The Washington Post's oped page is a safe haven for hawks.  Today we have Michael O'Hanlon and Richard Cohen fighting for the war in Afghanistan.

O'Hanlon is for generals respecting the president's policy decisions, except when he isn't -- cases where the general is obviously right, in that he agrees with O'Hanlon. (To me, this McChrystal incident shows the robustness of civilian control. McChrystal spoke too freely and got rebuked. The Republic seems OK. So does the Army.)

O'Hanlon's other goal is to attack those who want to limit the objectives in Afghanistan to counter-terrorism. To do so, he imputes his nation-building goals to the less ambitious strategy. He says we tried the narrow mission under Bush and it failed.

A. Not really. Does this, for example, sound like counter-terrorism?

B. It only failed to achieve the counterinsurgency strategy's (maybe impossible) objectives of a stable, centralized state in Afghanistan. A counter-terrorism (or go small) strategy sacrifices some probability of heightened stability for less cost in blood and dollars. We have been doing fine at counter-terrorism all along, largely because al Qaeda is overrated. Afghanistan is not a terrorist haven anymore.

O'Hanlon also says that we won't collect as much intelligence without a full-scale counterinsurgency. Again, this is true, but insufficient. A smaller footprint provides benefits (less radicalization, less cost) that we exchange, in a sense, for lost intelligence-gathering opportunities. In any case, intelligence needed to target airstrikes can come from allies on the ground, intercepts, and overhead surveillance, as in Pakistan. Progress in surveillance and strike capability and the will to use it means that a rerun of the 1990s, where al Qaeda was safe in Afghan camps, is a phony nightmare. O'Hanlon also claims that absent a large U.S. ground force, we would have to offshore all UAV bases that range western Pakistan. This is a pessimistic assessment; we could defend most of our airfields with a limited force in Afghanistan, and we have at least one UAV base in Pakistan.

Cohen calls Obama soft for letting McChrystal run amok, ignoring the fact that both the Secretary of Defense and the National Security Adviser publicly rebuked him.  Cohen approvingly cites Obama's foolish claim that Afghanistan is a war of necessity. One can't say enough that this is senseless; even wars of pure self-defense aren't strictly necessary, and Afghanistan, at this point, isn't that.  He then drops the dominos. Should we leave, he says, the Taliban will take over Afghanistan and then Pakistan, grabbing nukes. India then invades Pakistan, and we get 1947, but nuclear. He doesn't say how the Taliban columns advancing on Kabul will suppress our airpower. The widespread Afghan and Pakistani hostility to the Taliban -- especially among the non-Pashtuns who support and dominate both governments -- doesn't impress him. He doesn't mention the fact that the Pakistani military keeps close hold on its nukes, no matter who is officially in power.  One could go on, but suffice it to say that there is an equally plausible worst-case scenario that results from following Cohen's advice and expanding the war.

To be fair though, Cohen is a clear-eyed realist compared to Daniel Twining, who writes for Foreign Policy's Shadow Government blog. Twining sees the war in Afghanistan as a means to keep Russia in a box, China down, India up, world trade humming, and the current international order, whatever that is, intact. I'm not going to bother to explain how all this works, but I picture the causal diagram as somewhat psychedelic. It's almost like a parody of Jack Synder' work on imperial myths, like he missed the part of the story where it says these aren't theories you copy but BS people use to sell wars.

For a relatively coherent version of the idea that we should fight in Afghanistan to help Pakistan, read Steve Biddle. Biddle has two arguments that he thinks are one -- that instability in Afghanistan will spread to Pakistan and that Taliban power in Afghanistan (not the same thing as instability) will provide extremists a base to attack Pakistan's government. One problem with the first claim is that historically Afghanistan's troubles have not destabilized Pakistan. The second argument struggles with the fact that we likely cause insurrection among Pashtun Pakistanis by warring with their Afghan cousins. And, as Justin Logan and Matt Yglesias note, the Taliban was in power for Afghanistan for years while Pakistan did OK, and many Pakistani elites want it back.