It turns out that when you torture someone until he tells you Iraq was training al-Qaeda operatives in chemical and biological warfare, he'll tell you want you want to hear even if it isn't true. Mark Kleiman and Kevin Drum both point to this sort of problem as the "pragmatic case against torture." It seems to me, however, that this is more like the pragmatic case for torture. The Bush administration, among many other flaws, has embraced confirmation bias with remarkable gusto. It seems to be their main epistemic method.
And that's precisely the sort of thing torture is really good for. If you already know what the truth is -- perhaps because it can be deduced from regime-type rather than boring intelligence gathering -- but just need some more evidence in order to convince others, then torture is a really, really, really good way of getting that kind of evidence. That's always been the main historical use of torture -- you have your prisoner, you want a confession, so you torture him until he confesses. It's not, after all, as if the administration was genuinely wondering about Iraq/al-Qaeda ties. They knew what they wanted to prove and they needed to make the case. Torture was an excellent way to get the job done.
To Save California, Read “Dune”
9 hours ago