Serves Me Right for Talking About It

Cross-posted at Balkinization

In response to my post a few days ago lamenting the post-bin Laden urge to rehash debates about torture’s efficacy, Ben Wittes writes to disagree.

“Pearlstein is right, of course, that we will never know with any certainty whether any specific piece of information that the CIA program developed would have been developed had the program used no coercion–or had the program not existed at all and had the interrogators in question been military folks. One can never establish a but-for relationship between coercive practices and any valuable intelligence that we learned following its use. But I do not think the effort to understand the value added of the CIA program is useless. In fact, I think it’s critical. If one believes that the program contributed nothing–or little–of value that could not have been obtained in a manner that did not trouble our collective conscience, one will and should feel very differently about it than one will and should if one believes it provided critical intelligence that saved American lives and led to the capture and killing of key Al Qaeda figures. While it may be impossible to get certainty on the answers to these questions, it should not be impossible to get some more solid understanding of the matter than we currently have as a society. A lot of people are quick to opine on this matter in one direction or another, but it has received virtually no impartial study.”

Various things to say. First, I quite agree that impartial study of intelligence methods and their effects is of great value. That (among other reasons) is why I have long supported the establishment of an independent commission that would look back over intelligence practices of the past decade and evaluate not only questions of history, but also questions of strategic impact, small and large. It is also why I thought it worth looking closely at the findings those impartial studies that do exist – like the 2006 report on Educing Information by the Intelligence Science Board – which I’ve often cited for its conclusion, among others, that “knowledge of behavioral indicators that might assist in the detection of deception is very limited and provides little reliable information that could assist intelligence collectors” with current populations of interest. In other words, 4 years in to our dramatic program of coercive intelligence collection, no one knew in any real way how to secure the revelation of accurate information from an individual. My point in this week’s post was not that such investigations, when rigorously and ethically done, lack value. Just the opposite, it was to note that post hoc assertions based on incomplete information about a particular anecdotal case are no substitute for rational study. Put differently, for every “maybe some guy in Gitmo said something useful” story, there’s a “some guy in Gitmo said something false that lead us to war in Iraq” story. Dueling anecdotes do nothing to advance – and plenty to mislead – the inquiry about efficacy.

Second, and more broadly, let’s imagine for a minute that the closest to “truth” about efficacy we will ever be able to get is that sometimes people say useful things under torture and sometimes they don’t. In other words, let’s imagine that study will not be able to conclusively foreclose the possibility that on some occasion, someone will say something true if you torture them. (With more study, maybe one could discern what kinds of torture are more likely to produce true answers than false ones, even true answers not otherwise obtainable, but I’d be fascinated to hear the scenario for doing this in any ethical way.) The far harder question still remains: What policy conclusion follows from this result? How does one weigh this information in a cost-benefit analysis that includes tactical gain vs. strategic loss, effects on our own forces, variations among individual responses, designs of training and education, alternative options (including a finding that sometimes, someone will say something true if you give them a Big Mac), or all of the important arguments from philosophy, morality, law, medicine and religion that say efficacy is entirely beside the point, and we can decide as humans that there are some kinds of things we simply won’t do? Science can be enormously valuable. But I think it’s a mistake to imagine in thsi case that it will get us out of – or even much clarify – the otherwise far deeper question. In all events, for those who have not yet settled this matter in their own minds, this week strikes me as a particularly inopportune time to try.