More on Interrogation

A useful Senate hearing on interrogation. Also, I attempt to uncloud some issues Amy raised about my earlier post on learning to interrogate effectively.

Senators Sheldon Whitehouse and Lindsey Graham led an informative hearing on interrogation today. See witness statements here. See video.

Senator Whitehouse’s opening statement is very powerful. At the end, Senator Graham says he is afraid that if we start putting CIA interrogators on the stand and they tell us how much great information they got by torturing people, we as a country will learn the wrong lesson. That’s why we need a calm process to establish the facts of what happened before considering culpability and sanctions. The hearing itself is an indication that Congressional hearings can be helpful in getting at the truth, and that Graham’s concerns may be overblown.

On the whole, the main actors at the moment interfering with such a calm process are those on the right, including former Vice President Cheney who is loudly claiming that the Obama administration is making us less safe, those who are attacking Nancy Pelosi’s supposed complicity, and those such as Frank Gaffney who are claiming that President Obama is actively seeking to release more torture pictures on Memorial day (when both this and the release of the OLC torture memos were required by judicial processes, and just today the administration has announced that is moving to re-litigate the requirement that these photos be released).

Also, there is a real possibility that our covert relations with other intelligence services and governments would be disrupted by publicity of what was done via extraordinary rendition, and this should not be a trivial concern in deciding how to proceed.

Amy complains that I have clouded some issues in a horribly common way.

Taking her points in turn:

1) Amy titled her post that I reacted to “What if Torture Worked” and addressed another blogger’s discussion of the need for social science research to evaluate the effectiveness of torture. While she said “forget waterboarding for a minute,” she did not exclude other coercive techniques. I never thought she intended to torture undergraduates, but my inference was not a rush to a dark corner, it was an understandable perception of the plain meaning of a piece about torture and the social science study of its effectiveness.

The national debate is about the ethics and legality of coercive and dehumanizing techniques and about the effectiveness of such “enhanced” techniques compared to rapport-building techniques (not excluding deception). You can’t settle the relative effectiveness of these two broad types of interrogation by only studying non-coercive techniques. Amy has to choose between devising some sort of research design that would address coercive techniques, or agreeing that experimental social science will be silent on the subject that is denoted by the title of her post.

2) “Good research starts by assuming nothing.” Good luck with that. You can’t begin a research project without millions of assumptions, some explicit and some not. I didn’t know there were still any pure positivists out there. Amy’s philosophy of science would have you treat phrenology on the same plane as the experience of successful neurologists. I didn’t say you just accept the results of experience, but it’s a good place to start. Of course if the “best practice” is the application of leeches, then assuming that there is no best practice is fine. But going where something seems to be working and looking for variations in results, questioning why the technique works sometimes and not others, and looking for the underlying mechanisms and reasons for failure and success is usually a good way to proceed. Often the result is a completely different practice.

Amy has some common-sense understanding of who better drivers are, and knows that her grandmother is not among them. Presumably if you were trying to understand good driving you would start with experienced drivers with few accidents who were in their prime, and you would contrast them with a population of drivers with bad records but no particular defects such as blindness or poor reflexes.

There is now more recent military experience with interrogation than there was in 2002. I would not exclude the experience in Iraq as the source of expert inferences that can yield scientific hypotheses.

3) The difference between anecdote and data. I of course agree. That’s why I said

A serious attempt at finding the best methods usually starts from the best current practice and conducts empirical and theoretical work from that base.

I’m not sure how Amy read a post entitled “Learning to Interrogate Effectively” as being against effectiveness research.

In fact, I began my post with the intention of amplifying Amy’s call for more empirical research in connection with the military’s ongoing efforts to elicit information from those in its custody, whether this custody is transitory or more prolonged.