Categories
podcast

Can gaze direction reveal clues about lying/truth-telling? A talk with deception researcher Tim Levine

Maybe you’ve heard that you can get clues about whether someone is lying by what direction they look when they talk. The most common form of this idea is that if someone is looking up and to their left, they’re more likely to be accessing real visual memories (associated with truth), and if they’re looking up and to their right, they’re more likely to be constructing visual images (associated with lies). But there is no basis for this; in fact, many studies have found evidence against that claim. This idea and other more broad ideas about eye movement clues were popularized by NLP (Neuro-Linguistic Programming), a school of thought whose core ideas have been debunked time and time again.

In this episode, I talk to Tim Levine, a respected deception detection researcher. We talk about: the eye direction idea; the huge amount of bullshit in the pop behavior analysis space (e.g., shows like the Behavior Panel); reasons why the spreaders of this bullshit are so popular and successful; what the science says about using behavior to detect deception; why it’s so difficult to use behavior to detect deception; the idea that you need to establish “baselines” for people to aid you in reading them; how behavioral patterns in games/sports can differ from more real-world non-game scenarios; confirmation bias in the behavior analysis space, and how even smart researchers can be unreasonably biased in favor of their own ideas; Paul Ekman’s work; and more. 

Episode links:

Resources related to or mentioned in our talk:

TRANSCRIPT:

(note that transcripts are rough and will contain errors)

You can find a lot of people who will tell you that you can get clues to what people are thinking, and clues to whether they’re deceiving you, based on where they are looking: for example, whether they’re looking up and to the right, or down and to the right, and so on. 

Here’s one common version of this idea: 

[TikTok video: https://www.tiktok.com/@mandrae/video/7086540364655562030 talking about how you can get clues about whether someone is lying based on where they look. If they look up and to their left, it’s visual memory, meaning it’s likely to be tied to real memories, but if they look up and to their right, it’s imagined visual, meaning that there’s a good chance it’s false and fake.]

These eye-direction ideas come from the world of neuro-linguistic programming, also known as NLP. Here’s an image from an NLP training site where it labels someone looking up and to their left as “Remembering pictures” and looking up and to their right as “Constructing pictures”. https://www.nlpworld.co.uk/nlp-glossary/e/eye-accessing-cues/ 

The upper left and upper right idea is just one specific iteration of the more general claim that you can get reliable information of some sort about what someone is thinking by the direction of their eyes. There are other assorted variations on this; including some more subtle and nuanced-seeming ideas. 

Here’s a clip from the popular YouTube show Behavior Panel talking about using eye direction to gain information. This was from the first Behavior Panel episode in 2020, in which they analyzed the behavior of Carole Baskin. Baskin was featured in the documentary Tiger King; some people suspected she had killed her husband, which is why they were examining her: 

[https://www.youtube.com/watch?v=QKpjC8rwZW0: 27:30 

Chase: You don’t see a lot of eye movement going hard right or hard left or hard anywhere they go a little bit to her left but not a whole lot just they’re barely going back and forth. So instead of accessing which would look if my face is it clear here accessing would would look more like this and her eyes were just like this.

Greg: But you can go to different parts of your brain when you’re accessing. 

Here’s another part of that episode, where they’re talking about another moment of Carole Baskin’s interview (21:50 https://www.youtube.com/watch?v=QKpjC8rwZW0):

Greg Hartley: If you notice in every other place go watch when she’s describing her father building these cages and she’s accessing and she’s remembering it’s for buy and she’s using data yes he’s recalling her eyes are drifting right she uses that as an illustrator as well she’ll make her points by doing this this is the only time I see her eyes go to her left as she’s describing the number of accidents he had and he was getting dementia if you go back and look at it her eyes deviate from that baseline fairly significantly as well as breaking contact. Chase: I hundred percent agree her baseline has her doing recall of looking up into the corners and when she broke I don’t know if you want to just maybe play this again in the in the final video but when she broke this time her eyes stayed towards the middle they

were still focused on an object that was off-camera.

This idea, that you can get reliable information about what someone is thinking or about their likelihood of deception, is repeated by many people, in a variety of iterations. It’s even been taught in some law enforcement and interrogation trainings.

A 2021 paper titled “Investigative Interviewing: Research and Practice” includes information about the use of NLP-associated ideas in investigative work: 

Criminal investigators describe NLP as useful for developing rapport in an interview or Interrogation, where the focus is on the interviewer matching an interviewee’s nonverbal behavior, the manner in which they speak, and their choice of words. More often, NLP has been proposed as a way of helping an interrogator discern truth telling from lying in criminal interviews and interrogations. Here the focus is on an alleged relationship between eye movement and thought: for example, if right-handed people are visualizing an imagined event (i.e., something they are lying about), they are likely to look up to their right; if they are visualizing a remembered event (i.e., presumably something that they are not lying about), they are likely to look up and to their left. 

One example of this: a 2012 paper by the interrogation and interviewing consulting firm Wicklander-Zulawski was titled “Misconceptions about Eye Movements: Part 2.” https://www.w-z.com/wp-content/uploads/2012/10/Column-LPM0312-Interviewing.pdf  In it, they described how noting person-specific patterns about which eye direction was linked to accessing memory can lead to hugely reliable deductions about someone deceiving or not. They describe an interrogation where a subject’s eye movement patterns allowed them to conclude she was lying about denying using drugs on the job. They concluded that example by writing: “The subject’s eye movements during this exchange helped in gaining an admission.”

Because these ideas are so common, it’s understandable that many lay people think there’s validity here. Many people will conclude, “So many people are talking about this idea; there has to be some valid information there, right? It must be that eye direction gives us valuable clues to what people are thinking and whether they’re making stuff up, right?”

And yet, no, there is no evidence that one can use someone’s eye direction to get useful information about what they’re thinking about, or deduce whether they’re telling a lie or not. There’s actually good evidence against that claim, as numerous studies have found no useful correlations. Now, to be clear, this is not to say there are not person-specific tells; people can have all sorts of idiosyncratic tells and patterns when it comes to behavior, and we’re not talking about that. We’re talking here about the idea that these eye-direction ideas can be used for the general population; that there are common types of patterns that can be found amongst the general population that help us read them and get clues as to whether they’re likely to be telling the truth. And again, there is no evidence for that idea. (If you disagree with me on this, I invite you to send me a message, as I will be doing a deeper dive on this topic in the future.)

Now many people who spread these ideas are, I think, major bullshitters; many of them are on the highly deceptive, unethical end of the spectrum; for example, you’ve got people like Chase Hughes, who claims to be an expert at behavior and influence and whose immense amounts of deceptions and unethical behaviors I’ve examined in other episodes. Other people who spread these bad ideas I think are true believers of various sorts; I would count Greg Hartley, also of the Behavior Panel, in the true believer group; he and Chase’s ideas on behavior are heavily influenced by NLP, and these ideas are commonly held by those who embrace NLP trainings, even though NLP ideas have been thoroughly debunked. 

Some people, like the Wicklander-Zulawski organization I referenced, I think have done good work on interrogation patterns and strategies in other contexts; I actually interviewed David Zulawksi when I first started this podcast https://behavior-podcast.com/tips-on-interrogating-people-for-information-and-confessions-with-david-zulawski/ and I found it an interesting conversation and it was one of my more popular episodes. I think they’ve just made the mistake, as many in law enforcement and interrogation work have, of defending some old, outdated, but common ideas that have no basis in evidence. 

This is just to say; my criticizing these ideas is not meant to imply that I think everyone who spreads the various forms of these ideas is purposefully lying or being unethical; it’s a spectrum, as with everything. But the behavior analysis space, especially the more pop-behavior-analysis space where people like the Behavior Panel make claims of frequently getting reliable information from assorted interviews and speeches, is full of bad information; the people in this space, and in NLP, have incentives to exaggerate what’s possible with interpreting behavior. They even have internal incentives to persuade themselves of some of these ideas — and it’s easy to persuade ourselves of faulty ideas, especially when it involves an ambiguous and high-variance information source. 

Later I want to do a much deeper dive into this topic, as I’ve done a lot of research into it and it makes for an interesting history of how these ideas came to prominence back in the 70s and how NLP was involved in that, and why these ideas are still so popular, and the various iterations of these ideas from the more easily debunked to the harder to debunk. And I’ve not seen anything like that deep dive elsewhere, that delves into these ideas in such depth. If you’d like me to spend time working on that deeper dive, please let me know in the comments and send me a message, as the more encouragement I get, the more likely I’ll work on it. But this episode will not be a deep dive in that way; it will be a talk with deception detection researcher Tim Levine; Tim and I talk about the eye direction idea, and we talk about the huge amount of bullshit in the pop behavior analysis space, and some of the reasons the spreaders of that bullshit are so popular and successful. We talk about what the science says about using behavior to detect deception. We talk about why it’s so difficult to use behavioral information to detect deception. We talk about the frequently heard idea that you need to establish “baselines” for people and that this will help you read them. 

We talk about poker tells and how behavioral patterns in games can differ from more real-world non-game scenarios. 

We talk about confirmation bias in the behavior analysis space, and how even very smart researchers can fall pray to the mistake of being biased in favor of their own ideas. We talk about Paul Ekman and whether some of his teachings about behavior have gone awry due to his burning desire to prove his own theories. 

If you’re interested in behavior, or you just like debunking bullshit, you’ll like this episode. This is actually my second talk with Tim Levine and if you like this one you should go back and listen to that one; in the first episode with Tim we talk about his truth default theory; basically the idea that we will tend to believe people unless something triggers us and gives us a reason to not believe them; this helps explain why we can be so gullible about so many things; why it’s so easy for us to fall pray to scammers and con artists – which, by the way, include some of the same people out there who make grandiose claims about being able to teach you how to read and influence and manipulate people.  

You can learn more about Tim Levine at https://timothy-levine.squarespace.com/. I’ll read a little bit from his bio on his site: Levine has published more than 160 refereed journal articles reporting original research related to communication and he is an internationally recognized leader in deception research.  He is the author or co-author of Information Manipulation Theory, Truth-Default Theory, the Veracity Effect, the Probing Effect, and the Park-Levine Probability Model.  His research on deception has been funded by the National Science Foundation, U.S. Department of Defense, and the FBI.  His current research focuses on what makes some people more believable than others, the prevalence of lying, and on effective interrogation strategies. Levine’s book, Duped, describes his program of research on deception relevant to Truth-Default Theory.

In this talk with Tim Levine, we focus mostly on the most extreme and easily debunked form of this idea; the idea that someone looking up and to their left is associated with real visual memory; and the idea that looking up and to one’s right is associated with visual construction. That is still a common idea despite being so clearly false and debunked. But our talk also pertains to the more quote “sophisticated” version of these ideas; and I put sophisticated in quotes here; the idea that people will tend to have one of two patterns; looking one way for recall and the other for construction, and that you just have to figure out what their pattern is; I call that the more “sophisticated” version because it’s harder to debunk and can come across like a more advanced, sophisticated version of the old, more clearly false idea, but again, there’s just no evidence for any of it. There are no studies supporting these ideas, even as so many speak as if they are extremely reliable. And that should be a red flag, as Tim and I talk about; when people act as if a behavior pattern is highly reliable and contains a lot of information, and yet nothing has been found in studies despite people looking for it, and actual behavior researchers don’t think it’s legitimate, these are all red flags there’s likely some bullshit involved. 

Okay here’s the talk with Tim Levine…

Zach: [00:00:00] Hi, Tim. Thanks for joining me again. 

Oh, you’re welcome. Happy to be here. 

Yeah. Good to see you. Uh, so yeah, uh, maybe we can start with, you know, you’ve been in the deception research field for quite a while. I’m wondering if you can maybe give a synopsis of, of how, uh, what your views are. On the whole like neuro linguistic programming associated idea of what quadrant you look in upper left, upper right, that those things can be tied to.

You know, somebody accessing their recall, making it more likely to be truth or accessing their more creative side, making it more likely to be made up in a lie. You know, we see these ideas a lot. So I’m wondering if you have a A rough summary of how you view those ideas in your many years in the field.

Tim: Uh, so, uh, were [00:01:00] you intentionally looking up while you were talking to me? 

Zach (2): No, I 

Tim: wasn’t trying 

Zach (2): to 

Tim: do. What should I infer from that? 

Zach (2): I’m going to be thinking about this too much now. Oh, sorry to do that to you. Got to throw it in there. 

Tim: So I’m not in any way, an expert. on eye behavior as it relates to what people are thinking about.

My area of expertise is all in, uh, deception. And, uh, to the extent that, uh, the neuro linguistic stuff and eye behavior is linked with truth and deception, uh, then I can say with pretty good confidence that there doesn’t seem to be any relationship at all between eye behavior of any sorts and whether or not people are honest or not.[00:02:00] 

Zach: So, yeah, the surprising thing to me is, I mean, that’s my understanding. I’ve looked at a good amount of research on this. debunking it. You know, there was the well known one from 2012, Richard Wiseman debunking it and some, some others. Uh, and I just keep seeing the idea, and it, it is really surprising to me because I, I hear people saying it quite confidently, like, you know, for example, this guy on the behavior panel, uh, Greg Hartley, will say, like, it’s very tied to, you know, uh, recall and, and, uh, or very, very tied to, to, uh, deception, highly correlated, and it just kind of surprises me because I’m like, you would think if something was so highly correlated, they would have found something A little something in the research, but you know, it’s so, it sounds like, yeah, you, you have not heard of any.

research that stood out saying like there’s something there in those terms. 

Tim: Uh, no, and I have, um, so [00:03:00] as part of my deception detection work, I’ve, uh, been collecting these, uh, uh, videotapes of people. Uh, I, I bring participants into lab, you, you know, of this work, uh, and give them, uh, a reason to lie. They, uh, they’re playing a trivia game and they get an opportunity to cheat.

And it’s up to them whether or not they cheat or not. Um, but if they do, they might be in trouble because it’s a university setting. And then we interview them about this. And I have 485 tapes. And, um, out of that, uh, there’s one liar. With the sort of 

up, 

Tim: right? So there might be people who do this. Um, but it wouldn’t ever show up in social science because most people don’t.

And, and even that one, uh, was an international student, uh, who [00:04:00] was, uh, obviously speaking in second language. And so it might’ve had nothing to do with the fact that she cheated and lying. And, uh, it might be that. you know, the communication task was quite difficult for her. And, and I have met individuals who do eye things when they’re thinking.

Um, my wife does, um, my wife has a visual memory and, um, when she is recalling things, her eyes go up, um, and, but this isn’t. You know, this is a thing that’s unique to her. Right. And, and people have particular kind of eye behaviors and eye patterns, but I know of no evidence that those sorts of claims are general across people.

Zach: Right. I [00:05:00] think that’s an important nuance because it’s like. Yeah, I mean, to tie it to my Pokertel’s work, it’s like, there are Pokertels that you can use that A, are pretty valuable and are not studied, like there’s no formal studies, so it wouldn’t be surprising to me to learn that there were patterns that just haven’t been studied or haven’t been studied well, but then, and then there’s also player specific patterns, right, like that you wouldn’t find in a general, like that you wouldn’t find in a general.

Population, but yeah, for that first case, like it wouldn’t shock me to learn that there was some kind of upper left upper right. It’s a general pattern, but like for the people that speak as if it’s like a really highly correlated pattern. That’s what gets me. It’s like, it’s, it, it kind of clearly can’t be that, uh, highly correlated a pattern if you, if all these studies, multiple studies have not found anything and, and for, you know, so I think, I think, I think that to me is the important part because like, yeah, sure.

There [00:06:00] could theoretically be something there, but if someone’s talking as if like, this is a very highly reliable, Yeah. Clue to something related to deception. Like that’s to me where you get into like, you should really red flag should really be waving. Cause that’s, there’s nothing supporting that. And if it was such a highly correlated thing, somebody would have found something by now.

So that’s kind of where I stand on it. It’s like, it’s not that I’m like skeptical of everything out there that somebody says, but I think it really, when, when some of these people speak in the really highly confident ways. That’s what really bugs me because I’m like, there’s no way it can be, you know, highly correlated like that for a general population, but yeah.

Tim: You know, I’m a pretty experienced researcher and when things are highly correlated, if you’ve seen lots and lots and lots of data, it probably only takes about 20 people to see the pattern. 

Zach: Right. Highly correlated. Yeah. Right. And 

Tim: if I’ve [00:07:00] done 400 something and I’m not seeing the pattern. Right? Then there’s either, it’s a tiny correlation that might be real, but you only need big data to see it.

Zach: Right. 

Tim: Um, or it’s not there. And either way, it’s not going to be at all practically useful in any way. 

Tim (2): Right. 

Tim: Right? So I, I think we can really, really confide. at least as related to deception, rule out the idea that there’s a strong correlation there. I think, you know, just, just the way statistics work and, um, you know, if strong correlation was there, it would, it would show up in data and it would show up across data sets.

And it’s very clear it doesn’t. 

Zach: Yeah, so that’s and I think that’s a real important point to a general point. I mean a lot of these behaviors Studies that you find or just maybe just research in [00:08:00] general. They’ll find a correlation, but it will be very weak, right? So it’s like even if there is a correlation there the chances of it being like Practically useful even if it even if it is there which sometimes is in a doubt if it’s a really small, you know small correlation Sometimes the research papers make it, like, just if you read the research papers, it makes it sound like there’s something that could be meaningful there.

But a lot of the stuff, and I’m not even talking about deception, because that’s, you know, that’s a specific area too, but, it just seems to me like sometimes the, the, the papers, and then the way the media reports on it, will make it sound like there’s this, like, large correlation there, but you actually read the paper and it’s like, Oh, it was like, you know, a few percent more likely in this one scenario, but I’m curious if you have thoughts on that, like, do you see that often happening in, uh, like science reporting for behavior?

Related things where it exaggerates how much correlation is there? 

Tim: Uh, [00:09:00] yes, absolutely. And I was writing about that, uh, just before we got on together. Uh, so one of the, one of the problems is that researchers rely on something called significance tests. And the way they’re most often used is what they’re being used to statistically rule out.

the idea that there’s nothing there. And it’s real easy, even if there’s something there, right? If you can rule out nothing there,

that means there’s something there. But that there’s something there doesn’t mean that there’s much there. So when I’m teaching to this, my students, if I have a dime in my pocket, It means I’m not broke. I don’t have zero money, right? But that dime’s not going to take me very far. It’s not going to buy me a cup of coffee or a beer or even a candy bar, right?

And there’s a big, big, [00:10:00] big difference between having a dime and being a millionaire. Um, so ruling out broke doesn’t necessarily tell us much. In the context of deception, uh, the latest data is that the best tell to deception. Uh, has to do with the number of details in an account. And on average, honest people provide more detailed accounts than people who are deceiving or lying.

And this effect is somewhere between, uh, one third of a standard deviation and half a standard deviation. Um, so let’s, let’s think about what that means. So if somebody gives a detailed account, does that mean that they’re honest? No, of course not, right? And if somebody [00:11:00] doesn’t provide much details, but if you really wanted to use this in an instance, you would need to know what I would call a cut point.

So how many details does it take before it proves that somebody’s honest? Right? And the second you start to think about this, you realize how absolutely silly it is to take a statistical finding that occurs, you know, across a large number of people in very tightly controlled situation with all other things being equal and try to apply it outside of that.

For example, uh, I don’t have a visual memory. I mentioned earlier that my wife has an incredible, she’s probably like one in a thousand, one in 10, 000 visual memory people. Right. You’ve got the 

Zach: aphantasia thing, which is, you know, which is how I would describe my own. And 

Tim: yeah. But I have no visual memory.

So, uh, [00:12:00] I, I can’t pull up. So if you’re asking me to describe something visual, I’m not going to be able to provide, I give you gist, but I can’t give you any details. Uh, this doesn’t mean I’m lying, 

Tim (2): right, right, right, 

Tim: you know. Now, if you want to give me, if you want me to give you details about, you know, my latest study, then I can give you all kinds of details, right, right, right.

Um, and I’m, and if you’re talking to me, I can give you a lot more details about like the last study I read than the average professor can about the last study they read. Yeah. Um, but I can’t like tell you. The people, the faces of the people in my classroom, 

Zach: right? The level of detail, it might have a correlation, but it’s not practical for all, for all intents and purposes.

It’s not practically useful for, for practical. Yeah. Yeah. 

Tim: Cause what that correlation means [00:13:00] is right. All things being equal in a carefully controlled environment across large numbers of people and just all bets are off in particular situations when there’s other factors at play. 

Zach: Mm hmm. Mm hmm. But I can, can you still hear me?

Yeah. Um, yeah, the, um, though to play devil’s advocate, I can imagine like if you, if you were in a police interrogation scenario, I imagine outside of the lab, if it’s like a consistent thing where somebody is not able to give you like pertinent details and when they’re giving a report, I can imagine scenarios where you’re like, You know, okay, well, this seems to make it significantly more likely based on this specific scenario that this person’s, you know, um, making stuff up or, or telling a lie, but then I think the pertinent question, even if you think that is like, what do you do about it?

Because, and usually I think in interrogations, that’s not going to be the only clue, right? [00:14:00] Like you’re not, I don’t think it often happens where they’re like basing a big decision based on like, You know this guy left out some pertinent details. I think that’s what gets left out is like, you know, investigators may have a feeling one way or the other, but like rarely is that like making up a, you know, a decision point of where they spend their time.

Usually there’s going to be some like other evidence involved or some reason to go down a path, right? They’re not just like, right. He left out some details or other some other behavior thing and they’re gonna they’re gonna like go down this path based on that and I think that often gets Left out. It’s like in the case in the big picture of things all these things are can be pretty minor Even if you think they are A factor, at least that’s how I view it.

Tim: Exactly. And let’s, let’s think about your particular example. So we got the detective is interviewing a suspect and listening for details. It probably matters a whole lot how long ago the thing was, right? So I’m going to have [00:15:00] a, I’m going to have a lot better memory of a recent event than a distant, and that’s going to affect how detailed things are.

It’s going to matter how smart I am. If you’re, you know, Interviewing somebody who’s two standard deviations above average versus somebody who’s two standard deviations below average, you’re going to get very, very different degrees of specificity. 

Zach: There’s all sorts of context. Yeah. There’s all sorts of things in the mix of like.

Whether somebody would judge this was abnormal or, or not. 

Tim: Yeah. Was this a typical thing or something that was really memorable? 

Zach: Right. 

Tim: Right. What was their emotional state when this was going on? Right. There’s so much in there that’s going to affect how detailed you are or where your eyes are moving than just the fact that you’re lying or telling them.

Zach: Yeah. Well, let me, we’ll get back to the general behavior things, but I, to get back to the eye direction. [00:16:00] Do you have a sense of where those ideas come from because I, I get this, I get this rough sense that, you know, I think there is some evidence that looking up people generally like to look up when they’re recalling things because basically because it’s like a clear field of vision and it’s like a place that’s not distracting.

So you generally, people will naturally sometimes look up because it just is somewhere to look. I mean, so I can understand that part of it. And then I kind of understand where they, the NLP people got the left and right idea because of the left and right hemispheres, one being tied to more, you know, concrete things and one being tied to more creative things.

So I got it, I kind of get back to, and then there’s also these things where they’re like, if you’re looking in the middle sphere, it’s more auditory. And they also have this thing, if you’re looking down. It’s more, um, sense related, like, uh, tactile, which kind of makes sense because you’d be thinking about, like, your hands touching something.

So I could, you know, just to say, I can see where they got the ideas, at [00:17:00] least that’s my rough understanding, but I’m curious if you Have your own thoughts on where those ideas might have come from or originated from. 

Tim: Uh, I think a lot of ideas come from legitimate observations of people, uh, where we don’t realize how idiosyncratic the observations are.

Right? So we notice certain trends, and we really notice them. Right. But then what we do is we over, over extrapolate. Yeah. Yeah. Over extrapolate, overgeneralize. Right. Uh, I don’t think people appreciate just how, how different people are person to person. This is a huge problem in, let’s say like brain scanning, right?

Because not everybody’s brain works the same way. . 

Zach: Mm-hmm . [00:18:00] Yeah. No, it makes sense. And the, the, um, the, the, uh, bias to, once you start thinking something is a, is a clue, you’re more likely to notice it. And, uh, I mean, I can see how that’s, especially if you’re not a, a scientist and you, you start having a theory about something and you’re like, oh, it really seems to hold up.

You know, but you’re, you’re just biased and you’re remembering, you know, like we all do for various things. We’re just remembering the times it worked and the. And forgetting the times it didn’t work, de emphasizing the times it didn’t work. But I can kind of see how that played into the NLP neuro linguistic programming where those guys kind of thought they were geniuses, which I think the narcissistic element of some of this stuff, you know, kind of plays into like, Oh, I have this theory that is going to make me.

You know, show, show my genius. And I’m, they’re really motivated to like find the evidence for it. And then they start, they really start believing like, Oh, this is, this is so important, you know? Um, [00:19:00] So 

Tim: I have these different categories of researchers. Um, I think there’s, there’s some researchers who are really all about the science, right.

And they’re trying to learn stuff. And, um, for them, it’s, it doesn’t matter about being right yourself. Right? It’s about trying to, um, find 

Zach: the truth. 

Tim: And if you’re wrong, hey, you know, then, you know, you learn something too, right? And there’s no, no difference between whether you’re right or wrong, as long as you learn something, right?

And even if you know, you don’t know, it’s also important. thing of knowledge. Uh, then there’s, uh, researchers out there that are just playing a publication game to, uh, get tenure and get a job and to be good in the race pool. And, um, [00:20:00] uh, for them, it’s, it’s just kind of a thing. It’s a, it’s a grind. 

Zach: Yeah. Yeah.

Tim: Um, and then there’s a third kind of people, which I call the crusaders. Or the true believers. And, uh, some of them are very, uh, intent on being the genius and being right. Others of them have a particular cause, right? And they don’t want to let data stand in the way of either their brilliance or their cause.

And so there’s all kinds of, uh, tricks that researchers can play out of view, uh, to make their findings, um, look supportive when they present them. And when somebody else tries to do the study, um, it doesn’t come out. Um, and then they have, and then they have to like 

Zach: search for reasons why it didn’t come out because they’re really [00:21:00] invested in the idea behind it, right?

Tim: Yeah. And they get into this, uh, uh, circular logic that goes something like this, um, I’m right. If you find something that doesn’t agree with me, it’s because you didn’t do your study right. I know you didn’t do your study right because you didn’t find what I know is right. 

Zach: Right. Yeah, this almost seems like a good idea for a study, studying researchers for the different types of, um, you know, motivations and separating the, um, ideology or, or something from the, the true, the, the, the true questers for, uh, truth and such.

Um, do you have, do you have thoughts? Are you willing to share, like, you know, how that maps over to some researchers, specific researchers in the behavior space? I mean, cause I’ve heard like, Let’s take Paul Ekman, for example. Obviously, he’s done some good work, but I [00:22:00] also have seen, you know, criticisms that he is a bit too, you know, uh, set in these, some of these ideas, and that even when they turn out to not be true, they still, you know, him and his, I think it’s the Ekman Foundation or Institute or something, they still promote some of these ideas that are, uh, that seem to be, I’m curious if you have any thoughts, whether it’s on Ekman or anybody else, would you like to share thoughts about the behavior and or deception sphere at all?

Tim: I’m, uh, very likely to, um, get into particular names, but I can say with. Uh, that I’ve only kind of met him, seen him talk once in person. And, uh, when he did his presentation, uh, he was very explicit, um, that what he was trying to do was [00:23:00] stick to his guns, no matter what. And so I don’t have any problem calling him out on this because if I heard him correctly, this was self, a self classification in that camp.

Um, 

Zach: And I’ve heard that criticism from other people. Yeah. So it’s. That doesn’t surprise me. Like he, yeah, I was kind of shocked 

Tim: that he would say it aloud. Um, I thought, I thought that probably took a whole nother level of arrogance. And maybe I can get 

Zach: the exact, is it, was that a recorded thing or just something you, you heard?

It might’ve 

Tim: been, it was at, uh, the second Decepticon at Stanford. 

Zach: Hmm. Okay. Well, I can probably find something equivalent if I can’t find that one. Cause I think that’s, I’ve heard people make that criticism. So, um, but maybe we, yeah, maybe we can. pivot to the, you know, when it comes to the general, uh, behavior for use for deception, do you want to [00:24:00] share any of your general thoughts about, you know, how, how useful that is, or anything bugging you in that space that you’ve seen recently, anything like that?

Tim: I don’t know so much recently, but there’s this longstanding belief in the deception literature, uh, that there are these cues or tells to deception. And my reading of the whole literature, as this has pretty clearly been debunked, I think there are ways And that’s what you and I 

Zach: talked about the first time I interviewed you.

Yeah, 

Tim: and there absolutely are ways to detect deception, but it’s not by reading people’s cues. It’s not by listening for details. So details are a great example. So the Q people all want to count up details. And if you’re giving me a really detailed account, um, then you’re probably honest. And if you’re, seem to be avoiding [00:25:00] details or can’t bring them up, then you’re probably lying.

In my view, what you want to do is you want to listen for what the details are and see if you can fact check them. Right? So if you’re giving me details that don’t align with the truth as I know it. 

Zach: Right. The actual evidence. Yeah. 

Tim: Yeah. 

Zach: Yeah. 

Tim: So, so. The point in the deeper point is usually in deception detection.

We’re not interested if somebody’s lying or not, we want to know what the truth is. Um, and then the question is if what they’re saying Once we know that it’s not truthful, then we can ask ourselves, are they lying to us, or are they just misguided? Um, I, I wondered some of this with modern politics. How much, you know, when, when people are, when politicians are saying things, uh, that are truly false.

And obviously false, and easily fact checkable and provable false. Uh, are they believing their own bullshit? [00:26:00] Or, right? Or, or are they, uh, you know, and being sincere in this falsehood? Or are they just, uh, duping people? And, and of course it might be a mix of both. Right. Right. But, but I think usually what we want to do is we want to know what’s true.

And, um, you know, it’s only if I really have a relationship with you that I care. Are you being honest with me? Uh, otherwise it’s good enough for me to know what’s true or not. 

Zach: Yeah. And you bring up an enter a good point there because I mean, people can fool themselves into, you know, people’s people often say things that they really believe that are clearly untrue.

And I, and I actually. When it comes to, I mean, there’s studies showing that, that narcissists, people with more narcissistic personalities, especially, can convince themselves of things that are clearly not true, or, you know, things that most people would say, like, that’s not true. So, it opens up [00:27:00] this space where it’s like, yeah, it’s, sometimes it’s even hard to tell if someone is knowingly telling a lie, even when you know that they’re Telling something, saying something that’s not true.

So, um, yeah. And I, so I wanted to ask too, I think, cause I think the, uh, I think the devil’s advocate response or the people in the behavior, the pop behavior kind of space who I often criticize, they, you know, they often do this thing where they’re like, Well, you need to get the baseline, and if you have the baseline, you know, then you can tell deviations.

So they might say, like, well, yeah, maybe it’s not that valuable, but, uh, maybe, you know, once you study them for a while in the, in the context of an interrogation or something. But to me, you know, I think this is often just, uh, just covering up for bullshit. Because what, what they’re really trying to do is have it both ways.

They’re trying to acknowledge. that a lot of this, the behavioral cues stuff isn’t that useful, right? So they’re saying like, well, you gotta check the baseline. But in any practical, you know, sense, like if you’re studying some [00:28:00] interrogation footage or some speech of somebody every time, like, you would have to rack up such a huge amount of baseline to like, judge things like small behaviors or things like small verbal things.

Like, you would have to rack up, you know, for all these things that happen pretty infrequently. Like, but the, but the way they talk about it in these kind of like pop behavior videos, they’ll act as, they’ll speak as if it’s highly reliable, but then they’ll occasionally be like, well you gotta check the baseline.

But it’s like, that’s just, to me it’s just a way to wave away the fact that these things are, are, are barely reliable, if they are reliable even, like, because a lot of things they say are just like the eye direction thing, it’s like, that’s kind of clearly bullshit, and like, Yeah, so I’m curious what you think of all that, this whole, like, you got to get the baseline 

Tim: stuff, 

Zach: you know.

Tim: So the baseline has been researched way more than the I stuff. I mean, there’s a lot of research on this dating back, uh, to one of my former [00:29:00] professors, Jerry Miller, um, in the 1980s. And it is clear. Uh, baseline doesn’t help much. Mm. I mean, best case scenario, it might move you from like 54% to 58%. Mm-hmm

Mm-hmm . Um mm-hmm . But, but that’s, I mean, that’s best case, right? Um, it’s, it baselining doesn’t help much. Mm-hmm . Uh, the argument I’ve heard, uh, against kind of the, uh, go with evidence and go with facts is, uh, well, sometimes evidence and facts aren’t there, and at least in a. Criminal Context. My answer to that is, what do you do when interviewing people going in cold with no facts, right?

It’s not like police pulling random people for no reason, right? Yeah, exactly. I don’t want 

Zach (2): to go out in the street and know nothing about them. I got to go only on his [00:30:00] vibes, you know. Yeah, 

Tim: I can just see like the seals. You know, randomly, um, going in and snatching up random people from, like Yeah, I mean, I think that’s the, I mean, 

Zach: I think that’s the, that’s the really important part.

Cause like, it’s like, in most, in almost any interrogation interview kind of setting, you’re gonna have, like, there’s a reason you’re doing that, and you’re looking for actual evidence. You’re not just going in blind, being like He looked a certain way or did this thing and therefore I’m going to change the direction of my investigation based on this little, little thing he did, you know, like it just, it’s kind of, it’s kind of, uh, leaving aside how these things actually work in the real world.

Uh, oh yeah. So, but the, yeah, to, to think of, to, to mention a specific thing, like I was, I’m remembering the behavior panel. You know, one of their first ones, actually their first one was, uh, analyzing, um, the Tiger King. And so they were analyzing Carol [00:31:00] Baskin’s, you know, behavior, how she spoke and where she looked and, you know, this is like a common thing.

They’ll them and other people in this space will be like. Oh, you know, she deviated from her baseline there when she was answering, answering that question. She looked, she looked, uh, a different way and she talked faster or whatever the thing is they say. But it’s like, yeah, to, to reiterate that point, it’s like, there’s all sorts of reasons why she could be doing that.

Like it’s a slightly more emotional thing. She’s, her emotions change. It’s uncomfortable. She’s thinking about whatever, like there’s, A bird flies by. Yeah, there’s, there’s, there’s literally so many things. And for, and for, And to say, like, that that ever could be meaningful, you would literally have to put Carol Baskin in and study her for, like, you know, a hundred hours or more to, you know, get a sense of, like, what actually are her patterns and when does she deviate?

And then that would even imply that you could even That can easily link like, lies to, you know, which, which is itself hard to [00:32:00] do, like we said, like it’s hard to tell lies from something someone has fooled themselves to believe. All these things. So there’s just so much massive complexity, but that all gets waved away and these Pop Behavior analysis people’s things where they’re like, Well, you got to get the baseline, but still all speak as if this thing is highly valuable and it’s like, well, when is it valuable.

If you just said you had to get the baseline and how are you going to get that baseline? You know So this is just why I mean, I’m, I’m trying to emphasize, like, there’s just so much bullshit in this space. It’s wild to me. It’s, it’s wild. 

Tim: But there’s a market for bullshit. There really always has been. 

Zach: Yeah, there really is.

Right? 

Tim: This is, this is all through human history. 

Zach: Yeah. Right? 

Tim: There’s been bullshit and there’s been market for bullshit. 

Zach: Yeah. What do you think explains the market, the demand for the behavior stuff specifically? 

Tim: Um, wishful thinking, mostly. 

Zach: Do you think it’s, um, But wouldn’t [00:33:00] it be cool 

Tim: if it was true? I mean, what, what all of this almost always has in common is wouldn’t it be cool if it was true?

Zach: Do you think, uh, It’d be really 

Tim: helpful. 

Zach: Do you think there’s a I mean the way the way I think of it is it’s almost like people want to have a special ability like they’re questing after like I’m one of the special ones that can recognize this stuff. And I mean, I see that I see that in this, uh, like the behavior panel Facebook page where the fans are like posting a video.

And they’re like, saying all this bullshit about what they think happened in this video. And it’s just like, it’s just their biases and their dislike of people coming out. And so they’re, they, they might hate some royalty, member of the British royalty. So all their dislike, you know, they’re, they’re filtering all these behaviors through their dislike of these people.

And so they’re just like, she moved her head that way. It shows she’s a, you know, a filthy liar, blah, blah, blah. They’re just, they’re just, and, but they’re, they like to embrace the idea that they’ve learned some interesting or powerful [00:34:00] Tool that is amongst the special set of people, but that’s how it strikes me as why the demand for that is there 

Tim: Well in their defense now i’ll play devil’s advocate.

Okay. I like that. Um, I think I have some special knowledge right and um I think it’s kind of cool that I do, you know, and because I have the special knowledge, you know, you’ll invite me onto this podcast and, and there’s, there’s like good stuff that happens when you have special knowledge. So I, I can understand why people, I think that, that feels psychologically good.

And there’s, you know, some advantages in life. If you can convince other people that you have it,

right? Nobody [00:35:00] wants to go see the, uh, the doctor who really doesn’t know what they’re doing, you know, but if they think you, you know, you’re the, the specialist who really is going to cure my cancer, um, you know, there’s a lot of people willing to pay big bucks for that. 

Zach: Yeah. And I guess that’s. I guess we would say how you approach the quest for special knowledge is important.

You need to be hyper, hyper skeptical and not fall prey to the, uh, desire to have special knowledge quickly, basically, I would say. Um, yeah. 

Tim: Yeah. And you have to kind of. A, work hard to get it and B, be willing to have it to independent tests,

right? You know, so if my ideas and my findings don’t hold up on other labs, look at them, you know, then, uh, [00:36:00] 

Zach: Yeah. Uh, I don’t know if you want to, uh, get into this, but I think some people, and I talked about this in my last interview with you, but I don’t know if you listened to it cause I just added a note.

But I was trying to tie in because I think a lot of people would be just surprised at me who I’ve worked on. You know, I was a former poker player. I wrote books on poker tells a lot of people would be surprised that I agree with you, that deception detection is like very hard with behavior, if not mostly impossible, but the clarification there is.

I think that I think poker tells the ones that are useful are not really about deception because people didn’t think like bluffing is deception. But what usually happens, we’re just talking about patterns that, you know, for example, there’s different categories, but one category is people are just much more relaxed when they have a strong hand.

I mean, that’s one category, and that has nothing to do with deception, right? Like somebody. Making a big bet in poker and they’re very relaxed. They just do things that a [00:37:00] buffer wouldn’t do a buffer’s more You know more tied down and like so a lot of the more relaxed behaviors from bettors are tied to very highly correlated to You know, having good hands, and that’s one example, and then there’s other examples of attention, like, you know, somebody, like, uh, staring at their cards, for example, they, they, people tend to not look at their strong cards long, because they like to hide their They’re, they’re, they’re power or their treasure, right?

Which means that somebody that looks at their cards for a while tends to equal a weaker hand that they’re just not that interested in. And that has nothing to do with deception detection either. It’s just kind of an attentional thing or like a, a desire to hide value when we have something good. So just to say there’s these classes of tells that are quite valuable.

Uh, and, and I’m not the only one who thinks this. Experience poker players who think these things but these things have nothing to do with deception detection, right? So I think that’s I just want to throw that in there because I think people would be surprised that I agree with you on the [00:38:00] on the behavior and the deception front But I don’t know if you have any thoughts on that or if you anything comes to mind about uh, the different You know, these other categories of attention and so on.

Tim: Um, no, I think, you know, there are behavioral things. You know, we’re more likely to smile when we’re happy. Um, you know. 

Zach: Yeah, and a real smile that, like, affects our eyes. Yeah. These kinds 

Zach (2): of things. There’s all these, yeah, there are all these patterns, yeah. 

Tim: Right, and when you’re engaged, you know, the fact that we’re having a good time talking to each other will be Evident to anybody who watches this, right?

Cause we’re doing all this, like engaged communication stuff. Yeah. Um, that’s. This is, this is a real thing and this is, this is how people act when they’re, uh, into it. We’re not hating this. You can get the read that we’re not, we’re not hating 

Zach: this interview. 

Tim: No, no, no, no. This is, this is like a good back and forth and, and this is communicated, [00:39:00] uh, in part non verbally.

So, you know, non, non verbal stuff is pretty good at, uh, conveying emotions. People can seek to hide those, but, but generally speaking, um, you know, things like how engaged you are in the conversation, and, uh, do you like somebody or dislike them, or what are your major emotional states, um, are you tired, are you energized, uh, these things, these things come off.

And they’re, they’re Real things it, it, you just got to be really, really careful in reading them to use your poker example. And you would know much more about this to me, but my guess is the elite players know this stuff and that these things are going to be much more useful on less elite players than more elite players.

Exactly. 

Zach: Yeah. Yeah, exactly. [00:40:00] Um, yeah, that’s, uh, Do you have anything else you want to fit in here while we’re on the call? Do you think we’ve, we’ve, I know we’ve covered the things I want to cover. I don’t know if you want to say anything else about the iDirection things that, that we haven’t talked about anything else.

Tim: Um, no, I was just, uh. It was just a blast. Oh, it was fun talking to you again. It’s been a while. How long ago was it? 

Zach: Oh, that was like a couple years. Yeah. Yeah, it was a while back. But yeah, I want to yeah Anybody listening go check out the first interview because that that was I don’t know if I ever told you that was one of the

It’s just a term that people search deception and things like that, but yeah, I think people really like that. 

Tim: Well, cool. Yeah. Well, it’s nice to be on with you again. Thank you for inviting me. 

Zach: Thanks Tim. I’m 

Tim: reaching out. 

Zach: All right. Bye bye. See 

Tim: ya. All 

Zach: right. I’m going to.