CRF Blog

Nothing But the Truth

by Bill Hayes

In Nothing But the Truth, the Marshall Project reports on a “radical new interrogation technique.”

The trouble with modern interrogation technique … is that, despite its scientific pose, it has almost no science to back it up. Reid and Inbau claimed, for instance, that a well-trained investigator could catch suspects lying with 85 percent accuracy; their manual instructs detectives to conduct an initial, nonaccusatory “behavioral analysis interview,” in which they should look for physical tells like fidgeting and broken eye contact. But when German forensic psychologist Günter Köhnken actually studied the matter in 1987, he found that trained police officers were no better than the average person at detecting lies. Several subsequent studies have cast doubt on the notion that there are any clear-cut behavioral tells. (Truth tellers often fidget more than liars.) In fact, the more confident police officers are about their judgments, the more likely they are to be wrong.

But the scientific case against police interrogations really began to mount in the early 1990s, when the first DNA-based exonerations started rolling in. According to the Innocence Project, a group dedicated to freeing the wrongfully imprisoned, about a third of the 337 people who’ve had their convictions overturned by DNA evidence confessed or incriminated themselves falsely. These and other exonerations furnished scientists with dozens of known false-confession cases to study, giving rise to a veritable subfield of social psychology and the behavioral sciences. (At least one confession elicited by John Reid himself  —  in a 1955 murder case  —  turned out to be inaccurate; the real killer confessed 23 years later.)

Researchers have even broken down these false confession cases into categories. There are “voluntary” false confessions, like the many presumably unstable people who claimed credit for the Lindbergh baby kidnapping in order to get attention. Then there are “compliant,” or “coerced,” false confessions, in which people are so ground down by an intense interrogation that, out of desperation and naïveté, they think that confessing will be better for them in the long run. The third category, “persuaded,” or “internalized,” false confessions, may be the most poignant. Here, the interrogator’s Reid-style theming is so relentless, the deployment of lies so persuasive, that suspects  —  often young and impressionable or mentally impaired  —  end up believing they did it, however fleetingly.

And yet, even in the face of these documented cases, police and prosecutors have resisted admitting that false confessions are even possible. In court, they routinely move to reject expert testimony on the phenomenon by saying it goes against common sense that an innocent person would ever confess to a criminal act. But a wealth of research since the 1990s has shown that false memories are remarkably easy to implant. And in 2015, Julia Shaw, then a psychology PhD candidate in British Columbia, conducted a study that took direct aim at the idea that ordinary, innocent people would never confess to a crime they didn’t commit. In fact, she found that people can be made to do it quite reliably.

In just three one-hour sessions, Shaw was able to convince 21 of her 30 college-age subjects that they’d committed a crime when they were around 12 years old  —  assaulted another child with a weapon, for instance  —  and had a run-in with the police as a result. She supplied details that were recognizable to the subjects  —  the location where the assault supposedly happened, who the other child was  —  drawn from information their parents provided in a questionnaire. Shaw tells me she designed her study to mimic the techniques used in some false-confession cases. “I’m essentially marrying poor interrogation tactics with poor therapeutic tactics,” she says. The results were so strong, in fact, that she stopped administering the experiment before she had run through her full sample. [more]