Brain-scan lie detectors don’t work, study finds

Here’s an excellent story from Pacific Standard magazine:

June 10, 2013 • By Lauren Kirchner

It sounds just like something out of a sci-fi police procedural show—and not necessarily a good one.

In a darkened room, a scientist in a white lab coat attaches a web of suction cups, wires, and electrodes to a crime suspect’s head. The suspect doesn’t blink as he tells the detectives interrogating him, “I didn’t do it.”

The grizzled head detective bangs his fist on the table. “We know you did!” he yells.

The scientist checks his machine. “Either he’s telling the truth … or he’s actively suppressing his memories of the crime,” says the scientist.

“Dammit,” says the detective, shaking his head, “this one’s good.”

But it isn’t fiction. Some law enforcement agencies really are using brain-scan lie detectors, and it really is possible to beat them, new research shows.

The polygraph, the more familiar lie detection method, works by “simultaneously recording changes in several physiological variables such as blood pressure, pulse rate, respiration, electrodermal activity,” according to a very intriguing group called the International League of Polygraph Examiners. Despite what the League (and television) might have you believe, polygraph results are generally believed to be unreliable, and are only admitted as evidence in U.S. courts in very specific circumstances.

The brain-scan “guilt detection test” is a newer technology that supposedly measures electrical activity in the brain, which would be triggered by specific memories during an interrogation. “When presented with reminders of their crime, it was previously assumed that their brain would automatically and uncontrollably recognize these details,” explains a new study published last week by psychologists at the University of Cambridge. “Using scans of the brain’s electrical activity, this recognition would be observable, recording a ‘guilty’ response.”

Law enforcement agencies in Japan and India have started to use this tool to solve crimes, and even to try suspects in court. These types of tests have not caught on with law enforcement in the U.S., though they are commercially available here. That’s probably a good thing; the researchers of this study found that “some people can intentionally and voluntarily suppress unwanted memories.”

The experiment was pretty straightforward, and the participants were no criminal masterminds. Ordinary people were asked to stage mock crimes, and then were asked to “suppress” their “crime memories,” all while having their brains scanned for electric activity. Most people could do it, the researchers found: “a significant proportion of people managed to reduce their brain’s recognition response and appear innocent.”

Not everyone could, though. “Interestingly, not everyone was able to suppress their memories of the crime well enough to beat the system,” said Dr. Michael Anderson, of the Medical Research Council Cognition and Brain Sciences Unit in Cambridge. “Clearly, more research is needed to identify why some people were much more effective than others.”

Separate studies on guilt-detection scans, conducted by cognitive neuroscientists at Stanford University, had similar findings. Anthony Wagner at Stanford’s Memory Lab had study participants take thousands of digital photos of their daily activities for several weeks. Wagner and his colleagues then showed sequences of photos to the participants, and measured their brain activity while the participants saw both familiar and unfamiliar photos.

The researchers could identify which photos were familiar to the participants and which ones were not, with 91 percent accuracy, Wagner said. However, when the researchers told the participants to try to actively suppress their recognition of the photos that were theirs—to “try to beat the system”—the researchers had much less success.

Scientists still don’t know how this “suppression” actually works; like so many questions about the inner workings of the human brain, it remains a mystery. But the fact that so many test subjects could, somehow, do it on command, led the authors of both the Cambridge and Stanford studies to come to the same conclusions.

In short, brain-scan guilt-detection type tests are beatable, their results are unreliable, and they shouldn’t be used as evidence in court. Except on television.

2 responses to “Brain-scan lie detectors don’t work, study finds

  1. It’s been quite some time ago now BUT I once read an article that the current polygraph test could be controlled by the examinor of IF you passed or failed!! I guess IF the examinor can decide IF you pass or fail the polygraph test…. then how reliable is it as well!?!??? :-) :-) :-)

  2. Experiencing a recent court case where people were under duress during the hearing, observation of the supposed guilty party under duress is a good indication of guilt or innocence.
    Lack of eye contact, unusual nervousness and flushing are certain give away that something needs more investigation.
    I am no scientist, but it works for me.

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s