Friday, March 03, 2006


Telling whether people are being truthful or lying is important in life, generally, and in particular areas of life such as the justice system and the war on terrorism. According to a former judge, Bill Pincus, in the AFR today (here, subscription required), people are not at all talented in spotting lies.

He claims that research shows many people do not identify lies from truth more than 60% of the time. Moreover many people, including police officers, overestimate their own accuracy in picking a liar, leading to dangerous mistakes in such things as laying criminal charges.

Steve Van Aperen and Andy Shea have written a book on detecting lying. Lying can be detected by changes in body language, in expression, in hand-to-face gestures and in muculature changes. Women are better liars than men in the sense of better being able to conceal their lies. Also the socially-skilled generally make better liars. One sign of lying is pausing while you speak and male liars pause much more frequently. Its generally easier to lie in speech than in writing because verbal and body language cannot help conceal a written lie.

One approach to detecting lies is via interrogation. Here the most successful approach is to induce a confession rather than 'get in the face' of a subject. Another approach is use of the polygraph although there is no scientific evidence confirming the success of such devices.

Another possible approach to detecting lies could be using fMRI (functional magnetic resonance imaging scans). Basically the brain scan of a liar shows activity in the hippocampus region of the brain as illustrated in the picture above.

The US Army Office of Research have supported studies in Philadelphia to use such techniques to detect terrorists. Claimed accuracy rates of the approach from two studies vary between 78-99%. Two US companies are now proceeding top market fMRI tests for lie detection. The technology is expensive and so initially would be mainly directed towards detecting 'high-risk' behavior such as terrorism. Nothing seems to be done in Australia along these lines at present.

Of course even if this technology does develop efficiently there is no guarantee that it will be able to deal with a more prevalent form of misrepresentation in our society, namely bullshit. Princeton philosopher Harry Frankfurt's On Bullshit distinguishes between bullshit and lying. Bullshitters misrepresent themselves not by deliberately making false claims. Indeed, bullshit need not be untrue. Rather it is an attempt to convey a impression without being concerned about whether anything is true so there need be no specific brain impact. This can undermine a practitioner's capacity to tell the truth in a way that lying does not. Liars at least acknowledge that it matters what is true so bullshit is a greater enemy of the truth than lies.

It might also be difficult for any technology to determine when people are being evasive or dismissive, as skilled politicians are, when they seek to avoid difficult questions.

Incidentally, economists also use fMRI imaging. A basic survey of neuroeconomics, which looks at the role of the brain in evaluating economic decisions, risks and rewards, is here.

1 comment:

Bring Back EP at LP said...

He's lying