An Iraqi army soldier closes the door of a cell in Abu Ghraib prison near Baghdad in 2006. (Khalid Mohammed/AP)

During his campaign, President Trump repeatedly said that “torture works” — that it gets important information out of terrorists. He backed the use of “enhanced interrogation” techniques such as waterboarding.

But his view isn’t widely backed either by those with experience or by social science research. CIA Director Mike Pompeo has expressed conflicting views on enhanced interrogation. More skepticism about those techniques comes from seasoned national security professionals ranging from Defense Secretary Jim Mattis and 176 retired flag officers to experienced interrogators such as retired Colonel Steve Kleinman and former FBI Special Agent Ali Soufan.

Social scientists also doubt the effectiveness of these techniques. As we will explain below, research has found that torture and enhanced interrogations are likely to be less effective than noncoercive interrogation methods in persuading individuals to cooperate.

Here are five things you need to know about enhanced interrogation techniques.

1. What do we mean by “enhanced interrogation”?

“Enhanced interrogation” is a term coined by the George W. Bush administration to refer to interrogation that includes physically coercive interventions, such as waterboarding, sleep deprivation, facial slapping, forced standing, and so on. Proponents of enhanced interrogation have tried to distinguish these methods from torture, defined by international law as “cruel, inhuman or degrading” treatment.

Enhanced interrogation uses physical control to try to win over a subject’s compliance. By contrast, traditional interrogation methods try to get the source to cooperate by choosing to provide information.

2. How can we even study whether torture works?

Social scientists can’t directly study enhanced interrogation because subjecting people to these harsh methods would violate ethical principles of research. However, they can study interrogation methods that induce psychological pressure to cooperate.

Researchers have used two approaches to study interrogation: first, conducting experiments that simulate interrogations; and second, observing recordings of actual interrogations.

A body of scientific evidence has been gleaned from more than 100 studies of interrogation. This knowledge can help illuminate which is more effective: enhanced interrogation, or such noncoercive methods as building rapport, confrontation, elicitation, negotiation and offering incentives for cooperation.

President Trump recently provided an example of noncoercive method when he quoted Mattis as saying, “Give me a pack of cigarettes and a couple of beers and I do better with that than I do with torture.” This strategy would develop a relationship with a source, who might then be more willing to negotiate about favors in exchange for information.

3. What kinds of interrogation tactics get the best information?

Interrogations can be categorized as either “information gathering” or “accusatory.” Information-gathering interrogations involve rapport, direct and positive confrontation, and open-ended questioning. Interrogators initially listen respectfully while sources tell their story and then later confront sources about parts of the story that don’t make sense or seem suspicious.

That approach is quite different from accusatory interrogations, such as enhanced interrogations, in which interrogators try to control, manipulate and accuse the subject to win obedience. This would be familiar as, say, the fictional character Jack Bauer’s approach in the TV series “24.”

During information-gathering interrogations, sources talk more. That means interrogators walk away with more details that can be fact-checked against other evidence. Sources then become still more cooperative when confronted with inconsistencies between their statements and the evidence.

Further, studies find that when interviewers show acceptance, respect and empathy — referred to as “rapport-building” — sources use fewer counter-interrogation strategies, such as refusing to speak to an interrogator or changing the subject. What’s more, they provide more information.

When intelligence interrogators are surveyed, those who rely on rapport-building methods are more likely to say their interrogations were successful. Treating a source respectfully can go a long way in getting him or her to cooperate.

If information-gathering interrogations work better than accusatory interrogations, then we can reasonably conclude they might also work better than enhanced interrogation, which are by nature accusatory. Inflicting physical harm and control gets in the way of developing a productive, respectful relationship, which is the goal of rapport-building.

4. Which approaches produce more reliable intelligence?

Interrogators are looking for reliable information. To get there, they must ask their questions carefully, to help ensure the source’s recall remains accurate. They have to watch out for two pitfalls in particular: coercing sources so much that they give false information; and asking questions in a way that makes it harder for the source to remember information accurately.

Three types of enhanced interrogation may make sources more likely to give false information, according to research: sleep deprivation, social isolation and intense interrogator pressure. In enhanced interrogation, repeated threats of physical harm, such as waterboarding, impose more severe pressure on the source than most other interrogation methods.

Interrogators have only a limited ability to tell true statements from false ones, especially when they have no other evidence but the prisoner’s word. And even if analysts have other sources that they can check against, false evidence from a high-value source may lead them astray. 

But even if sources intend to cooperate, several components of enhanced interrogation may prevent accurate recall. Psychological and physical stress, sleep deprivation and repeatedly asking questions that rely on misleading information can lead to inaccurate or incomplete answers. Stress and sleep deprivation interfere with clear thinking. Mentioning misleading information, such as events that did not occur, may “fool” a source into believing that the information is true.

Scientists have evaluated several noncoercive methods to see which best helps subjects to remember things accurately. The Cognitive Interview has emerged as the most promising. In this type of interview, the interrogator assigns sources tasks that increase recollection of accurate details — such as having them remember specifics about the event’s context, like the weather, what they were thinking or feeling, or the layout of a room.

Evidence suggests that, even if enhanced interrogation did make sources willing to talk, it would get in the way of eliciting a full, accurate recount of details.

5. Which approaches make it easier to tell when a source is lying?

Research has long established that most people are generally poor judges of deception. Liars don’t seem to behave much differently from truth-tellers.

But some tactics can make it easier to spot a lie. For instance, when asked to tell a story in reverse order, liars hesitate and speak more slowly than truth-tellers. That works best when the source is ostensibly cooperating voluntarily.

Enhanced interrogation methods make tactics such as these less feasible, because liars and truth-tellers probably respond the same way to physical harm. That makes it harder to recognize who’s telling the truth.

Social science has a great deal of evidence suggesting that enhanced interrogation isn’t the best way to get reliable intelligence.

Misty Duke is a lecturer and legal psychologist at the University of Texas at El Paso whose research has been funded by the Federal Bureau of Investigation’s High-Value Detainee Interrogation Group (HIG). The opinions presented in this paper do not necessarily represent the views of the FBI or HIG.

Damien Van Puyvelde is assistant professor of security studies at the University of Texas at El Paso.