Does It Matter Whether Torture Worked?

The Washington Post's intelligence and national security reporting team make the case that, in the case of 9/11 planner KSM, enhanced interrogation techniques -- EITs --or torture -- facilitated his cooperation with the CIA.  The story has produce a violent reaction among supporters and opponents of using the practice, with supporters crowing and opponents accusing the Post of letting Dick Cheney man their editing desk.

Well, we know that whether torture worked should not effect the moral case for or against it, but in the mind of the public, which seems to look at the practice through the "24" ticking-time-bomb lens, its effectiveness does seem to be related to its appropriateness in extreme situations where lives are at stake. Given this distortion -- most every instance of torture did not take place with a threshold-level "24" scenario in the offing -- it is quite comprehensible why proponents and opponents of torture are so invested in proving that it never works, or that it almost always works.

There is a difference, though.

It's become an article of faith for opponents of torture that it doesn't work because most of the evidence available to us -- the testimony of many military interrogators, psychologists and FBI agents, the incompleteness of the interrogation record, the mental projection that a tortured person would say anything to avoid being tortured, and the historical association of torture with bad people -- suggests that its benefits are dubious. But most every engaged opponent of torture doesn't care whether it worked because the moral wrongness of torture exists in nature prior to an evaluation of its effectiveness. The heuristic is: torture never works, and even if it did, it's still never right.

Proponents of torture techniques (who may object to my use of the term "torture") are more likely to base their judgment on evidence (or hints) that the tradeoff between using the techniques and the value that is derived from them is worth making. The heuristic is: torture is usually wrong, except in circumstances where we can use it to prevent deaths; don't take this tool out of the toolkit.

The standard for proving whether torture worked ought to be the same standard for proving that it didn't work. And proving this is kind of impossible: it is vulnerable to the post hoc ergo propter hoc error. It involves human behavior and its interpretation, intelligence (the value of which is unknowable in isolation) and the fallacy of single causes. There's no double blind study available.  

So opponents and proponents look at the evidence and make reasonable inferences. If KSM didn't cooperate until after the EITs began, and if it doesn't seem as if anything else happened at the same time, then, for whatever reason, his torture by waterboarding did work. Or, rather, it didn't NOT work.  It didn't work 100%, as he still apparently gave false intelligence. Even among those who received the intel that KSM provided there is a debate about whether he really provided much value; there is no debate that, in addition to whatever value he provided, he also provided a lot of useless information.

 Using the same standard, Abu Zubaydah began to give up actionable information well before he was tortured, and he apparently gave up some information after he was tortured. In his case, it's reasonable to say that torture did not facilitate his cooperation. 

Was torture the only way to facilitate KSM's cooperation? Unknown, but it seems like other methods were tried and failed. Did his cooperation, once established, provide actionable intelligence that saved lives? Not quite clear, but it seems like most of the people who were directly involved in his interrogations believe that it did. Assuming that the answer to the above questions are yes, then torture opponents would seem to have a problem. As much as torture is abhorrent to them -- and I'll reveal my bias here -- abhorrent to me -- the general public does not separate the right and the good, and they make gradations based on the intent of the interrogator/torturer and the effects of the practice. The heuristic: we are no better than the Nazis if we torture works on me as a moral argument, but it does not work for many politicians, and it does not work on most Americans. An American CIA interrogator whose techniques yield valuable information is much less reprehensible than a Gestapo torturer whose techniques resulted in the death of Jews or gypsies. Doesn't mean the CIA guy was right, but it's still hard to disagree with that sentence. 

If opponents of torture (like me) want to change the minds of Americans, we have to be ready to accept that our preconceptions may need some changing, too. The burden of argument rests on us -- we need to persuade people that the act of torture in a democratic society is always wrong, that the ticking time bomb scenario is rarely -- if ever -- the situation interrogators face, and that even if torture works in a few cases, it is not worth the moral (and tangible) costs to our country. If there's evidence that torture worked in certain cases -- or if it wasn't counter-productive -- even if the evidence isn't complete or is fragmentary, it's not going to advance the anti-torture cause to ignore it.

To effectively change the minds of policy makers and the American people, torture oppoents should be prepared to accept -- subject to rigorous standards of doubt, of course -- that torture might not always be counter-productive.  It's usually counter-productive -- and might have been almost always counter-productive in the setting of Bagram, the secret prisons and GTMO -- but even as I criticize their moral judgment, I can't dismiss out of hand the conviction of those who participated in the interrogations that the EITs were necessary evils, and that other means were tried and found wanting.  No doubt, though, that some of the later claims of the effectiveness of the EITs stem from an attempt to justify what in retrospect are morally questionable practices.