic education levels of Pakistan. A political settlement requires a reasonably strong permanent government. The best argument against the surge, therefore, was never that a US operation without an adequate Afghan government partner would be unable to defeat the Taliban—though it won’t. Nor that the attempt to strengthen the US campaign will intensify resistance, though it may. Nor because such a deployment of over 100,000 troops at a cost of perhaps $100 billion a year would be completely disproportional to the US’s limited strategic interests and moral obligation in Afghanistan—though that too is true.

Instead, Obama should not have requested more troops because doing so intensifies opposition to the war in the US and Europe and accelerates the pace of withdrawal demanded by political pressures at home. To keep domestic consent for a long engagement we need to limit troop numbers and in particular limit our casualties. The surge is a Mephistophelian bargain, in which the President has gained force but lost time.

What can now be done to salvage the administration’s position? Obama has acquired leverage over the generals and some support from the public by making it clear that he will not increase troop strength further. He has gained leverage over Karzai by showing that he has options other than investing in Afghanistan. Now he needs to regain leverage over the Taliban by showing them that he is not about to abandon Afghanistan and that their best option is to negotiate. In short, he needs to follow his argument for a call strategy to its conclusion. The date of withdrawal should be recast as a time for reduction to a lighter, more sustainable, and more permanent presence. This is what the administration began to do in the days following the speech. As National Security Adviser General James Jones said, “That date is a ‘ramp’ rather than a cliff.” And as Hillary Clinton said in her congressional testimony on December 3, their real aim should be to “develop a long-term sustainable relationship with Afghanistan and Pakistan so that we do not repeat the mistakes of the past, primarily our abandonment of that region.”

A more realistic, affordable, and therefore sustainable presence would not make Afghanistan stable or predictable. It would be merely a small if necessary part of an Afghan political strategy. The US and its allies would only moderate, influence, and fund a strategy shaped and led by Afghans themselves. The aim would be to knit together different Afghan interests and allegiances sensitively enough to avoid alienating independent local groups, consistently enough to regain their trust, and robustly enough to restore the security and justice that Afghans demand and deserve from a national government.

What would this look like in practice? Probably a mess. It might involve a tricky coalition of people we refer to, respectively, as Islamists, progressive civil society, terrorists, warlords, learned technocrats, and village chiefs. Under a notionally democratic constitutional structure, it could be a rickety experiment with systems that might, like Afghanistan’s neighbors, include strong elements of religious or military rule. There is no way to predict what the Taliban might become or what authority a national government in Kabul could regain. Civil war would remain a possibility. But an intelligent, long-term, and tolerant partnership with the United States could reduce the likelihood of civil war and increase the likelihood of a political settlement. This is hardly the stuff of sound bites and political slogans. But it would be better for everyone than boom and bust, surge and flight. With the right patient leadership, a political strategy could leave Afghanistan in twenty years’ time more prosperous, stable, and humane than it is today. That would be excellent for Afghans and good for the world.

Meanwhile, Obama’s broader strategic argument must not be lost. He has grasped that the foreign policy of the president should not consist in a series of extravagant, brief, Manichaean battles, driven by exaggerated fears, grandiloquent promises, and fragile edifices of doctrine. Instead the foreign policy of a great power should be the responsible exercise of limited power and knowledge in concurrent situations of radical uncertainty. Obama, we may hope, will develop this elusive insight. And then it might become possible to find the right places in which to deploy the wealth, the courage, and the political capital of the United States. We might hope in South Asia, for example, for a lighter involvement in Afghanistan but a much greater focus on Kashmir.^*

I began by saying that “calling” in poker was childish and that grownups raise or fold. But there is another category of people who raise or fold: those who are anxious to leave the table. They go all in to exit, hoping to get lucky but if not then at least to finish. They do not do this on the basis of their cards or the pot. They do it because they lack the patience, the interest, the focus, or the confidence to pace themselves carefully through the long and exhausting hours. They no longer care enough about the game. Obama is a famously keen poker player. He should never be in a hurry to leave the table.