 Hi, and welcome back to this video course on biological psychology in this video video 4.4 We're going to take a look at working memory or sometimes called short-term memory Now so what is working memory in the previous video? We've seen that all basically all visual stimuli that we see Briefly form a sensory memory, right? They're sort briefly Reverberate in your visual brain areas and the same is true for sounds that briefly reverberate in your auditory brain areas, etc As soon as we perceive something we have a very brief sensory memory of it Now this sensory memory as we've seen Disappears after about one second but what you what you can do is you can pay attention to some of the things that are in your sensory memory and This paying attention Keeps the representation of these these these items active and that is working memory So imagine for example, you see a display with 12 digits on it, right? Those 12 digits are too much you work you will forget them But what you can do is pay attention to only a few of those digits and then you can rehearse say three or four of those Digits that you see keep those active in your working memory, right? So what happens then is that you see 12 digits You have a very brief sensory memory of all 12 digits and then you form a working memory of only three or four of those digits That's the idea of working memory Now and working memory is roughly equivalent to the term short-term memory I would say it's mostly a fashion choice of fashion whether you want to call it short-term memory It's a bit old-fashioned maybe or working memory, which is the term that most researchers use nowadays Now working memory has a very limited capacity And I gave the example with that you're able to remember only three or four digits It depends kind of on the sense that you write the sense that you are using for your working memory So and also on how you measure the capacity of your working memory, but roughly speaking Visual working memory sometimes called the visual spatial sketchpad has a capacity of about four items So you can sort of in your mind's eye you can sort of memorize three to four items what they look like visually Well your auditory memory sometimes called the the phonological loop auditory visual Auditory working memory has a somewhat higher capacity of about about seven items, right? So if you hear for example a phone number of ten digits, that's too much You will not really be able to remember that but seven digits to a lot of people is barely doable, right? You start to you maybe you're able to rehearse those seven digits in Effectively in your working memory Auditory, right? So for vision seven items would be too much, but for for hearing for audition it would be okay Now this cognitive framework that you have That you have different forms of working memory and especially the the visual form of working memory Which he badly called the visual spatial sketch visual spatial sketchpad or the auditory form of working memory Which he called the phonological loop was developed as I as I said by Ellen badly in the 1970s It's a very influential model of working memory, but I think it's somewhat outdated It's not really incorrect as such I would say but nowadays we have more refined Ways to think about working memory than just thinking it thinking about it in terms of boxes that have a particular Modality attached to them and a particular capacity now in the previous video about sensory memory. We saw this partial report paradigm that Sterling used in the in the 1960s to test sensory memory and you can test almost the same paradigm Test working memory with almost the same paradigm The only thing that you have to change is add some delay between the stimuli and the recall queue recall queue so imagine that so you imagine that we're Memorize we asked a participant to memorize these three by four Letters right and digits actually Then if we show the letters very briefly and then immediately afterwards within one seconds ask the participant Okay, report the top right location. Then we're tapping into sensory memory, right? Because we're still in the time range of sensory memory about one second But if we do the exact same experiment, but we wait a little bit So we present all 12 letters and then we wait say five seconds Then all those letters have faded from sensory memory and in order to remember any of the letters You need to actively pay attention to them and rehearse them, right? So you need to use your working memory and then you will find that performance is much worse because the capacity of working memory is much lower than that of Sensory memory, right and people participants will be able to recall maybe three or four letters Another way to tap into working memory is to mask stimuli So that would mean that you first present these three by four letters and digits and then immediately afterwards you present 12 hashtags, for example at the same location and then what you're essentially doing is Overwriting the sensory memory with those hashtags and then the only thing that is left is your working memory that you actively rehearse All right, so it's a very simple way to test the capacity of working memory of people using this partial report paradigm How exactly neural neuraly working memory works is still quite unclear But I think there are three main hypotheses And the traditional hypothesis which we'll take a look at first is that the frontal cortex has Areas that are dedicated to working memory, right? So your frontal cortex is the part of the brain that's in front and that is not really Directly involved with one specific modality And there would be sort of special brain areas that implement working memory. That's the traditional traditional model a More modern way to think about working memory is that working memory is essentially a form of perception without actual Any sensation coming in, right? So in that case with working memory Corresponds to you could say sustained activity in sensory brain areas So to make it a bit more concrete It would mean for example that if you working memory of something visual a visual stimulus would basically trigger the same kind of activity In your occipital cortex in your visual brain areas that actually seeing that same stimulus also Triggers, right? So we sort of reusing our visual brain areas for visual working memory And we would be using reusing our auditory brain areas for auditory working memory, etc So it's quite an attractive idea And I would say a state of the art model is kind of a mixture of the two and To my view, but it's a this we don't know it's just my opinion But to in my view this sort of state of the art mixture model probably approximates the truth best And the idea is that working memory can be either active or passive so that there are different ways in which you can keep things in working memory We'll take a look at that later And that when you keep something actively in working memory It is indeed represented in your visual brain areas if it is a visual item or in your auditory brain areas if it's an auditory item, etc But if it is not visual But if it's not active so if you're keeping Multiple items in working memory most of these items would be past represented in a kind of passive way And that would them maybe indeed be done in your frontal cortex, right? So the idea being that you have one active item in working memory and multiple sort of passive items in working memory We'll take a look at some evidence in favor of this idea in a minute But let's start with the idea that there is a link between working memory and the frontal cortex, which there certainly is And there are a set of classic experiments and one very nice one was what done by Goldman Rakic and colleagues in which they trained monkeys to make memory-guided saccades So a memory-guided saccade is simply the monkey looks at a fixation dot like this and then you flash for example Stimulus here and then after you ask the monkey to your train the monkey to remember to to to be able to do this task After say a few seconds the monkey makes an eye movement to the location that was cute, right? So the monkey makes an eye movement, but not immediately after a little delay So there's an element. It's just like very very simple working memory, right? With the set size of one just remember the location that you're going to make an eye movement to So they see a cue then there's a delay and then the monkey makes a saccade an eye movement to the cute location Now at the same time while the monkey is doing this task there the Goldman Rakic and colleagues were measuring Activity of neurons individual brain cells of that monkey in the prefrontal cortex So they really had right so these are the type of experiments where they really open up the skull of the monkey insert electrodes into the brain and measure individual Individual neurons, right? That's a very powerful technique and it has taught us a lot About how the brain works regardless of how you might feel about that kind of animal treatment from an ethical point of view And then the following came out. So this looks really complicated, but I'll walk you through it. It's not that difficult So say that the monkey was preparing an eye movement down So the monkey was initially fixating here and preparing an eye movement down to this dot here Then what they found was during the delay period So while the monkey wasn't new already that it was going to make a second downwards But did not yet execute that's a cat, right? So it's a cat is an eye movement They saw that some neurons specific neurons started to fire Encoding essentially for the downward position On other trials where the monkey would make an eye movement to the left a different set of neurons would start to fire If they would a monkey would make an eye movement to up a different set would start to fire, right? So different sets of neurons encoded for different positions the memories of different positions So clearly these these neurons in the frontal cortex had some information about what this monkey was keeping in in his or her working memory, right because You can tell from this pattern of results right if you look at these neurons you can say, okay This monkey is his head has a downward cicada in working memory So this information is very simple form of working memory was clearly represented by these neurons in the frontal cortex of the monkey a very striking very very profound finding So that's where the idea comes from that the frontal cortex is involved in working memory Now we also have evidence that sensory brain areas are involved in working memory Right and that comes from many types of experiments But for example the following kind of experiment say that you now we're not measuring monkeys But we're measuring human humans in an fMRI scanner So we are measuring their brain activity using functional magnetic red resonance imaging And then we asked a participant to keep for example a line in working memory and the line can for example Be tilted like this or like this and a participant has to say it has to remember what what the orientation of the line is now then With the brain activity that is measured Through fMRI and some computational processing right in a computer You can determine what the participant is memorizing keeps is keeping in working memory in other words You can sort of read you read the mind of that participant You sort of decode the content of the working memory of that participant based on brain activity and more specifically Based on brain activity in the visual cortex not in frontal cortex, but in the visual cortex in other words The visual cortex in this type of experiment contains some information about what the that person is keeping in working memory So visual cortex therefore encodes working memory at least in some cases right So the studies by goldman rakeach and colleagues using the monkeys Shows that working memory is related to the frontal cortex And here these studies show that working memory is also related to the sensory cortices right and I think they're both quite Compelling findings they both show tell us something about how working memory works But they tell us something different Right, so what is the role of the frontal cortex and what is the role of the sensory cortex? So Well, let's go back to the sensory cortex this kind of suggests that we reuse brain areas Right, so it suggests that if we have a brain area that's involved in vision So our primary visual cortex here Then we use that same brain area for visual working memory And if we have an area that's involved in hearing we use that for auditory working memory, etc A very elegant idea. I think very for very parsimonious Now one question that you can ask yourself is if you are remembering Multiple things that use essentially the same visual brain areas. How can you do that? Right? So for example, how is it possible that we can remember multiple objects that were presented at the same location? So these would these these objects would activate similar activity in your visual brain areas So it's difficult to see how you can remember all of those different things in the same visual brain areas Right in a sense, they just don't fit that we our visual brain areas are just not Don't have the representational capacity to represent all the things that we can keep in working memory or Maybe they do but that is one that is at least one argument that you can raise against this model of working memory So more recently People have tried to reconcile this this idea that you use your frontal cortex and use your visual cortex using kind of mixture model And this mixture model suggests that you can generally keep one active item in your working memory really active That's the item that you are going to use immediately Right. So imagine that you're in a supermarket and you have a lot of things on your in sort of in your mental shopping list The then one of the things that you have on your mental shopping list is the thing that you're looking for right now Right. So for example, I want to I want to get I want to get a beer. I want to get a sugar and I want to get milk And then but right now I'm searching for the milk So the milk would then be active in my working memory and the sugar and the beer would be still in my working memory But less active and the ideas of this this type of model is that that active work active item in your working memory Would be represented in your sensory areas Whereas multiple items that are not active are represented somewhere else possibly in your frontal cortex Right. So that there are different ways to represent things that are in your working memory Kind of goes against the idea that visual working memory or working memory in general is one thing Right because it suggests that there are different forms of working memory Which is not really attractive to most scientists. We like to reduce things make things simpler, right? So to make to basically Reduce all different forms of memory into one form of memory Because it makes life easier it makes life we if things are simple That's more elegant as an as a scientific theory And what this mixture model does is actually make things more difficult because it takes a concept working memory that we thought of as one concept and Tears it apart into different kinds of things, right? So it makes a messy world even messier But that does not mean that it's not true. Obviously um, so in uh, uh, researchers such as christian olivers and masut hussein and others have done quite some work Uh, suggesting that this mixture model has some truth to it. I think Now, um Let's take a look at one specific experiment, which uh, I think supports Somewhat supports this idea that there is a mixture model So what in this experiment participants have to do? I walk you through it This is this is really if you're not familiar with kind of neuroscience experiments This may be a bit tricky, but I think it's interesting if you if you follow me So what participants have to do here is remember two motions. So first they see a dot So cloud of dots here green that's moving for example a little To the to the bottom right and then there's a cloud of dots that is moving upwards Um, and then after so they see them both first the first and the second and then they have to reproduce these uh the orientations um Now they are presented one after another and the assumption that they make which is quite an assumption But there's something to it is that the last thing that is presented is active Right so there you can have things that are active in working memory and less active And the thing that is presented less the red the red dots in this case would be active According to them Then what they do, uh so, uh So at the end of the trial, right when participants are reporting the the motions the first motion the green one in this case Would be inactive and the last motion the red one would be active Now what they then do is apply tms. Now. What is tms? Tms is essentially a magnetic pulse that you can apply to the brain of a human or of anything But also of a human because it's kind of harmless and it disrupts Brain activity So if you apply tms over your visual cortex, you will start it will disrupt visual brain activity And you will start seeing all kinds of things that are not there right sort of hallucinations And what they did is they applied tms to area m m t plus Which is and that is very crucial and this is an area that is evolved in the perception of visual motion Why is that important? Well, because the participants here where we were memorizing visual motion, right? So say that both these these motion the the this motion and the upward motion here would be represented in area m t plus Then applying tms to area m t plus should disrupt performance of both Both stimuli right both the green and the red one But what they found Was that the memory was actually impaired only for the last item. So the one that was active And improved for the actually the inactive item Suggesting That this active item was represented in area m t plus right So do you see the logic they sort of disrupt activity in your sensory brain areas with with TMS And then they they look at what that does to your working memory and in this case They find that it disrupts the working memory only of the active item the last presented item the red one in this case And hence they conclude that only the active item was represented in your in your sensory brain areas That's the logic that they have here and that kind of suggests indirectly And there's certainly much more to be said here and much more to be replicated also This is I think the kind of finding that you need to replicate But it kind of suggests that there really is a qualitative difference between things that are active in your working memory Which are represented in sensory brain areas and things that are not active in working memory Right because only the active item was disrupted by tms Okay, now with that fairly technical story, which I hope you enjoyed Let's move on to the next video video 4.5 in which we're going to take a look at long-term memory