Why our memory fails, and how to improve it Our memories are easily contaminated. We can be made to believe we rode in a hot air balloon or kissed a magnifying glass — even if those things never happened. So how do we know which of our memories are most accurate? This week, psychologist Ayanna Thomas explains how memory works, how it fails, and ways to make it better.

Did That Really Happen?

  • Download
  • <iframe src="https://www.npr.org/player/embed/788422090/788632007" width="100%" height="290" frameborder="0" scrolling="no" title="NPR embedded audio player">
  • Transcript


From NPR, this is HIDDEN BRAIN. I'm Shankar Vedantam. Almost a century and a half ago, a young German philosopher named Hermann Ebbinghaus picked up a curious book at a used bookstore. It was called "Elements Of Psychophysics" and it described how to gather data on invisible mental processes. The book gave Ebbinghaus an idea.

AYANNA THOMAS: And he thinks about, OK, well, we all have this experience of learning information and forgetting information. How can I study that?

VEDANTAM: Psychologist Ayanna Thomas - she says Ebbinghaus wanted to answer a big question. How does memory work?

THOMAS: He really wanted to understand memory at its most basic core. How quickly can we learn new information that we've never been exposed to? And how quickly does that information degrade?

VEDANTAM: This was the late 1800s. There was no protocol on how to run a psychology experiment. Most insights about memory came from philosophers and theologians. So Ebbinghaus came up with a plan. He would run an experiment on himself. First, he needed something to memorize. He came up with the idea of nonsense syllables - three random letters strung together. He put them on some cards.

THOMAS: So more than 2,000 of these.

VEDANTAM: Then he shuffled the cards, divided them into groups and set about learning them in a variety of controlled conditions. Sometimes he read the cards aloud...

UNIDENTIFIED ACTOR #1: (As Hermann Ebbinghaus) Leg, muf, vok, dal, sen, cap, nud, suk (ph)...

VEDANTAM: ...Forcing himself to read them at the exact same rate in the same soft tone of voice keeping time to a ticking metronome.


UNIDENTIFIED ACTOR #1: (As Hermann Ebbinghaus) Dal, sen, kep, nud, suk, kas, leg (ph)...

VEDANTAM: He would read a group of cards aloud once, hide it, and then try to recite it from memory. Every time he made a mistake, he would stop, note it in his records...


VEDANTAM: ...And start over.

UNIDENTIFIED ACTOR #1: (As Hermann Ebbinghaus) Leg, muf, vok...


VEDANTAM: He kept doing this - stopping and starting over, stopping and starting over - until he could recite a group of syllables perfectly.

UNIDENTIFIED ACTOR #1: (As Hermann Ebbinghaus) Suk, suk...

VEDANTAM: He repeated this process periodically to see how much he could remember 20 minutes later, an hour later, nine hours, a day and so on.


VEDANTAM: Stopping, starting over, stopping, starting over - for more than six months.

UNIDENTIFIED ACTOR #1: (As Hermann Ebbinghaus) Vok, dal, sen, kep, nut (ph)...

VEDANTAM: The experiments were highly precise and monotonous and required a huge amount of self-discipline. His greatest enemy was boredom. Years later, Ebbinghaus told a group of psychologists about the tedium of his most celebrated accomplishment.

UNIDENTIFIED ACTOR #1: (As Hermann Ebbinghaus) I shall not go into the agonies that months on end of such study engendered in me. I attempted to...

Lek, muf, vok, dal, sen, kep, nut, muf, ton, gen (ph)...

And with that respect, the magnitude of the task seemed staggering to me.

THOMAS: And you might ask, why would anyone want to do that? Well, his motivation was he wanted to see what information was remembered but also how much information was forgotten.


VEDANTAM: This week on HIDDEN BRAIN - the science of memory from the early ideas of Ebbinghaus to the latest discoveries of modern psychologists.


VEDANTAM: How do our minds hold on to the events of the past? What are the most common ways our memories fail us? Can we ever really know for sure that our memories are telling us the truth?


VEDANTAM: How we remember, why we forget and the simple lessons we can all learn to make our memories sharp and vivid.


VEDANTAM: Hermann Ebbinghaus remembered thousands of three-letter strings and then tried to see how many he remembered the same day, the next week, the next year. He discovered something that many of us experience. Over time, he started to forget things. But since he was running an experiment, it allowed him to actually measure how much he was forgetting and how quickly. It led him to discover something that is today called the forgetting curve.

THOMAS: That forgetting curve, if you look at it, you see an exponential decline in what he can accurately remember as time passes.

VEDANTAM: This, again, is Ayanna Thomas, a psychologist at Tufts University who studies memory.

THOMAS: You see a really quick drop off, so by the time you get out to a week and then a month, there's a lot of information that's been lost. But I also think what's really neat is a month later, there's quite a bit of information that still has been retained.

VEDANTAM: What I find really interesting about this story, Ayanna, is that it reveals two things that at some level feel intuitively true to us. All of us have this experience of remembering something and then finding a week later we can't remember what we remembered. And also realizing that if we try and relearn what it is that we have forgotten, it is quicker the second time around than the first. So at one level, he is telling us things that we all experience in our life. But was he also, in some ways, putting on the map the idea that you can actually study these things empirically, that there was a science to memory, that it was amenable to measurement and to precision?

THOMAS: Indeed. He was the first to bring this in to a laboratory where he has a hypothesis about how memory is going to operate. And he operationalizes the construct of memory. So I think about memory as a really broad construct - a really broad term. And that term can encompass so many different things. Right now, I'm using memory to communicate with you, right? I'm bringing things up from my long-term memory and I'm using my working memory to communicate that information. But that's not how Ebbinghaus operationalized it. So he thinks about, OK, well, we all have this experience of learning information and forgetting information. How can I study that?

And in order to study memory and what Ebbinghaus introduced us to is that we have to figure out what our construct is, we have to define that construct, and then we have to develop tasks that are appropriate to the definition of that construct. You can think about memory in a whole host of different ways. I know how to ride a bike, but I can't tell you exactly how to do it. I just know how to do it. That's still an element of memory and not the one that Ebbinghaus was studying.


VEDANTAM: After Hermann Ebbinghaus, other researchers came along and said, OK, we can study memory systematically. Through careful observation and measurement, we can see how memory works and when it fails. In the 1930s, Frederic Bartlett set up an experiment that revealed a very interesting aspect of memory. The experiment revolved around a story - a Native American legend called "The War Of The Ghosts."


UNIDENTIFIED PERSON: One night, two young men from Egulac went down to the river to hunt seals, and while they were there, it became foggy and calm. Then they heard war cries, and they thought, maybe this is a war party. They escaped to the shore and hid behind a log. Now canoes came up, and they heard the noise of paddles and saw one canoe coming up to them. There were five men in the canoe...

VEDANTAM: Frederic Bartlett tested the volunteers on what they remembered of the story, sometimes hours later, sometimes days, even months and years later.

THOMAS: And what he found was there were these systematic changes to the way these individuals were recalling. They forgot details that were foreign to them. So it was a native story about these ghosts warring during dreams. And so many of the subjects just didn't remember some of these elements that were foreign to their cultural understanding. Many of the subjects also made systematic changes where a word like canoe was changed to boat. And so what he found was that the subjects, when they were retelling this story, they were changing the story so that it aligned with their prior experiences because those prior experiences help us, generally, to remember information from our autobiography. But what those prior experiences are doing is they are putting those new experiences into a just-like story, where we have an idea of how things unfold, but we don't remember the specific details.


VEDANTAM: In other words, the forgetting wasn't just random. Hermann Ebbinghaus had suggested that memory was like a sieve. You put stuff in and it falls out gradually over time. Frederic Bartlett said, no. There are certain things that get remembered and certain things that get forgotten. What you already know, what you're familiar with - that shapes what you remember, what you can recall and what gets forgotten.


VEDANTAM: Effectively, what the experiment showed is that when we recall something, what we remember is only partly about the thing we are remembering. Our memories are also shaped by all the other things we know. Things that are similar to what we already know are more likely to stick. Things that are less similar are likely to drop off or get modified so they fit better with what we already remember. Memory, in other words, is not a simple process of opening a mental file drawer and taking things out. It involves a process of reconstruction.

THOMAS: And I always think about it like a paleontologist, where a paleontologist uncovers a fossil just as we uncover a memory. We remember a piece of an element. And we have a piece of the puzzle that we start putting together with other fossil pieces that have been uncovered. And we reconstruct a - and, you know, from the paleontologist perspective, maybe a skeleton of a dinosaur. From our perspective, we reconstruct a memory of a past event. But that paleontologist doesn't have all of the pieces. And what that individual has to do is fill in the gaps with best guesses and prior experience with that person's expertise on what could go there, just as we fill in the gaps of our - these memory holes with our best guess of what could go there.

VEDANTAM: I mean, that is both wonderful and terrifying at the same time, Ayanna.

THOMAS: Yeah, I think it's really interesting. I think it's really interesting to think about why we do these things, why we misrecollect our past, how those kinds of reconstruction errors occur. And I think about it in my own personal life - I share my memories with my partner. And many of us who have partners, we have these sort of collaborative ways in which we recollect. But those collaborations often result in my incorporating information into my memories that were suggested by this individual, but I never experienced. And so I might have this vivid recollection of something that only my partner experienced because we've shared that information so often. And so that's how we can distort memories in the laboratory. We can just get individuals to try and reconstruct events over and over and over again. And with each reconstructive process, they become more and more confident that that event has occurred.

VEDANTAM: So I want to look at this process of mental paleontology a little more carefully. And there have been a number of experiments that have been conducted along these lines which ask people in some ways to place themselves emotionally in events. And these experiments find that we change what we remember as a result of these interventions.

THOMAS: Yes, there have been so many experiments now - I guess we're talking about 25 years' worth of research - looking at how people come to believe in things that never occurred. And so we see that people can come to believe that they were lost or that they took a hot air balloon ride or spilled punch on the parents of a bride at a wedding when they were young. And generally, the way that these studies work is we get some sort of information about kinds of events that had happened to them or had not happened to them before the age of, let's say, 10. And then that individual goes through some sort of suggestion procedure.

VEDANTAM: Researchers wanted to see if they could induce volunteers to develop false memories. In several studies, they described events to volunteers and then told them that they had drawn these accounts from conversations with their parents or siblings. Another study used a more high-tech method.

THOMAS: One of my colleagues went as far as distorting pictures using Photoshop and inserting participants into the picture of a hot air balloon...


THOMAS: ...And finding that that resulted in, of course, creating this whole false memory of an event that never took place. And that's quite a distinct event - right? - taking a hot air balloon ride. Even - you would think you would remember that. And if you didn't remember that, you would think that probably never happened. So the fact that - this was Kim Wade and colleagues - the fact that she was able to distort individuals with this technique really demonstrates the suggestive nature of pictures.


THOMAS: And so personally, you know, I have this experience when I look at old photo albums that my parents have collected that I go back and I'll look at these old pictures of myself and, you know, I'm at 5 graduating from kindergarten. And I'll see these pictures and I have no memory of these events. But you've had this experience. We've all had this experience where you see a picture and you start to think, oh, yeah, I kind of remember that. So it's like you're with your wife and she's suggesting something. Pictures are even more distorting from that perspective because we can mentally generate that image. And that mental generation is a big cue to us.


VEDANTAM: And in some ways, stories function the same way. I remember my parents telling me a story - and I don't have any personal recollection of this - but telling me a story when I was a very small child and there was a family gathering at our home and many people had come over. And as with many Asian households, people left their footwear at the door or at the entrance to the house before coming in. And apparently, I sort of wandered off from this gathering and I essentially stole everyone's left shoe or left slipper and basically took all the left shoes and left slippers and basically moved them to a new location. And I have obviously no recollection whatsoever of doing this, but I've heard the story so many times that, to me, I can actually sort of vaguely see myself doing this. And I feel also vaguely proud that even at the age of 2, I could pick out just the left slippers from the right ones.

THOMAS: I think they're making that up. I don't see how a 2-year-old could have really done that.

VEDANTAM: (Laughter).

THOMAS: I think that you might have stolen a few shoes and this story has changed so much over the years. Can you imagine (laughter)?

VEDANTAM: So that's very sad, Ayanna. You're ruining one of my precious childhood memories.

THOMAS: Well, if it really did happen, that's pretty amazing.

VEDANTAM: (Laughter) I'm going to go with that version. I'm going to go with that version.



VEDANTAM: What Ayanna and other researchers have found over and over again is that our memories are fallible. And the implications of this extend far beyond how we think about our own childhoods. They extend to serious settings like the criminal justice system, where we constantly ask people to make recollections or remember things under oath. Do people remember things accurately? In the early 1970s, Elizabeth Loftus and John Palmer ran an experiment to find out.

THOMAS: What they were really interested in was how people recollect a witnessed event and how the way in which questions are phrased influence our recollection.

VEDANTAM: The researchers showed their volunteers short films of car accidents. Then they asked the volunteers some questions about what they had just seen.

THOMAS: And there was a critical manipulation in the first experiment. What they did between groups of participants was they changed one word in the question.

VEDANTAM: Some people were asked how fast were the cars going when they contacted each other? Others were asked a question with a more vivid verb, like, how fast were the cars going when they bumped into each other, smashed into each other or hit each other? The people who heard the words smashed or collided recalled that the cars were going faster than the people who heard the word bumped. So changing just one word in the question changed people's memory of the speed of the cars. Then there was something else. The volunteers were asked whether there was broken glass at the scene of the accident. The researchers found that volunteers were much more likely to remember seeing broken glass if the word smashed or hit had been used.

THOMAS: And I think it's really important to note that there was no broken glass in any of the images that participants were shown in the study.


THOMAS: So not only did it change their estimates of speed, which is subjective in and of itself, but participants distorted their recollection in the act of recalling because of the word that was used. And so with no broken glass being present, participants remembered broken glass because that question was posed in a particularly suggestive way. So this has really important implications for criminal justice and how we question witnesses, how we get information about investigations from witnesses. The way in which officers at the scene of a crime, investigators during the context of interviews - the ways in which they question witnesses, whether witnesses are questioned together, whether witnesses talk to one another after the fact and now in the age of social media, now witnesses being able to do their own online investigations - all of these elements are serving to introduce factors that can distort the recollection of witnesses.


VEDANTAM: What is especially, you know, dangerous about this is that when people think back, they can't actually tell which of those memories is real and which was made up. After some time, I can't distinguish between the things I actually remember and the things that I have reconstructed. They both feel exactly the same.

THOMAS: And I think you're hitting on one of the issues, right? You think that some memories that you have are, as you say, actually remembered. But they're all reconstructed. The question is, which reconstructed memories are more accurate? And can you learn to monitor that process? And I think people can.


VEDANTAM: Coming up, we delve more deeply into Ayanna's research to understand the many ways our memories can fail us.


VEDANTAM: The stories we tell ourselves about past events can also change the way we remember the past. Ayanna once brought volunteers into the lab and asked them to imagine doing both ordinary things, like bouncing a ball, and bizarre things, like kissing a magnifying glass. I asked her why she did this and what she found.

THOMAS: We were really interested in whether people could accurately distinguish between what they imagined versus what they actually did. And we compared performance on very simple common kinds of activities that people would do in their daily lives - that they could easily think, of course I did that - as compared to things that they would never do. Like, we had kissing magnifying glasses and rubbing lotion on the chair - very strange things that people generally would not do in their daily lives. And so they did some of these things and they imagined some of these things.

But critically, what we did was we brought them back in on a second session and we just had them imagine a bunch of - some of these activities over and over and over and over again. And what we found when we tested them two weeks later was that when people imagined these actions, whether they were sort of usual common kinds of actions, like bouncing the ball, or unusual actions, like kissing the magnifying glass, irregardless of the bizarreness of the act, the more times people imagined it, the more likely they were to say that actually they performed it on that first day. So we were able to get people to believe that they did these strange things when they only imagined them repeatedly. And so you can extrapolate and say, well, can you get people to believe they've done very strange things in their lives that they would never have done? And I think the answer is, yes. You can.

VEDANTAM: Now, you did find that people were less likely to remember doing the bizarre things than the ordinary things. So it was easier to say, I remember bouncing the ball the first time I came in rather than I kissed the magnifying glass the first time I came in. But there were errors on both counts.

THOMAS: Indeed. So it's easier to distort people's memories if the information is plausible. The more common, the more usual, the more easy it is for people to imagine themselves in this activity and maybe pull from their prior experience to distort their memory. And so the more unusual, the less plausible something is, the less likely it is for people to come to believe that these events took place or that they engaged in some particular activity.

VEDANTAM: So it's interesting when I think back to that childhood memory of my taking everyone's left slippers, one of the things that I realize my brain is doing is that in my mental image of that scene, I'm painting a picture of myself dressed in the clothes that I've seen myself wear in a childhood photograph. And, of course, I have no idea whether I was wearing those clothes on the day that this particular event happened. And as you say, it might never have happened. But it's interesting that I'm fleshing out things that actually I have no idea about using different elements of what I know about the past, in a way, to create essentially this picture that makes sense of a past event that I don't actually remember.

THOMAS: And those pictures of our - when we were children are highly suggestive of the way that we construct our childhood memories. I've seen myself in I think one or two outfits in pictures, and those are - that's pretty much how I imagine prior events. I must have been wearing this dress. Obviously, I had more than two dresses - or maybe I didn't - but I don't know. In fact, I remember very little from those early days. So, yeah, we use our pictures. We use those stories your family has embellished and you stole all the left shoes.

VEDANTAM: (Laughter).

THOMAS: And those all serve to create a really fun story to tell but maybe not an accurate depiction of what actually took place.


VEDANTAM: The fact that memory often is not an accurate depiction of what took place can help us understand how two people can have very different recollections of the same event.


CHUCK GRASSLEY: Can you affirm that the testimony you're about to give before the committee will be the truth, the whole truth and nothing but the truth so help you God?



VEDANTAM: One of the things I wanted to ask you is, when you see stories breaking in the news that revolve around people's recollections of events, I'm wondering how you, as a memory researcher, process these stories? And the story that springs to mind is the one that occurred some time ago where Christine Blasey Ford accused now Supreme Court Justice Brett Kavanaugh of sexually assaulting her when they were both teenagers.


CHRISTINE BLASEY FORD: I don't have all the answers, and I don't remember as much as I would like to, but the details about that night that bring me here today are the ones I will never forget. They have been seared...

VEDANTAM: She said she had a very vivid memory of this happening.


FORD: Brett and Mark came into the bedroom and locked the door behind them. There was music playing in the bedroom.

VEDANTAM: He said that it did not happen.


KAVANAUGH: I categorically and unequivocally deny the allegation against me by Dr. Ford. I never had any sexual or physical encounter of any kind with Dr. Ford. I never attended a gathering...

VEDANTAM: And in some ways, the country was divided over these dueling memories. What would you have said if you were on the expert witness stand about how to understand these two very different recollections about an event?

THOMAS: Well, I think that it's important to note that both individuals could be recalling that night as accurately as they could. But the interpretation of saying that this couldn't have happened because I would have remembered it is an inappropriate interpretation of how memory works. Of course, it could have happened and you have forgotten it because often we will remember things that are particularly salient to us. And so sometimes we forget things that maybe are really important to other people because they weren't that big of a deal to us.

And so I use the analogy of my brother and I. My brother and I have lots of shared childhood memories. He's three years younger than me. And he will tell me stories that he remembers very well and I have no recollection. And it's because it wasn't that big of a deal to me, but it was a big deal to him. And so I think the important thing in the Blasey Ford and Kavanaugh case is they both - neither necessarily were lying about the - that past. It's just that memories are going to differentially degrade and differentially be distorted for different individuals. Do I know who was accurate in that retelling? I don't. Could both have been distorted? Most definitely.

VEDANTAM: What was interesting to me, based on what you've told me for the last several minutes, is that I think many of us believe that the more vividly we recall something, the more it must have happened. And I think when a lot of people watched the disagreement unfold on national television, we said, well, she clearly believes it did happen and he clearly believes it didn't happen. And only one of them, therefore, must be true and the other one must be lying because if they both vividly recall this, clearly they both would have remembered something as astonishing as this happening or something as out of the ordinary as this happening. And in some ways, what you're suggesting is actually more difficult to stomach, which is that they both could have been telling the truth or they both could in some ways have been making up something without actually realizing it.

THOMAS: Exactly. You know, you hear these people say, oh, I believe this individual or I believe this other individual. And the question is not about belief. Both have their own versions of the truth. And the - what people have to understand is that there is no getting at exactly what happened on a night 30 years ago or 40 years ago. There's just no way to get at that information. Unless there was some video recording, we do not know. I always think back to - have you ever seen that movie "Rashomon?"



UNIDENTIFIED ACTOR #2: (As character, non-English language spoken).

THOMAS: "Rashomon" is a great movie to think about from this context because you have a retelling of an eyewitnessed event. And, you know, some of the witnesses to this crime have motivation to lie, but some don't. And they just tell a different version from their own perspective. And something, you know, these - this is really what's going on, generally. When we are telling about a past event, even if it's traumatic, we're telling it from our perspective. And that perspective alters our recollection.


VEDANTAM: Just as we are sometimes too confident about our memories, Ayanna says there are other times when we're too insecure - when we jump to the conclusion our memories are failing us. Take, for example, the experience that many older people report in forgetting where they left their keys or where they parked their car.

THOMAS: And I used to think that older adults actually are demonstrating more of these memory failures than their younger counterparts. And to some extent, that's true. There is age-related decline in attention processes and memory encoding processes that likely result in these kinds of errors and will increase the frequency of these errors. But I also think that older adults are just more sensitive to these kinds of errors and are more likely to remember the error than their younger counterparts.


THOMAS: Because many individuals, as they get older, become more sensitive to potential problems with their cognition and their aging, whenever they have these kinds of failures, those failures are more salient and so they remember them more. So I think there's also a bias in reporting, where there's underreporting for younger people and overreporting for older people.

VEDANTAM: And you and others have also found that if you, in some ways, frame challenges as being about, you know, declining memory, you're more likely to see errors among older people in a phenomenon that's sometimes called stereotype threat.

THOMAS: Right. And so - and I do this work in collaboration with my colleague Marie Mazerolle. And we were really interested in how when you either activate or alleviate threat in older adults, how that influences memory and other cognitive processes. And so what I mean by that is there is a negative stereotype associated with getting older. And implicit in that negative stereotype is that you're going to demonstrate these cognitive deficits. And in fact, the majority of time I bring people into the lab that are, let's say, over the age of 65, they're coming into the cognitive aging and memory lab. And the first thing they say to me or the researchers who are testing them is, well, you know, I have a terrible memory, or I'm really worried about my memory. They're already feeling threatened...

VEDANTAM: Mmm hmm.

THOMAS: ...About their memory. And that psychological experience of just worrying about one's memory tends to negatively impact performance on memory tests. And so you - when you reduce that anxiety and you alleviate that threat, older individuals do better.


VEDANTAM: When we come back, how all of us can improve our ability to remember things. Because memory works in predictable ways, it turns out there are specific things you can do to make your own memory better.


NELSON DELLIS: Four hearts, four of diamonds, jack of diamonds, six of hearts...

VEDANTAM: This is HIDDEN BRAIN. I'm Shankar Vedantam.


DELLIS: ...Clubs, eight of clubs, nine of clubs, two of hearts...

VEDANTAM: What you're hearing is the voice of a man named Nelson Dellis.


DELLIS: ...Six of clubs, eight of hearts...

VEDANTAM: He's what's known as a memory athlete.


DELLIS: ...Five of spades, eight of spades, seven of clubs...

VEDANTAM: Memory athletes memorize huge quantities of information, like strings of numbers, lists of words and decks of cards. This is Nelson at a memory championship in 2014.


DELLIS: ...Four of spades, jack of hearts, 10 of hearts...

THOMAS: I really love watching memory competitions and watching these memory athletes be able to rehearse back long strings of numbers and recite back long passages from books. They are just like us. And we could do what they do. And what they're doing is they're using memory like a skill that can be honed and perfected.


VEDANTAM: So how do memory athletes imprint all this information in their minds? One of the most popular techniques is called a memory palace. You create a mental image of a familiar location, like your house or the street where you grew up. Then you plan out a path through that space, creating conspicuous stops along the way. Once you've done that, place all the items you want to remember at each of those stops. This mental image - the location, the path, the stops with the items on it - this becomes your memory palace.

Here's an example. Let's say you want to remember four things in a list. First, pick a location like your house. Now make a path with four stops along the way. You might start at the front door, then stop at your coat closet, followed by the kitchen then the bedroom. Now go back and mentally place the four things you want to remember at each of those locations. Once you've done this, remembering your list gets easier. Just walk through your house in your mind and look at the four things along the path. The memory palace is based on the idea that to remember things that are unfamiliar, you want to hook them onto things that are familiar. It's an ancient technique going back to Greek and Roman orators who often had to learn long speeches by heart.

THOMAS: What individuals would be doing when they were giving speeches is they would associate elements of the speech to different individuals seated in the room. If people were sitting in the same place in specific rooms, they would be able to associate, let's say, the first line of the speech with a person sitting in the first row on the right and so on and so forth. And so all of - all these speakers would have to do is look across the room to be able to remember the different elements of the speech.

VEDANTAM: What's so fascinating about what you're saying, Ayanna, in some ways is that the very same processes that in some ways can contaminate memory, which is our past experiences - the things that we know well can contaminate the things that we learn, the things that happen to us because memory is reconstructed - you can actually use this now to actually improve your memory. You can actually use it to remember things that happened by associating them with other things in your past that you're familiar with.

THOMAS: Yes. So we talked about how prior experience can distort memory. But prior experience can also scaffold memory and we can use it to learn new information.

VEDANTAM: I'm wondering if I can give you a list of five things to try and remember and have you essentially put them on a familiar path and have you do this sort of live so we can see how this works? Would you be game to do that?

THOMAS: (Laughter) We can try. I will maybe fail miserably.

VEDANTAM: Well, we - remember the stereotype threat research, though. You shouldn't tell yourself that you're going to fail miserably.

THOMAS: (Laughter) Yes.

VEDANTAM: All right. So the five items I want you to try and remember are milk, eggs, olive oil, garlic and pink cake frosting. So can you build a memory palace or one of these walks where you're familiar with where you place these five items? And I can tell you them - I can tell each of them to you as you go along the path so that you will remember them in the future. Again, milk, eggs, olive oil, garlic and pink cake frosting.

THOMAS: OK. I will try.

VEDANTAM: Do you want to try and tell us what you are doing in your head so that we can follow along with you, whatever the path is that you're setting up?

THOMAS: Well, for me, actually, I'm thinking about what I would cook with these elements. So I'm making all of these elements interact. And you threw a little curveball in there with garlic. But fortunately, garlic is a distinctive element in that list because I can make a cake with most of what you've - what you already talked about.


THOMAS: So that helps because now the four elements that were potentially unassociated are now associated...


THOMAS: ...Into one unit. So I've made something distinctive - a pink frosted garlic cake. It sounds terrible.

VEDANTAM: (Laughter) It does sound terrible. But what's so interesting here, in some ways, is you're employing imagination to boost memory. And we talked earlier about how imagination can undermine memory. Isn't it fascinating that it can play these dual roles?

THOMAS: Yeah. Our imaginations are highly effective in allowing us to remember these disparate pieces of information. And so it is associations that allow us to remember information. And when we are able to make associations and when we're able to rely on interactive, unusual imagery, we're able to remember unrelated lists of items.

VEDANTAM: So the analogy that's springing to my mind is that to think of, you know, memories - individual memories are like islands. We consolidate those memories but also potentially contaminate them by essentially building bridges to those islands from other islands that we are more familiar with. Some people talk about this as the difference between learning somebody's name is Baker (ph) and that's someone's profession is baker. If you ask that person to remember, you know, do you remember the person's name or do you remember the person's profession, it's usually easier to remember the person's profession two weeks later than to remember that somebody had a name Baker. Talk about this idea, that in some ways the context that matters enormously in terms of what it is that we're remembering and some things just come with more context than others.

THOMAS: I really like the analogy of thinking about memory as islands. We think about it as - memories as nodes, right? So you have a particular concept, like baker. And when I say baker, a number of related concepts that are connected to those nodes or connected to those islands, those associative bridges that you're talking about, are activated. So you think about bread and you think about maybe the time of day that you might go to the bakery and you think about your time in France, which I just got back from. So now I'm thinking about France. And you think about all of these contextual elements that allow you to strengthen that memory representation. But those nodes and those associations can also lead to error.


VEDANTAM: Ayanna has also studied ways in which students can get better at retaining information. One of the most effective methods - taking practice tests. Closing your books and forcing yourself to recall information turns out to be a powerful way to remember things you've learned. Another insight - space out learning and avoid trying to cram everything the last night before an exam; also try to get some sleep.

THOMAS: Memory consolidation occurs during sleep, and you have to have specific high-quality sleep in order to consolidate these kinds of memories. It's interesting. I just read a recent article about sleep and consolidation and how - and during the context of sleep cycles, we actually select particularly salient memories. And that gets preferential consolidation treatment while we're sleeping.

VEDANTAM: And in some ways, there is an analogy here with physical activity as well, right? Physical - your trainers will tell you, yes, you know, try and come in more than once a week. But they will also tell you if you try and do four hours of physical activity on one day rather than, you know, half an hour every day, you're not giving your body time to rest and actually build up strength and stamina and everything else, that actually the rest is actually part of what makes the physical activity pay off.

THOMAS: Right. So it seems that our body requires a balance in order to engage in the variety of interconnected processes that are going to result in physical change. And a memory and memory formation is a physical change in the brain. And memory retrieval is the act of reconstructing some of those changes and bringing them back together to make a conscious experience of a recalled event.

VEDANTAM: All right. So I am going to try and see if you remember that list of things that I gave you maybe about 15 or 20 minutes ago. And let's see if you can use your recipe technique, Ayanna, to remember the five items on that grocery list.

THOMAS: Yeah. I'll tell you, I actually thought you were going to ask me this pretty soon. And so I mentally retrieved it right about the time we were talking, and I was like, do I have it? I think I do. So it was milk, eggs, olive oil, garlic and pink cake frosting.

VEDANTAM: That's spectacular. And I'm sure in years to come, we will both remember the time that you made some delicious garlic cake for us both.

THOMAS: (Laughter) I'm very happy with myself.


VEDANTAM: Ayanna Thomas is a psychologist who studies memory, why our recollections do not always reflect reality, how memories degrade and what we can do to retain things better. Ayanna, thanks for joining me today on HIDDEN BRAIN.

THOMAS: Thank you. It was a pleasure.

VEDANTAM: This episode was produced by Laura Kwerel and Parth Shah and edited by Tara Boyle and Rhaina Cohen. Our team includes Thomas Lu and Jenny Schmidt. Special thanks to Jonathan Burley (ph) and Jonas Harnow (ph) for their voice acting. Our unsung hero is Jaap Murre, a theoretical neuropsychologist at the University of Amsterdam. Jaap conducted a replication of the Hermann Ebbinghaus study from the 1800s and he generously shared his time to help us understand how the first study was run. Thank you, Japp.

Since this was an episode about memory, here's one thing I want you to remember. It's a website - donate.npr.org/hiddenbrain. That's where you can support your local public radio station, which is the best way you can show your love for HIDDEN BRAIN. Again, that's donate.npr.org/hiddenbrain. In case you forget it, we've posted that link on our Twitter account. Giving to your local public radio station helps make our show possible. So thank you.


VEDANTAM: I'm Shankar Vedantam, and this is NPR.

Copyright © 2019 NPR. All rights reserved. Visit our website terms of use and permissions pages at www.npr.org for further information.

NPR transcripts are created on a rush deadline by an NPR contractor. This text may not be in its final form and may be updated or revised in the future. Accuracy and availability may vary. The authoritative record of NPR’s programming is the audio record.