The lying game

But gathering reliable research data is a tricky proposition. To begin with, a set of lies and truths need to be collected. Ideally, participants should not be aware that they are taking part in experiments that are dealing with the subject of truth and lies. So Dr. Street and his colleague devised an ingenious and well-intentioned deception of their own that involved hiring a film studio in London and persuading passers-by to be interviewed for a "documentary" on tourism.

They were told by research assistants placed outside the studio that the film makers were running out of time and asked if, in addition to describing genuine travel experiences, they would talk about places they had not actually visited. Inside the studio, the speakers were then interviewed by a director who -- they supposed -- was unaware that they had agreed to lie on film.

"The idea was that they were lying to someone that they could potentially deceive. They were lying on behalf of another person, but the lie was spontaneous and told with an intention to mislead," said Dr. Street. The sequence of filmed interviews that resulted from the experiment constitutes a valuable body of material that is being made available to other researchers in what is still the relatively new field of human lie detection.

For more than 30 years, the standard approach to tapping the unconscious has been to use the "indirect lie detection" method.

"People are asked to rate some behaviour that is indirectly related to deception," explained Dr. Street. "For example, does the speaker appear to be thinking hard or not? The researcher then converts all thinking-hard judgments into lie judgments and all not-thinking-hard judgments into truth judgments."

The fact that these indirect judgments give better accuracy than asking people to directly and explicitly rate statements as truth or lies has been taken as evidence that people have innate, unconscious knowledge about human deception. Dr Street and his co-researcher and author Dr Daniel Richardson, of University College London, have developed a different explanation, which they explore in their new article in the Journal of Experimental Psychology: Applied.

"Indirect lie detection does not access implicit knowledge, but simply focuses the perceiver on more useful cues," write the authors. It is an argument that could have real-world significance, in the training of interrogators, for example.

"There has been a push in the literature suggesting that indirect lie detection works and the reason is that it is unconscious -- so people should not be making reasoned judgments but relying on their gut feeling," said Dr. Street. "But if our account is correct, that is a very bad way to go."

He readily concedes that human lie detection -- while a fascinating subject -- requires a great deal more research and is a long way from infallibility.

"Typical accuracy rates are around 54%, reaching up to around 60% with training. So there is unlikely to be a one-size-fits-all strategy that gives us accuracy rates anything like what we would want in a legal setting. The field needs to start considering how to improve clues to deception, how to prevent raters from using less reliable clues, and to better understand how information about the current context plays into that judgment.

"We often think of nonverbal behaviour when we think of deception," continued Dr Street. "But it would be better to focus on the content of the tale people are selling us, and asking if it is consistent with other facts we know. But even then there is a large amount of room for error."

If human lie detection has a long way to go and there is probably a cap on the accuracy that can be achieved, could the polygraph machine fill the gap? No, asserts Dr Street, adding that the British Psychological Society is one body that has dismissed the polygraph as a tool that will never be useful.

It purports to work by detecting anxiety. "But are liars more anxious than truth tellers?" said Dr Street. "The reality is no, because often the reason we lie is that to tell the truth would be very difficult and more anxiety-provoking than a lie."