Torture Does Not Work: The 2014 Senate Report

The Senate Select Committee on Intelligence spent five years and reviewed more than six million documents. The executive summary of its study of the CIA's Detention and Interrogation Program was released on 9 December 2014, redacted, at just over five hundred pages. The full report runs to about six thousand and remains classified.

Its first conclusion, of twenty, is the one this chapter is about: the committee found that the use of the CIA's enhanced interrogation techniques was not an effective means of acquiring intelligence or gaining cooperation from detainees. The committee examined the twenty most frequently cited examples of intelligence supposedly obtained through the program and concluded that in each case the information was either already available from other sources, was obtained before the coercive techniques were applied, or was not in fact accurate.

The CIA disputed the report's methodology and conclusions. A minority of the committee dissented. The factual record of what was done — waterboarding, sleep deprivation for up to a week, stress positions, confinement boxes, rectal rehydration, threats against detainees' families — is not disputed.

What this guide can add is the mechanism, which is not a matter of politics and was known in advance.

Extreme stress does specific, measurable things to the memory system. Cortisol and catecholamines rise; sustained elevation impairs hippocampal-dependent retrieval and degrades prefrontal executive function. Charles Morgan and colleagues demonstrated the magnitude of this in studies of US military personnel going through SERE survival school, where trainees undergo a realistic stressful interrogation. Twenty-four hours later, a large proportion could not correctly identify their own interrogator from a lineup — misidentification rates in some conditions exceeded those in the eyewitness literature that the field considers alarming. Morgan also found substantial reductions in recall accuracy for information presented during high-stress interrogation.

These were healthy, trained, motivated soldiers who knew the exercise would end.

So coercion damages the instrument. The interrogator's objective is to read a memory, and the method degrades the memory system and the executive function needed to report it, while raising suggestibility.

And it produces speech. That is not in doubt, and it is why the technique persists. Pain reliably makes people talk. It does not make them accurate, because the speech is aimed at ending the pain and any statement that achieves that is equally successful from the speaker's point of view. Ibn al-Shaykh al-Libi, under torture in Egyptian custody, produced the claim about Iraqi training of Al-Qaeda in chemical weapons that was cited in the case for the 2003 invasion. He later recanted, saying he had said what he thought his interrogators wanted. The claim was false.

Which returns to the question left open in chapter 102. Moran had this right in 1943. The evidence has only accumulated since. The persistence of coercive interrogation is therefore not explained by a belief that it produces better intelligence, and the report offers the alternative explanation implicitly: it was adopted quickly, by people under enormous pressure to act, on the advice of contractors with no interrogation experience, and then defended because it had been adopted.