There’s a lot to recommend in Peter Beinart’s column on McChrystal yesterday, but this paragraph gets something important wrong:

Obama’s problem isn’t that McChrystal is talking smack about him. His problem is that McChrystal isn’t pursuing his foreign policy.

Well, now McChrystal is gone and Petraeus is in and the policy is the same. So you could conclude, as Matthew Yglesias does, that there’s a broader circle in and around the military, that isn’t pursuing Obama’s foreign policy. Or you can conclude, as I think you have to, that in Afghanistan, Obama isn’t pursuing Obama’s foreign policy.

The Obama Doctrine, as laid out in the National Security Strategy, is about sustaining American power through a rules-based international order. Any strategy has tensions, because the world is never a neat and orderly place upon which a coherent intellectual blanket can be placed. In Obama’s case, the principal tensions concern what to do about violent extremism, since any security strategy has to address the most acute sources of immediate threats. And the greatest area of tension surrounds the war, an enterprise that he seeks to mitigate through counterinsurgency and then, after counterinsurgency, a very gradual extrication of military power giving way to long-term positive-sum relationships with Afghanistan and Pakistan. So it’s fair to conclude that the strategy sees the war — and the drone strikes in Pakistan, Yemen, Somalia, etc. — as near-term emergencies to address in less-than-desirable ways. But then after the unpleasantness recedes, the administration can fold the areas of crisis, from Afghanistan to Pakistan to Iraq to Yemen to Somalia, back into a rules-based international order and the doctrine can proceed on course.

Does that sound massively optimistic to anyone else? My personal view is that the administration has responded to the mess it inherited in Afghanistan by designing and implementing the least-worst strategy on offer. And that strategy still might be doomed. But strategy can’t treat its exceptions as temporary deviations that will give way to inevitable harmonization. It has to be up front that sometimes, the strategy either just doesn’t apply, and it might not ever apply in the case of a persistent conflict, or that the strategy has to change to accommodate such persistent exceptions and reexamine its assumptions.

I suppose the other option is to embrace incoherence in strategy-making, treating different areas of the world and different issues in whatever ad hoc manner is on offer and shrugging your shoulders when observers decry your inconsistency. While that might have the benefit of convenience, it’s no way to run a railroad. And there’s been way too much of it, in practice, across decades of American foreign policy.