Consciousness Live! with Javier Gomez-Lavin (S1 E8)

0:00 okay we are live joining me today very special guest Javier Gomez lawmen I want to thank you for joining me so hobby for those who don't know you can you just tell us a little bit about who you are and what you do uh yeah thanks for inviting me on first Richard this is awesome I am a cheese I guess I'm a philosopher of mine philosopher of science x-5 former neuroscientist I'm doing I'm going through a lot of transitions here but I'm starting as a postdoc at the University of Pennsylvania in the philosophy

0:32 department working in part with Mira which is actually Lisa maracas group which looks at issues of philosophy mind philosophy of science and artificial intelligence so I'm starting in a few weeks I'm super excited I just finished my dissertation at CUNY on working memory and consciousness and yeah I mean that's kind of how it feels like at the mo m moving right now so apologize for this first of all well I've been here for a while and mine's just about a sparsity went to the michele condo like a car toy thing and I was so you say you

1:07 said you had a background in neuroscience what was that you did that as an undergrad or was there a graduate degree above there no so yeah I mainly did that so full disclosure my mother is like an old school electrophysiologist so there was this kind of I remember when I was like in first grade like her version of like not babysitting but like oh I have to take the kid for an afternoon was taking me to the lab refusing a rat errand like taking brain slices of I remember like this is an old school like with the oscilloscopes would

1:42 print out these large sheets of the various wave dynamics of the voltage clamps that they would use and you know on this on the neurons and she would make me go with a ruler and I would have to like highlight the peaks you know I just remember the smell of formaldehyde still takes me back yeah that's old school yeah I mainly did an undergrad but I feel like I had a lot of experience with it but in undergrad I started doing a lot of work on cognitive science and on neuroimaging so I was

2:16 really fortunate and I got a grant from the NSF to go for the summer to Carroll seekers lab at Colorado State which is an fMRI lab looking at basal ganglia structures and their how they how they work and how they might in dancing mechanisms related to expectation and learning and so that was my first real dive into it was I had six weeks to develop a fMRI study and a really good team there with Curt brown liqu and Carol Seager and we developed a paradigm looking at whether or not there were

2:49 different we whether or not we could show a differential recruitment of brain structures depending if you were using categorization rules or not so we were looking at the differences of memorizing and trying to keep in mind a category versus specific low-level visual features of an object and that eventually was published in neural image in 2015 or so but I don't if it's not great like but it was really cool to learn how to do it and it was hard and I have total respect for no imagers it just made me realize like I didn't want to spend all my workweek kind of looking over massive matrices of data being like

3:25 where is this signal we can happen I wanted more to just like sit and write those discussions section since I think that's what led me more to gravitate towards this intersection of looking at it more and a theoretical level like some people might like Richard Feynman might say I'm like a janitorial neuroscientist I like I clean up people but I think that there's value in that would be would be like I don't know like maples during a strike you know and I think we're at a point where we're excited a lot of the old guard is changing in philosophy and we definitely before we get all excited and just pick

4:00 up where the neuroscientists are and the money is that we clean up those concepts something yeah that's an old school so you're you're a practitioner of the handmaiden to Simon's view a philosophy is that I don't necessarily know know I think like it's tough right I think when I remember being in the lab and there was this really smart kid he's I can't remember his name is like William or something and he was like doing it and like super intent and I realize like he's a philosopher he's doing epistemology he's just doing it bad

4:34 let's look at it the opposite way I think scientists are philosophers they're just doing it sometimes a little bit on a drone level without thinking about oh man if I use certain words and we could get into this later like activation yeah the way I build my theories and I'm picking out the word activation from a methodological like constraint like I because I use a certain instrument then all the problems with that instrument are now baked into your theory and you have to make sure that you can have the conceptual space

5:06 and where we call to separate this no that's actually the correct yeah no I've worked in a couple of neuroscience labs as well I almost thought I was gonna do a PhD in neuroscience but I couldn't do the chopping of the rats that was my downfall and I didn't really like the the EEG work like neither cheer nor there but I do remember that I was in the lab one time and the P I said you know he didn't have any theories and so I have no theory we have no theories here we're just collecting data I was like holy [ __ ] are you kidding that's all we have our theories like there's a

5:39 theory here somewhere the more you deny it the more its dominating probably what's never got to level were you chopping humans you were comfortable with that but not the rats well I never chopped humans no no the worst I did was scalp recordings from undergraduates which was already pretty bad actually but I never got to do fMRI though actually I did I did ERP which was funny yeah no definitely sorry I didn't mean to cut you off no that's fine no but I think you're right it's it's

6:11 really interesting to see even when I was there in Colorado within like a few months how all the mathy modeling stuff which I was like oh man I can't do that like you know how that just revamped the way we were doing it in the span of a month like I remember when I got there we had to do this so tedious process of you would have the anatomical image so this high-resolution image of somebody's brain first you would have to stretch it into a kind of standardized States the tower at space and then you would have the functional data which was like much more of a fuzzy

6:43 cloud and you had to like literally use your own machine learning algorithm aka your brain it's like map it on perfectly in three dimensions and it took like hours you would have to do this very participant and by the end people had models and algorithms that would do it and now I think the biggest advent the thing that I think I want to flag for philosophers is like if you're reading an fMRI paper and it doesn't have the words or the letters MVPA or a multiple multi voxel data analysis you need to stop because it's probably unless you know exactly what you're doing you're

7:16 probably relying on stuff that is basically not null and void but pretty obsolete and we have to be very careful when you look at proves a philosophy to paper and they're citing all this neuroimaging from before maybe 2007 at the earliest they're really 2012 13 14 it's you have to take it pretty critically because you click the whole univariate model this idea that you know which we go back to you know functional dose associations and Penfield going in and like probing people's like sensory motor your cortex is and making them

7:48 move their arms and stuff that's been nuanced and I think you know there's some people who are very explicit about this lab Postel in Wisconsin I think that's some really good work nuancing this but a lot of philosophers aren't aware of that division in this idea that like yes we know that the sections of the brain do specific things so the you know fusiform face area you know home tends to light up when we see a face but we can't be ignorant to the fact that it's not doing it in a little glass while it's doing it in the context of this insane structure yeah the office doing all these other things and really

8:23 you have to look at how that brain state together I don't want to use it word globally but together constitutes a particular mental state and I think with the previous techniques where we were subtracting out everything it's at the highest signals we were losing the kind of connections that are really important to creating particular mental state for instance and I think that's what you get when you're losing something MVPA you get better resolution and so like you don't filter out all the what you thought was success information get a better picture of just how the brain state itself looks

8:58 throughout the brain in a given moment and I think that's really important and I think that's some cool stuff that I'm excited to see the develop where it goes and I'm excited to see philosophers pay more attention to yeah I couldn't agree more I think multivariate pattern analysis is is one of the most important developments and fMRI and I mean fMRI studies before I mean there's still kind of highly underpowered but they were incredibly underpowered before and it's some very dramatic kind of claims were made on the basis of this kind of crude method which for a long time people

9:30 thought was the most advanced method that was you know that's where you get all the big grant money for him but I think it's important to keep that kind of stuff in mind my own lab work with with the animals was we're looking at you know single-cell recordings in in living animal so I was you know coming from come from that background people thought you know fMRI ERP NSF was just too wishy-washy if the data wasn't good you want to get in there and record from the actual neurons and I think you know it's gotten better so I think I think this MVP a isn't definitely one of the

10:02 ways in which it's gotten better so I'm glad you brought that up that's very interesting so I didn't realize that about your history I think you know it's important that philosophers have this kind of background and the actual sciences that they're that they're interested in philosophize your mom so it's important let me ask you because I know that you're I mean I want to get to talk about working memory but one of the things that struck me by reading your work is how much attention you pay to the history of this of the topic going back to looking at the way the working memory models have was started and then evolved over time can you tell us a

10:36 little bit I want to ask you about the specifics of that but before talking about that can you tell why do you think that that's important to go through the history of this kind of stuff rather than just talking about where we are now totally I think it's it actually hangs a little bit on your last point about the the importance of loss aversion if they're interested in doing a particular science thing get their feet wet that doesn't mean they have to get a degree necessarily but there's a sociological component and and doing the scientific work I mean it you know I know a lot of people might poopoo like tellin Lanza no and these people who like really take that seriously but I think if you don't

11:10 if you don't have that experience going into a lab it really is easy to miss yeah that there is a deep sociological component and things get ossified and so ways articles are written are written for audiences that if you're not aware about the dynamics just like you know how is the scientists must be like wait why did he send this to consciousness and cognition and not mind or like you know there's all these dynamics that go in but if you're not aware of the context you might miss out what the the negative space around the article or claim and how that frames it and I think then you might be willing or tempted to

11:44 supplant your own and then you might get it wrong and I think we could talk about this a little bit I think Carruthers does a really good job at actually bringing a lot of these things to bear that are new and exciting and MVPA inactivity silent working memory but it's like it's also being careful about that negative space around it so to me the most exciting thing about writing the dissertation or not writing but researching it was actually going every library I would go to that I would have time whether it was when I was teaching at C of C or at Brooklyn College or at because the grad center library let's be

12:16 honest is a little yeah the old psychology section where they would have the tones of these like hand books from the 70s you know and that's like sociologically like that's how they did it they would just be like the seventh volume of the handbook of learning and memory you know for 1974 and you know they hadn't been open since like Aidan look at it and it smells like that wonderful smell of old books which is probably bad for you and just going and cruising and looking at how the debates about these concepts were on we're

12:51 really a live issue and like realizing that working memory though it's now like battle a is God and like yes he's given us a Testament from the mouth and it is the standard multi-component model and if you go to any I was doing this thing when I was writing the dissertation where I was treating like this taxonomy animals are learning in like an mode to look to the a Aristotelian salat like this index la quorum of like Oh all the references and all the definitions of working memory and almost every paper trots out basically the same structure it's like working memory is

13:25 capacity for to maintain and manipulate limited information over small durations and you know the most standard models Baddeley's melted component model whatever and just I got ten thousand words of just people just repeating that mantra and you realize that going through these old books and seventies like bad lays is one of many theories that's jostling for attention that's jostling for not even dominance but especially earlier like 74 75 you know people were more talking about Broadbent's primary memory they were

13:57 talking about a whole there's like I remember this one I think it's 1976 Crick and Levi they do this beautiful table that just like ghosts are 12 different theories of primary memory you know the memory used short-term memory wherever however they wanted to call it this dissociation from long-term memory and they spent like two paragraphs talking about battle A's multi-component model they spend most of the time talking about Atkinson and Shiffrin model which becomes I think which badly took inspiration from so it's really interesting to go back see these things

14:29 as live debates allow yourself the conceptual freedom to be like what if they are wrong you know what why do you think I mean so I mean I want to get to the specifics again in a second but sociologically then what how did the dominance of badeley come around you stink what what actually did people just get tired of debating it what happened no I think it was I think no I don't think that's the case it's it's a good question I had to like kind of stop myself from going past 76 mainly because I was just trying to show the how there was a previous model that was used by

15:04 Baddeley and the connection between their two models I think the reason why ballets been successful this B's battle a is a really and let's the other guy oh my gosh I can't believe I forgot hitch there are really good experimentalist so they had and I think we have to it sociologically again realized this was still like if you go to these books they're like hand drawn grass right and they're trying to be like we're a real science like they have hand drawn or kiss renditions one of my favorite uh which we can put it up on the screen it's just like hand-drawn information processing

15:39 model the brain and literally is like a bunch of boxes some of which are transparent with like an eye at the front of it and like a hand coming out the back and it's just like really nice but you gotta realize I think sociologically psychologists were like we're real scientists and I think the reason why they were successful is because they really were good experimentalist and they showed this dual-task property right to look like rather if you have a unitary short-term store then if you're trying to remember for instance auditory and verb like word

16:16 cues or number cues you're trying to keep those visual cues in mind and auditory cues in mind if you're trying to keep both of the mind it should deplete it equally right it should complete that that that short term store equally and you should be able to see deficits and the way they tested this is they have this kind of phonological loop mechanism proposed and a visuals sketchpad proposed and they found that actually people could maintain those different modalities of information without it necessarily affecting the other when they were under cognitive love where they would ask them like

16:49 true/false questions about the stimuli so they show that there was a kind of modal distinction there whether you could keep both visual information in line and auditory information mind with it without depleting it at at this kind of general store Frei as a general store so I think and they were able to show this very robustly and reliably and they gave a good mechanism for it and I think that's what attracted people to move away from the unitary model I think the danger with battle-ax is and it is a good proof I think that Union can store it's a kind of proof that I'm attracted

17:22 to this idea that the brain can sort also can maintain all sorts of different representations they kind of restricted it to two at the beginning but they give no principled reason why I shouldn't be more and I think the problem is because they were stuck with this model of these two slave systems so you can tell it's written by white people systems sorry that's all right call it what it is you know it was their model and they didn't want to change it so it said they started adding these special vague things like the Central Executive and this episodic buffer which

17:56 does everything else and it's just like you know you start to lose a theoretical grass to be able to critique I think once it got that people were happy they started adopting it and then it becomes a kind of thing where I remember being in an fMRI lab and wanting to critique the stuff and their being like you can't that's speculation you can't put that in the paper but that's not what our research shows and I think you know there's these ossified sociological systems that make it so now we're all beginning all our working memory papers which if that covers such a huge range

18:28 of phenomena from olfactory responses visual responses motor responses from different methodologies whether it's EEG fMRI oscillatory dynamics behavioral measures all this stuff and we're like it's all working that we have to take a pause and be like wait maybe that's not the best way to think about so yeah no I've definitely heard some more people when you make these kind of points they say well isn't it all working memory it's like you know in a way that's what I mean is your thesis maybe get big is that because I think that you where you end up with and maybe we're jumping

19:02 again but I want to come back to some of the details the second but I think what you end up is that working memory is just a name for cognition in general a bunch of a bunch of different ways of keeping information relevant and tasks ready so that is just another word for what we mean by cognition is that is that that's the deep worry that like at Lisa tree describes I mean cognition again it now it depends like are you an embedded and body person or do you think like the ear is doing cognitive work when it's like deflecting sound signals and so forth but I think especially when we're interested in like thought and

19:35 decision-making you know whatever kind of cognition deals with that maybe some people call core or central or higher right debate why one is better or none or good but that kind of cognition I think the problem that we slept in that we slipped in even earlier than battle-ax and it's just been ossified and preserved at the very heart of this construct is that really working memory just read ascribes a phenomenon it's meant to explain which is how we're able to think and deliberate and it's very obvious when you even like I remember I

20:08 think in Baddeley's 2001 book the first line is like working memory underlies all our ability all the human uniquely in the uniquely human abilities for complex thought which again I have a problem with but still you know it's like and then you know it's really funny I remember I was in Yakima's facilities their class on de anima and just reading the data and reading all the debates on fantasy ax there's like a mechanism that supposed to accompany every thought that kind of preserves it that's essential for memory and realizing like this anxious I don't want to be like

20:44 Aristotle discovered working memory but I want to be like look we've been in terms of our architecture and understanding of the mind we've been dealing with this for thousands of years so it's it's not a surprise of nothing like that battle-ax even better no I think it's it's kind of part of our theoretical pre commitments about how the mind should be and I think that's what takes that's what gets scary it's like wait a second yeah I don't want to go folder on it where we're like we'll never know how the mind works you know I don't know Steven Pinker doesn't know God even if God told us we wouldn't understand I think you know there is

21:17 some room for not for optimism but we do have to even go back to those you know pre assumptions as intuitions we have about the mind because I think it's embedded in there that we want this controller we want this thing that structures our thoughts and their decision space and we can rely on other people having and that we have internally and that grounds thought and decisions and also all those heavier ticular concepts like rationality and deliberation and agency and I think we have to maybe check ourselves and ask

21:51 wait what do we actually have what do what does the research show and I think right now it shows that and I think I'm really indebted to the touring and visiting and talking to a lot of people in John Dylan Hayes lab over Berlin school of mine and brain and Shahid day especially Thomas Crystal fell and these folks over there because I think they're the ones you know their latest paper in 2017 or so they're like we got a restart thing about this it looks like the whole brain is capable of maintaining information where it's created and it's like yeah I mean that's that's a different model of the mind

22:25 than we've had before so any I know I'm rambling but no that's no it's totally fine I was gonna yeah I was gonna bring that stuff up anyway that sort of need to have a central place a central thing that is us or that is the doer of all the things that we think are important and that's that's seem to be what's behind a lot of the enthusiasm for these sorts of models so going back to again sort of the more specific step so you think that out of all this the two kind of things which counts most like that

22:59 look most like what they should be called a working memory or maintenance and manipulation of information but then you also think that those things are found ubiquitously in the brain like I think you said that in the retina maintains and manipulates information if your argument that maintenance is a kind of manipulation is correct so I wonder if first of all so if we're looking for something that is a natural kind those would be the hallmarks of it is that your view so I don't know I you know I kind of had to I feel like I use the framework of natural kinds because it

23:34 was a way for philosophers to get their teeth well you know to whet their appetite and kind of figure out this this this concept and I feel like as I go forward I want to rely less and less on that and I want to motivate philosophers because I feel like there's old people philosophers have a lot of hang-ups with that and I think we need to realize like even if it's not a natural kind whatever that is and I use the most I call it the most congenial theory Khalidi Siri we're like you know you haven't been slow bar I think like look an espresso is a non natural kind

24:08 but it's still important we can still talk about it we could still theorize about it so that doesn't I don't know I feel like we have too many hang-ups about why that's an important thing we've packed too much into it so I beyond that I guess what I'm saying is I think you look at the literature when you look at the way people operation lines what this capacity to hold things in mind into that's important for thought actually should do is that it should and this is just I this is like this taxonomy I build up 10,000 words and I think something like two hundred ten

24:43 definitions and almost I would say 80% of them and maintenance or manipulation and some white built into it and I think if you go back you see this even with like Atkinson and Shiffrin model which is the first real model of working memory this unitary model which is very beautiful it's like they have stimulus coming in then it goes through the stimulus registers like haptic visual etcetera then it can either stop or it can go through this temporary workspace they call working memory in which case it could be maintained through rehearsal or you can do stuff like this so you can

25:16 manipulate it and then either it gets we because either behave based on that information or it gets encoded into long-term memory and I think it's just at the core of the concept as it's been operationalized in science and psychology over the last 50 years that maintenance a manipulation you have key pillars of it the problem is I think there's not there's some there's some efforts to do this kind of caste dependency so like maybe we can cash out what maintenance is by giving it a kind of specific tasks profile or made a manipulation especially but I feel like

25:49 in general people haven't spent enough time thinking conceptually about like why in may even be the case that those two things are separable like isn't it you know doing the philosophy thing for a second isn't it the case that maintenance like this if it's not just a static thing which even at the beginning in Atkinson's model they assume that it's not a static like image that's retained in the brain it's an actual rehearsal process it's an active thing that disappears it comes back disappears it comes back and you know this has been shown in terms of like how the

26:21 oscillatory dynamics works but also things like an activity silent working memory states which I want to talk about later but it's an active process and so unless you're making this kind of weird ad hoc do or post hoc the visions between the two you need a kind of conceptual reason by thinking that maintenance isn't itself a kind of manipulation and I don't even think that kind of work has been done and I think that what if we're also finding these terms in and I think activity is another one like active maintenance and what that means right you know we're we're like basically betting and committing

26:56 ourselves so framework then might not be the best and peer to base empirical investigations on in philosophical discussions on so that's that's interesting but I wonder if so I mean I think active maintenance is probably the way that I would be tended to tempted to characterize what I think is right about the intuitions behind working memory that you're actively maintaining the information what is it about the active maintenance that entails manipulation it's just that is it just that you're

27:31 doing something with it when you rehearse it so if you just take the classic case of remembering a number you know we could repeat it over and over your head over and over and your head trying to remember the series of digits or whatever how is that manipulation on your view or how is that something not we would think of as manipulation so I think we need to figure out what what what the desiderata for something to be manipulated is right so if it's an actual change in the contents right then I think we can start theorizing a little bit about how that might be different but the thing is that if it's so assume that you have some representations that

28:05 you're trying to keep in mind like a representation of an apple or just you know we could go with a and B like whatever a and you want to manipulate that by adding because you've learned something else so you have some background knowledge about it you want it to be like a prime and so the question is what are the what are the kind of abstract processes that might be involved in that that that journey from a to a pun is it the case that you there is a mechanism there that sticks the prime onto the a is there like an actual mechanism that changes that

28:38 representation and it's so I think that's very interesting we might have to talk about that as a separable system but my intuition is that's gonna be very depending on modalities specifics and modality dynamics it probably isn't gonna be one universalizing mechanism that does that but even so in the process of remembering in the process of memory now let's say we have whatever magical system turn turned it from a to a prime now we're trying to remember that so in some sense maintenance is built into the prospects of manipulation almost as a necessary condition so I think seeing them as separable in a

29:12 strict sense without postulating like what we want maintenance to accomplish minimally is a bit dangerous so I think it's not so much that I'm thinking every instance of maintenance is gonna be I don't I think if anything manipulation of a minimal sort is a kind of a fundamental property of the brain and so far as like I call it at one point like a currency of cognition I think we're gonna see it everywhere because it's just some product of physical systems it's like that's just a bare-bones like you know kind of relational property

29:45 that exists in the world that other systems can take advantage of and I think it's an important relational property for how the brain works and so we're gonna see it everywhere if we want to then derive specific manipulations we need to realize that that might be this whole separable class of processes it's gonna be modality dependent that's gonna be very distinct it's gonna be very important for us understanding how the brain actually processes information but we can't necessarily see it in the simple binary framework though so the public doesn't make it easier but it doesn't make it but I wonder so what

30:19 would you say if you wanted to add things like resist distraction or something like that so what if you know sometimes you hear working memory described in these terms that you keep information online and you can resist distractors and then you know answer a question about it later or something like that so those kinds of things aren't going to help adding to the definition to make it something that's interesting so it's not that it's not interesting I think so in one way but not in another so I think the more kind of operationalized terms that you add on

30:52 to this whatever view working memory so let's call it where you know the specific order working memory the more like caveats and tasks dependent properties that you add on of course you're gonna find something that does it and you're gonna find some people that are better than others and you're gonna be able to say these brain regions seem to support it but then you're not gonna be able to come back around and say okay so we found the key just because you found something that's past sensitive that works with all these optimizations it's not a clear foregone conclusion that that thing's supporting the general

31:26 properties that we've already associated with working memory of thought and decision-making right so it's almost like we have two choices either we can separate it and become narrow and and drive down and call working memory just one of the many ways we can maintain and manipulate information however we decide to that's be so and so robust so and so you know ask me for chunks it has to be at least for second I've got to be such a suction robustness and it's gotta work you know it's gotta have this dynamic I'm sure you're gonna find that it does exist I'm just not going to be convinced that that's a uniquely unified separable

32:00 system of the brain that itself is important in how that self is fundamental in the same way that we've discovered in your experiments and the way we think and reason and and decide things I think we have two choices here we can either say yeah there are going to be so many mechanisms and ways in the brain and they're gonna attach and combined together and innovative in different ways to allow us to do things like remember things with distractors present but that might not itself underlie or be the key to solving general problems of cognition or thought

32:34 right so it's either gonna be so specific that it's kind of uninterestingly general it's not generalizable or it's going to be this other thing that's just more general and so also not interesting for helping us theorize about the nature of a manatorian an unexplained Ettore and I think that I think it's what I would like people to take away from this is yes maintenance and manipulation are important but they're gonna be instantiated in all sorts of different ways we have to be sensitive to the conceptual distinctions at play and I'm not saying I solved it but I'm saying like look I just want to show that you

33:07 know at least from a base like simple philosophical level it's clear a priori why these things should be distinct especially we have this notion of active maintenance because it you know it might be the case that when we actively maintain something it's like you know not to get into occasional ism and Descartes but it's like is it is it a new representation every time is it the same representation is that a silly way to even think about it but in some sense I think we need to do some of that work and realize that there's going to be really a lot of mechanisms that support these these these these are

33:40 probably not going to be natural kind terms in themselves they're gonna be too generalizable and we're gonna have to break them down in ways that are way more tractable to the fact that there's going to be a whole host of modality sensitive components that are doing all of these operations that are but the brain does and that's okay you know so you're not a skeptic about about the other normal candidates for memory like long-term short-term memory I mean you're not I or does this get this an extent that those guys are not a little bit I think I mean I'm a skeptic of the ways you know I almost feel like and

34:14 this is you know probably more fault of my own vanity and pride like it's it's kind of like theoretical physics before punk and I'm not saying I'm Punk but it's just like you know we're like we got it this is easy we're done everybody you know there was this amazing lecture when I was in neuroscience 101 the guy was terrible I'm not going to mention his name but he worked on like crab Anatomy and he like showed a diagram of like the 32 neurons that control the max of the jaws of the crab and how they napped on every dendrite every synapse

34:47 every connection every neurotransmitter phase change like and they still couldn't successfully model it and it's because they didn't take the context of the glial cells involved and the other things this is 32 nah yeah I get it and so it's more just I want to urge a bit of conceptual flexibility but I think it's to be very important but also the idea like of course we can do all these things we can think we can reason we can deliberate we can remember things we can remember things at different time scales but it might be a spectrum of processes that's involved in this and at many different levels and I think as

35:20 long as we don't commit ourselves prematurely to an ontology that isn't flexible or I guess if we do do that that's bad basically and so we just need to have an ontology that is more flexible and sensitive to the fact that meaning it's a manipulation I mean those things probably exist in all sorts of forms all over the nervous system these it's just how information gets processed minimally in a biologically realized system so sure they're the cornerstones of working memory but they're the cornerstones of long term memory they're going to be the cornerstones of like about how you get the mock battle lucien and the neuron

35:52 and i mean in the retina yeah like it's it's gonna be you're gonna find this property throughout so there it's more about the second order properties that derive from the specific mechanisms that we might track for example how does the mac find allusion work oh it turns out it's this lateral inhibition of your of the cells in the retina that are causing these kinds of edge contrast changes and that's a good explanation and that's the second order it doesn't mean that all maintenance is lateral inhibition right right like there's many ways to instantiate this and I think really

36:26 opening ourselves up to that and I think neuroscientists know this I mean every time I have this talk more scientists alike we know that working memory isn't like a real thing is much broader thing and you know it's funny you read a working memory article and you to me you just go to the second paragraph cuz that's really what they're testing saliency maps in motor cortex it's like cool that's awesome let's do that that's cool stuff you know or they're looking at beta wave beta for a beta phase synchrony in oscillatory dynamics during tactile responses that's cool let's do

36:59 that but to show them all into one thing is I think that's how you get grants sociology is important yeah yeah you got to talk that stuff up so so given I mean III find your kind of considerations very convincing at least the the – of cold water is very refreshing so what practical implications do you think that this has for like theories of consciousness so I know that you go

37:31 through pretty much all the major theories of consciousness and say that they have problems with this but it just generally and in particular if you want to talk about specific once what do you think the issues are with those theories and their reliance on working memory well I would say first order theories I can't do all of them right so yeah I think first order theories are the ones that are particularly and I don't know we should talk about this a bit because I know that you know coming from the CUNY hive mind and having higher-order thoughts flowing around there there's like maybe they also involve working memory but

38:03 maybe they take care of it in a more nuanced way I'd like to think more about that I really focused just on first-order theories because I think they're the most again this is very vain but I want to disabuse philosophers of a reliance on terms themselves that are potential black boxes as Agnetha theories and I think the clause was most at risk for this or people who are committed to kind of first order theories of consciousness so that is to say these naturalistic minimalistic neuronal or cognitive theories so people like the Han and global workspace people

38:38 like Carruthers and global I mean global workspace is basically the name of the game and of course there's a few other ones like Jesse's air theory which i think is also just global works fixed I think they're the most at risk so yeah I mean I think it's interesting because they also just to give three examples of them one counter Jesse's air theory you know basically says you're conscious when you have an attended intermediate level representation that's made available to working memory so working memory is a structural component of that

39:11 theory then you have people like de hann and Carruthers well looks over at the hand who says stuff like or looks over the cousins actually because he thinks a working memory is this workspace of the mind that actually instantiates the global it is identical to this thing that's a global workspace and to be conscious it has to go through or at least the contents of working memory are going to be necessarily conscious I know you know it's not the super strict claim but it is somewhere in that ballpark and then you have people like the hand who are much more careful they're like wait no

39:44 no no working their ways a separable thing because I know I'll get yelled up on my friends down the hall like what do they describe theirs the function the localization and the well basically the function the the where and the how and the whys of the global workspace is pretty damn close to working memory and so and it's weird especially to see them try to wiggle their way out of it but if you put pressure on that it becomes very obvious that either you have two things that do basically the same thing or they're

40:17 really talking about the same thing and you know them just basic fundamental parsimony requires us to take it seriously that working memory is an active component and their view of the workspace and again there's some there's some silliness there I mean they do talk about consciousness being global and they don't want to commit to it being in a particular area but at the same time it's and it's necessary you have it's necessary to maintain information and maintaining information as a byproduct of consciousness consciousness is a the workspace is generally and realized by

40:51 these frontal parietal neurons that have long axons and so it's a messages to each other or innervate the brain that is to say what do you really put pressure on it either they have to accept that there are two weird things that do very similar things or really it's it's one general system and because I think working memory packs in so much because and this is again from this crack and living review in 1976 when they're talking about actin sand and chip-ins model they say I can send and chip-ins model this isn't this kind of

41:23 where you know you have sensory buffers and then you have this temporary workspace and then you have behavior for like the temporary workspace is not this kind of platform for learning it it's all of cognition ok well if that's been baked in and priced into working memory in various ways though people have been more careful about it and so if you're committed to the view the consciousness is dependent robustly on this thing called working memory and if you don't really specify how working memory doesn't just read ascribe most robust cognitive acts then you're just saying

41:57 consciousness just as a byproduct of cognition which is seems trivial in some sense for a materialist and it's not explanatory and I think if we want to actually convince people like Chalmers or anybody it needs to be explanatory at the very minimal okay so that's why I take particular target those people because I think they've already made the mistake to some degree by pricing in working memory to their models and then LeMay is another one that's slightly different or Lama because well he's nuts right but it's

42:30 good oh what's he saying like hey it's this minimal recurrent processing it could be anywhere sorry I don't ask the hard problem because it's not working memory and I'm like wow that's exciting so I mean you brought up a lot of interesting things there but I wonder so at the sort of hone in on some of the stuff you just said so if Jesse hit you know attention makes these things available to working memory you object working memories gotta be fleshed out more it's not done or whatever it's not

43:03 a real thing in some sense that's important what if Jesse just switched and said okay attention makes these representations available to cognition which really is what you think his view says if he explicitly said that then you would say your theories not explanatory anymore but I wonder what's been lost I don't want to defend Jesse's view I'm not a big fan of it but I think it's worked on in enough detail that we can ask these kind of questions because most of what's going on there I mean I mean I've talked to people who say she doesn't even need to use the

43:36 word attention really you know it's these certain kinds of oscillations these frequencies and 40 Hertz they get entrained the informations broadcast to another area how does it being the prefrontal cortex so whether it's called attention or whether it's called working memory why does that in your view why does that drive the explanatory value of the theory that's a really good point so I'm wrong now I'm just getting a little more sense of the critique of Jessie I mean there's a reason why it's 60 pages but yeah that's long it's it's I mean I

44:10 love Jesse's my adviser but he's very careful it's really pinning him down takes a lot of work so yeah this to make a long story short I think it wouldn't make too much of a difference because the problem I think is more in this two things one is the specifics of what constitutes the right kind of attention like what constitutes a tantrum isn't this generic thing right it's a very very specific thing it's basically gammab and modulate you know its vector waves that are modulated

44:43 by gamma band activation in the right way and there's this key point where like he sneaks it in three or four times in a theory and in the right way means it has to be controlled by the right structure yeah and then one point even says ask me controlled by the central structure see central brain structure so it's not just any vector wave modulated by gamma van oscillation because then he would commit himself to all these cases where we can find that and it doesn't exactly look we have time about no solutions when we sleep right it's an amazing amount of work if you look into

45:14 it there's all it's like very clever work showing that gamma band like maintenance because it's again this might just be a real azure of one of the realizes of mainland is everywhere it's just a generic property of the brain that it uses along with other properties to calculate and actually process information and that's not surprising so he has to hone it in and he's aware of that because he's Jesse smart and so he hones it in the problem is in the vagueness by saying it has to be controlled in the right way by the right structure so then you have to ask you said okay well what are these right

45:47 structures what are the functions of them and and basically you can go two ways and he can either say they're not the same structures as working memory or they are if they're not the same structures of working memory then you have two central kinds of brain structures you have the working memory ones and any of these other magical ones and then he gets into conflicts about explaining just the rest of this theory in terms of availability and all these are the properties I'll get into in just a second if they are the same structures then it turns out that working memory doesn't just sit around passively being like oh I'm just available it actually engenders the

46:21 right kind of attention to working memory then engenders consciousness and that's just that is something that he's committed to not saying for very many reasons and one of it is because he doesn't want to give a weird disposition with account of availability right though right it can't be this like Ouma Syrian state where you have this vector wave it's modulated in just the right way and what makes it conscious is that it could in the future be used by working memory right that's is like a weird and I think Burge has very good critiques in this already about like why a dispositional cognitive view is not not not tenable so he doesn't he says to be available is

46:58 just is just to be modulated in the right way so it's a it is a kind of disposition wasn't but at the at the level of the modulation right so it's actually you know it's more like to be available to working memory just means that it has to be modulate in the right way but that modulation in the right way is actually controlled by working memory so now it's a source of complete into one kind of working memories doing all of the robust work and he can't really keep up all the divisions he needs in order to not make it just a variant of

47:30 global workspace including this non-centrality desiderata and on locality desiderata and but it still would be I mean so not to cut you up too quickly and I want to hear the rest of your thought but it still would be distinct from at least some versions of global workspace theory because they think that these thing these states need to be sort of Rhian coded in the in the higher-level I mean what they would call working memory and where jesse doesn't think that thinks they don't need to actually be encoded there so isn't that's a pretty big distinction so I

48:03 guess that would mean two things perhaps there might be some wiggle room there I don't even think that the I don't I'm not necessarily convinced about the re-encoding story on global workspace you I think but beyond before we get to that point I think with Jesse's view it's a case that if working memory if making a given a air representation available to working memory is just it being modulate in the right way and being modulated in the right way it just means that it is basically under the

48:35 control of working memory however that looks like then in some sense there has to be I wouldn't say I don't know if it's a rien coding but there is an effect of working memory actually the communicating with the structure and creating top-down signals on to it to amplify it maybe to amplify it so that it can actually be accessible to other consumer systems I think once you get more of that story involved you see that there's actually it's it's a maybe it's an identical to but it's a short stride away from something like a more minimal global work space theory where with global work space again I mean I

49:08 challenge the theorists to be a little more cautious about how they use the difference terms between crowd broadcasting and maintaining in the work space I think these are thrown around in a very messy way that's yeah very difficult it makes it very difficult to give a charitable critique of them but I think when you have global work space you have you know the generic bars view where you have this representation and what happens is it's able to be broadcast by a given system so that all these consumer systems can can see it kind of like the schoolteacher putting something up on the blackboard then all

49:40 the children at their desk kind of work on whatever it is that they need to and then some of them might go hey I have the answer and she'd out back out so I mean in some sense yes there's there's differences but I think it's within the family and I even Jesse kind of makes this that makes it kind of quip at some point in his uh in his uh in his book later on where he's like it is a kind of global workspace the one that's non central right that's the key difference for him that he sees between himself into him and the non-centrality means that it's not controlled by these

50:12 special workspace II areas that it just can be available to them and he's trying to give the whole explanation of how the the broadcasting and the and and not only the broadcasting but the attentional modulation that then begins the broadcast and he wants to put that into one component on the kind of supplier side rather than having it on the central side and I think the problem is he can't maintain that unless he explains the relationship between these supposedly important central brain structures and the vector wave or the

50:46 representation that's actually intentionally module so I mean there's nuances in both and I think it's one thing that people haven't given Jesse enough credit is to really work through carefully his view of attention which is weird and probably shouldn't be called attention yeah yeah no that's something I've heard before is that people wonder why does he called this attention and does it look like whether other people production so so if he agreed with your critique of working memory and recast this very without it and you think the same problems would

51:21 result so it doesn't I mean what I'm getting from you is is it sounds like it's not really a problem with working memory so much as that it's a problem with the way that he thinks whatever control structures are modulating the signal is going to collapse it into just a version of other kinds of theories but I'm not hearing a challenge what I thought was the beginning challenge which was the challenge of the explanatory this yeah to the theory no I think you're right so there's I think I approach Jesse's view more critically in terms of this kind of mythic enos about

51:54 it being internally consistent I think it is and I think that's because it's so close to being like a cool like whoa maybe he's given us this mechanistic real view that we can test about consciousness and then it just gives it away at the very end by adding this vagueness and these central components which I think there is an explanatory problem there namely that if the central components are the same as working memory and if working memory is just cognition then it's just kind of not saying anything interested be right this is almost a notational variant at that point like there's a separate critique

52:26 of being like look if it consciousness is just engendered by cognition what does that tell us explanatorily I mean nothing more than if we say it's engendered by working them it right I think that it's more like I'm feeling you know revealing like what we're actually saying when we think we're saying something explanatory by substituting a scientifically vetted concept like working memory in place of cognition yes right and I mean you know going back to Bernie and you mentioned his aversion of before he certainly you know some book was called a cognitive

52:58 theory of consciousness that's what he thought he was doing it's just that what they took cognition to the relevant cognition to being was working memory so now that's kind of taken up but I think the real insight that those guys had was that consciousness is cognitive it's somehow cognitively constructed or built up by the cognitive system so I'm just wondering you know if you convinced all these people that there are these problems with working memory what would they lose besides the term I mean the structures that they think are there the dynamics

53:31 that they think are there are still there and they still might play the role and it still wouldn't be cognition and they said what have a cognitive theory of consciousness so I just wonder what would be so devastating to the theories if they were if they adopted your view right so I mean if they just substituted I guess so but we need to be a bit careful here I think cuz I think we're giving them a little bit too much credit okay by giving a cognitive theory of consciousness is that they can decompose it into like these relatively testable modules and and that they make ensure

54:05 that the communication between the modules is what's making it happen and I think the way they do that is by our reliance on this kind of thing of working memory as one of the key components in order to actually give an explanation of how things are controlled and how things are maintained and how things are manipulated in the brain now the problem is if you get rid of those components if you say look in fact those components that actually describe because I think I'm willing I'm buying everything until you say like look the components and structures are still the same it's a no look it's like that's gonna the fun when we blow up working

54:39 memory we're not just saying oh just replace the term with consciousness and retain the central executive retain the phonological loop retaining this kind of fronto-parietal model where the representations are taken from sensory systems and like ferried over to this protective place where they're maintained and where we can use it as this workspace where a thoughts from across the cognitive economy can come together like a blackboard that we can combine in interesting ways I think no I stop no like it's it's it's not only that explanatorily it's basically just

55:12 saying cognition but it's more like the box ology that they've come up with is flaws right at the deeper level of critique I think Pease once you it's not just like working memory just means cognition and everything's fine we can change all the words it's more like the way the reason why working memory means cognition is because these structures don't actually they it's not that they don't exist we can clearly do the tasks that people want us to do but I don't think that we do the tasks that people want us to do because we have a dedicated unique and very specific and

55:46 specified kind of realize over this box ology in the brain I think it's because it just so happens the brains a kind of thing that can maintain and manipulate and we could combine it and those various structures and mechanisms that are maintaining and manipulating in the right way to perform a task I mean that's kind of what the brain does so it's not surprising to me that we can do these things it's just surprising to me when people say ah my operationalize task dependent version clearly unveils the central structure that is the thing that grounds thought and decision-making cognitions that know right because I

56:21 think we have to take a step back and realize no no no no no like the whole all these with all these mechanisms and systems and subsystems that are probably responsible for maintaining manipulating all the ways that we have thought of and haven't thought of so this is kind of where I want to flag activity silent stuff to talk about in a moment because I think that's been a big surprising thing philosophers have to pay attention to too when we see all these components there's a huge diversity and of ways that the brain might be able to perform

56:54 a task and it might not be the same between you and me and might be something that changes over the life span over enculturation practices and might be something that's actually much more socially embedded than we've expected before and so yeah it's not just like working memory to substitute for term for cognition and everything's fine it's like the Box ologies that these theories depend on in order to give the explanatory weight of how these systems interact a communicator in order to produce a phenomena at hand consciousness that ain't true right I say

57:29 so I skeptic in general of any box ology or any correct theory of how cognition works or not to get it go off footer again cuz you said you didn't want to but it sounds like you're I mean you're leaning towards a pretty hardcore kind of skepticism that no theory is gonna capture I don't know if it's an in principle skepticism I think it's more like a man I fallen off the camel and I've seen the light a little bit like really what things to the test it's like

58:01 these aren't better Factory and I'm actually I used to think that all box ologies were horrible I have this one slide where I show like the original 1970s box ology which has like the ayah a like a Han artistic drawing eye that's sticking out of the box and a hand that's sticking out of the box as well and I show the progression over 50 years and to the Box ologies of today where it's like oh my god fun AHA she's PFC version yeah it's real explain anything when on the craver bandwagon and be like

58:37 no it's not illogical different we need mechanisms and realizers and components but I recently I just saw and I were I can't remember her name to choose amazing she's a student of Michael Pollan's gave a really good talk in Berlin where she tried to show that the face recognition box ology could map on to some of the explanatory desiderata of mechanistic theories so I'm willing to like stop for a second be like maybe some will help and be worthwhile I think the current kinds of box ologies it's just like it's amazing I think the Coens one which is really spectacular is like

59:13 here's working memory and it shows this vague blob and this is like your entire repository of knowledge right there's a little circle outside the blob and it's the central executive and then it beans down onto the block of attention and that's working memory and I'm just like no I feel like this is an explanatory it's not the pessimism I just want to check ourselves and be like wait a second do we what do we want with a theory of cognition what exactly what

59:49 kind of concepts are crucial to it how might we actually given the way that the brain is not this von Neumann machine right it's not of this kind of it might not even be a Joycean machine it is it's a weird fricken kind of little machine learning thing with like all these different networks that were just beginning to crack open so we need to give ourselves a conceptual flexibility to be like wow there might be many components that are involved in thought obviously and they also might not be identical between people between creatures between all and there that's

1:00:24 okay there might be different ways to realize it from the components that we have I think coming up with a more flexible theory so I I remember seeing this beautiful set of studies by Curtis clay it's like his lab for the last five years including one by Judah and 2012 where they do a bunch of MVPA studies where they show that they do some very very very clever minimal paradigms as I like to call them where minimally they're just trying to get people to remember a location of a given object or to covertly attend to the previous

1:00:59 location of a given object overtly attend to the location that I get an object or make a plan to make a motor saccade between like a an eye movement between various objects right and what's very cool is like these things have different conceptualizations right so one's memory one's attention once covert attention one's motor preparation or motor intention as the psychologists call it with a T and what they found it's like oh it turns out if they train a classifier so if they train you know this algorithm very simple

1:01:34 algorithm on the results from one it can exactly predict the the kinds of dynamics that are going to happen in the regions recruited on all the other ones so it turns out there's just like you know the brain is flexible and it's not seeing these conceptual divisions right using very similar mechanisms to deal with all of these problems and then we need to even take a step back I think I'll let you stop in just a second bit you know there was this moment a long long time ago where I was talking to David jansch's a neuroscientist he was a UCLA does like really hardcore

1:02:07 electrophysiology on monkey so like hardcore stops right yeah and he was like working memory stupid I was like we mean and he's like well think about every single psychological study has you remember these cues and instructions in order to successfully perform the study so you're always doing maintenance and manipulation of active information in order to be successful in the task so how can they ever like ferret that out how can have they ever subtract it out of the thing if it's implicit and every freaking psychological study we have it's there even in studies where we're not looking for it so it's not surprising that we're

1:02:40 getting this like oh look there's this same structure that's popping up over and over and over it's like because part of the task demands of all these studies are the same so anyhow yeah I fallen off the camera I've seen the light I don't want to be Fodor I want us to pay attention to the dynamics that could exist realize that these terms need but really better cooperation between philosophy and psychologists actually come up with real empirically testable like this is a challenge I'm most like we need to come

1:03:12 up with ways like how can we actually test the means here or not what would that look like you know right no and that's important work so you've been talking a few times you've mentioned activity silence working memory and I want to make sure that we have a chance to discuss that because I think that's extremely important and exciting stuff I guess I have recently sort of found reason to be a little bit skeptical of that so I wonder what you think about some of this works coming out of it was actually buy it from dehance lab I think basically what they suggested what they

1:03:46 what they basically I think they found was that it's not I mean as soon as you try to use it it's not silent anymore so that's the idea that there's maintenance or manipulation in activity silent networks it's a little bit hard I mean it just pretty theoretically it seems kind of far fair so you can manipulate information without having any activity like that which is what activity silent is it seems crazy when you think about offers and we need to like me like what would Aristotle say you know this is unde agree because I think it is nuts a

1:04:20 word it's an oxymoron I've to be silent like it's like complete nuts so I don't think it's like that I think it's so for those who might not be aware I think Lewis peacock and especially people associated with Brad Postel but I think also you have Stokes and sprog more recently and neuron have shown that representation so if you're giving people you have to you have to ask people to maintain two modality distinct representations so it might be the orientations of a line and a work right and then they'll be asked to do two tests where they have to basically

1:04:55 compare it to a new stimuli and decide if it's sufficiently similar or different based on whatever rule so the slants might be like it has to you know they have to both still be acute and they can't be you know obtuse or whatever and the word like if it's think then it has to be ponder and you know it's sufficiently similar then it matches and you click the match button on that much it's so bad have you ever done one of these studies with these kids it's like it's how it's torture I mean you're like oh my god I'm making you do two hours so what happens is you get them you have to retain them then

1:05:32 you get these retro then you get these retro keys right so then it says basically a little arrow will come up and if it's at the top that means you have to remember and we're about to test you on the angle so let me test you on the angles and then another arrow will come up sometimes it'll be at the top sometimes as well if it's at the top you can test it on the angles again if it's at the bottom then you have to go back to the word bring it up and remember it now what they do is they track via EEG fMRI all these other methodologies that are standard in us and they're imaging

1:06:04 how the representation is treated of those two separate representations and what they find is that so after the first cue so you're cute on the lines that representation is is basically very high and the other one drops off so it goes back to baseline then simultaneously what happens is if if it's cute again you know you're successful in the performance but if the second one is cute on the one that you work that one the timid II drop down to baseline and how to explain how they do

1:06:37 a baseline later if the word is cute as a second Q then the activity comes back up and you're able to perform basically as if it was cued first and what I mean by baseline is when you get the second Q so let's say the thing you get them you get the angles cute and then the angles are cute again you basically forget the the word right and so that's the baseline that they're establishing it looks exactly like if they just had forgotten and in the end so somehow this representation which looks like if you

1:07:08 just discarded it comes back a lot and their explanation for this is that there are sub-threshold population dynamics that keep that representation alive and souks is the one who came up with word I can be silent and I think it's terrible it's more like Lois and peacock started with this notion of unattended and attended but I think that's bad too because attention we if you if you think working memories man man we had tensions worse attention is like that's like good luck I'd like Wayne who's version but like any and we can talk about that

1:07:41 later so yeah it's just like these sub-threshold population dynamics are what's keeping it alive what does that mean that means it is active it's just not something we can detect and basically what I think about this is I think two things one is that these sub-threshold dynamics may be things like calcium kinetics may be things like short-term changes in the synapse is and dendritic pruning things that happened the things that happen also that we see with more clear phenomenon things like long term potentiation and so forth the learning and memory consolidation it

1:08:14 turns out that those dynamics might be happening writing short term and that means that somehow this kind of working memory is we want to call it that this ability to maintain information is actually contiguous more with that kind of mechanism then the one that's is kind of the neuronal axon potential spiking model that we've had before that means there might be two routes to maintain information and I think that that's probably the best test case that shows us like wait maintenance isn't a single mechanism in the brain there might be many ways to keep a representation alive in the brain I think this the second

1:08:48 consequence that we have to be careful of is that it shows that there may be type 2 errors in neuroimaging which is like a duh but yeah we're pricing into our theories notions of activation because if we're pricing in to our theories these notions of activation and were referencing activation in this kind of methodological specific way then we might be ignoring the fact that turns up like the brain is off you know it's not this 10 percent metaphor it's all active all the time you know so that might be the wrong word to use when we're kind of

1:09:23 crafting these theories and also just because you didn't find activation in a given brain region doesn't mean it's not doing anything right I mean that's a touching tribute obvious but I think that's something that we need to pay attention about ya know it's it's that's a real danger that's a real danger today I'm almost getting distracted by you bringing up this 10% metaphor which I want to just scream about but let's let's let's ignore that so I mean the activity silent steps is to get back on this kind

1:09:57 of threat here is important for you because what you think it shows is that there's another model or a possible another model for how main this a manipulation could occur but it's more about maintenance rather than manipulation right and that's why I have to like at the beginning of the dissertation be like maintenance is a proper kind Amanda and it is it is clear by my guess and this is I don't have the neuroimaging to prove it but I'm sure there are much cleverer people than me that can that

1:10:29 when we figure out what these sub-threshold population dynamics are we will be able to see that like the kind of thing that happens in like memory like long term memory where things these kind of more morphological ways to maintain information a change over time that might be a kind of way to manipulate it I think anything that gets people on manipulation is something that was explicit in early models so this voluntary control right it was really it's part of it was the Colin Terry control and the seat of consciousness

1:11:01 both those things come up and up and up in the Atkinson and Shiffrin model the association with consciousness has continued to the present but voluntariness dropped out yeah that's a pretty obvious reason because we're trying to find the explanation for von to like agency and all these is a very big and we don't want to place them in at that level so I think that's the manipulation now is kind of a cold phrase for like this kind of voluntary you know I'm really doing it myself activity and I think that I we need to challenge ourselves to think about like

1:11:34 okay what would a real well how do we really cast that out in the testable way like it's it's tough I don't have the solution because I mean that would basically be having a solution for how we find agency in the brain which is I think that's come I think we're a little away from so I just want to be able to have a flexibility to say like look if these morphological mechanisms can create if what we mean by maintenance is like changing of a representation in terms of its content then look these mechanisms do it trivially insofar as

1:12:08 they do it when we like misremember things down the road and Philippe Dibrugarh has a lot of really cool stuff on how remembering is actually a prescriptive future for future looking phenomena and I think that's that could be kind of manipulation too so I'm not super worried about in a way yeah that's interesting I well let me let me press on because I noticed where I don't want to keep you here forever I know you have another thing that's coming I can talk to you forever I know it's a really good conversation but so I'm wondering well

1:12:46 I'm getting bogged down by something that we were just talking about but I wonder so finally good moving on to what some issues about reportable oh man that's a whole nother can of worms I know so he you you you you want to really denied or we put pressure on the link between reportability and consciousness and I'm wondering what the motivation for this is so it's not so subjective report is often touted as the

1:13:21 most or the best or the only or one of those select your favorite one we check on what's going on with the consciousness of a subject you deny that is that right or what is exactly your beef with reportability so I guess man this is like late in the dissertation and I was kind of seeing this word come up and up and up again as it's kind of a you know it's a kind of evidentiary gold standard for first-order theories of consciousness is that they should be

1:13:54 able to access consciousness requires that we can report it you know this is kind of the minimal block eonline amazing yeah and so the question then becomes like okay well how do you actually devise a test report and it turns out when you think about that and I think Lama makes it very explicit and like a 2007 trans review paper when he's like this is crazy because depending on what you count as report you're gonna price you're gonna build in your model all these different higher cognitions

1:14:26 and it can really be I think at one point he says by that logic you know the whole any part of the brain or probably the entire part at all different parts of the brain of all you've been implicated in consciousness depending on your realize their report and I kind of want it to be like men is this a single thing like what unites reportability as like an actual not quite natural kind but just like conceptual kind turn right and cuz like people every time my like mo right now is like every time people use a concept I kind of want to do the Josh nope thing and be like wait what are the intuitions

1:14:59 behind it not necessarily go and like review and interview people but be like are the intuitions and the pre and the pre theoretical assumptions that we have baked into this concept are they good and I think with report we've been relying on this kind of like we'll die it's just like behavior but but hey here is a very very thing and so I was very surprised to look at different cases where for example and Hydra encephalitic children hydro and encephalitic children so and children and some other people

1:15:35 who have suffered massive brain damage who are to our ticket to be able to report about conscious experiences and I think that when we look at those cases that actually put some pressure on the only thing that I think unifies report it into a conceptual kind so I kind of do this little Jannetty and thing round like is report unified by being and its neural basis well no because trivially depending on what you count as a report it's gonna it's gonna implicate different regions is it unified in a kind of normative way it's like yes but

1:16:10 it turns out that you know people are willing to say we can be wrong about what we report because that seems like something you need like people can be mistaken that's right something even these kinds of intentional concepts and then I'm like well I think the only thing that really unifies report as a concept that's useful in the first for sort of theories of consciousness is it it's really closely tied to consciousness itself it's like that it we hook it up to seven implicitly to subjective experience in a really tight

1:16:43 way so what I mean by that is even if we're wrong about what we report and this is the Hannah and that and that and that I can't pronounce this lesson no we may be wrong about mulberry report but even that report as they say it's not a cut and copy or cut and paste scene a cut and paste copy of a visual scene it is itself a conscious comment it's a comment right it's not just comment so even if we're wrong about when we're reporting that wrong report is still itself it's conscious and so that's why you can tie it to consciousness because

1:17:19 report is just right so it's this kind of deep deep deep connection that I think is at the concept of rapport ability and why it's often used in for shorter theories so what's interesting then is if you look at the cases on the periphery so I remember in February that came out this case about giant magnet so this is a young black woman that was going in for routine tonsillectomy and it just it's a really tragic story that deals a lot with like inequality and medical

1:17:51 services and like embedded racism in like the medical practice but basically she never came out of the surgery she had to be put on a ventilator after a few days she was declared dead and her family think or convinced that she's still alive and a petition and it's it's a dramatically heartbreaking story but they actually brought in a lot of these neurologists who had been more skeptical of the previous orthodoxy of brain death and they do these kind of mini studies where they show like oh our heart rate increases when she hears your mom's

1:18:26 voices of us other voices she actually started showing signs of electrical activity after enough time it wasn't like her brain was a soup she just didn't have the diencephalon she didn't have like her brainstem and thalamus and these lower base structures that we think are very important for information processing she could signal like her mom would ask you what's the [ __ ] you thing and she would like kind of flick her middle finger and the way they recorded this enough times and people are like maybe there's like this is a conscious comment now okay you might be skeptical but the hannah is willing to say that

1:19:02 the vegetative state patients that have been interviewed in Owen and other earlier studies where these vegetative locked-in patients they asked them imagine playing tennis is that right they're playing tennis are going through your room and in fact in I think there's a monty study it was a follow up where they do this with 52 patients identified the ones that can answer yes or no questions by imagining playing tennis or going through the room and then a few of them I think very few of them like five were able to answer autobiographical questions at a high rate of accuracy like 80%

1:19:34 higher and those people actually managed to recover so here it is that we're taking they're not even saying anything they're not doing they're not flicking nothing it's like they're just we see some brain blobs on the fMRI sling and we're taking that as a conscious comment on their mental representation now saying if we're willing to do that we need to be more sensitive about people who might be plucking their finger so with joy cool now to finalize it compare it to a case of hydro encephalo with a child basically due to many factors it's really hard or horrifically tragic but

1:20:07 basically their cerebral cortex of the neocortex fails to develop where they're left basically a brain stem maybe some diencephalon maybe some colonists so all the things that global the frontal parietal networks working memory all this stuff is all linear cortex primarily and they don't have they don't have that they don't have most of the limbic system at all right yeah a lot of these children can display affective responses they can light up they have favorite toys if you give them special because they have severe motoric

1:20:39 limitation if you give them special switches they can call out for their favorite toy they can do they can respond and effective ways and anon and a child without that conditions we would count as possible rapport like States and so now we have to be like man if we need to dissociate what we count as report from its strong implication to consciousness because otherwise we can't make sense of these cases without drastically reframing our theories so so yeah and I agree with all that I wonder what you would say about someone who

1:21:11 wanted to restrict the kind of report that was privileged to verbal report so what clearly we have to I mean I I think that there are non-verbal reports and these sort of things were talking about account as that example of them but I would be skeptical that those things are linked to consciousness automatically without a lot of further consideration but certainly verbal reports should have a special place if you say if someone says do you see and you say yes I see it then that should be an indication that you're consciously experiencing or

1:21:44 subjectively aware of it or something like that so the real skepticism it seems to me on your on your part comes from the fact that a lot of these theorists wanna count these nonverbal methods as indication of access or something and then want to exclude other versions of it is that the real kind of underlying or is it even subject a verbal report that you're worried I don't know I mean I think that that would be probably operation lame why is at the moment I think there's two things that I want to push again so one is that so so the problem is if we just include verbal

1:22:21 report I mean that means that like people would that count like people pressing buttons and responses to screens like No okay so that basically it's like okay hands look right I like you know that's like okay we got to start over yo I mean that would be I think a big bullet for these guys to bite maybe it's the right bullet to go forward if this is actually part of their methodology I'd have to think a little bit more about that I want to give them a bit more options I think the other way to go about it is realizing that report can't is evidence but it can't be the only kind of

1:22:55 evidence we need to have theories of cannot just cognition but we need to be able to be like what like like maybe this is kind of I know llama is like everybody makes fun of it but at least he's putting a bunny a boy he's saying recurrent processing so if we see report and recurrent processing chances are we might be conscious and that's of course pretty crazy because it could get it all over the place and when you're sleeping in all these ways maybe you're conscious when you're sleepy I don't know there we need a kind of conceptual flexibility even that level and I think another thing that motivates that is yes the

1:23:29 search for this capital P phenomenal variant of consciousness and its neural constituents is important but we should just limit ourselves to that I think when we see children like that it there might be a hole we need a kind of conceptual humility as well because there might be a lot of different kinds of experiences ranges of experiences that by count in some way is conscious and that have different instantiations in the brain have different behavioral realizers and so what I think to be a little more cute we need to have

1:24:04 a little more humility before we bite off some of this stuff and I think it's a whole package having the conceptual flexibility and the humility kind of lets you able to think about okay how just why we want to count some nonverbal reports as reports and and what might be some a priori ways that we might theorize about what we ought to see and then oh we didn't see that okay so let's move on you know and so yeah I agree but I mean I the same time I would be really cautious about wanting to build theories based on you know an encephalic children having you know smiling when they see

1:24:38 their mother or picking special Tory I don't want to you know I want to discount that date I want to be open like you're saying to these possibilities but it seems to me that if we're trying to build a theory we should started with the easy cases or the good cases that the non-controversial wants and those student me to be based on verbal report so I think a lot of consciousness science should be sort of more beholdin to verbal report as the gold standard at least for checking you know these sorts of things and then we can extend out from there so that's a more conservative methodology I think

1:25:11 but you know I mean just to bring up one kind of example you know binocular rivalry when I was a student you know that was the rage Frank Tong and his work was taking over the world and everyone's like this is all you study consciousness you do in binocular rivalry and then it seems that we showed that well you could have rivalry even unconsciously when the subjects don't report that they're aware of the switches and then you find people saying well now here's an objective measure we don't look for a report anymore we look for you know I movements and cicadas and heart rates and stuff like that and then

1:25:43 those correlate with the switch of the percept and now we have an objective measure that's not based on verbal report about what the person is subjectively experiencing with the houseless base but if you really believe that there's unconscious rivalry then the objective measure doesn't answer the question whether it's conscious now you still need the subject saying I see I see a house or I see a face in order to really know what the data you're looking at is telling you whether something that's conscious there or not so I just don't for me I don't really see how we're ever gonna get away from

1:26:17 verbal report being the the gold standard and these other things that are called report I'm happy to call them nonverbal reports but you know taking blank type Wednesday patients can press a button to indicate that a thing is there but they can verbally report that it's there but when you are ideal we can we can make the verbal report so it just seems that that the more reliable that the non skeptical way to go forward is to take verbal report more seriously I think maybe III so I'm hesitant to say absolutely yes but I think the my my

1:26:52 worry so my worry would be that how much of what we're detecting is actually just the mechanisms involved in verbal report so that's kind of what makes me a little nervous because that's not a trivial process right so like actually forming a sentence you know figuring out how those all the language parsing that goes into it preparing a motoric response waiting and like inhibiting yourself before you need to say it saying it at the right time dealing with timing then actually

1:27:25 producing it requires like a good chunk of brain structure right and maybe we can like isolate it out but it seems to be the case we would need to isolate verbal reports out we would need that basically create a parrot of somebody and just make them say it right really subtract and this this becomes them I think also quite difficult because if we're if consciousness has been more elusive and it's a bit more about how things are connected throughout the brain then we're those things are gonna fade out and we're just gonna get why

1:27:56 look we have this structure but that's structure I'd be more skeptical of saying that's a consciousness structure I'd be like that's all this machinery that's associated with all these complicated things this verbal report is complicated and I think we can't be trivial about that's not something easy it's even understanding the mechanisms that are involved in that is not we're not there yet right maybe but we have to be conscientious about those worries and then I think like the blindside think it's interesting cuz I think the more I think about you know and passing his work on the induced blindsight it's like maybe that induce

1:28:31 kind of blindsight mechanism plus kind of finding a way to figure out the you know how to subtract out the the verbal report and what might be associated with it because then we can phase people in and out of the space where they can verbally report and then we can phase them in and out of if we can do the the same thing for we're rule report itself then then we might be able to isolate kind of a real candidate I think so that's something I think I haven't thought about that I think would be pretty cool and I would love to hear how fun would probably be like no that's not

1:29:03 possible yeah that's a different story I'm noticing we're creeping up on on time here so I want to there's many things we could do well on but I want to want to sort of stop being pessimistic and and negative so we as a lot of the and a lot of you and you're pretty open about this a lot of your work is sort of clearing away this kind of you know the I don't know the confidence and the ease with which people invoke these kinds of

1:29:35 things without really doing no conceptual work that I think your work carefully does so and that's important work and we don't want to dismiss it but at the same time I'm curious about what you're more positive views are so where what are you excited about in terms of something which might be a path towards figuring out a theory of consciousness if it's not this sort of stuff I yeah I don't know I'm so kind of working through that because I think part of the anxiety of coming from a science background is that you know you're like

1:30:08 oh I'm in philosophy mode I can only do the thing I really know and and and creating positive projects and positive views is actually it's much harder to do confidently and that's something that takes some time but I'm kind of working through what I might want it to be like I think it's easier for me I think the consciousness but I need to I need to be humble myself and realize that like we need to I need to learn more about higher order theories associated theories I think kind of getting more grasp on these things and maybe having that

1:30:40 influence my conceptual humility and really kind of considering even cases like this like having people push back be like verbal report matters what might that look like what might a real test for River report and the mechanisms associated that look like but I think apart from that because I to me consciousness was a test case where working memory was doing this black box II work and I wanted to disabuse people of that and I think I don't know if I was successful but it was a good test case he worked for me now I want to be able to say okay what are the other core

1:31:15 fundamental properties of the mind that we associate with this kind of central workspace metaphor and realize that like a lot rides on that and you know you have to go a little bit more away from the science of people like I'm thinking about McDowell and Brandon and that's pretty far away from the Sun right yeah it's weird like I started my clothes off of career reading random and no it's coming back people we're talking about rationality is kind of existing in a particular way in requiring a workspace of ideas in order for us to really be

1:31:49 rational agents we need to and even if you look at really good people like Phillip paralysis work on on this kind of esoteric theory of and that's not the right word the questioning theory of attention and everything you know it's just a supposition that like oh there's some mechanism in the mind that allows us to consult all our background beliefs at once in order to pick them out put them in the blackboard and that's what rationality is like everyone has that we all have the same kind of mechanism that's why we can trust each other to be rational and I think going more in this

1:32:22 direction to be like wait no that metaphor is the same one that's gotten us in trouble with cognition and consciousness maybe I can we need to like revise those deeper thicker epistemic concepts of rationality of what deliberation is of what agency is these things that depend on their being of their their in us that is this controlling central workspace that allows us to consult things and is like the core of rationality and agency in this car

1:32:53 a tional Center in the mind sorry I'm talking like Jessie but you know and start critiquing like what would deliberation look like if we didn't maintain that people needed to have a specific works facing the mind to get it done is it this case and I think this is probably the the truth is that we do have work spaces where we can put things together but they're ones that we create either with ourselves over time and this is maybe through kind of scaffold and embedded in the body ways so like writing stuff down or that we create amongst each other in time like really in a deliberation a deliberation is not

1:33:27 this like pure thing that happens maybe some people can do it but mostly it happens when you're arguing with somebody in a room it's a yes so it's you phenomena and so I think really kind of relocating a lot of those core concepts of rationality agency and thought and cognition in the more social space that respects in fact there were social creatures so the liberation is probably something that happens not because we have a work space in our mind but previews we create work spaces where thoughts from our mind can go and be in a public represented space or you know even by talking to ourselves you know

1:34:02 and the kind of a got skin way but also things like rationality why do we have all these crises and failures of rationality and dual process theory it's like it turns out that maybe it's a case that we're making too much into the individual and what the individual should do and so that's kind of the direction I really feel like going I do want to continue with the I want to throw my stuff out into the consciousness people get a sense of like maybe we can create new ways to kind of parse this language and create new box

1:34:35 ologies that are better testable and give scientists the tools that they need conceptually to like look for these things but I think my goal on the next like 10 20 years ago for for the big guns you know my so it sounds like a lot of you I mean way too time wasn't the whole assumption methodological solecism that we can learn about the mind by studying an individual in isolation in the lab what we need to expand that to I mean my colleague LaGuardia would be very happy they mention Bhagat scheme

1:35:08 which you know no one I guess I'd graduated actually dorsum Bhagat Singh's but the idea that we're social creatures I think is really important and overlooked in a lot of theorizing so I think that's very interesting but you think you're gonna get a theory of rationality and consciousness from looking at it us embedded in our social environments or just that we're gonna be able to add some elements to existing architecture yeah I think it's about actually reframing some of the vocabulary that we've used that not to

1:35:42 get into the super intense phrase a way to phrase it but that we've been you know kind of using talking about rationality since I can't and I mean I think it really is kind of revising that kind of methodological solecism of rationale that's implicit in the notion of rationality is like we're all a little rationally don't attack the Enlightenment man sorry what's the same Jonathan Oh God don't mention Joe to Peter said goodbye good luck to [ __ ] expect to be like I'm gonna come up with him John Condy theory rationally I'm

1:36:16 actually really excited working with Lisa cuz she's an epistemology she's actually we're on board and a lot of these things we have a lot of similar ideas about how things are wrong and so I think I'm not the one who needs to create a framework for that but relying on other people who have similar skepticisms and worries but also wanna be scientists about it and I think there is a lot of promise it's just about freeing ourselves from these kinds of preconditions that we may have set without realizing about how the mind works know how the mind ought to work how how mental ontology should look like

1:36:49 just because we have as philosophers this desire to see there being this rational checking thing in the Met and the brain maybe there really is it maybe that's something that happens a discursive look and that's okay it's like so not to open up a thirtieth thousand can of worms here but it's not like there's a lot of skepticism about folk psychological notions behind what you're saying so yeah is is that accurate to say that you're gonna so that you probably tend towards a limited ISM about folk psychology I mean I don't know not in the church limb

1:37:21 waiver we're like not talking about I believe there's coffee in the fridge and replacing that with like st. P 42 like no obviously not I think it's not so much I'm in a limited this is that I just want to critique it when it comes into our theory building about rationality about deliberation about thought and I think that's an okay perspective too because it's already our folk psychology constructs are already filtered through so much and they're filtered through I don't think they're

1:37:55 universal I don't think they some of them are like belief and desire probably but a lot of the way we might think about thought rationality deliberation decision-making those things are in cultured those things change over time and I think you can see that in the progression of progression of philosophy over the last since Aristotle basically and I think that we need to realize that sometimes we've imported weird stuff into that and we need to be willing to free ourselves from that not just to create a more a ballistic let's say ballistic kind of structure of like what

1:38:29 cognition or consciousness might look like but also I don't know I think I'm not motivated a lot by claims of human uniqueness so like hi every time I sense something there I get pretty skeptical and I think a lot of these things that were talking about thought rationality agency these have a lot of that human uniqueness built into them and in a way that makes me uncomfortable and makes me want to be critical of those concepts because I don't think we've proven like without begging the question these these facts necessarily like yes we're

1:39:02 uniquely like terrible I think yes no we're social creatures were weird we do weird things deliberation and I you know like Tad's list he has this whole theory of I think these are not new ideas I just wanna reference that I think what's dangerous is by importing certain psychological constructs and using them substituting them in to our philosophical theorizing just because they're hot and in psychology not going to solve these bigger problems that lie underneath the surface and we've been dealing with for years so

1:39:35 yeah yeah okay or there's a lot I want to ask you talking to you but I can't we gotta end at some point so I'm wondering is there anything that hasn't come up what you want to talk about or mention before we close and sign off and the answer can be no by no I mean I think it's been a lot of fun I mean to me it's just I want people because I find that people when I present this especially scientists get like they're like what like how is this what is this guy talking about what is he all about how can they believe that there's no working

1:40:07 memory and that it's like of course we can think that I want people to take a moment and pause and if they do feel that way if they do because I think the longer the hardest things for me to do was to I felt like this would never get taken seriously by psychologists or philosophers and part piece philosophers be like that psychology and psychologists would be like that's not that's ridiculous you're not giving me an experiment and I think it wasn't until tonight was terribly difficult to get published and that's kind of stuff until Carothers got his book out and then he kind of whet people's appetite and made people realizes there were full

1:40:39 of awful things to discuss here I think so if people have that reaction that they're like this is crazy this is enough I just want them to and you know not to use ableist language but I just want them to take a step back and be like wait what what am I really committed to what rides on this my view of our cognition being right and if they explain that and I mean I'd love to hear it because I think that's there's something in there that is priced in to the commitment that we have to these views in the first place okay with crazy people like me we're gonna say no yes

1:41:14 exactly I think that's exactly right they should pay more attention I mean you know if there's an argument there it should be addressed that's always been my view whether you agree with it or not whether you think it's beyond the pale you know a lot of this had to do with the data so I think a lot of this is going to be played out with a book on data we have but also designing better experiments and being more open to new kinds of approaches it's amazing how all kind of dogmatic scientists and science itself can be so I think it's good to have people like you out there doing what you're doing

1:41:46 certainly going to be looking for your future publish work I think it's gonna we're gonna see great things from you so I just want to thank you for joining me today this is a great conversation it's been a lot of fun and I hope I'll talk to you again soon definitely Richard thing yeah so invitation no my pleasure so hang on just like I'll talk to you afterwards briefly but sure okay