FEATURE1 October 2010

About face

Dr Paul Ekman specialises in analysing facial expressions to read emotions, and is the real-life inspiration for the investigator played by Tim Roth in the TV drama Lie To Me. Robert Bain plucked up the courage to meet him face to face.


Dr Paul Ekman doesn’t seem like a particularly intimidating figure. At least not at first. A stocky American in his mid-seventies, Ekman is calm and slow in his words and movements. The reason he is intimidating is that he has dedicated his life’s work to studying people’s facial expressions in order to read emotions – and detect lies. Fortunately in this instance he is the one being interviewed.

People are often nervous when they meet him, Ekman says, and he usually does his best to put them at ease. Not always, though. “When I interview people for a job, if they don’t already know it, I tell them that I am expert in not only reading emotion but spotting lies,” he explains, “and I hope that they’re not in any way tempted to either test me or try to mislead me.”

Recently Ekman has become better known outside the scientific field by providing the basis (“with a lot of exaggeration and imagination thrown in”) for the character played by Tim Roth in the TV series Lie To Me – the “human lie detector” who solves crimes by reading people’s expressions.

“I can’t tell you how to get to London Zoo with facial expressions, but unless I’m a poet I can’t tell you how I feel at this moment as well as my face will display”

Ekman first became interested in what the face can reveal from observing group therapy sessions and “seeing that so much of what was occurring was not in words, but in gesture and expression”. After reading Charles Darwin’s book The Expression of Emotions in Man and Animals, Ekman wanted to test Darwin’s belief (largely disregarded at the time) that the way the face portrays emotion is universal across cultures. To do this, he travelled in 1967 to the highlands of Papua New Guinea to find communities who were isolated from the outside world. “I was the first outsider they had ever seen,” said Ekman, “and their expressions were the same as anyone else”.

Individuals vary in the strength and speed of emotions and their tendency to suppress them, but the expressions themselves and how they relate to emotion are involuntary and universal, making them “the primary source for knowing how people feel, moment to moment”. “I can’t tell you how to get to London Zoo with facial expression,” Ekman says, “but unless I’m a poet, I can’t tell you how I feel at this moment as well as my face will display. We’re all capable of doing a rather bad job, unless you’re a good actor, of falsifying emotion, but there are changes in expression that occur without choice.”

Ekman also found that tiny, fleeting expressions, lasting sometimes only a twenty-fifth of a second, can reveal concealed emotions and that, with practice, anyone can learn how to recognise them (it makes watching politicians speak “extremely entertaining”, he says). His system for doing so is now used by everybody “from the animation studios to advertisers” and Ekman has found himself working with security services, law enforcement bodies and health professionals all over the world, especially in the wake of the success of Lie To Me, which has seen Ekman receive requests for work “fairly often” from as far afield as Russia, Iran, Syria and China.

Compared to technologies like brain-scanning, Ekman says, his technique (which can be learned at face.paulekman.com) is straightforward and unobtrusive and gives more precise moment to moment detail.

What Ekman can’t do, of course, is read your thoughts, and he insists that “if anyone ever developed a machine that would allow us to read people’s thoughts as they have them, I would be among the first to want to destroy it.”

When asked what difference it would make to the world if we could all read each other’s emotions – even the ones we try to conceal – Ekman plays it down. “It would certainly mean we’d better understand each other than we do now,” he says. “I like to believe that it would be of more benefit than harm.”

“There are some companies that are turning to this as a source, rather than focus groups which have been used in the past to try to get at what people feel about particular products”

How can facial expression analysis be used in market research? So far most of Ekman’s work has been in law enforcement, anti-terrorism and health. “That’s just because those are the people who’ve knocked on my door, and the problems are fairly urgent,” he explains. “Whether you buy Colgate or Crest matters an awful lot to the companies that make it, but doesn’t matter much to me. Whether a terrorist gets on the plane I’m going to take tomorrow matters a lot to me, and to everyone else.”

He is, however, looking at opportunities in the commercial sphere, although he refuses to work for firms involved with alcohol, tobacco or gambling – an obligation he feels partly as a result of having spent much of his career supported by the National Institute of Health.

He’s done some work with Procter & Gamble on new product development and how facial expression analysis could be used “to evaluate consumers’ reactions to existing products and advertisements”.

Ekman says: “There are some companies that are turning to this as a source, rather than focus groups which have been used in the past to try to get at what people feel about particular products or product advertisements. This is another tool. Will it give you added value or give you things you don’t get from a focus group? That’s a reasearchable question, too.”

One of the people putting Ekman’s techniques into practice in the market research world is John Habershon of Momentum, who analyses video recordings of discussions with consumers and their reactions to products and ads. He believes that facial expressions help answer one of the big challenges facing research. “People say to themselves in market research, we can’t just follow what people say because that’s just their rational brain, but the means to get beyond that have not been available. It’s like, we know this is the wrong way at the moment, but what do we do now?”

Although the interpretation of expressions comes down to a subjective judgement, Habershon says there is usually a high degree of agreement when different people look at the same recording. “I did a workshop last year where I tested it on the audience. I have to admit I was a little bit nervous because I hadn’t tried it in public before – and these were all qualitative researchers. I asked them to look at the video, discuss it in groups and say in which of the instances they saw engagement. And they all got it right. So I think the main lesson for me is that it’s not something there’s going to be a lot of disagreement about. You get the same thing with clients – you look at it, the client looks at it, and 99% of the time there is agreement. It’s not a case of saying, ‘trust me’, it’s pretty universal.”

Having managed to get such rich findings from facial expressions, Haberson believes that those looking at high-tech neuroscience techniques to understand the subconscious are barking up the wrong tree. “Apart from having to put someone into an FMRI scanner, with all the noise and so on, the brain is actually much more complicated than we think, and this idea that certain parts of the brain ‘light up’ is a gross oversimplification.”

“In a typical project the results from the non-verbal analysis are much more decisive than the ones from the verbal. You’ll get just one or two things standing out as triggering an unconscious reaction, whereas once you get into people’s rational brains they can spread their good feelings over a whole range of things.”

“For clients it offers a more decisive form of qualitative research,” says Habershon. And although it requires a certain “jump of imagination”, it is becoming more mainstream.

Ekman is currently in the last stages of negotiating with “a British business group operating in a variety of fields”, which is looking at applying his work in the corporate world, both in the UK and overseas. “It’s certainly going to be useful in personnel decisions, negotiations, in bargaining, and market research,” he says. Companies like this, who are looking seriously at how facial expression can answer business questions, are still outnumbered by those who are looking for silver bullets, he says. Sadly, “there aren’t any”.



Lie to me

Over the last couple of years Paul Ekman has been spending a fair chunk of his time working on the TV series Lie To Me, in which an investigator played by Tim Roth catches out liars and solves crimes by reading facial expressions. The show is based on Ekman’s work, and he reviews all the scripts to make sure they’re a fair reflection of real science. He also writes a critique of each show in his blog, and these will be included as a special feature on the DVD of Season 2, which is out in November.

How loosely is Tim Roth’s character, Cal Lightman, based on the real-life Paul Ekman? “My contract said that he could not be the same nationality, could not be the same age, could not have the same build, could not be married, could not have two children, and could not have the same personality. They could use my science but he couldn’t resemble me a bit. I’ve gotten to know Tim Roth, and the character he plays is a very different person than me. He actually is the kind of person that nobody would talk to in real life, but he’s entertaining.”

“Most of the science is robust, except for the fact he does things with more certainty and more speed”

Ekman seems pleased with how they’ve treated his work, although it’s a “mixed bag”. “There is false information that they purvey,” he says. “Most of the science is robust, except for the fact he does things with more certainty and more speed. He’s got 42 minutes to solve a problem. There are very few problems I’m willing to solve where I only have 42 minutes. I’ve been trying to get them to do an episode based on him making a mistake, calling someone a liar who’s actually not a liar. And that can happen. I think that’s the worst piece of misinformation that the show purveys, that it’s perfect. But it’s much more scientifically accurate than other shows.”

Ekman does his best to make sure the actors get the expressions right but accepts there’s only so much he can do. “I do little video episodes for them to look at. But some of them like to look at it and some of them don’t. Tim has never looked at any of them, to the best of my knowledge. I don’t think he wants to, he wants to do it his way. As he has said, the show is about Dr Lightman, not Dr Ekman. I’ve once or twice coached an actor on the set before they did it, but I don’t have the time to spend on that. If they get it badly wrong then it’s in my critique, so people who are seriously interested can learn a lot.”


11 years ago

I still watch lie to me on Netflix. Have u ever made a cameo appearance?

Like Report

10 years ago

But the real question everyone wants to know is "how can I tell if a girl likes me?"

Like Report