Matthias Michel: Is a Science of Consciousness Possible?

Okay, and we are live. Welcome to MindChat. I'm Philip Gough. Hello, welcome to MindChat. I'm Keith Frankish. Are we recording, Philip? We are. We're there. Great. Good. Good. That's good. Well, we're honored today to have with us Matthias Michel, who's currently Bersoff Faculty Fellow in the Department of Philosophy. in New York University. His research focuses on foundational methodological issues in cognitive neuroscience of consciousness.

and studying how scientists investigate consciousness on the methodology of consciousness science. And he's a young researcher, but he's already established himself as a... Very important figure in this field. And we're delighted to have him with us today to talk about the possibilities for a science of consciousness. So, Matthias, welcome.

Could you perhaps begin by saying a little bit about yourself and about how you got into this field? How did your interests focus on this area of the science of consciousness? Yeah. Okay. So I started being interested in consciousness when I started doing analytic philosophy in France, because one of my first classes in analytic philosophy was... on philosophy of mind and obviously on consciousness as well. And then I realized that a lot of neuroscientists were working on these topics.

as well on consciousness, and that there was an actual science of consciousness. So as I've always been... interested in cognitive science and neuroscience as well, I started studying, thinking about, oh, what are these people doing? So that's how I got interested. in the science of consciousness in particular. Yeah, I think one of the things that really pushed me towards the approach that I have right now is Liz Irvine's work, in particular her book, Consciousness as a Scientific Concept.

in which he raises a number of methodological challenges against the scientific study of consciousness. And I wrote my dissertation on... these challenges, showing how despite all these difficulties, maybe we could have a rigorous science of consciousness. Brilliant, thanks. Well, that's exactly what we're wanting to focus on today. But so just before we get to the main discussion, just.

you're freezing some listener questions but don't confuse but towards the end questions up and hopefully we can give Matthew questions so and if you aren't enjoying these videos please do like and subscribe to help us get the content out to a broader audience but okay so our topic for today is is is a science of consciousness possible is are there some reasons that consciousness is

a more challenging phenomenon to investigate scientifically than other empirical phenomena. And Matthias, you wrote a really interesting paper, the subtitle of which is A History of Endless Debates.

consciousness science underdetermined, which is open access, I think, isn't it? So people can read it if they want. But I mean, we want to get to the kind of the substance of these naughty questions. But the paper starts off with just... really interesting history of how these debates, these same debates have resurfaced over the last 150 years and on some key questions that doesn't seem to have been any consensus reached or any.

or converge towards so i think it'd be good maybe just to start off by if you could share with listeners or viewers some of that really interesting history yeah sure so um As I think about it, the history of scientific research and consciousness can be traced back to the 17th and 18th century, in particular to the conflict or to the... to two different philosophical traditions, two ways of thinking about the mind, and this has been described in an article by

Alison Simmons in 2001, a fantastic article entitled Changing the Cartesian Mind. And so she describes two philosophical traditions. On the one hand, there's the Cartesian tradition. and the English traditions. On the other hand, there's the German tradition or the Leibnizian tradition. So in the Cartesian tradition of the mind, the idea is that consciousness is the mark of the mental.

If it's mental, it's got to be conscious. And a kind of corollary of this is that everything that's in your mind is accessible to introspection. So introspection... is the capacity that you use in order to know your own conscious mental states from the first-person perspective. All that's happening in your mind is conscious, and all that is accessible to introspection. On the other hand, there's another tradition, which started with Leibniz, who started thinking about the mind...

not as characterized primarily by consciousness, but as characterized by representationality. Representationality is the mark of the mental, not consciousness, according to this Leibnizian tradition.

So the main job of the mind is to produce mental representations that indicate stuff about the external world, and that's what the mind is doing. And the kind of revolution... that this brings is that once you define the mind in terms of representation and not in terms of consciousness, you can start thinking about unconscious mind because.

In the Cartesian tradition, that doesn't make sense because if it's not conscious, it's not the mind, so it's not a mental activity. And so that's how the conscious versus unconscious distinction can start.

because now we think about the mind in those terms. Following this, scientists started being interested in consciousness. scientists and philosophers. At the time, the distinction was not so relevant in many fields. And so they started having a series of debates, I think. four main debates. So the first debate is a debate about the unconscious mind. The second debate is a debate about unconscious perception, which is a particular mental activity.

There's a debate about the richness of consciousness, how rich consciousness is, is there consciousness, how many items can you be conscious of at once, and then there's also a debate about the physiological conditions of consciousness, what kind of brain stuff or nerve activity, as they say at the time, is sufficient for consciousness. These are the four main debates at the time.

Right. And so and so this I guess this this idea arising that we could have unconscious mental states connects with. the contemporary discussion of of the the what's called overflow or the overflow thesis um and you know it's really interesting because you say a lot of people often say oh the science of consciousness starts in the 90s and one big

fixture of the modern debate is this overflow thesis. But as you nicely say in the paper, people in the early 19th century were discussing pretty much exactly the same thesis. So maybe should we lay that on the table? What is the overflow thesis? And why does this create really specific challenges for the science of consciousness? Sure. So the overflow thesis, like the recent version, was developed by Ned Block. And the idea is just...

that there is more to your conscious experience than what you can notice at any given moment to be present in your conscious experience. The basic idea is that what you experience... What you feel overflows the limits of what we call working memory, which is what's really present in your mind at any given moment. So, phenomenal consciousness, as Ned Blox says, overflows. access consciousness. And what I can access is what I can think about, basically, at any given time.

This distinction was actually something quite close. It was actually developed very early on by Condillac in 1746. where he distinguishes between two kinds of consciousness or two kinds of conscious experiences. He says, there's the kind of conscious experiences that I remember and the kind of... conscious experiences that are so faint that I don't remember them, even if they've just passed, because they are just so faint. Think about, for example, the weight...

of your clothes on your shoulders. It's a kind of faint sensation. It's not so clear that you always have... Then there's the question of whether you always have that conscious sensation. or not. So overflow theorists will say that even when you don't think about it, you have that sensation, and people who don't believe in this kind of phenomenal overflow will say that...

You might have that sensation, but completely unconsciously. Right. That's a nice example of clothes on the body or is it connected? I get David Armstrong's famous example, you know, where you're driving along in a familiar road and you're just. thinking of something totally different and then you suddenly are aware of the road in front of you but of course you've been driving accurately so the question is when your mind was on other things were you

conscious of the road in front of you, but in such a way that you weren't really attending to it? Or does the fact that you're not attending to it mean you weren't really conscious of it at all? So somebody who accepts the overflow thesis will say you were conscious of the road in front of you when your mind was on other things, you just weren't attending to it.

Whereas someone who denies the overflow thesis would say, well, that doesn't make sense. How can you be conscious of something if you're not aware you're conscious of it? So for some people, that seems kind of contradictory, right?

Right, that's the debate of whether, so a very closely related debate is the debate about the relation between attention and consciousness. If you believe in the overflow view, you'll tend to think that attention is not necessary for consciousness. You're conscious of more than what you attend to at any given moment.

But still, I think it's important to differentiate between two debates. There's the debate about the unconscious mind in general, and then there's the debate about unconscious perception. And then there's another debate about perception without attention. Because you can believe in unconscious perception even if you don't believe in perception without attention, in conscious perception without attention.

You can believe in unconscious minds even if you don't believe in unconscious perception. So just lay out those three things again. Right. So there's the debate about the unconscious minds in general. And this debate is pretty much solved, as far as I can tell. And that's a debate in which... We have made a lot of progress. And basically, the idea is that we've realized that in order to explain a wide variety of phenomena, of mental phenomena,

we have to appeal to unconscious mental processes. And these processes are mental processes. So that's relatively uncontroversial, then, I guess. That, I think, is whether there's unconscious perception. right so there's the debate about unconscious perception and then there's another debate about perception without attention right or conscious perception without attention right

So why do these, I mean, they're kind of fascinating debates in their own right. Some people think that just seems contradictory, that you can have a conscious state you're not aware of or you're not tending to. But why does it matter for consciousness science? Why is this important for trying to study consciousness? They sound like kind of abstract philosophical conundrums, but why do they impact real-world experimental consciousness science?

Well, first of all, it seems like a really interesting question to determine the number of things you can be conscious of at once. So that's one thing. But second, it's also really important to be able to demonstrate

unconscious perception if you want to study consciousness. Why? Because if you want to study consciousness, one very nice thing that you could try to do is compare perception when it's not conscious to perception when it's conscious, and then you study the difference between the two.

And so the kind of stuff that you can get with consciousness will help you understand consciousness. For example, you can put people in a... an fMRI and measure their brain activity between those two conditions, and this should reveal some physiological basis for consciousness, or at least that's one of the hopes of consciousness research.

If there is no unconscious perception, or if we don't know whether there's unconscious perception, that could raise a lot of problems for... for the scientific study of consciousness, because then you're not sure whether you're contrasting conscious versus unconscious perception or just conscious. strong conscious perception versus weak conscious perception that participants don't report.

yeah sorry i think i'm probably going to ask the same question but just jump in the way so we have unconscious perception okay which is i guess which is your sensory systems taking in information about the world and it's using it to control behavior at some kind of low level, but you are not noticing this, you're not in a position to think about it and tell other people about it and so on. So it's having some effects on your psychology, but not widespread effects, not effects on...

what we think of as ourselves, as it were. And then there's the kind of conscious perception that is perceptions that have attended and are conscious in this strong way where we can report them and i can say yes i'm perceiving this and so on but then there's supposed to be a middle position that is stronger than the the unconscious one it is conscious

but maybe I'm not thinking about it, not noticing it, not reporting it. So what's the difference between that middle one and the completely unconscious one? What's the character? This one's very different from that one, right? Because this one's conscious and that one isn't. But this one's not being noticed. It's not having the sort of effects that the strong folks...

What's the distinctive thing about this middle one that distinguishes it from the unconscious? Right. What distinguishes it from the unconscious is that it feels like something, or at least it's supposed to feel like something. Even if... if it's not in the focus of your mind, so to speak. And the feeling like something, is that a matter of it having certain effects on me? Or is it just something that's intrinsic to the state?

Does it have to have any effect on the rest of my psychology in order for it to feel like something? I would tend to think that it has to have some effect. But what matters is that it doesn't have an effect on your cognition.

Namely, you cannot think about it. But maybe it has some other effects that are not effects on cognition. So can we take a concrete example? Could I be, for example, feeling anxious? So I'm really feeling anxious. It's a conscious experience, but I can't think about it or attend to it. For someone who accepts the overflow thesis, could that be the kind of thing we're talking about?

That's probably not the most intuitive example. That would be a very strong overflow in which I could have an excruciating pain right now, but I don't notice it, I cannot think about it, and it has no... effect on my behavior. Let's put that aside for a moment. A more concrete example would be again the weight of my clothes on my shoulders. the kind of feeling of my back, the pressure of my back against the chair, background noises, stuff like that, that would be a better example.

kind of very intuitive if you have trouble understanding the overflow view. An intuitive example is change blindness. In change blindness experiments, there's an image... and then the image changes, and it goes on and off, and there's a change that's introduced in the image. in a zone of the image that's outside of your attention. And basically, it's very surprising because people can miss very, very... very strong and big changes introduced in the pictures.

It's the gorilla one, isn't it? It's the gorilla one. The gorilla one is a kind of inattentional blindness. It's not exactly the same effect, but it's a related effect in the sense that it depends on attention. If you know that... there's going to be a change in the picture, you're going to be much better. If you know where the change is going to be in the picture, of course, you're going to notice it immediately. That shows that this is an effect of attention.

What people who defend the overflow view believe is that you're actually consciously feeling what's changing, but you don't think about it. And a very intuitive way of making people understand this view is to... It turns out that you can do the same thing, change blindness, when... the change is extremely slow. So, for example, there's a good example that was developed by, I think, Atas and colleagues in 2014, in which there are three people...

and one of them is smiling, and his entire facial expression is changing, and people don't... You tell people that something is changing, and they are completely unable to tell you that this guy's face... is changing from happy to sad. So they're unable to think about what's changing, but in this case, it seems like it would be...

slightly nuts to say that you don't see it because you're just looking at the guy. You're looking at his face the entire time. And so that's a way of making the overflow. view a bit more plausible. I disagree with it, but I think it's a plausible view.

What you take to be intuitive, if you don't understand the overflow view, you take what you take to be intuitive in the case of slow change blindness. And I think that's very intuitive that you see something changing, but you're unable to think about it. At least you're not thinking about it. as changing, and then you take what you think is plausible in this case and you just apply it in all the other cases.

And then this is problematic for science coming back. So that's a really nice concrete example. But coming back to why it's a problem for science of consciousness, because is it because we rely on report? And so we ask people to report. But if you accept the overflow thesis, then what someone is able to report comes apart from what they're experiencing. Whereas people who deny the... the overflow thesis think what what you report goes perfectly hand in hand with with what you're experiencing so

I mean, you call in your paper this detection procedure. So there's in order to study consciousness, we need these detection procedures that link. external indicators such as report to conscious experience, but because of controversial things like the overflow thesis, how those detection procedures work.

is going to be controversial, and so the empirical science of consciousness is going to be controversial. Is that right? Right. So I think it's important to distinguish between two versions of the overflow. One is stronger than the other. There's the overflow view that says, that's quite intuitive, that's the one I mentioned, that says that you can miss stuff and you can...

be in a position in which you're not able to think about your experiences in cases in which you don't attend to the relevant visual feature, for example, in vision. So, that's a failure of attention. And it seems intuitive that if you don't pay attention to something, maybe you experienced it, but you cannot report about it because you didn't pay attention. Then there's the stronger view that says that...

People who are doing experiments, just sitting in a chair, attending to something, they say they don't see it, they know what they're supposed to detect, and yet they say they don't see it. These people could never do it. nevertheless experience it so we have examples of both of those so uh the the first view would be the kind of inattentional blindness or change blindness we were talking about you're not attempting

you're not attending to what's changing, and for this reason, you don't realize that you're not experiencing it. Another case could be, for example, visual masking. In visual masking, you present a stimulus, which could be, for example, a square or a diamond. You have to discriminate between the two. You have two buttons, and you press either square or diamond. Then the scientists ask you whether you saw it or not.

In that case, you know that you're supposed to detect either a square or a diamond, so you're expecting that. That's what is going to be shown. And in addition, you're just attending to the stuff. So if in those conditions, you say that you don't see the stimulus,

This seems like a condition in which your reports should be taken as much more reliable than in the first condition. So my intuition is that if you want to postulate... if people say that they don't see the smuse in this condition, where they just attend to the stuff.

the price that you're going to pay is much higher in a sense. It's a much more costly hypothesis because it's much more counterintuitive that you could... fail to notice experiences that you're having even when you're just attending to something, when some attention is available.

could we just i mean if we could just um sort of pull back the focus a little bit here because you mentioned at the beginning uh liz erving's work um was skeptical about the possibilities for a science of consciousness and we We've been just discussing some reasons, I think, for thinking that. But could you sort of lay out that sceptical challenge a little bit more broadly? There are some people who think that we just can't.

can't have a science of consciousness for methodological reasons it's not the sort of thing we can we can study using the scientific method and so what's the what's i mean we've touched on but what's the general argument for that

So there are actually a series of arguments. There's no one argument. The main argument, and I think the main problem, is actually something that we find in other areas of science. where that's something that Mar called the problem of coordination in 1896 when he was studying temperature. And it was also mentioned by...

in 1927 in his philosophy of space and time. So the problem of coordination is something like the following. Let's imagine that you don't know anything about temperature. You're completely ignorant about it. and you want to study it. Let's put ourselves in the people who started studying temperature as a scientific phenomenon. These people didn't know anything about temperature, they wanted to study it.

In order to study it, it's important that you're able to measure it. So in order to measure it, you're going to try to find some indicators that correlate with increases or decreases in temperature. And now the question is, how do you find those indicators of temperature, given that you don't know anything about temperature? It's not like you can find some indicators.

and say, oh, look, these indicators vary at the same time as temperature because you don't know anything about temperature. So that's the problem of coordination. And you have this problem in many scientific fields. And I think we have this problem. with consciousness science right now. The good news is that we've solved the problem in the past for other scientific constructs. The bad news is that there's no systematic way of solving that problem.

If there was a systematic lesson that we can draw from these cases of problem of coordination across history, it seems to me that we have to accept some unreliable foundation. before we can go further. We have to accept some imperfect procedure at the start of our scientific investigations before we can work towards more reliable methods. That's the...

only way to solve the problem. For example, in the case of temperature, they accepted that sensations of hot and cold could vary at the same time as temperature. Even though they were not completely stupid, they knew that... that sensations of hot and cold are not reliable indicators of temperature. In the same way, we can accept some procedures, like for example, verbal reports and so on.

as indicators of consciousness, even though we know that maybe those indicators are not perfectly reliable indicators of consciousness, and maybe later on we'll revise...

our foundations and find something else. So that's my general response to this problem. But Liz Irvine really showed that We have this problem. And basically, she raises a number of challenges against detection procedures, so procedures that we use in consciousness science in order to know whether people see the stuff consciously or not. That's basically it. So that's what a detection procedure is. And one of those problems is what we call the problem of the criterion.

That's slightly technical, but basically the idea is that in a psychophysics experiment, when you show stimuli and you try to find, for example, unconscious perceptual effects... It turns out that when the signal is very weak, when the stimulus is low quality, because we have to mask it, for example, or make it a bit weaker, otherwise the subjects would see it all the time.

So we have to weaken the stimulus. When subjects are not so sure whether they saw the stimulus or not, they tend to become more conservative, we say. So that means that they tend to say that they don't see the stimulus. even though they saw it. And that's a problem, and so Liz argues that there's no real way to solve that problem.

That's one of the challenges that she... Would it be right to say that something like this, that the challenge is that putting it very simply, that whatever it is we're measuring when we put... person in a certain experimental situation and we measure their response in some way what they say or whether they press a button or where their eye gaze is or something like this and then the problem is

how do we know whatever reaction it is that they produce, how do we know whether that was produced consciously or unconsciously? And we might assume that certain reactions... can only be produced consciously, but how do we know that? Because after all, as you say, we have to accept that there's a lot of unconscious perception, unconscious mental activity.

which could be producing the same, could even perhaps be producing a report that we would normally take to be a sign of consciousness. And that does look like a very...

a very difficult problem, doesn't it? You think we can... If there is a really clear distinction between reactions that are produced by conscious and reactions that aren't, we can sort of creep up on it gradually by starting with some imperfect way. and then gradually improving that way of detecting it. That's one problem for conscious science, but there are others too.

in addition to that one, did you say? Yeah. So for example, there's another measurement problem. That's actually not from, she doesn't call it like that, but that's what Erin tell. who's a philosopher of measurement called the problem of quantity individuation. So the basic idea is this. Let's imagine that I have two ways of measuring consciousness and those two ways of measuring or detecting.

would be a better term, detecting consciousness disagree in some cases. One says you're conscious of 90% of the stimuli. The other says you're conscious of 80% of the stimuli. The question is, whether those two measurement procedures or detection procedures measure different things or the same thing with different degrees of accuracy. And that's a huge problem.

So that's a problem whether you don't know whether the difference that you observe is due to a difference in accuracy or validity. Validity is whether your measurement procedure actually measures what you're intending it to... to measure. So that's another problem. Now, I suppose that the key question here is, is this just as you suggested, just another case of

another example of a problem that we've met in other areas of science and managed to overcome? Or is there something special about consciousness that makes this problem particularly, either particularly difficult or insoluble? I mean, a lot of people think consciousness is pretty special. So it wouldn't be perhaps too surprising if it did pose a very different kind of problem.

to the things that temperature or whatever poses uh now i guess your position is that it's not so quite so special is that is that right you think we can get a grip on it in the by these methods that have worked in other areas of science

right exactly right uh right right um i i i tend to think that i tend to agree with you there but i think that involves re-conceptualizing consciousness a bit to make it to sort of bring it down to earth a bit in that way so that we can get that grip on it i think the way some people conceive of it well it takes it out of that realm i would tend to think that at least at the start

we have to accept that some indications, even if you think they are imperfect, some indications are better indicators of consciousness than others. So we have to accept this. Otherwise, it's impossible to study consciousness. I think that doesn't imply that those indications will always be recognized as the better indications of consciousness.

It's just that's where we start. Does this way of thinking assume that consciousness is a really kind of sort of sharp edge thing? I mean, if we take this from unconscious perception through to... conscious but sort of unnoticed through to sort of reflective consciousness where you're aware of it. The idea seems to be that there's a sort of sharp point where it switches from being unconscious to being.

conscious and we need to detect whether that line has been crossed or not but what if it's just a matter of i mean unconscious perceptions have effect presumably a psychological effect and then conscious but unattended perceptions have effects and then attended conscious perceptions have effects so what if it's just a continuum of increasing degrees of psychological effects and there is no sort of

boundary where it goes from being unconscious to conscious. If that was how it was, then we wouldn't be able to find distinctive features that were distinctive of conscious. We'd just be able to say, well, it's more to that end than to this end. Is that a possibility, do you think? Or does the approach that you're considering assume that there is a fairly clear sort of signature of consciousness? Well, I don't think it can assume this.

Because it seems to me that that's something we will discover, whether consciousness is graded or not, which is also a big topic. One problem here is that It's not so clear what it means for consciousness to be graded. That's an area where we need philosophers. Because I think there's a real conceptual problem here. I think Philip's probably wanted to come in here. I can see him. Well, just so I guess...

I am inclined to think there's something different about consciousness. And so what kind of perennial theme I try to press is that consciousness is not... publicly observable. You can't look inside someone's head and see their feelings and experiences. Our fundamental way of knowing about consciousness is not from public observation experiment. It's not something we discovered in an experiment.

our fundamental way of knowing about it is just through our immediate awareness of our feelings and experiences if you know putting overflow to one side for a second if you're in pain and you attend to it you're just directly aware of it and so now of course Science is used to dealing with unobservables, but it seems to me in all other cases, we postulate unobservables in order to explain what we can observe so that when it comes to the data, it's all data of...

public observation experiment whereas in the unique case of consciousness i would say the thing we are trying to explain is not publicly observable or if i could if i could put it more carefully at least Not all the data is publicly observable data. Now, we could have a nuanced discussion about there obviously is publicly observable data relevant to the science of consciousness, obviously, but there's also pretty important data that's not.

not publicly accessible data, namely just one's immediate awareness of one's feelings and experiences. So this does seem to me, because people think, you know, the job of science...

is to explain the data of public observation. If you've got a theory that can account for all the data observation experiments, job done. But I would say in this unique case, there's more data. There's the data that... arises from our immediate awareness of our feelings and experiences and actually so just to finish that up i mean i mean i had a public if people are interested i had a public exchange of letters with um massimo pigliucci on this but also

You and I, Matthias, a while back had this mini Twitter debate on this topic. I mean, very brief. People could look that up as well if they wanted to. And you seem to be disagreeing me on this. But having read your paper, a couple of your papers here, this seems to me to account for. why these challenges arise because the phenomenon is not publicly observable and so we have to work with these detection procedures.

we have to rely on external markers of consciousness, like report, and then it gets controversial how to map them up. So doesn't this thesis... my thesis of you know it's not just my thesis obviously but the thesis i signed to of why consciousness is so different also explain why these perennial challenges arise that you've over the last 150 years that you've so well recorded So let's see. I'm not so sure that's what explains it. Again,

I'm not so sure I see the difference, the fundamental difference between consciousness and dinosaurs, for example. Let's say that I study sauropods. That's the phenomenon that I want to study. I want to explain how they lived millions of years ago. Sauropods are obviously unobservable. What they left are traces, so the traces are the data.

that I can use in order to study sauropods, which is the phenomenon I'm interested in. In the same way, I would tend to think that introspective reports are traces. that are left by the subject's consciousness on the detection procedure that we use in consciousness science. Just as sorobots left traces, consciousness is leaving a trace under the...

under the form of an introspective report. And that's what I can use. Just as paleontologists use traces to learn more about sauropods, that's what I can use in order to learn more. about consciousness i don't really see the problem with with that good okay so that's and that's what you said in the in the twitter response so and that's a really nice putative counter example i mean where i would say at least one difference so there's

Difficult to hear how exactly we understand the evidential significance of report. But just to put that on one side, what I could say is, in the case of consciousness... Not all the data is data. In the case of dinosaurs, all of the data is publicly accessible, fossils or whatever. that's the all of the data is publicly accessible whereas in the case of consciousness yes there are these there are reports and um other things that people in the no report paradigm appeal to but there's also

This other data, which is just my immediate awareness of my pain, my immediate awareness of seeing color. And it's that phenomenon that I want to explain that I'm directly aware of.

so and that's not relying on my reports of it it's relying on my immediate awareness of it and i would i i think we know stuff about the nature of consciousness in that way we know red experiences have this qualitative character that can't be conveyed in the purely quantitative language of neuroscience um i mean that's a step that's that's another step right but but

you know there is this way of knowing about it which which is the this is what i want to explain i mean some people deny this right so daniel dennett says you know all the data is just publicly observable And once you've explained that, that's job done. But I guess what I would say is the more common sense of you is that there's what I want to explain is this, you know. So your idea is, so Philip's idea is that, yeah, we do all this work about, you know.

theorizing about dinosaurs from the fossils and so on. But we also have this kind of vision. We've had this vision of dinosaurs, which, you know, we just we just intuit that dinosaurs existed. And so we know they existed. And then we sort of do the science from the fossils and so on to try and find out a bit more about them. But the certainty of the dinosaurs existing just comes to us immediately. That will be the analogy for consciousness science.

Your feeling that you have that intuition is an interesting bit of data itself. But the question is whether we trust that intuition of yours. It's like someone who's read the book of Genesis and thinks, I know all this. the only question now is to sort of see how that the data anyway go on go on let's assume for a second that um there are subjective data

In that case, I would tend to think... So that means that consciousness is observable in some sense, right? At least I can observe my own consciousness. So if consciousness is observable in that sense... that should make consciousness easier to study. Because dinosaurs are unobservable in all sense of the term.

observable, they're unobservable from the first and third person perspective. So if we have this infallible access to our own conscious mental states, to their quality, I would tend to think that that should make consciousness science easier than other scientific fields in which the phenomena that they're studying are completely unobservable, even from the first-person perspective.

Yeah, in a way, I mean, yeah, I mean, just answer Keith's part, I suppose it's a question of starting points, you know, it's like, I mean, how do we even know we're not in the matrix? You have to just make some starting points that I can trust my senses, I can trust my memories or whatever.

And I guess I think Descartes was right that the starting point of our immediate awareness of our feelings experiences is better known than the external world. But yeah, so why doesn't that make... Well, it doesn't, it doesn't, doesn't it? In a way, it makes the phenomenon... more certain, I would say. And I think there are things we can know about it. But then the trickiness arises from matching up, I would say.

from the inside, you know, with things like, like the overflow thesis, you know, do, do what I know from the inside, does that, does that go beyond? what I can report and so on. So yeah, in a sense it makes it easier, but it also seems to me a good explanation of these detection challenges because it's how to match up what you know from the inside and what you know from the outside.

Because you only have this... Sorry. It's just a matter of what you know from the inside and what you know from the outside. It's what you know from the inside by the special means of intuition, of acquaintance with the brother. And then what you know through...

cognitive processes, you know, you've got to link up that immediate intuition with the rest of your cognition. You've got to think about that and reflect on it and remember it and, you know, reason about it and so on. And all those are cognitive processes, okay, which, you know, are mediated then.

they don't that immediate awareness how do you try it's not just how you you line that up with what scientists are doing it's how you line that up with the rest of your psychology uh and as soon as you start thinking about it you're reacting to it it's just you're just like an outside observer reacting to your own reports anyway i don't want to get too off track here but um i think that that's i think what i was sort of getting when i when i was saying to michelle um matthias about um

bringing consciousness down to earth. I think you've got to separate it from those intuitions if you want to do this kind of science on it. Because you've got to think that we can approach it in the same way that we can approach other scientific problems.

I mean, what do you think, Matthias? Do you think there are subjective data? I think me and Keith disagree. And I guess we think if there are these special subjective data, then that's why consciousness is very different to other scientific phenomenon. I think there are, he thinks there aren't, but what do you think? Do you think there are subjective data, but it doesn't make it... I guess you think it's not really radically different from another scientific phenomenon, so does that mean...

How do you square that with the reality of subjective data if you do? Okay, so I don't think there are subjective data, but we have to define subjective data first. So to me, subjective data are data that only I can have from the first-person perspective that I obtain through introspection by being aware of my...

own conscious mental state, so I'm aware of my own conscious mental state. That's different from some people who would say that you cannot observe your own consciousness, but all you can do you know. Some people have defended what we call the transparency thesis, that consciousness is transparent when you introspect what you're actually pointing out. attending to are the qualities in the external world. So, for example, when I attend to my phone, what I'm attending to is...

When I attend to my experience of my phone, what I'm attending to is actually the external object. It's not like there is a kind of veil of consciousness between me and the external world, and what I'm attending to is the stuff on the veil. So if you want to defend the view according to which there are subjective data, you have to reject this view of introspection according to which you are conscious.

that you are introspectively aware that you have mental states with conscious qualities by attending to external objects. So for example, In a sense, I can be aware that you have a headache. I can be aware that it's probably painful for you or something like that. That's a piece of data that I can have. Of course, I cannot be aware of your headache, because that would require me to introspect in your mind. So there's the question of whether even I can be aware of my own mental states.

in that sense, of the quality of my own mental states, instead of being aware of the qualities of the external objects. If you reject this view, then there are no subjective data. And so I tend to agree with this kind of transparency.

thesis that we cannot directly attend to our own experiences. What we do is that we become aware that we have conscious mental states by being aware of external qualities of objects.

But let's make that concrete. Right. Like the yellowness of this cup on your view, when you attend to the yellowness, it's the surface quality of the cup. It's not something in your mind. Right. So I cannot acquire any data that... you could not acquire in principle because all I can be aware of is aware that I have such and such mental states, but I can be aware that you have such and such mental states as well.

Oh, sorry. I think we can put that debate aside because we are probably not going to solve it. There's a second question. Let's assume that there are subjective data, namely that I can be aware of.

my own conscious mental state. There's the other debate of whether subjective data counts as scientific data. Because, for example, I can have data that... I can obtain data that's not scientific data in favor of some hypothesis, for example, that the hypothesis that brandy is a cure for the common cold. I could have data about that, but it would not be scientific data. I just have a few experiences, and I say, that's data. No, that's not scientific data.

So the question is whether subjective data, if they exist, are more like unscientific data or more like scientific data. And I would tend to think, again, that even if they exist... Subjective data are not scientific data. So they're just more like hunches or something? Right. So, one thing is that there's also the difference between the logic of discovery and the logic of justification. It's really important to see the role that subjective data...

should play. For example, I'm not completely against if subjective data exists, I'm not against subjective data playing a role in motivating your hypothesis about consciousness, in motivating research about consciousness, but that's the logic of discovery. That's not the... logic of justification of your hypothesis about consciousness. For example, the famous example is Keculi's dream that led him to discover the molecular structure of benzene.

Discovery was made after a dream, but obviously, that's not the data that he used in order to justify his hypothesis. In the same way, if there are subjective data, maybe they can lead us to produce... hypothesis about consciousness, but then the hypotheses are justified through purely empirical means with those good old objective data. They are not justified by the subjective data.

Can I just say just one thing about this external quality stuff? Just right. So. Tale of flopsy bunnies. Yeah. So I would say, I think, I think. What I attend to is the kind of pinkish quality on the surface of this book. But in being aware of that, I'm aware of myself having a state, a conscious state that represents that quality.

And derivatively, the conscious state has that property of representing that quality. So is that okay? That would be... objective data to me because I can aware that you're representing that property as well but I still think I still think I still think you can't I still think I have access to the character of that quality And that can't be articulated in the purely quantitative language of neuroscience, physical science more generally. So there's something about the conscious state.

that I know about from the inside that can't be characterized in the data of physical science, even though it's the quality of my conscious state.

represents that the you know and but that's a property of the conscious state that it represents so charm has caused these edenic qualities right you know these like colors as we naively conceive them on the surface of objects so i think my conscious state right now represents this has the property of representing this identity quality but that that property of representing that identity quality cannot be conveyed in the language of um

quantitative physical science. So there's something about the experience that I know from the inside. So I don't think, yeah, I don't think we have to have the naive kind of, you know, all these qualities in my mind. A more nuanced view is that... the conscious state represents qualities out there, but still. It represents them by having the quality itself.

Well, the conscious state has the property of representing that pinkish quality out there in the world. Well, it could do that without having any quality itself though. Yeah, but it really does have the property of representing that. Oh, so does the word pink. Yeah.

I guess I would say that's derivative from the mental representation of that quality. Yes, but it's not clear that the mental representation has any mysterious... inexplicable, ineffable thing to it any more than the word pink does even if the mental representation is the is the foundational one i mean you need to say there's something about that mental quality that's that's peculiar in order for it's just a bit unclear whether whether the word pink

what the word public language word pink refers to. But if we're talking about Edenic pink, this, you know, colors as we naively tend to be, I don't think you can understand that property unless you have a pink experience. You can't convey in the quantitative language of neuroscience to a colorblind scientist what... So the experience itself has to have the quality? No, no, it has the property of representing that Edenic quality.

I'm quite happy to talk about mental states representing qualities. I think that doesn't seem particularly... I mean, you have a naturalistic theory of representation. But I don't think we could give a, for these reasons, I don't think we could give a reductive account of my brain representing Edenic pinkness, because to do that...

We'd have to characterize the property of representing Edenic pinkness. And you can't do that. You can't convey Edenic. You'd have to convey Edenic pinkness in the language of... quantitative language? I could have a property of representing all sorts of things like games and family likenesses.

all sorts of things that I can't define, that I can just pick them out. I can recognise it's one of those birds again. What sort of, I don't know, it's just one of those sorts. I just recognise it. I can't give you any account of... what exactly you know i can't give you a specification of the thing i'm recognizing it's just i just can recognize it when i see it it's not peculiar that's not so you might have some

That's because we're getting rather technical here, aren't we? I suppose that you're thinking of cognition extending outside the head. But somehow the physical story has to determine. reductively account, if materialism is true, has to reductively account for our representation of these Edenic properties, I don't think it can.

We are getting off topic here a bit, but does that mean that for a panpsychist that electrons have representational states because they're supposed to have this same thing? No, not necessarily. I think you could maybe have non-intentional, non-representational mental states. I just, I mean, I agree with you both. I think that it seems wrong to say that the pinkness is a quality of...

The experience, the pinkness is represented to be on the book. But I don't think that's the problem for that. That undermines the anti-reductionist argument. It just means we can't give a physical reduction of. the property of representing Edenic pinkness. I'm sorry, I've gone back into it. We are, we're getting away from it. Can we just, maybe just to try and bring it back a bit, isn't one...

Just on your view, really, there is no science, in a sense, there's no science of consciousness to do because we know the nature of consciousness itself in the most immediate and complete way we could ever know it. And so there's nothing else to know about the nature of pinkness. That's just revealed to us. All there is to do is...

to tell some sort of story about how that's related to the rest of the stuff that's around us. But there's no way we need to advance our knowledge of pinkness as we might advance our knowledge of dinosaurs by doing science. That's that bedrock. knowledge the data that's unrevisable data and that's where we so in a way you you're denying in a way that there's a science of consciousness

You turned to me. Sorry, I think you were addressing my question. He asked to fill it, but he was trying to link up with the debate about science. Yeah, let's try and bring it back to Matthew. So, yeah, the problem is we know this stuff from the inside. We know loads of stuff from the outside. And to me, the challenge is how to link these up. And that gets very challenging because...

is the overflow thesis true or not? Who the hell knows? And I mean, so just, I mean, just coming back to that matters to bring it, I mean, how do we solve this problem about, we need to, we need to say whether the overflow thesis is true or false to do the science, I think you're saying. How do we get around that problem? So it's just, to me, it's just a matter of inference to the best explanation. In that sense, I agree with Ned Block, although I disagree with him that

the inference we should draw is that there are overflowing contents. But I agree with the methodology. So can you roughly gesture at how you deny the overflow thesis by inference to the best explanation? Oh, okay. My denial will be a bit complicated, but let's just explain. Let's just explain how Ned would defend his view. What's really, really important is to recognize that you don't have to... It's kind of counterintuitive.

But you don't have to test a scientific hypothesis directly in order to gain support in favor of that hypothesis. So if you look at the history of science, that's how we gain support for the stuff like the theory of evolution. For a long time, it was very difficult to test the theory, but still it provided better explanations than the alternative creationist hypothesis, and therefore we ended up...

thinking that the theory is right. In the same way, there's a way in which we could at least maybe two ways in which we could gain support for the overflow hypothesis, even if we don't have data that directly indicates that there are overflowing contents, obviously, because people cannot report about them. The first way is to develop a theory and test all the aspects of the theory that are testable and see what that theory explains about the phenomena where we can test the theory.

At the end of the day, we discover that this theory is super, super good, that it explains all the phenomena it... answers all the questions we have about consciousness. It explains why consciousness disappears during dreamless sleep. It explains why we have the feeling that we see colors in the visual periphery. It explains why visual imagery doesn't feel exactly like consciousness.

conscious perception. It explains all this. It explains visual masking, motion-induced blindness, all the stuff that I was mentioning at the beginning. It explains all that, all those phenomena related to consciousness. But there's just one thing. also makes the hypothesis that there are overflowing contents, namely contents that feel like something, but you're not directly attending to those contents.

then I think that a reasonable thing to do is to accept the theory, even in the cases in which we haven't tested the theory, and even in the cases where we cannot... get direct empirical evidence just because it's so good in all these other cases. The second method is just to produce new empirical phenomena that are very difficult to explain. but just without any overflowing contents. That's the strategy that NetBlock uses when he talks about the Sperling experiment.

In the Sperling experiment, you have three rows of four letters that are flashed for something like 250 milliseconds to 300 milliseconds or 400 milliseconds, so it's really fast. and you ask subjects to report as many of those letters as possible. And it turns out that subjects can report about four or five letters. But then you do something different.

Just after you flash the letters, you present an auditory cue that indicates whether you should report the highest row of letter or the lowest row of letter or the medium row. of letters. And it turns out that if you do that, subjects can report any given row of letters, which means that they saw all the letters. So now you have two explanations of the phenomenon.

Actually, you have a third explanation. You have two main explanations of the phenomenon. The first explanation is that they saw all the letters consciously, and then they could access only four of those letters. And the second explanation is that they saw some letters consciously, the other letters they saw unconsciously. And then you add the fact that, at least it seems...

Let's assume that's the case, that people, when they see those rows of letters, they report having the vivid feeling that they saw all the letters consciously. Now you have those two hypotheses. And it seems like in the second hypothesis, you have to... appeal to a lot of unconscious stuff going on. You have to appeal to what they call summary statistics representations on top of those unconscious representations.

then you have to make a lot of hypotheses that's quite inelegant. In the other case, you just have the hypothesis that, okay, that's just conscious perception, they just forgot. because they could not transfer the representations in working memory. That's why they don't report about it, but you can show that they are conscious of all the letters. That doesn't test the...

overflow view directly, but this seems like a good hypothesis. It explains this phenomenon quite well. The third explanation, which is the explanation that I prefer, is that when you actually don't encode all the letters... what happens is that there's a kind of post-dictive effect. These are fantastic, beautiful effects, psychological effects, in which an event, the way in which you perceive an event at time t...

depends on what's going to happen at time t plus 1. So in that case, what really happens is that the auditory queue, even if it comes after the... the letters are presented, the auditory queue actually changes what you experience. And so that's how we can explain the effect.

So you have these different hypotheses, and ultimately, you don't need a direct test. That's the key point here. So, Mr. Skeptic, Keith, why isn't that an answer? You're skeptical of, you know, science of phenomenal consciousness. Doesn't this kind of inference the best explanation? Doesn't that do the job?

I'm still not clear exactly what it is we're trying to explain what it is that makes a perception conscious as opposed to being unconscious but having a lot of the same effects as a conscious one. I'm not sure what sort of extra source consciousness is adding here. I mean, we can tell a story about the processing of information in the brain and the kind of effects that each... stage of the processing has. And we could tell that story in great detail using this kind of

mentioning things like, you know, predictive effects and so on. We could tell this story, but then I don't see how it helps to try to find a place in that story where, you know. We put the label conscious on the events and that story. What we need to know is what's happening, what bits of information are being processed, at which times, what effects they're having, and so on. And we can map all that. Adding that one and that event was conscious.

where that is giving us any extra information in addition to information about what kind of information was being encoded and what kind of effects it has. That doesn't seem to me to be adding anything extra to the story. I think we could probably do all the psychology without mentioning the word consciousness at all. What am I missing out on that story, Matthias, if that's what I...

So we can certainly tell why one of the things we might explain why people want to say that some of it is conscious, that will be interesting, maybe. But what is actually missed out of the psychological story? So it seems to me that, first of all, I just... I don't want to define consciousness in a way that doesn't imply any effect. I think that consciousness has effects. I just think that we don't know what those effects are at the moment. But we are going to study it, and then we'll find out.

what those psychological effects are. I think we should maintain this term in part because it's an important term for people. It makes a difference to... People, for example, in, I don't know, let's say, take one example, animal welfare, whether fish have just nociception. and react as if they have pain-like experiences, or whether there's a conscious sensation of pain. Because we know from the human case that there's a difference between nociception and the experience of pain.

And that's actually how pain is defined by the association for the scientific study of pain. They define it as a feeling, whereas nociception is defined entirely functionally. So, I think it's important to keep the term for many, many applications, most notably therapeutic. applications. For example, there's some very nice work by Hakuan Law on fear conditioning that's done completely unconsciously. So that's fantastic work.

And here, the big difference, why it's important is because it's done unconsciously. People don't see the conditioning stimulus. And if they saw it consciously, it would... it wouldn't make any sense. So the reason why these therapies and all this stuff is important is precisely because you're doing something without consciousness.

And what I mean by without consciousness here, you can define it operationally as it's just like there's nothing for people undergoing these therapies, representing the stimulus. feels like not representing it. It just feels exactly the same. You can define it operationally like that. Yeah, I mean, I'm inclined to agree with you that...

For everyday purposes, the notion is very useful and it makes distinctions that we find useful in talking about ourselves and talking about other people and interacting, but perhaps not as useful as we think. I'm just not convinced that it's doing any work. in science that we could simply, we could replace it with functional notions, notions of states that have certain specific kinds of effects where we try to specify what those effects are and say, you know, does this...

Does this state have this kind of effect, that kind of effect? And we could capture everything that we need to capture from the point of view of psychology and cognitive neuroscience in those terms. And I'm inclined to think that consciousness is just a folk. term that is kind of useful to us like meaning maybe you know how do we pin down meaning in a scientific it's useful and belief and this sort of thing it's it's useful and we shouldn't just throw it out but

We shouldn't try to science it too much because it can't bear the weight and it just crumbles when we try to put the weight on it. I think that's Liz's view as well. I don't want to put words in her mouth. Can I just ask you, Matthias, there's a very simple question that people might be asking. Do you think that there is anything about consciousness, say the consciousness of another creature?

an octopus or a closer creature like a dog or a cow that we can't know, that we can't study scientifically? Is there anything that is sort of systematically beyond our knowledge? about consciousness in other creatures or other people? Who did it all? I'm not so sure. I'm not sure. Ultimately, it will depend whether we can study consciousness in non-human animals. So we'll see. You are optimistic about having a science of consciousness. So that would suggest that you think the answer should be...

you would at least hope that the answer was yes. You would be working on the assumption that it was yes. Right. What we can do notably is study qualities in terms of relation. in relation with discrimination. For example, if there are two stimuli that you cannot discriminate between those two stimuli, I would tend to think that those stimuli feel the same to you. So that's something that I can know even in octopuses.

Ultimately, we won't be able to know. If you're extremely good at discriminating those two stimuli and you're conscious, it's probably because those two stimuli feel quite different. Again, I can imagine... a skeptical scenario in which that's not the case, in which conscious experiences are completely separated from any capacity to discriminate, but... the probability that this skeptical scenario obtains is probably lower than the probability of the non-skeptical scenario.

For all practical purposes, we should prefer the non-skeptical scenario for now. I think we should reconceptualize consciousness in a way that those skeptical scenarios become incoherent. discriminations and the reactions. There's no extra source there. But anyway, that's me, so you know what I think about it. I think the skeptical scenarios are quite irrelevant for the scientific study of consciousness.

It's quite interesting how we conceptualize consciousness. Again, it's a matter of accepting that the foundations... of the scientific study of consciousness are not going to be immune to those kind of skeptical scenarios where you could imagine, oh, well, but what if people just say that they have experiences and in fact they have absolutely...

no experiences? Or what if, you know, you can put a lot of what if those skeptical scenarios are not, they can be relevant because they can drive scientific research in a sense, but I don't think that's the kind of scenarios that we should... We shouldn't put too much weight on those scenarios, in part because they prevent the scientific study of consciousness.

of making further progress. Instead, what we should do is accept that the theoretical foundations are not entirely reliable and try to build on those foundations.

We should probably move on in a minute. There's a couple more things we want to talk about. But I just want to ask one more thing. It just occurred to me while you're talking. So, you know, you've said it's like it's kind of like dinosaurs, you know, where we can't observe them. or um but why is i mean that we can what what what's unique

Why we have to do that in the dinosaur case is dinosaurs are in the past and you can't observe the past. And then there are troubles, you know, studying the subatomic world because you need very high energy levels or something or studying things far away. Why is why is it especially tricky in the consciousness case? You know, why do we have to have these sort of indirect inference? The best explanation if I mean, for example, I understand Daniel Dennett, if he just says.

What we're trying to explain is publicly observable. data you know report or you know what's going functional process inside the head i think keith would say the same you know and then there is no special problem we don't have to have these inferences to the best explanation um uh but so why

in the case of dinosaurs it's because it's in the past and i would say in the case of consciousness it's because our fundamental way of knowing about it is in this very special way and i you know i know about my own mind in that way and i wonder does an octopus have experience in the way i do uh and you can see there why it's tricky to do and we have to sort of think of these indirect why do you think it is tricky it seems to me if you don't have my view

that there's this special way of knowing about the phenomenon, it shouldn't be tricky. You're just explaining publicly observable data. It should be just easy problems, as Chalmers would say. Why is it tricky? It's tricky because of the problem of coordination.

that I was mentioning at the beginning. It's extremely difficult to find reliable indicators of some phenomenon that you want to study if you don't know anything about that phenomenon. But again, that's a problem that we had with temperature. That's a problem that we had with time. How do you measure time? That's a huge, huge challenge. At least it was a huge challenge. Now, we do it all the time. tough problem to solve, and obviously we shouldn't expect to solve it in a snap.

Okay, that's a really interesting analysis. I won't come back. That's very clear to me now. It's a very interesting and kind of unique, I think, analysis of the problem and how we address it. So just getting a little bit more specific now, we've talked in quite general terms. So I want to ask you about, you know,

the state of science of consciousness as it is now. And we've referenced the two interesting things you've written. Well, there is this... co-authored paper um opportunities and challenges for for maturing signs of consciousness for how many people was wrote this 27 or something was it i know 46 or something 46 sorry i don't know where i got it maybe so yeah It's a short article. Again, people can look at it. It's open access.

But it's also, I guess, the spirit of a public letter in a way. It's sort of raising certain concerns about the state of the science of consciousness and making some proposals about how to move forward. So that's one thing. And also this... paper you've written with haquan lao the philosopher and neuroscientist um um what's that called now a sociological take on the meta problem so this is the special issue of

of the Journal of Consciousness Studies on David Chalmers paper on the meta problem of consciousness. So this is the challenge of explaining why there is a hard problem of consciousness. and Chalmers sees that as an interesting interdisciplinary task that whatever you think about consciousness um you can engage on but but you and haquan give a sort of sociological account of that in terms of certain

corrupting influences in the science of consciousness. Well, anyway, maybe we could take both of those in turn. So just with the co-written paper.

With the 46 people, I mean, what could you tell us? I think people will be interested about the concerns that are being raised there and the proposals. Sure. So the main... The main goal of that paper was to attract attention to the fact, first of all, that serious people work on consciousness, that there's a lot of serious work being done, most notably at the ASSC, which is... the Association for the Scientific Study of Consciousness.

One of the main problems, especially in the US, it's not so true in Europe, it's extremely difficult for scientists to get public funding. in order to work on consciousness, even though I think this work can be quite rigorous. So we mostly wanted to promote this rigorous work that's being done. in order to attract funding, or at least just put the idea out there. that the field is at a point now that some public funding would be very much welcome to make progress.

So why do you think it's having these difficulties? So in your... paper with haquan you have this vicious cycle of corruption not corruption's too strong sorry that's not but starting off with consciousness science being a taboo and this leads to certain gurus uh with revolutionary views getting in on the scene and then that leads to back scratching that like you know you give me some

Right. And then that leads to break down a peer review and then back to more taboo. And so, yeah, gone. And you think this kind of explains why we have these troubles of consciousness anyway. And maybe why funding is a challenge.

The basic idea is that we don't present this as a sufficient condition for the meta problem of consciousness. We just think that It's a necessary condition, probably, that if science doesn't make any progress, then people will tend to think that science is unexplainable, that consciousness is unexplainable.

So it's a necessary condition. It's not the only reason why we think that we have a problem with consciousness. But perhaps if we made more progress, people would not think that consciousness is so... mysterious now irrespective of whether you think that's a good hypothesis you can take our hypothesis about this kind of visual circle seriously in order to explain why we haven't made

as much progress in the scientific study of consciousness compared to other fields that were revived, let's say, at about the same time, like social neuroscience, for example, or neuroeconomics. You see a lot of public funding for these two fields, and consciousness doesn't have as much. Although, you know, from kind of purely, again, from a... the therapeutic point of view. From a practical point of view, it seems like consciousness seems quite important. So the vicious circle is the following.

Because of behaviorism, mostly consciousness was considered a taboo for a pretty long time. It was very difficult to study it. And in the 70s, following the cognitive revolution, people like, I don't know, Tim Chalice, for example, and so on, started studying consciousness again as a scientific construct. And then you had the work by Larry Westkrantz, for example, on blindsight, the work of Brenda Milner, distinguishing between two kinds of memory, procedural memory, which is unconscious and...

which can be acquired unconsciously and episodic memory. And consciousness was out there at the time. And one problem is that the kind of... air of the idea that consciousness is mysterious attracts people who have big reputations. and who don't necessarily know the entire literature and the work in psychophysics that was done before them, and want to solve this problem in order to get a second Nobel Prize or something.

And so these are what we call gurus. That's not an insult. That's a term that comes from Dan Sperber. In general, a guru in this sense is someone with a very high level of authority. The example that we take is John Eccles. So John Eccles was obviously a very important scientist. He did most of his work on synaptic transmission, how neurons communicate.

and then went on to work on consciousness. And his work on consciousness, I think we can all agree that it's quite clearly pseudoscientific. That's one of those cases where... Well, I mean, maybe not pseudoscientific, but not serious scientific work. And then, you know, one problem with these gurus... is that with people with a lot of unearned authority, at least in this domain, is that they attract a lot of attention.

And because they attract a lot of attention, they attract the media, and funders will just take John Eccles as, at least at the time, as the kind of... kind of representative of the work that's being done in consciousness research. And if people do that, then it's reasonable to think if you're a founder that...

you're not going to give funding to consciousness research because it's not serious. And so that's the kind of misrepresentation of consciousness research that's misrepresented by these... these figures with a lot of authority, but a kind of authority that's not earned within the field of consciousness research, that's earned in other areas. That's the general idea of the paper.

Yeah, so it's like a pop star coming to your own politics or something. They often make more sense than politicians. Yeah, so it's not, that's interesting you say it's not a sufficient condition. I was kind of thinking this doesn't seem plausible to me, an explanation of, you know, I mean, given these intuitions go back to Descartes or whatever, to the explanatory gap and so on.

But yes, you're not... Well, it could be... It's probably... Then it depends on how much weight you put on this. I certainly think that, again, if... we had a theory of consciousness that could explain a lot of consciousness-related phenomena.

which doesn't require solving the heart problem. For example, if you have a theory of consciousness that explains binocular rivalry, motion-induced blindness, visual masking, change blindness, all that stuff, and that also explains why consciousness disappears during sleep.

the difference between visual imagery and perception and so on. You have a theory of consciousness that explains all that. I would tend to think that people will start just identifying consciousness with whatever is the main construct. of that theory. And maybe people's intuitions change over time. That's something we see. For example, there's a fantastic a fantastic article by Direct and Deux in 2007, in which they showed that over time following Newton's theory, people started just accepting

action at distance. They just didn't see the problem anymore. It's just a theory, just like people... might accept today. Copenhagen's interpretation of quantum mechanics. Maybe some people just don't really see the problem. Why do we want your theory to be deterministic? And stuff like that. So people who grew up with the new theory don't consider it less or more unintelligible than any other theory. So that could happen with consciousness. I'm not confident enough.

in my intuitions that I think that those intuitions will just continue forever, that people will always have those intuitions, I don't know, and I don't think we can know that. So our conception of consciousness could...

could be reshaped quite radically by by science you think yeah it happens all the time take um take for example kant uh he believed or at least that's what I'm told, I'm not a specialist, but I'm told he considered Euclidean geometry as necessary. We found that that's wrong, but he was evidently... unable to think of a kind of geometry that would be different. He was completely wrong. It was inconceivable to him. So obviously here are scientific theories made.

changed our intuitions. The thing is that I think we are extremely bad at knowing first of all in which direction the sense will progress. If you look at the record of scientists and philosophers trying to predict the future of science, they've been more or less always wrong. But also... Just before the creation of quantum mechanics, some very important scientists said, that's the end of science.

We just have to fix some details here and there, and that's the end of science. But obviously, they were completely wrong. These people were very intelligent, probably more intelligent than... than me, so if I try to predict the future of consciousness science, I will probably be wrong, so I cannot be so pessimistic. And also, the second thing I forgot. Sorry, go ahead.

Sorry, I caused you to lose it there. I just wanted to say, I think that consciousness is peculiar in that people do have, that people feel that they... I feel a certainty about consciousness and an authority about it, that they just don't feel about any other aspect. People may have ideas about subatomic physics.

the nature of time or whatever but they don't feel most of them and don't feel that they have any authority to really pronounce on it and to say how it must be that can override science but they do feel that way with regard to consciousness and i think that itself is a very interesting fact about consciousness. Whatever it is, it's whatever we're talking about here. It's something that produces this sense of absolute

conviction that one knows it and is intimately acquainted with. And I think that's a fascinating fact, but I still see that as something that is open for explanation. Yeah, I think it's an important fact. There's some amount of authority, although, you know, when I teach consciousness and I present... Schwitz-Gebel's example in which you admit that you could be wrong, at least. Most of my students, it seems to me,

have no problem admitting that they could be wrong about the quality of their visual imagery, for instance. And think about just the fact that that we have this question about the richness of consciousness. Shouldn't it be obvious? Because we know consciousness from the first person perspective. If we were so sure about our conscious experiences, we wouldn't have that debate. People might not be linked to that word richness. the debate about whether all of these things like the

feeling my clothes is part of my experience. But I mean, I guess I've actually just this morning writing a reply to the physicist Carlo Rovelli, who's written a response piece to my book, Galileo. There is a special consciousness studies revised and so on.

Sorry, you froze for a second. Oh, yes. Am I back? Yeah, you're back. So he's pushing this kind of line that... our intuitions about consciousness and the hard problem can be revised so i guess i mean i guess my point is i'm totally on board with that in all other cases Because in all of the cases, we're completely beholden to public observation experiments. So, you know, who cares my folk notion of time or free will or whatever, because.

We should mold those to account for the best theory of publicly observable data. But I suppose, you know, again, I would say in the case of consciousness, our fundamental way of understanding the phenomena. is through our immediate awareness of it, of our feelings and experiences and the qualitative character they involve. And I don't think we're totally infallible. I'm totally up with Eric Schvitz-Gable's thoughts that we make mistakes, but I don't think I could be wrong. I mean...

What it is to experience that my experience, that my colour experience involves qualities. You know, I don't think I could be wrong about that. And so... That's why I think there is something that there is data that's not publicly observable data. And so we haven't just got the publicly observable data.

to mold our concepts. But I mean, linking to that, I suppose, look, so I mean, I've had conversation, email conversations with Hakwan La about this. And, you know, I appreciate the concern about needing funding for science and, you know, showing you're scientifically respectable and so on. And I appreciate that concern. But at the same time, from my perspective, these are philosophically contentious issues. From my perspective, you can't separate.

the scientific experimental questions from the philosophical ones. You know, so just to take an example, you know, if just to take an example, if you're a David Chalmers style duelist. So if you think, you know, conscious states are non-physical and Chalmers thinks that they're brought into being by psycho, fundamental psychophysical laws of nature that link up, you know.

physical processes with consciousness. And so you're looking for those psychophysical laws. Now, that's not my view, but if you had that view, for example, I think it's pretty plausible that you should be looking for non-vague correlates of consciousness, because it's pretty plausible that fundamental laws of nature are not vague. And by vague, I mean in this technical philosophical sense of not admitting of borderline cases.

So, you know, there's whether someone is tall or not, the people who are definitely tall or people who are definitely not tall. But there are these fuzzy borderline cases. I'm probably neither tall or not tall. With baldness as well, I'm probably in the borderline of sort of being bald and not bald. But if we're talking about fundamental laws of nature, I guess a lot of philosophers think this kind of fuzzy vagueness is to do with language. It's a vagueness in the word tall.

you know, that it sort of doesn't totally define its referent. But if we're talking about fundamental laws of nature, they ought to be utterly precise. And so if you're a dualist, I would have thought you'd be looking for... correlates of consciousness that are utterly precise now you might think global workspace theory looks like that's going to end up being a vague correlate of consciousness it's a you know

So it tells us that we get consciousness when we get information that's broadcast, broadly broadcast throughout the brain. I mean, how broadly broadcast or not is presumably going to be some fuzziness there. Whereas integrated information theory, for example, gives you totally precise cut up point when there's more integrated information in the whole and the parts.

You've got that cut up point. So you might think as a property due list, well, that gives me reason to go this way rather than that way. So I appreciate from one philosophical perspective, you know, yours, perhaps you can hive off the experimental. situation but from my philosophical perspective where there is substantive facts we know about the nature of consciousness from the inside and that might have a bearing that might have a constraint on the experimental theory you just can't do that so

Yeah, I mean, and given this is so philosophically contentious, I don't know. I don't know what to do. So that was very long winded. But take that. So let's imagine that that you have a theory that. predicts and makes the prediction that consciousness is completely determinate. I have the intuition that consciousness has to be determinate in all cases, that something is either conscious or not, there's no...

borderline case of consciousness in that sense. I have this intuition as well. I would tend to think that if a theory respects that intuition, it's better. but it's just one factor among many, many others. If that theory is incompatible with some things that I know... So, for example, if that theory is completely incompatible with the existence of binocular rivalry, for example.

It seems to me that I will not hold on to that theory just because it satisfies this intuition. After all, either maybe my intuition is wrong or... there's another theory that's going to make the same prediction that's going to be better than the one I currently have. So, of course, I agree with you that there are some intuitions that we should respect.

these are maybe philosophical intuitions. In that case, I'm not even sure it's a philosophical intuition. It mostly seems to be a kind of common sense, pre-theoretical intuition. I'm not sure you need to do a lot of philosophy to get the idea that consciousness has to be determinate in all cases. I would tend to think it's a good thing if your theory respects those intuitions. But at the same time, I think there are so many other factors.

that could lead you to reject that intuition that it's just one small piece of evidence. I'll just say one more thing and then I'll step you up just briefly. Be careful with this argument Philip, you know, it led you to universalism in the past. Be careful, I don't want to combine it with panpsychism. You don't want to go back there, do you?

but the point is my point is if this is controversial and if you are a property duelist maybe you'll put more weight in that intuition than if you're not uh so we can't all just say oh let's all just keep the science clean of the you know you can't separate them you can't separate the experimental task from the so i mean i think we just need to persuade the

scientific community to take philosophy seriously? Because rather than just trying to pretend you can hive off consciousness science from the philosophy. Well, I think they do take it seriously. At least if you go to the Association for the Scientific Study of Consciousness, the annual meeting, there are a lot of philosophers.

who do good philosophy on consciousness. And it's not the philosophers go to the philosophy talks and scientists go to the scientific talks. It's really interdisciplinary. So I wouldn't think that. the scientists don't like philosophy or don't care about it. I think they do care about it.

Matthias for a long time but I hope he can spare us just a little more time because before he goes I want to get him to talk about theories of consciousness and you know come on tell us you know rate the different theories of consciousness that are on the market at the moment and tell us which ones are, from your perspective as a philosopher of science, someone who knows the science.

of consciousness. Where's your money on these? There's lots of theories around, there's integrated information theory, there's global workspace theory. Come on, give us the lowdown, give us the betting odds on these. Just before Matthias does that, if people would be like at this point to ask questions and then I'll gather them. And after we've discussed this very specific question, then we can turn to those questions. Matthias.

Okay, so let's take the main theories. Obviously, there are many, many theories. So first of all, it's, I think, important to distinguish between two kinds of theories. There's what I call consciousness studies. with theories that are not cognitive neuroscientific. And then there's the cognitive neuroscience of consciousness, which is really...

the set of theories that think that consciousness has something to do fundamentally with the brain creating mental representations. That's the basic idea. consciousness studies, you have all the quantum mechanics studies of consciousness, for instance. I don't know anything about that, so I never talk about quantum mechanics. Let's take the cognitive neuroscientific study of theories of consciousness. Again, a theory of consciousness is a set of explanatory hypothesis.

that aims to explain a coherent set of explanatory hypotheses that aims to explain as many consciousness-related phenomena as possible, as well as possible. And so I think... currently the theory that's most supported by neuroscientists and psychologists working on consciousness. There was a recent article by Julian Franken and colleagues.

indicating that I think global workspace was the theory that scientists found most promising. The second theory that scientists found most promising was probably higher order theory, although that's a group of theories. not a single theory, which is in agreement with the results that we had previously with HAC1 in a study that we did on what consciousness scientists believe. I must say that it's probably a biased population because we did this mostly with people who go to ASSC.

Probably the third one was IIT, or maybe it was the first one among scientists. So IIT is Integrated Information Theory. Personally, I tend to think that higher-order theories are more promising. I tend to think that there's a kind of intuitive appeal to the idea that what it is for a state to be conscious. is for you to be aware in some sense that you're in that state. So that's what I think David Rosenthal called the transitivity principle. Is it?

No. What's the name of this principle? Yeah, I think so. Is it? Yeah, I don't remember all of a sudden. That's the X principle or the higher order principle. So if you believe that, you'll tend to think that, or if you think that there's a good probability that it's true, you'll tend to prefer higher-order theories. And then the question is, you noted that I said... a state is conscious if I'm aware of being in that state. The question then is, what is aware of?

How do we define it? And then you have multiple ways of being aware of something. I can be aware of something because I think about it. I can be aware of something because I perceive it. and so on. And so you have multiple ways of developing higher-order theories. So these are the theories that I think are most promising.

And then, of course, I really like the global workspace theory. I think it's a good theory. The main problem with global workspace theory, I think, is what we call the... performance capacity confound. So it's the problem that basically most of the support for the global workspace theory comes from studies in which you compare conscious conditions versus non-conscious conditions.

And obviously, consciousness is not the only difference between the two. One big difference between the two is that when it's not conscious, you cannot do much with the stimulus. For example, if we measure your performance in discriminating the stimulus, your performance is going to be very low, like, for example, 55% performance. chance level is 50%. In the conscious condition, you're going to have 75% performance, let's say.

The idea is that the kind of brain activity that corresponds to global workspace activation actually correlates with this increase in performance between the... the unconscious and the conscious case, not to consciousness itself. And what we can try to do is match performance between the unconscious and the conscious condition. And when we do that... we can see that the kind of activity that's elicited by conscious stimuli is not as widespread as global workspace theorists would predict.

So yeah, that's the kind of, I really like the global workspace theory, but I think there's evidence against it as well.

So personal question, which studies do you think that have tried to do control for the extent of the effect? The performance capacity confirmed, so the main study... was Lau and Passingham in 2006. So it turns out that visual masking, at least some kinds of visual masking, metacontrast masking, So, you present the stimulus, and then you put a mask around the stimulus, and the mask is going to mask to prevent you from seeing the stimulus consciously. It turns out that this effect...

is described by a u-curve. So if you present the stimulus exactly at the same time as the mask, you're going to see the stimulus. But then you can wait some time. you can put some time between the stimulus and the mask, and then you're going to have no visibility at all, and then visibility is going to increase. So what they found is that performance for

level A of visibility, you can get the same level of visibility at two points of the curve with different... No, sorry. You can get the same level of performance at two points of these... that are not equal in visibility. So you can get different performance, equal visibility. And you can do something different. You can do equal performance, different.

different visibility. So when you do that, you get equal performance, different visibility. What they found is that prefrontal cortex still correlates, increased activity in prefrontal cortex still correlates with consciousness. or at least the increase in visibility, but the activity is much less widespread.

than what global workspace theorists should predict. You have something like that that was done, but slightly less, it's not as rigorous, and it was also the very start of neuroimaging by Larry Wisecrans. and colleagues, I think it's Saraje and colleagues, 1997, in which they did the same thing with blindsight patient GY. So in blindsight, patients are able to but they at least report that they don't see anything.

As far as they are concerned, they are blind, at least in one side of the visual field. What you can do is match performance between the blind visual field and the sighted visual field. visual field in which they report that they see the stuff. And when you do that, you find the same thing, that the main difference between the two conditions when you match for performance is increased activity in dorsolateral prefrontal cortex, so in the front of your head.

But again, it's not as reliable as the other one. So it's a big challenge. It's very difficult to do to match performance with different levels of visibility. It's extremely difficult to do. What do you think about the Templeton-funded adversarial collaboration between IIT and global workspace theory? We were talking to Christophe Koch about last month.

I think the adversarial collaborations are a good project in general. I'm actually part of the other one on first-order theories versus high-order theories. There's also an adversarial collaboration on that. The collaboration on IIT versus global workspace, I don't have so much to say about it, except that I tend to think that some of the predictions made by IIT, if I read correctly, have already been falsified by some of the studies, notably that Jean-Rémy King did a while ago.

Yeah, so it's a bit puzzling at times. But otherwise, I don't have much to say about it. I think it's a good project. It's a good project, especially to the extent that it really forces people to make predictions. Maybe it forces them to think about their theories in different ways, which is always good.

But this is a point, I can't remember where you make this, but you make it in one of your papers, that the trouble is that when you get a, when you, a theorem makes a prediction that seems to be... this confirmed you can always seem to uh you know particularly with consciousness that you can you can always do a bit of

revision to your theory and simply say that well you know uh that response wasn't produced by god wasn't was it consciously produced or something like this it's very easy to insulate theories of consciousness from empirical refutation i mean iit has some very strange predictions about sort of things that would be conscious, including just a set of simple switches that are connected up in the right way. And you might say, well, that's a disproof of it because that's not conscious.

Well, they can say, well, yes, it is conscious. And who can disprove it? Because we have these strong intuitions about the privacy of consciousness, you can insulate practically any theory from refutation. And that does seem to be a serious worry. the consciousness science? Yeah, tend to think it's a serious worry, although I worry a bit less about it than I used to. Mostly because refuting a theory is not

Just not how science works. It never works like that. You will never get, almost never get a moment where they just say, oh, my theory makes this prediction. Don't tell it's wrong. Ah, forget it. I just spent 20 years of my life on this stuff. What happens instead is that your theory attracts less and less people and ultimately dies down.

So it tends to work, except in some cases. In some cases, it doesn't work like that. We really have people who change their mind radically. But most of the time it's... It's just a slow decay of the theory, which also makes sense. You don't want to just reject your theory too soon. To some extent, having some degree of confirmation bias is rational.

because you don't want to just reject your theory at the first difficulty that you encounter. Instead, you're going to revise your theory, and some degree of ad hoc hypothesis can be fine. For example, that's how we discovered Neptune. So Neptune was first an ad hoc hypothesis, and now we are glad that they decided to make this ad hoc hypothesis instead of rejecting the Newtonian theory, because otherwise we wouldn't know about Neptune.

Or at least by now, I hope we would have known, but that's how progress was created. But I do think that even if you cannot just reject a theory, I do think that...

making predictions that turn out to be wrong will force you to revise your theory. And perhaps by insulating your theory from empirical evidence more and more, your theory will be more and more in conflict with some intuitions that we have, some other phenomena that we've discovered, and then the problem if you do that is that your theory is going to make...

more and more extraordinary claims such as for example the claim that's made by iit that um an inactive inactive set of logic gates could not only be but infinitely more conscious than the human mind. That's an extraordinary claim. So in order to defend it, you need extraordinary evidence. or evidence that you would tend to believe at a higher degree of probability than the probability that logic gates are conscious in your...

in your mind. And the problem is that the evidence just isn't there in the case of IIT. I think that's the main problem. The main problem is not directly that it makes extraordinary claims. The problem is that if you make those claims in order to convince people, you need extraordinary evidence, and there's no extraordinary evidence in the case of IIT. The main evidence that they have...

consists in comparing unconscious states, namely comma, for example, versus conscious states like wakefulness. But of course, there's a huge confounding factor. or at least a number of huge confounding factors, which is that consciousness is not the only difference between comma and... and wakefulness. So for example, one difference is that when you're in a coma, you cannot do math, for example. You cannot think about your grandmother. So why think that phi correlates with

with consciousness instead of the capacity to think about your grandmother. And so as long as you did not rule out all those other confounds, as far as I'm concerned, you haven't really provided any good reason to think about your theories.

Correct. There was also the problem that, yeah, go ahead, sorry. I mean, you're saying there's no extraordinary advice, but IIT justifies itself. I'm going to... Try and resist going off, you know, partly in partly empirical considerations, but also partly on philosophical considerations. They claim there are these five axioms that are known through introspection and then infer five physical correlates. I mean, I'm not saying I'm totally on board.

that but one might be totally persuaded by that philosophical argument about what you know about consciousness from the inside and take that to be extraordinary evidence um that that overweighs um outweighs empirical evidence, certain kinds of empirical evidence. I mean, where I think it gets more convincing for me is the idea that consciousness is not vague. If you think that's a sort of solid fact, you can know about consciousness from the inside.

not just an intuition. It's something you can know about consciousness. Especially about the inside, Philip. There's a brain in there. There isn't any magic in there. should we should we bring in some questions before we just just one thing i think i think you're right to um give some weight to those intuitions maybe i don't think the axioms are

I think the axioms can be debated. There was a good article by Tim Bain about this. But let's say that you accept the axioms. Then the question is whether you think that your credence, your level... that your level of belief that these axioms are true is higher than your intuition that logic gates are not conscious. In my case, given that I'm not so confident in the axioms...

I would tend to reject the theory instead of rejecting my intuition. That seems equally strong that logic gates, maybe they're conscious, but certainly not as conscious as the human mind. So it's a case, again, in which you have to... put all these evidence on balance and see what you think is more plausible. Okay, brilliant, that's fair enough. So I guess what, should we have like... 10 minutes of of questions see how many we've got quite a lot of questions here lukash got him first uh

When will the problem of consciousness be solved? Will it be debated for another 150 years? And then he also said, does Matthias think there's an explanatory gap? So the problem will be solved in 2075, let's say. Thank you. Way too early. This is recorded now. This video will be around 2075. People can check it out. So I think that's on Axel Claremont's CV.

He had listed on, at least it was on his website at some point, that he had the year of his death in the future. He wrote that he would die just after, right after solving consciousness. Do you think there's an explanatory gap? That's a good question. I have no idea. I don't have very strong opinions about these issues. I tend to think there's an explanatory gap here. I would tend to agree with Janet Levine.

that you had the other day on phenomenal concepts, that even if consciousness is material, there's an explanatory gap, or at least there seems to be. We have different ways of thinking about consciousness. But at the same time, I see the appeal of the other view, the other materialist view that says that there's no explanatory gap. Ultimately, I think...

When I think about it in this way, I tend to think that our intuitions will just change and we'll just not see the explanatory gap anymore. After we have the right theory of consciousness. So these are the two materialist positions that David Chalmers called type A and type B. Exactly. Type A is Dennett. There just isn't an explanatory gap or at least maybe more Patricia Church and will one day overcome that kind of thinking. Whereas.

Janet Levin, who we have on type B, says, no, there is a kind of expansion because of the very different concepts, but it doesn't have any metaphysical implications. It's still a straightforward idea.

Yeah, I just tend to hesitate between those two versions. Right. Yeah, good to be open-minded. IcarusKK asked, what is science, in inverted commas, and what is the difference of your concept to earlier understandings of science? Thanks a lot. What was the second part of the question?

Well, I suppose as an implied, has the conception of science changed? Is your conception of science different to how other people think of science? Maybe your views don't necessarily hang on a particular conception of science. I think that's... maybe suggested by the corporation? No, they don't. So I have no definition of science. It's extremely difficult to find what's the difference between science and pseudoscience. So I tend to avoid...

making claims that some things are pseudoscientific for this reason, because I don't know what's the difference between the two. I certainly think though that although we are unable to find a definition, we are able to recognize it when we see it, at least most of the time, and that we have no problem identifying, at least in most cases, identifying what's scientific and what's not scientific.

So I would just define it by examples. So what's scientific is more like physics and more like, I don't know, sociology and history. when it's done right, then astrology, for example, I could give. Is mathematics science? Yeah, I tend to think so. At least they are funded by scientific...

It's a good way of defining it by the funding bodies. I'm getting confused by which of the questions and which of the chat. Maybe one more question here. This is a quote, I guess a statement of Theodore A. Hopp. And he's asking if you agree or disagree. Here is the statement. Modern neuroscience is capable of correlating brain states to conscious states. Empirical verification that the mind is produced by brain processes and is dependent on our evolutionary history.

Does that sound good to you? That sounds right, yeah. We're able to do that. Is that the claim, that we're able to do that? Yeah. I think we were able to do even more. Yep. Even more being what explain and not just correlate, but explain what it is. You're optimistic. Oh, yeah. I think we can explain consciousness-related phenomena, at least. For example, global workspace can explain why inattentional blink happens.

why inattentional blindness happens, why change blindness happens, all that. So we can answer all of our questions about consciousness with scientific theories. Maybe at the end of the day, Philippe will be right, and we'll stop there. We'll have answered all the questions we could have.

be able to explain why it disappears during dreamless sleep and all that stuff that will be unable to explain why there's something it's like rather than nothing. To me, it's good enough if we can answer all those other questions. And there are still.

interesting philosophical problems and methodological problems that we should solve. But you don't, I guess you don't think there's a hard problem really. So it's not like... it's not you say it's good enough but i mean if you don't think there's a kind of hard problem either because there isn't an explanatory gap or because it doesn't have real metaphysical implications it's just a conceptual artifact

So they are the only questions, really, if there's no hard problem. Is that right, do you think? We'll see. We'll see. Fair enough. If we are still completely unsatisfied. I don't know if we could all agree on this, but I think what Mattias is saying is that we shouldn't get bogged down in the hard problem. We should get working on the problems that we can work on, which we know how to deal with.

They're not easy, but we know we have procedures for them. They're uncovering, they're illuminating the subject for us. They're changing the way we think about it. They're broadening our understanding of it. They're taking us forward. just sitting in our armchairs thinking about the hard problem isn't taking it forward and then maybe re-evaluate the hard problem after we've done say 50 years work on the easy problems and then see where we are. I would tend to think that these are not

So there are not easy problems in the sense completely exactly in the same described by Dave Chalmers. Instead, they're explaining consciousness related phenomena. So there are still consciousness is in there.

It's just that those phenomena like binocular rivalry, visual masking, all that are phenomena that we've produced, that we've discovered, that are related with consciousness. Once we explain all that, we've explained with... answered all of the questions that you could possibly ask about consciousness that are except the hard problem, which is, is there something it's like rather than nothing?

When we've solved all those other problems, maybe when we look back at the hard problem, we'll think, well, what was it anyway? Right, exactly. That would be my prediction. Although, again, predictions are extremely... risky in this field. But it makes sense to work on what we deal with the tractor. Also, note that in general, I would tend to favor a theory that...

that is empirically adequate over a theory that's philosophically nice but not empirically adequate. So I think that empirical adequacy comes first and the capacity to... help us answer some really philosophical problems comes second. But that's your philosophical view. Well, the problem is that a theory that doesn't solve...

That's not empirically adequate. It's just wrong. So a wrong theory cannot be true. Well, as you say, it's philosophically wrong. As you said yourself, it's rarely adequate or not adequate. It's pressing in one. And you're saying your view is you weigh heavier the philosophical, sorry, the empirical intuitions of philosophical intuitions, but that is itself a philosophical.

viewpoint you can't demonstrate that empirically you know so we we have and I can demonstrate that people who took their intuitions seriously in the history of science always lost. So not always, but that's not approved, but that's inductive support. In the same way, I think one of the best arguments in favor of physicalism.

functionalism is precisely that it's an argument here's a counter example to that the people who tried to explain mathematics empirically didn't do very well so if like me you think consciousness is not a purely empirical phenomenon it's partly known about in this very different way then it's a little bit like i think you know

if you're a platonist about numbers you know you have if someone gives an empirical theory of numbers there's something going wrong there because we we know about numbers in in through mathematical intuition not not through just empirical so if like me you think of consciousness a bit like that then yeah i mean people you shouldn't trust your intuitions on purely empirical phenomenon which are just subject to a public observation experiment but if you don't think consciousness is like that

Well, modeling our epistemology of consciousness over the epistemology of mathematics seems like a dangerous ground because the epistemology of mathematics is as messy as... the epistemology of consciousness philosophy is nasty what you're essentially saying is that it's as mysterious as this other stuff I don't think we should duck the challenge and pretend it's sort of not that. I mean, obviously, could you ever imagine, you know, as a result of, you know,

The sort of work that has been talking about after another 50 years or 100 years of this. Could you ever imagine those intuitions of yours changing or becoming weaker?

I'm curious about their status, because I mean, I feel their force. I think everyone feels the force of these intuitions when they reflect in a certain way. But I find myself, you know, I kind of, yeah, that's how it's sort of presented to me. That's how it's, you know, at least when I tend to it in the right way. But then I think, yeah, but...

could be you know it could be misleading it could be a the way it seems to be presented to me my sense of what's real is a psychological construction of my brain my brain tells me certain things are real and then you know that's That's a psychological process and maybe it's misleading me.

How can you be sure you can sort of see through your own psychology, as it were, to the raw reality of consciousness? That's what puzzles me. We come back to starting points all the time, don't we? How do you start knowing about reality?

somewhere and then you just try and improve it well yeah one way is trusting your senses about to tell you about external reality I think there's also another way which is to some extent trusting our immediate awareness of the reality of feelings and experiences you just start by trusting something foundationally you just start somewhere and then keep exploring and see you know OK, fair enough. But I can't see. But I can't. I'd have to have reason to. But, you know, I have more credence in.

what I know about consciousness from the inside. And I can't see what empirical facts might emerge. You know, I'm open-minded, but it's hard to see what empirical facts might emerge that might... override that, just as it's hard to see what empirical facts might emerge to make you think, identify numbers with physical objects or something. It's... I think we'd better... Right, one final question. Dystopiana Banana says...

Matthias, the ban on the study of consciousness was not a global phenomenon. Cultural disinformation, I guess they're accusing you of cultural disinformation. It wasn't just, it wasn't a global phenomenon. It was maybe a Western phenomenon.

I don't know anything about this. I don't talk about it. Okay. Oh, sorry. We've been talking like almost an hour and a half. Thank you. That's really... enlightening discussion uh do you have any anything that didn't come up that you wanted to talk about or no all good thank you so much Are you happy? Any what? Brilliant. Well, Keith picked out that. Well, thank you very much for joining us, Matthias.

Very enjoyable discussion. That was really enlightening. I think I would encourage listeners, if they're interested, to read some of Matthias' work. I think most of it's available, at least in e-print form. I think someone's on your website, isn't it, Matthias? And it's fairly easy to find on the internet, at least in preprint form. And it's really worth reading. And particularly, I recommend the paper, the historical paper.

fascinating stuff Matthias has been through a lot of work in 19th century neuroscience tracing connections between ideas that are around now and that seem very contemporary and ideas that were already being explored then it's really interesting stuff and well look at all his work it's very uh accessible i think and very important so please go and search out his uh his work thank you thank you so we'll say goodbye to matthias and me and keith will have another little chat but

Thanks for joining us and hope to meet you at some point. Yeah, sure. Thank you. See you. Great. Okay. What do you think, Keith? We've got to stop hijacking the discussion with guests, you know. Yeah. It's hard to resist, doesn't it? No. Like Morecambe and Wise, we get a guest on. And, you know, we get this really famous guest like Glenda Jackson or something. And then we start fooling around. Like, you know.

We've got to stop it. I know. Yeah. Who is Ernie and who's Eric? I'm going to be Eric. I'm the straight guy. Which one's the straight guy?

Well, they're both, well, Ernie, I guess, but they're both comedians. They're both the comic, aren't they? Anyway, I was good. I enjoyed that. I thought, he's... still a young researcher and he's going to make a big impact well he is already making a big impact on this field and he's going to make a you know one of the dominant figures in the field i think over the coming decades he's doing really important stuff i think um and i think it's stuff that

Whatever philosophical position, I think it's stuff we need to take account of because the stuff they're investigating is real. We can build our theoretical constructions around consciousness, but it's got to be consistent with the sort of work that's been done there, because that's...

That's a sort of constraint, I guess, on any philosophical theorems. Absolutely, yeah. Yeah, I really enjoyed reading his work in preparation for this. And yeah, I mean, I think I learned more about his position as we talked. I think I really got a sense about halfway through of his analysis. When I asked him, why are things tricky? His answer to that.

Yeah, I guess I always think you either buy all this hard problem stuff and you think things are tricky, or you don't buy it and you don't think there's any special problem here. And it was interesting, I got a sense of a middle way option here that there are, we haven't found a way of calibrating data we have.

There is a kind of special problem, but it's a thing we've been over before in thinking about temperature and so on. And yeah, that's a really interesting middle position. He's more of a realist than I am, but he's... I think he thinks that the real thing here is it's...

I don't think he thinks it's simply a theoretical construction, but he thinks a heavy theoretical, I don't want to put words in his mouth, heavy theoretical constraints on our notion of consciousness. Whereas I guess you think that it's... it's kind of revealed to us in a way that's completely pre-theoretical.

Whereas I want to say that it's, you know, it's just a theoretical notion altogether. And, you know, it's something that, you know, it could be a theoretical posit we don't need. But yes, so yes, I think he does have a sort of which.

I guess, is a sensible position. I mean, what I liked also about the position was it was very cautious. It was like, well, OK, let's, you know, we start with intuition. OK, we start with intuition. There's some, you know, some it's a basis to begin with. And then we. we explore it and then we maybe revise it and perhaps reject it and so on. I think it's a very level-headed approach, you see.

You know, it's not as English who are the level-headed, sober-sense ones, it's the French, you see, that's why. It's English people like us who have these crazy extreme ideas. Yeah, yeah.

What went wrong? Yeah, well, I'm following Descartes, the original French writer. Yes, you see, the English are all followers of Descartes. What's happened? The French are followers of Leibniz, you see. We were the empiricists, weren't we? The evil European rationalists against the... British empiricists, but... Yes, but we trusted... That leads you to Barclay and idealism, if you're not careful there, you see. Ah, yeah.

Well, it was a really good night. You said an hour and a half, but I think we've been two and a half hours. I know, we've gone on a bit long. Well, we have to talk before you go about your... Your gig at Glastonbury, the philosophical equivalent of the gig in Glastonbury doing, you did a plenary session on a... the joint session of the mind in our alien session, which is the big annual UK philosophy conference. And that was Saturday morning, just gone.

It was online, yes. I wasn't there in person. Nobody was. But yes, it was done well. The online... sessions worked worked well i i kind of i'm coming around you know to online conferences that yeah i think they're working a lot better than they used to and there's a lot of nice nice features to them i think um yeah it went well eugene

I think it was supposed to be pro-panpsychism, but it actually turned out to be rather critical, although he is, I think, very attracted to panpsychism, but he thinks it meets a lot of problems.

And that was what he sketched in his paper, outlined a dead end for panpsychism. I don't think he thought that it actually showed panpsychism was false, just that... it might be true but then it took us to a place where we couldn't really make much progress on the problems it faced like the combination problem and then I

I had a more radical sort of attack on pan-psychism to follow up on that. So really we were beating up a bit on pan-psychism. I thought maybe it would have been better to have had somebody who was a bit more wholeheartedly committed to it. But yes, it was a nice session, some good discussion, I thought. And I got to show the photo of my wisdom tooth extraction again. Yeah, that's going around talker. Yeah.

Well, you were both very respectful about plans, like, you know, not saying, oh, this is just rubbish, you know. And as you say, I mean, my question to you was, is all this consistent with... You know, my argument, for example, that it's the theory that's most likely to be true. He said yes. And so, you know, in a way, he was quite happy with perhaps it might be the most, you know, the most.

theory that's most likely true but he just thinks we can't sort of cognitive dead end as he put it we can't find out what the consciousness of particles Because it's like we can't really find out how they combine to make our consciousness. So there's certain details we can't fill in. But still, it might be that there are certain philosophical arguments that it's probably true.

He seemed open to. But so that was, again, an interesting... I'm happy to sort of endorse a conditional that given a certain conception of consciousness, then it's probably that it's very plausible, perhaps the best theory, given a certain conception. I mean, I do a sort of... modus tollens on that and say that I think that that's the wrong conception to start with. But in a way, I'm more sympathetic to friends. You see, the thing is that conception, I think, is one that's widely shared.

by people who have no time at all for panpsychism and might even think panpsychism is ridiculous. Now, I think I've got more respect for the panpsychists who say, look, you know, we're starting with this conception and look, if you take that conception seriously, you kind of end up here, you know, at a panpsychist view.

All of us say, well, we have that sort of conception, but we're not going to go and say that we're not going to go down that road. I think it's a reasonable route to go down once you start with that conception. I think that's a reason for not starting with that conception. That was the point of the paper.

I, you know, if it is this, if consciousness has no conceptual connection with any psychological processes in a functional sense. then why think that only creatures with conflict psychologists have it? So, yeah, I think in a way what you're doing is you're nicely... sort of taking phenomenal realism and sort of you know really taking it seriously and saying okay you know if I'm a realist okay this is this is where you go and I think that's useful either because it's

either whether it's true or whether it's a useful illustration of where you go if you take it seriously in both cases it's useful to explore that bit so yeah yeah Yeah, well, I'm happy for you to defend the idea that either illusionism or psychism is the, because... Yes, yes, I think you should have the same attitude to illusions, of course, that if your conceptual consciousness is wrong, then...

But why didn't you give your thought experiment that was accompanying the tooth extraction? Because I guess that's where you were your reason for thinking this. this conception of consciousness that is i guess pretty widespread it is problematic yeah i mean yeah the idea is that if you if you know if you if you detach

consciousness, the notion of consciousness from any psychological effects, then it's hard to see why it should matter. I mean, if it's not having, if it's, if it's the, I mean, do talk of it sometimes, the intrinsic nature of a brain state, let's say the state that realizes pain.

And let's say it is that. Well, okay, but what's that to me? Why is the intrinsic nature of one of my brain states of more concern to me than the intrinsic nature of one of yours? Only if it has some further effects on... the rest of my cognitive system i think i was it's something that's occurred to me partly during the the debate the discussion we had with matthias that it's kind of where are we in this picture i mean we talked about unattended

So there's conscious perception going on. And I think, as you or Matthew said, our mind is somewhere else. Okay, so there's this conscious perception, but our mind is somewhere else. So where am I in this picture? Am I with my mind, with these conscious reflections, or am I over here with my perceptions? Where is my locus of interest in this rich?

pattern of what's occurring inside my head and am i somehow the intrinsic nature of all this or am i the sort of the sum of all the dynamic uh all the activity in there and i tend to go with the activity So I think in a way, the debate about consciousness is much about trying to locate the self in all of this as it is about the nature of some property, some mysterious property.

So do you want to give the thought experiment explicitly? Oh, shall I give it here? Oh, yeah, okay. Yeah, it's okay. So you're having a wisdom tooth extraction. You're going to be in the dentist's chair for... as i was for a couple of hours while he basically saws at your jaw and uh pulls bits out of it all right

I mean, I was anaesthetised, but it was still pretty traumatic. So you're going to have a choice of two anaesthetics. You can't take both, OK, because if you do, you'll die. So the first one, it, what is it, which is the first one? The first one, it. leave okay so when you when you have the operation you're

pain receptors in your tooth or whatever will fire and they will send the signals to the brain and your brain will enter the state that sort of registers pain. Let's say it's your C-fibre's fine, it's not, but whatever it is. And now... one of the anesthetics, what it will do is it will stop that state having any effects on the rest of your brain or any other brain systems. So you won't believe that you're in pain.

You won't be able to think about the pain because you won't have any information about it. You won't have any memories of the pain, not even from a millisecond.

You won't have any emotions. No emotions will be triggered. You won't feel fear. It won't have any effects on your behavior. You won't be struggling. You won't be begging for it to stop. You won't have any... wish for it to stop because you won't have any information about it there will just be this intrinsic feeling to the to the brain state so the brain state itself will be sort of screaming in pain but the rest of your brain will know nothing about it and then as i put it there hi

won't know anything about it because I'm just the sum of all those brain processes. Now the other one, the other anesthetic will do the opposite. It will dispel the intrinsic nature of that. brain, say, let's say if it's made up from the intrinsic natures of all the little particles that compose it, as panpsychists say, it will dispel that, it won't stop them all combining. So the state itself will have no intrinsic...

quality to it, but it will still have all the normal effects of pain. So you'll still believe that you'll be absolutely convinced that you're being desperately wish for the pain to for whatever's happening to you to stop, you will uh all the emotions associated with pain will be triggered you'll you'll be you'll feel afraid of a feel there will be explained in functional terms uh you'll have

You'll have horribly traumatic memories. You'll be struggling. You'll be screaming inside. You'll have all the effects of pain, but without the intrinsic paininess. Okay, so in one case you have the intrinsic, or your brain state has this intrinsic paininess, but no effects of pain at all.

psychological, behavioral, physiological, and the other one you have the opposite, you have the intrinsic nature. You don't have intrinsic nature, but you have all the effects. And so you're supposed to choose which. one you would take. And the idea is it's forcing you to think about where you are in this picture. Are you the intrinsic nature of your brain or are you the sum of all the things that the brain is doing? Where do you locate yourself?

My intuition is that the more I spell out the richness of the psychological effects, if you just say, oh, you'll think you're in pain, if I spell it all out in complex detail, my intuition is that... And I suspect that people will go more and more with wanting to quell all the psychological effects. Because after all, you're in a state, it seems to me that, well, so long as this...

the intrinsic nature of the pain state is not having any effects on me. It's nothing more to me than the intrinsic nature of one of your brain states. You know, why do I not... Well, I mean, I do care about your brain states, obviously. But, you know, I don't care about the way I care about mine because I don't know about them. And if I'm being sort of hived off in that way, why should I care about it? So my objection to you on this was...

which is uh the objection i raised in the session itself is yeah just to just to get that on the record um that um you know you're relying on a sort of purely causal my theory of what we call it mental representation, things like beliefs, desires, hopes. fears a purely causally defined notion so these are just like like we might define

poison in purely causal terms. It's, you know, whatever kills you when you ingest it. You define it as effects or flammability. You're defining beliefs and desires and hopes and fears in purely causal terms, in terms of their effects.

So then you can say, because the effects are the same, you don't have beliefs about your pain, you don't have that. But many people on my side of the debate actually reject that kind of... causal analysis of beliefs desires hopes fears we think these are themselves kinds of consciousness highly evolved forms of consciousness to believe i'm in pain to fear my pain to the emotional reaction to desire to hope to fear these are

forms of consciousness so um so then the picture gets much more complicated and you use this word yeah to mean this I say, okay, so you've got this emotion or this thought or whatever, and it has this intrinsic aspect to it, and then it has all the effects. And so that's supposed that we could have all the effects without the intrinsic nature. Wouldn't that be just the same?

you know just moving the intrinsic essence about within the brain or whatever isn't going to you know just making the intrinsic essence sort of track the causal properties isn't really going to make it the important thing it's the causal it's the causal

story that's the key one. But I should write, I said in the talk that I thought I'd maybe done with panpsychism, but...

Maybe I should write a bit of... I mean, it's just occurred to me right now, a connected thing might be Terry Horgan, one of his arguments for my kind of view that beliefs and hopes and desires are forms of consciousness. He has a kind of... partial zombie thought experiment where he thinks we can strip off we can imagine creatures um that are like us but they don't have agentive phenomenology so they don't feel like it's

it's i am reaching out for the cup they just sort of see their hand reaching out for the cup they might desire it as well but they uh and and we can and then he strips them all away to the end he thinks you know if If all consciousness was just colors and sounds and smells and we behave just the same, but our consciousness was just colors and sounds and smells rather than, you know, I like cherry hog and I think my consciousness is.

seeing a face that's part of the character of my experience and you know seeing a table and a chair if it was all just colours and sound we'd sort of be like non-human animals we wouldn't have any kind of concept our conceptual understanding of reality is built in to our experience right now if I look here I'm actually my father I'm staying with my in-laws I'm in my father-in-law's recording studio at the bottom of the garden

If I look, I'm seeing a microphone. That's part of the character of my experience. and yeah i'm happy to agree with that and that's you know that's your phenomenal phenomenology but i'm still you know i'm still going to tell the same story about that phenomenology that we cashed that out that sort of you know richer sort of conceptually infused phenomenology, again, in terms of effects, in terms of reactions. It's not a debate about what the phenomenology is.

about how rich phenomenology is or whatever, or how conceptualized it is, but about how we cash out those claims, whether it's in intrinsic terms or in relational terms. But you think those sorts of things are not phenomenal consciousness, right? You can't meet UO as an argument for that.

if you want to say that you want that you want to talk about those those those things in the sort of vocabulary that we use for experiences the vocabulary of you know if you want to produce sort of phenomenological reports that stress the similarity between cognition and sensations. Yeah, okay. And to say there's an element of sensation in all cognition or something. Yeah, okay, that's interesting. But it's not going to force me to change my view on what sensation is.

That's not the view. You're not allowing me my view. My view is not that they're similar. It's that they are states of phenomenal consciousness. I suppose they are. That's how you characterise. But I still deal with them in the same way that I deal with all others.

the states that you and i deal with them in the same way that you know i deal with a pure experience of the blueness or something you know it's I guess it'd be harder to set up a part experiment, wouldn't it, if these richer things come apart. Are you still there? I'm still here. Sorry. Just drifting off a bit. Well, we've been going on for a while.

Maybe this is something we both need to think more about and we could revisit it. I just don't see how sort of introducing... sort of more phenomenality does anything to undermine my story about what we're talking about when we talk about phenomenality that you know it's not really you know phenomenality when the claims about phenomenality should actually be cashed out in terms of reactions and so on.

just positing more of that stuff. I don't see how that, you know, I just tell the same story about the cognitive states, it seems to me. So they have the, they have... They have the element that is captured in the standard story that doesn't see them as phenomenal. And then they have some extra element that you want to talk about as phenomenal, but which I capture in terms of a further set of distinctive reactions on top of those ones.

But if beliefs and desires are states of phenomenal consciousness, then you can't describe the thought experiment in the way you do. Well, I can still describe it in terms of reactions. I mean, maybe I can't characterize those reactions simply as beliefs in the pre-theoretical sense, but they're beliefs star or whatever. And then we tell the story about what beliefs are as being beliefs star plus some extra reactions that turn them into actual beliefs or something like that.

It's something like that, I think. But it was, I mean, it was nice to have, you know, the first time in history, I think, a plenary session on panpsychism, which is, you know... you can't imagine that as eugen said in his talk you can't imagine that 20 30 years ago so it was you know you were both slagging it off it was quite nice to you know take it seriously enough to slag off

Well, yeah, well, I'm philosophers in no position to preach about what about taking things seriously. I mean, the philosophers have taken all sorts of things seriously over the years. And yeah, this is I don't, as I say, I think of the. phenomenal realist theories. I think it's okay. I think it's far less...

I think it's far more attractive than property dualism, which has a huge element of arbitrariness in it, it seems. Property dualists don't seem to take their own view seriously that this thing is... This property is characterised in completely non-psychological terms because they keep having this psychologising bias that it only goes along with psychological processes. Well, why should it if it's nothing to do with it? If it's a pure feel, why should it only?

I should just, you know, hang around only with psychological properties. Peculiar. Much more elegant, I suppose, that it's everywhere. Well, there's a point of commonality between us. And I think I agree. That, you know, honestness is wherever it is. And nowhere else. Hey, we're getting better at this, you think, aren't we? We did it. I can't remember how to stop.