The TV show Lie To Me focuses on the exploits of an expert in lie-detection as he solves perplexing crimes in his high-tech Washington laboratory. It's actually fun to watch, especially since it appears to make some effort to get the science right (a real-life expert on lie-detection, Paul Ekman, serves as a science adviser on the show).
One of the show's premises is that only highly-trained experts (most importantly, its protagonist, Cal Lightman) are capable of sniffing out a well-schooled liar. This too is based in fact. Most of us are very bad at spotting liars, taking their seemingly earnest facial expressions as the real thing. Ekman's research, along with many others, has shown that it's possible to detect subtle differences between inauthentic emotional expressions and the real thing. Since telling a lie invokes its own distinctive emotions, it's possible to see remnants of these emotions by carefully watching a liar in the act of deceit, even when the liar masks his or her true feelings with a feigned emotion.
But what if there was a shortcut in sniffing out a lie, relying on our own instinctual behavior? Would it be possible to improve the lie-detecting abilities of ordinary people without all that training? A team led by Mariëlle Stel had a hunch that our tendency to mimic the physical and facial expressions of the people we are speaking to might help us to tell when they are lying.
They recruited 92 volunteers to participate in a very short conversation. The volunteers were paired up randomly, and one person from each pair was randomly assigned to be the truth-teller or liar. This person was asked before meeting the other participant if he or she would like to make a donation to Amnesty International, and then, randomly, told to either tell the truth or lie about it, with a one-euro reward if they could convince the partner they were telling the truth.
The partners were divided into three groups. The first group was told to mimic the liar/truth-teller's facial expressions and body movements. The second group was told NOT to mimic. And a final group was given no instructions about mimicry.
Finally, the partners met, sat down, and talked about why the truth-teller/liar chose to donate (or not donate, depending on what they actually did and the instructions about whether or not to lie). After the conversation, the partner became the lie-detector, and was asked to rate on a scale of 1 to 7 whether they believed the other participant was telling the truth about donating to charity. This brief conversation was recorded on video, and trained coders watched each conversant separately, coding their facial expressions and movements. These were then compared, and if the truth-teller/liar's movements or expressions were repeated by the lie detector within 10 seconds, it was counted as a match. Did the instructions make a difference? Here are the results:
As you might expect, those instructed to mimic did so significantly more than the other two groups. In fact, the amount of mimicking done by each group was significantly different, reflecting the fact that we do naturally mimic the movements of conversational partners.
So did the mimicking have an effect on ability to detect lies? Here are those results:
As you can see, both when the lie detectors were given instructions to mimic and when they were given no instructions, they rated lying as significantly more truthful than telling the truth (on a scale where 1 = "totally not truthful" and 7 = "totally truthful"). Only when lie detectors were explicitly told not to mimic did their truthfulness ratings actually correspond to the behavior of their partners: they rated lying as significantly less truthful than truth-telling.
So while the dozens of tricks employed in Lie To Me can help true experts detect lies, this simple study seems to show that simply telling interviewers not to mimic the behavior of the people they are talking to can make them much better at detecting lies. Unfortunately, that probably wouldn't make for as exciting a TV show!
Stel M, van Dijk E, & Olivier E (2009). You want to know the truth? Then don't mimic! Psychological science : a journal of the American Psychological Society / APS, 20 (6), 693-9 PMID: 19422628
- Log in to post comments
More like this
When dealing with liars, I use one of two approaches.
If I want them to think they're succeeding, I act like I'm not disagreeing with anything they say, but maybe not fully buying it all. This may bait them into doubling-down, so to speak, revealing a bigger plan.
When I want them on the spot, I stare at them with a dead expression, and never blink: this unnerves them.
(Yes, I can do that. I used to have a Doberman and he would initiate staring contests, which he always lost, either blinking or turning his eyes away.)
I expected the people who mimicked most to be the most accurate so this is a very interesting result.
Perhaps people were just taxed too highly, cognitively, with trying to mimic other people? I know I've tried mimicry and gave up because it was too many things to juggle in my awareness.
When we're not worried about it at all (the "no instructions" group), we are easily taken along by empathy or being interested, and might be thus more susceptible to falling for lies. We're kind of "conversationally entranced", sort of speak.
When we are given instructions NOT to do something, unless it's an overwhelming urge, we tend to manage it just fine. Sure, we are still constantly thinking "I can't do this, I can't do this", but it's just one thing. When actively trying TO mimic, we might be thinking that and also stuff like "What did he just do? Was that a full or a half-smile? Is my timing ok? He raised his arms â should I mimic that? Am I being obvious?" etc.
I think it would have been a more interesting study if a second control was told to perform calculations or some other kind of heavy taxing brain work instead of mimicry. Right now I really can't feel comfortable attributing the results to mimicry or no-mimicry alone.
Oh SWEET! Another way for half baked "experts" to decide that my autistic flat affect is evidence that I'm "lying". Doesn't help to have supporting documentation that proves that autistics lie significantly less often than the general population -- and in fact if you try to make me lie, my body is so offended that I vomit.
Cultural and neurological differences will always land us on the side of "liars", until more people are willing to accept there is more than one "look" of veracity.
Ashkelon: I'm not sure you're interpreting this right.
It seems that it is the person *with* the 'flat affect' that decides the other person is lying more often; I don't see anything about how the liar confronted with a flat-affect judges the other.
(Although this is a good followup question: if the liar consciously mimics, lies normally, or flat-affects, how does a normally responding person perceive? Offhand, I'd predict conscious mimicking would make the lies more effective, but not as much if the victim is mimicking, normal would be neutral, and flat affect would be neutral or harmful, as you say. But I don't think this study actually supports or opposes the third prediction.)
I think another element here is that, sometimes at least, in order to be lied to we need to lie to ourselves (not in the study cited but in life generally), we tend to feel in alignment when emotionally when we mirror other people's expressions, body movements and language.
So part of us may know we are being lied to instinctively, but barely registered consciously, even as we over ride this little unconscious nudging by enjoying the sense of rapport and shared perception that mirroring/mimicking can bring-the best liars tend to be very personable
Mark, I agree with you; I actually was thinking of that exact word, rapport. When we feel any type of bond with somebody we are more willing to trust them and are more likely to be duped.
Interesting study. I wonder if the non-mimic'ers are more skeptical because they are consciously breaking rapport with the liar, thus gaining more objectivity.
I'm getting really annoyied at all this lying lark. we all lie if you say you don't your lying. lying is an essential part of human life the past liars get the highest paid jobs an are normally better socially. lying is a skill that should be learn't well if one ever wants to progress in any society. we feel guilty about lying because most of us are brought by being told not to or in later life morals and religion. more work needs to be done to explore the dialectic between society and self.
This is probably related to attention. If you are trying to mimic someone else's behavior you don't have the cognitive resources left over to detect cues to lying. To be a meaningful contrast you would have had to ask those told to not mimic to count something or do something comparable to mimicking. The authors may wish to interpret this as mimicking leads to empathy which leads one to believe another's statements, but you'd have to rule out other interpretations to conclude that. And in fact, the result should be that those mimicking should be much higher rating the truth tellers but they are not, they are impaired. So it would make as much sense to conclude that if you mimic behavior then you distrust the authenticity of your own judgments that someone is telling the truth because you know you are engaging in effort to simulate their state.
Since those told not to mimic have nothing to lose and are not being judged on their success, they invest no effort in consciously avoiding mimickry, so they are attentional resources to devote to listening and watching the other person. This is a flawed study because you cannot conclude anything strongly about why this is happening because of its poor design, in my opinion.
You suggest that Eckman's research establishes this sort of lie detection is simply "fact." That is professionally irresponsible (assuming that you are really a scientist). In "fact," there is an enormous amount of scientific controversy about this, with many serious scholars (I believe the majority) holding that Eckman's research findings are highly dubious.
The comment was made, "...simply telling interviewers not to mimic the behavior of the people they are talking to can make them much better at detecting lies."
This is a rather narrow statement... This assumes the deceiver is going to be completely passive, stagnant, and unaware of the situation while occurring. Realistically, more likely, and not in an academic setting with 9X random people chosen that are all more than likely more honest than not, your average strategic deceiver is going to be significantly more skilled than your average detector at pulling a fast one by virtue of practice alone. The study doesn't appear to take into consideration whether the participant's skills or inclination was toward deception or honesty - and strategic deceivers aren't uncommon, but they're not typically the type to volunteer for scientific experiments. Anyone with half a brain that's read for more than 15 minutes about deception would have a clue to follow suit when someone breaks rapport - most deceivers can figure that out even without the 15 minutes of reading just through trial and error from childhood on up.
Purposefully breaking rapport is just going cause the strategic deceiver to follow suit, until they've spotted the pattern and rhythm - then when you next break off you'll just find the deceiver has preempted you and the rapport remains. That is, if you notice you're under their spell at all before it's over and done with...
Making the statement that interviewers should "simply not mimic" to aid in lie detection "simply" puts those people at just as much risk to being deceived by even common strategic deceivers as before.
Then again, the world's misinformation is to the benefit of all strategic deceivers... In that case, all y'all, think whatever you like. :)