Consciousness Live! with Richard Passingham (S1 E2)

0:00 joining us today with me today is a very special guest dick passing him from University of Oxford so for those who don't know you could you say something about who you are and what you do I'm cognitive psychologist who took to brain imaging very early on in 1988 when positron emission tomography got going and I've made a special study of the prefrontal cortex and I've had research

0:35 groups both in Oxford and in London so the London group are at the film which is the Welcome Centre for human neuroimaging and I'm in the psychology department in Oxford okay very good thank you for that so that I'm very excited to talk to you this morning and I know that you yeah you say you adopted brain imaging very early on so when it was one you're one of the first people to use it actually I believe is that correct well I shouldn't make too much of that what happened was I heard a talk

1:11 in Oxford given by Richard Peto GLIAC and he was at the MRC cyclotron unit at the Hammersmith Hospital in London and he and Terry Jones were doing PET scanning at the time when the Karolinska was doing it and it was being done in one or two places in the USA and I heard the talk that Richard gave and I went

1:44 down and said could I collaborate so from then on I had a research group both in London and in Oxford and the London group did the imaging and for some years we did post-trauma mission tomography but the physicists there didn't believe when fMRI came along they didn't believe that the signal was genuine they thought that it was due to movement and in the our movements that distort the image there are artifacts but this

2:23 meant we couldn't actually do fMRI at the MS a cyclotron unit because the MRI people didn't believe in it so Richard Walkowiak to him with a group of us went to Queen Square in London and built what became known in England as the film and is the Welcome Centre for human neuroimaging and for many years that was one of the premier centres together with

2:59 Montreal together with University Washington and so on and I've run out of things to say let me ask you some of the dangers in be because I know cognitive neuroscience now is fm rise you know there's hundreds of studies done thousands probably and there there are some dangers associated with

3:32 interpreting the data from there and I know that you have some cautionary principles that you would like people to take to heart what do you think we're getting from the fMRI images exactly what is it telling us about the brain at the tail I'll call it the film the Welcome Center in London they brought him what we call project presentations so if you were going to do a project you had to present it to the whole group and those happen every Friday and many other

4:08 labs have taken up the same and I constantly used to ask why do you need to do the imaging study why can't you tell from the behavior alone in other words what does the imaging tell you that you don't already know from the behavior right but that's problem number one problem number two if you see let's say somebody's depressed I'm gonna take an example that I've used in a book and you

4:42 scan them then you find two areas of the brain in which the activity is not normal and it's very easy to conclude from that that depression is a brain disorder now of course in one sense it's a brain disorder you wouldn't be depressed if you didn't have a brain but to say it's a brain disorder is to imply that it starts with a problem in the brain whereas this an

5:18 alternative which is the Germans broke up right and as I point out in the book if you scan people and you simply ask them to think sound thoughts while they're in the scanner the same two areas show abnormal activity so what you seem to be doing is just finding with a very expensive tool what you know already which is that they're not really making sad thoughts now I'm not suggesting that depression might not be

5:56 caused by problems in the brain we know that there are genes 44 genes that increase the risk of depression and those have to act by changes in the brain but they probably hurt by changing how you react to stressful events they make you more reactive in other words but you might not have those genes your marriage is just broken up and of course

6:32 if I scan you I'm gonna find you depressed but it's told me nothing that I didn't know and so what I'm saying is just because I see those changes in the brain that doesn't tell me whether what you've got is a brain disorder or not and the same applies for lots of other things and to use another example I've used in a book if you scan gay people you find and you show them pictures then you find that if

7:06 they're men and you show the pictures of men there are changes in the hypothalamus that you don't see in heterosexual people but if you then the problem is that all you may be doing is telling the showing what you know already which is that they find these attractive wrangles if you show men who are heterosexual women you find the same

7:42 changes right yeah so it's another example where people think the images and telling them something when actually all they may be doing is telling you what you already know from the behavior and by the way we have a brain by the way by the way and I'm not suggesting that everything is the brain you've got a body as well you know but but so

8:22 that's one problem there's another problem which has become much more there's much concern about which is that imaging studies in the old days we used to have six people in a group and we then did statistics yeah sickly in one group and six people in the control group I needed statistics and then people thought oh well whenever I can we only use six because positron

8:57 emission tomography involved radiation and we didn't want to radiate too many people sensible but whenever I came along it was very expensive so twelve people and so the norm became that you did studies with old people but then that became a crisis and the crisis first happened in fact outside the imaging community dan Kahneman pointed

9:34 out that the priming studies that social psychologists were doing where let's say you showed them words the synonyms for aging and then claimed that they walk out of the room more slowly right he challenged them to replicate these studies there was a replication crisis over priming studies but then Minako showed that if you look at studies with genes and you try to find genes that are

10:10 associated with psychiatric disorders you find them but then other people fail to replicate so the began to be a replication crisis across biology and it then hit imaging and the main reason it hit imaging is because we were using groups that were much too small they are underpowered for statistics so the statistics were much under part and respond track at Stanford has been one of the people he's not the only one but

10:46 both in his Bronx and in books has been one of the ones who's really been pushing that we must use many many more people and in the connectome project that David Van Essen dogs after in America where they scan 1,200 people of whom many were twins both identical and non-identical you've got sufficient power and there are now even studies where what you do is you

11:20 have many people studied in one lab many other people studied in another many others studied in another and you put all the data together and you get huge groups so imaging which looked as if it was going to be glorious had problems of interpretation and now has problems of replication the cosmic relationship so

11:59 that seeing a particular area is differentially activated between two conditions so people then make all sorts of inferences about what the cause of the behavior is so this area must be the thing that's producing it now seems to be not what's uh wanted by the by the energy no without that that's in a sense the depression point you see change these changes which are in what are called sub general cortex and the amygdala but actually they may not be the cause the marriage breakup may be

12:33 the cause and there are many other problems one is that the Pulsar is very easy to say area a is not activated when in fact the truth is that if you put electrodes in and record it from the cells you'll find they're active and you simply don't have sufficient power yes yeah I agree that's I mean my own background from laboratory work I did ERP stuff but also

13:08 single-cell recordings and animals and I just fMRI is an important tool but there is this problem I think where people don't find the activation then assume well then there isn't any activation there whereas the tool may just be too coarse-grained to actually be picking it up yeah a specific example we did a study very early on on prefrontal cortex we taught people finger sequences so they had to learn sequences of finger movement so we gave them feedback so

13:41 that they learnt it and we taught them and then we studied them when they had learnt it and it was automatic and what we found is during the learning prefrontal cortex was very active but when the finger sequences were automatic we saw nothing in prefrontal cortex and it's very easy to conclude that there's no activity there now there's a single unit study done in old Miller's lab in

14:17 which they looked at a similar situation so they were teaching they weren't sequences but they were teaching tasks and this is in this case two macaque monkeys and there were two conditions one of which they were learning anew and one in which they had familiar pictures and there was a delay it was what's called a delayed matching task so you were shown a picture and then after a

14:50 delay of shown two more pictures and you have to pick the one that was the same that's the task and if the pictures were novel then during that delay when they're remembering the first picture there was activity as measured by electrodes if the pictures were familiar there was very little activity during the delay but there was activity just before the two new pictures the

15:25 the two response pictures are presented and of course in imaging you'd miss that in other words that's the difference between tonic activity and phasic activity and on the Halal my suspicion is that imaging is very good at picking up tonic activity and very bad at pick-up phasic activity so my suspicion is that if we had done the sequence study with monkeys and we recorded then

16:00 just before each finger movement even though it was automatic we would have found activity even though we did the series in the imaging interesting yeah well I'll ask about something in a second but so I wondering since you mentioned the prefrontal cortex if we could talk about that for a while so I know that's your area of expertise the special expertise you've done a lot of work on both humans and in monkeys I think my that's right so very interestingly that the prefrontal cortex

16:35 seems to sit at the top of a couple of hierarchy so I know you guys you think it's at the top of the action hierarchy also the top of the vision higher right well I'm just fishing sense to me right so what if you could give us your take on what the function of the prefrontal cortex is what is it doing that it needs all these inputs simple question right I'm just working on the second edition of the book this and a dancer that you would get if you

17:15 asked the neuroscientist would be that prefrontal cortex is involved in decision-making that's the standard answer that you would get from people who were recording in animals and people who wouldn't studies in people but as you pointed out it sits at the top of a sensory hierarchy so in other words there's a whole chain of information processing going on before you get to

17:51 the prefrontal cortex and then it doesn't project straight to the motor cortex but there's a whole chain of areas through which it then sends information so the motor cortex so it's performing some transformation between those sensory inputs at a high level and those motor outputs at a high level now in a book I wrote 93d I think I thought that what prefrontal cortex did was

18:32 support what I call conditional behavior if sensory contexts day and you won't how come B then do X but the problem is that prefrontal cortex strictly defined is unique to primates but rats can do problems like if sensory cortex ah if sensory context they and our desire can be then do X so

19:14 it looks and I'm now depending very much on the work that Earl Miller's lab has done uh and it might help if I give some examples when I told you this it looks as if the representations in prefrontal cortex are very abstract so the rules conditional rules are very abstract and so for example you know minusma have

19:52 Jonathan Wallace and Earl Miller recorded in monkeys while monkeys did delayed matching I've told you you're shown a pattern and then after delay you've shown to mourned you have to make the same pattern and they found selves that fired when you're shown pattern a and you picked bad name but then they found at the same cells far for the general rule matching in other words he

20:26 didn't matter what the inputs were it didn't matter what the patterns were they were firing for a very general rule so it looks really as if the key to prefrontal cortex lies in thinking of it as supporting abstract rules tall behaving now it gets all these sensory inputs and then it gets very direct

21:03 information about desired outcomes from what's called the orbital frontal cortex which is the lower part of the prefrontal cortex and then it's got a it's got some outputs as I say indirectly to the motor cortex so I think that we will know what to say that it supports very abstract behavioral rules you would say that the same in the case

21:38 of the century case as well as when we're not talking about behavior yes now at the moment I'm I just started revising a monograph that I wrote with Steve wise he said NIH and I'm going to change my mind over some things and in there in the preface which I've written because I always believe in starting from the beginning and then working to

22:11 the end that's how I've written the preface that for this the second edition I point out that you no good as a scientist if you've never been wrong one of the things that we said in the first book first edition was that prefrontal cortex was important where what you did depended on a single event something that had just happened but both Steve and I now think that we over ate that

22:48 and that it's really not just single events it's more recent events so that's a sort of finesse that we want to put and it's an important characteristic of the monograph that we wrote then we put the hypotheses in italics and in the last chapter we actually try to say how those hypotheses could be falsified because it's a real problem I think that

23:26 in neuroscience in an imaging neuroscience in particular many of the things that people say are not stated clearly enough so that it would be easy to falsify them and I quote Minsky who said that if he's believed something for five years he doubted it and I think that's a very good rule and I knew David Marr very

24:04 well because I miss a graduate student at the same time and in the same place as he was and David Marr who founded or was one of the founders of computational neuroscience in his early papers had a starring system so that of the claims that he made he'd give them three stars two stars or one star and if three stars turned out to be wrong the whole thing falls one star is wrong well that's a

24:41 branch and in his original cerebellar Theory theory about how the cerebellum works it took ten years before somebody could test the 3 star prediction and sadly David not had died by then he died early um so we felt in our book that it's very very important to make sure that whatever we say about prefrontal

25:16 cortex can be falsified and for that you need to be clear and the trouble with being clear or is it becomes of this when you're wrong as like on reading the book given that as I hope it's clearly written it will become more and more obvious where we rock yeah I think that's a very important point do you have any examples of things

25:49 that you think are sort of too vague to be too not clear enough yes it's often said that prefrontal cortex performs executive functions yeah what's that how do you falsify that what's what's not going to be an executive function and yeah so I I think that this desire to be clear and to suggest how things

26:27 could be falsified the the reason we came up with it was that I had a training atrocity I did philosophy and psychology together because at that time in Oxford you couldn't do psychology on its own well I was taught in the 1960s by dick hair became professor of ethics in Oxford and by Toni Kenny and I was taught theory of mind

27:02 sorry I was taught philosophy of mind by Toni Kenny so just two comments about that one in a tutorial and in Oxford students have tutorials which means that two people say write a paper which is then discussed in the tutorial with the tutor and Toni can t just written a book on philosophy and mine published by rat Ridge and a question I was given to

27:36 write on was talked about in his book and I looked in his book and I served it up in my paper and after I'd read it he said I don't agree being young I was taken aback because he just written it and he said yes I changed my mind and I've never forgotten

28:13 that he genuinely on this particular issue had changed his mind now another thing that really influenced me was having tutorial as we dig hair and the reason was that Oxford philosophy in those days was it was the premier school of philosophy this was before are basically America took over when logicians

28:46 particularly in America became very well known and not should so lost its its were ranking but this was a time when the whole time people were saying what do you mean right ordinary language philosophy exactly and I I was brought up with that so dick hair would constantly say what do you mean and that

29:21 led me to read birth from Russell in particular because Russell obviously I never read principia mathematica but Russell in his pop writings right forget the fact it's funny he writes clearly yes and I for many years I I read Russell and I worked on

29:56 writing so that I would every sentence I would prepare beforehand and I put ABCDE and so on and I make sure the first sentence of the next paragraph was a logical link from the previous one and I did this for years and then it just became part of me so I taught myself I hope to write clearly and the reason is that's honesty yes yeah and

30:34 the reason that I don't like I want I shouldn't mention any philosophers because I don't want to offend anybody over the air okay philosophers who are very well-known who write a lot but he doesn't have the source of the logical precision that dick hares books had and I went when I'm asked about this I

31:14 recommend people to read Meadows review of the book the phenomenon of man by tear de Shonda now tear to Shanna was a French priest and he wrote this book the phenomenon of man a meddler was an English a monologist in Oxford who won the Nobel Prize and he was very influenced by popper a wrote about the philosophy of science himself and his review of tentative shadow is the most

31:51 dramatic tearing apart but you've ever seen and if you haven't read it it's it's a it's online and I'll put a link to it in there yeah find it online it is absolutely devastating and the main point is that tear de chartres did right clearly because he couldn't think clearly right yeah they're connected

32:25 definitely and we all know and if you want to think clearly the best thing to do is put it down in the old days you wrote with a pen but now you write on the computer but that's the best way of clarifying your thoughts yeah I agree I often tell my students that they don't really know what they think until they try to write it out exactly yeah so philosophy heavy training philosophy I think had an effect for years I said I

33:01 wish I hadn't done philosophy very rude about philosophers as I say the truth is that I benefited from doing philosophy and I knew it okay good I have no problems at all with sitting in the armchair and thinking about issues of morals issues of law

33:41 issues of the logic of maths these are all things that can only be done sitting in an armchair and thinking my problem comes when philosophers do philosophy or physics or philosophy of neuroscience now of course I don't know what

34:18 philosophers are physics say and do and so what I get when our physicists may be sour grapes yeah a little bit but what physicists tell me is that when I asked them have the philosophers actually contributed to physics is is no now if on the other hand I think about Einstein then I have to worry because Einstein

34:54 sat in his old chair and he did thought experiments right he never did experiments and did he contribute to physics I think so I tell me he did so I think I have to leave physics because I have to admit that you know they were going to be people like Einstein I don't know if Schrodinger or deep dinner did

35:28 they'd done experiments I I doubt it themselves it's quite clear that just sitting and thinking you can contribute in a major way to physics so then the question becomes does just sitting and thinking can you contribute to neuroscience and obviously the biggest problem that philosophers have thought about his consciousness from Descartes on and so the question then becomes had

36:03 philosophers contributed to our understanding of consciousness and it's then that I get worried but there has to be a rider which is that I don't read what philosophers are saying at the moment much so then there may be lots of people saying lots of things that I really don't know and that are contributory but I have read banned in it and I have read yeah I've read some

36:36 of the obvious things I know you've read Ned block and David Rock I've read so and so on and um so as I say the question becomes is is the an issue like consciousness one we're simply thinking about it he's going to get us further forwards now wait when people have done in Oxford philosophy and psychology and then asked

37:14 me what they should do afterwards and they said they want to go on doing both I said no no no don't do that drop the philosophy just go and doing the science and the reason is that science changes very quickly you know for any young scientist all that matters is what's happened in the last five years things are moving a pace philosophies a bit different yeah I talk to philosophers and they still talking

37:49 about Plato so I say no no no go into science now as you know Hackman Lao worked with me as a graduate student and there's a post up afterwards and he was one such that that I said no no no no you know drop the philosophy just do psychology but as you know and obviously you know him and you've interviewed him a Hackman has in fact had up his interest in philosophy and he writes in

38:25 his blogs pieces which can reasonably be thought of as being philosophy yeah so let's get back to question-there sitting in the armchair and just thinking about consciousness can he get anywhere well I just read Tim Bane's article in which he looks at two noni and others

39:00 integrated information theory yeah the axioms of IIT exactly and I read that and I see it is a very good solid piece terrifying whether these genuinely are axioms from whether the deduction said genuinely deductions but of course I don't at the end find the solution to consciousness at the end of the article right so what he's done is a very very

39:37 good job because philosophers are trained to think very clearly and they're trained in logic so let's get back to consciousness does a training in logic help us understand consciousness well the alternative is to study it so I'm comparing sitting in the armchair thinking about it is supposed to get in

40:11 the lab on doing experiments so let's take one of the experiments that thun only did well he showed that as you wake up what happens in the brain is that areas that were not talking to each other not communicating with each other begin to communicate with each other that's where the integration integration comes from

40:48 and he did he didn't experiment using transcranial magnetic brain stimulation which which showed this effect and there are other people in Oxford who've shown that as you go to sleep so areas stop talking to one another and so on so experiments are beginning to tell us something about what happens when we go from sleep to wake or wait to sleep and then their

41:24 experiments of course in which you present stimuli which people either do see or only see sometimes because they're very dim or whatever and you looked at see what happens in the brain when they see something as opposed to when they don't see when they miss it and we know from recent experiment bystander in in Paris that if you look in the brain in a monkey

41:58 these are macaque monkeys and you compare the first visual area to which vision cortical area to which vision projects and then a later one called v4 and then prefrontal cortex so he's recording in these three areas and he's recording with electrodes and you compare what calm hits with misses so the monkeys asked to detect targets and

42:33 presses a button when it's detects a target and of course some of the targets it will detect and some because then they're dim they'll miss and you can compare hits with misses and if you look in primary visual cortex or the visionary of before on there's not very much difference in the activity but you look in prefrontal cortex and there's very little activity for missus in other words what's

43:11 happening in prefrontal cortex seems to be closer to what you and I are aware of than his activity earlier in the system now those are the sorts of experiments that seem to me to begin to tell us something about consciousness and my worry is what is sitting in the armchair egg to do now philosopher to say to me

43:52 well look as a philosopher I'm trained to think logically am I reply to that is I can think logically you don't need to tell me why can't I do that thinking where is the room for somebody suggested they're thinking right now I think that's a very important point and wonder if I could kind of trust you a little

44:25 bit on this and follow up so it so yeah I think you're kind of unique with this philosophical background that you have as a scientist I don't know a lot of scientists who were doing psychology and philosophy at the same time and I think maybe there's a risk of especially when it comes to consciousness but what the mind in general I think there's a risk of people thinking they're experts already because they have a mind they are conscious and so there there's a risk of scientists sort of taking their

45:00 naive pretty theoretical assumptions about what these phenomena are I think to Nonis in a special example of this kind of thing because he starts with these axioms which really just encode his pretty theoretical intuitions about consciousness this is how he thinks of the phenomena as being unified having cause and effect power all that kind of stuff so without doing the training first of being very clear about what assumptions are what justifies assumptions which I don't think

45:32 scientists are really taught to be honest there's this you run the risk of importing your naive assumptions into your science do you do you think that that's a fair characterization or no I think my challenge is tell me one thing that a philosopher has found about consciousness well I mean so is it that they have to find something or can they

46:06 just contribute to clear thinking about it so that others can be clear about what they're looking for why can I think clearly now you're saying – no he doesn't think yeah and now today I mean he probably was a mathematician or what I don't know what his training was you know yeah it's not that I'm not saying well I do but it's not that I'm saying that he doesn't think clearly because I've met him I know he thinks very clearly but what what I'm saying is that there's a danger in importing assumptions and taking them as kind of bedrock because you know he

46:45 has conscious experience in his experience he it seems this way to him so then he goes and tries to formalize that into some kind of axioms but the axioms are really they aren't axioms in the sense that you find in mathematics or logic they're just assumptions about the way consciousness works which would also need to be tested not not empirically and not used as the starting point of constructing a test i I think I mean III I don't have any sympathy with the weights Melanie's going about it as

47:18 I say I would mess with her the sorts of experiments simple experiments that I've been talking about and I'll come to one critical one in a minute you know was I think – no need did do an experiment very early on which showed that areas communicate with each other when we become aware as we go from sleep to waking and I think that was important but many other people are working on

47:54 communication between areas and in different states I I wouldn't want to try to study consciousness by setting up a set of axioms in the way he is I would simply want to do some very simple experiments like ask where in the brain is the activity closest to what the person reports now I've told you

48:33 that the the answer to that from the stand-ins experiment is it looks in a monkey as if it's the prefrontal cortex now there's a very simple test if that were true and if prefrontal cortex were removed from both hemispheres the question is would the monkey be aware

49:08 and the question is can we test that now Allen karien Petra story did an experiment with monkeys and it went like this they taught the monkey if you see a light touch it on the screen if there's no light touch a square and then they removed the visual cortex from one hemisphere and that means that the monkey couldn't see anything on the

49:44 opposite side of space they then tested that the monkey still knows the rule so if you present in the space which the monkey can see because that's supported by the intact hemisphere if you present a dot the monkey touches it if there's no not the monkey presses the square so the monkey still knows the rule now if you present the spot in the blinds part

50:20 of space quotes the monkey always presses the square he always says he he but he/she doesn't see it so we have a way of asking the animal did you see it so if prefrontal cortex were critical for awareness and if we remove

50:53 prefrontal cortex in principle it ought to be possible to ask the monkey this way whether it sees things now a long time ago in 1962 Pat Goldman Asha de Noirs did remove prefrontal cortex in both hemispheres in monkeys and they could still discriminate between visual

51:29 stimuli that is they could tell the difference between a plus and AD swear and they were not blind by which I mean they didn't bump into things so the question is does that show that you don't need prefrontal cortex for awareness or as hack one loud things were in fact these monkeys showing what

52:04 Larry West grant called blindsight in other words the ability to do the task while actually you're off aware of the stimuli mm-hmm so I haven't found a way I'm sort of trying to move through a way of trying to find I'm not going to do this experiment but I'm thinking now what would the experiment be that would tell us whether you needed prefrontal cortex

52:39 for awareness because as you know in the field there is a dispute between whether it's sufficient to have activity in early visual areas or whether you also need to have activity in prefrontal cortex way and a critical test would be what would happen if you had no prefrontal cortex now I don't see in this sort of conversation how philosophy

53:19 is going to contribute except in the following way it worries me in the experiment that Alan Kerry and Petra story did it's very easy to conclude that what the monkey was doing was telling you something about its awareness but what it may have been doing is simply telling you something

53:56 about the external world right and saying there wasn't a spot so that brings up the issue of whether by awareness we mean what we call metacognition do we mean when we say that we are aware that we are aware of that so we know that hakalau with a philosopher wrote a paper on

54:37 metacognitive theories of awareness and that's an example where you know a philosopher I think deep contribute to be issue of how should we define awareness does it necessarily involve ethical a meta Commission or not I don't you'll be on lassies well you know pick hoc one and I have a paper on this where yeah I might be that philosopher your

55:20 Osito okay yeah David yeah I studied with David at the Graduate Center yeah yeah so I mean my own view is that these are it's it's an empirical issue but and I I don't really know but my view is certainly the evidence is seems to be not in favor of just the sensory cortices doing that work yes oh so it seems to me I did if I had to bet right now I'd bet that the prefrontal cortex is sufficient was it serious which I've accused hacker on a defining awareness

55:57 in such a way that it has to be metacognitive and he comes back at me as hack one does and says no no he's not defining it so I think this is one issue where philosophers could help us when we mean I am aware of X in the external world does that or does that not involve metacognition and you don't think that's

56:36 merely a definitional issue III worried that it might simply be that Hackman had defined it so but it's the sort of thing well I mean that's exactly philosophers training right right so saying that I don't like philosophers I'm challenging them to come up with I'm challenging them to move the field on

57:10 and there's a problem which is the way I did philosophy in Oxford the philosophers of mind were simply not interested in empirical science and it has become a tradition in America in North America in general through Ned block and Searle and and many others that you do take note of empirical things but when I read Patricia

57:49 Churchill's book sorry she's written more than one but the one a fret brain trust her I just felt that she was just giving an account of neuroscience I didn't see where the philosophy came in yeah I mean Patricia's work is is trying to use neuroscience to deal with philosophical problems which she calls neural philosophy so I guess that's she would take that as kind of yeah and that's what she used

58:23 but it's not supposed to be just philosophical it's supposed to be neuroscience trying to answer questions that philosophers have traditionally tried to answer well obviously there are many problems that philosophers have tried to answer like what is life and and there was then a long beetle and all the rest would have turned out to be issues to do with DNA and copying and development and so on and philosophers are worried about

58:58 consciousness and it is my feeling that that's now coming out of the philosophy fields and into empirical science yes now there are other issues though than consciousness that philosophers think about and talk about and then discuss empirical data for example soul discusses the Labatt experiment yeah

59:34 it's the famous experiment done in 18 1989 when he showed that if you ask people when they're aware first aware of pressing their finger it's about 400 milliseconds before they press it whereas if you look in there eg you've studied the brain waves from electrodes outside the brain you can find a signal 500 milliseconds or even a thousand milliseconds before and there's been a lot of discussion on whether the

1:00:14 experiments right or whether the interpretations right and so on and philosophers have contributed to that discussion by the way as citing what I just read that the limit experiment was recently passed it some replication test someone replicated it pretty well actually well how can hack when did you to my lab on a week we got 400 milliseconds and so on and no it it

1:00:51 passes his test but there are worries the question is when will you first aware but in mark Hallett's lab he thought of another way of asking when you were aware and he had beeps coming on every in our name and when you've heard a beep you had to think were you aware of getting ready to make the

1:01:25 finger movement and he got times which were much earlier but he didn't have to be a philosopher to do that let's get back to so this issues of free will does that abet experiments showed that my brain is in some way pushing me about well that's ludicrous of course because I am my brain in my body there's nothing for my brain to push about then Searle

1:02:00 says oh well for free will we have to have some sort of random generator in our brain and if we're not careful we'll get into quantum physics and all the rest but clearly there is an issue an interesting philosophical issue of the difference between free will and things

1:02:38 that we do are not free we we do them I can sell used to just put his arm up in elect to demonstrate Freebo he still does that it's not free now it's automatic yeah he should be wiggling his nose but it is an issue that animals not just us

1:03:13 animals do things of their own are called they don't just respond we don't just respond to external stimuli there are internal changes in the brain that lead us to doing things and so does that bring up a question of free will I I don't see why the fact that there are internal changes in my

1:03:49 brain that led me to doing things means that I'm not free the argument presumably is that there is a causal chain leading up to those internal changes and if there's a causal chain where's the freedom and my worry is that philosophers will go on worrying about that till the next century so I this is very interesting what one question I

1:04:22 would ask is well I have two questions but one that I'll ask right now is so if you take some of the work like from John Dylan Hayes Haynes yes lab where they they try to extend some of these limit findings and as far as I have understood it they they get like the with with multivariate pattern analysis they can predict like almost up to three seconds in advance decision John claims it's much earlier but but but but let's say three seconds just just for the moment okay so if yeah good Oh

1:04:58 I say so just one way to try to press this free well question might be in terms of prediction rather than in terms of cause and effect so if if you could predict if you in principle could predict what someone will in fact decide to do well before they have decided it and doesn't that seem if that were established empirically wouldn't that challenge traditional notions of free will or even there do you think that there's a well free did siegfried has recorded in the brains of patients when

1:05:32 they're doing that abet task and he finds cells firing well before the movement the prediction was at 57% right barely above 50% is but of course it's not fair for me to say that because techniques will improve if you if Jim Haxby shows people different pictures

1:06:06 like a chair scissors a person and so on he can using multivariate patent analysis get predictions up to 80% or 90% of what they're seeing from studying the pattern of activity in the brain so it will be fair to say what would happen if John's methods improve and he gets that prediction up to 80% and you're asking me given that you can predict

1:06:41 what they're going to do does that mean they had to do it or would it put tarnish I wouldn't put pressure on the more traditional notion of free will as as lay persons consume it because I think there's also this issue that philosophers like to talk about compatibilism that you know there's a kind of free I know Dan it's work is involved heavily in this that there's a kind of what called control system in the brain which may be completely determined but still senses internal to us and so forth it's a sense in which we're making our own decision

1:07:16 so I think that's one area philosophers to contribute is by distinguishing the traditional notion of you know the random I can do anything screw the laws of physics kind of free well from this other well isn't there some kind of control mechanism inside the brain which produces our responses which may be determined or not but so what am i question so that's a different issue but my question was let let me just stop either okay I want to agree with you in other words what you've done is picked up another issue as in the case of

1:07:50 awareness where we need to be clear what we mean by awareness we need to be clear what we mean by freedom by freewill and I don't want to stop philosophers thinking about that because I think that is indeed their job and that's going to be helpful if we understood but especially because of issues of law right because it's you know so you find

1:08:26 somebody and they've got a tumor in the amygdala and they go through the top of the tower and they kill people and are you going to send them to the electric chair or not were they free or not these are issues where clearly we they're like issues of moral philosophy where where they're central to philosophy so yeah I get what's deny that's an understanding the concepts the use of these words is

1:09:05 going to help us now I interrupted your question and I'm hoping you forgot me [Laughter] I mean my question was just whether if we did so if the kind of work from John's lab was really accurate and could predict up to five seconds I've heard some Donna Singh claims you know from him but so say you could predict in advance what someone will in fact to do doesn't that sort of put pressure on the traditional notion of free will so so

1:09:39 whether or not so to sort of bypassing the question of yeah are there causes which precede effects that's sort of obvious brains massively connected obviously some things happening before I decide to do something that's kind of the trivial but isn't there still an issue here that could impact the more philosophical question of do we have this kind of so-called libertarian free will ability to you know have done otherwise or these other more you know but but I think now what you're saying is that findings in neuroscience are challenging some of the concepts that we

1:10:19 or the lay people have had in the past yeah that is one place or even philosophers and that and perhaps that's why patricia churchland is doing what you call Neera philosophy exactly she's saying that there are findings in in neuroscience which are mean that that philosophers need to change some of the concepts that they have or clarify some of the concepts that they have or whatever yes but but my word is the

1:10:54 other way around it's it's what a philosophers contributed and and I've admitted the Einstein case in philosophy of physics so he's but what do I have to I can't say that sitting in the armchair somebody is not going to make a contribution to the study of consciousness I can't say that and that will be like saying before Einstein that sitting in the armchair he couldn't have

1:11:27 done it right I'm just saying I'd like to know what this contribution is so far yeah I I think that's a fair point I mean in fairness you could make the same point a lot a lot of psychologists from the early like for instance Freud I think probably comes to my mind as someone who not exactly from the armchair but from single case studies at least generalized quite a bit from abnormal cases to the general population a bunch of in my opinion wild speculation which can't be

1:12:03 tested in any way that I mean in one it is tested is shown to be mostly false except for a couple of things you know projection that I think is a stress this asterisk total disaster and in fact I in Oxford there was when when I was an undergraduate there was a post palm readers this was under being a professor it was just being under them and there was a wild reader in philosophy mental

1:12:38 philosophy it's now actually a professorship in philosophy and it was in the psychology department and the person who gave lectures talked about Freud and his whole point was that it was not falsifiable and I think it's a Pauling disaster that in France and in America Freud still influencing psychiatry and Italy in a way that is

1:13:14 less typical in the UK hmm I'm not saying that that there is no no psychoanalysis but but it's more predominant and if you look at the development of the diagnostic tool for mental diseases DSM you can see the imprints of psychoanalysis in some of these terms yeah yeah though don't get me on the

1:13:49 subject I don't want to keep you all day so I want to let you go but I wonder if I could just ask you one kind of more general question about the state of the field and it has to do with we've been talking a lot about consciousness in general but I wonder what you think about unconscious processing yes do you think so do you think that that has been demonstrated to a sufficient degree the that you hear a lot about the cognitive unconscious not Freudian I don't want to the problem was that these priming

1:14:28 experiments that were done in the Netherlands or Denmark I forget have turned out to be very difficult to replicate I'm worried about it in the following way I have a lot of my best ideas at three o'clock in the early morning I wake up and problems that I've been thinking about now I've solved of course where I wake up at eight o'clock

1:15:06 in the morning they're wrong this goes on and on or not so the worry is it looks as if I'm thinking about these problems over not now of course we know that the brain is active overnight oh not in the same stages during the day but then I say to myself how do I know that the into

1:15:40 alternative explanation isn't right which is that when I wake up my mind is clear and then I eat on the hit on the answer right so it doesn't seem to me to prove that actually I was thinking about those problems overnight now I'm not saying I wasn't thinking about them I'm saying that sort of demonstration and there are famous ones in science where people overnight dream things all the

1:16:17 double-helix you know that doesn't show it the truth is that I'm off my patch when you ask me the question and especially when you get old there's a huge danger that you pontificate about things which are not your expertise so you must be a permanent decent question

1:16:51 which is about unconscious processing and I think I have to say that I I'm not an expert which sounds very weak but when I see sort of physicists talking about microtubules and consciousness and it worries me and I knew I heard Eccles who was a very distinguished physiologist yeah I couldn't kind of said something strange in person in there I heard him when he was an old man and mr. danger just look at the bald cap

1:17:28 I appreciate the honest answer and you know going all the way back to Socrates this is a been a problem that was Socrates main criticism in one way of thinking about it was that people who who were experts in one area thought that that entitled them to say things about another area of which they knew very little so they certainly appreciate that honest answer but I wonder so given that though isn't there a problem for how you set up how we should study consciousness since there it seems like one of the main things you want to do is contrast cases where you know subjects

1:18:04 just take backward backward masking or something like that so the the subjects were port that they don't see it and other cases report but they do see it and then we look in the scanner to see what the difference is and we say well when it's not conscious we see this when it is conscious we see that but that sort of already tends to assume that we have actually a general case of someone conscious processing going on well I think no I mean we can take that very simple case if you record selves during backward masking we know what happens

1:18:40 what happens is the signal is simply much smaller does the cell activity goes on for less time and if you record the growth signal it's much smaller if the mass comes very soon after the stimulus that you're meant to be seeing so here we're simply talking about how big does the signal have to be the internal signal has to be for me to say I saw it

1:19:17 ah hits a supposed to misses and that's a simple criterion presumably of amplitude but really your question was about home conscious thinking right and that's what I'm saying that another next Photon I think you know as I think I can cope with what's happening in backward masking I think that's simply a signal size problem and if we were able to look trial by trial a hits and misses my

1:19:57 guess is that you would see that the missies the signal was always below a criterion level that experiments not been done trial by trial but that's what I would guess somebody tried to do it now so that's that's that's good and I yeah you're right about what I was doing so I yeah that's a useful correction but I wondered so the last thing I want to ask you about it let me just one thing because it worries me this signal because Cochrane now has made a

1:20:33 particular issue that if you're studying consciousness and you've got a conscious and a non conscious state then the performance must be managed and he with me did an experiment in which performance was matched with backward masking and in one condition they were more likely to say they saw it than the other now my worry is that the signal wasn't matched yeah that's very

1:21:13 interesting I don't know that the signal wasn't matched as you've got to do this trial by trial experiment but it's just a worry I have anyway yeah that's almost gonna derail me from like just make sure that hack Ron doesn't watch this year he already emailed me and says he is you'll probably get a scathing call from him soon just as a final question I wanted

1:21:50 to ask one thing that I think is missing from this conversation also from your introductory book which by the way which I highly recommend if you can see this here I think my is my feed gone dead over here can you hear me yeah okay sorry so one thing I think is missing from the discussion is a general kind of theory of what's going on in the brain a theory about what all the firing is doing all these neurons so there's a lot about connectivity what parts connected to what there's a lot about what parts

1:22:23 are differentially active when we're performing various tasks but I wonder what you might think about attempts to kind of give an account of what that firing means and I'm think is particularly like predictive coding and these kind of Bayesian models that are becoming very popular so what I just want to get your reaction to that and also the more general question of how important is a theory of brain functioning to doing cognitive neuroscience I'll allow you to ask the question so long as you don't use the terms free energy okay you use I did say

1:22:56 that the aim of neuroscience is to understand how the brain works and there are 89 billion cells neurons and we know that the astrocytes and so on are doing some important things too and we don't know what it is and there can be up to 26,000 connections on to any one self so we're now came to understand that brain

1:23:34 without producing computational models and in fact lutea Viner and i published a book edited a book which came out last year in which various computational neuroscientists tackled what the cerebellum does and how it does it what the neocortex starts and Harry glasses and what the hippocampus does and how it goes it now the way I see it is this take any one area it has a unique

1:24:13 pattern of connections and it's gotten it unique the inputs that is an a unique pattern of outputs now the cell connections within an area are fairly similar across areas but not identical so we're to think that each area performs some transformation from input to output

1:24:47 given those set of inputs the internal wiring and those set of outputs of the areas it can influence and it's clear that our job is to find out what that transformation is and we cannot say that we understand what area any does unless we've specified that transformation but then you've got the fact that the brains

1:25:23 interconnected since the brain we're trying to understand so now we really are in a big computational model now Henry Markram was given huge amounts of money in Switzerland to try to solve the brain and he uses blue brain which is IBM's huge computer right and he's worked out the wiring of what's called

1:25:57 whisker barrels in rap in rats and mice so they've got whiskers and in the primary area for touch there are what are called whisker barrels and he's worked out the wiring in intricate detail and my problem is how are you then going to go from that to understand in the brain you can't simply add a lot of these together and say you've got a brain you've got to connect them in the

1:26:33 way that the brain is connected now don't see us doing that so I think the better way is the way that deep mind is approaching it that is build circuits which have some resemblance to the brain and by teaching them by giving them

1:27:12 experience with faces or whatever teach them to discriminate faces or voices or how to play go or whatever and what really excites me is this there was a paper recently that came out in which deepmind looked at how you find your way and we know from rat work in particular done by John Okeefe and Moses and they

1:27:49 got the Nobel Prize from yes we know we know something about rats help rats find their way and there are cells with different properties place cells boundary cells and so on and the extraordinary finding was that when deep mine taught their program to find its way and then they looked inside it they found that the way it had found to solve the problem was in fact the way that you

1:28:25 and I do it Wow because of course the worry is that you might teach how to play go but it's nothing like the way that people do go or you might teach how to play chess but it's nothing not the way the Kasbah does chairs and that's the way that early programs did it yeah that's the way IBM approached it but the way dig mines doing it and demis hassabis who'd worked with Helena McGuire on wayfinding at the

1:29:03 Imaging Center in London and he's very influenced by what he knows about the brain but he's building these huge connectional nets with particular properties and if it turns out that the way they're learning it is similar to or it they found the same way of doing it then we're getting someone yeah then

1:29:37 we're that's a really that's a really cool finding I'll have to look that paper up and include it in the links down below as well well if you send me an e-mail I'll email you the that's why thank you because I had been raised that 60% or something of the I know I won't give up figure anyone good I was gonna say my statistics teacher famously told us that 98% of all statistics are made up but I was gonna say is less than 0.05 Yeah right I was

1:30:18 just gonna say I was kind of surprised that they when they opened up they found something because my understanding was they were having kind of problems figuring out what representations if any were going on inside deep mind especially when I came to like deep dream and look I I may have got it wrong um you know I'm not saying that understand them no I'm they're going wrong but I think that that's what they found and as I say oh I'll send you I'll send you the link if and if I'm wrong I won't send you the

1:30:52 link that's the way to go yeah very interesting you know I wouldn't try to build a brain with the identical connections and so on not mind root right yeah no that makes sense so on that note is there any final comment or anything you want to say before we sign off yeah sorry [Laughter] very good see totally sincere well I

1:31:36 want to thank you very much for joining me I know your time is valuable and I really appreciate you taking this time to talk with me it was a great conversation I really appreciate it very nice okay I'll talk to you soon I'll send you that in they'll shortly okay cheers Cheers