The science of torture
The Senate Intelligence Committee report on its years-long investigation of the Central Intelligence Agency’s immoral torture-based interrogation methods says the CIA got no information that stopped terror attacks. Which is not surprising. Scientists have been telling us for a long time that torture is a lousy way to get people to tell you the things you want to know.
“The scientific community has never established that coercive interrogation methods are an effective means of obtaining reliable intelligence information.” Martin Robbins leads off his indignant post at The Lay Scientist with this quote from a 2006 report of the Intelligence Science Board, formed to give scientific advice to US intelligence services. The Board was abolished in 2010, ostensibly for the sake of efficiency and the budget. I can’t help wondering if it was dumped because it told intelligence agencies–a misnomer if ever there was one–things they didn’t want to hear. Such as: torture doesn’t work.
The idea that torture is effective is deeply ingrained, surely at least in part because pop culture fictions tell us so. Robbins points out that there’s even a TV trope called Torture Always Works, summarized as: torture is an instant source of 100% reliable information. I never watched 24, but I gather it was built around that trope. I am watching Homeland, which takes the moral issues with torture more seriously. Quinn really hates torturing people, just hates it. He suffers almost as much as the torturee. Still, he does it, because it’s the only way to, you know, Prevent Explosion of the Ticking Time Bomb.
Robbins also notes that the techniques torturers employ, like pain and sleep deprivation, are if anything counterproductive because they are almost guaranteed to interfere with memory. At SciAm’s Observations, Joshua Krisch expands on that point, arguing that, when torture subjects lie, the lie may not even be intentional. It’s a product of torture’s effects on the brain and memory.
Krisch links to links to past SciAm articles on torture. He also links to documents claiming that CIA interrogation techniques were not really torture and that the information gleaned thereby saved countless lives. (Note that this vindication site has a .com domain name, not .gov.)
The Senate report took many years to put together and is said to be 6000 pages long. We’ll probably never know because we’ll probably never see it. What the Intelligence Committee finally released on Tuesday may be the longest Executive Summary on record, 525 pages. A PDF can be had a number of places, for example CNN.
The CIA itself knows torture does not work–or at least it knew that, and said so, before 9/11, according to Robbins. So how and why did this horrid stuff happen?
At Mind Hacks, Vaughan Bell explains, under a hed that says it all in brief: Snake Oil salesmen selling torture. The salesmen were two former Air Force psychologists, James Mitchell and Bruce Jessen, whose previous experience with interrogation techniques consisted of trying to train soldiers to resist them. They secured a generous long-term contract to run the CIA’s interrogations program and even collected $81 million before it was cancelled in 2009.
The Mitchell-Jessen program was based on a misunderstanding (willful?) of the concept of “learned helplessness.” The concept was developed by psychologist Martin Seligman, who found that subjecting dogs to repeated electrical shocks that they can’t escape eventually makes them passive; they hunker down and just take it. (Seligman is distraught over this distortion of his work and has said so emphatically, according to Jesse Singal at Science of Us.)
Mitchell and Jessen theorized that prisoners forced into this passive state would be more cooperative. “This, to be frank, is just bizarre. The theory predicts the opposite would happen and this is, rather grimly, exactly what occurred,” Bell says. He also notes that the report says CIA staff, including staff psychologists, argued repeatedly against the Mitchell-Jessen approach, but were overruled.
So the interesting question is why these two bozos were put in charge and backed up so strongly even in the face of reasonable criticism. The new report does not appear to have answered that question. And neither, really, does the 2009 NY Times piece that chronicled the rise and fall of Mitchell and Jessen. (The fall came via the newly elected President Obama, who shut the CIA interrogation program down.) The Times piece says the two were persuasive, especially Mitchell. Is that really all it was? That the CIA needed an interrogation program and these two turned up and said, “Yeah, we’ll do it. Scientifically.”
What does work?
Psychologist Wray Herbert, at the Huffington Post, describes the work of the High Value Detainee Interrogation Group, which has been conducting field and lab studies on effective interrogation techniques.
It has supported psychologist Laurence Alison of the University of Liverpool, who described what he does at The Conversation. “In our own work, based on hundreds of hours of observation of field interviews, we found that interrogators that used approaches more akin to methods used in therapy were more effective at both decreasing detainee disengagement (including “no comment” interviews) and eliciting useful information and evidence,” he says. Alison notes that these methods are nothing new, are widely used by law enforcement too, and are backed by studies showing that rapport-based methods also work in clinical settings, such as addiction treatment.
Posts at the British Psychological Society’s site describe individual research projects that back up this point. One retrospective study of techniques employed by interrogators from several countries with suspected terrorists found that rapport-based techniques generated far more information than coercive ones.
Even some TV writers have gotten that message. You can now occasionally see a cop show where the interrogator–often a woman, fancy that–brings the prisoner a soda, sympathizes with his problems, lays off the accusatory mode, and gradually draws out useful information. It’s lots less dramatic than torture and threats and doesn’t feed our involuntary desire to see the bad guys tormented for their badness. But maybe it’s the start of a pop culture lesson on how to get people to tell you what you want to know.
Having excoriated psychologists for participating in this evil and pointless practice, let us turn briefly to the medical profession.The report is searchable, and so bioethicist Craig Klugman searched it for terms like doctor, physician, medical etc. He tells us at the Bioethics.net Blog that physicians agreed to be involved with torture–administering torture, judging whether “detainees” were healthy enough for torture, designing interrogation, and failure to report it–despite the fact that these things are forbidden by medical ethics and explicitly by medical professional associations.
Klugman is dismayed that his fellow bioethicists have not spoken out strongly against torture, which he argues is wrong by the standards of nearly all schools of ethics. Because torture does not yield reliable information, he says, even utilitarians should be outraged.
Finally, I urge you to read Michael Hare’s post “On doing bad things, being a bad person, making a living, and having a voice” at The Reality-Based Community. It is only tangentially about the CIA report and involves science hardly at all except for mention of William Shockley’s racial opinions. But it is exceptionally sane and enjoyable withal, and I got the idea of using the Boteros as illustrations here from him.