How the Brain Makes Memories Scientists have been eavesdropping on the sleeping brain to find out how it makes memories of the day's events. We discuss this new research.
NPR logo

How the Brain Makes Memories

  • Download
  • <iframe src="" width="100%" height="290" frameborder="0" scrolling="no" title="NPR embedded audio player">
  • Transcript
How the Brain Makes Memories

How the Brain Makes Memories

  • Download
  • <iframe src="" width="100%" height="290" frameborder="0" scrolling="no" title="NPR embedded audio player">
  • Transcript


You're listening to TALK OF THE NATION SCIENCE FRIDAY. I'm Ira Flatow.

A little bit later in the hour, we're going to talk about a potential new malaria vaccine. But up next, a couple of brain studies in the news this week. One of the biggest mysteries of the mind is how long-term memories are formed. And a new study eavesdrops on the sleeping brain and comes up with an answer to that question. Researchers recording a rat's brain activity during non-dreaming sleep found that the animals actually replay the day's activities nearly simultaneously in parts of the brain that process initial memories and lay down permanent ones. We'll talk about this study.

And another one out in the - that's is out in the January issue of Nature Neuroscience, the one that tells us about how our brain makes memories.

Plus another study in the same journal that shows that humans can track a scent on the ground as well as a dog can. Now, here's something you can try at home, I'll tell you how the scientist did this. You just have to put your nose to the dirt. We'll talk about the students who actually did that and maybe they're being congratulated these days.

Our number is 1-800-989-8255. Nicholas Wade is a science reporter at The New York Times here in New York, and he joins via phone. Welcome back to the program, Nick.

Mr. NICHOLAS WADE (The New York Times): Thanks, Ira.

FLATOW: Let's talk about these couple of new brain studies that came out this week. This first one, the researchers were looking at how the brain forms memories what - give us a little thumbnail sketch of that study.

Mr. WADE: Oh well, I think this is really quite an important advance and gives us a keen new insight into how the brain, as a whole, works. It's based on the study of rats learning to run a maze. And we've met sometime that these - that the initial memories of learning a maze are laid down in cells in the hippocampus, that's a substructure deep in the brain.

And they are recorded in what the scientists refer to as place cells, because as the rat passes each point in the maze, a different one of these place cells will light up. So you couldn't tell where a rat is in the maze by looking at your recording systems and seeing in which of its place cells is active. So this gives you a very good handle on memories, because it turned out that these - that the memory of running a maze is played back, as you just said, when the rats pauses for a while or when he's sleep.

You can see these place cells firing. Very interesting - it's in reverse order of the route that is traveled. So you see the place cells are firing again in a very distinctive cascade in the opposite of the way the maze was run. So that's what the - Matthew Wilson of MIT reported. I think it was in February of this year.

And now he's looked to see if he can see the same trace elsewhere in the brain. So it was thought the initial memories were laid down here in hippocampus. But they sort of seem to disappear from there. And then we see - we know that long-term memories reside in the cortex, that's sort of sheet of cells on - of cells on the outside of the brain which is the sort of seat of consciousness in the higher cognitive functions. So it's been a big question as how do - how do memories can get transferred, if that's what's happening, from the hippocampus to the cortex.

So (unintelligible) reported last week, what (Unintelligible) his colleague Dr. G(ph) did, was they - this time they put electrodes both in the hippocampus and in one part of the cortex - the visual part of the cortex, the one that deals with sight. And interestingly they found the distinctive memory, the sort of reverse, rewind was placed both in the hippocampus and in the visual cortex during sleep. So this is the first time that we've known that the cortex also has its equivalent of place cells. So these would have been the cells, I guess that the rat is - it will be the places where the rat is sort of seeing things in the maze corresponding to the actual location that's in the hypothalamus.

FLATOW: So the rat is replaying this in fast-rewind.

Mr. WADE: Yes.

FLATOW: So the fact that you're seeing this after the rat is either asleep or has done this, implies that it's consolidating what it has seen into memory?

Mr. WADE: That's probably the case with the hippocampus in the - when the rat does this thing in the waking state. So what's the interesting to sort of ask about, is what's happening during sleep? So it certainly could be a consolidation of this memory.

But Matthew (unintelligible) also thinks there's another component of this and it may be - it's not simply a question of a memory being transferred from the hippocampus to the near cortex, to the visual cortex. It might rather be the cortex is trying to figure out what's going on.

It's trying to make a model of the brain and it's drawing data from the hippocampus - saying sort of, what happened, what does this mean. And he's trying to create a model that will guide the rat's behavior the next day or in future. (Unintelligible) - the distinctive aspects of how our brain are like computers that can record masses of data but have a greater - a very hard time saying what does the data mean?

Our brains are very good at extracting meaning from the data. So maybe what this experiment is capturing is the cortex in the process of, sort of interrogating the hippocampus during sleep and trying to sort of figure out to (unintelligible) hippocampus has experienced.

FLATOW: Could it be figuring out whether this is something we want to remember? We want to record permanently?

Mr. WADE: Yes, so the hippocampus maybe records all of the, sort of, raw data.

FLATOW: Like a buffer in there in the beginning.

Mr. WADE: Right. And the new recordings will maybe be sort of extracting from it, just the data the near cortex figures is meaningful. Of course, it will be the hippocampus's memories are being sort of stored in the - in a vary sort of distributed way all over the cortex. After all, there are bits of the cortex that deal with hearing and bits with smell.

So maybe the initial memories are recorded in all places based in the cortex and in hippocampus. But somehow rather the information in the hippocampus is essential for linking them all together.

FLATOW: Is the fact that the visual areas are involved mean that it's sort of watching a movie like - or what's happens during a dream. You see, you know, you create some sort of visual concept.

Mr. WADE: I mean, if they could be - I mean, they just chose to record from the visual cortex, I guess because it's the limited number of electrons you can put in the rat's brain. But it could be that the - that the cortex just has lots of frames in the movie. But to play them in the right order with the right timing, you need the information in the hippocampus. I mean that could be this link between the two.

FLATOW: 1-800-989-8255, talking with Nicholas Wade of The New York Times about this interesting study with rats in a maze and recording their brain activity. This study was done with rats, Nick, did the researchers think the results possibly will hold true for humans also?

Mr. WADE: Yes they too because they are the same. The rat brain and the human brain have the same structures and the (unintelligible) molecular biology level - that they work the same way. And it's very unlikely, with something so basic as memory, very likely evolution would have evolved to totally new mechanism for us. So almost certainly, these findings will be applicable to people.

FLATOW: Let's talk about this other study in the Nature Neuroscience that put the human nose to the test. Where they actually had their students, who always used in these tests, actually put their nose to the ground and sniff a trail?

Mr. WADE: Right. This is the craziest experiment this year, I think. You see, I think, the actual purpose to the experiment was to ask whether having two nostrils, just a little bit apart, it helps a mammal of any kind locate a smell. And we know the fact that we have two ears a little bit apart helps us locate a sound because the source of the sound because the sound arrives at slightly different times at the two ears.

So does the - do the two nostrils of the nose perform a similar function for us. I think the researchers thought first, of testing this on a dog, but dogs object to having one of their nostrils blocked because this is sort necessary procedure to make sure that you need two nostrils not one. So they chose more amenable experimental subject, which was the undergrads in the psychology department. I'm not sure how - quite how voluntarily their participation was.

FLATOW: And we've all been there. And so they laid down a chocolate scent on the ground, and they actually had them put their nose to sniff to the ground, and the surprising part is that they did as well as the dogs did.

Mr. WADE: Well, they - yeah. They think of the chocolate was the scent that would appeal to the undergraduates. They laid down a scent on the ground. The poor undergraduates were sort of blindfolded and their ears are blocked off too so it would just be nothing except smell would be coming into their brain.

They had to put their nose, literally, to the ground so there's not many things like crawling along with your nose in the ground and your bottom in the air. But they managed to do it. I mean, it wasn't quite as good as the dogs. I mean, they were really slow at it, and they couldn't sort of deviate very far from the track or they would lose the scent.

Where as if you would follow a dog doing the same thing the dog's senses are so good that it can run fast, deviate quite a long way from the track and still come back to it.

FLATOW: Right, I can see the next episode of “Monk” where (unintelligible) hologram following the trail. Did they discover that the stereo effect was necessary?

Mr. WADE: Well yes, that's right. So when they blocked off one nostril in their subjects, their performance degraded considerably.

FLATOW: It did?

Mr. WADE: Right.

FLATOW: I mean, is that because they've lost the sensor? How do they know it's the stereo? You know, you lose half of your sensory organ.

Mr. WADE: Well, that's a good point. And they could follow it some, I guess. But it just made them move a lot slower because I guess the nature has one nostril trying to track the scent.

FLATOW: If people want to watch, see this action, they can go to our Web site at We have the video, there's actual video done of this, where there's a link to the nature Web site showing one of the study participants following the scent.

Do you think there's a follow up to this study, Nick?

Mr. WADE: I mean, maybe it shows that our sense of smell is not as bad as we think it is. And we're always told how we've lost all these smell genes, and mammals have a sort of standard repertoire of genes with perfecting odor, and I think about half of ours have become inactivated. So I think we assume, therefore, that our sense of smell is very poor compared to a dogs.

And it probably is for all those odor genes we have lost. But in terms of just raw performance that you can compare us with some smell that we can still detect. Then as the experiment is shared our performances not so bad, and the sort of basic architecture the system is still there and operative even though we haven't used it for tracking for the last five million years or so.

FLATOW: We're talking with Nicholas Wade about smell this hour on TALK OF THE NATION: SCIENCE FRIDAY from NPR News.

Let's see, if we go to the phones. Get a call or two in for Nick. John in Overland Park, Kansas. Hi John.

JOHN (Caller): Hi.

FLATOW: Hi there.

JOHN: Thank you for taking my call. I'm a college student and I don't always get the most sleep. I was wondering on the first study, whether the amounts of sleep has any possible lengths that can be drawn from that first study to the amounts of data that can actually go into lumped memory, that's actually rewound and put onto the cortex.

FLATOW: Hmm, Good question.

Mr. WADE: We'll that's an very interesting question. I don't think they have done enough to sort of quantify exactly how much sleep you need to do what. But certainly, this establishes a very real function for sleep.

One the people had guessed beforehand that the purpose of sleep is to consolidate memories. But this I think is one of the first sort of clear, proofs that is indeed what's going on. And in just to me, it takes place not curing REM sleep is the sort of periods of from which subjects can always report dreams when you wake them up. This takes place during slow wave sleep, which I think much more sort of mysterious part of the sleeping process.

FLATOW: There was a study out there recently, I remember reading in the journals about people who have sleep apnea, having bad memories because they're not asleep, their memories, like mine, are just gone. But if they gave them this positive airway mask to use that within a matter of weeks, their memories improved dramatically. Are you familiar with that one at all? Did you come across that?

Mr. WADE: I'm not familiar with that study, but it's very interesting and it seems to point in the same direction that sleep is very necessary for whatever this purpose is. And clearly, when one needs to take the brain offline so they can perform this processing, and they didn't take it offline, then the system gets degraded?

FLATOW: Yeah, we've had Robert Stickgold come on, he's a sleep specialist and say that if you don't get a certain amount of - number of hours of sleep a night - and again, he's done it with his undergraduates. I think it's seven - was the magic number.

If you don't get seven hours of sleep a night. You don't consolidate a lot of talents like playing an instrument, new kinds of things you want to learn unless you get a lot - enough of that sleep at night. It's certainly amazing how well he shows these things to happen in his laboratory.

So anything from here, Nick, I know you're not the researcher on this one. Do you expect any follow up to the brain? I mean, is it possible to do this in people that we do not how to put electrodes in our heads, but maybe on the outside in some way - non-invasive way?

Mr. WADE: Right. I mean, we allow electrodes if you have to go the brain for some other reasons, for some - to treat seizures or something. And then, you have a few minutes in which you can put electrodes in. But I don't think you can get to these recordings from the exterior of the brain. You need to put the electrodes right in. I think there will be lots of follow-ups, and this is just - I hope we have this handle, which is you know looking for these distinctive fast rewinds.

I mean, this is a field of research they can take a long way. I think they can try and learn, exactly what is the nature of the dialogue between the hippocampus and the cortex. And this is a consolidation of memory as you're suggesting, or is it the cortex trying to build up models in understanding the outside world. I mean, the answers to these questions may be emerging for the first time.

FLATOW: I want to thank you Nick for taking time to - out of your busy schedule - to be with us today in a holiday season. Have a happy holiday to you.

Mr. WADE: The same to you. Great to talk with you. Bye-bye.

FLATOW: Your welcome. Nicholas Wade is a science reporter at the New York Times and he joins us today by phone. We're going to - we'll take our short break, switch gears, and come back and talk about more surprise and research.

This one is about a possible new malaria vaccine, which researcher's say, it actually packs the malaria inside the mosquito, and I think that there's a potential here that it works to wipe out malaria from whole geographic regions.

Stay with us, we'll be right back. Talk about it lots more.

I'm Ira Flatow. This is TALK OF THE NATION Science Friday.

Copyright © 2006 NPR. All rights reserved. Visit our website terms of use and permissions pages at for further information.

NPR transcripts are created on a rush deadline by Verb8tm, Inc., an NPR contractor, and produced using a proprietary transcription process developed with NPR. 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.