2006 LECTURE SERIES
Watching the brain at work:
Imaging the formation and retrieval of memories
Dr. Michael D. Rugg
University of California, Irvine
Tuesday, March 21, 2006
--which is a vote by the presentation of that word. Some of this neural activity will be due to recognizing this word and having to respond to it, and presumably some of this neural activity may reflect something that’s going on that causes this word to be encoded into memory. I can do that for a whole series of words. I’m illustrating four words here in our experiments. In general, of course, we use many, many more words, but illustratively here, we have four words. So you lie in the scanner, and you do something with these four words, one at a time. Maybe you decide whether they represent living or nonliving objects. There is then a delay, when we take you out of the scanner, much to your relief usually, let you go to the bathroom, things like that. And then, there’s a surprise memory test.
In most of our experiments--this is an important point. It’s a little digression, but it’s an important point. In most of our experiments, people do not know that we’re going to test their memories at the time that we actually present them with the information that we’re later going to test their memories on. There’s a reason for this. If you think about it, most of your time is spent committing things to memory without trying to. You do not go round the world trying to remember things. If I said to you, for example, how many of you in this audience can remember what you had for breakfast this morning--I won’t ask you to put your hands up because it’s embarrassing for those of you who have to leave them down, but I would guess that at least 90% of you could remember what you had for breakfast this morning. Now, when you were eating breakfast this morning, I very much doubt that you said, oh golly, I must remember what I ate for breakfast this morning because Dr. Rugg is going to ask me tonight.
It’s quite unusual for us to actually be effortfully trying to remember things. We remember things as part and parcel of our interaction with the world as we deal with it online. It’s only in relatively artificial situations like tests, either at school or at college, that we actually effortfully and deliberately try to commit things to memory. By using this kind of a method, we’re trying to capture what happens in real life. Memories occur incidentally. They don’t occur because we try to make them happen.
We give people a surprise memory test. We can then look at the words or identify the words that we presented to people, and we can classify them as according to whether people could remember those words or whether they forgot them on our memory test. Our assumption is that those words that are later remembered are words that were better encoded when people were in the scanner than those words that were later forgotten. We now have words that were well encoded and words that were badly encoded or relatively badly encoded. We can now go back to the brain activity, that we recorded in the scanner, and we can average together the brain responses that were elicited by those words that are later remembered, and we can compare it with the brain responses that were elicited by those words that were later forgotten. We can ask where in the brain does activity differ at the time that people are looking at these words as a function of whether the words are later remembered or later forgotten. In effect, we’re asking the question, where in the brain does activity predict whether a word will be later remembered or later forgotten? And that is our index of encoding.
Let’s turn that into a real experiment. People are lying in the scanner. They see a set of words. And in this particular experiment, they were simply asked to judge whether those words represented a living or a nonliving thing. This is a meaning-based task that we’re asking people to do. They come out of the scanner, and after 15 minutes we test their memory for these words, and we can then separate the words that are later remembered from the words that are later forgotten. We can then look in the brain and say where do we find activity that predicts successful memory?
I’m going to show some data here. Here are some data. This is now a picture again of a random subject. This is just one of the subjects in our experiment. We had, I think, 16 in this experiment. We’re looking at the left hemisphere, the front, the back. What we see is that, in the cerebral cortex, there is a concentration of activity here in the left frontal cortex, the so-called inferior frontal gyrus. This tells us a number of things. The first thing that it tells us--and this is actually quite important in its own right--is that activity in the brain that determines whether you’re going to remember an event or not has already started with the presentation of that event. In other words, things that occur within just a few hundred milliseconds of the presentation of an event are already helping to determine whether you’ll remember that thing 15, 20 minutes later, in some of our experiments, 48 hours later. We see activity here.
Also, in this and in many of our experiments, we see activity in the hippocampus. Now we’re looking through the brain from the back. This is the medial temporal lobe. Here is the hippocampus. Sure enough, the hippocampus is more active for a word that you’re later going to remember than a word that you’re later going to forget. This tells us that the hippocampus, whatever else it does, seems to play a crucial role in the initial encoding of information. It’s more interested in words that you can later remember than those that you later forget.
Now, in looking at these data, which have been replicated and are replications of other people’s work, we were struck by the fact that this region of the brain in the cortex here, which we see here as a memory encoding effect, is actually a region of the brain that has been shown in many studies to be very interested in the processing of words for meaning generally. In other words, if we weren’t doing a memory experiment, and we just asked where in the brain do we find activity that goes with processing a word for meaning compared to not doing so, this area would light up. This led us to think that maybe what we’re looking at here is not some region of the cortex which is somehow dedicated to encoding memories, but rather, what happens is that regions of the cortex that are recruited to process something online--we’re asking people to make these meaning-based judgments--is also contributing to the encoding of that information. In other words, what you remember is a byproduct of what you do online.
We tried to test that experiment--that idea. I’m just going to give one example of how we did that. We did an experiment where we compared the activity that we saw using two different kinds of encoding tasks. What we’re seeing here now is data from another experiment, where again we asked people to make a living/nonliving judgment on a set of words, and once again we see activity in this inferior frontal gyrus. We also see a bit of activity up here in a more sort of anterior region of the frontal cortex. But basically, we replicate our original findings.
What we did was to have another task in the scanner. Instead of asking people to make meaning-based judgments on words, we asked them to make a sound-based judgment. And, in fact, it’s a killer of a task, this. We asked people to decide whether the number of syllables in a word was odd or even. Now people don’t pay attention to the meaning of a word, they haven’t--I’ll tell you, this is a murderous task to do--they don’t pay attention to the meaning of the word. They’re working flat out to try to figure out the sound structure of the word. And we know that this particular task depends upon regions of the brain in the back here, in the so-called parietal regions, and we know that this particular task, the syllable judgment task, does not depend on these frontal regions. The reason that we chose the syllable judgment task is because we knew, ahead of time, that to do this task, you have to recruit regions at the back of the brain here.
What happens? Well, this is what happens. We find now that the regions of the brain that predict whether a word, when you’re judging it for syllables, the bits of the brain that predict whether that word will be later remembered, are in the back of the cortex here. They do not overlap with the regions of cortex that predict later memory in our animacy task, in our meaning-based task. You can see here essentially no overlap. We can move around, in the cerebral cortex, where the activity is that predicts later memory simply by manipulating what people do with the things that we’re hoping that they’re going to remember.
There is nowhere in the cerebral cortex that is dedicated to remembering things. This makes good sense. As I said before, you do not go round the world trying to remember things. And, in fact, your cerebral cortex is an extremely important resource for you. You need it to react to, to identify, and to appropriately respond to events as they occur in real time. It would be perhaps slightly peculiar to dedicate this very valuable resource to trying to remember things. And, in fact, if you think about it, if you were, I don’t know, a hunter-gatherer of 40,000 years ago, it would not be a good idea if you were confronted by a snake or a tiger or something, as you went round hunting, to have to say golly, I better remember where I saw that snake or that tiger. You’re better off dealing with that emergency as it occurs, and your brain is built so that you will remember that later, even though you didn’t try to do so. If you devoted processing resources to trying to remember that thing, you wouldn’t be around to have to remember it later basically.
Now, these sorts of findings, along with theoretical ideas, have led to some specific ideas or specific hypotheses about how episodic memories are encoded in the brain. And, in fact, they’ve also led to ideas about what it is that gives our episodic retrieval this content that it has. One of the things that’s striking about retrieving an episodic memory is that you bring back phenomenally part of what it was that you actually experienced. When you remember having breakfast this morning, there’s a sense in which you’re back having breakfast this morning. How do we get this kind of content to our memories? Well, there are some specific ideas about how that’s done that come out of this kind of experimentation.
I’m going to do a little cartoon show now, that explains how this idea works. We’re looking now at a cartoon of the brain. This is the surface of the left hemisphere, and this is the inside of the brain, and this red oval here is the hippocampus. Imagine that an event comes along, and we encode this event in memory. What happens? Well, the idea is that the event comes along, and we process that event. In processing that event, we evoke particularly strong activity in particular regions of the cerebral cortex. I might get activity here, here, and here. The idea is that what happens is that the hippocampus records and forms a representation of this pattern of activity, so that when the event is over, what we have now in the hippocampus is a representation of the pattern of activity that was evoked by this event.
Now, this is a very important point. What is being stored in the brain is not some kind of literal, camera-like copy of what happened to us. What is stored and represents that event is the pattern of activity in the cortex that was evoked by the event as we processed it and dealt with it in real time, and what the hippocampus is doing is storing that pattern of activity. Another event comes along. It evokes a different pattern of activity. The hippocampus records that activity. And when the event goes away, we now have a record of that activity too. The idea is that, as we go round the world, as events occur, they engender a pattern of activity in the cerebral cortex. That activity is encoded into the hippocampus as a memory representation.
What happens when we retrieve information? The idea is that when we retrieve an episodic memory, what happens is that you get a reactivation of the hippocampal representation, and this feeds back to the cerebral cortex and reinstates in the cortex the pattern of activity that occurred when the event was originally experienced. In other words, the reason why we can reexperience an event when we remember it is because we are, in some sense, reinstating in the brain the pattern of activity that was there when the event was originally experienced.
The second event comes along. The same thing happens. When we experience two different events, we have these--we experience these as two different events because different patterns of activity in the cerebral cortex are evoked by these two events. When we remember two events, the reason that these memories have different contents is because we’re reinstating different patterns of activity in the cortex. The idea is that there’s congruence between the activity that we get in real time and the activity that we reinstate when we’re recalling these events.
Now, the way I’ve described this sounds too good to be true, and it is too good to be true. We all know that memory is not perfect. The way I’ve described it suggests that we should always remember everything perfectly. The hippocampus just records everything, it plays it back. How can memory ever go wrong? Well, memory can go wrong for many reasons, and what probably happens in real life is something rather different. If you imagine an event that evokes this pattern of cortical activity as you experience it, what happens when you remember it is probably that you get a very degraded version of the original cortical activity back. And, in fact, you may even get spurious activity in regions that were never activated when the event originally occurred. To the degree that we get a lack of correspondence between what actually happened originally and what we bring back, of course, memories will be inaccurate or they will lose detail.
There are many reasons why this can happen. It can happen because the hippocampus is not perfect at separating the representation of different events, particularly if those events are very similar to each other. It can happen because over time the representations degrade either through simply the function of time or disease or the influence of other events that come in. There are many different ways in which things can go wrong. The bottom line is that this idea gives plenty of room for why memories don’t always work, but the basic gist of the idea seems to be what is going on.
Now, if this is true, it makes a clear prediction, and the prediction is, to the degree that we recollect two different kinds of memory that have different kinds of content to them, then we should find patterns of cortical activity that differ as a function of what we’re recollecting. In other words, we can do an experiment to test these ideas. We can go back from these kind of cartoon-like notions, and we can actually do an experiment and ask whether it’s true. Is it true that when you remember different kinds of things, different patterns of activity occur in your brain? We can test this hypothesis.
I want to describe now an experiment that does that. This experiment is based on other people’s work initially, that demonstrated something quite important for our experiment. We’re looking here at sections of the brain taken as if we were looking down from the top here. It’s these sorts of slices that you see in this cartoon here. The eyes are here. The important point about these data is that--they’re not our data at all, they’re from two different laboratories, but what they demonstrate is that when people process words, visually presented words in real time, there are specific regions at the higher visual cortex, so-called fusiform cortex, which are specifically activated by the demand to process visual words. By the same token, there are other regions of this fusiform cortex, which are different from the word region, that are activated in real time when you have to process pictures. In other words, we know that if you’re having to pay attention to process words versus pictures of objects, then we have characteristic patterns of focal activity in these nearby regions of the fusiform cortex.
Our experiment takes advantage of these prior observations. What we did was as follows. We had people look a series of words and pictures, a random series of words and pictures. And what people had to do in this experiment was simply say whether or not the word corresponded to a living thing or whether the thing wasn’t living. Penguins are living, barbecues are not. Balloons are not living, babies are living, etc. People just classified each of these stimuli, as they came up, as living or nonliving. Then, while we scanned them, we gave them a memory test. And the memory test consisted of the presentation of just words. People didn’t see pictures at test. Some of these words are words that they’ve never seen before; some of these words correspond to words that had been seen in the study phase; and some of these words correspond--I’m desperately looking for one as I say this--some of these words correspond to pictures.
People see the series of words whilst we are scanning their brains. For each word, they had to make one of three responses, one of three judgments. If they saw a word and they remembered seeing the word at study, and they could remember specifically the details of the presentation of that word or object--in other words, if you see tree here, and you remember--oh, sorry--if you see windmill, for example, and you say, oh yes--people think, oh yes, I definitely remember seeing the word windmill at study, then they had to tell us that they remembered that word, meaning that they could recollect specific details of when I first saw this. So apple comes up. I remember seeing an apple at study. Then I’d say remember.
By contrast, if they see a word and they just have a feeling that this is a word that they saw earlier, but they can remember nothing about the specific details of it, then we asked them to say this word feels familiar. I know I saw this, but I can’t recollect anything about the experience. And the point about this distinction is that we would regard this as being true episodic retrieval. You don’t just have a vague feeling of familiarity about something. You can remember specific details about when and where this thing happened. Here, you just vaguely know the thing happened. If you don’t distinguish between these two kinds of memory, you get into trouble. And, of course, people had to classify items they’d never seen before as new.
We have two critical questions in this experiment. The first is where in the brain is activity greater for recollected items than for items that are only familiar? In other words, are there regions of the brain that are particularly interested in detailed episodic recollection, when something comes back that has content to it as opposed to some vague feeling of familiarity? And obviously, what we’re hoping to see is the hippocampus. The idea is that the hippocampus is particularly important for detailed recollection. Secondly, we’re asking the question, where does activity differentiate recollected words and recollected pictures? In other words, if we look at the activity evoked by these items which correspond to words as opposed to pictures, where do we find differences in the brain, depending whether you’re recollecting a word or recollecting a picture? Those are our two questions.
Well, as you can imagine, I wouldn’t be presenting this if we didn’t find something interesting. It’s unusual for people to have a buildup like that and then say the experiment didn’t work, on to the next one. It has been known. Anyway, first off, if we simply ask, where do we find more activity when people look at a test item and say I recollect something about this test item, as opposed to this test item is just familiar to me? Among other areas, we get the hippocampus--in this case, on the right side of the brain, we get a nice activation within the hippocampal region. We predict that. It’s very nice to see it. More critical, though, for our hypothesis is what happens when we contrast the recollection of pictures with the recollection of words. This is what see.
These are brain regions which are more active when people are recollecting words than recollecting pictures. Don’t forget they’re recollecting on both occasions. This isn’t contrasting good memory and bad memory. This is contrasting good memory for words and good memory for pictures. And we see activity in fusiform cortex here. And I’ve circled this particular region here where we get more activity when people are recollecting a word than a picture. Over here we see a region where people are recollecting--we get more activity when people recollect a picture than when they recollect a word.
Again, you can see where this is going, I trust. These regions are embarrassingly overlapping with the regions that other people have demonstrated to be specifically interested in the real time processing of words and the real time processing of pictures. In other words, when you recollect a word versus recollect a picture, part of doing that is reinstating activity in the bits of the brain that were active when you actually processed these things in the first place.
What do we conclude from all of this? First off, we conclude that encoding, as I’ve already said, is a by-product of experience. It depends on the interaction between functionally specialized regions of cortex and the hippocampus. Any bit of the cerebral cortex can contribute to encoding. It just depends on what you’re doing with the event when it comes in. What you remember is a by-product of what you do to deal with that event in real time. Memories are stored in terms of the pattern of brain activity that’s evoked by an event as it’s experienced. There is nothing that corresponds to a film, a video film or something in the head, that takes a literal picture of what’s going on. What we store in our memories are patterns of activity that reflect what we did with something as it happened. I seem be laboring that point, but if you don’t remember anything else from this lecture, it’s a good thing to remember.
And finally, retrieval involves the reinstatement of the patterns of brain activity that were evoked by the event when it was experienced. The contents of our memories are determined by the nature of this activity. What you remember depends on what you did in the first place. Okay, so that’s the sort of basic bit about what I wanted to talk about. In the final sort of 15 minutes of my presentation, I want now to turn to two slightly more applied aspects of memory, human memory, and show how we can use these same methods and these sorts of ideas to understand things that are of great interest, I think, to all of us.
The first thing is of especial interest to me, as I grow older, and that is the relationship between memory function and aging. This pretty well illustrates what it is that we’re dealing with here. These data are from a very large-scale study of, I think, virtually the whole population of a Swedish city. This is a heroic study. It’s a combination of longitudinal study that followed people over time along with memory testing that was done of people of different age groups. Along here is age. So here are people in their late 30s, and here are people in their late 70s. Up here, as you go up these axis, memory performance gets better and better.
Now, there are a couple of things to say about this. The first is things are not all bad news. You’ll notice, for example, that short-term memory here--that is the ability, for example, to look in a phone directory, remember a phone number, and hold onto it while you dial the number--short-term memory is pretty unaffected by aging. You carry on doing that pretty well. Likewise, our measure of implicit memory, priming, shows no systematic relationship with age at all. In the case of semantic memory, you’ll notice that, in fact, you carry on increasing your semantic memory function at least until your late 50s. I like to think it actually goes like in reality. In other words, it’s true what they say. As you get older, you get wiser, basically. The trouble is you don’t remember how you got wiser.
If you look at episodic memory, you’ll see that, unlike these other forms of memory, there is an early and a steady decline in episodic memory performance with age. I’m sorry. That’s the way it is. I myself can no longer be a subject in my own experiments. I can’t be a pilot subject in the experiments that we run in our young undergraduates because I can’t do the tasks anymore. I’m now a pilot subject for our older subjects in our aging studies. It just happens. We would very much like to know what is going on in people’s brains as they grow older that leads to this decline in episodic memory function. We don’t just want to know that because we’re curious about what’s happening to us, but understanding the changes that occur in the brain that cause these changes in memory function is going to be crucial to an understanding of what goes even worse wrong or how things go wrong even more in people with neurodegenerative diseases, such as Alzheimer’s disease and other diseases of old age, which have an impact on memory function.
I want to give an example just of one experiment that tries to get at this issue. The findings I’m going to present are actually illustrative and pretty typical of what’s now being found fairly routinely in these sorts of experiments. They’re slightly surprising. They’re particularly surprising if you think that one of the problems as you get older is that you don’t use your brain enough, that things sort of fade away, so to speak. Nothing could be further from the truth, as you will see.
In this experiment--we have a bunch of pictures--what we asked our subjects to do was to pay attention to a little cue. Here is an asterisk and this is a plus sign. They had to pay attention to a little cue that came up just before the picture. If the cue was an asterisk, people had to make a living/nonliving judgment on the picture. You have to say, for example, that this is a living thing. If the cue was a little cross, then you have to say whether or not this object would fit in a shoebox. A barbecue does not fit in a shoebox. A European badger, if you squeeze it and have a very big shoebox, it might just get in.
Now, the point about this is not just to entertain ourselves and our subjects. We’re actually asking people to do two different kinds of things with these pictures, and there’s a point to this. When we tested people’s memories, and this is when we’re scanning them, we asked--we showed people the pictures they’d seen before and we showed them new pictures. When they saw a picture they’d seen before, they didn’t just say yes, I’ve seen that picture before. They had to tell us whether they remembered the judgment that they made on that picture. So if you saw this badger, it wasn’t enough to say yes, I’ve seen this badger before. To be correct in this experiment, you had to say I’ve seen this badger before and I made a size judgment on it; or I saw this penguin before and I made a living/nonliving judgment on it.
The reason for this is again we want focus in on memory for episodic detail. We’re not interested in whether people have some vague notion they’ve seen this before. We want people to remember what they did with it, and that tells us that they’ve got a clear episodic memory of the original episode. Now, there’s a trick to this experiment that I’m not going to go into, but let me just say right now that we arranged this experiment in such a way that the performance of our older subjects and our young subjects--our young subjects were in their 20s; our older subjects were in their 60s and 70s--the performance of our subjects was matched. We set this experiment up so this memory test was equally hard. The data I’m going to show you are not muddled up by or confounded by the fact that older people were struggling to do the experiment and younger people were cruising through it. They were performing at the same level. If I tell you how we did that, you wouldn’t be so pleased.
There are two questions. Where in the brain is activity greater when you recollect a picture rather than judge a picture as being new? And more importantly, how do these differences vary with age? This is the pattern of activity we see in our young subjects, and this is pretty typical of what you see in these experiments. These are the regions of the brain or the regions of the cortex which are more active when you successfully recollect a picture compared to when you say this is a picture I haven’t seen before. You see a lot of activity, particularly in the left cerebral hemisphere, frontal cortex, parietal cortex. Nothing is news here. We’ve seen this pattern many times.
These are our older subjects. What you see is essentially recruitment of pretty well the same brain regions, but to a greater extent. If you statistically compare these patterns of activity, these are the regions where older people show more activity when they’re recollecting pictures than young people do. There was not a single region of the brain where our older people showed less activity than our younger people. Every difference that we found between our older and our younger people took the form of more activity in the older group.
Now, for people who think the trouble with old people is they don’t use their brains enough, I think these data show that that is not the case. I have to say, that this is something that surprised me tremendously, but it is the result that is coming out now consistently when people do experiments this way. In other words, I would have to say when they do them properly. In performing at the same level as younger people, older individuals activate the same brain regions, but they do so to a greater extent, and we’ve found this in other experiments. Other people have found it too.
Here is the $64,000 question. Is this helpful? Is this increased activation compensation for the decline in the efficiency with which you can remember things as you get older? In other words, is the older brain putting more resources into doing things, so as to kind of compensate for what would otherwise be even worse performance? Alternatively, is it part of the problem? Are we seeing here the fact that as we grow older, we’re less able to optimally allocate resources of the brain to a particular problem? In other words, we throw everything at it, if you like. Less is more--more is less, sorry. I’m getting old. More is less. We simply don’t know. At the moment we simply have a correlation, and we don’t know which way it goes. If you ever invite me to give a talk again in the next few years, I will tell you the answer to this, because we’re actually addressing this question right now in collaboration with colleagues in the Alzheimer’s disease research center here at UCI. But right now we don’t know the answer to this question.
The last thing I want to talk about, just for the last five minutes, is memory for emotional events. Everything I’ve told you so far involves experimental material that is deliberately intended to be emotionally bland and not to be very variable in its emotional significance. There’s a reason for this. We know that emotionally arousing events lead to the formation of stronger and more lasting memories than nonemotional events. Therefore, when we’re doing the kinds of experiments that I’ve been describing, we want to reduce the variability that people have in their memory due to this kind of a factor. But, of course, in everyday life, many of the things that we remember and many of our enduring memories, if not all of them, have tremendous emotional significance, either positive or negative. Of course, in people with, for example, posttraumatic stress disorder, their problem is not that they remember too much, it’s that what they remember is traumatic and distressful to them.
The question is what can we learn, from the sorts of methods I’ve been talking about, about what happens when people are retrieving emotionally significant information rather than the kind of bland material I’ve been talking about so far? Now, we know a great deal about what goes on that promotes the formation of enduring emotional memories. We know a great deal about that, not least because of the contribution of Dr. Jim McGore [phonetic] over many years, who’s research is simply seminal in this area. We know, for example, that the ability to form these very lasting emotional memories depends upon a particular structure in the brain known as the amygdala. The amygdala is situated in the medial temporal lobe of the brain.
Just follow this big black arrow here. It’s this structure here that sits just in front of the hippocampus. When an emotionally significant event comes along, the amygdala is activated, and this sets in train a whole series of biochemical events that last for quite a long time after the event is finished, that modulate and strengthen the storage of the memory of that event. We know a great deal about that. We know considerably less about what happens in the brain when emotional information is retrieved. That’s what we wanted to find out.
Now, according to the ideas that I’ve given you, these ideas about reinstatement, if the amygdala and other emotional circuitry in the brain is activated when you originally process an emotional event, then when you recollect that event, you should re-evoke activity in that circuitry. If you evoke activity in emotional circuitry, you presumably are going to feel an emotion. Perhaps what it is is that the reason why our memories of emotional events carry with them such a strong emotional component is that we’re essentially re-evoking or reinstating in the brain the emotional reaction that we had at the time that we experienced the event. I want to quickly describe an experiment that tests that idea.
What happens is that when people are in the study phase of this experiment, they are presented with scenes like this. This is a fire hydrant. I didn’t know that until I came to the United States because our fire hydrants don’t look like this in the United Kingdom. They’re much more discreet. However, this is a fire hydrant, and I now know you’re not allowed to park within a certain distance of it and things like that. This would be an example of an emotionally neutral scene. What we ask people to do is to look at these pictures and to rate them for how emotional they were. People would say that this a neutral scene. Up comes then a picture of an object, in this case, a camera, which is simply superimposed on the scene. The task of the subjects was to say or to imagine where in the scene this object might have been found.
We carry on. Here we have a car wreck. People would judge this as being moderately emotionally negative, unless you’ve been in a car wreck, in which case, you would have a different opinion, I suspect.
Most people would judge this as being moderately negative. We have much more negative pictures than this, but obviously I’m not going to show them here. And so, people would make that judgment. Then up comes a picture and it’s superimposed, and so on and so on. Now, the critical point of our experiment is that it tests when people’s memories are being tested, and we’re doing the scanning, all they see are the pictures. They do not see the background scenes. All you see are these pictures. Your task is simply to say whether you saw the object earlier in the experiment or not. The emotional information now is not presented to you when your memory is tested.
The critical question is whether or where in the brain objects that were paired with emotional scenes evoke more activity than objects that were paired with neutral scenes. For example, this barbecue here was paired with an emotional scene. The camera was paired with a neutral scene. If we get more activity when people recognize this barbecue than when they recognize this camera, that can only be for one reason. It must be because this is evoking a memory of the emotional scene compared to a memory of a neutral scene, because there is nothing else that distinguishes these items at the time that your memory is being tested. That’s the critical point.
What do we find? Well, first off, we find that regions of the brain, the hippocampus and the medial temporal lobe, are more active when you recognize pictures that were paired with an emotional scene than pictures that were paired with a neutral scene. In other words, something is coming back along with that picture that’s causing more recollection, even though you’re not asked to recollect these emotional scenes. You’re just told is this old or new--or you’re told to say is it old or new. More importantly, we see activity in the amygdala. So when you’re recognizing pictures that were presented along with an emotional scene, you get back amygdala activation, which was there when you were actually processing that scene in real time. We also see activity in other regions of the brain that are known to play a key role in the control of emotion.
This says, even when you’re not represented with emotional material and asked to remember it, when you remember it involuntarily, you get back the kinds of emotional response that you had when you originally saw or experienced the emotional event. Retrieval of emotionally arousing event can occur automatically in response to an emotionally neutral reminder of the event. It involves enhanced activity in regions that underlie episodic memory retrieval generally. It activates regions that are also implicated in the processing of emotional events in real time. These data, I think, have important implications for our understanding of conditions such as PTSD and depression, where people have this involuntary recollection of emotionally negative experiments. In fact, right now, the only grant I still have running in the United Kingdom is a study of PTSD-- people with PTSD and depression using these experimental procedures to address these sorts of questions.
Let me finish. This is Jane Austin’s quotation again. I just want to point out one thing at the bottom here. She ends up--very prescient though this quotation is, there is something that I hope I persuaded you to disagree with. She says our powers of recollecting and of forgetting do seem peculiarly past finding out. In other words, she thinks that things are so mysterious, we’re never going to figure out how memory works. If nothing else, I hope I’ve convinced you that that’s unduly pessimistic. Maybe we have made some progress in that respect, and there’s hope for the future. It just remains for me to acknowledge the people who do all the work and the people who pay for it. Thank you very much indeed.
|
