Human obsession with lie detection is nothing new. One of the earlier forms of lie detection existed in India 2,000 years ago. A potential liar was told to place a grain of rice in his mouth, and chew. If he could spit out the rice, he was believed to be telling the truth. If he could not, it was understood that the fear of being caught had parched his throat, and his deceit was confirmed. Although lie detection technology has advanced monumentally since then, a perfect test still does not exist. The need to ‘crack the science’ behind modern lie detection is fairly understandable in the context of criminal investigations. Perhaps, however, the obsession is fuelled more by a refusal to accept human fallibility. This article, Part 3 in the Lie-spotting series, explores society’s obsession with the truth and our ultimate incapability to accurately recognise its opposite. This article first appeared on FirstRand Perspectives. – Nadya Swart
Lie-spotting series: Even more lies Part 3
By Alistair Duff
In the modern era, there has been the propensity to demonise lying and to judge those caught in the act of deceit. From business leaders like Madoff to truly global leaders like Clinton, their misdeeds have dominated news cycles, splashed across the print headlines and are the subject of days of televised debates. We have been forced to accept a post-truth world in which meaning and intent should be considered as more important than fact, as the current president of the United States seems to be addressing daily via Twitter.
It would seem that most people still think facts are important and that lies should be punished. There seems even more of an obsession with truth and dishonesty, empowered by an endless discovery of everyday occurrences throughout our world. Whereas we were previously shocked by a high-profile scandal, we have evolved to more of a lottery pick on who will be caught next.
Culturally, this obsession heightens public interest, which means any individual with some ideas around this topic will author a book, give a speech (or many speeches), stand in front of any camera available and, in this new digital world, the sudden appearance of a surprising number of apps.
Each app is masterfully represented by intriguing icons of fingerprints, heart monitors and piles of lie detector drawings. So mysterious and alluring, and all available for a mere $14.99; why choose only one or two? If each one offers nearly 100% accuracy, surely a combination can supply 300% or 400% accuracy and the potential to impress friends, embarrass others, and the possibility to author a book. Obviously, the success of this book will be severely hampered by the rank and file of the American intelligence community who have all penned a similar tome with the additional magic ingredient of ‘field experience’.
One of the key beneficiaries of this heightened awareness and interest in lie detection has been Pamela Meyers, the main subject of our previous article on fibs. Her books and lectures have proven to be bestsellers and have entrenched her reputation as the best human lie detector. People were amazed that through simply studying behaviour, they could increase their ability to assess truthful statements from 50% to 90%. To be fair, this sudden ability is not the subject of a money back guarantee in any of her work. The irony of the human lie detector lying about lies would prove way too meta (physical) for many. Subsequent research has garnered far less attention, but then again it has unilaterally failed to show similar remarkable results. The bucket of cold water for many aspirant Sherlock Holmeses is the number 54, or 54% to be precise. If the work conducted by Meyers suggests 90% accuracy, it suggests the participants were made up of around 90% functioning, highly powered telepaths (someone call the X-Men). All other research in this regard seems to settle comfortably around 54% or very close to this figure.
Fortunately for us, researchers do occasionally read previous studies. So, in a single project, 206 previous studies were analysed. This massive undertaking included 24,500 judges with over 6,600 messages collated and assessed with regard to progress in detecting lies. Don’t worry about a lack of field experience, as 3,000 recognised experts were also included. The average success rate among all 206 of these report rates was, you guessed it, 54%. An average of 54% accuracy can be easily contextualised as it is remarkably similar to the outcome of simply tossing a coin.
Upon deeper analysis of these results, certain facts became clear about the mystical art of lie detection. When simple fact checking is not possible, it seems we all but ignore the actual statement, relying more on behavioural cues that seem to suggest lies. These physical quirks have been generalised over time and are now branded ‘tells’. The popularity of this belief drives the public to learn and look for the most accepted tells and in general, the majority of talks and books focus on exactly that. This is not to completely discount them, as they are accepted norms for a reason and may at least suggest dishonesty. However, the great obstacle to a uniform approach to detection is the simple fact that the human race is made up of individuals both in terms of behaviour and detection. Attempts to decipher the honesty of an individual based on suggested norms fails to recognise key fundamentals that affect the person in question. Mood, motivation, habits and appearance have all been found to massively influence reactions to questioning.
It would also seem clear our interpretation may well be based more on our own views and a strong case of confirmation bias than the actual evidence. Confirmation bias is a human condition in which we actively seek support for our existing beliefs while ignoring contradictory evidence. The strongest example of the inaccuracy of lie detection simply made use of appearance. In a study that categorically disproves the ability to accurately and objectively detect dishonesty, the real truth becomes undeniable. In an ingeniously simple experiment, respondents were shown statements by people who naturally looked sincere vs those who may seem less sincere. Unsurprisingly, these judgements are made on pre-existing geographical or cultural biases. This approach was described as a demeanour-veracity match condition comprising shifty-looking liars and sincere-looking honest people.
The trick was also to include shady-looking purveyors of truth and those who seemed innocent but were in fact lying. The study included students from different cultures, well-seasoned university professors, and not to be left out, our trusty lie detection experts. The results are as follows:
The conclusion suggests that when apparently innocent people tell the truth or when shady-looking customers lie, we are pretty good at recognising it. However, if the appearance and the veracity of a statement doesn’t match, we all seem pretty awful regardless of culture, experience or most shamefully, claimed expertise. The experts show the greatest gap in ability to look past appearances. While you would naturally expect professionals to look past veneers during assessments, their score is embarrassingly low. It seems there is an even greater tendency to jump to conclusions without considering individuality and context. In the final analysis it would seem that finding a trustworthy lie detection expert may well prove remarkably difficult.