How does Embodiment Science shift how we think about Memory? (with Guy Claxton & Emily Poel)

Welcome to the Future Learning Design Podcast. Hi everyone, welcome to this next episode in the series on the implications of embodiment science in education. I'm your host, Tim Logan, and after episode 1 on intelligence and episode 2 on learning, this is the 3rd installment where Guy Claxton and Emily Poole share their take on what embodiment science has to say about how we understand memory and knowledge. This is a hugely important call to those who have forgotten that the dominant way we talk about memory in education is actually a metaphor, and a bad one at that, and the science is now telling us that it needs to be replaced with something much more biologically plausible. I started out by asking Guy, how does embodiment science change the way we think about memory.

It follows from the way we've been talking, developing alternative ideas of intelligence and learning. We need a biologized way of thinking about memory. To have a theory or a model or an image of memory is to have a way of conceptualizing the effect that experience cumulatively is having on our ability to behave intelligently in the world. You know, you have to remember, like anything else in science, the first attempt to understand, you know, what is memory, it usually comes from some kind of metaphor or some kind of analogy. And the analogy that's been in most in favor, certainly since the 1960s or before the 1960s, but strengthened by the arrival of the digital computer, is that memory is a store. And there are still people on the planet who don't realize, who don't understand that memory as a store is a metaphor. It's a way of looking at memory. Memory is not a store. Memory, for some purposes, might be conceived of usefully as a store, but it isn't. The idea that memory is a warehouse, memory is a hard drive, memory is a library, those kinds of ideas which ran psychology from around about 1960 to around about mid-1990s, they've passed their sell-by date, as all metaphors do. You know, we used to think that the atom was like a pebble. Then someone said, no, it's more like a current bump. And then someone said, no, it's more like a mini solar system. And then somebody else said, no, it's more like a kind of smear of something in space, and so on and so on. It's like metaphors and analogies, usually grounded in some kind of concrete experience, back to the embodiment and embeddedness of thinking, provide a way of bootstrapping our understanding But that understanding then always comes to that metaphor, always reveals its limitations. And now we've got to the point where the limitations of memory as a store are becoming apparent because that store image does not connect, pays no attention to the development of our understanding of how intelligent our biology is. So we need an image of memory now that is biologically plausible. So in the book, we developed the idea that what we've been taught, the memory, the sedimentation of experience is what we call the life scape. You can imagine it as a plain which is kind of hilly, and memory is the way in which that landscape has been sculpted, eroded by the weather so that when it rains again, the effect of previous weather is to create runnels and rivers and channels. hills and valleys, which guide the flow of our neural energy as it trickles its way through the Mississippi Delta of our brains. So that gives us a different starting point. And it's hilly because it's not binary. It's not digital. It's like everything is probabilistic. You know, most cats have fur, most cats have 4 legs, most cats mew, most cats like fish, but not all of them. Some cats have 3 legs. Some cats have recently been shorn, some cats, et cetera, et cetera. So concepts are fuzzy, are inherently fuzzy. So the livescape is fuzzy. But then when you're about 1 or so, you start to realize that there are different ways of accessing the livescape that you don't know yet, but they're called words. So you have a fuzzy concept of the family dog, and then you learn that family dog has a name. Called Fido. So you can activate the Fido bit of your livescape by hearing a word. So you begin to develop another plane of words, which is another way of accessing the livescape. And it derives from the livescape. You couldn't know what Fido referred to or the word dog referred to unless you already had a fuzzy concept of the dog. But once the wordscape begins to develop. So words now can join themselves up and they join themselves up in ways that may no longer correspond exactly to what's going on down in the livescape. You can have an image in your mind's eye of something called a unicorn that, as far as we know, doesn't exist in the experiential landscape, except it does in Disney films now. But you can start— the wordscape becomes semi-detached. but never completely detached because all the conceptualization in the wordscape ultimately needs grounding in the livescape. So if you ask someone to start talking about an abstract concept like love or democracy or truth or something like that, almost immediately they'll start telling you about anecdotes or protest marches that they went walking on or sexual feelings or whatever it might be. It's like words would not have any meaning. This is the old Chinese room problem. Words would not have any meaning if they weren't capable of being constantly grounded in what's going on in the landscape. And a lot of the embodiment science work in experimental psychology has been documented about how even abstract concepts activate parts of the brain, which were the ground of the original metaphor, which has now become more abstracted. John grasped the hammer fires up the bit of your brain that is about clenching your fists, but so does John grasped the argument. If you pop someone in an MRI machine and ask them to make a judgment about whether John grasped the argument is a good sentence or not, the little bit of their brain that is to do with clenching your fist, to do with grasping, automatically lights up, even though that's completely irrelevant understanding the meaning of John grasped the argument. So those connections remain there, maybe not all the time, maybe not exactly as they always were. But that grounding, that earthing of our conceptual knowledge in the life-scape is always important. Gives us a very different image of memory, doesn't it?

It does. And the thing that I think is so exciting about that is the reminder that we've lost ourselves in the map or in the wordscape or in the metaphor. We've confused the metaphor with the reality. And so when people talk about short-term memory, long-term memory, the bottleneck of the working memory and all of this cognitive architecture language about memory, yes, they have forgotten that that is a metaphor. And as you have just said, it's an outdated metaphor. But it is, even if it was contested but still relevant, it is still a metaphor. And I mean, in a way, your Lifescape wordscape is still a metaphor.

Absolutely it is. Quite right.

But it's reminding us that it, there are other ways to conceive metaphorically of our reality, that maybe some of them are more, I don't know, are more generative than others.

Yeah. I just think it's bad science these days in 2026. If your model doesn't have some biological plausibility, forget it. We need to be grounding our models in what we know. And of course, there's still what Donald Hebb, the father of this kind of thinking back in 1949, referred to as the conceptual nervous system. We're developing a language which is biologically plausible, but of course the brain remains 100 orders of magnitude more complex than the livescape and the wordscape. They're just images. They're tools for thinking, but they're not, you know, you'd be foolish to suppose that that was the be-all and the end-all of what's going on in your brain. But they have more of a biological plausibility and that gives them brownie points.

But can I ask, how would you then articulate the idea of the limits on working memory that is used often in an educational context for designing instruction. How does that change with embodiment science?

Well, here's another— lots of people are writing and talking about working memory now as the temporary activation of different areas within the livescape or the wordscape, whichever you like. In other words, short-term memory phenomena are to do with the way you manage separate centers of activity that need to be kept coactive whilst the clever processing modules in your brain are trying to find a way of meaningfully integrating them. So short-term memory phenomena tend to arise under circumstances where you can't immediately create a sense of meaning or a sense of the gist. What does this mean? So strings of digits, For example, are the prototypical material, or people talk about items in memory. That was the experimental language that people used to base those models on. And it's quite difficult. It's quite demanding, I think, at an energetic level. The brain is not built to keep 7 separate little pockets of activity going simultaneously. without finding a way of integrating them, without finding a way of resolving them in some way or other. I think my simple metaphor is it's like spinning plates. You know that old-fashioned kind of circus thing, that spinning plate? You have to keep running around, you have to keep respinning the plates, otherwise they fall off. And so I think there's a whole different, a more embodied, a more physiologically plausible way of thinking about short-term memory. Which is, you know, part of the problem is that the brain doesn't like trying to keep the different areas of itself activated but unintegrated.

Like energy conserving or something.

Exactly. So it's a bit like, here's another of my simple metaphors. The brain has built-in mechanisms around switching off areas of your brain, you know, that haven't been refreshed for a while. A bit like those irritating public toilets where you walk in and the light comes on. Then halfway through you're doing your business, the light goes out because there's been no movement in the situation. So I think the brain has mechanisms like that around conserving energy or stopping too much activation, just like stopping a kind of irresistible, like throwing a bomb into a box of fireworks. Everything kind of starts firing off. That becomes an epileptic fit where the brain loses the ability To keep itself under control, to keep itself corralled in a way.

It's fascinating to me how self-fulfilling so many of these prophecies are, right? Because your point about this, the items, I am not curious about this string of 7 digits that I have to learn. And therefore, of course, there's a whole bunch of short-term memory capacity issues around doing that because I don't really want to do that. And this, you just get these self-fulfilling prophecies within the current education system? Because then it seems like this short-term memory bottleneck theory holds true because you see kids struggling to memorize irrelevant bits of data.

But the problem are the irrelevant bits of data.

Exactly. That's exactly my point. Exactly.

Well, there's a name for these kind of unintegrated, ununderstood things. They're called Facts. So you shrink your image when you have that image of memory as being the storage of facts. Then you see the kind of the damage that that does to the curriculum. You know, you then start thinking about the curriculum as the storage of facts. That's your image of memory.

But then you can't argue against that because you can't argue against people knowing stuff about the world. And because knowing stuff about the world is defined in the language of facts, And so is memory. And so again, it's just another self-fulfilling— it's like a closed system that self-fulfills, right?

Well, there's a difference between knowing facts just because someone tells you they're important and knowing facts in the way that David Perkins says, you know, there are other kinds of facts which are useful navigationally, things that have meaning or mental skills like, you know, reading a bus timetable or whatever it might be. You know, they, they're embedded in some sense. They have some sense of purpose and functionality.

We hope you've enjoyed this episode. Please feel free to continue the dialogues with our guests, with us at Good Impact Labs, on our blog or on social media, or within your own networks.