The Neuroscience of Consciousness – with Anil Seth

- Two of the three fundamental mysteries about our place in the universe have already been solved. The first is literally about our place in the universe. Many years ago Copernicus told us that we were not at its center, that we were just a point suspended in the abyss. This is an image of Earth taken by the Voyager 1 probe as it left the solar system about six billion kilometers away. The entire history of humanity, the entire history of life on Earth, has taken place on that pale blue dot. The second mystery, Darwin then revealed that humans are just one branch, or twig, of a wonderfully rich and delicate evolutionary tree.

And that much of the machinery of life is shared with even the humblest of our fellow human beings. The third mystery is that of

consciousness

, our inner universe. Now, at the beginning of this year, for the third time in my life, I ceased to exist. As the anesthetic propofol flowed from the cannula in my wrist into my bloodstream and then into my brain, a meltdown occurred. A blackness. An absence. And then I came back. Sleepy and disoriented, but definitely there. And when you wake up from a deep sleep, you may not know what time it is, especially when flying somewhere, but you will know that some time has passed.

There seems to be a basic continuity between your

consciousness

then and your consciousness now. But coming from general anesthesia, it could have been five minutes. It could have been five hours. It could have been five days or five years. It just wasn't there. A premonition of the oblivion of death. And general anesthesia not only acts on the brain. It doesn't just work in your mind. It works on your consciousness. By altering the delicate electrochemical circuitry inside your head, the basic fundamental state of what is meant to be is temporarily abolished. And in this process lies one of the greatest mysteries that still remain in science and philosophy.

How does consciousness occur? Why is life in the first person? He leaves and comes back. The modern incarnation of this problem can often be traced back to Descartes, who in the 17th century distinguished between matter, res extenso, the stuff these desks are made of and what clothes are made of. But brains and bodies are also made of material matter. And res cogitans, the stuff of thought, of feelings. The stuff of consciousness. And in making this distinction, it gave rise to the now infamous mind/body problem, and life has never been simple since. But Descartes actually created even more damage with his beast machine doctrine, which I'm going to mention now, because he anticipates where I'm going to end up when the bell rings when I'm done in an hour.

Before Descartes, people commonly believed in something called the great chain of being, with rocks and plants at one end, and other non-human animals, a little higher than humans, and then angels and gods at the top. And this great scale of being was also a scale of moral virtue, so that humans had more moral virtues than animals and plants, and then rocks, etc. Now, Descartes, in making this division between mind and matter, argued that only humans had minds and therefore moral status, while other animals did not have minds. They were mere physiological machines, or beast machines, morally equivalent to plants and rocks.

And from this point of view, the physiological mechanisms that give rise to the property of being alive were not relevant to the presence of mind or consciousness. Now I am going to propose, at the end of this talk, the opposite. That our conscious sense of self arises because of, and not despite, the fact that we too are beast machines. So to get there, let's return to the apparent mystery of consciousness. In 1989, quite a long time ago, but not that long ago, Stuart Sutherland, founding professor of experimental psychology at my university in Sussex, said the following. "Consciousness is a fascinating but elusive phenomenon.

It is impossible to specify what it is, what it does, or why it evolved. Nothing worth reading has been written about it." It is a rather pessimistic point of view. And that may have been true then. I don't think it was true then, but in any case things have changed a lot since then. And around the time Sutherland made these comments, we can see the birth, or rebirth, of the study of consciousness within the

neuroscience

s. And a nice milestone is this paper by Francis Crick and Christof Koch, published in 1990. And they start their paper by saying that it's notable that most of the work in cognitive science and

neuroscience

doesn't reference consciousness or attention at all.

And then they propose their own theory about what the neural correlates of consciousness are. What is there in the brain that is responsible for being conscious. And since then, over the past 25 years, there has been first a trickle and now an avalanche of research into the brain basis of conscious experience. Some of this work is carried out in my laboratory, the Sackler Center, the science of consciousness, which was founded six years ago with Hugo Critchley, my co-director. And now there are even specialist academic journals, The Neuroscience of Consciousness, which I started last year at Oxford University Press.

And this is a real change of direction. When I started more than 20 years ago, it was thought to be professional suicide to want to study consciousness scientifically. And it may still be, we don't know. We'll see. So while the brain basis of consciousness remains a mystery, it is, in a sense, an accessible mystery. And I think the author, Mark Haddon, put it very well. He said that the raw material of consciousness is not on the other side of the universe. It didn't happen 14 billion years ago. And it is not hidden deep in an atom. The raw material of consciousness is here, inside your head, and you can hold the brain in your hands.

But the brain won't reveal its secrets very easily. The extraordinary thing about the brain is not so much the number of neurons, although there are around 90 billion. It's not even the number of connections, although there are so many that if you counted one every second, it would take you about three million years to finish counting. What is truly extraordinary are the patterns of connectivity, which to a large extent are not yet known, but in which everything that makes you you is inscribed. The challenge then is this, at least from my point of view. How can the structure and dynamics of the brain, in connection with the body and the environment, explain the subjective phenomenological properties of consciousness?

And considering things this way, we run into what the philosopher David Chalmers has often called the difficult problem of consciousness. And the idea is this. There is an easy problem. The easy problem is to understand how the combined operations of the brain and body give rise to perception, cognition, thinking, learning, and behavior. In other words, how the brain works. The difficult problem is to understand why and how all this should have something to do with consciousness. Why aren't we just robots or philosophical zombies, with neither in the universe? Now here's a tantalizing intuition: even if we solved the hard problem, even if we solved the easy problem, the hard problem would still be as mysterious as it seems now.

But this seems like a mistake to me. It may not be necessary to explain why consciousness exists in order to advance understanding of its material basis. And this for me is the real problem of consciousness; how to explain their various properties in terms of biological mechanisms without pretending that they don't exist at all, as you do if you solve the easy problem, and without trying to explain why they are parts of the universe in the first place, which is the hard problem. And in the history of science, we've been somewhere similar before. It's hard to say if this is exactly the same situation.

But in our understanding of life, the eminent biochemists of the time found it completely mysterious how biological mechanisms could give rise to the property of being alive. And things like elan vital and vital essence, and all kinds of other things were proposed. And although we still don't understand everything about life, this initial sense of mystery about life has largely dissolved as biologists have just begun to understand the properties of living systems in terms of mechanisms. An important part of this story was the understanding that life is not a single thing, but rather a constellation of partially dependent and partially separable processes, such as metabolism, homeostasis, and reproduction.

Likewise, to advance the real problem of consciousness, it may be useful to distinguish different aspects or dimensions of what it is to be conscious. The space of possible minds, if you will. And a simple classification is at the conscious level, which is the property of being conscious at all. For example, the difference between being in dreamless sleep, or under general anesthesia, and being awake and conscious as you are now. And conscious content, when you are conscious, you are conscious of something. The infinite images, sounds, smells, emotions, feelings and beliefs that populate your inner universe at any given moment.

And one thing that you are aware of when you are aware is the experience, the specific experience, of being you, and this is the conscious self. And it is the third dimension of consciousness. Now, I don't claim that these distinctions mark completely independent aspects of what it is to be conscious, but they are a pragmatically useful way of looking at the problem a little. So let's start with the conscious level. What are the fundamental brain mechanisms underlying our ability to be conscious? And we can think of this, at least to a first approximation, as a scale that goes from being completely unconscious, as if in a coma, or under general anesthesia, to being awake, alert and fully conscious as you are now.

And there are various states between being drowsy, being lightly sedated, etc. The important thing is that, while being conscious and being awake usually go together, this is not always the case. For example, when you dream you are asleep, but you are having conscious experiences. The conscious experience of your dreams. And on the other side of this diagram, there are pathological states, like the vegetative state, where physiologically you're going to go through sleep/wake cycles, but there's no one home. There is no consciousness happening. So what are the specific mechanisms underlying being conscious and not simply being physiologically awake?

Well, there are several possibilities. Is it the number of neurons? Well, actually, probably not. There are more neurons in the cerebellum, in this part of the back of the brain, than in the rest of the brain combined. In fact, there are approximately four times as many neurons in the cerebellum as in the rest of the cortex. But if you have damage to the cerebellum, yes, you will have some coordination problems and whatnot, some cognitive problems, but you won't lose consciousness. It's not just the number of neurons. It doesn't seem to be any region in particular. In fact, there are regions that, if you suffer damage, you will lose consciousness permanently; in the thalami, nuclei and thalamus deep in the brain.

But these seem to be more on-off switches than actual generators of conscious experience. It's not even neural activity, at least not simple types of neural activity. Your brain is still very active performing unconscious states during sleep. And even if your brain is highly synchronized, one of the first theories of consciousness was that it depended on neurons firing in synchrony with each other. If your brain is too synchronized, you will lose consciousness, and this occurs in absence epilepsy states. What seems to be the case is that being aware of everything depends on how different regions of the brain communicate with each other in specific ways.

And this was groundbreaking work by Marcello Massimini in Milan about 10 years ago. And what he did here was stimulate the brain's cortex with a brief pulse of electromagnetic energy, using a technique called transcranial magnetic stimulation, or TMS. And then he used EEG electroencephalography to listen to the brain echoes. A bit like tapping the brain and listening to its electrical response. And what he noticed when he does this, and he can see that on the left he's asleep and on the right he's awake. And this is very slowed down. When you stimulate the brain while you are asleep, there is still a response.

There is still an echo, but the echo remains very localized up to the point of stimulation. It doesn't travel much and it doesn't last long. But when you stimulate a conscious brain, there is a spatially and temporally complex response. This echo bounces around the crust in very interesting ways. Furthermore, the complexity of this echo can be quantified. Some simple algorithms can be applied to describe how complex and rich this pattern of interactivity is. This is also from the Milan group. And what they've done here is basically watch the echo as it moves through the brain. And they see to what extent one could describe it, the minimum length of the description.

How much can you compress the image of that echo? The same way algorithms create compressed files from digital images on your phone. And they came up with an index called the perturbation complexity index. and whatWhat you discover is that you now have a number that you can associate with your level of consciousness. I think this is really intriguing, because it is a first step towards a real measurement of the conscious level. This graph at the bottom shows this measure applied to a variety of states of consciousness, ranging from pathological states of consciousness, such as the vegetative state, where there is no consciousness at all, throughout the lock-in syndrome and then a healthy awakening.

And you can immediately see that techniques like this might already have clinical value in diagnosing the potential for consciousness that patients might have after a severe brain injury. Now at Sussex we continue working along these lines. In fact, rather than hitting the brain with a sharp pulse of energy, we want to see if we can get something similar by simply recording the brain's spontaneous activity. So we look at the spontaneous dynamics of, in this case, waking states and anesthesia. This is work with my PhD students Michael Schartner and Adam Barrett. We measured its complexity and, in fact, found that we can distinguish different levels of consciousness simply by the spontaneous activity of the brain.

In some ways this is not so surprising, because we know that several things change. The balance of the different frequencies of your brain activity changes when you lose consciousness. But this has nothing to do with that. This is independent of that. There is something specific that is being detected in these changes in complexity. More recently we have applied the same measures to sleep, in this case taking advantage of recordings taken directly from the human cerebral cortex with colleagues in Milan. These are implanted electrodes. And we see practically the same story. If you compare where the two areas are, you compare the complexity of waking rest and early non-rem sleep, where you don't dream much.

You see that complexity is very deceiving. The interesting thing here is that if you compare waking rest to REM sleep, where people often report having dreams if they are awakened, the level of complexity is very similar to that of the waking state. There's something else going on here, which is that the complexity in the front part of the brain appears to be greater than in other parts of the brain. And that's something we still don't fully understand. I just wanted to give you some fresh press, so to speak, which is where they've also been applying these measures now to data taken from people under the influence of psychedelic drugs; psilocybin, ketamine and LSD.

And what we found, at least in our hands to begin with, is that the level of complexity actually increases compared to the baseline state, something we haven't seen before in any other application of these measures. So the important thing about this way of looking at things is that it is based on a theory that attempts to explain why certain types of brain dynamics go hand in hand with consciousness. And put very simply, the idea is this, and it goes back to Guilio Tononi and Gerald Edelman, people I worked with in the United States about 18 years ago: the idea is very simple.

Consciousness is extremely informative. Every conscious experience you have, have had or will have, is different from every other conscious experience you have had, are having or will have. Even the experience of pure darkness rules out a vast repertoire of possible alternative experiences you could have, or might have, in the future. In any conscious experience there is an enormous amount of information for the organism. At the same time, every experience you have is highly integrated. Each conscious scene is experienced, as if it were a single piece, it is united. We do not experience colors and shapes separately in any way.

It is conscious experiences at the level of phenomenology that combine these properties. On the one hand, they are very informative and are made up of many different parts. On the other, united in a unified whole. And this motivates us to look for mathematical measures that have the same property, that do not lack information either. On the left, you see a system that is all connected to each other, so it can't enter many different states. On the right is a system that is completely dissociated, so they can enter states, but it is not a single system. We want measures that follow this middle ground of systems, that combine integration and differentiation.

And several of these measures already exist. Here are some equations, which we can talk about later if you want, that try to aim for this middle ground. And time will tell whether, by applying these more phenomenologically grounded measures, we arrive at even more precise practical measures of consciousness. Now, why is this issue of measurement important? And I want to make a general point here, which is that if you're trying to naturalize a phenomenon that seems mysterious, the ability to measure it is often one of the most important steps you can take. And we have seen numerous examples of this.

The history of our understanding of heat and temperature is a very good example. Here is an old thermometer from the 18th century that used air expansion. But of course there are many problems with generating a reliable thermometer and temperature scale if you don't already have fixed points. And if you don't know what heat is, you get trapped in a kind of vicious circle that took a long time to get out of. But people managed to get out of this, and the development of thermometers catalyzed our understanding of heat from something that flowed in and out of objects to something identical to a physical property.

The average molecular kinetic energy of the molecules of a substance. And having that concept of heat now allows us to talk about temperature far beyond the realm of human experience. We can talk about the temperature on the surface of the sun, in a sense, in the same way that we can talk about the temperature of interstellar space, close to absolute zero. None of these things make any sense in our phenomenological experience of hot and cold. This brings me to my first take home message. Measurement is important, and consciousness, the conscious level, depends on a complex balance of differentiation and integration in brain dynamics, reflecting the fact that conscious experiences themselves are highly informative and always integrated.

Now, when we are conscious, we are conscious of something. So, what are the brain mechanisms that determine the content of consciousness? And the hero of this part of the story is the German physicist and physiologist Hermann Von Helmholtz. And he proposed the idea that the brain is a kind of prediction machine. That what we see, hear and feel is nothing more than the brain's best guess about the causes of sensory stimuli. And the basic idea is, again, pretty simple. The brain is enclosed within its bony skull and has very indirect access to the external world. All you receive are ambiguous and noisy sensory signals, which are highly and directly related to this external world of objects, and so on, if there is an external world of objects out there.

They know about it. From this point of view, perception is necessarily an inferential process in which the brain interprets these ambiguous and noisy sensory signals with respect to some expectations or prior beliefs about what the world is like. And this constitutes the brain's best guess about the causes of the sensory signals that impact our sensory surfaces all the time. What we see is the brain's best guess about what's out there. I want to give you a couple of examples that illustrate this process. It's pretty easy to do, in a way. This first example is a well-known visual illusion called Edelstein's Chessboard.

Now here you will see two patches. You'll see patches A and B. And I hope they look like different shades of gray to you. They? They appear to be different shades of gray. Of course, they are the exact same shade of gray. I can illustrate it by putting an alternative image and joining those two patches. You will see that it is the same shade of gray. You may not believe me, so what I'll do is change it and you'll see it even more clearly. There are no sharp edges. It's the same shade of gray. What's happening here, of course, is that the brain is unconsciously applying its prior knowledge that a cast shadow obscures the appearance of the surfaces it casts on.

Therefore, we consider patch B to be lighter than it actually is, to account for that effect. And this is, of course, an illustration of the success of the visual system, not its failure. The visual system is a very bad physical light meter, but that's not what it's supposed to do. It is supposed to, or something it is supposed to do, is to interpret the causes of sensory signals in terms of meaningful objects in the world. It is also an example of what we sometimes call cognitive impenetrability. Even if you know the patches are the same shade of gray, when I remove that bar, they again look different.

I can't do much about it. The second example simply shows how quickly the brain can absorb new prior information to change the nature of conscious perception. This is the so-called Mooney image. And if you haven't seen it before, hopefully what you see here is just a series of black and white blobs. Does everyone understand that? Black and white spots? Some of you may have seen this before. And now what I'm going to do is complete it and you'll see something very different. What you will see here is a very significant scene that involves at least two objects, a woman, a hat, and a horse.

Now, if you look at this for a moment, I won't leave it for too long, but if you just look at it for a moment and then I remove that image again, you should still be able to see. the objects within that image. To me this is quite remarkable, because the sensory information has not changed at all. The only thing that has changed is your brain's previous expectations about what that sensory data means. And this changes what you consciously see. Now, this also works in the auditory field. Here are two spectrograms. This is something called sine wave speech, and what you see here are two time-frequency representations of speech sounds.

The one above has all the crisp acoustic features that eliminate normal speech. A bit like putting a threshold on an image. And the background is something else. So I'll play the top first and see how it sounds. And now I'll play something else for you. (BOTTOM SOUND - MAN'S VOICE): Jazz and swing fans like fast music. - I hope everyone understood that wise advice. And now if I play the original sound again... - Yes? This is exactly the same. Again, all this change is what we hope that sound means. - Once again, just to be lucky.

It's not just a bunch of loud whistling, it's speech. Now, this typical framework for thinking about these types of effects is Bayesian inference. And this is a form of probabilistic reasoning, which is applicable in all kinds of areas, not only in neuroscience, in medical diagnosis and all kinds of things, like finding lost submarines. But in neuroscience we talk about the Bayesian brain. And it's a way of formalizing Helmholtz's idea that perception is a way to guess better. And the idea is that sensory signals and prior beliefs can be represented as a probability distribution. So, for example, this yellow curve is the probability that something will happen, perhaps that you've briefly glimpsed an object moving to the right.

Sensory data may say something different. You may have a chance of peaking at a different angle of movement. Maybe it's going in a different direction. And the optimal combination of the prior curve and the probability, the yellow curve and the red curve, is this green curve, which we'll call the posterior distribution. And that represents the best optimal combination of these two types of evidence. And the idea is, well, that's what we perceive. Thinking about perception in this way is quite foreign to the way that, classically, in neuroscience, people have thought about perception. The classical view is that the brain processes sensory information from the bottom up or forward.

This is an image of the monkey's visual system, and the idea is that information comes through the retina and then goes through the thalamus. Then it goes to the back of the brain. And as sensory signals filter deeper and deeper into the brain, they encode or represent progressively more sophisticated features of objects. So you start at the early levels of the visual cortex with response to luminance and edges, and then further up to objects, including other monkeys. The important thing here is that the perceptual heavy lifting is done by information flowing in this upward or forward direction.

Now, the Bayesian idea of ​​the brain says something very different. He says what's really important are the top-down or inside-out connections that flow from the center of the brain outward. And we've known for a long time that there are a lot, a lot of these connections, and some descriptions flow rather the other way around. But the function has been quite mysterious. Thinking about the Bayesian brain gives us a good way to interpret this. And they are exactly these connections abovedownward or inside-out that transmit predictions from higher levels of the brain to lower levels, to lower levels, and back to the sensory surfaces.

So these blue arrows convey the brain's predictions about the causes of sensory signals. And then what flows in the forward direction or bottom-up, outside-in, is simply the prediction area, the difference between what the brain expects and what it gets at each level of description. So in some formal frameworks, this is often called predictive coding or predictive processing. And the idea is that prediction error minimization occurs at all levels of this hierarchy at the same time, and what we then perceive is the consequence of this joint prediction error minimization. We can therefore think of perception as a kind of controlled hallucination, in which our perceptual predictions are checked at all points by sensory information from the world and the body.

There are now quite a few experiments showing that something like this is probably happening in the brain. These are a couple of examples. And since they are... I was looking for the best example so that they don't come from my lab at all. This is from Lars Muckli in Glasgow. He has shown that, using advanced brain-reading techniques, which I won't describe, you can decode the context of what a person sees from parts of the visual cortex that don't even receive any input. What's more, you can decode better when you do it from the top of the crust, which is supposed to receive predictions from higher levels.

This suggests that there are predictions that are feeding each other. And in another study by Andre Bastos and Pascal Fries in Germany, they used a method called Granger causality, which is sensitive to the flow of information in systems. And they find that top-down signals and bottom-up signals are transmitted in different frequency bands in the cortex, which is what would be predicted from predictive coding. One last experiment that I find particularly interesting is one conducted by Masanori Murayama's Japanese group. And they were used to optogenetics, which is a way of using lights to selectively turn neural circuits on or off.

And in this experiment they showed that simply turning off the upper superficial levels of the somatosensory cortex in a mouse brain, the part of the mouse brain that is sensitive to touch, could affect the mouse's ability to make tactile discriminations. Those top-down connections came from a motor cortex. There's a lot of evidence that top-down connections in the brain are important for perception, that's the basic message. But what's quite strange, and what I'll tell you next, is that this is all very good, but predictive processing is not a theory of consciousness. None of what I have said has anything to do with consciousness, at all.

It has to be... it's a very general theory of how brains do what they do. How they perform perception, how they perform cognition, how they perform action. Somewhat counterintuitively, I think that's exactly why it's a great theory of consciousness. And the reason I believe this is because it allows us to ask all kinds of questions about the real problem. About what it is, what's going on in the brain that underlies what you're aware of right now, without getting sucked into the metaphysical hole of why you're aware in the first place. In other words, it provides a powerful approach to look for neural correlates of consciousness, those things in the brain that go along with being conscious.

Because now we can take advantage of a very general theory about how the brain does what it does, instead of just looking at this or that region. So what specifically does predictive processing or the Bayesian brain say about consciousness? Well, many years ago, some influential experiments revealed a very strong connection between top-down signals and conscious contents. In this example by Álvaro Pascual-Leone and Vincent Walsh, what they did was they had people observe visual movement, examples of visual movement. And they used TMS, this intervention technique where you can deliver a shock to the brain very briefly. I mentioned it before.

But they used it here, specifically to disrupt the top-down signaling that was evoked by this perception of visual motion. And the result was that if top-down feedback was specifically disrupted, conscious perception of visual motion would be abolished, even if bottom-up signaling was left intact. So that was one of the first keys. Now, more recently, in our lab and in many other labs everywhere, we've been asking some other questions about the relationship between what you expect and what you consciously experience. One of the most basic questions you can ask yourself is: do we consciously see what we expect to see?

Or do we see something that violates our expectations of what we expect? And a recent study from our group, led by Yair Pinto, used a method called continuous flash suppression to address this question. It is illustrated here. You see different images in different eyes. In one eye you see this rapidly changing pattern of Mondrian squares. And in the other eye you see a face or a house. And they change the contrast like this. So initially the person will just see this random pattern and then they will see a house or a face. And you just, you just ask them to wait to see... you just tell them that it's more likely to be a face or a house.

And what we found in several studies is that we see faces more quickly when that's what we expect to see. It may seem obvious, but it could be the other way around. At least in these studies, we see what we expect to see, not what violates our expectations. That's the data. And the same goes for houses. These types of studies support the idea that it is top-down predictions that are really important in determining what we are aware of. There is another experiment that I will only mention. We did practically the same thing. This is called motion-induced blindness.

If you are in a lab instead of a conference room and you stare at this central dot, then this red dot might disappear from time to time. And what we did was after it disappeared, we changed its color and made people expect the color change to be one thing or another. And again, it reappeared more quickly if it changed color as expected. Again, I am confirming that once your expectations were validated, it accelerated your conscious awareness of something in the world. That's just behavioral evidence. That's simply asking people what they see and when they see it.

We have also become interested in the brain mechanisms that underlie and shape how our expectations, what we consciously see, change. And we've been particularly interested in something called alpha rhythm. Now, the alpha rhythm is an oscillation of about 10 hertz or 10 cycles per second. This is especially prominent in the visual cortex, at the back of the brain. In a study, led in this case by a PhD student, Maxine Sherman, with Ryota Kanai, at Sussex. What we did here was manipulate people's expectations about what we're likely to see. And it was a very boring experiment. The only thing they could see was what we call Gabor patches.

They are simply patches of very faint lines that are blurred at the edges. But the visual system loves this kind of thing. They activate the early visual cortex very, very well. And people expected that there would be a patch there or that it wouldn't be there, in different conditions. And while we were doing this we measured brain activity. And to summarize, what we found was that there were certain phases, certain parts of the cycle, this 10 Hertz cycle, where their expectations had a greater effect on what they said they saw. So there was a part of this cycle, like the alpha rhythm, there was a part where your expectations dominated your perception.

And there was another part that was the complete opposite, where your sensory cues were more important in determining what you were seeing at that moment. This suggests that this oscillation in the back of the brain is orchestrating this exchange of predictions and prediction errors. And that's the kind of cycle that could be the neural mechanism of conscious vision. And other theories about what the alpha rhythm does, there are many. One is that it doesn't do anything, it's just the brain inactive, and I think this is at least a more interesting way to think about it. In another experiment we did with another PhD student, Asa-Chang, we showed people these very rapidly changing luminance sequences.

And it turns out that your brains will learn to predict the specific changes in these sequences that change you very quickly. And the signature of this learning, again, is in the alpha rhythm, and it suggests that this oscillation has something important to do with the way the brain learns and updates predictions about sensory signals. But we don't go around the world looking at Gabor patches or quickly changing things like this. We go around the world looking at people and objects. And that is what our visual world is made up of. So can any of these ideas say something about our everyday visual experience?

And I think that's a very important challenge that we need to overcome in neuroscience. Get out of the lab and think about real-world experiences. So we've been using virtual reality for the last few years to try to come up with some of these ideas. This is an Oculus Rift, which I believe is now available for purchase. And we've been using them to address some of these aspects of real-world visual perception. And one of these aspects of the real world is called perceptual presence. And this is the observation that, in normal perception, objects really seem to be there, rather than being images of objects.

And this is, of course, what Magritte plays with in his famous painting, The Treachery of Images. For example, this is an object. I think it's there, and in a sense I can perceive the back, although I can't see the back, although the back is not giving me any sensory input, I perceive it as an object with a back. How do you explain that? Well, one idea that may occur to you within this Bayesian brain framework is that the brain doesn't just predict the possible causes of sensory signals arriving here and now. But it also predicts how sensory signals would change if you performed certain actions.

If I took this object and moved it, or just moved my eyes from one place to another. There is a long article. I wrote about it and... please don't read it. - But that's the basic idea. And how do you test an idea like that? That's why we've been using some innovative virtual reality methods, or augmented reality methods, with my postdoc, Keisuke Suzuki. And what we do is we have virtual objects, and these virtual objects behave like a normal object would. They are all strange and unknown objects. But they can behave like a normal object would, so you can learn to predict what would happen.

This one is weird. It always shows you the same face, a bit like having the moon on a plate in front of you. And then there are other conditions where objects respond to your movements, but they do so in strange and unreliable ways. So the question is: what does the brain learn about these objects and how we experience them? Do we experience them as objects in different ways when they behave differently? And we're still doing those experiments. Another way we can use virtual reality is to investigate what happens in visual hallucinations of the type experienced in psychosis and other more pharmacologically induced conditions.

What we're doing here is combining immersive virtual reality—imagine you have a headset strapped to your head again—with intelligent image processing that models the effects of hyperactive backgrounds on visual perception to generate a highly immersive experience. This is actually the Sussex campus, but it looks quite different now than it did at lunchtime today. I'll tell you that. What we have done is record this panoramic video that we can feed back through a virtual reality helmet. And we've run this video through one of these Google deep dream algorithms that you may have seen, which can generate sort of animal-like bowls of pasta.

And this can seem like a lot of fun. It's fun, but there's a serious purpose here, because it allows us to model the effects, to model very unusual forms of perception and how they might play out in different ways in the visual hierarchy. And I think understanding how visual hallucinations can occur and the broader effects they have on the mind is a very important part of the study of visual perception. This leads us to the second conclusion, which is that what we consciously see is the brain's best guess about the causes of its sensory information. Normal perception is a fantasy limited by reality.

Now, before moving on to the last section, I want to pay tribute to an unlikely character in a talk about neuroscience, which is Ernst Gombrich. ErnstGombrich was one of the most prominent art historians of the 20th century. And it turns out that Gombrich's approach to understanding art had a lot in common with Bayesian ideas and the Bayesian brain. And more specifically, with the idea that perception is largely an act of imagination or construction on the part of the perceiver. And this is most evident in his concept of the viewer's part, which emphasizes that the viewer brings so much to the act of experiencing a work of art.

As he put it in his 1960 book, Art and Illusion, "the artist gives the viewer 'more to do,' drawing him into the magic circle of creation and allowing him to experience some of the thrill of making that was once his privilege." of the artist." I think to me this is very powerful when looking at, especially, things like impressionist art. And here, one way to think about this is that the artist has reverse engineered the entire process of perception, so that what is there are not the objects, the end points of perception, but the raw materials; the light patterns that involve our perceptual machinery in doing its job.

And for me, this might be why paintings like this are particularly powerful. Now, the final dimension of consciousness that I want to talk about is the conscious self. The fundamental experience of being someone. Be someone like you. There are many aspects to our experience of being a conscious self. There is the bodily self, the experience of being and identifying with a particular body. A little piece of the world is with you in the world all the time. There is the perspective self, the experience of seeing the world, or experiencing the world, from a particular first-person perspective, usually somewhere in the body, but not always.

There is the volitional self, the experience of trying to do things and making them happen in the world of agency. And these ideas, of course, are often associated with concepts of will. Then there is the narrative self. This is where, only so far we don't have to worry about the concept of self, but when we get to the narrative self, there is now and self. There is a continuity of experience of the self from hour to hour, from day to day, from month to month, and from year to year, with which you associate a name and a particular set of autobiographical memories.

And finally, there is a social self. The way I experience being me depends in part on the way I perceive you perceive me. I am only going to talk, in the remaining minutes, about the bodily self. This is something we are working on a lot at Sussex. The experience of identifying with and possessing a particular body. And the basic idea I want to convey is, again, very simple. It's just that we should think about our experience of possessing a body the same way we think about our experience of other things. That is, it is the brain's best guess about the causes of body-related signals.

And the brain is always making this inference. It is making inferences about what in the world is part of the body and what is not part of the body. But it has access, in this case, to other types of sensory signals, not only visual or tactile signals, but also proprioceptive signals. The orange arrows here. These inform the brain about the configuration and position of the body in space. And then also, and often overlooked, are the interoceptive signals. These are signals originating from inside the body that inform the brain about the physiological state or condition within the internal physiological environment.

And you can think of the idea as being that our experience of embodied identity is the brain's best guess about the causes of all the signals together. Yes, that's just to emphasize interoception. An important part of this idea is that interoception, the feeling of the body from within, should work on the same principles, the same Bayesian principles that we were thinking about earlier: vision and hearing. That is, our experience of the inside of our body is the brain's best guess about the causes of the signals coming from inside our body. So we can think, again, of top-down predictions that contain predictions about what the body state is like, and bottom-up prediction errors that report differences between what is happening and what the brain expects.

So what is our experience of the inside of our body? Well, back in the early days of psychology, William James and Carl Langer proposed that emotions, emotional experience, were actually about the brain's perception of changes in its physiological state, rather than its perception of the outside world. So in this classic example, seeing a bear does not itself generate the experience of fear. Instead of seeing the bear take off, a series of physiological changes prepare for fight and flight responses. And it is the perception of these bodily changes in the context of the bear's presence that leads us to experience fear.

So the Bayesian perspective simply generalizes that idea and says that emotional experience is the brain's best guess about the causes of interoceptive signals. And this fits very well with a lot of evidence. And this is just one study done by a Finnish group. And all they did here was have people report where in their bodies they felt various emotions occurring. And then you feel anxiety in a part of your body. You feel fear in another, and so on. So our experience of emotion appears to be intrinsically embodied. Now, another part of our experience of being a body is the body as a physical object in the world.

And this may seem quite easy to take for granted, since our physical body is always there. It goes with us, it changes over the years, unfortunately. But it's always there. But it would be a mistake to take our experience with ownership of an organism for granted. And there are some classic experiments that demonstrate how malleable our experience of body ownership is. This is the famous rubber hand experiment. Some of you have probably seen this. What happens here is that a volunteer has his hand hidden under a table, and the fake hand is placed on top of the table, and then both hands are stroked simultaneously with a paintbrush.

And it turns out that simply seeing a hand-like object where a hand might be, feeling the touch, and then seeing the object being touched, is evidence enough that the brain's best guess is that the fake hand is, in fact, part of my body. . Some kind of part of my body. This is what it looks like in practice. Here you can see the fake hands, focusing on them. There are the real hands, without focusing on them, the simultaneous caresses... and there are several ways to try it. - I found that doing this works even better on children, by the way, if you do that.

That's interesting, because visual and tactile cues are used to convince the brain that this object is part of your body. In my lab, we've been interested in whether these signals from inside the body also play a role. So we created a virtual reality version of this rubber hand illusion, where people wear these glasses and see a fake virtual hand. And we also record your heartbeats. And now what we can do is make the virtual hand blink in time or out of time with your heartbeat. And we asked the question: Do people experience this virtual hand as belonging more to them when it flickers in time, rather than out of time, with their heartbeat?

And the answer is yes. And this is just some data, basically that, bigger than that, which means that they actually experience the hand as more of their own. The way we actually measure that is we can ask them first. That's the easiest way. Then we can also ask them to point to where they think their hand really is, and we can see how far they go from where their hand really is to where the virtual hand is. And that provides a more objective way to measure the strength of the effect. This is what it looks like in practice.

Again, if you can see this, that's the real hand. That's someone's virtual hand. Again, imagine you're wearing headphones so you can see it in 3-D. And I hope you can see it by clicking to read and return. And you can also do other things with these virtual reality rubber hands that you couldn't do with real rubber hands. For example, you can map the movements of the real hand to the virtual hand, so you can start asking questions about how far the virtual hand moves while predicting that it will move. How much does that affect the degree to which it feels like a part of my body?

You can make it change color. So you can have someone embody a skin color associated with an outgroup cultural group and see if they become less racist as a result. And then my favorite is where you can actually change the body size. And that comes here. So what we do here is we can scale up and down the size of the manual telescope. And again, this may seem like fun, and it is, but it has a serious purpose. There are several conditions. In fact, there is a condition called Alice in Wonderland syndrome, in which people report that all parts of their body do, in fact, increase and decrease in size.

And in a more subtle way, there are many body dysmorphias, subtle misperceptions of body shape, that could be associated with eating disorders. And so, these types of techniques allow us to address, in a very detailed way, how people can misperceive their own bodies. That brings me to the third important message about self. And with apologies to Descartes, the message we should take away is that I predict myself, therefore I am. In the last nine minutes, before the bell rings, I want to come full circle and return to this Cartesian idea of ​​the beast machine. To try to convince you that our experience of being a conscious self is intimately linked to our beast-machine nature.

And to do this, I need to mention one last aspect of perceptual inference, which has a lot to do with Karl Friston, who has done a lot of work on the UCL Bayesian brain here in London. And if we think that the brain is responsible for minimizing prediction errors, this can be done by updating our perceptual predictions, which is what I have been talking about so far. And this is what Helmholtz said. Or we can minimize prediction errors by taking actions. We can change what we predict, or we can take an action to make our predictions come true.

You can change with sensory information, or you can change what you believe about your sensory information. One point of doing this is that you can take actions to discover more about the world around you. And this is what Helmholtz has in mind when he says that every movement we make, by which we alter the appearance of objects, must be regarded as an experiment designed to test whether we have correctly understood the invariant relations of the phenomena before us. . Which Gregory, much later, said something similar when he talked about perception as hypothesis testing. The point of this is that we make eye movements and other types of movements to understand what the world is like.

That, indeed, there was tomato there, for example. But there is another way to think about active inference, and that is that, when we minimize prediction error, what we are really doing is controlling a sensory variable. We are preventing it from changing, because we are making our prediction about what it is come true. And this is the use of active inference, to control or regulate something, rather than understanding what the causes of that something are. And this brings to mind a very different tradition: 20th century cybernetics. And this is Ross Ashby, who was a pioneer of this way of thinking.

And he, with Roger Conan right at the end of his life, wrote an article. The title of the article was "Every good regulator of a system must be a model of that system." The idea here is that if you want to regulate something very precisely, then you need a good model of the effects of that system. Now you could also apply this idea to the outside world. When you try to catch a cricket ball, you are actually trying to control the level of the angle above the horizon. But I think it applies more naturally to the internal state of our body.

So really, what matters about my internal physiological condition, I really don't need to know exactly what the inside of my body is like, nor care about it. But I need to control it and my brain needs to regulate it. Therefore, this way of thinking about active inference applies most naturally to interoception. Think about it this way. Having good predictive models is always useful, but we can have a pendulum that swings, on the one hand, towards control. We can use these predictive models for control, and that applies more to the state of our internal body. Or we can turn the other way and think about perception, understanding.

We might think about instrumental and epistemic ways of thinking about the role of action and perception. And this reminds me... I mentioned Karl Friston. He came up with something called the free energy principle. And I can only salute the tremendous amount of work he has done here on this topic. With the motto that organisms, in the long term, remain inthe states they expect to be in, by virtue of having good predictive models about their own internal condition. This brings us back to Descartes, but in a very different way. As I said right at the beginning of this lecture, for Descartes our physiological reality was quite irrelevant to our minds, our rationality and our consciousness.

This is a quote from a 1968 article about his beast machine argument: "Without minds to direct their bodily movements, animals must be regarded as thoughtless, unfeeling machines that move like clockwork." Now I think if you try to think about how this idea that our predictive models control our internal physiological states and the resulting experiences, that perceptual content that might generate, you can argue the opposite. And the opposite case would be that conscious identity arises because of, and not despite, the fact that we are beast machines. And I think this is a deeply embodied view of consciousness and self, and it speaks to this fundamental link of continuity between what it is to be alive, what it is to have a mind, and what it is to be a conscious self.

So, again, the third conclusion should make even more sense now. I predict myself, therefore I exist. And I am a conscious self because I am a bag of self-fulfilling predictions about my own physiological persistence over time. Now, why does all this matter? These are a lot of interesting ideas, but why should we be interested in studying consciousness? Well, it is something very interesting, I hope I have convinced you. But there are also many practical reasons to be interested. In the United Kingdom alone there are between 20 and 60,000 patients suffering from disorders of consciousness. You are in a vegetative state, in a coma, or in some other seriously abnormal state of consciousness.

Having better measures of the conscious level, as I described at the beginning, is really going to change again how we treat people like that. And of course, in psychiatry. Psychiatric disorders are increasing in prevalence in all modern societies, with an estimated one in six of us, at any given time, suffering from a psychiatric condition. And understanding the mechanisms underlying conscious content and conscious self, because many psychiatric disorders include alterations in the way we experience our body, although that might not be the most obvious symptom, can help us understand the mechanisms involved in psychiatric disorders. . not just the symptoms.

There are also some more general reasons to study consciousness, which raise some ethical questions. When does consciousness arise in the development of conscious newborns? Or does consciousness even begin in the womb? Perhaps different dimensions of consciousness arise at different times. Are other animals conscious? Well, I think that may be a good argument for mammals and primates, but what about the octopus? The octopus has more neurons in its arms than in its central brain. They are very intelligent creatures. Here you have to ask yourself not only what it is like to be an octopus, but also what it is like to be an octopus arm.

And finally, with the rise of artificial intelligence, we should start asking questions about what it would take for a machine to have some kind of subjective experience. I don't think we're close to that yet, but we should consider what science can tell us about its possibility, because that would raise some very, very difficult ethical questions. But fundamentally, consciousness remains fascinating to me for the same reason it has motivated people throughout the centuries. I mean, Hippocrates, the founder of modern medicine, put it one way. He said that from the brain and only from the brain come our sorrows and our joys.

And he also had his first insight into psychiatry that madness comes from dampness. And then Francis Crick, in the 1990s, who I mentioned at the beginning. He gave rise, if you will, to the modern neuroscience of consciousness. He said almost the same thing in his surprising hypothesis. But there is still the mystery and wonder of how the biological machinery inside our heads gives rise to the rich inner universe that makes life worth living. And despite this mystery, modern science is advancing. I hope I have given you an idea, although we do not understand how consciousness occurs, we can begin to understand its mechanisms.

Therefore, we should not be afraid of naturalizing consciousness. It is not bad to understand its basis in the material world. As is often the case in science, with greater understanding comes a greater sense of wonder and a greater realization that we are part of the rest of nature and not separate from it.