The folks at The American Spectator giddily report on the revelation that some of the key intelligence relied upon in foiling the recent London bombing plot was obtained by torture in Pakistan. Conclusive refutation, say they, of the argument that in addition to being inhuman, torture doesn’t work. But I think they seriously misapprehend that argument. Here’s the Specator:
The logic being that if you threaten to abuse a detainee until they talk, eventually they’ll start making up anything just to get you to stop. While it leaves the moral debate unchanged, the Guardian story, if accurate, pokes serious holes in the argument that torture doesn’t work. In this case, it may have very well saved thousands of innocent lives.
But the argument was not that genuine terrorists who have useful information will never disclose it under torture. The problem is that people who aren’t terrorists, or terrorists who don’t know much of importance, will offer up plenty of bogus stuff as well. Since the media don’t tend to run a whole lot of stories on every wild goose intelligence analysts chase and every dead end they run into, though, this cost tends to be invisible if you’re just following the papers. If torture were the only way to extract the necessary information, it might be worth it in strict cost-benefit terms despite producing a lot of hay for every needle it elicited. But I haven’t seen any evidence offered for that thesis, and it doesn’t seem to be the conclusion the old hands have come to.