Short Wave - How Do You Spot A Liar?

Episode Date: October 8, 2024

For over a century, we've been inventing technology to catch liars in the act. To this end, the polygraph was invented and became wildly popular in the mid-20th century. Then, there was an era of "micro-expression training," which claimed person could be caught lying through a skilled analysis of their face. Now, there's talk of using artificial intelligence to analyze the human voice. But do any of these methods even work? And if not ... what are the risks? Emily and Gina investigates how deception research has changed and why it matters. Check out our episode page, where Emily linked to the experts she talked to and the papers she discussed.Got another human behavior you want us to investigate using science? Email us at [e-mail] — we'd love to hear from you!See pcm.adswizz.com for information about our collection and use of personal data for sponsorship and to manage your podcast sponsorship preferences.NPR Privacy Policy

You're listening to Shortwave from NPR. Hey, short wavers. It's Regina Barber. And Emily Kwong. Okay, so between the two of us, I feel like you love TV more. Yes, that's absolutely true. It was my first and for the longest time, my only friend. And now we're a friend, so that's cool. Yeah, now I have two. But I recently found myself going down a TV rabbit hole. It's technically a reboot of a reality classic, The Mall.

Welcome to Malaysia Where players are working together as a team To carry out a series of missions To prize money in a pot That only one of them can win So you chose the cash, right? I 100% chose the cash. Within their ranks is a mole Go go go go go, go! Yes!

Someone hired by the producers to sabotage the team. Should the players trust me? No. So this is like the amazing race meets among us. Yeah, but it's not on a spaceship. This show is to in the real world. That's cool.

I'm Ari Shapiro, and I'm going to be your guide for this incredible adventure. Shout out to our news roundup buddy Ari. He hosted this season. It's our very Ari. I was so proud. And I had to ask him. Before hosting the show, did you consider yourself someone who was good at spotting a liar? Absolutely not, and I still don't.

I do not think I got any better at it in the course of the show. Tell me about that psychological unraveling for you. because people can be nervous for all kinds of reasons having nothing to do with the fact that they're lying. Yeah, I mean, I am a terrible liar. I just cannot do it. But would you say you're good at spotting a liar? I think after all these years, I'm actually better at it, but I'm still not good. Gina, you are not alone. There's this thing called Truth Default Theory, developed by Tim Levine at the University of Alabama at Birmingham. he suggests that in communication, honesty is the default.

And people tend to believe each other. Because humans are social. And everything in our lives wrap around us being able to communicate and convey information to one another. But if there is a reason for people to lie, they do. So there's a whole body of lie detection and deception research spanning multiple scientific fields. And when I told Ari this, he lit up at the thought and he gave me like an assignment. I want a playbook. I want some scientist to tell me what do I need to do to figure out who the liar is.

Okay. Like, fairy nuts and bolts, checklist, follow these guidelines. Okay. I mean, I love his, like, faith in science, for one. I like that he wants this playbook of spotting the liar. Yeah. But I also wondered, does a lie detection playbook even exist? Right.

So today on the show, Emily goes down the molehole of deception research to discover how lie detection has changed and determine if it's even possible to catch a liar in the act. You're listening to Shortwave, the science podcast from NPR. Okay, Em, where do you want to start here? I want to start with the polygraph. The machine used for lie detection. Yeah, and oddly, the early inspiration for it was a medical device to record vital signs. So a patient's like pulse or blood pressure.

I mean, that kind of makes sense, but I mean, that's not for interrogation at all. It's not an interrogation tool. No, no. That happened around 1921 at the Berkeley, California Police Department. A rookie cop and physiologist named John Larson got a hold of a psychology paper by William Moulton Marston. Okay, I think I know this guy. William Marston also created Wonder Woman, right, and her lasso of truth. Yeah, he was a little obsessed with, like, sussing out who was lying.

And he thought there was a link between vital signs and emotions, and that a machine, the polygraph, could prove that. Dr. William Marston demonstrates a complicated device whereby he claims he can determine and compute comparative emotions of blondes, brunettes, and red hair. Now, Gina, the science to back up Marston's claims was scant. But at the time, it was convincing enough for the Berkeley police chief who saw an opportunity. He wants to make the police themselves more law-abiding, give them some. science instead of violence as a tool.

And the lie detector fits into that program because he wanted to change the way interrogations took place. Yeah, this is Ken Alder. He is a historian at Northwestern University, wrote a book about the history of the polygraph. He says its use as an interrogation device started in Berkeley, but it then spread to police departments around the country and into the commercial sector. So by the 60s and the 70s, you know, millions of Americans just getting an ordinary job would have to take one. And it was constantly in the news and in politics because people who were mistrusted or somehow, you know, doubted would actually end up, you know, offering to take polygraph tests to prove their innocence. And Gina, this runaway implementation was happening without really any scientific evidence to back up the machine's accuracy, especially in real interrogation scenarios. Yeah, that doesn't surprise me, though, right?

Because, I mean, throughout history, polygraph data has not been. admissible in like most U.S. courts. That is true, but there is a catch. While its results may be not admissible in court, if in the process of a lie detection exam you confess the crime, that confession is admissible in court. And so the police use it to basically entrap people. That's awful. Yeah, the polygraph is not a lie detector.

experts have told me it's more like an anxiety detector. I would fail. Your anxiety would be palpable. Yeah. And ultimately, Ken told me, you know, the work falls on the examiner. There are professional polygraph communities with membership numbering in the hundreds, and they stand by the machine. But studies performed outside of the polygraph community have found high false positive rates. So the polygraph has many critics. Alder Dvry, a professor of applied social psychology at the University of Portsmouth in England, is one of those critics. Because it's very naive to think that in interview settings, the true tellers are completely calm because they're not. Very often in these interview settings, if you're interviewed for a long

time about a certain crime, the police believe you're guilty, but you're not, you're innocent. Yeah, I mean, like regardless of whether you're guilty or you're innocent, I mean, just being around police could be nerve-wracking. Right, so critics are rightfully cautious about these machines. Even as pop culture makes light of them, like Vanity Fair is posting videos of celebrities getting polygraphed. Yeah, I mean, I've only seen the Pedro Pascal one and I laughed my butt off. Yeah, the Kiki Palm one is pretty funny when she couldn't ID former Vice President Dick Cheney.

I hate to say it. I hope I don't sound ridiculous. I don't know who this man is. I mean, he could be walking down the street. I wouldn't know a thing. Sorry to this man. Is she? Tell them the truth about that? That's true.

All right. But is the polygraph still being used for like, I don't know, job screenings? In the security sector, yeah, for hiring law enforcement officers. But the method has largely fallen out of favor. Since the late 80s, a federal law and many states have banned employers from requiring lie detector tests with very limited exceptions. But that hasn't stopped the deception field from trying to develop new methods. there are other lie detection technologies out there. Okay, like tell me the other kinds of ways we're trying to figure out if people are lying.

Oh, it's every tech you can imagine. I mean, scientists have tried to measure lies with brainwaves, with fMRI machines, with AI. And for a while, the deception field was captivated by the idea of micro-expressions that a lying person could be caught by a skilled analyst of the human face. Yeah, I've heard of this, like being able to tell somebody's line if they're like blinking too much or they're not looking at you or their face is twitching. As like a really anxious, like kid in person, like I read people's faces obsessively, like trying to figure out if they're mad at me. And you probably have psychologist Paul Ekman to thank for that. I mean, he championed this idea for decades.

But Regina, I am here to tell you that like the polygraph, micro expressions, everything you mentioned about like blinking all that. It is also an unreliable form of lie detection. That sounds like a good film script. And in fact, it is a film script because there is no research showing that really works in that way. Here's Alderd again, the psychologists at the University of Portsmouth, who's been studying deception since the late 80s. And he, along with the majority of deception researchers, reject the micro-expression's model. And in a recent survey polling 50 researchers, a good number of them, like over 80%,

of them rejected another idea that a liar cannot look you in the eye. The popular stereotype that a liar cannot look you in the eye, this is one of the few things that experts really agree just does not hold any water. We've got about a century's worth of data to demonstrate that this is the case. This is Timothy Luke, a psychologist at the University of Gothenburg in Sweden, and he was the lead author on that survey. And like this is, you know, I know this goes against a lot of like what people think, but for me it's not that surprising that like science doesn't really totally back like how to find a liar, right? Like anxiety, gaze aversion, like your face moving a little bit. None of these things are like universal. So is there any hope for Ari? Like is there a step by step process for detecting a liar? Like what do the experts say? So it's an important question to ask. And it really depends on who you ask. I can say that the field is changing and moving in a direction away from nonverbal lie detection to verbal lie detection.

So paying more close attention to what people are saying. And then on the examiner side, it's changing from like passive interviewing where you're like in a monotone like, tell me everything you remember to active interviewing. Like asking follow up questions and analyzing specific details. Yeah, what details are they analyzing? It totally depends on the researcher. Aldert in the UK, he believes that liars and truth tellers use different cognitive strategies. So he says that in like an extended conversation, liars will keep their stories simple. And truth tellers are more forthcoming.

Truth tellers are willing to tell it all. But initially it doesn't really work that way. Because truth tellers initially don't say that much. So what you need to do, you need to ask questions. and these questions, if you ask the good questions, what you will see, the truth tellers, volunteer more information. Now, Timothy Luke, over in Sweden, he agrees that a lack of detail is a promising cue, but he actually questions the idea that there's any surefire cues at all.

It's been a conversation in the field pretty much since 2003. There was this big meta-analysis led by Bella de Paolo at the University of Virginia. And Timothy has further argued in this paper called, called Lessons from Pinocchio, that if there are cues, they may be over-exaggerated. It's not going to be a very powerful indicator. It might be an indicator in a weak, probabilistic sense. It might be detectable in a research context when you have recordings of people and you're able to carefully observe and carefully document that. But the real world has so many variables. And Timothy thinks the very best liars can trick

an interrogator. It also makes him really uncomfortable whenever law enforcement starts building policy around ideas that might not hold up to scientific scrutiny. If you have a policy where you are encouraging security officers or law enforcement to rely on potential cues to deception, that might be weak, might be unreliable, and also might be highly subjective in their perception, that is going to open the door for potentially a lot of really problematic discretion. It opens the door for a lot of bias to creep into judgments. And that can be both completely innocent and by accident as well as potentially more nefarious. So maybe we shouldn't be messing around with things that are this unreliable. Timothy thinks, and there's a lot of

disagreement in the field, but he says the best method for detecting a liar, like the only technique he advocates law enforcement use is called strategic use of evidence. It's an approach to interviewing. Where you as an interviewer ask questions in a way to get them to address the information that you have without revealing that you already have that information. That way you have a kind of untainted statement to check against the facts that you have. Wait, so basically fact checking. Yes. Fact checking, I guess is the best lie detector. of them all. Yeah, what we do every shortwave episode. And Timothy Levine at the University of Alabama at Birmingham agrees that it is the only true way. And anybody who's tried to fact check knows it's

really, really hard and it's real skill. But to the extent that you can triangulate with multiple sources and to the extent that the communication is about something factual, a good journalistic fact-checking will get you about as close as you can get. So that's the best we can do for Ari is fact-checking. Emily Kwong, thank you so much for reporting. I now invalidated that it was right that I trust no one. I wish there wasn't some truth to that, but maybe. This episode was produced by Hannah Chin and edited by a showrunner Rebecca Ramirez.

Tyler Jones checked the facts. Kwayze Lee was the audio engineer. Beth Donovan is our senior director, and Colin Campbell is our senior vice president of podcasting strategy. I'm Regina Barber. And I'm Emily Kwong. Thank you for listening to Shortwave, the science podcast from NPR.