I actually have spent much of the last 6 or 7 years studying torture and its utility for information gathering. Short answer: it works, but the effectiveness depends heavily on who’s asking the questions. An interrogator who is working from a seriously flawed point of view will only get information that supports that viewpoint. There are, on the other hand, examples of the use of torture in interrogation that did yield good evidence, but that was on account of the skill and openmindeness (I know — that seems like an odd word) of the interrogators.

Interestingly enough, there are studies that show that much of the sort of “interview” techniques used to question children who have allegedly been abused operate on a psychological level in much the same way as physical torture. And the results show that the more ideologically blinkered the “therapist” the more likely the results will be totally wrong.