A fair point from Michael Cohen in partial response to something I wrote about considering civilian casualties to have strategic implications for the Afghanistan war:

If we’re not willing to accept the fact that civilians are going to be killed in war – and that lives will be upturned by the determination that our perceived interests in southern and eastern Afghanistan trump those of local civilians – then we have no business fighting this war in the first place. This isn’t intended as a dig on Spencer (who I’m sure understands this) but there is something frustrating about the way we seem to talk about civilian casualties in Afghanistan as if our very presence and our very decision to go to war is not a fundamental part of civilian suffering in Afghanistan.

I’ll meet halfway, because I suspect Michael and I don’t actually disagree here. Yes, the U.S. deciding to continue to prosecute the war will mean civilians will unfortunately die. Anyone who considers the war to be in the national interest, like myself (with caveats), must carry that recognition as fundamental in the name of basic intellectual honesty. No aspect of that recognition mitigates the U.S.’s obligation to minimize civilian casualties with all due effort. Indeed, to take rigorously the idea that civilian casualties will a major factor influencing the Afghan people’s “decisive” perspective on which coalition — Karzai’s or the Taliban’s — better secures their interests, then minimizing casualties becomes a strategic objective in the war.

Here’s where those who base their opposition to the war its promotion of human suffering have to meet halfway as well. If the U.S. stops prosecuting its end of the war, civilian casualties will not end. What will end is the civilian casualties we directly cause. The Taliban-led coalition will continue its insurgency until victory or negotiation, with all the acceleration of civilian casualties that will entail. (I would think it’s likely that the Taliban would greet an abrupt U.S. withdrawal, in the absence of a capable Afghan security apparatus, as a disincentive to negotiate, since its coalition will perceive itself to be winning. Negotiations would become a venue for the Karzai government to capitulate.) It’s additionally possible that Afghans will consider a U.S. withdrawal to constitute abandonment, and then hold the U.S. responsible for the casualties that the Taliban-led coalition subsequently induce.

Now, you can argue that such a circumstance ultimately benefits the U.S. national interest better than an indefinite, bloody and expensive war. Or you can argue that the counterinsurgents are wrong, and while civilian casualties are to be avoided in general, they don’t have strategic implications. But you can’t simply argue that a U.S. withdrawal comes with a pony for every Afghan citizen, since that overlooks the United Nations’ documented increase in the proportion of civilian casualties for which the Taliban are responsible. (This isn’t intended as a dig on Michael, who I’m sure understands this.) These are hard choices, without easy answers. A recognition of that doesn’t diminish the obligation to choose, but we should reject simplicity from whomever offers it and for whatever purpose.