Chris Matthews, who hosts the MSNBC news program Hardball, is convinced that torture works. In the last two days, in the wake of the Senate Intelligence Committee's damning report on CIA interrogation practices, he has allowed that so-called "enhanced interrogation" may be illegal and immoral, but he keeps reiterating versions of his belief that, for all that, it's effective in eliciting information from the enemy. That is, like it or not, it works.
Except that it doesn't. Matthews is wrong about this. It's perhaps unfair to single Matthews out, since he is not alone in this unshakable belief that if you hurt somebody enough, they will tell you the truth. Many Americans hold this same mistaken belief. But there is a wealth of solid science demonstrating that cruel forms of interrogation do not elicit good intelligence. In fact, this kind of interrogation often leads to resistance and falsehood.
But there are techniques that do work, quite well, and the evidence is not hard to run down. Much of it is summarized in a report from the High Value Detainee Interrogation Group, or HIG, which for the past four years has been working to create a science of intelligence gathering. Coordinated by psychological scientist Christian Meissner of the University of Texas at El Paso, the group has been conducting both field and laboratory studies with the goal of developing evidence-based interrogation techniques. Its substantial findings can be found here and here.
Here's some background on HIG. In 2006, the Intelligence Science Board investigated the existing science on interrogation, and basically found that there was none. That surprising and alarming conclusion led to recommendations that the government investigate effective, ethical interrogation techniques. And this led in turn to an executive order, issued by President Obama early in his first term, creating HIG, which comprises FBI, CIA and Pentagon entities. HIG launched a five-year, $12 million research program to support the mission, and Meissner (now at Iowa State) became the principal investigator.
HIG's scientific work is unclassified and peer-reviewed, and subject to all the usual ethical oversights. Its mission is broad and its findings range over many topics related to interrogation. But on the main question -- does it work? -- the results are unambiguous. The highlights are:
Rapport-building works. Laurence Alison of the University of Liverpool, UK, has studied a large sample of video-taped interrogations with alleged terrorists -- including some with al-Qaeda connections -- and he has found that the most effective strategies are demonstrating empathy and acceptance and allowing autonomy. When interrogators use these rapport-building strategies, detainees reduce their own counter-interrogation strategies -- silence, scripted responses -- and they provide more useful intelligence. Scientist Melissa Russano of Roger William University has interviewed highly experienced interrogators -- experienced in questioning "high-value" targets -- who confirm the importance of rapport.
Controlling strategies do not work. Meissner and his colleagues compared the effectiveness of rapport-building techniques and control-based techniques -- those used to instill fear and magnify the accusation and potential penalties. Controlling techniques -- taken from the Army's field manual -- increased anxiety, but supportive techniques elicited both positive reactions and higher quality information. Other studies have demonstrated convincingly that accusatorial methods -- those that increase anxiety and stress and use false evidence -- are more likely to produce false confessions. Other investigations have found that when interrogators are trying to confirm their presumption of guilt, they are more likely to use coercive methods that extract false statements from the innocent.
Strategies for detecting liars. Psychological scientist Aldert Vrij has been honing interrogations techniques for distinguishing truth-tellers from liars. For example, a consistent finding is that liars prep themselves for interviews by rehearsing possible answers to questions they expect to be asked. Interrogators can throw them off by asking unanticipated questions. Having to fabricate plausible answers on the spot boosts the liar's "cognitive load" which is apparent in the quality of the answer. Vrij and others have demonstrated the effectiveness of this and other cognitive strategies for exposing deception.
Tell me a story. Some of the new science verifies what experts is the real world have discovered through practice. For example, scientist Par Anders Granhag of the University of Gotherburg, Sweden, has been studying the techniques of one of the most renowned interrogators ever, World War II Luftwaffe officer Hans Scharff. Scharff was a master of the conversational interview. That is, instead of asking explicit questions, he told long, meandering and highly detailed stories. This story telling created the illusion that Scharff knew everything already, and it offered prisoners the opportunity to add missing details or correct errors. Scharff built a reciprocal relationship with the prisoners, who shared in the narration. These seemingly small corrections and confirmations were the intelligence that Scharff was seeking all along.
Follow Wray Herbert's reporting on psychological science in The Huffington Post and on Twitter at @wrayherbert.