The problem with looking for clues to deception is that you need to know what you’re looking for before you start. Should we count the number of smiles a person makes? How long they hold the smile for? Whether the smile is of a certain size? Whether it is faked? It requires some expectations, which we know aren’t always accurate.

By using an automated system, we were able to capture overall bodily motion (from the waist up) in a way that did not require us to select a particular behaviour to examine.

In one study, people came into the lab under the guise of a maths study. The study was rigged so that the second maths test was much harder than the first maths test. But we offered a cash prize for those who did better on the second test. Would people claim to have done better on the much harder test in order to get the cash prize? around 60% of people did, while 40% were honest about having done worse on the second test. We captured their body movements while lying and telling the truth.


There was a second opportunity to lie. Half-way through the study, a friendly research assistant accidentally knocked over the experimenter’s laptop while the experimenter was out of the room. At the end of the study, the experimenter checked his laptop and found it not working. Would people drop the friendly research assistant into trouble, or would they lie on the research assistant’s behalf? Again, about 60% of people lied, and about 40% told the truth.

Whether people were lying for a self monetary gain or for the benefit of another (not being reprimanded by the experimenter), we found that people moved less overall than the truth-tellers.

Making use of dynamical systems mathematical techniques (assessing motion recurrence and multiscaled entropy over time), we were able to show that it was not the case that liars simply clamped down their behaviours, as though ‘caught in headlights’. Quite the opposite: although liars moved less overall, there were subtle changes in how they behaved, showing greater flexibility in their patterns of behaviours than truth-tellers. This would seem to suggest that when we lie, we have flexible strategies that allow us to adapt our behaviour to the demands of lying. That is – people in general are skilled liars that have strategies preventing them from giving away their lies.

Selected References

Duran, N. D., Dale, R., Kello, C., Street, C. N. H., & Richardson, D.C. (2013). Exploring the movement dynamics of deception. Frontiers in Cognitive Science4. doi: 10.3389/fpsyg.2013.00140. [pdf]

Eapen, N. M., Baron, S., Street, C. N. H., & Richardson, D. C. (2010). The bodily movements of liars. Proceedings of the Thirty-Second Annual Conference of the Cognitive Science Society. Austin, TX: Cognitive Science Society, 2548-2554. [pdf]