Benjamin Bestgen

Last month, I sketched out some definition problems we encounter when considering what a lie is. I also noted that humans are bad lie detectors. Research tells us that even supposed experts on lying – police, lawyers, psychologists, HR personnel – are not great at classifying truth and lies correctly, with an accuracy only slightly above 50 per cent.

Based on feedback received, lie detection seemed to spark some interest, so I dug deeper into the subject – naturally not with the aim of making anyone a better liar. Most of us need help with spotting lies better, not telling them. How can we do that?

The gold standard for lie detection is to tell accurately when a specific individual lies about or in response to a specific question or scenario. A mere guess – educated or not – that someone may be lying is not good enough.

Can technology help?

Humans have pondered lie detection techniques for centuries. Some believed liars would be unable to recite certain prayers or touch holy objects. Others considered that liars are anxious and cannot produce much spit, so a suspect was forced to stuff their mouth with dry rice and spit it out again (tricky even if your mouth isn’t dry, I’ve tried).

Nowadays, an entire industry wants to help spotting liars more reliably. In the US, polygraph (“lie-detector”) tests are conducted not only by law enforcement but can form part of job interviews for the police or paramedics, being an estimated US$2bn/pa business.

Polygraphs have been debunked repeatedly as completely ineffective, are hardly used outside the US and are generally inadmissible in court. But they exemplify the common belief that liars display physical signs (“tells”) that truth-tellers don’t and a polygraph helps detecting some of those. Blood pressure, perspiration, breathing, eye contact, pupil size, pulse, a dry mouth, various facial micro-expressions are all physical signs allegedly involved in lying. But they can also indicate stress, illness, medication, fatigue, excitement, concentration, surprise etc. A truthteller might be stressed trying to recall an event accurately or nervous because of an intimidating interrogation while a liar could be well at ease. Polygraph methodology and results are so ropey and open to interpretation as to be useless for telling truths from lies.

More recently, fMRI (functional magnetic resonance imaging) analysis explores if lies could be detected by analysing brain regions associated with the cognitive processes involved in lying. Problems of methodology aside, the brain areas most closely associated with lies and deception are also involved in planning, working memory, attention and inhibition control. fMRI also does not measure “brain activity” itself. It follows the flow of oxygenated blood: more oxygenated blood in a specific area indicates more neural activity. But blood flow and oxygenation levels are also influenced by age, medication, fitness or vascular capacity. The influence of anxiety, tiredness, stress or fear on fMRI results are also poorly understood to date.

In their current state, experts note that fMRI lie detection attempts are unsuitable for use: it is highly debatable how fMRI results should be interpreted, what neural activity really indicates, how reliable the tests are or whether lie detection tests done in laboratory conditions even work in real-world settings.

How about truth serums?

In law enforcement or intelligence services, learning the truth can be a matter of justice, security or life and death.

Intelligence agencies combine well-structured interview techniques with breaking down a person’s ability to concentrate or refuse cooperation. In narcoanalysis, suspects might be given sodium thiopental or amobarbital which induces disorientation, drowsiness and makes the subject talkative. 

Another technique involves depriving a suspect of food and sleep for several days, followed by a hearty meal and a large glass of wine. The exhausted body is overwhelmed by the food and alcohol, blood flow goes towards the stomach and a skilled interviewer may find it easier to elicit unguarded responses.

The ethics and legality of such techniques aside, they are not that reliable either. Narcoanalysis or alcohol tend to make subjects more talkative generally, eliciting both true and false information but do not reliably help to divide truth from falsehood. Similar points apply to all kinds of torture, which may well make a person talk but does not guarantee truthful statements.

But not all is lost – in part two of this article, I will present some psychological research that appears more promising and does not require technology, drugs or “softening up” a suspected liar through torture and inhumane treatment.

Benjamin Bestgen is a solicitor and notary public (qualified in Scotland). He also holds a Master of Arts degree in philosophy and tutored in practical philosophy and jurisprudence at the Goethe Universität Frankfurt am Main and the University of Edinburgh.