In Our Time - Memory

Episode Date: May 29, 2003

Melvyn Bragg and guests discuss the function and significance of memory. The great writer of remembrance, Marcel Proust, declared “We are able to find everything in our memory, which is like a dispensary or chemical laboratory in which chance steers our hand, sometimes to a soothing drug and sometimes to a dangerous poison”. The memory is vital to life and without it we could not be the people we are, but can it really contain the sum of all our experience? Is it a repository constantly mounting events waiting to be plucked to consciousness, or if not, then under what criteria are memories turfed out?With Martin Conway, Professor of Psychology at Durham University; Mike Kopelman, Professor of Neuropsychiatry at King's College London and St Thomas’ Hospital; Kim Graham, Senior Scientist at the Medical Research Council’s Cognition and Brain Sciences Unit.

This BBC podcast is supported by ads outside the UK. What makes people want to believe in aliens? I'm Tristan Redmond, one of the hosts of the Global Story podcast from the BBC. Donald Trump last week announced that he'd be releasing the US government's UFO files. So why the renewed interest in life out there? And what deeper spiritual meaning might people be searching for? Check out the global story. We are serious journalists on BBC.com.

or wherever you get your pods. Hello, the greatest writer about memory, in my opinion, Marcel Proust said, quote, we are able to find everything in our memory, which is like a dispensary or chemical laboratory, in which chance steers our hand sometimes to a soothing drug and sometimes to a dangerous poison. Can memory really contain the sum of all our experience?

Is it a repository of constantly mounting events waiting to be plucked to consciousness? Or if not, then under what criteria are memories erased? me to discuss the science of memory. Martin Conway, Professor of Psychology at Durham University, Kim Graham, senior scientist at the Medical Research Council's Cognition and Brain Sciences Unit, and Mark Copelman, Professor of Neuropsychiatry at King's College, London, and St. Thomas's Hospital.

Martin Conway, when we perceive or experience something which is going to become a memory, what are we doing? What's the initial process? I think if I could answer that completely, it would have solved most of the problems of the human mind. but we do have some evidence and there are some things that can be said. First of all, we don't really have any conscious control over what we encode.

We can try to remember things, but by and large it's out of our conscious attentional control. So there are a set of unconscious processes that operate to encode experience. And those processes are sensitive to the relevance of the experience, how it maps onto our goals, how it fits with our self-concerns, our self-image, and also how much effort we actually expend in comprehending the particular experience,

what we attend to and what we choose not to attend to. These are all factors which influence will eventually get stalled in long-term memory. To break that down into real simplicity as at this stage, are you saying that you can't force yourself to remember things? Because a lot of people would say, well, what was I doing at school all that time? Indeed. Indeed. I think we need to step back and confront another complexity before we can really get to that question. We need to think about what it is that's actually getting into memory.

If we're learning facts, then that's rather different from recalling, for example, episodes. Fact being the capital of China. Indeed. Episode being what? Well, an episode might be your visits to the Great Wall. Right. So let's say you remember, well, you don't consciously remember. you can bring to mind the fact that Paris is the capital of France.

It would be quite difficult to remember the learning episode in which you acquired that concept. And that's a type of information we call semantic memory, and that's hopefully the sort of knowledge ones acquiring in education, conceptual semantic knowledge, which is free of the context in which it was learnt. You implied in your opening statement that we didn't have much control of this. Now, before we move on, can you just... Can you go into that a bit more?

because when I said about what do people spend time at school for, it wasn't at all facetious. I mean, we're drilled into remembering our Moa, Massa Mat, six times six is 36, and so on and so forth. Now, so we can consciously say, well, teachers can tell it, you have to remember that, and we say we're better if we don't pass the exam and get kicked out, that kind of thing.

So what about that sort of memory? Well, it's quite interesting if we think about knowledge acquired from school. Let's take a Moa Massa mat. Can you remember what those words mean? Yes. Can you remember what the coefficient of... linear expansion is, which is a concept you're most certainly acquired in physics. No.

But you recognise it? I recognise the word. Really what we're trying to do when we're acquiring conceptual knowledge like that is build up a knowledge base in long-term memory that allows us to understand the area, that allows us to operate creatively with that knowledge. On the other hand, we may acquire knowledge not to a level, not in a form, which allows us to operate in that creative fashion with it. So you might know that the coefficient of linear expansion

is a concept from physics that you once learnt, but that's all you know about it. On the other hand, you might have a vast knowledge of Latin terms, which you can use and effectively construct creative sentences in Latin. Good, well, there's obviously a lot more to go. For my government, the theory, there's a theory, what people would generally say, that we know about two sorts of memory,

there's long-term memory and short-term memory. Can you develop that a little? What do they mean in your terms as a neuropsychiatrist and so on? Well, can we... Yeah, I'd just like to say, I'd like to defend the umbrella term memory. Sure. Because although there are all these different components within it,

and autobiographical memory is a different sort of knowledge from semantic memory, can you just keep people into what semantic is again? So we're all... Semantic is knowledge of words, concepts, and facts, as opposed to autobiographical memory for incidents and events. So sticking to Martin's idea, semantic is, there is a great wall of China,

I know that, and episodic or autobiographical is, I once went there and met a friend from Wicton, that sort of thing, yeah. But if you look at any, it's possible to take extremes and say these are very different, but if you take any particular aspect of memory, there probably is a continuum of different types of knowledge. And so that some things that we, might be asked to remember, may have an autobiographical

and a semantic component to how we remember it. Say, a picture of Princess Diana's car crash, well, you've got semantic knowledge about where that happened in what circumstances, but you also may help to retrieve it by autobiographical knowledge of what you were doing at the time and other things that help you remember about Diana. So I would defend the use of an umbrella term, then we break it down into different concepts within that umbrella term.

And one distinction is between autobiographical and semantic or episodic and semantic. Another distinction, usually used within episodic, is between short term and long term. And the answer to the Proust quotation is, I'm afraid, he was in this context wrong, that we don't remember everything, and we may get on to discussing why it would be a disavit. if we remembered everything a bit later. But the traditional view, at least since William James and probably before, was that an awful lot of our memories are held for a few seconds, or perhaps a bit longer, in a short-term store,

now often referred to as working memory. And then a lot of that is filtered out, is the traditional theory, and consolidated or selected for, Through some kind of process, some items are stored in a much more longer-term system, which has greater capacity and holds memories for a lot longer, but takes only a select amount of what is going through the short-term store. But the brain systems that are involved in short-term and working memory are different from the brain systems which are critical for,

laying down memories in the longer-term store. Those systems, the ones that lay information down more permanently, consist of circuitry in the sort of middle regions of the brain, so-called limbic diankophallic structures. And it's when those systems are damaged that amnesic patients have great difficulty in absorbing and acquiring new information. Kim Graham, could we take on this different systems idea? Can you discuss it more?

Is it possible you to talk about if we, you know, lifted up the lid of our head and saw the brain, can you say that bit's got to do with that sort of memory? And I can tell, because I put it on the microscope or other, and I can see this happening, and that bit is to do with the different sort of memory, and that's bit, is that possible or is this ridiculously simplistic? Well, I mean, it is rather simplistic. I mean, I think in terms of kind of the way we're talking about memory here, what we're really doing is differentiating between different types of memory that differ according to certain factors. So some of those might be kind of capacity for persistence,

such as a difference between short-term and long-term memory. So one's very kind of... One's like writing down on a jotting pad, and it only lasts for seconds. Another one can last for minutes to decades and so on. Or kind of the characteristics of mental experience, so sometimes memories are kind of consciously available and sometimes are not consciously available.

Now, given that we kind of distaste, distinguish between these memories according to kind of psychological factors. I mean, these are things that we think differentiate between them. And an interesting issue is whether in the brain, the brain respects these psychological factors that we think differentiate these types of memory. Now, in certain circumstances, they do. So you could, so Mike was talking about medial temporal lobe structures, the limbic dikephalic structures in the middle of the brain.

And we know that patients that have damage to those become profoundly amnesic. They have problems laying down new memories. We also know that patients who have damage to those structures can do short-term memory tasks so they can basically recall digits back to you or they can learn new skills such as learn how to backwards, mirror, draw and so on. And that tells you that the structures are involved in laying down new memories that are different from the structures involved in shorter, shorter term memory. But on the other hand, if we look at other regions in the brain such as the frontal lobe, which is right at what's called frontal lobes because it's at the front of the brain, If we look at the types of findings that you get from neuroimaging studies where you put healthy subjects into a scanner

and you ask them to perform memory tasks, what we often see is that the same types of regions in this area are activated during both working memory tasks and also during long-term memory tasks. So what that tells you is that these particular regions are maybe not storing memory per se. They're not storing short-term or long-term memories and they're sites of these memories.

But instead, what they're involved in is the reactivation of the memories and that they're playing some kind of retrieval role and that retrieval role is common across different types of memory. So just kind of in summary, in certain circumstances, there are regions which are differentially involved in the different types of memory that we identify, but in other circumstances there's brain regions such as the frontal lobes which may be in common

to these different types of memory. Have you any views, I'll ask you Mike briefly, on the relationship between conscious and things you take in consciously, like I want to learn the six times table. And things you're taking unconsciously, like all the people around me in the class at that time were wearing dark blue blazers. And so, or whatever it was.

Is one more powerful than the other? Is one going to be more lasting than the other? Do we have any sort of fix on that? Well, actually, if you look around everybody in the rooms wearing dark blue blazers, you're probably taking that in consciously. Whereas when you've learned the six times table or whatever, there is a point at which you will produce the response,

perhaps not completely unconsciously, but in a rather automatized habit-forming way. But there certainly are memories that are held at a less conscious level of awareness. And these are sorts of things like perceptual motor skills. Once you've ridden a bicycle, 30 years later, if you haven't ridden a bicycle, you get on it and you can still ride. Whistling's a good example as well. Yes, and musical skills.

And this goes into the part of the memory that will stay forever. They just locks in somewhere. That's called a... Well, it may not stay forever, but it stays for a long time. And there are other kinds of less conscious forms of memory, conditioned responses, but also the so-called priming phenomenon, which is when you've seen something before, you may not remember that you've seen it before,

but it speeds up or facilitates a subsequent response. Mike said earlier that Proust was wrong that you couldn't remember everything and this cannot go unchallenged but I don't suppose I get any supporters with you a lot but still do you have a comment on that? Why can't why? Well, would you respond to that

because there's a feeling that everything's around somewhere isn't there? It's part of our it's part of what Freud and others intimated that if you ferret around

someone's those memories will come back and if they don't stay, where do they go? I mean, do these things evaporate and do they evaporate inside the skull? Is that way you get headaches? I mean, anyway, there you go. So why can't you... Why cannot we remember everything?

And is that a good thing? Well, I mean, I entirely concur with Mike. I mean, it's very clear that we can't remember everything that we experience. And I think there's very good reasons for why we don't want to do that. So I think the types of kind of forgetting errors that we make are actually kind of by-products, of an adaptive system, the kind of the costs of the benefits, as they say. And so, I mean, a kind of example of that is that if you think about, you know,

kind of walking down a very busy London street or so on, there's so much information and so much kind of contextual detail that you're processing. If we actually remembered all that contextual detail, then we'd end up kind of filling up our system and it would all get very confusing and so on. So we have systems which basically want to encode unique events. It's a fast kind of learning system. We want to hold on to those events. The other thing we want to be able to do is we want to be able to acquire just information about events.

So we talked about this before about building up a database of knowledge about the world. And one of the ways of doing that is basically across repeated events, pulling apart the things that are common about them, and then basically learning information about these circumstances. One of the reasons why we may actually not kind of concentrate on these contextual details, but instead attempt to kind of pull apart commonalities is so that we can build up this other system. So we have this kind of fast learning system

which encodes episodes very rapidly. But we also need to have a slow learning system where we build up knowledge about the world and we do that slowly and carefully so that we don't cause interference with concepts that we learn. And that may well be one of the reasons why we actually kind of have forgetting. So you are actually picturing the brand to my mind

is almost like a sort of railway track system of things passing there, the main lines, branch lines that are get banned by beeching and that sort of thing. Hopefully it's a bit more efficient. Yeah. What sort of cues does memory use, Martin Connery? Well, it uses lots of different types of cues. But if I can just dwell on this point just a little bit longer about selection of information.

I mean, consider this. Supposing when you remembered an experience, you remembered it in such detail that the remembering took as long as the actual experience. Clearly, that would be dysfunctional. Also, I mean, I'm just, it's me being anarchy. Please go on, but I don't see much particularly dysfunctionally. It's one way to spend a life, really. Well, that's exactly what Proust did, didn't he locked himself away in his flat in Paris and didn't come out for 30 years.

He said by the gentleman on my left. Okay, well, I'll mention one more fact to do with it. There was a very famous person, not patient, person, a neimanist called S, who was studied by the Russian neuropsychologist Lou Rhea in the earlier part of the 20th century. Now, S had an absolutely phenomenal memory. He could remember 20 pages of random numbers that he'd seen once 40 years earlier.

And he was exactly right every time. Now, he actually ended up suffering quite severe mental illness, partly related to his memory problems. He probably also had some personality problems anyway, if you could spend your life memorising 20 pages of random numbers, there must be something wrong. But there have been other examples, of people who've sort of trained their memories up

to be these incredibly effective. efficient devices for remembering essentially meaningless material, and they usually end up having some sorts of mental problems later on. So there's some indication there that it's probably dysfunctional to do that. But memory is a device which allows us to operate on the world. It allows us to have a self. It allows us to communicate. And it's got to be a device which supports inference and isn't a very slow, informationally wieldy system. Can I go back to Mike? As you have... had this huge assault, which I expect a lot of letters for,

and he's probably going to follow you for the rest of your life on Proust. But he did say, he was very precise about the Madeline moment, he said the sight of the little Madeline had recalled nothing to our mind before I tasted it. Now that brings up the question I want to ask, is there a hierarchy among the senses that tells us, that queues memory? I mean, do we remember more because of what we see, because of what we hear, because of what taste, touch? is there any hierarchy? Right.

Well, there's certainly a hierarchy within memory, but can I point out, I quote it to Prousta, with approval in a recent review. I'll give you an anecdote which partly supports what you and he are saying, but the taste is interesting because what this guy, S, that Martin was referring to, did, was he had powerful synesthesia

When he heard a word, he would say that he not only heard it, but he would imagine a colour associated with that word, and he would taste the item related to that word. He had very powerful sensory input, which he used to help process and retain his memories. Can we talk about this hierarchy? And then can we talk about storage? And does sleep help storage?

Well, I mean, I think that there is a hierarchy. I think that vision is very, very important to us. And I think most of the memories that we end up, kind of, you're walking down the street and you see something, and that cues an event from 20 years ago or so on. And I think that... Can we register to this. Is vision more powerful?

Is there a way that it is registered more powerful in the brain? There's been an exciting talk of neurons firing. and so it's nice to have these subversive volcanoes inside these placid skulls. When you see this, when you look at the thing, is vision causing more activity than hearing? Can you prove it, as it were? I think we just don't know the answer to that, and I think there needs to be a lot more experimental work

in order to address that type of issue. But I personally think that vision is a very, very important cue for us, and I think that hearing as well, or smell, there's often occasions where you may retrieve an autobiographical vent from your past on the basis of a particular smell or hearing, I think taste is probably less, that's less likely, less likely to happen. But certainly, I think the important thing about this is in all the memory tasks that we ever do, either in the laboratory or the memory that we're doing as we, you know,

wander around and interact in our environment, the more cues we have, the more information to help reactivate the memory, the more likely we're going to be able to do that. So that we basically go around using all the information available to us, if that's vision, if it's hearing, if it's sound, if it's knowledge about the environment to help us kind of solve the problem of interacting with it and memory is just one part of that. So I think

there's a combination of cues often that ends up in the situation where suddenly we remember an event from the past. Yeah, I think I'd want to just make a point here about some patients that have recently come to light who suffer brain damage to the back of the brain, the occipital lobes, and they lose the

ability to generate visual images. It turns out these patients often become amnesic for their life prior to their accident. And this is because a lot of our memories, by far the vast majority, 99% are encoded in terms of visual images. And if you lose the ability to generate visual images, then it become amnesic. Interestingly, you still retain conceptual knowledge of your life. So you might remember you went to St. Mary's grammar school.

But if you've suffered this brain damage, you won't be able to generate any images of specific episodes from that. So I think imagery probably does really come top of the list when we're talking about the sorts of information that memories encoded in and retrieved in. Yes, that's particularly true of right hemisphere damage and particularly for visual cues for episodic memories.

You can get the opposite way around. I had a patient who's a professional artist. He got damaged to the left side of his brain so he has problems with his semantic memory in naming things, in verbal memories and in reading, but he was a professional artist, and his artistic skills were still intact, are still intact. And in fact, I used to have an office facing Big Ben,

and he would come into the office every day and for five or six years and say, you're going to ask me the name that bloody clocked her. He couldn't remember, and yet within weeks of his illness, He did beautiful pictures of Big Ben and he later became rather obsessed as part of his rehabilitation process

in doing paintings of his own MRI scan, his own brain with the damage which became more or more abstract as the years went on as with his sort of self-generated rehabilitation. Martin, I wrote something that you wrote about the use of objects in queuing memory. Could you just address that for a few moments?

Well, we often use just if we're studying in a laboratory people's autobiographical memories, we'll just give them words that name objects, chair. Tell me a memory that reminds you of. What tends to happen, which is quite interesting, is that people usually bring to mind an image of an object in their current environment, and then they use that as a cue to probe memory with. And this has led us to the view that memory is sort of extended out, all the objects around us in our working environment,

and our home environment are kind of imbibed, imbuted with memories. And if we lose them, if we break them, if we move house, for example, then we, in a sense, lose a bit of our self, a bit of our memory that's been projected out into objects in the world. You've implied about exploring memory. If you use the word chair to somebody, can you keep going on with chair until it goes back to the first high chair they sat in? Is that something that can be driven through and uncover?

Do you really want to talk about childhood memories? Do you really want to talk about childhood? memories. I'm not driving to it, but it's something to do with it, doesn't it? It's a great omission so far, but we can't do everything in a single programme, but it is an omission which actually was the territory moved into most particularly by Freud. Yes, you can get people to recall their earliest memories. There is a period of child in amnesia. People rarely remember much below the age of about

three, three and a half years, which is curious, if not for the fact that if you take children who are that age, they actually have lots of memories. So it's not that the memories aren't there to be remembered, it's just that the adult doesn't remember them. We've been doing some work recently which might interest you on asking people to recall their very earliest memories, the very first memories of emotional experiences. So what's your first memory of feeling frightened? And what's interesting about that is people normally only get back to when they're about six or seven years old. Now you might want to say, well, maybe they didn't have the word, but in fact they did have the word because we have age

of acquisition data on the words we use. And we know that they had the word earlier than their memory. So there is something curious about childhood memory. I think everyone's aware of that, an adult's memory for their childhood, I mean. We don't really have completely decisive data on that. And yet there's a school of thought which believes

that the experiences we have up to seven are themselves decisive. Absolutely. And is that maybe why we don't remember them so well? Well, that may well be the case. I personally think, if you're my view, I have a very strong view that memory is a motivational

system and we remember things that relate to our goals. And I think the goals of a six, five, four, three-year-old are really very, very different from those of the older child and certainly the adult. And it's perhaps for that reason the goal incompatibility between the goals we currently have and the ones we had as an infant and a young child that prevents us from getting at the memories. I just want to go back to sleep with Kim and then come back. You didn't mention sleep. And in sleep, is that an enforcement of memory, it is unraveling of memory. There was some suggestion somewhere and some of the thing I've been reading for this programme that we had to sleep because of memory. Well, I mean, that's quite a, it's quite a controversial issue, but there is a school of thought that sleep does play a very important role in consolidating our memories.

So, I mean, we've talked about kind of encoding and storing and retrieving memories. We haven't really touched upon consolidation, but the process of consolidation is thought to be about where you've encoded a memory and there are changes. that can occur to that memory long after that stimulus has kind of disappeared so that we may make more kind of permanent or strengthening of these memory traces over time. And what's likely to determine that will be about the emotional salience or the importance of that event and so on. And it's kind of a filtering out of our experiences.

It's a way of the brain trying to work out what's important to hold on to and what's not. Now, one of the schools of thought is that in order to consolidate these memories and to make them stronger and to lay them down in a more permanent form, we reactivate them. And obviously we can do that via conscious reactivation. Someone can ask you what you did yesterday, or you want to tell someone about the dinner that you were at last night or so on. But the other view is that sleep plays a role in this.

And there are some data in animals which suggests that the neurons, the pattern of neuron firing that you get while they're kind of a rat is running around a room is replicated at night when they're asleep. And some people have argued that that means that what's happening is that these animals are basically replaying these memories and that that's a way in which we can consolidate our experiences. It used to be said that REM sleep, rapid eye movement sleep, which is when you dream, was important in consolidating memories.

More recently, some people have argued that, slow wave sleep, which is when you're not dreaming, is important in consolidating memories. My team have done some research on this and the findings are only preliminary, but to date we've found no difference between them and indeed not much evidence that it's important in consolidating memories. I think this is still very much an open question, but most of the research has been done in animals and there is a lack of good. human data on this topic. So you don't expect to find that actually we need sleep

because we need memories to be sorted out. Well, there is a little bit of evidence that if you learn something just before you go to bed, it's better remembered the next day than if you learn it and then go and do other things. And I remember reading this just before university finals and got some personal evidence that it might be working. Well, thank you very much to Martin Conway, Mike Copleman and Kim Graham

and next week we'll be talking about the Lunar Society. Thanks for listening.