Conventional wisdom has it that when people talk, the direction of their eye movements reveals whether or not they're lying. A glance up and to the left supposedly means a person is telling the truth, whereas a glance to the upper left signals deceit. However, new research thoroughly debunks these notions. As it turns out, you can't smell a liar by where he looks.
Researchers in the United Kingdom investigated the alleged correlation between eye direction and lying after realizing it was being taught in behavioral training courses, seminars and on the Web without the support of a shred of scientific evidence. The idea has its roots in a largely discredited 1970s theory called Neuro-Linguistic Programming (NLP), a set of techniques intended to help people master social interactions.
In one experiment, the scientists monitored the leftward and rightward glances of 32 study participants as they told a mix of truths and falsehoods. The participants — all of whom were right-handed, in case eye movements are reversed in lefties — were equally likely to glance upward and to the right and upward and to the left, regardless of whether they were lying or telling the truth.
In a second experiment, the researchers found that a group of 25 participants who were informed of the alleged relationship between eye movements and lying were no better at detecting lies than a second group of 25 participants who were not taught the rule of thumb. [How to Pass a Lie Detector Test]
Finally, in a third experiment, the researchers investigated a video archive of 52 individuals making public pleas on behalf of missing relatives, half of whom are known to have been lying (based on the outcomes of the cases) and the other half, found to have been telling the truth. When the researchers tallied the number of upper-left and upper-right glances among the speakers, there was, again, no difference between the directions of the liars' and truth-tellers' eye movements.
"Our research provides no support for the idea [that certain eye movements are a sign of lying] and so suggests that it is time to abandon this approach to detecting deceit," said study co-author Caroline Watt of the University of Edinburgh in a press release.
The researchers have detailed their findings in a new paper published today (July 11) in the journal PLoS ONE.