Speaking of Psychology: Why can’t you remember being a baby? With Nicholas Turk-Browne, PhD

Why can’t you remember your first birthday party? Or the house you lived in at age 2? Nicholas Turk-Browne, PhD, talks about new studies that suggest that babies and toddlers may form early memories; why we aren’t able to retrieve them as adults; and how evolving brain imaging techniques could help answer these questions—and help us understand more about what it’s like to be a baby and how babies experience the world.

Nicholas Turk-Browne, PhD, is director of the Wu Tsai Institute and a professor of psychology at Yale University. He studies how humans perceive, pay attention to, learn from, and remember the world around them. His lab has pioneered brain imaging techniques for studying memory in infants.

Kim Mills: Can you remember your first birthday party? Or your first words? Or the house you lived in when you were 2? If you can’t, and I bet you can’t, why not? The mystery of what scientists call childhood or infantile amnesia has long puzzled researchers. The infant brain is an amazing learning machine. It takes us from helpless newborns to walking, talking preschoolers in just a few years. But for some reason, virtually no memories of those formative years stick with us into adulthood.

Today we’re going to talk to a scientist who studies infant learning and memory. He has looked into why we don’t remember our own earliest experiences. Are baby’s brains forming memories at all? And if so, why can’t we access them later in life? How does the infant brain make sense of the world? What insights can studying infant learning and memory give us into learning and memory more broadly? And how are scientists using new brain imaging tools and techniques to explore all of these questions?

Welcome to Speaking of Psychology, the flagship podcast of the American Psychological Association that examines the links between psychological science and everyday life. I’m Kim Mills.

My guest today is Dr. Nicholas Turk-Browne, a professor of psychology and director of the Wu Tsai Institute at Yale University. He studies how adults, children, and babies perceive the world and how we learn from experience and store information in memory. His lab has pioneered techniques for brain imaging in infants and toddlers. His work has been published in leading journals and featured in major news outlets including the New York Times and The Atlantic. He’s received young investigator awards from APA as well as from the Vision Sciences Society, the Cognitive Neuroscience Society, and the Society of Experimental Psychologists, and he recently received the Troland Research Award from the National Academy of Sciences. 

Dr. Turk-Browne, thank you for joining me today.

Nicholas Turk-Browne, PhD: Thank you for having me, Kim.

Mills: What is the earliest age that most people can remember? Now, some people claim to be able to remember things from their babyhood or very early toddlerhood. Is that even possible?

Turk-Browne: Well, to figure out what early memories people have, scientists sometimes will conduct surveys or poll adults about what they remember from early in life. And typically the earliest memories that people report are around 4 to 5 years old and up. Importantly, that’s regardless of how old you currently are. So if you’re 20 years old, that’s true, if you’re 70 years old, that’s true. There’s a period of the first few years of life that we are largely amnesic about. I never say never. People sometimes report early memories and they may be real, but the science suggests that memories from that period before 3 or 4 years old are unlikely to be pure memories that you’re recollecting from a specific moment in your life.

Mills: You talked about asking basically people who are adults, how far back they can remember. But what if you talked to a 7-year-old or a 5-year-old? Can they remember being 1 or 2?

Turk-Browne: Yeah, sometimes they can. I can say that from personal experience. I can also say we are currently helping curate an exhibit at the Peabody Museum in New Haven where there’s a station where people can write or draw their first memories, and it’s a wonderful exhibit and station just to see kids interacting with their parents, talking about their early life. And so I think it’s possible and scientifically understudied—for how long in childhood can we retain early experiences and memory? I don’t know for sure that we know the answer to that question, because often people don’t have ground truth about what was actually experienced early in life that they can go back to and verify whether a memory was accurate or not. And so that’s where the scientific research comes in, to try to establish on firmer ground that a memory is authentic and accurate.

Mills: Well, let’s talk about the question I asked in the introduction just a few moments ago. Do we make memories as babies that we just can’t access when we get older? Or do babies not form memories at all, at least not in the same way that adults do?

Turk-Browne: Yeah, this is a profound and fascinating question that I think we have all thought about in one way or another as parents and as people thinking back on our early lives. I mean this has fascinated scientists from more than a century. Freud popularized the idea of infantile amnesia and had his own ideas about it that are not accepted today. But basically there are two general kinds of explanations which you’re alluding to. One is that maybe the infant’s mind and brain are immature and unable to store durable memories. And so our lack of memories later in life is because there’s nothing available for us to pull up. The other possibility is that the infant brain is storing memories and that over time we lose access to those memories. And in some abstract sense, they may still be there in the brain if we could access them somehow.

And so these are the two sort of classic hypotheses about infantile amnesia. The first having to do with a failure of encoding or storing a memory. The second having to do with a failure of retrieval, of being able to access the memory. And this is a longstanding question, but really excitingly, there’s been amazing scientific progress in the last few years that have shed light on this question. The most compelling recent research actually has come from animal studies in rodents.

So to walk you through that briefly, imagine that you have a baby mouse and you give them some kind of memory. So for example, you put them in a maze and they have to learn how to escape from the maze. If you then let that mouse grow up, and in the case of a mouse that’s not 18 years, that’s 20 weeks or so, and you put the mouse back in that same maze, they won’t remember how to escape. That is they show something like infantile amnesia. They don’t remember something they learned early in their life.

So the recent research though used very fancy modern molecular methods to figure out how the initial infant mouse’s memory had been stored. And then after the mouse had grown up, to stimulate the neurons that were part of that memory. And they do this using a technique called optogenetics where you shine light on neurons and it causes them to become active. So if you knew what neurons had stored a baby memory and then you could activate those neurons in adulthood, what they found in these studies is that even though the mouse couldn’t remember how to escape from the maze normally, if you activated the memory through this kind of stimulation, then the mouse remembered and could escape. And so what this suggested was that actually the memory was there, it’s just being back in the environment wasn’t sufficient to access it. It needed to be directly stimulated.

Now, we can’t do that in humans, and I could tell you about the research in my lab that’s trying to get at a similar question, but that evidence is suggestive of more of a retrieval explanation for infant amnesia—that we may have latent forgotten memories that are still there in some fashion in our brain.

Mills: And is the reason we can’t do it in humans because we basically have to open up their heads and go into their brains and start stimulating?

Turk-Browne: Yeah, indeed. There’s a couple of reasons. The tricks to sort of tag which neurons were part of a memory requires genetic and other sort of molecular tools that are not safe or ethical in humans. And the other is to stimulate neurons in the brain deep down in the brain—and the brain region that’s really essential for this kind of memory called the hippocampus—you’d have to open up somebody’s head. That’s done in certain clinical conditions, for example, in neurosurgery for epilepsy, but for basic science studies, that kind of research where you knew what neurons were active in a baby’s brain and then you stimulate them in adults is not currently possible in humans.

Mills: Some people think that you can hypnotize people and regress them back far enough to retrieve these memories. Is that scientific in any way at all or is that just a lot of hocus pocus?

Turk-Browne: My initial reaction is hocus pocus, but what I will say is that it’s absolutely true that one thing that helps us to recall forgotten memories is to put us back in the same sort of mindset or context that we were in when we initially encoded the memory. This is the idea of context that when you store a memory, it’s not only stored in terms of the content of what you’re experiencing, maybe the people at a birthday party or what flavor the cake was and so on, but also kind of where you were in space, what time of day or what time of year it was, what’s going on in the world and so on. So our memories are deeply embedded in the context in which we experience them, and it absolutely is a very effective strategy to kind of guide yourself or other people back to the same mindset in order to prompt retrieval.

I don’t know that hypnosis is particularly good or bad at doing this, but this is the classic—if you’re trying to find your keys, think about where you’ve been today. And so we do this ourselves and professional memory athletes do this as well, using techniques like the method of loci where you sort of trace steps in your mind to reconstruct how things had been stored. So there’s some truth to that. Whether it’s possible to provide contextual cues in an older child or an adult that would help them sort of get back into the same state they were in as a baby is a much more challenging question.

Mills: So what happens to the memories that babies might be forming? I mean, where are they? And if they exist somewhere in the brain, could they somehow be retrieved?

Turk-Browne: Prior to our research on this topic, which used functional magnetic resonance imaging, most research on infant memory was behavioral research. And so there’s very clever paradigms for several decades now in developmental psychology research to try to get at memory in babies. This is very challenging. We can’t ask a baby, what do you remember? Most of the babies under a year are preverbal and unable to answer that question or even maybe understand what the question is. And so you have to use indirect behavioral methods. So Rovee-Collier and many others have developed really ingenious methods for measuring memory based on what babies look at, how they move their bodies and so on. And so that’s one way of studying baby memory. And when you do that, you see that there’s a transition somewhere around 9 to 12 months old where babies seem to be able to start forming memories that can be expressed in these indirect forms of behavior.

Our research really asked the question, well, what is the nature an infant’s memory abilities? We know from adult research that memory is not one thing, it means several different things. There are five or six different brain systems that are involved in storing memories in different ways. So for example, a really critical brain structure is the hippocampus, which is deep in the brain. This is the structure that was resected in the famous patient, Henry Molaison—HM. It was removed surgically in his case, and he lost his ability to store what are called episodic memories, memories for specific events that happened at some location at some moment. So that’s one example of a memory system. There are other brain systems like the amygdala that are involved in storing emotional memories, the cerebellum that’s involved in storing motor memories, the striatum or the basal ganglia that seemed to be involved in storing reward related memories.

And so really when you’re measuring an infant’s behavior, you don’t know which of these brain structures or perhaps other developing brain structures might be responsible for what infants can start remembering. And so our research using functional MRI was able to look under the hood and say, what’s happening in a baby’s brain when memories are being formed? And we focused on the hippocampus in particular because it’s deep in the brain, it’s—in adults, the only way to access it through brain imaging is with functional MRI—a part of my lab focuses on adults. So we naively at the time thought, well, why don’t we do functional MRI experiments in babies, it’s safe.

MRI’s been used in babies for a long time, but they’re almost always asleep. And the reason is that babies are not good participants for functional MRI studies. You have to stay completely still. It’s like taking a photograph at night. If you move, the image is blurry. And usually you need to collect an hour or two hours of data in order to find the signal in the noise. But babies have a short attention span of a few minutes if you’re lucky. And so we had to develop techniques to figure out how to do these sorts of experiments in infants. But when we did, we were able to study the hippocampus and we showed that beginning around one year of age, the infant hippocampus campus is participating in the storage of episodic memories that last for at least a few minutes. Overall, we saw this across our whole age range from 3 months to 2 years, but the effects were strongest in babies starting around 1 year of age—where greater activity, brain activity in the hippocampus when they were experiencing something new was associated with better memory later on.

Now we measured their memory in this case based on increased looking at things they were familiar with. So that’s one clue that actually the core memory functions of the brain involved in episodic memory, those basic mechanisms for storing these really rich memories that we can feel nostalgic about and reflect on that that sort of core machinery may be coming along online around one year of age. What we know, we published this work a couple of months ago and it’s really opened up a lot of questions because as we talked about at the beginning, people don’t tend to report many memories before age 4 or 5, but the machinery seems to be starting to work around age 1. And so what happens to those memories that are formed 1, 2, 3 years old that makes them inaccessible? And we have different ideas about why that might be that we’re starting to test.

Mills: I have to ask, how do you keep a 1-year-old baby still enough to do an fMRI?

Turk-Browne: I still have a lot of imposter syndrome as a developmental psychologist, I trained as a cognitive psychologist, which is a blessing and a curse because I was naive and foolish enough to want to try this. And I convinced some really adventurous graduate students to help me with this. So Cameron Ellis was one of my first graduate students, who’s now a professor at Stanford, and the lead on this memory study that I was talking about is Tristan Yates. She’s now a postdoc at Columbia University, and we have a large team of students and postdocs who’ve been working on these projects. So there’s no short answer other than a lot of trial and error, almost over a decade of efforts now.

I started with the thought that we needed to engineer some really clever solutions for this, but as I should have known, it actually all comes down to the soft stuff, having parents feel really comfortable, getting them to participate in the research, to help in the procedures, setting up the scanner environment to be really comfortable for the baby, having them bring comfort items like pacifiers and blankies and stuffies, giving them bottles and snacks during the scans. I have found goldfish hiding in our MRI room. It’s really variable.

In some ways it’s not what you want in science because no two sessions are alike. We’ve scanned—we’ve done probably closer more than 400 of these sessions now, and I don’t think any two of them are the same. So this is not repeatable science, but this is very adaptive, flexible science where we find ways to keep the baby happy and comfortable long enough to engage them in some short and statistically powerful tasks to get enough data to do something. And it’s now feasible. It’s not easy. But before us, there were a couple of groups who are doing this kind of work, Rebecca Saxe at MIT and others. I’ve been pioneering these methods. And so this is a growing field. I think it’s extraordinarily exciting. There are unlimited questions about the baby’s mind where by looking under the hood, you might be able to reveal something about how their mind works before they’re able to express themselves or behave in more sophisticated ways.

Mills: I’m glad you’ve assured our listeners that no babies were harmed during this research.

Turk-Browne: Absolutely not. We started this research right around when I had my own babies, and although they were never formally participants in our studies, because that felt like a conflict of interest, they guided everything we did. And in fact, often babies are unhappy when they come out of the machine because they’re looking at really interesting things and they’re very comfortable. I’d say about a third of the time, the babies fall asleep and have a nap after 10 or 15 minutes in there. So they’re quite comfortable. And actually, we’ve had parents ask us if we could just let them sleep a bit longer in the machine.

Mills: But isn’t there that really loud noise through the whole experience? Are they dealing with that?

Turk-Browne: That’s right. So MRI machines have a lot of moving parts and they clang. And so one of the challenges early on was how to, I mean, this sounds trivial, but it took a couple of years of work how to apply hearing protection in a way that would be durable and safe and redundant. And so we came up with a system that involves three layers of hearing protection, and we can attenuate the sound levels down to everyday sound levels that a baby might encounter on a sidewalk or in daycare and so on. We know that it works because they don’t startle when the machine turns on, they’re quite comfortable and babies are not shy about telling you when they’re surprised about a noise. And so that part’s working quite well. And really excitingly, we just developed with a company noise canceling headphones for babies. In the past, we’ve used a lot of hearing protection to make sure that the machine noise wasn’t too loud, but now we have headphones that can play sound, so it’s not just silent to them. And we can look at really interesting questions around language and multisensory processing and so on. So we’re still evolving our techniques, but that’s a great example of something that seems not especially technical or sophisticated, but that makes all the difference for their comfort.

Mills: What else might brain imaging research be able to tell us about how babies experience the world? I mean, could it give us insight into what it’s like to be a baby and experience the world as a baby?

Turk-Browne: Yeah, indeed. I mean, William James and others have had this idea that the infant’s experience is really different, the blooming buzzing confusion and other ideas. And it’s very hard just trying to study what a baby remembers when you can’t ask a baby and when you can’t give them a complex task. Indirect measures, whether they’d be behavioral—looking time or reaching behavior—whether they be based on brain measurements are the only option. And so it’s quite hard to know about the nature of an infant’s experience.

Thomas Nagel, the philosopher, has this classic example of trying to understand what it would be like to be a different animal. What does it feel like to be a bat with very different sensory systems, for example? And there’s an analogous question in infants, what’s it like to be a baby? It’s hard for us philosophically to look at a baby, as an adult, and sort of understand the nature of their experience. But we can make some steps in that direction using advanced analysis approaches that have been developed in adult cognitive neuroscience research to try to infer something about what baby experiences.

So to give you one example, one of Tristan’s other studies was looking at a phenomenon called event segmentation. And this is the idea that although we experienced the world or—the information we receive from the world is continuous, we’re constantly flooded with sensory information—we experience the world actually in more discreet events. So for example, Jeff Zacks at Wash U. has a beautiful line of research over many years showing that if you present a movie to participants and you ask them to press a button when they feel like something has changed, something has happened in the movie, there’s something new happening, people press the button at basically the same time. That is, we perceive a continuous world in discrete ways. Our mind sort of carves up our experience into events.

And this is one way of asking about infant experience. Do they experience the world in events or in the more James kind of idea—it’s all an undifferentiated mess, a jumble, a soup of experience. And so Tristan applied brain imaging methods that had been used to figure out when does the brain decide that a new event has started and applied that to infants watching a movie. And we had different ideas. So maybe infants are just really, the way they experience the world is just totally dependent on the outside world. If there’s a loud noise or a bright light or color or something changes in the world, they immediately sort of pivot and they’re reactive rather than having an inner narrative of what’s going on in  the movie. And adults, we can use these methods, you can see that they might separate events every minute or two. And so there’s a question, do infants have shorter, more reactive events?

In fact, we found the opposite that infants tended to have more stable brain activity over time. They carved up watching an animated movie into fewer events. They seem to integrate more over time and have longer timescale experiences. So that’s a question about the perception of time, about the continuity of experience. It wasn’t noisier, it was just longer. And so that’s the kind of thing we could do. There are other techniques for brain decoding where you could try to figure out, well, here’s what a baby’s looking at, but what is being processed in different parts of the brain? And we can use machine learning or other AI-like techniques to infer the content of what they’re experiencing. So yeah, there’s many possibilities in that direction. Ultimately, it’s going to be hard to validate that that’s the conscious experience of the baby without being the baby, but at least we can start to reveal some of what might be the contents of their experience.

Mills: Where is that moment when you say, Aha, that’s a memory. The baby just remembered something. How do you know? Because showing different images, the kid’s in an MRI machine, so you’re not able to do eye tracking. I mean, the child is in there. So how do you know, bingo, there’s a memory?

Turk-Browne: So in this case, we are doing eye tracking in the machine. So I didn’t mention it, but there’s an MRI safe camera that’s positioned over their face, so they’re lying in the machine. If anyone’s ever had an MRI before for your head or your brain, you’re lying on your back. And so that’s where the babies are too. And they’re looking up and there’s a screen over their face. And so that’s where we’re presenting the images that they’re looking at, and there’s a little camera sitting there that’s tracking their eye movements. So we adopt a technique that’s widespread in developmental psychology and infant cognition research called visual paired comparison. So we’ll present the baby with a series of individual images. It could be a photograph of a face, a dog toy, a landscape, a series of these individual images to test their memory for those images after delay and after other images.

We have test trials in which we present two images at the same time. So for example, the face that they had seen before alongside another face that they hadn’t seen before on the other side of the screen. And in this kind of a task after they’d only seen the face once before we hypothesized that they would look more at the familiar face if they remembered it from before. If they didn’t have a memory of seeing that face before, they should have no reason to look at one face more than the other. And so that’s actually the behavior that we’re measuring in that study, which is to measure how much they look at a repeated image on one of these test trials and then go back and say what was happening on the brain for images where they showed this strong familiarity preference at the test.

Mills: How did you get interested in studying this topic? You mentioned you’re not a developmental psychologist by training, so what got you on the path of studying memory in babies?

Turk-Browne: My lab started as a cognitive neuroscience lab focused on the adult or at least mature brain and in particular we’re interested in trying to understand the coordination of different cognitive functions. In cognitive psychology, we often study different aspects of the mind separately, you might study perception or attention or memory, but ultimately in the real world, all of these functions are intermixed. We remember what we perceive and what we remember influences what we pay attention to or perceive. And so we started in my lab studying the interaction of different brain systems in support of cognition. And so I’ve long been interested in how humans learn and remember, and this isn’t a profound insight, or at least it’s not a unique insight, but we were studying this in adults, and I think there’s learning that happens throughout life—I like to think that in my 40s. But if you’re really interested in learning, you can’t ignore the first couple years of life.

This is this magical period of plasticity where we learn language effortlessly, we learn how to walk, we form social relationships, develop food preferences. Focusing just on adults if you have an interest in learning and memory is leaving a lot of the variance off the table. And so that’s kind of how I approach this to say, well, I’m also interested in developmental change and so on. But even as a case study of saying, here is a brain of an infant that’s learning in astronomically sophisticated fast ways, and what do we understand about how that works? We started from understanding how the adult brain learns and remembers. I talked before about some of the brain regions that are involved in that. When I got interested in working with infants and toddlers, there just wasn’t a lot of research on those brain regions.

Those are very well understood in adult humans, they’re very well understood in animals, both young animals and older animals. But there just wasn’t a lot of research on memory systems in early human development. And the reason is is that most of them are deep in the brain, like the hippocampus, the amygdala, the stratum, they’re far away from the surface of the brain. And most of the existing methods for measuring the infant brain used scalp based measures like electroencephalography or EEG or functional near infrared spectroscopy, FNIRS. And these are really powerful and good methods for measuring the cortical surface but cannot resolve deep brain signals. So the shorter answer to your question is this is a fascinating period of life for learning and memory. And because of my training and background in adult cognitive neuroscience, I had expertise with the tools needed to measure the key brain systems in infants.

Mills: Do you think that your findings have implications for parents or preschool teachers or anyone else who is interacting with very young children?

Turk-Browne: I do. I mean, the research was done for basic science purposes, trying to understand the nature of early memory and memory in general. Memory is a fundamental part of who we are as the basis of our personality, our relationships, our expertise. It’s identity defining. And so I think in a basic science way, understanding how early experiences are learned and remembered is fundamental for understanding human nature in a more applied way.

I mean, I think parents are quite interested in understanding what they should do with their kids. I think about this all the time, even things like when you should take that big vacation, are they likely to remember going out West if we go on a vacation before they’re old enough? Do they remember meeting their great grandparent or their grandparent who’s since passed away? I think parenting often, we put a lot of pressure on ourselves to try to sculpt the experience of our kids in order to build the kind of people that we think we want them to be.

And so I think this research suggests actually that even though we later in life don’t remember those first few years, actually, there are memories being formed that are influencing our identity and our behavior, even if we can’t consciously recollect them. So I think this emphasizes the sophisticated cognitive abilities of infants and of toddlers, and it suggests that we shouldn’t think about memory as being a black hole for the first few years, but rather it’s there. Trying to understand where those memories go, how they get combined, and how the influence our behavior, those are some of the questions that we’re working on. So yeah, that’s the short answer.

A slightly longer answer. I won’t go on too long, but just to say that many of the core things that we want kids to learn in more of a daycare or a school setting are being learned via these memory systems. You talk about language acquisition or learning about concepts or categories, learning about social relationships. That learning happens in these brain regions. And so if we want to understand both normative, how does the developing brain learn in an ideal case, in a healthy way? Also, when things go awry in development, either early on or later, how does that impact knowledge and skills, language delays and so on? If we had a better understanding of the underlying brain systems that absorb our experiences and build memories that might point to some potential ways of either diagnosing or treating developmental disorders in the long run.

Mills: What’s next? Are you continuing this same line of research? Are you expanding moving into other areas?

Turk-Browne: The study that I referenced before, which was published in Science in March, is a starting point. This is showing that infants have the capacity to form episodic memories in the hippocampus beginning around one year of age. Because we were doing both the encoding and the retrieval in the same study, it was only a few minutes before they were tested on their memory. So we know that they can store a memory, but we don’t know how long it lasts. And this is a really critical question when we think about our lack of early memories, we’re really talking years. It’s our autobiography. And so it’s an open question about how durable these early memories might be. Maybe that’s part of the explanation.

It’s also unclear if they are durable, why can’t we access them? The mouse study I mentioned before suggests they’re there. I’m not suggesting we should go around stimulating the brains of humans to try to remember being a baby, but maybe we could figure out some clues from that mouse study. Just putting the mouse back in the maze wasn’t enough to recall the memory. That is the sensory cues, the landmarks in the environment, the geometry of the maze, maybe the smell of that room—that wasn’t enough to pull up the memory.

And I think something very much like that happens in us that as a baby we haven’t experienced—maybe you go to a grandparent’s house and you’re, I don’t know, maybe a year old and maybe you’re sitting up or lying down. So you have a certain vantage of the environment you’re in, you’re hearing conversation, maybe parents and grandparents talking, but you don’t really understand much of the language or you certainly don’t understand the concepts being expressed, the narrative of the story. You’re missing certain categories to make sense of the objects in your environment, but your hippocampus is storing a memory. That’s what our findings suggests.

So it’s creating a snapshot of the sense that the rest of your brain is making of that experience. The hippocampus gets input from the cortex and just takes a sort of a picture of what the rest of the brain is processing. Now imagine you grow up and you’re 5 years old or 20 years old, and you go back to your grandparents’ house. It might be the exact same room, the exact same people. If it’s like my family, maybe the exact same arguments from 20 years ago. It’s the same sensory information, but it’s not calling up that infant experience that we had. And the hypothesis that we’re pursuing is that it’s because how the rest of the brain develops influences the input that the hippocampus gets. So now you can understand the conversation. You’re taller and you’re walking around, so you’re seeing different things. You have certain concepts and categories that you didn’t have before. That becomes the input to the hippocampus. That’s like what you’re searching for in a search engine. Those are the terms that you’re using to try to look up the memory, but that’s not the content of your original memory. You couldn’t make sense of what was happening.

So the hypothesis there is that because of the development of the rest of the brain related to language and vision and concepts and thinking that actually there’s a mismatch between how the memory was initially stored and how we’re processing the retrieval cues that could help us to pull up that memory. And there may be ways to trick that system into sort of get back to some of the earlier cues, maybe some cues like smells, olfaction might be more potent at retrieving these early memories because there’s more consistency in how the brain processes them from early in life to later in life.

Mills: Like Proust’s madeleine, right, the flavor of the cookie. But you’re sure the memories are still there, or is that something that you’re really trying to ascertain?

Turk-Browne: That’s what we’re working on. Yeah. Trying to ascertain. In some sense, they’re absolutely there because memories are accretive—you kind of build memories on top of memories. So language is a set of experiences that have been layered on top of each other, our knowledge of our hometown or our childhood home. It might not be a specific memory of that time—I had a temper tantrum and ran up to my room and slammed the door and the turtle fell out of its aquarium, or very specific things. But I do remember in a general way where I lived and how many bedrooms our house had. And so we have memories from early on, but really the deep question is, can I remember specific events from my life that happened when I was young? And we don’t currently know whether those kinds of really specific episodic memories persist much beyond early childhood, although that’s something we’re working on in longitudinal studies.

Mills: This is great work. Dr. Turk-Browne, I want to thank you for joining me and telling our listeners all about what you’re doing.

Turk-Browne: Thank you so much for interest. I really appreciate it.

Mills: You can find previous episodes of Speaking of Psychology on our website at speakingofpsychology.org or on Apple, Spotify, YouTube, or wherever you get your podcasts. And if you like what you’ve heard, please follow us and leave a review. If you have comments or ideas for future podcasts, you can email us at speakingofpsychology@apa.org. Speaking of Psychology is produced by Lea Winerman.

Thank you for listening. For the American Psychological Association, I’m Kim Mills.