transcript

Speaker 1:
[00:00] This is Hidden Brain, I'm Shankar Vedantam. We've all been there. You're at a party and someone says hello. They look vaguely familiar. You've seen them somewhere. But where? A coffee shop? A conference? Maybe just a passing glance on the subway? Or maybe you ask a couple of friends to pick up a bottle of hot sauce on their way over to your home. You tell them there's a cartoon fish on the label. They go to the store, find the bottle, but the label doesn't show a fish. They call you to tell you what they picked up, and you just know they've picked up the wrong sauce. But then you look at the bottle, no fish, no cartoon, just a plain red label. At the University of Chicago, psychologist Wilma Bainbridge studies these strange quirks of the human mind, how we remember faces, forget details, and sometimes invent memories that were never there. Wilma Bainbridge, welcome to Hidden Brain.

Speaker 2:
[01:04] Thank you so much for having me.

Speaker 1:
[01:06] Wilma, some time ago, you were scrolling on the online social platform Reddit, and you came by an interesting post. Tell me what was going on in your life at the time and what this post was about.

Speaker 2:
[01:17] Yeah, so I had just moved to Chicago to start my position as a new assistant professor at the University of Chicago. So to be a professor for the first time and start my lab. And sort of before going to bed in my nerves in my new apartment, I was scrolling through Reddit, and I came across this really interesting post that links to a blog post where it showed a series of pictures, and you had to guess which was the correct version of the picture. So for example, it would show the Monopoly Man, and there would be a version with him with a monocle on, or a version without a monocle on. Or it would show the Berenstein Bears, this really popular kids book series, with Berenstein spelled in a few different ways. And there were probably about 20 of these different examples, and it was a quiz where you had to click on which one was the right one based on your memory of the character.

Speaker 1:
[02:11] For our listeners who haven't played the Monopoly board game, who was the Monopoly man, Wilma?

Speaker 2:
[02:18] Yes, the Monopoly man, he's actually the central character used to market the board game Monopoly. He's on the box, and he's on all of the artwork for the board game. And I think his name is Mr. Moneybags, actually. He's like this rich older man.

Speaker 1:
[02:35] And you had two versions of the Monopoly man that you had to pick, and one of them was the correct version. To the extent that you remember this blog post, what was different about these two images?

Speaker 2:
[02:45] Yeah, so both of them showed a rich older man, but one had a monocle and one didn't have a monocle.

Speaker 1:
[02:51] And what was your guess about which was the real Monopoly man?

Speaker 2:
[02:54] Oh, it was definitely the guy with the monocle.

Speaker 1:
[03:00] Another two of these images were the character C-3PO from Star Wars. Who was this character, and what were the two images like, Wilma?

Speaker 2:
[03:10] Yeah, so C-3PO, he's a sort of humanoid-esque robot from the Star Wars movie franchise, and he was in the first movie. So he's one of the most famous, popular characters of Star Wars. And so for these two pictures, there is C-3PO, completely gold, and then there was this other version of him with a silver leg. And for me, it was clear that he was a completely gold one.

Speaker 1:
[03:39] The last set of images had to do with the character Waldo. Again, for people who are unfamiliar with the context, who is Waldo?

Speaker 2:
[03:46] Yeah, so Waldo is a character from a popular kids book series called Where's Waldo? Or Where's Wally? In some other countries. Basically, for this book series, you see a really complicated, noisy scene, and you have to find Waldo within this scene. And Waldo is very distinctive looking. He's this tall man with a red and white striped shirt and a red and white beanie. And so in this blog post, the two versions were just him as his usual self or him holding a cane. And at least for me, I was confident he's usually not holding anything.

Speaker 1:
[04:22] All right. So you have these three sets of images. You remember the Monopoly man is having a monocle. You remember C3PO as being an all gold character. And you remember Waldo as not having a cane. How accurate were your memories, Wilma?

Speaker 2:
[04:38] I was completely wrong on everything. The Monopoly man does not have a monocle. He has no eyewear. C3PO actually has a silver leg, and Waldo is always holding a cane.

Speaker 1:
[04:58] Now, the fact that you made these mistakes is interesting, but what's even more interesting is that you were not the only one who was making the mistake. Part of the reason this blog post probably existed is because many people were making the same mistake.

Speaker 2:
[05:13] Exactly. That was what was really shocking, and why this Reddit post went viral. It was because everyone made the same mistakes. Everyone thought that the Monopoly man had a monocle. And so we were surprised, how can we all have this same false memory? Like, you know, what is reality?

Speaker 1:
[05:31] In some ways, this made you think about a related phenomenon that was called the Mandela Effect. Tell me how the term Mandela Effect came to be in the first place and the connection that you made with what happened in this Reddit post.

Speaker 2:
[05:46] Yeah, so the Mandela Effect is a term coined by Fiona Broom. She is a paranormal researcher, and she coined this term when she realized she had the same false memory as many other people about Nelson Mandela. So she had a false memory that he had died in prison in the 1980s, when really he was actually still alive when she coined this term. And after his time in prison, he had served as the president of South Africa. But it turns out that not just her, but many other people had the same false memory. And so from this discovery, the Mandela Effect came to be known as this term, describing when many people have a similar false memory. And that's what was happening in this Reddit post too. We were all having the same false memories for these really popular icons.

Speaker 1:
[06:38] So the interesting thing here, Wilma, again, is not that people are making a mistake. It's fine for someone, one person to remember that Nelson Mandela had died. They're making a mistake. It's just one person. What's interesting here is that many people were making the identical mistake. You were really struck by this phenomenon, and you decided to test it in the lab. You started with 40 different cultural icons, altered the images, and then showed them to people. What were you looking for?

Speaker 2:
[07:05] Yeah. So previously, the Mandela effect had only been studied as this paranormal concept. There's actually a big group of people online who think the Mandela effect is evidence that we have leaped into a parallel dimension. That actually, the Monopoly man did have a monocle when we were kids, but then we leaped dimensions, and now we're in this strange but similar world where he doesn't have a monocle. Yeah. And so as a psychologist, I thought this is a really fascinating phenomenon about our memories. And so first, I wanted to test, is this Mandela effect really something so pervasive, or is it just something like a little small community on Reddit feels similarly about? And then I also just wanted to understand what does this mean about memory, and can we find evidence maybe for this or against this idea about leaping across dimensions?

Speaker 1:
[08:01] So you basically showed a number of people a series of altered images. Can you describe the experiment and what you found?

Speaker 2:
[08:08] Yeah. So first, we ran an experiment almost like that blog post I had seen. We ran an experiment where we tested hundreds of people online and showed them examples, multiple examples of 40 different cultural icons. And we included many that we knew had this Mandela effect, so like the Monopoly Man and C-3PO. And we asked people to guess which is the correct one based on what you remember of this character. And we also asked them how confident are you about your choice and how well do you know this franchise or this character?

Speaker 1:
[08:41] Two of the images that we haven't discussed so far were Curious George and the Fruit of the Loom logo. Describe again what these images were and how you altered them.

Speaker 2:
[08:52] For the Fruit of the Loom logo, one version that we presented was a bunch of fruit surrounded by a cornucopia, which is what many people think of when they hear Fruit of the Loom. Another version we presented is a bunch of fruit on a plate, which is not what a lot of people think of. And then the third version we presented is a bunch of fruit with nothing containing it, which is the actual Fruit of the Loom logo. But most people have this false memory that there is a cornucopia holding the fruit. Then for Curious George, the three versions we presented were Curious George with different types of tails. So either he had a long tail, a sort of bushy tail or a no tail. Most people think that he has a tail, but he in fact does not.

Speaker 1:
[09:37] So again, here you're seeing the same pattern that you saw with the Monopoly man or with C3PO, which is it's not just that people are making a mistake, they're making the same mistake systematically.

Speaker 2:
[09:49] Exactly. And what's more, what we found in our experiment is that they tend to be pretty confident about the wrong answer, even if they're very familiar with the character. So even if they're like, I wear Fruit of the Loom shirts every day, they still get it wrong.

Speaker 1:
[10:04] So you and your co-author Deepa Shree Prasad call this a visual Mandela effect. And I think I'm seeing the connection here because again, people are remembering these images in a false way.

Speaker 2:
[10:16] Exactly. And there are examples of the Mandela effect that are not visual, like they're more verbal. So for example, Berenstein Bears is spelled differently from what people think. The ending of Berenstein is not S-T-E-I-N, but it's S-T-A-I-N. But we specifically focused on visual examples of the Mandela effect or ways in which it's manifested in images because you can do all these interesting analyses on images, look at what people look at in the image.

Speaker 1:
[10:46] So, I understand that's exactly what you did in one of your experiments. You tracked people's eye movements as they were looking at these images and then tested them again. What were you hoping to find, Wilma?

Speaker 2:
[10:56] So, we wanted to test many different hypotheses for how this Mandela effect might happen. And one hypothesis is that people are just not looking at that part of the icon, and so it makes sense they wouldn't have a memory for it if they don't ever look at it. So, for example, for C-3PO, when you're watching the Star Wars movies, you're probably looking at his face, you're probably not looking at his legs, or maybe his legs are not even on screen a lot of the time. And so it then makes natural sense that you would fill in that his whole body should be the same color. So, to test this question, we ran an experiment where we showed people the correct versions of the images and then saw where they looked at when they explored the image. And then after that, we tested their memory to see if their looking patterns predicted whether they'd get things right or wrong.

Speaker 1:
[11:46] And what did you find?

Speaker 2:
[11:47] Yeah. So what we found was that even if they looked right at the correct feature, they would still get it wrong. So even if we saw that they looked at C3PO's silver leg, when we tested them, they would pick the wrong C3PO who is entirely gold. Or they would see that Curious George doesn't have a tail, but they'd still pick the wrong Curious George. So it's not just that people weren't looking at the full image and making that mistake.

Speaker 1:
[12:22] So given that the problem didn't seem to be that people were looking in the wrong place, you next searched the internet for images of each icon to see how they were presented to the public. Maybe the reason people misremember the Monopoly Man, for example, is that they had seen a version of him with a monocle from some outside source. Now, that sounds like a plausible explanation. Did it turn out to be correct?

Speaker 2:
[12:47] Yeah, so exactly, we ran this other experiment, wondering if maybe people's memories were just for a different version of the image where it did have that feature on it. So, as you pointed out, maybe actually there's fan art of the Monopoly Man wearing a monocle and mine, and maybe people have seen the fan art more than they've seen the actual game, and so it's not that they have a false memory. They actually have a true memory, but for an incorrect image, like an image that a fan made. But what we found was that actually still, for all of our icons, a majority of the images online were of the correct version, the version that people don't remember. Now, as the Mandela effect is becoming more popular on social media, the correct versions are starting to get more replaced by the Mandela versions if you look online.

Speaker 1:
[13:38] All right, so you've potentially ruled out yet another explanation for the errors, which is it's not that people were systematically looking at made up images that had the wrong image. Another potential explanation for the findings is something called Schema Theory. Can you explain this to me, Wilma?

Speaker 2:
[13:59] Yeah, so there's this predominant theory in memory research called Schema Theory. The idea is that we have schemas or templates for categories. So for example, your schema or template for a kitchen might be that it should contain a refrigerator, a stove, and a microwave. And so some of these Mandela effects might happen because we're having false memories that make an image align better with our schemas. To give an example, perhaps a Monopoly man, he's a typical rich person. So we then fill in details from our schema of what a rich man should look like. So that might include having a monocle, along with having the top hat and the handlebar mustache.

Speaker 1:
[14:43] And the same goes with Nelson Mandela. We know that he spent many, many years in a brutal prison in South Africa. Perhaps our schema would suggest that, well, the chances are he probably died in that prison.

Speaker 2:
[14:55] Yeah, exactly. Maybe our schema is that then that person must have a more tragic life rather than having made it out successfully in the end.

Speaker 1:
[15:10] Was there evidence to support the schema theory as an explanation for this visual Mandela effect that you were finding, Wilma?

Speaker 2:
[15:18] So we think it's possible that schema theory could account for some of the examples, like the Monopoly Man example, but we think it's unlikely to account for every possible example. So for example, with the Fruit of the Loom logo, we see bunches of fruit all the time in our lives. And when we see it, it's usually maybe on a plate or in a bowl. But when have you ever seen apples, bananas, and grapes in a cornucopia? The only time I know I see cornucopias is in the fall with squashes and pumpkins. So I don't know why people's schema for fruit would include a cornucopia.

Speaker 1:
[16:01] So, an alternative explanation that you've proposed is something that you call memorability. Can you explain what memorability means and how it differs from schema theory?

Speaker 2:
[16:11] Yeah, so one thing we're finding in our work is that even though we are all unique individuals, surprisingly, there's a lot of similarity across people in what we remember and forget. And so there are just some images that are intrinsically memorable, there are some that are really forgettable, and then in this case, there might be some images that just cause a lot of false memories. And so that's a question we're really interested in looking at right now in our lab.

Speaker 1:
[16:37] I understand, Wilma, that you tested this idea. You conducted a study on people's faces and what was memorable about them?

Speaker 2:
[16:44] Yeah. And so this was actually one of my first studies in graduate school that has now formed the basis for a lot of the research in my lab now. Basically, what we did is first, I created probably one of the largest and most diverse face databases at the time. Previously, a lot of psychology research used these very homogenous face datasets, but here I created one that really captured the rich demographics of the United States that had over 10,000 images. And then I put these 10,000 images into an online memory experiment with close to a thousand participants. And I wanted to see what are the faces people remember, what are the faces people forget, and what are the ones that people have false memories for.

Speaker 1:
[17:28] So when you showed these volunteers these thousands of images, were there in fact some images that turned out to be more memorable than others?

Speaker 2:
[17:37] Yeah, that's exactly what we found. So this is pretty surprising because of course our previous experiences should impact what we remember, right? Like I will remember someone who looks like my mom, maybe more than you might.

Speaker 1:
[17:50] Right.

Speaker 2:
[17:50] But what we found was that surprisingly a majority of what impacted people's memory was something about the image itself. So there were some faces that everyone just remembered, and there were some faces that everyone just forgot.

Speaker 1:
[18:09] You found something interesting about faces that were rated as more irresponsible or faces that were kinder or faces that were unhappy. Tell me about some of these findings, Wilma.

Speaker 2:
[18:19] Yeah, so we got a collection of different attributes for the faces, and we wanted to see are there any attributes that make a face memorable? And we did find that there were some attributes correlated with a memorable face. So unfortunately, some more negative things like seeming more irresponsible or less kind was related to having a more memorable face. But we also found that this was not like an overwhelming effect. You can still have an incredibly responsible and kind face. Like really, the takeaway was that we found there was no one attribute that made a face memorable. And even combining all the attributes we measured, we couldn't really predict memorability just from specific attributes. So it's not a matter of, you know, if you frown, people will remember you. One other thing people often ask me is whether attractiveness is related to memorability. We find a generally pretty weak relationship. In fact, being attractive is more related to false memories, this feeling of haven't I seen you before. But being memorable, you don't have to necessarily be attractive or unattractive.

Speaker 1:
[19:32] So this is a really striking finding here, Wilma, because you're saying that it's not just about the attributes of the face. It's not just whether a face is good looking or not, a face is angry or not, a face is sad or not. These attributes, in fact, explain only about a quarter of the variance in memorability that you were finding.

Speaker 2:
[19:51] So exactly, this was pretty surprising to us because intuitively, it feels like most of what determines our memory should be about our own previous experiences. Or if not, it should be a really obvious attribute like, oh, okay, that face is really attractive, I'll remember it. But what we found is that memory was not fully determined by either of those things. There's really something about that image or that face that's separate from specific properties of the face and separate from your own previous experiences that determines what you'll remember. And what's also interesting is that if people tend to remember and forget the same images, that means I can predict your memories based on the images that I'm showing you.

Speaker 1:
[20:39] When we come back, Wilma explores how these curious findings shape our artistic preferences and sensibilities. You're listening to Hidden Brain. I'm Shankar Vedantam. Hi there, Shankar here. We know you have a million things competing for your attention. Work, family, and a never-ending email inbox. The fact that you not only choose to listen to Hidden Brain each week, but also support us financially, we don't take it lightly. Each episode we create is a labor of love. It takes lots of research, thinking, editing, and yes, plenty of caffeinated drinks. But we do it because we believe in the power of science and storytelling, and because we know you do too. Your support tells us that curiosity still matters, that thoughtful conversations about rigorous science still have a place in our world. What we're building together is something worthwhile. Thank you for listening, for making this work possible. Thank you for your support. We are deeply grateful. This is Hidden Brain, I'm Shankar Vedantam. What makes something memorable? At the University of Chicago, psychologist Wilma Bainbridge studies this question. She has found that there's a surprising consistency in what people remember that cannot be explained by outside forces. She calls it memorability, a distinctive, inherent quality that makes something stick in our minds. This applies to images we encounter, faces we meet, but also our artistic preferences. Wilma, you wanted to look at the concept of memorability and how it applies to the art world. You tracked people's impressions of over 4,000 paintings from the Art Institute of Chicago. Some of these paintings you had people look at on a computer, but you also followed some people through the museum itself, encouraging them to explore each floor freely?

Speaker 2:
[22:51] Yeah, so the Art Institute of Chicago is a fabulous, huge art museum here in Chicago. I encourage everyone to visit it. It's right downtown. But the Art Institute gets over 1.5 million visitors a year, and it's humongous. But we were curious, what would people remember if they visited this institute?

Speaker 1:
[23:15] I'm wondering whether there were certain paintings that ended up being more memorable for your volunteers.

Speaker 2:
[23:24] Yeah, so we had a set of participants explore this American art wing at the Art Institute of Chicago, and we basically told them, just explore it freely. You can go with your friends. You can chat with them. You can go in any order that you want, looking at the pieces. Our only rule for you is that you see every piece at least once. And we didn't even always have an experimenter there because it was during the pandemic. Then after the participant explored the gallery, they would then sit outside on a bench and then take a memory test on their smartphone, where they would see each of those pieces from the gallery and answer whether they remember seeing it or not. But the tricky part is that these pieces were also intermixed with similar pieces of artwork from the same artists, same time period, same content types. And so the memory test wasn't that easy. Like half of them were actually not pieces they had seen.

Speaker 1:
[24:20] And what did you find?

Speaker 2:
[24:22] What we found was that actually there were still some pieces of art that everyone remembered and some pieces of art that everyone forgot. And this was surprising to us because you would think when you visit a museum, you're really motivated to remember everything. It's like an important valuable experience for you. But even so, there were some that people remembered and some that people forgot.

Speaker 1:
[24:49] You found that there were some key factors that influenced what paintings people remembered. For example, there was a correlation between the size of a painting and how memorable it was?

Speaker 2:
[25:01] Yeah, so actually, bigger is better. So the larger a piece is, the more memorable it was. We also found this interesting interaction where if you happen to have a small piece of art, it did get better remembered if it was surrounded by large pieces of art.

Speaker 1:
[25:18] You also found something curious about where the paintings were located within the art museum itself.

Speaker 2:
[25:25] Yeah, so the American art wing was actually split into two different floors, and they had two very different styles. So the first floor was a bit more homogenous in style and time period, while the second floor was more diverse and had more modern pieces. And so what we found was that the more diverse and modern floor tended to be generally better remembered.

Speaker 1:
[25:48] So it might not have been the location of the floor. It might have actually been the quality of the paintings that were curated on each floor?

Speaker 2:
[25:55] Yeah, so we think it was actually the types of pieces that were curated on each floor. And this is not super surprising because there is a lot of work in the memory field showing that context matters. So if you don't have a lot of things competing for your memory, then you'll do a better job at remembering overall. And so we sort of took this result as an example of, okay, the floor with less competition, like fewer pieces, and they looked more dissimilar from each other, were the ones that were better remembered.

Speaker 1:
[26:25] I'm trying to think back to the last time I visited a museum. In fact, it might have actually been the Art Institute of Chicago. And I'm trying to think if the pictures that I remembered were the pictures that showed the most emotion, or the most color, or the most beauty, was there any correlation between beauty and emotion and color and people's ability to remember the paintings?

Speaker 2:
[26:49] Yeah, so we ran a separate experiment online, getting those attributes for each of the pieces in the Art Institute's collection. So we had separate people rate, how beautiful are these pieces? How emotional are they? How familiar are they? And we also got these computerized measures of color and clutter within the images. And basically, what we found was that a lot of these things didn't matter. So there was no relationship between beauty or emotion with what was memorable. So you could have a totally beautiful, emotionally evocative piece, and it could be entirely forgettable.

Speaker 1:
[27:27] I'm wondering what this says about abstract art.

Speaker 2:
[27:30] So in that study, we did not find a clear relationship with content type, like abstract, still life portrait and memorability. But in some recent work, we actually do find that pieces of art that are easier to understand, that have something that we can wrap our heads around, tend to be the ones that are best remembered.

Speaker 1:
[27:52] So you're finding that as people go through the museum, they're going through it at their own pace, looking at paintings in their own time. But some of these paintings end up being more memorable than others. But you don't really seem to have a clear explanation of why this is the case, Wilma.

Speaker 2:
[28:07] Yeah, so actually, our hypothesis for what makes something memorable is that it's something that's easy for the brain to process. So a piece of art that is easy for us to understand, that sort of fits in with templates in our mind, might be the one that we best remember. But it still has to have something interesting or something unique about it to set it apart.

Speaker 1:
[28:32] If the human mind makes these predictable choices, including predictable mistakes, it should be possible to build a model of this. I understand that you've constructed a machine learning tool. Describe this tool for me and how it works.

Speaker 2:
[28:46] Yeah, so we've actually created an artificial intelligence based model called ResMem, where you can upload a picture and it'll tell you how memorable that picture is. So like one means 100% of people remember it, while zero means no one will remember it.

Speaker 1:
[29:02] And in some ways, this is building on the fact that human minds look at pictures and think of certain pictures as memorable, and they do this in the systematic fashion. And so what you're doing is saying, let's take those findings from human beings and create a machine learning model of this. And now the machine now starts to predict which images are memorable.

Speaker 2:
[29:22] Yeah, that's exactly correct. Because, yeah, the idea is that if everyone's remembering the same things, that means a given image has this inherent memorability and that's a value you could train an AI system to learn.

Speaker 1:
[29:37] I'm wondering what this might tell us about which paintings become popular or which paintings become famous, Wilma.

Speaker 2:
[29:44] Yeah. So first one thing about resmem is that we train this model purely based on regular images, like general photographs. We didn't train it based on artwork. So we thought it might not do a good job at predicting people's memory for something as rich, subjective, and complex as artwork. But in our experiment, we tested whether resmem could predict what pieces of art people would remember when they explored the museum. And to our surprise, we found that actually resmem could predict what people remembered or forgot when they're exploring the museum.

Speaker 1:
[30:20] In other words, what you're finding is that if these patterns of memorability in some ways are predicting which paintings are remembered and therefore which paintings become famous, the machine learning tool now is essentially able to reverse engineer the process and predict which paintings in fact became famous.

Speaker 2:
[30:40] Yeah, so one thing with the Art Institute of Chicago is that it has all of these incredibly famous pieces like a Self-Portrait by Van Gogh or the American Gothic by Grant Wood. And so actually for our experiment, we did not include those pieces because we knew people would remember seeing the super famous Van Gogh Self-Portrait. But Resmem, our AI model, doesn't know anything about fame or culture or history. So we were curious, what would Resmem do if we showed it famous pieces of art and then not so famous pieces of art, both from the Art Institute's collection?

Speaker 1:
[31:15] And what did you find?

Speaker 2:
[31:16] What we were surprised to see was that Resmem predicted that famous pieces were more memorable than non-famous pieces. In other words, what this means is that pieces that end up becoming famous have something about their visual properties that just make those pieces more memorable. One of our takeaways is that possibly part of what makes something famous is that it sticks in your memory, and then maybe you share it with other people, or it sticks sort of in a cultural memory, and that's what helped it become famous.

Speaker 1:
[31:54] I mean, in some ways, this challenges the idea that art is very subjective, because when we look at a beautiful painting, we might experience it as a deeply personal experience, but it might be the case that, in fact, we're following these predictable cognitive patterns, perhaps shaped by ancient algorithms in the brain.

Speaker 2:
[32:13] Exactly. It's actually really fascinating and often jarring for many people. So in a lot of our research, we find that people cannot predict what's memorable or forgettable. So actually, in a new study, we find that ResMem, our AI model, can predict what pieces of art you'll remember better than you can predict what you'll remember.

Speaker 1:
[32:34] Beyond the world of art, Wilma, I think that this work has implications for clinicians and families who are dealing with Alzheimer's disease and other dementias. Can you explain this idea, this body of work?

Speaker 2:
[32:46] Yeah. Well, first of all, if you know in events what people will remember and forget, then this gives you the power to design worlds that are more memorable. So imagine you could use more memorable images to help someone remember to take medication or put more memorable labels on medication. We also have a separate set of works showing that most people tend to remember and forget the same things. But when people diverge might be when they're starting to develop an early cognitive decline like Alzheimer's disease. We published a study a few years ago where we found that there are some special images that are memorable to healthy people, but forgettable if you're in the early stages of Alzheimer's disease or dementia. And these images, you can think of as diagnostic images, because if you get them right, you're fine. But if you get them wrong, then you might want to go do a clinical memory test. And these diagnostic images can actually predict if you're on an early stage of cognitive decline better than if you test hundreds of randomly chosen images.

Speaker 1:
[33:58] So in other words, the big idea here is that for reasons that are not entirely clear to us, certain images are more memorable to healthy human brains. But you're saying that people with certain kinds of dementias might in fact not see those images as memorable. And so if you use those images almost as a diagnostic tool, it could become an early warning signal of potentially a brewing dementia.

Speaker 2:
[34:24] Exactly. And some of why we think this is happening is because when you start to experience early stages of dementia, specific parts of your brain start to lose some mass. And those parts of the brain are associated with certain types of image processing, like processing complex details within a scene image. And so we think the images that we've identified really tap in to those parts of the brain that are experiencing loss. And so if you do poorly at remembering those images, it's because those parts of your brain are starting to degrade, and that you might be along the trajectory for dementia.

Speaker 1:
[35:04] Wilma Bainbridge is a psychologist at the University of Chicago. She studies perception and memory. Wilma, thank you so much for joining me today on Hidden Brain.

Speaker 2:
[35:13] Yeah, thank you so much for having me, Shankar. It's been a great pleasure.

Speaker 1:
[35:20] Hidden Brain is produced by Hidden Brain Media. Our audio production team includes Annie Murphy-Paul, Kristen Wong, Laura Quarell, Ryan Katz, Autumn Barnes, Andrew Chadwick, and Nick Woodbury. Tara Boyle is our executive producer. I'm Hidden Brain's executive editor. Since you're a member of Hidden Brain+, it's clear that you're interested in the ideas and research we feature on Hidden Brain. If you'd like even more thought-provoking ideas about human behavior, there are even more ways to engage with our work. Check out our new YouTube channel, and don't forget to subscribe to our newsletter. You can find it at news.hiddenbrain.org. I'm Shankar Vedantam. See you soon.