The Neuroscience of Consciousness - Anil Seth

two of the three fundamental mysteries about our place in the universe have already been solved the first was literally about our place in the universe this is a photograph taken by Voyager 1 on its journey out of the solar system from a distance of about I think six billion kilometers and all of human history in fact all of the history of life on this planet took place on that pale blue dot that's the famous pale blue dot photo that's the Earth from far away now this just emphasizes the point that we've known since Copernicus that we are not at the center of the universe we're just a tiny speck suspended somewhere out there in the abyss yeah we're not so central the second mystery is about our relationship to other forms of life and Darwin pointed out that again we're not so special we're related to all other creatures and we're just some branch or some twig of a beautifully rich and delicate evolutionary tree sharing much of the machinery of life with even the simplest of our fellow creatures and the third mystery is the mystery of our inner universe and that's the mystery of consciousness now consciousness for me is always been absolutely fascinating it's the one thing that cannot be an illusion we could be mistaken about everything else we could be mistaken about the existence of the external world at all what it's made of whether it's there it can be mistaken I could even mistaken about you know not wanting brexit to happen but yeah I could be mistaken about such things that's so fundamental but I can't be mistaken about the fact that right now I'm having a conscious experience people might want to question that but we'll hold questions at the end if you disagree with that we'll be in trouble I mean I think well I'm glad about that even if it's seen but you'd be surprised actually that that quite a number of philosophers have spent a lot of ink trying to argue that we can be mistaken about the fact that we are conscious which i think goes down as one of the the sort of silliest sort of forms of philosophy that I've come across even if it seems like we're conscious that's enough we have I don't know whether you are but I certainly know that I am conscious right now how does this happen this is another of the central mysteries that we face and it's similar to the other mysteries in the sense that's challenging in some way our perceived importance in the universe which is one reason I think this is something to bear in mind so there's one reason why people often resistant to a scientific account of consciousness because just like Copernicus are just like Darwin it sort of naturalizes what we find no special about ourselves and some people don't like that and I think by the end of the talk I'll hope to convince you that that's not something to be worried about it's something we should rejoice it but this is the third mystery how does consciousness happen you know somehow within each of our brains the combined activity of many billions of neurons is giving rise to a conscious experience and not just any experience your experience right here and right now how does this happen why is life in the first person so the brain is really remarkable object there's anything that's going to do something remarkable maybe it's the brain it's not so much the number of neurons there are about 88 billion at the last count it's not even really the number of connections there there are so many connections in the brain that if you counted one every second it would take you about 3 million years to finish counting it's more the intricate patterns of connectivity that are still not fully known but within which are inscribed everything that makes you you now you sometimes hear and I often hear that we know nothing about how the brain and body give rise to consciousness that there's this uncrossable explanatory gap between conscious experience in the physical world the origin of this view is often traced to Descartes who back in in the Enlightenment some hundreds of years ago he divided the universe into two things on the one side there's res Extender the stuff that things are made material stuff the stuff that tables chairs and made of the stuff that bodies and indeed brains are also made of and on the other side there's res Kaja tans the stuff of thought of the realm of the mental and the realm of conscious experience and by dividing the universe into this way he generated the problem of how you put them back together again the philosophy of dualism and nothing was ever the same or simple since then and they cannot so in preferred solution some of you may have heard was the pineal gland the pineal gland is this part of the brain deep in the middle of it that were supposed to be this interface between these two realms of existence now this is a kind of a strange idea certainly the pineal gland is reasonably important but it's not the seat of the soul probably the reason Descartes chose it was that unlike most parts of the brain there's only one of them because it's right in the middle most of our brain is sort of you know we have left and right hemispheres you find structures on both sides so if you're looking for a place where one realm of the universe interfaces with another it's sort of parsimonious to only have one of them rather than two but that's probably the only good reason if it's a good reason for the pineal gland but this view then that this dualistic view that on the one hand you have consciousness and mind on the other hand you have physical material this has persisted in science and more laterally in psychology in neuroscience for a very long time I started my undergraduate in the early 90s and just before I started fortunately I didn't read this at the time in the international dictionary of psychology the international dictionary of psychology Stewart Sutherland who is the founding professor of psychology at my current University of Sussex said this he said consciousness is a fascinating but elusive phenomenon it's impossible to specify what it is what it does why it evolved nothing worth reading has been written on it this is quite depressing state of affairs to be sort of canonized into the in the dictionary of psychology but this indeed was the environment at the time and I remember studying psychology and neuroscience in the early 90s and in Cambridge and it was really not allowed to discuss consciousness yeah maybe you could talk about memory and attention a perception but not consciousness that's philosophy or religion and this is so now that these these cartoons I am very grateful to jolly entre shank oh he's done always beautiful cartoons about neuroscience this is another sort of idea about the impossibility of saying anything sensible about consciousness that you can talk about and maybe hopefully hopefully this won't happen by the end of this talk but you can have somebody talk about as much about neurons as they want but there will still be some magic process by which it all all chemically is transformed into the redness of red or the sharpness of the pain or the actual experience of what the fundamental nature of a conscious experience and I'm not saying this mystery has gone away at all but what I want to convey during this talk is that we shouldn't worry too much about this mystery about how consciousness yeah why it's part of our universe in the first place so long as we can do a good job of explaining its property scientifically and in indeed this is the time to change in over the last 25 years there's been a rehabilitation and resurgence and now first a trickle and now quite a deluge of scientific work on consciousness this is my lab from a year or two ago if you come through of my lab and many others around you find scientists of all different disciplines into now said it nice introduction meaning we work with physicists mathematicians virtual reality engineers neuroscientists psychiatrists and philosophers all trying to understand how consciousness works and what happens when it goes wrong get to this at the end but studying consciousness is not only interesting because it's one of the greatest remaining mysteries it's also I believe of enormous practical importance for the insights that can give us into things like depression other psychiatric disorders which for the for the moment for the most part don't benefit from a very rich mechanistic understanding of what's going on in the same way that other branches of medicine deep thoughts feel and the approach that we take and the approach I want to argue for today is to treat a consciousness a little bit like life now this isn't an analogy that's been used many times it's an imperfect analogy but I think it's still valuable now it wasn't that long ago that eminent biochemists thought that life could never be explained in terms of mere mechanism that have much you understood about a material system there's something special very special about living systems that required some special source so an alum beetle some spark of life that would account for the difference between the animate and the inanimate but as biologists got on with the job of accounting for the properties of living systems things like metabolism reproduction and homeostasis in terms of physics and chemistry the basic mystery of what life is started to fade away now we don't understand everything about life people don't really know how a cell works but this basic metaphysical sense of the unknown that we that life is in principle beyond mechanistic explanation now that has faded away people no longer think that that mystery is as mysterious as it once was and the hope is that as with life the same thing will happen with consciousness that as we start to account for its properties in terms of things happening inside brains and bodies the basic mystery of what consciousness is will also start to fade away and people will worry about it less and the reason I mean I'll give you some examples of how this is actually happening but then and I'll return at the end I think to some of the ways in which people still resist this this approach why do we still think there's something you know there's going to be some residue left I mean if you think about it even physics physics is prob the best example of a science that's able to shed enormous insight into the way things are in in ways that are very counterintuitive but also very predictive very explanatory but physics still doesn't tell us why there is a universe versus not it's still you know we don't really know what matter is there are still some basic things in the domain of physics that remain mysterious but that hasn't stopped physics revealing a great deal about the nature of the world we live in and I think the same will turn out to be true the consciousness now there's another way to put this which I want to flag since some of you may have heard of this and this is a sort of the the the more recent incarnation of the Cartesian divide that's divided into res extends and res khajit and this is between what people have been calling following the Australian philosopher David Chalmers the easy problem in the hard problem how many people have heard about this this distinction easy problem hard problem a few of you okay can you say what psycho synthesis okay we can talk about that later so for Chalmers he he in a very influential set of papers about twenty years ago he distinguished between the so called easy problems of consciousness and a hard problem now the easy problems are not easy at all they're incredibly difficult but they're basically the problems of figuring out how the brain does what it does how it works how it's complicated circuitry under writes all the functional capacities that we humans and other animals have how we perceive how we use perception to guide action how we make decisions things that you can measure from the outside what reports we would say about what we see but basically the easy problem is about figuring out how brains do what they do in ways which we can observe validate from the outside as external observers the hard problem is figuring out how and why any of this should have anything to do with conscious experience at all you know that magic step why are the lights on than off how do conscious experiences qualia as it philosophers say why and how do they come about given any kind of explanation of the activity of neurons or neurotransmitters or whatever level of detail you want to go into how and why should that ever generate the redness of red that's that's the hard problem for David Chalmers and the intuition here is that solving the easy problem figuring out how brains do what they do would leave us completely in the dark about the hard problem no explanation in terms of mechanisms neurons or anything would shed any light on this hard problem of how my consciousness happens now there's already this analogy with life hope you can see that it's it's pretty I think unproductive to make claims about what we still can't understand when we haven't done the hard work of trying to understand and it also Cleaves the problem in an unhelpful way so with sort of slightly to annoy David Chalmers I've been talking about what I call the real problem of consciousness the real problem is the problem of accounting for phenomenological properties that's properties of consciousness in terms of mechanisms so it's not the easy problem because the easy problem focuses on behavioral and functional things things you can measure from the outside so I want to try to explain properties of what conscious experiences are like know what's the difference between a visual experience and an olfactory experience and an emotional experience and an experience of free will or volition so it's targeting properties of consciousness and it's not the hard problem either because in order to do this it's not necessary to explain how and why consciousness comes to be part of the universe in the first place we know that it exists we're all conscious so we can map its properties onto properties of mechanisms begin to explain them if you can explain predict and control you've basically done everything that a scientific approach can do so there is one of the things you can maybe look at if if you're interested after the talk is there's a sort of public article in Eon which is a online magazine called the real problem which so argues for this in a bit more detail but the question that's motivated me in my research over the last couple of decades is how can the structure and dynamics of the brain together with the body and environment account for the subjective phenomenological properties of consciousness so that's the question so we're trying to map mechanism not just of the brain but also appreciating that if we want to understand anything really about mind in general but consciousness specifically brains don't exist in isolation they're continually interacting with the body which is embedded in the world brains are embodied in bodies are embedded and we need to understand these these interactions - how can those interactions shed some light on what it is like to have this experience or that experience so that's the question and what then are the properties of consciousness that we might like to explain what are the explanatory targets now there's many different ways you can you can skin this cat and for today I want to just divide it into two three and this this is a heuristic division I'm not saying these aspects of consciousness are completely independent or that this is a consensus or the only way to do it but this is the way that will provide us a nice narrative for this talk conscious level we can start with conscious level conscious level is simply the difference between being conscious at all and not being conscious as is the case in general anesthesia or coma or dreamless sleep you can think of you know a level of consciousness that you gain at some point and then maybe there are different gradations of kind of level of consciousness when you are conscious drowsiness compared to you vivid or work vivid awake alertness then there's conscious content when you are conscious you're conscious of something the sights sounds smells emotions thoughts beliefs that populate your own universe at any one time not everything that our brain senses is reflected in our conscious experience of the world in fact what I'll argue for is that our conscious experiences are only indirectly related to things in the world in interesting ways and then finally and I think most interestingly is conscious self one of the things you are conscious of when you are conscious is the specific experience of being the subject of experience the specific experience of being you or of being me this isn't something to take for granted that sits behind all the rest of our experiences this is another experience itself that has a particular character and understanding of that I think is the key is where people often encounter the most resistance for instance if I try to give you an account of how and when and why you experience certain actions as voluntary that is getting for many people to the heart at what they think should resist any kind of scientific explanation because that's the core of them you know the origin of voluntary action of free will but if we can begin to understand that then I think we get to the real implication of a science of consciousness which is not to explain why red things look red but how we experience ourselves as part of the world around us and our role within that world so that's where we're going and we'll start with conscious level what are the fundamental brain mechanisms that underlie being conscious at all I mean we can think of this as I've said as a scale at one end with zero consciousness let's say in coma or general anesthesia through deep sleep light sleep all the way up to conscious wakefulness you can think of it yeah this is an approximation but to a first approximation we can think of a single scale here but you can see this is along a diagonal because the first important thing about conscious level is it's not the same thing as physiological wakefulness or physiological arousal it's often correlated but it's not the same we notice when you are asleep you're often dreaming so you're having conscious experiences but you're physiologically asleep in physical arousal levels are low on the other hand there are pathological states like absence epilepsy or the vegetative state which is now called the wakeful and awareness state which you can go into if you suffer very bad brain injury where you will behaviorally you will go through cycles of sleep and wake your eyes will open you look around you may make odd movements but there seems to be nobody home seems to be completely without conscious experiences at all so the mechanisms that underlie being conscious at all not the mechanisms they're separable from those mechanisms that control whether we are awake or asleep and this is of course clinically very important because one of the big challenges I won't have much time to talk about this but with the developments of intensive care people are surviving injuries much more frequently that would otherwise kill them and ending up in what the neurologist Adrian Owen has called the grey zone where they are diagnosed as being in a vegetative States and often it's unclear whether they have any residual awareness or conscious experiences at all and of course this is something that we really need to know because it informs their treatment and it informs how one deals with their families and so on and so forth but anyway we need to know what the mechanisms underlying conscious level are and so you might think there's a few simple things to consider first how about the number of neurons is being conscious simply a function of the number of neurons involved well probably not I've said we have about 88 billion in our in our brains this is the cortex I'm sure you notice by the end of the day now it's cortex here this is the cerebellum anybody know how many neurons are in the cerebellum compared to the rest of the brain half much more it's about four times more neurons in the cerebellum than in the rest of the brain put together which is a fact that I still think is extraordinary and the cerebellum is a little brain hanging off the back of your brain it has the vast majority of all the neurons that you have now if you have damage to your cerebellum lose it somehow through a stroke or there are rare cases of cerebellar agenesis where people don't the cerebellum at all now they have lots of problems in moving around in sequencing their actions even in sequencing their thoughts but they are not unconscious it doesn't seem to affect consciousness at all so it's not simply the number of neurons you can lose three-quarters of the neurons in your brain and your conscious experiences will be roughly to a first approximation unaffected at all is it any particular region are there you know is there a seat of consciousness somewhere in the brain if it's not the pineal gland is it somewhere else well again probably not there are parts of the brain that if you suffer damage to them you will lose consciousness forever certain nuclei deep inside the thalamus what you can't see here is there's a set of this is the cortical surface but there's part of the brain called the thalamus which is a sort of relay station from the outside world to the brain in between different brain regions - if you if you have damage there you will lose consciousness forever but it seems to be that these are more like you know unplugging the TV it's not going on off switch rather than that's where you know consciousness happens if you have damage to other parts of the brain you may lose specific dimensions of your conscious experience for example the inability to experience colors or the inability to experience faces is different from one another or movement and Oliver Sacks writes about this beautifully in his various the neurological case studies but again you won't lose consciousness altogether of course if you wipe out large parts of the brain then yes you do that's what happens when people end up and coma vegetative state but no single region seems to be responsible and it's not even a simple form of neural activity now when you fall asleep and lose consciousness yes metabolic activity and firing rates of neurons in the brain go down but they certainly don't shut off and they don't go down by very much actually at all and even sort of more slightly more interesting kinds of activity like if you look at the extent to which different parts of the brain are synchronized an early theory of consciousness talked a lot about synchrony that if different parts of the brain are firing together at let's say 440 Hertz the so called gamma band then you know somehow that leads the consciousness well in fact if you have too much synchrony in your brain if the brain all the neurons are firing in lockstep with each other what will happen is you'll have an epileptic seizure and lose consciousness it's not yeah synchrony is not the answer so what is the answer I'm going to sort of skip a lot of intermediate research and bring you up to do some exciting stuff in the present day and it seems to be to put it very abstractly being conscious it all has to do with the different ways in which different parts of the brain speak to each other it's to do with the patterns of information flow causality and communication between different parts of the cortex and one way we can begin to look at that is by using techniques such as this this is a lizard my lab at Sussex and this is my lab manager who's being a guinea pig here and what we do here is we use transcranial magnetic stimulation how many people know about this this techniques called TMS for sure what it is it's basically just the way of injecting a very short sharp pulse of electromagnetic energy into the brain and then this is he's also wearing an EEG cap electroencephalography which measures the the electrical activity of the brain now does so at a very coarse way these are individual sensors if you think about the complexity of the brain that I've described now 90 billion neurons all wired up this the this is sort of it's been described as trying to figure out how the brain works by you know dangling and microphone above London from you know about two kilometers up and trying to figure out what everybody is saying to each other it's it's not very very easy you can probably pick something up if everybody shout something at the same time so you might pick up football games and I annoyingly realize that this schedule this talk to clash with United against Arsenal which so really annoyed about but anyway never mind it's another story so what you can do with this kind of setup is you can you can inject a pulse and then basically record the echo of that pulse a bit like banging on the brain with an electrical hammer and listening to the the echo the Bell like echo of the brain after you've stimulated it with this with this pulse and you can sort of stimulate different parts here and there in interesting ways so this line of approach this was pioneered by a colleague a friend of mine Masumi knee in in Milan and what he did his first study was to just compare people who were in early stages of sleep so not dreaming and people who were awake and what you'll see in these movies which is very they're very heavily slowed down you'll see the TMS pulse and then you'll just see a movie of the echo as it what happens to it over the cortical surface and so you can see the pulse if you look at what's happening when the person is asleep it will it will leap around there's still a response but it stays very locally to the state of stimulation but when the person is awake the pulse moves around in complicated spatial and temporal patterns across the brain yeah this is like this is about 200 milliseconds so still a fifth of a second these things happen quite quickly here we go again so you can see this there is a response but it doesn't move around the brain very much and when you're awake it does what this suggests is that in the conscious awake state there is much more functional connectivity or what we call effective connectivity in the brain that is different parts of the brain when active can cause activity in many other parts of the brain but again the whole thing doesn't just sort of take off you don't have an epileptic seizure if there's some sort of precision to it so what's happening there and what's been very exciting about this is we can start to quantify the difference between these two patterns put a number to it and this is what's shown the bottom here this is called the the PCI which stands for the perturbation complexity index this is important because what this is what this is it's the first version of the consciousness or meter of a thermometer for level of consciousness now it's still very crude but it's the first time that that people have actually been able to put a number that has some empirical robustness and and purchase what you can see here is that these are the PCI soup the way this number is derived is by basically you can imagine unwrapping this pattern over time so millisecond I should have had a figure for this but millisecond by millisecond you unwrap it by time and then every millisecond you've got a kind of pattern of which brain areas are part of the echo and which aren't you have basically this kind of complicated pattern of black and white that evolves over time what you want to know is how random is that yeah if the brain is not active at all it's it's a very simple thing it's just white if the brain is always active it's always it's black so to quantify how complicated that pattern is you basically try to compress it using an algorithm which is just the same as the algorithm that your phone will use when you send a digital photo it finds the minimum description length of an image how what's the minimum string of ones and zeros that you need to recreate a picture an image it will always be smaller than the image itself because in you know think about the digital photo or even this nearby pixels are always correlated it's only when things are different that you need to say something so you can come up with a number simply by asking how what's the minimum way to length of ones and zeros that you need to describe this pattern of activity that's the PCI and it's yeah it's called PCI because it's depending on a perturbation in its complexity because you're measuring the compressibility in its index because it's a number go between zero and one and what's plotted here is the PCI for various people who have been diagnosed as being in the vegetative state in the minimally conscious state which is when people show transient signs of consciousness after some manner of recovery after severe brain injury and in what's called the locked-in condition where people are completely paralyzed but nonetheless have completely intact conscious experiences those have you seen the diving bell and the butterfly or read the book will know about the locked-in syndrome and these are healthy controls who are either awake or asleep or having some ketamine going on we'll come back to that but what you can see I hope is that this number tracks the emergence of consciousness quite well across individuals now this is cross-sectional it's not longer to Dan'l you know what would need to be done is for a given person to see how their clinical state varies as they individually move along this scale but even so it's quite exciting that we have a number that is a first approximation of how conscious somebody is and they're interesting new this person here is behaviorally diagnosed as being in a vegetative state that their pci is basically consistent with them being conscious that's somebody you'd want to take a closer look at in the hospital so this is exciting developments have been around for now about four or five years we've been doing some similar work in the lab in Sussex not but without using the TMS impulse just recording the spontaneous activity of the brain and again applying the same sorts of algorithms we just want to measure how complex that signal is how how diverse how compressible it is this is like raw EEG from awake stat awaking restful state and from loss of consciousness in this case under propofol anesthesia again what we do is we convert all these brain signals into ones and zeros just to simplify them calculate their complexity how much how far can we compress this string of ones and zeros and what we see is a nice correlation between high values of complexity when people are conscious and awake and low values under general anaesthesia intermediate values from mild sedation now for those of you who know a bit more about this stuff this important thing about here is the these changes in complexity are independent of changes in the spectral power of the brain signals so we know for instance that when people fall asleep you see an increase in slow wave activity in the low frequencies but we can characterize changes in complexity that are independent of these changes in spectral power we've also done this in in sleep here this is a different kind of data set this is a intracranial data set so this is in collaboration with people in Milan and what's happening here as we were able to record from deep within the human brain these are patients who have been hospitalized because they need brain surgery to remove part of the brain from which epileptic seizures start so it's sort of last resort for intractable epilepsy and in order for this surgery to be done that we need to localize where the seizures come from so basically you implant a bunch of electrodes and just leave them in for you know usually a few days or even a couple of weeks until that person has had a sufficient number of seizures that you can figure out where the bit of the brain is that you need to remove but of course you can record from the brain all the time and people will go to sleep and wake up again so we have some data from deep within the brain this just shows over a number of patients where the electrodes were and we can sort of again measure their complexity and what we see is pretty much the same thing that complexity values are higher during waking rest than early stages of sleep when people are less likely to dream but in REM sleep rapid eye movement sleep where dream reports and those frequent it's actually pretty much the same as when people are awake so this is now teasing apart wakefulness from consciousness and we see the values here the same for consciousness but not for wakefulness this thing here these are four different parts of the brain the frontal part parietal either we divide it into the four main loads occipital temporal parietal frontal frontal parts of the brain always seem to have more complexity going on regardless and we don't still fit trying to understand how and why that is so in both examples I've shown you so far they've involved the reduction of complexity from a baseline of waking State different ways of losing consciousness whether it's brain injury whether it's anesthesia whether it's falling asleep in the most recent study we've done it's now just coming out to a year ago we worked with people in London Robin cohort Harris and his group at Imperial who've been pioneering the brain imaging of people while taking psychedelics psilocybin ketamine in low doses in high doses ketamine's and anesthetic low doses it's a hallucinogenic and an LSD of course and they have again they have in this case mmmm eg Magneto photography which is basically the same as EEG it's all measuring the define electric electromagnetic fields that are generated by neural activity but what we can see here quite interestingly is that the level of complexity reliably all these this is zero it always goes up so there's a sort of increase in complexity in the psychedelic state compared to the baseline for all the different compounds and this was kind of interesting because we've been looking for a while were there any situations where we would see a measure of conscious level go above baseline rather than go down this is so far still the only example that that we found and of course because it came from psychedelic drugs we got all kinds of amusing headlines like brain scans reveal evidence for a higher state of consciousness and which is of course not not true not really true you know why do we need science to do that you know of course we know that that hallucinogenics give you a higher state of consciousness what we've actually found is that on a particular measure of neural signal complexity or diversity that there's a reliable difference in the brain that this goes the opposite direction from sleep anaesthesia and so on there's basically more disorganization in the psychedelic brain the Daily Mail also rated up but being the Daily Mail they had to accompany it by a very large sidebar pointing to all the dangers of LSD which I do find in a much as I love the Daily Mail I find them which is not very much find it quite annoying because the work of robbins group here and other groups too and one reason we're interested in is that there is enormous promise in facade in psychedelic drugs for treatment of things like depression and PTSD and I think it's hugely it's a bit tragedy for medical science that these this line of research has been until very recently closed down and is still very much under threat we we're still continuing this work actually this is just to show you some of the stuff that we're doing right at the moment it's still on the same data but what we've done here with my colleague Lionel Barnett is we've looked in more detail at the dynamics of the brain in the psychedelic state and we found something which I still fight which I find very interesting I'm still struggling to understand which is that for each of the psychedelic compounds the brain divided up into 90 different small regions so again it's m e G data and here if we just look at the correlation in activity between the different regions how much the you know the brain signal moves together we see it goes up compared to baseline red is more so the brain regions are more highly synchronized in the psychedelic state which is weird but then if we look at what this other measure and this I think is a measure that's not often applied in neuroscience this is a measure called Granger causality what it basically does is it measures information flow it doesn't look at how correlated different brain regions are it looks at how well you can predict one region from knowing what's going on in another region it's different so in it so it's really about the flow of information rather than how different brain regions are moving together and if you look at this measure you see it goes down blue is less than zero compared to baseline so this is evidence that the parts of the brain are speaking to each other less frequently or less intensively or less by this measure in the psychedelic state although they seem to be more correlated and in most experiments this is a bit harder to measure statistically so in many studies people only look at this and they say ah the brain is more synchronized so there's more communication going on between brain regions what we find is the opposite and so we need to understand that and see if that generalizes to other things like what happens when you fall asleep we don't know that yet hopefully find out soon well anyway this is the main result that this measured but back to the idea of measurement this measure of conscious level goes up when you're in the psychedelic state now the X I I think the the reason to focus on measures like this that try to characterize consciousness in terms of the complexity of the neural dynamics this is interesting because it's based on a particular theory about what consciousness actually is and how should relate to underlying mechanisms back to this idea of the real problem trying to map between mechanisms phenomenology and the basic idea is goes back to work by the Italian neuroscientist Giulio Tononi and his and Gerald Edelman who worked together Adelman was my old boss so I did my postdoc in San Diego in the early 2000s ableman was one of these people like Francis Crick who had won their Nobel Prize in a different area of biology and then thought right now I've won a Nobel Prize I can go and figure out consciousness and because in those days as I said you know the only way you could do it legitimately you couldn't do it as a student you had to kind of win a Nobel Prize but fortunately you don't have to do that now otherwise I would still never get around to it but the basic idea that that drove this theory which is now called the integrated information theory of consciousness is some very simple but I think very meaningful observations the first is that every conscious experience that you have is unique it's highly informative for the organism that you are every experience that you have had having right now is different from every other experience you've ever had ever will have or ever could have even the experience of pure darkness rules out a vast repertoire of alternative possible experiences that you could be having you could be having much more fun experience right here right now but you're having this experience of listening to me go on about experience that is not an experience that you will have again fortunately not at least not in exactly this way but this is formally what we mean by information information is how much uncertainty is reduced by a system being in a particular state and by UX having this experience right now and not any other experience that you could have there's a huge amount of information that's been generated you know compared to a simple photodiode that could either only ever be on or off depending on whether it's light or dark now for the photo dieting Zarn title dark because light or dark for eyes is you know it's in the context of many other possible discriminations that we can make if there's only ever light or dark there's just one or zero there's really nothing so consciousness is highly informative but it's also integrated so every conscious experience that we have we experiences all over peace we don't experience colors and shapes separately from each other you're experiencing a single unified conscious in right here right now now possibly there are cases where this might break down in various pathologies or in split brain cases and other weird cases but for the most part we have a single stream of consciousness this just shows this is one of these examples to kind of illustrate this a little bit that you can you know this is the Ruben phase of ours illusion you can either see it vars or two faces facing each other but usually not both at the same time in fact the more I've looked at it the more I can tend to see both at the same time but the point is that that really you should only be able to see one so we have this combined two properties and our conscious experience that they are both they're simultaneously informative they rule out a lot of possibilities and they're integrated there's somehow all of a peace and knowing that allows us to begin to write down mathematical measures that capture this middle ground that describes this phenomenology so systems that are not like this which is where everything is integrated but everything is connected to everything so there's not many state to this system can enter there's not much information all systems like this where thing every part of it can do something different so it can enter a lot of different states a lot of information but there's no integration everything is sort of falling apart and independent so we can come up with this middle ground just schematically looks a bit like this and figure out different ways to quantify that middle ground and that's what some of these measures of complexity that I've been describing try to do they don't respond to that case or that case but that case which is why there's reason to believe that they should be sensitive to changes in conscious level it also helps us understand one of the mysteries that I mentioned earlier which is why does the cerebellum have nothing to do or very little to do with consciousness but the rest of the cortex has a lot to do with it this is a sort of slice through part of the visual cortex and the only thing I want you to take home from this is that it's very complicated there's lots of recurrent connections between different parts of the of the brain and this is just the part of the brain that's involved in envision signals flowing laterally backwards forwards in all directions and this is certainly oversimplified if you look at the cerebellum it's much different yes there's full times more neurons but basically the cerebellum is seems to be like a vast number of relatively independent circuits it's organized like a crystal in two layers and each layer is implementing probably the same kind of computation but in a way that doesn't depend very much on what's going on in the layer next to it or in the layer next to that so what you have here is a REIT you can predict just from the anatomy of the cerebellum that it will not have a very high value of integration that it won't won't sort of underlie that property that we that we expect a measure of consciousness to have so this is a nice way of sort of understanding the relationship between Anatomy and consciousness now one of the things we've been doing in the lab is try to move beyond these the measures that I've described so far are relatively simple relatively blunt ways of characterizing this this complexity and for a number of years we've been work trying to develop more sophisticated ways which capture in a richer way this balance of or combination of integration information so like I'm sure you can you know just have a look at that equation and and we'll ask questions about it or ask questions about it later here's another one which is I think even more beautiful but basically there this is what they're trying to do they're trying to measure the extent to which a system is both capable of entering a lot of different states and the extent to which that assists each state is globally unified and they're just various ways one can do that mathematically and so that's where this this line of work is going and good zooming out for a bit this I think there's a lesson for this sort of approach which is a very positive lesson the history of science so I've been talking a lot about measurement coming up with a number well why you know when people were figuring out like they didn't worry about measuring the amount of life that a creature has it was I wasn't how the problem was solved but nonetheless I there are other this is what might be an imperfect analogy that there are other areas of science where measurement and very frequently measurement has proven to be the key in naturalizing something that was previously mysterious and measurement is really what empirical science is leveraged by you know without the ability to measure you can't make specific testable predictions very frequently and a great example of this is our understanding of temperature and heat we all now know what heat is it's the mean molecular kinetic energy of molecules in a substance but there was a time again where he was pretty mysterious it was this calorific substance that flew in and out through from hot things towards cold things nobody really knew what it was in order to try to naturalize heat to try to understand what it actually is it was necessary to find ways to measure heat and there's a beaut one of my favorite books in the hot in the history of scientist by hassock Chang called inventing temperature and it tells the story of how there was such hugely controversial activity in trying to come up with a reliable thermometer because how can you do it you know how can you validate a thermometer unless you've already got one yeah or a scale and so people were starting to design things like glass vessels but they would have impurities so you know they never showed the same reading twice and bit by bit accurate thermometers bootstrap themselves into existence oh yeah there's one thing I can't beautiful part of this story is that if you need if you try and develop a thermometer you need a fixed point you need a fixed point from which users reference and how do you get a fixed point without a thermometer well you have to have Sun a priori idea that something is not changing with respect to temperature so four people tried blood but you know blood changes and you sort of there a reason to believe that blood changes so for a period of time the zero point of temperature was taken to be a piece of cheese in a basement in Paris but now we have thermometers and now you know and this this development of modern thermometry has really transformed our understanding of heat from not just something that feels hot or cold to something that we understand the physical basis of and we can talk about now the temperature at the surface of the Sun or absolute zero in in interstellar space we know what temperature is in a way that that we had that we didn't before and so the the hope here is that maybe the ability to measure consciousness would lead to a similar transformation or at least be a step on the way to that transformation if consciousness is something that one can measure so that's conscious level to summarize that a complex balance of differentiation and integration and with that I think it's uh-oh the next thing I'll go onto is conscious content but let's take a five minute break and then we'll come back for the for the second it give you an example of something that highlights this issue of conscious content and so the idea of conscious content now we move on from level when you're conscious you're conscious of something and so this is a an excuse to show you one of my favorites of visual illusions and demonstrations how many people have seen the like chaser before this thing a few of you so what I would like you all to do is just focus your attention on this central black cross I don't this possible to take the lights down for a second if not doesn't doesn't really matter but if you just look at the black cross try not to move your eyes or blink too much what do you see green a green circle green circle moving around yeah what happens to the magenta patches this pick okay now blink need your eyes and have another look now back right so there is no green circle there is no there is no spoon there is no green circle yes there yeah there isn't no no there isn't what's actually going on here is a combination of at least three different visual effects the first thing that's going on here is something called trucks the fading if you're fixating somewhere in the visual field then things that are out there in the periphery don't have to be very far from the center of your vision that have indistinct boundaries will tend to fade away and be filled in by what surrounds them that's called trucks the fading the second thing that's going on is apparent motion when things come on or off in proximity to each other we tend to infer that there's something moving between them this is of course what happens in cinema and we so will tend to perceive something moving between them when of course there isn't anything just things turning on and off and the final thing that's happening is color opponents see which is that when the brain adapts to a particular color being there and it goes away we see in color space the opposite of that color and the opposite in color space of magenta is green now there's actually a fourth thing going on here which is I mean we know from Newton right that colors don't really exist out there in the world they're what the brain generates to make sense of electromagnetic radiation that impacts the eyes it's a way of picking out in variances in the environment so you know colors are a construction of the brain but magenta exists even less than other colors the reason being that you make magenta by mixing red and blue light you think about the electromagnetic spectrum you have red blue and green is in the middle so typically out there in the world if you're encountering red and blue lights you also expecting green and if you don't get green then what does the brain do this has to make something up and so it makes up magenta magenta is what the brain invents when it's expecting green and it's not getting it so the green you're seeing here is basically not not green so that's what's happening there and that just makes the point that what you consciously experience is only very indirectly at least in this case indirectly related to what's out there in the world and what I want to argue is that this is true all the time it's not just true these extreme circumstances so just imagine being a brain for a minute imagine you are your brain and imagine that perception is the business of trying to figure out what's out there in the world if there is anything out there in the world now that there's no sound inside the skull there's no light either and it's dark in there if it's anything all the brain has to go on are noisy and ambiguous electrical signals that are only indirectly related to things in the world whatever there may be very noisy and ambiguous so perception figuring out what's there has to be a process of inference in which the brain combines its prior expectations about the way the world is with this ambiguous sensory data to come up with its best guess it's best inference about the causes of those sensory signals and that's what we then perceive it's the brains best guess of the causes of whatever sensory data is impacting our brains right here and right now so let me give you a couple more examples that just illustrate this at this point this is something that's going on all this time this is a pretty famous visual illusion called Adelson's checkerboard how many people have seen this before many of you okay so if you look at the two patches a and B that highlighted conveniently here they should look to you to be different shades of grey yes does everybody see that yes yeah yeah good anybody not see it okay so they are of course the exact same shade of grey that's why it's an illusion and if you don't believe me I can show you by another version of the image here where I've joined up the two patches and you can see that it's one continuous shade of gray if you think this is not the same image and I'm cheating well I just move that gray bar across and you can see there's no division at all it's exactly the same shade of grey take it away and it looks different again so what's going on here what's going on here is that the brain is using its knowledge that's built deep into the circuitry of the visual cortex that objects on the shadow are darker than sorry a lighter than they appear things get darker under shadow that's why we see be as lighter than it really is because of the shadow cast by this so our brain makes an inference that this is actually lighter than it appears and the checkerboard just accentuates that so there's another there's an inferential process going on even if you know about this as well you can't help but um but see it here's another example have let's hope the sound works and some of you will have heard this before if you have then please just don't say anything until the end okay random sort of noisy whiskey things try again okay now have a listen to this which is the same thing but unprocessed I think sorry dude I'm not sorry actually now have a listen to the original and it's going to play exactly the same sound again that I played in first time okay so now you can all understand that right you can hear words which was previously just a bunch of noisy whistles one more time just for fun what's remarkable here is that the sensory input hasn't changed at all now all that's changed is your brains prior expectations were best about what caused that sensory information and that changes what you consciously perceive so our prior expectations are really really critical in shaping what we consciously perceive now there's a sort of mathematical way to think about all this which is Bayes theorem Bayes theorem is one of is the foundation of modern probability theory and it just tells us what we should do or what any system should do if it's trying to update its beliefs as new evidence comes in so you might have you know this is everything here is represented by a curve you a yellow curve here might represent your prior belief for instance that I said something about brexit and then the sense data may be ambiguous or say something else and then the green curve is the combination of these two curves it's called the Bayesian posterior and it's just the brain's best guess of how to combine sensory data with prior beliefs and so basically the claim is in perception this is the brain is doing something like this there are things in the brain that represent these sorts of probability distributions that have means and variances of course the sharper this curve that means the more confidence the brain will put in the sensory data the further the perception will be drawn towards the sensory data so you know it's not just taking them both equally they can be drawn more towards one or the other but the claim is the brain is doing something like the process of Bayesian inference when it's doing perception and this change is quite fundamentally how we think about how the brain does perception so here's the visual system of the monkey and typically we think of it in terms of a feed-forward architecture where signals come into the retina and they percolate deep into the brain with early areas dealing with things like lines and contrast and deeper areas dealing with things like objects and faces but the perceptual heavy-lifting here is done in this feed-forward or bottom-up direction now this idea of the Bayesian brain which also traces back to the German physiologist and physicist Hermann von Helmholtz says something very different it says that perception depends not only on signal sensory signals that come from the outside into the brain but as much if not more on perceptual predictions that flow in the opposite direction from the inside of the brain back out to the sensory surfaces these green arrows schematize the idea that what carries the contents of our perception is not the sensory signals but the stuff that's going the other way the stuff that's carrying the brains best guesses about the causes of those sensory signals so perception happens as much if not more from the inside out as from the outside in you can map this idea onto the neuroanatomy and neurophysiology of the brain this is a schematic of different levels of the visual cortex and the idea here now is that in each level the blue arrows are just conveying predictions from one level down to the the next level below so from high levels of the visual cortex right back to the first sensory relays and the bottom-up signals the the red arrows just convey the prediction errors so what we tend to think of as sensory signals coming and they're just the errors then mismatches between what the brain expects and what it gets at every level of processing and perception then involves just the minimization of prediction error continuously across all the different levels which has a really weird implication that we could still perceive things that we might perceive things most accurately when there's no signals flowing up through the brain at all because our predictions perfectly predict what the incoming sensory data is so and there's quite a lot of evidence for this view now this is actually an early study from from Vincent well she was here this morning and I don't know if he talked about this particular study those not here but if you use TMS again this transcranial magnetic stimulation to knock out specifically the signals that go backwards in the visual cortex you specifically knock out conscious awareness in this case of the movement and this is back and before this theoretical framework was available but this empirical evidence is still very meaningful and again in my lab we've been doing some similar studies which I'll just give you a flavor for we asked a very simple question do people consciously see what they expect to see or what challenges or violates their expectations and a simple experiment where we show people this is called continuous flash suppression so people wear a kind of stereo scope so they see each image in a different eye so in one eye they see an image which is either a house or face which gradually increases in contrast and in the other eye they see a changing pattern of shapes which gradually decreases in contrast and at some point the image will break through people will become aware of it and say I see it it's a house or a face and what we do we cue them to either expect to see a house or the face and just ask the question if they're expecting a face does it break through into their conscious awareness earlier and the short answer is yes it does over you know many control experiments and and so on but a short answer is that we see what we expect to see in other experiments we've tried to map this more closely now onto what's actually happening in the brain doing this kind of process and this is work with my PhD student Maxine Shannon and really dull experiments where people would just say whether this boring patch of lines was present or not on a given trial but while people make these kinds of decisions about presence and absence we can measure the electrical activity in their brain and what we find is in the visual part of the brain at the back there's usually a very prominent rhythm called the Alpha rhythm at about ten Hertz it's a much more prominent when you close your eyes but it's there even when you don't and what we found is that the phase of this alpha rhythm has a big effect on the extent to which your expectations influence what you see now that's complicated sentence but what it basically means is you can imagine this wave of activity going up and down in the back of the brain and that certain point of that wave may be the crest of it if you're expecting to see the patch and it's there you know you will be more like to say it's there so you're you're deception decision will be much more influenced by what you expect than by the actual sensory data and then then when the wave is that the trough at the other part of the cycle you know 50 milliseconds later then your perceptual expectation will have the least effect and the sensory data will have the most effect so the idea here is there's this constant oscillation between your predictions and then the sensory data and this may be orchestrated by this alpha rhythm in the brain now this is all well and good but this is all very sort of low-level experiments about you know it was that patch of light there or not there repeat a thousand times it's not doesn't have a much to do with our experience of the visual world around us which is full of objects people and the spaces in between them so what I what we've also been doing is trying to develop experimental methodologies that get a bit more about what vision is really like that for us most of the time out there in the wild and I just wanted to mention in this context another line of work that I've been thinking a lot more about which is art history instead of the older traditions in art history this is Ernst Gombrich who wrote the classic story of art among other things and he has a concept called the beholders share this is kind of unfashionable in art history these days but this is the idea that you when you're encountering a piece of art the observer brings to bear a great deal in the act of perception it's not all there in the painting the observer so I'll have to take questions at the end I think otherwise we're going to get through that it's the observers perceptual machinery that is involved in the impact that art has and I think this is really well reflected in things like impressionist art and a bit you know naive when it comes to art so I love this kind of classic stuff but you can imagine what is going on here is the artist is reverse engineering the perceptual system they're not painting the results of perceptual inference they're reverse engineering to paint the light the raw materials that your perceptual system engages with to generate the impression of the scene so here dirty palette scrapings on a canvas can convey a rich experience of of trees and River and movement and so on and if you think about this is a really phenomenal skill because to be able to do this the artist has to have an understanding of all the steps that the visual system is engaged in which extracts objects and meaning from mere patterns of light so I think this is as a beautiful example of two insights coming together from different directions and in fact Gombrich back in Vienna in the early 19th century was interacting with a lot of the early psychologists as well and talking about the parallels between these these approaches which sort of got sidelined later on unfortunately he said it's the power of expectation rather than the power of conceptual knowledge that mold what we see in life no less than in art which i think is a really pressing way of capturing some of the insights of modern neuroscience here written a long paper about this don't worry about it so we've been back to trying to capture some of this more naturalistic aspects of vision we've been using things like virtual reality to try to do this in some interesting ways in the lab and one of the things we've been focusing on is how and why we perceive things as being objects this is again his the famous painting by Magritte the treachery of images this isn't pipe it's a picture of a pipe it's in fact a picture of a picture of a pipe but what determines when we experience something as an objects now if I look at my phone now in some sense I perceive this as being an object as having a three-dimensional volumetric extent I perceive it as having a back and sides even if I can't immediately see them well why is that why don't I just perceive this as a sort of flat picture of an object well then the whole line of work in psychology suggests this is because my brain is somehow not only predicting the most likely cause of current sensory input but it's predicting how that sensory input would change if I were to make various actions rotate it or move around it if I move around to the back of a picture I no longer see the picture of the pipe but if I move around the back of a real pipe I still see the pipe so we can begin to play with these ideas in VR by generating situations where we have unfamiliar objects that behave in weird ways when people interact with them so here's an object that would that behaves normally as you spin it around and so on so it's it would it supports the generation of useful predictions this is like the moon this object always shows you the same face no matter what you do so it gives you a very canny feeling of is it really there or not and this is an object which behaves in weird unpredictable ways when you move and so the idea here is this is still ongoing work that we can manipulate these predictions about the sensory consequences of actions and that should impact the extent to which not whether people see or not but whether they see it as an object or not they try to explain something about more rich about the phenomenology of object at which again is in see represented in Picasso stuff here about what does it mean to be an object in the world another thing we can look at is hallucination so if you think of perception as this continual balancing act between prior expectations and sensory data what if the balances goes a bit wrong what if the balance in my brain becomes unlike the balance in all of your brains would I start to see things that you don't and the way we've begin to address this is by using these are these sort of things these are deep convolutional neural networks these are neural networks that have been used widely in AI now to do image classification basically you can show an image to this network and it will tell you if there's a dog in it and if so what kind of dog very good at classifying dogs what you can also do is you can run it backwards and some of you may have seen this it's called deep dream you can run it backwards basically you give it any image and you run it backwards effectively like saying to the network there is a dog there now change the image until you agree so you update the image until it the network is satisfied that it contains a dog and if you do this dog sprout up everywhere in an image so what we've done is we put this into we took a panoramic movie of Sussex campus and fed every frame through this algorithm and then you can put a headset on and you can look around and you will have an extraordinary experience of dogs coming out of everywhere in what yeah now some of you who sort of experimented with certain substances might find some resemblance between this kind of imagery and what you will have experienced under psychedelics so this isn't really like a schizophrenic hallucination but it is very much like a psychedelic alienation and it's not just like having dogs photoshopped onto the image you know they're really coming out of the image at all levels and so this something important here which is that it gives us a clue to the way we're generating this image it's picking up something about how the brain generates experiences of dogs in the first place and of anything really and also what might be happening when people are experiencing let's say a psychedelic Colusa nation and we can do other things we can fix other layers of the network and generate other kinds of hallucinations so this is a more low level so instead of seeing whole dogs here you just see kind of dog parts and eyes and cropping up in in weird places so you can start some model different kinds of hallucination this way and figure out what might be going on in the visual system of somebody that's experiencing a psychedelic Colusa nation versus let's say a psycho to a psychotic one and you know if we this is just a questionnaire thing if we actually ask people to rate their experiences and what we've been calling the hallucination machine along various dimensions like space distortions sense of peace intensity sense of you get a very similar pattern to what people have reported when taking in this case magic mushrooms and so there is a quite a subjective similarity here so think about this for a second I mean the the the point of that is that we can think about hallucination is a sort of uncontrolled perception where our perceptual predictions are not being sufficiently reined in by sensory data by the same token we can think of normal perception here and now as also a kind of hallucination but a controlled hallucination where our predictions are being regularly reined in updated constrained by sensory signals coming from the world and in fact we are all hallucinating all the time it's just that whenever we agree about our hallucinations that's what we call the real world so I'll finish up with just a couple actually yeah I'm running as usually when I thought I had more time I put a lot more stuff in and just to ensure that I'm still going to run out of time we can also apply this this idea to other forms of perception in this case time perception we don't have time sensors in the brain but we all have a perception of time passing of duration how does this happen yeah a naive idea of time perception is that we have a clock in the head we don't have a clock in the head I mean maybe we do at the level of 24-hour cycles I have jetlag at the moment but over the timescale of seconds and something not really we have a clock an ad so perception of time is also an inference and with work with warwick roseboom another my colleagues at Sussex we've been looking at how we can understand time perception as an inference on how quickly perceptual dynamics change over time and so we build another of these Network models that we show movies that have particular durations and we just observe how the states of the network at all the different levels change from frame to frame and then we use those rates of change as the basis for the network to estimate the duration of the un-- of the video so we have a network we have a set of videos the network will estimate the duration just based on an inference about how quickly the perceptions that the you know the inference has change and people watch the videos as well and we measure where they look on the videos we use eye tracking to measure where they look here's one of the typical videos we use we should do this with the psychedelic movies as well that would be interesting so at the moment we just have videos of things like cows and street scenes and getting a nice tour of Sussex as well it's part of this talk and what we find is that we can model very very closely human reports so this is how humans judge videos longer videos are longer but they're also characteristic biases they always judge longer videos as a bit shorter than they really are and short videos as longer than they really are and the model that we have basically does exactly the same so we can recover the exact the way people experience duration is well explained by them just making inference on how in this case their visual experiences is changing no need for a clocking head and again people use same biases so people judge busy scenes as lasting longer city scenes there office scenes where not much happens the model does the same thing we can even do it in real time this is by putting a camera on a laptop running that model and it will estimate the duration of events in real time in this case walking through a station I don't know where it is some station in London so time is also an inference now this brings to the second take home and I'm going to just briefly mention the third part of the talk and then wrap up conscious content is the brains best guess of the causes of sensory input this kind of control to this nation now the last part of the talk the last message is this same thing applies to our experiences of being a self - just as we perceive the outside world as a construction or we don't perceive it as a construction it is a construction of our brains the same applies to our experiences of being a self it's also a kind of control to loosen ation and in fact there are many different ways we experience being a self there's the experience of having a particular body and of being a particular body in the world there are experiences of perceiving the world from a particular first-person perspective of making volitional actions choosing to do this or that and only then of being a person that has a continuous identity from hour to hour day to day in year to year that's associated with an AI and a name and then even beyond that we can think about how I experience myself refracted through how I perceive you is perceiving me the social self I just mention a couple of things about the bodily self to show you that it's also a construction based on the brains best guess of the causes of body related signals another kind of control to loosen ation now you may have seen this before this is the the rubber hand illusion which is a beautiful illustration of the brain engaged in this process of continually figuring out what is and what is not its body so what happens in this in this experiment is that the person's real hand is hidden from view and a fake hand is placed in front of them and then the experimenter strokes both hands simultaneously with a paintbrush so the person is see a hand which is they know is not their hand but it's roughly where a hand should be looks like a hand they're seeing it being touched and they're feeling touch at the same time so this is quite a lot of evidence for the brain that even though it's not my hand it kind of it must be my hand and if you do this for a while you develop the uncanny sensation that this fake hand is in fact part of the body and there are many ways to test this but this is the only really good way to test it works really well on children as well I found so it works you know you can you can convince the brain that it's something as basic as what is our body is something that is malleable and of course many symptoms in following brain damage can we see this frequently people with phantom limb syndrome who've had limbs amputated will still feel the limb as being there this is the mirror image of that is something called somatic periphery near people whose limbs are still perfectly attached to their bodies they experience them as not their limbs and desire their amputation and this can even you can do you can extend this in many different ways this is Maxine and Rani at the British the Brighton know the British Science Festival last year what you do here is you have people wear headsets with cameras and you swap the feed so that now this person sees themselves through that person's eyes and again the key thing is when they shake hands you get that multi-sensory congruence which the brain takes is extremely powerful evidence and suddenly you feel like you're shaking hands with yourself from within another person's body it's a body swap illusion you can even have out-of-body experiences and I think there's a you can look at yourself and behind and do the same sort of thing there's an important message here which is that throughout the ages people have reported things like out-of-body experiences and I've often been sort of you know poo-pooed by it by science and this is wrong people are having these experiences and they should be respected but their explanations for these experiences should perhaps be not taken on board it's not that your soul has left your brain is floating around it's just that your brain has come to an unusual best guess about where its first person perspective is located by purely natural means now there's experiencing the body from the outside there's also experiencing the body from within and this is typically you know if you think about this is a very basic sense of salford that we all have this experience of just being a body and this highlights interoception which is an overlooked way in which the brain perceives the world this is the brain perceiving the internal state of the body a large amount of neuronal real-estate is dedicated to perceiving and controlling its internal state we think what brains are for they're ultimately just for staying alive that they're keeping the body in the brain alive perception of the world around just sort of follows indirectly from that so we wanted to know what the implication of this into reception is perception of the body from within has for experiences of being a self one thing we did was a version of the rubber hand illusion where now instead of stroking the hand with a paintbrush it flashes in time or out of time with your heartbeat and what we find is when it flashes in time with your heartbeat you perceive it more as part of your body so experience of what our body is is govern not only by signals from outside the body but signals that are coming from deep within the body even if you're not aware of when your hearts beating it's still influencing your perceptions of what is your body what is yourself this looks like this in there you know you did this in VR again now so here's somebody's hand it's a bit jaggedy because it's quite old and it flashes to red and back very gently so our experiences of what our body is it's also the kind of controlled hallucination and that leads to the third have take-home message that with apologies to Descartes yeah I don't think therefore am i predict myself therefore I am now the last point I want to mention is also about Descartes and it's really this idea that that I'm the Descartes the fact that we're made of flesh and blood was rather irrelevant to the presence of consciousness or soul yeah Descartes was keen for various reasons to reserve consciousness for humans and certain you know if animals bleed when you you know when you cut them that doesn't mean they have any kind of conscious awareness now I rather think this line of thinking is taking is the opposite point of view that we have conscious cells we are conscious not in spite of that because of we are living flesh-and-blood organisms because the whole reason for having is predictive machinery that enables perception is fundamentally about regulation and control of the internal state of the body so thinking again about perception as this reduction of prediction error we can either change our perception or another way of reducing prediction error is to change the sensation to make an action to change what we perceive you know if I'm expecting to see a friend and I don't see that friend I can either update my percept my prediction or I can go into where you know go to my friend's house and verify my prediction that way so I can make actions to make predictions come true even when I just move my hand or in another way think of this is I'm fulfilling a prediction about where my hand ought to be it's a self-fulfilling proprioceptive prediction so doing this enables us to control if I predict that my heart rate stays within certain range it will stay there as long as I use actions to make sure that prediction comes true rather than just update the prediction predictions can regulate and so here's that here's the last idea which is that interoception perception of the body from within is not really about figuring out what's there it's about controlling and regulating the internal state of the body and this explains something I think about the difference between the way we perceive the world in them when I perceive the world around me it seems full of objects and the spaces between them talked about objects already but the body I don't perceive my internal organs in different places you know I don't perceive being a body as a kind of objects I just perceive it as something that's either going well or going badly and I think this has to do with the different ways the brain is using predictions when the brain uses predictions for control we experience how well or how badly that control is going when we use predictions to figure out what's there we experience objects and things as the contents of our perception so we can begin to really explain properties of phenomenology now in terms of different kinds of mechanisms and this is related to a bunch of other frameworks by Lisa Barrett and Carl Kristen and is not what Descartes said but back to that point that conscious self would emerges because of and not in spite of the fact that we are what Descartes and Julia and a tree called beast machines it's very close ties between mind and life and this just makes the point that if that's the origin of our perceptual mechanisms then everything we perceive whether it's things out there in the world or things in our body are all grounded in these same basic predictive mechanisms that have their roots origin and ultimate explanation in keeping the body of life we perceive the world around us because of our living bodies with them through them and because of them so summarize the whole thing back to conscious level it's a complex balance and neural dynamics of information or differentiation integration conscious content is the brains best guess of the causes of its sensory input and conscious self is I think very closely tied up with the fact that we are living organisms whose internal physiology requires predictive regulation from moment to moment and I'll leave you with three implications of all this and just I won't go into them but just leave them out there for discussion one is psychiatry by understanding perceptions prediction we can start to get a handle on how perceptual experiences differ in pathological and distressing cases in psychosis depersonalization schizophrenia depression and so on we can begin to develop a mechanistic understanding and transition in medicine from just suppressing the symptoms to actually getting at the underlying causes so I think this has a lot of implications for psychiatry second thing is just consciousness is not just a matter of being smart we tend to think it is because we tend to think we're smart and we're conscious and you see this these days in the current discussions about artificial intelligence the idea that as you make things smarter they will at some point become conscious maybe then take over the world there's this weird association of intelligence with with consciousness I think much more important as the association between consciousness and life now you don't have to be smart to suffer but you probably do have to be alive to suffer so we should be much more concerned I think about the potential for suffering in other living systems rather than the potential for developing a particularly smart alphago player that is suddenly going to enslave the rest of us and finally the idea of consciousness and conscious self as a construction tells us that our way of experiencing the world and our way of experiencing itself is just one way out of many possible ways one might experience Salford I experienced my body as an object but that's because I have particular kind of sensory channels and make particular sensory inferences the octopus has three hearts completely flexible arms can taste with its skin it's experience of being a self if there is one is going to be radically different from ours there's a vast space of possible minds and we inhabit just one small region of that vast space I can't resist showing you this video of the amazing camouflage abilities of an octopus can you see the octopus you will it's there they had extraordinary camouflage ability and so you know what's what's their experience of being an embodied organism like it's really phenomenal so you know with I think all these disadvantage and understanding ourselves now they play into this bigger narrative that you know whether it's understanding our place in the universe whether it's understanding our relation to other forms of life and now with our understanding of consciousness in terms of brain activity with each of these developments I think they're helpful that they're kind of illuminating and enlightening because they allow us to see ourselves less as apart from nature as sitting on top of it somehow and different from it and more is continuous and a part of the rest of the natural world and if there's one message in general that science conveys it's that we are part and parcel of the natural world and not separate from it and when the end of consciousness comes well there is nothing at all to be afraid of so thank you very much [Applause] I'm afraid I over ran time but I'm presume we can still take some questions there's yeah one down here and and I should probably come along for a pint as well if you can always follow up there you speak up a bit yes why are we all lazy speak for yourself no I don't know right I am I empathize greatly it's it's I actually think we were promised a slightly different thing I mean I didn't see the other talks but my sort of take on this is we were designed and this was called an evolutionary psychology answer we were probably designed to be quite lazy it's only a real sort of creation of modern society over the last two or three hundred years that non laziness has become important and so it's no surprise that we struggle I think it's just not what not what we the environment we were built for that's guess Oh somebody begs to differ that's fine yeah yeah good question I mean I don't know that exercise that sounds very interesting so you actually perceive the object of changing in color though yeah yeah I mean the first thing to say from that is there's a lot of overlap between mental imagery and visual perception that they're leveraging the same sorts of circuits you can think of imagery as just the sort of imposition of predictions and trying to ignore as much as possible the sensory data you know why for most people it's easier to engage in vivid mental imagery with your eyes closed because then you don't have all this disk confirmatory evidence straightaway it's interesting that you can actually in in your case overcome that anyway now we've seen this in some examples with synesthesia where we can do or hypnosis where we can get people to perceive things differently through training or through hypnotic modulation so is this the same mechanism there's also there's also a number of people who a state called a Fantas ear so a non significant fraction of people just don't have mental imagery at all and they don't know they wouldn't know what you're talking about when they say you know it I can imagine this object with this color it's sort of for a large number of people they just don't do it but they still see that's the mystery that their perception isn't doesn't seem to be affected at least not in any obvious ways I mean where is it enable you know we're all the same distance on the evolutionary tree from this star so you know I tend to resist so the evolution of that you can think of one of the the things about the evolutions of brain that there was a period of rapid expansion of the brain and this is sort of reflected in the fact that if you look at different parts of the brain under the microscope they look pretty similar so it's almost as if the expansion of the brain was just you know replicate the same functional circuitry plug different sensory inputs and things into it and you'll get different capabilities so there has been there was a rapid expansion of brain size which then became limited by birth canal and so on and and people argue about relative brain size and humans compared to other animals there's a huge debate about always again some people trying to point out that we're special somehow and we have bigger brains and you would expect on the basis of body size and all sorts of things I'd tend to get a bit bored by these arguments I mean we we have reasonably large brains yes but they're know so many ways of other thing you know what proportion is devoted to the cortex versus other regions and the particular cell types that you see in human brains versus versus other brains one of the big mysteries here I think is that you know our brains have as far as we can tell remains pretty much the same for quite a while over human history yet our cultural environment has changed dramatically and their periods of human culture where people were basically making the same sorts of tools or decorative things and then change for thousands of years and the same roughly that's far as we can tell because you know we don't have preserved you know we just know that the kind of gross size but you see vast cultural changes when you can assume relatively stable near a physiological background I don't know if that answers your question I don't I mean what it means is that I think what brains do and how we experience the world around us is going to depend much more on the kinds of environments that we put brains into than on a any further biological evolutionary change of the you know the neural material itself there's some sense in which is collective in the sense that my experience of the world around me is very dependent on how I experience you you experiencing it so I you know I can't help but populate the world with other mental agents that I assume have conscious experiences now that's a sort of pretty deflationary way to think about collective consciousness there are the slightly more spooky ways which it says there's one kind of universal conscious being and we're somehow parts that I I don't I think that to me is where we step out of what we can say with science and we've become a bit more religious and metaphysical about it you know Pan psychism is a still a respected view in philosophy I think it's a bit crazy because it's not testable it's not even wrong but there is a you know there's a way to defend the idea that consciousness is somehow present throughout the universe that it has a similar status to sort of mass or energy or charge that's not to say that spoons think or perceive things the way we do but that you know whatever consciousness is at the most fundamental level is is sort of everywhere now I don't find this a particularly useful way to think because the world and what we would do with it basically looks exactly the same whether you believe that or not but from what we do know I think that the the substrate of consciousness is but is in the skull of an individual but the way we experience the world around us is a very collective and a very social thing I've lost track of who's who's next okay [Music] yeah it's a good set of questions on my colleague at Sussex Zoltan Deanna's has done a lot of work on both meditation and hypnosis and I mean the first thing to say there is that hypnosis is real it's not just some sort of circus stage trick you can you know one in ten people roughly is highly hypnotizable and so you can you can test this and certainly you can induce via hypnosis changes in perception changes in volitional intentionality so you can make people behave differently with actions that they make under suggestion compared to actions that they don't and analgesia as well so I know there's I mean I don't not something I've know terribly much about but I find it completely plausible that through hypnosis you can induce states of dissociation that are effective for analgesia thank ya I mean I've only seen the headlines about the pig brain in a dish thing and and actually I was wanting to look more into it because people are making these like oh of course it's not conscious and I kind of believe that but I would want to know on what basis they're making that yeah that that claim but sort of extrapolating out from that there's there's there's one thing which I think is probably not going to happen or at least not people shouldn't assume that it's sort of an inevitable direction we're going which is this idea that we can upload our consciousnesses to a computer or to the cloud at some point because in that assumption is a further assumption that consciousness is something that can be similar that it is something that one can that is just purely a property of the functional organization of a system and is not dependent on the substrate there's a sort of substrate independence now this may be true but it's a very very strong assumption and I think it's a date it's one that people take on board far too easily because they're sort of overwhelmed by the power of computers and computers really a general-purpose simulation systems so we simulate weather systems in computers nobody expects it to get windy or wet inside a computational model of a weather system however detailed that simulation is yet a simulation of something that plays chess is actually playing chess so where is consciousness on this we do I don't really know but I see no reason to think that it's something that we can be sure is something that can be simulated and I would not be signing myself up to you know uploading myself to to the cloud because I think that most you'll get a simulation of what it is to me at minimum you need to recapitulate the causal mechanisms whether you do that whether you can do that of something other than neurons I don't I don't know I actually think the more pressing concern will be exactly the the if you like hybrid things this line of research by these pig brains with you know we will be able to grow more and more sophisticated neural systems body things artificially and but that are made out of neurons and real biological stuff I think then there's there's definite cause for concern and the ethics needs to get out ahead of of that in a way that it probably hasn't yet thank you very much you have three main areas there the level of consciousness the consciousness of content and then the self right if you could eliminate the last two probably not yeah I don't think they're completely independent there's yeah I it's it's difficult to imagine what that would be I think in some sense conscious level probably boils down to the range of conscious contents that you have encompassing self as I think you don't have to have self consciousness so that they get back to some sort of trance and meditative States try to reach states of consciousness where one does loses at least some of the aspects of conscious self sentence some of the more abstract ones but yeah I don't I don't think would make sense to say that you can have conscious level independently of any content again probably not I think I think that they're related I think the the purpose of distinguishing them was more to to show that you you at this point you use different frameworks and different methods to to study them but for instance the fact that we applied these measures of level to the psychedelic state shows that a measure of level is in this case sensitive to a relatively global change in conscious contents you know people don't fall asleep or losing it when they're in under LSD but their global conscious contents are changing and we pick that up with a measure of level so I think the the interesting way to progress that is to ask what are their how what are the relations you know how can content vary and yet we still call it an unchanging level you know questions like that would be useful so there's this whole history from the 50s of these brain operations which were done mainly for again at the lepsy to cut the brain in half you stop the spread of epileptic activity throughout the whole brain so these operations are called calles X amis severing of the corpus callosum the amazing thing was people pretty much behaved the same after having their brains chopped in half compared to beforehand and it takes some subtle experiments where again you show different hemispheres different images and then you suddenly see divergences between what happens and typically one hemisphere is in control of language language is one of the few functions in the brain that is often reliably lateralized to the left hemisphere so if you show something in to the right eye of a split brain patient they will be able to verbally tell you what it is but if you show something to the left eye so it goes to the right brain they won't tell you what it is if you ask them but they will still be able to point to it with their left hand and so and if you show two different images then you can start to get situations where the right brain seems it was a left brain see to be saying one thing the right brain seems to be pointing at something else which leads to the idea that there are two separate conscious subjects within a single cranial container the odd thing about these experiments is that almost all that's been written about them was written on the basis of a relatively few case studies mainly done in the 60s and 70s 50s and as split brain operations became less common it's become harder to go at these studies with the rigor that one would apply nowadays when you're doing experiments on perception there's a group in Holland which have been doing this and they found actually the situation is not as simple as two separate things basically if you have a stimulus appear anywhere a split brain patient can point to it wherever it is so they're sensitive to a whole visual field but what they can't do is integrate information that crosses the midline so if perceptual decision depends on a stimulus half of which is going to one Hemisphere half of which is going to the other they will encounter problems making a decision about that stimulus so it's as if they have sort of one awareness but two separate areas of integration of perception I don't know what it would be like to be a split brain patient it's still unanswerable but it's not as simple as that the other sort of fascinating mirror image of that is what's called cranial Fagel twins so these are twins whose brains are melded together so it's the opposite of a split brain and you know this obviously happens very rarely and often most don't survive but in a couple of cases you find interesting situations where one twin is able to taste what another twin eats and so again are there two separate individuals or two separate and one super ordinate individual or what doesn't I don't know what the answer to that is but it's it's perplexing because they don't exist because esp phenomena don't exist so we don't know there's no evidence for it okay fine second it's up to you I'm happy to I think people are probably running out of energy but yeah you know I've never really understood what they meant by that I mean a hologram is awared embedding a 3d structure on a 2d plate so that you can it seems I always took it by some analogy to try to get a handle on how we understand the topology of the brain in three dimensions beyond that I don't really know it may have other analogies in terms of how you can recreate let's say a memory by probing a neural circuit with activity in different ways and regenerating some topological form but but it's sort of yeah you're right it's it's a phrase that's I remember and it sort of I haven't heard it very recently so I'm not sure it has any any particular legs right yeah that said yeah that may be more to it so gentlemen the front says I think very rightly that a better use of that term is the ability to reconstruct the whole from paths even when parts and parts may be missing so that the information is kind of distributed in a way across a network that's resistant to disruption of its parts which yeah it's again it's sort of by analogy isn't it but it's but it's a probably a more informative one than the one I said okay I think that's it thank you for your patience [Applause]

Channel: The Weekend University

Views: 42,712

Rating: 4.8135338 out of 5

Keywords: the weekend university, psychology lectures, neuroscience of consciousness, neuroscience consciousness research, neuroscience and consciousness, anil seth, consciousness neuroscience, neuroscience consciousness and spirituality

Id: Rz5b4dPhzIQ

Length: 109min 33sec (6573 seconds)

Published: Sun Jan 20 2019