Hidden Brain - Episode 52: Losing Face

Episode Date: November 22, 2016

It happens to all of us: someone recognizes you on the street, calls you by name, and says hello... and you have no idea who that person is. Researchers say this struggle to read other faces is common.... This week on Hidden Brain, super-recognizers, and the rest of us.

Transcript
Discussion (0)
Starting point is 00:00:00 If you're like me, you know this feeling. Maybe you're at a party or you're walking down the street. And suddenly, out of a sea of passing faces, one of them lights up looking right at you. This person starts waving, says hello. This person is glad to see you. And you, you have no idea who you're looking at. This is Hidden Brain, I'm Shankar Vitaanthan. Today on the show, we're talking about faces. Recognizing faces is a crucial skill. It's so crucial that there are regions of the brain
Starting point is 00:00:37 devoted to facial recognition. But although your mind is amazing at identifying your boyfriend or your child in a crowd, there are important limits to this ability. Some of us, like me, are extremely bad at it. Some of us are terrific. On today's show, we look at some people who are on opposite ends of the spectrum and talk about how our ability to recognize faces has broad implications in our lives. We'll start with someone whose job requires her to be quick with faces. She's a cop. My name is Alison Young and I'm a place of so-and-a-much for a place in London.
Starting point is 00:01:24 She started out several years ago working on response teams in East London. These are the cops who mostly just respond to 911 calls. Then, about three and a half years into that job, she and a bunch of her fellow officers were invited to take a series of tests at a university. You get given like three or four different faces and you have to memorize those faces. Then a new screen appears with other faces. These ones are obscured in some way, or heavily pixelated. One of the faces you saw earlier might now show up wearing a beard. So you have to try and work out which one of the faces is the face that you've seen before,
Starting point is 00:02:02 and it's that kind of thing. The test, of course, was measuring how good officers were at recognizing faces. A while later, Alison received her results. She came in second out of all the officers. And they asked me to come down to Scotland, you're odd. Scotland, you're out, of course, is the headquarters for London police. When she got there, she was told she was being added
Starting point is 00:02:24 to a new unit they were forming. It was called the Super Recognizer Unit. Yes, that's what that's the name that they've given. Do you feel like a Super Recognizer? I don't know, I don't, well I guess so, but I don't think I'd necessarily say that a lot because it's just worth super isn't it? It just sounds a bit super. I don't know, just the notion of the word super, it kind of brings out as if where some sort of super hero or something like that, That isn't the case. Here's how the Super Recognizer Unit works. This pool of cops with above average abilities to recognize faces sits in front of computers.
Starting point is 00:03:15 They have faces of criminal suspects shown to them and they try, in essence, to play a matching game. Well, they're the catalog of criminals, essentially, that are wanted by colleagues. And what they decided to develop was a thing called snapping which meant that we may not know who that person is but if I look at this face number one photo on this chart and then I continue to go through further and further and further through different photos can I find him in any other photos that he's wanted for, which then means that we've got him for one offense of, I don't know, theft. We find him for another offense to do his theft.
Starting point is 00:03:54 And you end up accruing this one person for around 25 to 30 different times. Now, computers are getting better at this kind of task. And maybe one day algorithms will come to replace human beings in making these kinds of matches for police. But for now, Scotland Yard decided to put its faith in humans like Alison. That's because there are some things that humans can do that computers can't. Last year, for example, the transit police came to Scotland to out for help. A 21-year-old woman and two girls aged 15 and 16 complained that a man had inappropriately
Starting point is 00:04:32 touched them while riding the bus. Transit police pulled security footage taken from the various buses. From the pictures and the witness accounts, it appeared his modus operandi was to get on the bus with the newspaper. He would sit next to the young woman and then attempt to fondle her under cover of the newspaper. And they were overtly young in their respect that some of the Murn school uniformed. The security footage was grainy.
Starting point is 00:04:58 The transit police didn't have an ID on the man. And because he struck at different times on different buses, they didn't know how to track him down. And they basically had said to us, we need to find this man. It's young girls, it's predatory, etc. So myself and my colleague, Texas Sergeant Elliott Porreck, like did some investigation. They studied the security videos, and eventually they figured out which station the man tended to frequent. We discovered that he had quite a specific route of generally being around Camden Town,
Starting point is 00:05:33 which is an area in northwest London. Camden Town is a busy neighbourhood. It's heavily populated with lots of shops and tourists and people always milling about. It's perfect in other words for someone to blend into the background. Alison and Elliot Poret knew what they had to do. So we made our way to Camden Town. From Scotland Yard, on a Wednesday, a current under the exact date, but I know it was a Wednesday.
Starting point is 00:06:01 This was supposed to be just a scouting mission to get a sense of the Camdentown bus station. Allison and her partner decided to look through all security footage in the CCTV rule. So we went to the CCTV bit, which was just behind a clear, per-spec glass, where people buy their tickets. So it's right by the entrance for a tube station. Detective Porid began talking with the transit security. Allison was looking through the glass at the commuters milling about the station. Detective Porid began talking with the transit security. Allison was looking through the glass at the commuters milling about the station. I just glanced up and through the crowd.
Starting point is 00:06:33 I just saw him. I saw him walk in, pick up a newspaper and leave. I will go to leave and at which point I, I mean I screamed because I don't know why I did it. I just screamed. I just thought I don't know, I can't quite work out why. But I just made quite loud noise and just said to Sergeant Porrette, he's outside. They both stopped what they were doing and rushed to catch up with the man. But it was quite difficult to get out because we had to go all the way back round, back round to the FOIA. So by the time we'd got into the main FOIA where he was,
Starting point is 00:07:12 we couldn't see him anymore. They ran out of the station to see if he'd left. They looked left, couldn't see him. They looked right. Alison caught a quick glimpse of a man disappearing around a corner. This fraction of a second was all a man disappearing around the corner. This fraction of a second was all she needed to recognize her target. The cops started running toward the corner. And as we turned, we looked just behind the wall, it was there. The officers approached the man.
Starting point is 00:07:39 As soon as we got up to him, I didn't fight to fight with him. The paragraph about myself, Detective Sergeant Porick, we were a hundred million percent certain that this was the exact same gentleman in all the photos. So he was fighting handcuffs immediately and explained to him what it was being arrested for, etc. And it was extremely noticeable that he was very nervous, his mouth just went completely dry. And he just wasn't able to speak. The man was recently sentenced to prison and is suspended from using London's public transit
Starting point is 00:08:14 after he's released. Alison Young no longer works on the super recognizer unit. She has gone on to other detective work, but she says the time she spent on this unusual unit was the first she ever realised she had an above average ability. The thing is, she is still not sure where the skill comes from. I think my mum was just like, oh you have that from me, modest as ever, my mother. My mum is very, very, very good with faces, very good. We're walking, just doing some shopping, and she'll see someone and go and speak to them.
Starting point is 00:08:56 And she'll have known them from primary school, and she'll remember them. And my mum's watching our 61, and she went to primary School like seven or eight with them and she'll remember them. So do you think this is genetic or do you think this is learned? I don't know, I don't think you could teach someone. I don't think you could teach someone to be able to I call. Coming up, how come in a zealousant skill? Are you a super recognizer? Can you learn to be one? That and more from a scientist who studies how we identify faces. This is Hidden Brain, I'm Shankar Vedantam. So the other week I was at the airport and just like everyone else, I showed my driver's
Starting point is 00:09:54 license to get past security. And it occurred to me that I was operating on an assumption that I think is widely shared. I assumed the TSA officer was pretty good at matching my face with the photo on my driver's license. So I asked Mike Burton, he's a professor of psychology at the University of York in the United Kingdom, if that assumption was true. It's not, although it is a very common assumption, most of us think we're pretty good at recognizing faces,
Starting point is 00:10:22 but when you actually test people out, particularly in the situation where someone who doesn't know you is checking a photo against you, it turns out people are really bad at this. Interestingly, even professionals are really bad at it, so we did some work with passport officers last year, where we showed that even passport officers find this a very difficult task and are often inaccurate. When I showed my ID at the airport last week, Mike, I had it over the ID and I noticed that the officer looked at the ID first and then looked at my face second and I assumed there must be some trick to this that it actually is you're able to make a better connection
Starting point is 00:11:01 if you don't actually look at the person's face that you start with the ID and then look at the face. Is there any truth to that? No, there's all kinds of techniques that the people who do this professionally use. And what's interesting is they all believe themselves to be performing quite well. But when you test them, just like anybody else, they're actually not very accurate at this.
Starting point is 00:11:23 It's a little terrifying what you're telling me because you're saying that this thing that we're relying on to keep ourselves safe to run security systems at airports and other places, that this is a fundamentally bad system? That is exactly what I'm telling you. I think that what's sort of interesting is that we have come to rely on this, but I think we've come to rely on it for an interesting reason. We are fantastic at recognising faces, those faces of people we know, we can recognise our family and friends across a huge range of conditions, you know, at distances in bad light all kinds. But we falsely assume that this means we're quite good at faces in general, and in fact we're not.
Starting point is 00:12:10 Mike can say all this with some degree of confidence because he ran a study to test for it. So we set up this little experiment where we asked people to match pairs of faces. They just have to say, are these two faces the same person or not? Mike and his colleagues ran this experiment in both the United Kingdom and in Australia. In both countries, they selected some faces that were likely to be well-known locally, but unlikely to be known globally. Yes, we use what we call be lists celebrities, so we check it out before I had, but we use people who are known very well by the local population.
Starting point is 00:12:48 These are people like newsreaders, local sports people, and they tend to be very well known by the local community, but not by people internationally. What we find then is that when the UK people are matching UK celebrities, they're really good at it. And they're really poor at matching the Australian celebrities, the people that they don't know. When you look at the Australian students looking at these photos, you find exactly the opposite pattern. They're great at matching Australian celebrities and poor at matching UK celebrities. So at this point, we know that there's nothing in the faces themselves that
Starting point is 00:13:25 make them easy or hard to match. It's just in the perceptions of the viewers. So far so good we know that people are better at matching familiar faces. But what we then ask is how well do you think other people will do on these faces when we give them the same task? And what you find is that the UK viewers think other people will find the UK faces easier, the Australian viewers think other people will find the Australian photos easier. That can't both be true. It must be that they are falsely generalising their own knowledge to other people. You make a very interesting point in the paper, and I was struck by it, which is that in
Starting point is 00:14:03 some ways this might be part of a general phenomenon in cognition where we do not fully understand how difficult a task is for someone else to do, and especially when we are good at something, it's very, very difficult for us to anticipate how much harder it could be for somebody else to do the very same task. That's right. It comes up in a number of areas of psychology, even something simple like general knowledge. If you happen to have read some books about Napoleon and be knowledgeable about Napoleon, you falsely generalise that and assume that other people know more about
Starting point is 00:14:38 Napoleon than they actually do. Of course, we all have our own different areas of specialist knowledge. And people turn out to be rather poor at understanding that and being able to generalise it. I'm wondering how good you are at facial recognition. I'm poor. On the tests I'm just a little below average. And doing these tests and studying this has it sort of changed the way you yourself trust yourself or your ability to recognize faces? Well, I do now know how poor I am. I certainly would try not to rely on my own ability to do it. But I think that nobody really knew until the last few years just how bad we all are with unfamiliar faces.
Starting point is 00:15:27 And it's just becoming clear now. Our chronic inability to recognize faces, coupled with our chronic overconfidence in our ability to recognize faces, has big consequences. One place that's especially true is in the criminal justice system, where eyewitness identifications are often central to police investigations. But this issue also shows up in lots of other settings with lower stakes, recognizing colleagues at an office party or a fellow parent at a school meeting. Julie Dorschlag from Washington, DC has had this problem for a long time. I've had a lot of uncomfortable situations for getting people and have been accused of being a snob or racist or I mean everything.
Starting point is 00:16:10 If Allison Young, the copy heard from earlier, was a super recognizer, you might call Julie Faceblind. Her whole life, she's been terrible with faces. When I was in college, you're on a campus, you meet a lot of people all the time. And there were people I met for a few minutes at a party or something, a meeting, I don't know. And I'd be on campus, a small campus, walking, and I'd walk past them.
Starting point is 00:16:33 Didn't think anything of it. Just smile and keep going. And they were offended. I've heard over time, people were offended. I got this little reputation of being this snobby person because I didn't greet everybody that I met or how are you doing. So, and normally would if I knew them or I thought I knew them. Julie struggles followed her as she left college and entered the working world. I worked for an architecture firm in Philadelphia and we had to go to a meeting and I swore
Starting point is 00:16:56 after going into the meeting and then coming back going back into the meeting room, I talked to this guy as if I knew him, he happened to be African-American and I was talking to him. He's like, I'm not that man. I said, I'm sorry. I could have sworn. He's like, oh, okay. So it became that. And it's not, it's not people of just color or different ethnicity. What do you do about it? I apologize profusely and usually the people just walk away from me. So I just stop. I just feel embarrassed. I'm just used to being embarrassed.
Starting point is 00:17:27 Research suggests, by the way, that people are worse at recognizing faces of people from unfamiliar groups. Many Americans are worse at recognizing the face of someone from a different race than a face of someone from their own race. Now, what makes Julie's dilemma especially acute is her husband, Marty. When I look at somebody, I never forget their face. If I spend about 30 seconds looking at somebody, I will remember their face for years and years
Starting point is 00:17:56 and years. Marty Doshlag, Julie's husband, is a super recognizer. For years, Julie's been keenly aware of her husband's superpower. One time they were in Las Vegas sitting down for dinner at a restaurant, Marty clandestapped at the waiter. He's like, oh, you waited on me in Columbus, Ohio, and X year, the guy just froze. And then he's like, oh yeah, and I don't know how you put it together, you named the restaurant, the time, the place.
Starting point is 00:18:23 And it was probably 15 years before. And he said, yeah, you're right. And so he does it a lot with servers and I think people in restaurant industry travel and he remembers them because they're server, you see their face. Marty said lots of encounters like this. Like at the Dallas airport, he spotted a man he sat behind at a University of Michigan football game three years earlier. Now, you might think that with this gift, Marty could at least be Julie's crutch,
Starting point is 00:18:50 but it doesn't always work out like that. If we're in a place, I'll always, sometimes, and I whisper in the back of the rear, that's Jim, who works at it, so and so. Well, sometimes he doesn't catch me in time. I think we were at one of your friends' apartments. A guy came in and I went up and hugged him and said, oh, it's so good to see you again. And Marty's friend, Lindoverin, said,
Starting point is 00:19:07 why is Julie hugging the caterer? So yeah. Oh, yeah. Yeah. I remember that too. And I thought that was funny. But I saw it coming, too. I saw Julie approaching the guy.
Starting point is 00:19:20 And I said to my friend, I said, watch, she's going to think that's your roommate. Because they look sort of a, they look, I mean, they had the same color hair, I think, or something. And they were the same size. And I said to him, here it comes, watch. And sure enough, she did it. It was,
Starting point is 00:19:33 but I touched him, I hugged him. So that big, that goes into another. You have to be careful. Oh. Julie's cringe-worthy ordeals hit close to home. Recently, I was watching a play. The lead actor looked familiar. I stared at his face for the better part of 90 minutes, but it took me until after the play was over to realise this was a colleague of mine from NPR.
Starting point is 00:20:02 I'd be absolutely terrible as a TSA agent. And I got me wondering, are there any solutions here? Julie's picked a simple one. I don't approach people with as much jua to viva. I don't touch them until I'm sure they want to be touched. And I know them. And also, I sort of create this verbal cue for them to tell me why I know them. So if I shake their hand, I'll say, oh, right.
Starting point is 00:20:28 Do I know you from somewhere? And they're like, if they say, I don't think so. I said, OK, you just looked a little familiar. I'd rather err on that side than not knowing them. And if they finish the sentence, I said, yes, that's right. Good to see you again. But I don't use again until I know that they've filled in the blank. For many thousands of years, humans lived in small, nomadic groups.
Starting point is 00:20:52 Everyone probably knew everyone else. People didn't have the information overload we have today. You and I can come into contact with hundreds, thousands of people in a single day. Our minds simply haven't evolved at the same speed as our societies. There are going to be outliers among us, people with extraordinary skill at recognizing faces. Some of them may end up as security officers, or gregarious socialites, or politicians. The rest of us are going to keep smiling awkwardly at office parties, at people we're supposed to know.
Starting point is 00:21:26 It's what happens when you stumble around in the 21st century with a mind that was designed in the Stone Age. This week's show was produced by Chris Banderer and edited by Jenny Schmidt. Our supervising producer is Tara Boyle, our team includes Raina Cohen, Path Shah, Laura Quarelle, Thomas Liu and Angus Chen. Our unsung hero this week is Alexander Diaz, who troubleshoots technical problems whenever they arise and has the most unflappable, kind disposition in the face of whatever crisis we throw his way. Producers at NPR have taken to calling him Batman because he is constantly, silently, secretly, saving the day.
Starting point is 00:22:13 Thanks Batman. If you liked today's episode, please take a second to share it with a friend. We're always looking for new people to discover our show. I'm Shankar Vedantum, and this is NPR.

There aren't comments yet for this episode. Click on any sentence in the transcript to leave a comment.