BI 246 Andrea Gambarotto: Cognition Requires Agency

[00:00:03] Speaker A: If you ask me, what is at stake for me was essentially an explanation of what is mind and how it relates to life. So for me, the main point is the relation between mind and life and the weird continuity, but also discontinuities that you have on this spectrum. Autopoiesis is the key is the fundamental requirement for cognition, but it doesn't give you cognition for free. Are cognition and intelligence the same thing? No, they're not. Intelligence is a special kind of cognition that has been going through historical and intersubjective processes of sedimentation and sedimented participatory sense making. If you ask a philosopher of biology, teleology is not a problem because, yeah, we have Darwin. Darwin naturalized teleology. Teleology equals design. So yeah, the story is over. Kant's distinction in the Critique of the Power of Judgment between intrinsic and extrinsic purposiveness pushes us to think, well, is the story really over? [00:01:24] Speaker B: This is Brain Inspired. Powered by the transmitter. Hey everyone, I'm Paul Middlebrooks. Welcome to Brain Inspired. Andrea Gambarotto is a postdoctoral philosopher and researcher at the University of Luxembourg. He is an expert in the philosophy of Jorg Wilhelm Friedrich Hegel, most commonly referred to as Hegel. More recently, he has been studying the relation between some of Hegel's ideas and those of modern theoretical biology regarding questions of autonomy and agency. Andrea argues that whereas Immanuel Kant believed we should explain biological stuff and inanimate stuff the same way via mechanistic explanation, Hegel believed to explain the biological stuff, we should leverage the fact that biological organisms have intrinsic purpose, I. E. Agency. And Hegel's approach is in line with what's called the inactive approach in cognitive science, which now has a fairly long history and continues to thrive. Andrea explains all of that during our discussion. One reason I invited him is because these issues get at the heart of what some of us care about, which is what are the differences and similarities between our natural intelligence and engineered artificial intelligence? Why should we care about those differences? A large language model isn't alive, but does it have a mind? Should we call what it does cognition? What are the relations between life, mind, cognition, intelligence, consciousness, those kinds of questions. While we even discuss why the famed octopus might be really intelligent but not conscious. Anyway, hold on tight, because Andrea definitely set a brain inspired record here for the most references cited in an episode ever, by far. I believe he's quite the scholar and I quite enjoyed our discussion. I hope you do too. You can get the full episode like you can get all Full episodes on patreon. Go to BrainInspired Co to find out how to become a patron to support the podcast. Thank you to my Patreon supporters. Thanks always to the transmitter. Here's Andrea. I'm going to ask you about your background in a second, but just to sort of frame the entirety of our conversation, or at least one of the main topics that we're going to talk about, which is Hegel and enactivism. So is it fair to say that the central premise of that is that we can use intrinsic purposeiveness, intrinsic purpose and agency as something to explain cognition rather than something to be explained and. And a particular kind of 4E cognition that is in activism. One of the 4E's is in line with Hegel's account of intrinsic purpose is. Is that it? In a nutshell? What would you correct there? [00:04:44] Speaker A: I'm not sure I would correct anything. I think it's pretty correct. So, yeah, my background is. So I come to 4 ecognition from a rather improbable direction because I was trained as what they call now, a continental philosopher. So I was trained in Italy. And normally people that are being trained in Germany, Italy and France and Spain. Spain. Spain is pushing right now towards more different model. But in general, people that work on the continent, they used to study history of philosophy. You study the classics, you study French and German philosophy mostly, and you're mostly engaged with understanding what this usually huge tomes say. And most of the time, not just [00:05:41] Speaker B: huge tomes, but difficult tomes, right? [00:05:43] Speaker A: Yes, huge and difficult. I mean, you kind of understand. Understand why there is a whole business behind it, because there's a need to understand and discuss what these people were saying. But then the risk of that operation is that, at least to the way I felt when I was doing that, is to be. To remain close within historical background and within the text, within the language, which, for example, I was hanging out at conferences of people that studied Hegel and those people speak a particular language, which is Hegel is. And only those people understand that language in a sense. So it's interesting for me to. I mean, if you think that particular stuff has some interest, then to kind of translate that into contemporary ground in a sense. Right now, for example, I'm teaching. I teach at the University of Luxembourg and I'm teaching a master course. I'm teaching the Heidegger module. And my, My. My course is called From Heidegger to Cognitive Science because of course, I'm teaching Being in Time. And I'm trying to then teach them first of all, teach them how to read this stuff because it's really, really weird, in a sense, and then how to use that material to apply to contemporary 4e questions, essentially. And so in my training, so my specific training in Padova was mostly focused on classical German philosophy. So Kant and Hegel, which was the main research and teaching topic. And my, my supervisor in Padova, Luka Letterati, was, is a big expert of Hegel's philosophy of nature and his. And of the concept of teleology and on the concept of organism in 18th and 19th century. And so that's where I started from. My dissertation was an incredibly, we could say boring dissertation about the rise of biology in Germany between the 18th and 19th century. And the thesis there was that the concept of teleology as intrinsic purposiveness, because Kant has this conceptual distinction between so intrinsic or internal purposiveness and which is external purposiveness. And the external purpose is what we, when we talk about teleology, we are used to think about which is design. Also evolutionary biology. Teleology is mostly thought through the lens of design. And Darwin is very important because he contributed to naturalized design. So we don't have to think about organisms as designed by God, but designed by a blind watchmaker, which is natural selection, if you will. And so in a sense, if you ask a philosopher of biology, like Michael Ruse, or like the mainstream people who actually grounded philosophy of biology as a field, for them, teleology is not a problem because, yeah, we have Darwin, Darwin, naturalized teleology, teleology equals design. So, yeah, the story is over. And so Kant's distinction in the Critique of the Power of Judgment between intrinsic and extrinsic purposiveness pushes us to think, well, is the story really over, or do we need to think about that a little bit more? And when I was starting my PhD, my postdoc project here in Belgium, my project was called Hegel's Theology of Biology. And the idea was, well, so essentially the form Hegel, what Hegel proposes is a form of naturalization of Kant's teleology. Because Kant is famously problematic for him, he sees this concept of intrinsic purposefulness as something that tells us something important about organisms. But he's also too afraid to really go wrong with the intuition and says, no, no, this is just a regulative principle with just the way we make sense of things, but it's not really a property of the furniture of the world, so to speak. [00:10:29] Speaker B: And he was, he was bought into mechanistic explanation as the explanation is that. Is that why? [00:10:35] Speaker A: Yeah, yes, yes. For him he was a Newtonian, hardcore Newtonian, and he was committed to the idea that yes, mechanics is the way to go if you want to do serious science. And then. But then you have some natural entities, organisms or as Kant called them, organized beings that do not respond very well to this. And for him was. And this is the famous what. What is called famous as the antinomy of teleological judgment. So on one hand everything in nature must be explained mechanistically. On the other hand you have some stuff that somehow doesn't seem to fit that program. And Kant leaves it there, says yeah, this is the way we make sense of things. And then we have to make sense of stuff mechanistically. [00:11:29] Speaker B: So from Kant he was basically saying I'm going to let future thinkers and scientists figure this out because we don't have the right tools. Or why did he leave it there? I mean just a paradox for him. [00:11:45] Speaker A: Or I think that there depends on where who you ask. And if you ask a Kantian, which I'm not, they will probably have very elaborate stories about how that fits within the system and they're going to try to defend that position as as legitimate and that they're going to be very sophisticated explanation or stories as to why he did that. Me personally, I'm a Hegelian. And so what Hegel says, and what I agree on is that yeah, it's just. He just didn't trust himself to jump towards uncharted territory which would imply us to question the all explanatory power of mechanism. So Hegel is very interesting in the sense of logic, has this, in the teleology section he has this very strong appreciation of Kant. It says this idea of intrinsic proposiveness is one of the greatest contribution Kant made to metaphysics. [00:12:51] Speaker B: But then he says here and I'll fix it. [00:12:55] Speaker A: Yes, you kind of stopped mid midway and it didn't go all the way. And this. [00:13:01] Speaker B: Yeah. [00:13:01] Speaker A: And he purportedly is the one who fixes the the problem. And he says very famous and also controversial or let's say infamous statement, according to which teleology is the truth of mechanism. So teleology is an explanatory framework that is in a sense higher of that mechanism because it's what allows us to explain Kaist. So mind and life and mind must be explained through a principle of intrinsic purposiveness, which should be understood as a property of organisms. [00:13:38] Speaker B: You said higher. Does that mean more? That's different than more fundamental. When you. So as from a reductionist background, when you think fundamental, you think lower. Right? But. Yes, so is. But higher could be fundamental. If he means that teleology is actually more fundamental than mechanistic stuff. [00:13:57] Speaker A: Yes. That's. That's kind of like the. The. Because in the way Hegel thinks everything is upside down. So you start from. From the more basic and in a sense wrong stuff and then you move on to the more complicated, complex and. But also truer stuff. Because it's. Because at the end you. The true. You get only at the end of the story, essentially for Hegel. Hegel works by elimination and reintegration of not exactly correct positions. We could say. [00:14:30] Speaker B: Yeah, okay, so. Yeah, sorry I interrupted you to. To ground my own understanding there. Yeah, so. Okay, well, I was going. Because we started by talking about your background. So I was going to ask you because you had this continental philosophy background and even in your super boring dissertation, as you put it, you were interested in Hegel's account of Hegel plus biology, essentially. Right. And is that at some point you met Matteo Mosio maybe at a conference or something, but I guess you had been. So your interest in biology was there from the beginning? [00:15:12] Speaker A: Yes, so I met Matteo. I was doing my PhD. My PhD in Paris and it's called Institut Histoire des Philosophies des Sciences Technique. And my story was again focused on the concept of teleology. And the thesis that I was defending in an historical argument was that, well, biology as a science in Germany was into a large part a result of a conceptual shift that happened between Kant on the one hand and the post Kantians. So both these gooding and people and the philosophers of nature. And for them teleology was to be understood as a constitutive feature of biological systems. And this is what differentiates a biological object from a physical object. And what justifies the label of biology as a science of living beings, which emerges and is formulated as such only in 1802. It was not there before. So was working on teleology in this classical German ground. And I met Mattel Moschio at this [00:16:24] Speaker B: point it's still purely from a sort of philosophical sense. Right. It's not like you're interested in cognition per se or. Or autonomy on its own, or explaining understanding that it's really like understanding the position of Hegel in this case and. And how different people thought of it and how to reorient things. [00:16:45] Speaker A: Yes, it was. Yeah. More than philosophical almost, you could say historical. Yeah, it was. I was Doing history of science in a sense and history of philosophy or as they call it today, HPS History and Philosophy of Science. So I was trying to do both. But. So this is the point where I, Mattel Moscow was working at the same institute. So I was working on one room and he was in the PhD student room and he was working in the, in the permanent people room. And we met over coffee and we were discussing and I was reading Varela already because I was interested a little bit. And he was working already. He was starting to work on the concept of teleology too, but in this completely different concept. And he ended up writing a paper that was published online in 2014 with Leonardo Bick called what makes Biological Organization Teleological. And, and it was striking to me the similarity between what these people were saying and what Hegel and Schelling were saying 200 years earlier. [00:17:58] Speaker B: And let me, let me just kind of summarize their position and then you can correct me maybe, because this is sort of how I came to you is because I read Moreno Mosio's Principles of Biological Autonomy or that book right where. And so. But that was 2015 after this paper that you're talking about. But the idea which has origins, well, I don't want to say origins, but refers back to people like Stuart Kaufman, Robert Rosen, et cetera. The idea is that autonomy can be thought of or is defined by almost what's called closure of constraints, which basically just means that although organisms are open ended systems thermodynamically far from equilibrium, they have to take in energy and release energy. They are wrapped up in themselves in that they produce their own constraints, so their processes and the materials that constitute the organism. The organism is producing its own constraints, its own enzymes which act on itself, for example, its own lipids which contain the intracellular milieu which is necessary to keep at a certain ph level and concentration of ions, et cetera, to allow all of these enzymatic reactions to happen and so on and so on. And so that there's this like sort of co. Determination of all the parts which co. Constrain each other. And that would be a whole and that would be like sort of the definition almost of, of an organism. I don't correct what I, what I just described. There's way more to it. I was trying to do it in a sort of pithy manner. [00:19:53] Speaker A: No, it's, it's very accurate. And you, you named the, the, the main people. So Principle of Biological Autonomy is from Varela 1979, just got republished by MIT Press with a New Apparatus, edited by Ezekiel Di Paolo and Ivan Thompson. Stuart Kaufman's Investigation. And even before that, the Origins of order from 1993 was Origins of Order were the first big book by Kaufman. And investigation in 2000, is the. The book that really starts this tradition that gets referenced most of the time. And of course, Robert Rosen, Life Itself and all the stuff from a bit earlier on, Life itself is from 1991, and other papers are from the 80s. What all these people have in common to connect this to what we were discussing about five minutes ago, is a fundamental reference to Kant, which is more explicit in Kaufmann, Kaufmann in Investigation, when he describes the kind of regime that you were talking about in which parts core constrain each other and essentially maintain each other in a way that, as Rosen will say, organisms are close to efficient causation because they generate a form of causation that sustain itself. And Kaufman calls this Kantian holes in investigations. And again, the reference here is Kant's third Critique, the Critique of the Power of Judgment, Most importantly, paragraph 65, in which Kant defines this. And Kant has this definition of an organized being as a system in which is cause and effect of itself. So the parts generate the whole, and the whole contributes to generate the part. And this is also, again, to reconnect what we just did, the historical background. We were just discussing a paper that for me was really the turning point, together with the paper by Matteo Monsio and Leo Bick on what makes biological organization teleological, is a paper that was published actually long before that, in 2002. And it was one of the last paper by Francisco Varela, which got published posthumously because Varela died of cancer in 2001. And the paper was published in Phenomenology and the Cognitive Sciences in 2002. And the title of the paper is Life After Kant. And then the subtitle is the Autopoietic Found Natural Purposes in there. [00:22:43] Speaker B: Yeah. [00:22:44] Speaker A: Yes, yeah, of course. Although, as we might discuss later, autopoiesis is not very much used these days anymore. And this is really, for me, what made me jump from the historical work that characterized my training and my PhD broadly to the kind of work that I'm doing today was this. This paper was really criticizing Kant for taking what they call an unstable middle position, which is exactly what Hegel was accusing Kant of 200 years earlier. And they say, well, Kant had this very good idea about teleology, about organisms being cause and effects of themselves, but then he couldn't bring the point home. So to Speak and what we can do, we have a theory about the poiesis and for them in that paper, Hans Jonas, the Biophilosophy by Hans Jonas, who was a phenomenologist and a student of Heidegger, who essentially had this idea that this very famous quote that is used in this literature, that organisms are what they do. There is this very important paper by Jonas in 1968 called the Biological Foundations of Individuality, which is really already putting this idea forward. Of course he was a philosopher, he was not a theoretical biologist. So this is only conceptual. But these people really take this, let's say post Kantian idea of teleology and they try to formalize it in the form of constraints and closure. And then of course what for me is a great achievement and also for Kaufmann. Kaufman cited that paper in his 20002021 book as a revolution. The paper by Matteo Moscio and Mael Monteville and Matteo Moscio 2015 in the Journal of Theoret Biology entitled Biological Organization as closure of Constraints, in which they provide what is to, to this day the most rigorous formalization of that idea. [00:24:55] Speaker B: Yeah, well, okay, so let me back up and just say explicitly what's at stake for me here, because I'm a neuroscientist. Yeah, I'm a neuroscientist. That's it. That's kind of a broad term. But so I became interested in these ideas of organizational closure, closure of constraints that we've been discussing here that you just cited, the Moreno Mossio 2015 paper. I'm interested in how principles like this can inspire my own thinking and perhaps research about the brain and how neural activity, how to think about neural activity with respect to cognition and mind. Because I fully accept the complex systems kind of approach, self organization, even people like William Wimsatt with his causal thicket idea that within these dense mutual circular causation regimes we have to think about things outside of a domino effect, mechanistic, billiard ball kind of explanation. Right. So what I'm interested in as a neuroscientist is how can I think about mind and its relation to brain, if there is one, right. In these kinds of terms, not necessarily how I can build up from thinking of the organism in this inherently different way from, you know, non mechanistically, how can I build that up to consciousness or something? And we'll get to, we'll get to the fact that these kinds of approaches which you talk about in some of your talks, you know, they don't actually buy us consciousness you can't get there because it has to be. It's something else. Mind is something else. It's dynamically uncoupled. But nevertheless, I want to understand, like how thinking about these things and some of the formalization that has been happening can basically inspire my own, my own thinking. So what's at stake for you? [00:27:11] Speaker A: It's a great question. So for me, as I was saying, I started doing historical work then, let's say since 2015, 16, I started working on this, let's say, dialectical biology approach. And I was working on this for a few years. And more recently I. My current project at the University of Luxembourg is called From Biological to Cognitive Autonomy. And so essentially brings us very close to the ground that you were just describing. And so essentially what generated all this for me was a paper that I wrote with Matteo Mosio a few years ago. And the key idea of that paper was that the theory of autonomy has, let's say, two different souls. And one soul is what Matteo has defined as organizationism, which is kind of like the field we were talking about earlier, Kaufmann Rosen, the theoretical biology approach that tries to formalize self organization and closure and autonomy in biological systems. And the other side of the autonomy approach is what is today known as enactivism. And in a sense they both, they are part of the same story because they both go back to Varela, because Varela published when he was, I think he was in his early 30s. It's really impressive. He published Principles of Biological Autonomy. And then in 1991 he published another book which probably became even more famous, which is the Embodied Mind with Evan Thompson and Eleanor Roche. And that book is the book that grounded what is today known as the approach to cognition as an activism. Known as activism. [00:29:16] Speaker B: Can you describe in a nutshell what an activism is? [00:29:20] Speaker A: Okay, let's say what they say in that book. They say, so they start from a theory of biological systems, and what they're criticizing there is again similar to what we were discussing earlier. This equation of teleology and design, these ideas. There's the sentence in which they say adaptationism is to biology what computationalism is to cognitive science. What does it mean? It means that this idea of biological goal directedness as something which is established extrinsically from the power of natural selection is what they are trying to criticize. And instead they want to promote an idea of teleology as intrinsic purposiveness, or as we call it today and in many important debates in theoretical biology and philosophy of biology. [00:30:23] Speaker B: Agency, agency. [00:30:25] Speaker A: Yeah, yes, biological agency. The fact that an organism is an agent, can be a more complex agent, can be a less complex agent, but is an agent. And on the other side, they were criticizing this idea, which is kind of like the other side of the same coin, that the brain is a computing machine, which has been optimized by natural selection to elaborate information via representation of the environment. And the inactive approach, as people like Ezekiel Di Paolo and Ivan Thompson like to call it, instead of enactivism, the enactive approach instead promotes a view of cognition as something that is always online. In a sense, it's always the product of a dynamical relation between the system, the organism and the environment, and is the dynamical stabilization of patterns of interaction, of technically, what they call it the organism. Environment coupling is a stabilization of that particular relation. And that cognition is not this idea that you have mind on the one side and the environment on the other side, and you represent the environment almost as a disembodied brain in a vat and you do computation on your representation. You're always. And this is, for example, something a phenomenologist like Heidegger gave to the inactiva protest, a very strong contribution. You never cognize stuff from the point of view of a God, as Jonas would say, a God recognize stuff like this because it's not a finite being, doesn't have need, doesn't have a body that is mortal. The way you and I or my dog relates to the environment is through a standpoint of need, because you need food, you need shelter, you need. So and this is what generates value, the activists call this cognition is fundamentally sense making, is the production of value by relating to the finite and precarious nature of the organism, generates by relating to the environment from a standpoint of need, essentially. So this is broad strokes how I will define inactive. [00:33:05] Speaker B: So I totally sidetracked us by requiring you to define an activism, which you did swimmingly. So thank you for that. But going back to what's at stake for you before I cut you off, [00:33:18] Speaker A: what at stake for me, so for me, I was always interested in. So the key element of the inactive approach, and this is what I think the inactive approach has a lot in common with Hegel, is that for both the inactive approach and for Hegel, the explanation of biological purposiveness is, so to speak, a means to an end. Because for Hegel, notoriously, what we want to explain is geist, or which you can translate either as mind or as spirit. Because like philosophies, Geistes in Germany is philosophy of mind, the kind of philosophy of mind you could be taught at Stanford. But Hegel philosophies Geist as a little bit of a heavier concept because Geist is not the individual cognition of a rational entity that is separated from everything else. It's, it's grounded in the fact that we are goal directed systems and it's expanded or extended in the fact that we are not cognizing alone, we're cognizing with other people. And your cognition is scaffolded by your being part of society, of a certain society, of having a certain history or having a certain coupling with your environment. And so for me, what, what if you ask me what it's at stake for me was essentially, yeah, an explanation of what is mind and how it relates to life. So for me the, the main point is the relation between mind and life and the weird continuity, but also discontinuities that you have on this spectrum fundamentally. [00:35:13] Speaker B: Okay, so you're, so you're becoming a cognitive scientist and I think you've already been cognitive scientist leaning for a long time now. So then maybe describe, I don't remember if you already started to do this, but so then you have an activism, you have Hegel. And so how, how are they? Why is it important to relate these two? And I guess that's related to what's at stake for you. But this is, you know, what you and Matteo Mosio do in this paper that I'll link to in the show notes. So how are they related to each other and why is it, why is it important? Here's an unfair question. Why would it be important to a cognitive scientist that Hegel thought along the same lines as people who are part of the inactive approach? I'll say to be kind to say it the way that they would want it to be said. [00:36:13] Speaker A: It's a provocative. But as the provocative questions, they're usually very interesting. So what is the relation, the relation between Hegel and activism? So the paper that you were mentioning with Matteo. So for me I started from, probably from a wrong premise in a way because I was associating what Matteo was doing with Hegel. And the paper in a sense was what was the result of Matteo saying, well no, that's not exactly true because what we're doing is more similar to what Kant was doing. We're naturalizing Kant's problem. And what Hegel is doing is much more similar to what the enactivists are doing essentially. And in that paper the key idea was for organizationism or for the theory of biological organization that Matteo Leonardo, Bic, Mael Monteville and Alvaro Moreno have been defending, among others. The point is to come up with an explanation as to how is it possible that a regime like intrinsic proposiveness is possible and how you can explain it in a way that is amenable to a serious scientific explanation, in a sense. So for them, purposiveness is an explanation. So you need to develop. The goal of the research work is to develop a framework that allows us to explain that particular feature of biological systems in a way that even a moderately reductionist biologist will be tempted to buy, in a sense. And for the inactivists, purposiveness is not the object of explanation, but is the tool of explanation. So in a sense it's an explanation. So for the. An activist, you don't have to explain biological purposiveness. You start from biological purposiveness and you use it to explain how cognition works. [00:38:33] Speaker B: Coming from a reductionist sort of background, as most, if not all scientists do, you have to buy into, you have to accept that you can do that first. And so is that the work that Hegel is important for doing? Conceptually, there is the Varela and the inactivist camp also. But you have to convince someone that, oh yeah, you can use that as a tool in explaining things. [00:38:59] Speaker A: So what we argue in the paper is that what we suggest is a fundamentally anti reductionist strategy in the sense that, well, you do the explanation with the best tool you have and then you think later about how you're going to have a theory of everything. So it's a form of. Now in philosophy of science, there's a lot of. There has been for some time a talk of perspectival realism. This idea that, yeah, you have different angles. Also you, you were mentioning whimsat. Whimsat talks about perspectives. So stuff is complex, nature is complex. So there's some, some stuff. Sometimes you just have to take it from different angles and see what you can explain from this angle. And then you might worry about how these two explanations tend to each other in a sense. And you don't have to worry necessarily about having a theory of everything or having one principle that explains everything. [00:40:03] Speaker B: Well, I think to whimsat, right. Like an explanation that comes, let's say, I don't want to say top down, but let's say from agency first explanation of some phenomenon that aligns well. So one of those that's sort of self contained and seems robust, seems viable if it is in line with a reductionist parts to whole explanation, if they're not contradicting each other. He would call that a more robust explanation then because you're coming from these two different perspectives and then it's like a stronger explanatory situation. [00:40:47] Speaker A: Yes. Another guy that was incredibly influential for me in the past 10 years is Dennis Walsh, is a philosopher of biology at the University of Toronto. And he is also developing a story that goes a little bit along these lines in which he says, well, of course, I mean, we're not going to discard mechanisms. You can do a lot of interesting stuff by studying mechanisms. So we, we don't have to give that up. But should we limit ourselves to this idea that that's the only way you develop interesting explanations? Probably not. And he has this idea of, yes, perspectival realism. You can look at reality from complementary perspectives and then maybe at some point in the long run these explanation will coalesce into the same explanation. But until they don't, I mean, why should we just explain stuff through one mechanistic angle and not have different options that allow and different frameworks that allow us to explain different aspects of complex stuff? So, but I want to come back to the. This was in a sense a big preamble to the question that you asked, which is Hegel and an activism. Why is it important? Why would it be important to a cognitive scientist? And this connects to the work I've been doing most recently. And so the idea, the basic idea there is the inactive approach was born by the idea of combining, as the name of the famous journal goes, Phenomenology and Cognitive Sciences. The idea that you can have an integration between first person and third person. Third person is the usual game we play in the lab, but there is this. Varela came up with this research program called Neurophenomenology, which they tried to complement these two aspects and, and in a sense give human experience its truthful, its rightful role within cognitive science. And so of course, the main philosophical references of the enactive approach have been, well, Jonas at the beginning in this paper by Weber and Varela, in which they say Jonas is very important. We should use Jonas more about integrating first and third person. There has been work on Donnan Heidegger, who was Jonas professor and teacher and mentor. A very important person is Merleau Ponty, who's also about this embodiment and sensory motor contingencies. But very recently there is a very important book that was published in 2018 by Ezekiel Di Paolo Elena Cuffari. And Hannah de Jaeger called entitled Linguistic the Continuity, Life and Language. And in that book there is a chapter, chapter six, if I'm not wrong, that says, well, maybe we should add a new item to the list that is useful for an active explanation of cognition. And that too is dialectics. Why is dialectics important? Dialectics is important because it allows you to formalize complex dynamics in which you cannot mathematize a priority. You can only, only explain a posteriori because biological systems, as Kaufman said already are fundamentally non ergodic. So you have, you cannot pre state the phase space in which you do the explanation because they're historical beings and when something new happen, you have, I don't know, a case of exaptation as a term that comes from Stephen Jay Gould. And. And you cannot pre state that mathematically. And you just have to take it as a fact and see what are the consequences, what are the new phase spaces that are opened by that shift. And so I mean, if you're talking dialectics, of course for them the reference is more the Marxist, the Marxian kind [00:45:26] Speaker B: of dialect or Marxist for sorry, can you define dialectics? Sorry to make you stop and define things. [00:45:34] Speaker A: Oh yes, it's very important. It's very hard. Dialectics is the. Well, Ezekiel Di Paolo and his co authors define it as the principle that you need to use to understand concrete and open totalities in a sense that you. So let's say you have two main components to the story of dialectics. One main component is the relationship between parts and wholesome, kind of like what we were discussing earlier. The part in a complex system, and especially in a biological system, makes sense only in relation to the whole. You have a dialectical relation between parts and whole and whole to parts. And this is what Varela called the dialectics of identity. In the sense that you have one identity that is dynamic and is maintained by this circular relation between parts and wholesome. And then you have another side of dialectics which was emphasized more by another tradition, which is called, so called dialectical biology. Richard Le Bontin and Richard Levins, who published 40 years ago in 1985, this book called the Dialectical Biologist. And there what they emphasize is what we might define as the dialectic of interaction in the fact that a complex biological systems is not only determined by the relationship between parts and wholes within the system, but also by a complex relation of co determination between the system as a whole and the environment. And this relation to the environment tends to integrate well also just metabolism is a dialectical phenomenon because you take something that it's outside of yourself and you turn it into identity. So you turn difference into identity in a sense, and on an evolutionary scale. But this is a big debate about the role of organismal agency in evolution. But if you believe that organismal agency has a role in evolution, well, what organisms do in phenomena like niche construction, for example, but also phenotypic plasticity, the way they take in, on the one hand, niche construction, they modify the environment, and on the other hand, phenomena like phenotypic plasticity or epigenetics, they take what the environment gives you and they modify themselves to fit the environment. So you have again, this back and forth between the organism and the environment. So a general definition I will give is the fact that the parts are not determined alone and that it's always that dynamic relationship between the system and the environment. [00:48:24] Speaker B: Okay, yes. I don't know if you run it or if you're one of the authors of the Dialectical Systems. There's a webpage with basically a blog with a bunch of authors, contributors rather to the blog series, which I'll link to in the show notes. Okay, so thank you for that. And I interrupted you to define it because you were saying that De Paulo and his co authors in the book Linguistic Bodies said, okay, we need maybe another ingredient here. [00:48:52] Speaker A: Yes, and that ingredient is dialectics. But if you're talking dialectics, there is one guy you need to talk to, and that guy is Hegel, because they refer mostly to Marx and the Marxist tradition. But Marx was a young alien in many ways, especially in his early days. But you cannot understand capital or other writings by Marx without understanding Hegel. That stuff doesn't exist without Hegel. So. [00:49:24] Speaker B: But only three people in the world understand Hegel, and you're one of them. [00:49:27] Speaker A: So a little bit more than that. I mean, there's a bunch of aliens out there, but. But yeah, so. And yes, so I happen, in a sense, was a form of serendipity. I happened to be trained in university in which this was the main business. And so for me the connection was almost unconscious. It was a direct connection between the two. And one case on which I've been working mostly in the recent years, in recent years with my colleague Thomas Van Ess was. So one of the areas in which you see the role of dialectics most prominently is social cognition. This idea that. So we wrote a couple of papers called Inactivizing Dialectics and another paper entitled An Active Account of Labor, which in that case relates to the Marxist relation. And it's this idea that, well, according to a more classical account, cognition is something you do in your own brain. You do it alone, you don't have relationship to what it's outside of you, if not as a relation of representation. On the other hand, and first of all, for example, social cognition has been understood through the lens of what is called as mind reading. So I have my own, I'm close within my own cognition and you are close within your cognition. And the way I understand you as another social agent is by, in a sense assuming or guessing what, what your thought process are. And in 2007, another, probably the most cited paper in the Native Approach came out again by Ezekiel Di Paolo and Andy Jaeger. And the title of the paper is Participatory Sense Making and An Active Approach to Social Cognition. In which they say, well, but social cognition does not work that way. It's not like we all have our own fully developed cognitive abilities and then we try to guess what the other person is thinking. But it's more like a participatory process [00:52:02] Speaker B: that happens in between interaction, constant interaction, closed, circularly, causally closed interaction. Right. [00:52:11] Speaker A: In a sense it's, yeah, it's the meeting of two autonomous systems that generates its own norm of self maintenance in the sense that if we are talking now and we are exercising a process of participatory sense making, but this interaction can go on only as long as you ask me a question, I try to provide a pertinent answer and then you react to that answer. So we have a back and forth. If I start looking out the window or I just start looking at my phone, the interaction is over. So in that sense, the inactive approach at that point was not expressing this dialectically, but it is in later iterations. Well, cognition works through dialectical tools or in a dialectical manner in many ways because we're not self enclosed identities, we are identities that co construct in dialogues. And if you think about the role that the parenting you had, the country you grew up in, the kind of relationships you had earlier on the, the mentors and professor that you had during your growing up process, they are all sedimented in your identity as a person, as a scientist and as a podcaster, that doesn't come from nowhere. So in that sense this is the dialectical component of social cognition. [00:53:48] Speaker B: So I'm going to just make a counter argument that's I think, weak, but. So astrology gives a lot of causal power to very distant stars, right? And constellations, et cetera, that would be affecting us very weakly. But the majority of what's happening in our daily lives can be explained by much more local gravitational pulls, like the Earth, for example, and what's happening in our environment as we go through our daily lives. And one could make perhaps a weak argument that, yeah, how I, the friend that I had when I was 7 years old and, oh, I don't know, pushed me into a rose bush or something and I got all scratched up. Yes, that is part of my current cognition. Right. But the vast majority of my current cognition is dominated by what's happening currently in my brain and in my local environment and stuff. So is that an unfair way to. To look at it? I mean, like, not, not all, not all causes get equal weights in terms of explaining what's happening at the moment perhaps is a much quicker way of saying that. [00:55:05] Speaker A: I mean, not all causes work equally. But I think there is a bunch of people talking to psychologists about their pasts, and that's. That alone is. Should be at least an indication of the fact that, of course, your friend pushing you in the bush might have had a lower or a weaker impact, but there might be other happenings that had a stronger impact on your psychological and cognitive development. Well, first of all, I mean, the kind of language you speak, and if you speak English or if you speak Urdu, your access to the way you formalize the world and the way you make sense of the world will be substantially different, in a sense. And just the language you speak is already. And this is something that you learn by your parents and then you learn by the particular society you grew up in. And also, things tend to connect to each other because if you had had a certain. If you have a certain parenting or you also have a certain genetic inclination without being a genetic reductionist. But that's part of the. You have, you have both nature and nurture that will push you to, I don't know, study neuroscience instead of studying comparative literature. And then. So it's a chain of. It's a chain of event, in a sense. And it's. The activists call it, they use quite a lot this poem by Antonio Machado, who was a Spanish poet, that, as the poem says, that no hay camino. El camino se hace alandar. There's no path. You make the path by walking. So this idea of path making, and the path making is a form of sedimentation or more biologically speaking, canalization of possibilities. So when you canalize some possibilities, there are some. Some other possibilities that are not really available to you yet anymore. And those. And so your story determines what Is the realm of possible outcomes or probable outcomes that will come after that. [00:57:37] Speaker B: So going back to Stuart Kaufman's life is non ergotic concept. [00:57:42] Speaker A: Yes, exactly. Yes. [00:57:45] Speaker B: Okay. I want to ask you something that you have no business answering. So you don't, you know, but I would love to hear you hazard a wild guess here. Right. So actually, two things. One, I mean, as a, as a neuroscientist, if I want to do the analog and approach understanding how brains work in relation to cognition and consciousness and ongoing, you know, lower level to higher level cognitive actions, should I even be recording and analyzing neural activity? [00:58:22] Speaker A: Well, so I have to say so, little disclaimer. The story is long and let's say the neural aspect of the story, for me, it's a very recent addition to the landscape. And for me, the person that is influencing me currently when thinking about this stuff is our common friend Luis Pessoa. And I was discussing with him because of course, for example, within this, the inactive approach is part of this. They call it radical embodiment family. There is this very famous book by Tony Chimaero. Tony Chimarrow is professor at the University of Cincinnati. And the book I think is from 2009 and is called Radical Embodied Cognitive Science. And this radical embodiment family basically joins two other families which are the inactive approach that comes from Varela and the ecological psychology that comes from Gibson. And in some cases there seems to be the idea that, well, you don't need computation at all. It's just direct perception. It's just relation between the organism and the environment. This is not true because there are people now working on radical embodied neuroscience. I'm not an expert on this, but Vicente Raja is and is doing interesting work on this in a sense that what I was discussing with Luis was, well, if you push it in this too radical way in which, well, the brain is not important anymore, the computation is completely useless, et cetera, et cetera. It tends to be at least what Luis was saying is, yeah, we. What we are left with not much in the end. [01:00:19] Speaker B: Yeah, very little. [01:00:19] Speaker A: Yes, we're left with what are you going to do? Well, and Louise example for me is, is, is Louis work is really a very interesting example of how you can use all the nice and interesting tools that neuroscience has. And also if it's a little bit what we were discussing earlier about mechanisms, you can be, you don't have to be an anti mechanist to be a reduction. An anti reductionist. You can, I mean, mechanisms do a lot of interesting explaining work. Why Would I want to do away with them? It would be silly. [01:00:56] Speaker B: It would be anti perspectivalist of you. [01:01:00] Speaker A: Yes, exactly, exactly. And so also computation, you can do computation, but as long as you're not doing what people like Hilary Putnam were doing, saying, yeah, the brain is a computer. If you're saying the brain is a computer and works as a computer, and you really should model the brain on the functioning of the computer, that I would probably not buy. [01:01:28] Speaker B: But as long as even that gives some explanatory power. Right. It might be weak explanatory power, but it's wrong. But all models are wrong. And maybe that's just a really weak model. [01:01:41] Speaker A: Yes, exactly. You have to take it as a model. And, and what Luis is doing is doing computational work that is doing very embodied. It's computational research that is doing very embodied work. Because at some point you also need to sell it to people that don't already believe the gospel, so to speak. And that's the way to. I mean, you need to provide, that's the business of science. You need to provide, provide results and evidence. [01:02:19] Speaker B: You know, Luis works, you know, in FMRI a lot, right. And so, you know, he has access to this whole brain system. And I know that part of like your work with Luis is talking about the difference between a hierarchy, like from Simon's old, from Herbert Simon, who introduced this idea of a hierarchy in complex systems versus a heterarchy. So in a hierarchy there's still sort of a controlled structure and in a heterarchy there's no main controller. And instead control is sort of an emergent property of the interaction of systems based on the context of what's happening in your environment and your ongoing needs, et cetera, etc. And in a sense this is inherently dialectical, right, because you have these interacting parts and it's all about the interaction and it's the relations between them that you could say that there's this emergent property of control, for example, if, you know, if you go down that road. And so what Louise is doing in some sense, since it's a natural fit to take this kind of approach. And having said that, modern day. So I record the activities of single neurons, many of them at a time. I happen to be only in a few brain areas, but more and more people are recording from more and more brain areas simultaneously, even though it's. You can't record all the neurons all the time. I mean, you're going to be under sampled in terms of recording neurons. So in some sense maybe the modern population Spiking neuron data collection is more amenable to this sort of heterarchal approach or asking scientific questions about it. So I'm just being very stingy in terms of what I want to answer with this particular data set that I'm working with, right from like two different brain areas and with, you know, a mouse performing a task. And I'm sort of. Because I buy. I really love all this work that you are a part of and we've been discussing, and I'm sort of struggling to figure out how to approach it from a sign, you know, what kind of questions can I ask if I take this Hegelian inactive stance, you know, So I know you don't have the answers to that, but this is sort of, again, what's at stake as a neuroscientist. It's what's at stake for, for me. [01:05:07] Speaker A: So I, I might have not an answer, but the, the maybe the beginning of an answer. And again, I don't take credit for this because this is largely something that I get from talking to Luis. So. But I'm going to take a slightly longer road. Not too long. [01:05:29] Speaker B: No, that's fine. [01:05:30] Speaker A: To, to kind of COVID all the, the interesting things you just said. So you started from hierarchy and hierarchy. So in a sense that formalization, to give credit where credit is due in that particular conceptual way comes from the work of the joint work done by my friend Leonardo Bick with Bill Bechtel, who's a philosopher of science at University of San Diego. He's emeritus now. And they've been doing an incredible amount of incredibly good work in the last five, six years on the concept of biological control and how control works. And there is a paper, I think the first paper they put out was from 2021 in the. I think was the Proceedings of the Royal Society about, about how was a special issue about basal cognition. And they try to define biological control as working not hierarchically, but hierarchically. So normally the idea is, and this is also where Luis comes in, when it comes to the brain, the received view and how people speak is at least as if the prefrontal cortex is really the manager of the office. And then the other people are the, the other areas of the brain are the people doing specific work, but who need to respond to the manager. So you have a very hierarchical view of how the brain works. And Bill and Leo are defending this idea. They're doing interesting work on C. Elegans, for example, of how actually the different parts of the brain take the lead in different situations. And Luis, although I don't think he really formalized it conceptually in this particular way, but he's a natural fit to this. And he's been doing incredible work all his career on the relation between emotion and cognition. And this, this is classic idea in which goes back to Plato, the famous metaphor of the charioteer, in which you have, in the Republic, you have the charioteer, which is the soul, and you have the white horse, which is reason, and the black horse, which is passion. And in the sense the soul needs to through reason dominate passion with the wild horse, that is bringing the system the wrong way. And Luis has been arguing with very strong empirical evidence that this is really not the case. And that the brain, you don't have a strict partitioning between the different area of the brain and who is doing the leading, who is the manager, largely depends on context. So for example, if you are sitting around at home and you're talking like this, then probably the prefrontal cortex will have more relevance because we're thinking about abstract stuff. But if I have a lion running after me, that's not exactly the case. So there'll be other areas of the brain who short circuit for survival reason and take the lead. And what's a way, what we were, in that talk you were talking about, and this is a paper we're trying to work on, but still, we're still struggling in a sense. One of the ideas that came up in discussion with Ruiz is this idea of network models. I think he has a paper from this year which came out on network models. Yes, exactly. Yes. And the idea is that, well, if the brain is a network, and you want to model that through the brain, through network models, but you want to take an heterarchical view, you cannot have a fixed network in which the nodes are strictly defined. Because if the brain, we are historical systems and our brain wiring changes, and of course then you have some forms of canonization. Our brain is less plastic than the brain of a two year old, but you're still reopening different connection between brain areas. And so in that sense, the brain is a network, but the nodes are not fixed. So the nodes are shifting. So you need some. [01:10:28] Speaker B: This applies both anatomically and functionally, because you could have a fixed anatomical structure and still run different dynamics on it. So I just wanted to throw that in there because we think of networks as. As structural, but they're also dynamic, functional. [01:10:44] Speaker A: Absolutely. And this connects again to this very dialectical theme that we were discussing about non ergodicity Kaufman and here there's a paper that is really arguing the same thing, but in which dialectics is really doing the heavy lifting, which is a paper by Richard Levins from 1998 entitled Dialectics and Systems Theory. Levins was a, was an ecologist by training, but he was also a Marxist by, by choice, in a sense. And, and what he says in that paper is that, well, systems theory, as in dynamical systems theory, is always working within a determined state space. But again, since we are historical beings and you cannot pre state mathematically beforehand, what are the paths that are going to be walked on? Because the part the path is done in walking dialectics is the tool that allows you to think about the impossibility of reducing. The structure of the network to a fixed state space. Because when then through evolution and organism environment coupling the state space, you're thinking about in this state space. And there is a very nice diagram in Luis paper in which you have the entire diagram, the entire axis shifts in a sense. And so then you have to rethink the whole model. So it needs to be fundamentally context sensitive and, and never fixed in a sense. [01:12:32] Speaker B: Well, so, but, but going back to. All models are wrong, even within a fixed state space or phase space, if you want to call it phase space. That's not to say abandon dynamical system theory approaches to modeling, for example, neural activity, because frozen in time it is a fixed state space. And so you know, you can approximate which all models are approximations. Alex Mayer would disagree with me on that. But it's perfectly fine to use a dynamical systems theory fixed state space model as an explanation of some neural activity. However, a different approach is, I guess, as you would say a dialectical approach is understanding that the state space itself can change and we can model that as well. Well, so I just want to say that it doesn't invalidate fixed state space models. It's just again, a different perspective, a different approach. [01:13:34] Speaker A: Yeah, I think it all comes back to this theme that came up a few times. One, to embrace this kind of approach, this complex systems dynamical or dialectical approach, if you want to push it a little bit further. Other you don't have necessarily to invalidate all reductionist models. You just have to remember that the map is not the territory and that the territory is extremely complex. And then you need flexible maps to navigate a very complex territory, you know, only very partially, in a sense. [01:14:23] Speaker B: So I'm going to bring it back to Hegel really quickly because I also sidetracked this just then. But again, I come Back to this. Why should I care about what Hegel thought? This is a terrible question to ask someone like you. [01:14:36] Speaker A: You don't. Honestly, I don't go to Hegel conferences anymore because I find them terribly boring. [01:14:43] Speaker B: Well, that's, you know, the way that progress is made is not by staying within one. Like, if you're a Hegel scholar your whole life, that is totally fine. And if you love it, that's great because you can reinterpret and try to re. Understand Hegel itself. But it's at least in the sciences, progress gets made. Usually when someone from an outside field with outside expertise comes into a field that they don't have expertise in and there's some interaction, like a dialectical process, but there's some interaction and then some inspiration from an outside field. [01:15:14] Speaker A: Field. [01:15:15] Speaker B: These days it's all like physicists and neuroscience. But this is one of the reasons why, like, I appreciate what you're doing, like kind of stepping into cognitive science and taking this philosophical approach and from what you understand with your deep background in like these German idealists, you know, from Hegel and Heidegger, and trying to understand the cognitive sciences and not just the cognitive sciences, but the biological autonomy and intrinsic purposes, purposeiveness in that modern sense relative to that historical philosophical sense. So I don't know. There wasn't a question there, but no, [01:15:56] Speaker A: but yeah, for me, Hegel is a tool in the toolbox. It's an incredibly powerful tool in the toolbox if you know how to use it, which is not in that sense. If you want to save also the historical work a little bit, you need people that are able to teach that very. It's a tool that is very costly and hard to manage. So you also need people that are able to teach students how to use that tool. So in that sense, also let's say if you want to be charitative to people doing hermeneutical or historical interpretative work on these texts. These texts are incredibly complex. This guy was not making it easy on you as a reader either. [01:16:46] Speaker B: So was that intentional? But some, some authors are intentionally abstruse. Right. And because it makes you look smart. Was he con. Was Hegel. Were these people concerned with looking smart or were they just doing their best to explain hard concepts? [01:17:02] Speaker A: Yes, I. I think that this is more the case. I mean, they're trying to. They're seeing something very interesting and very, let's say. I don't like the term, but let's say deep or subtle. Yeah. And then you have to express it in, in a language understandable to other humans. And for example, a very interesting case in this, because I was teaching Heidegger last year. I'm going to teach Heidegger next semester too. So I read Heidegger when I was 20 and I was really interested. Yeah, yeah, it was. It was like first. First semester of the university and. And it was like being slapped in the face. In a sense, you. You don't need. You don't even know what hit you because it's so. At the same time, it's super interesting and super. And you. So that's what pushes you to go on. But you also. You don't know what is even happening. And Hegel is even worse in that sense. And I was teaching that 20 years later last year, so I had a more mature approach to the text. And you really see, for example, in the case of Heidegger, he was working on this thing for 10 years. He was not a professor yet. It was like what you call an assistant today, probably. So he didn't have tenure. And so he was in his 40s, and his mentor and teacher, Husserl, who's the founder of phenomenology as a movement, was pushing him to submit something to get tenure. So he submitted the first half of the book, which is incredibly complicated because it's very idiosyncratic. But overall, if you have a little bit of training and you know how to read it, it's readable. And the ministry refused the tenure because it was not enough. So he was working on a second part. So he was pushed into the corner and submitted the second part too. And this is being in time. He was supposed to have six sections, we only have two, because then he changes his mind. But you really see the difference between the first part, which is complicated and weird, but overall readable, and the second part, which really is a bunch of notes, because in the end, these people are also humans who are having to deal with humanly stuff like getting a job, and they have a bunch of notes in the closet and then they say, yeah, I'm just going to publish them like that. And then it remains a classic, but in the end it was just a bunch of. I was reading the second part and it was not much peer review back then, but it was a very hard, revised resume. The ideas are great, but please write them in a way we can understand and organize them. So that's also part of the story. [01:20:01] Speaker B: I guess so. So we were taught. We've talked a lot about Hegel and you've talked about how you're teaching Heidegger. You've taught Heidegger and are teaching Heidegger. I mean, not speaking of tools. That's sort of. That's an untasteful joke about sort of his dark side. Right. But. But he's been intellectually, he's been like an important figure along these lines. And of course, Hubert Dreyfuss has these excellent online lectures about Heidegger and Marlo Ponty. And so I'm curious about your. How you see Heidegger fitting in along this same vein. I don't know if it is Heidegger a continuation of Hegel. How should we think of Heidegger in relation to Hegel and enactivism? Because he's very much related. His diet. Is it Dyson. Dyson. This sort of being in the world, like the concern with. With being and doing in the world when you're engaged in like, everyday activities. Right. For example. So anyway, I'm not sure what you teach about Heidegger because whether it's related to the cognitive sciences at all. But I'm curious if you have thoughts about how someone like me could. Could take inspiration from someone like Heidegger's work. [01:21:23] Speaker A: Well, you mentioned Dreyfus. Dreyfus. It was of course, the. The key reference for me because he, he was. He was able to accomplish something incredibly hard, which was introducing Heidegger in. In the United States, which. Which was for it's the realm of what is called analytic philosophy. And analytic philosophy hates Heidegger by definition because it's. The analytic philosophy is all about arguments and clarity of arguments and where every step of your argument should follow and your argument should be coherent, et cetera, comes from logical positivism. And Heidegger is. Is the completely different end of the spectrum because it's very evocative and he's changing completely. The language that he's using arguably a bit because it was idiosyncratic. But the idea is that he's trying to express things in a radically different way with respect to the way they were expressed in the former. According to Heidegger, 2,000 years. He was not shy man, let's say when it came to philosophical prowess and what Dreyfus's book. So Dreyfus is very interesting because he wrote a very interesting commentary to the first part of Being in Time, which is the part that I was saying is actually readable. And he also wrote a classic book. It was published in 72, before his. I think his Heidegger book was from the 90s. But the book that made Dreyfus famous is what computers can do and was a very early criticism of artificial intelligence, or what John Haugeland called the gofi, the good old fashioned artificial intelligence, which was very different from the current large language models, statistical structure. And what he was arguing fundamentally in that book is that artificial intelligence is not really intelligence because it lacks a body and understanding and cognition. And intelligence is not something done by a disembodied computing machine, but it's something done in the sense it is very, very inactive. Although never really, I don't think they ever came into fundamental contact. But it's a very embodied cognition leaning guy in the sense that all this idea is. So Heidegger, you mentioned this idea of being in the world. And this idea is the key concept of at least the first part of being in time. And it's this idea that, that what is the world, the world. We usually think about the world as a container in the sense that you are in the world like the water is in the pitcher, but you're not in the world in that particular way. Being in the world is you're relating to things as things that can be used. So I relate the picture as something that I can use to drink. I relate to the doorknob as something that I can use to open the door and go somewhere. And so in that sense, Dreyfus was doing. I don't know how explicitly, but there is a very strong connection between this idea of being in the world that's using stuff and using stuff that relates to me as an embodied being who wants to keep living. In a sense. This is what it boils down to in the end, for Heidegger, your relation to things in your field of cognition is the relation to affordances. You're not again, a disembodied intellect that relates to stuff in a purely theoretical level. Your relation to stuff is fundamentally practical. The baseline is, is practice and theory is a derivation of practice in a sense. And intelligence and cognition is in many ways a sort of, is the end result of our practical dealing with the world, connected to vital values, so to speak. And so this idea of cognition is a purely disembodied computational process. According to Dreyfus, relying on Heidegger is something that we should really discard. And I think I'm not by all any means an expert on artificial intelligence. I don't do a philosophy of artificial intelligence, but by intuition. This is something that is true for the good old fashioned artificial intelligence of the 60s and for large language models today. [01:26:51] Speaker B: That it's no different, right? [01:26:53] Speaker A: Yes. [01:26:53] Speaker B: Really? Yeah. It's a difference of degree, not kind. [01:26:56] Speaker A: Yes. [01:26:57] Speaker B: Well, that's okay. So I, I really, I can't not ask you now about artificial intelligence. And you know, from a Hegelian perspective or Heideggerian or an inactivist perspective, I mean, do you have thoughts about maybe the differences between artificial intelligence and biological intelligence or you know, is the word. Should we even be calling it artificial intelligence? I guess just largely. What's your view on all of the hubbub over artificial intelligence becoming conscious? Do we need to give artificial intelligence human rights? You know, welfare? What do you, how do you view all this? [01:27:43] Speaker A: I mean, the short answer is that I think it's nonsense. The longer answer is that, I mean, you also don't want to be necessarily a reactionary or ultra conservative by saying this will never happen or what I'm saying, I think, I don't remember whose paper it was. Maybe it was Ezekiel Di Paolo. And so the idea was, well, you might have an artificial, an agent which is not, was not born by. It's not a carbon based system. But to really have an agent and a cognizer and something that you will call it a general intelligence. This could not be something that is programmed or realized by statistical computing or statistical inference. It should have, have what you might say, a concrete history of embodiment. So it's something that should really form. Cognition needs can be formed only in, by walking the path which is the, your individual path of individuation of your relationship with the environment. Can it be done? I don't see it. I'm not, I don't want to say it's impossible in principle, but yes, I don't, I don't see it happening soon. [01:29:15] Speaker B: So I wanted to bring this to consciousness from the constraint closure perspective as well. Because my understanding, and you can correct me on this from people like Ezekiel depaolo and others, is that constraint closure is necessary for agency, but agency is something different than mind. Slash consciousness. Mind and life are not the same even though they are continuous, they are decoupled. Right. So, so wouldn't. What's. But constraint closure is necessary for agency, but is agency necessary for mind? Right. So I guess, guess going back to the artificial intelligence question, would you have to see constraint closure or some sort of organizational structure that is amenable to intrinsic purposiveness in order to accept that artificial intelligence is real intelligence? Right. [01:30:25] Speaker A: But also. So I think this, what you were saying about decoupling and the role of constraint closure in agencies. I think that the whole story is very long, but I think you can tell the story going back to the title of probably what is the most important work in this tradition, which is actually. It was actually a paper that appeared in Spanish first in the 70s and was published as a book. It was actually two papers, one by Maturana, one by Varela and others. And the book is called Autopoiesis and cognition is the foundational work of this tradition, so to speak. And since Autopoiesis and cognition was published in the 1980s, actually precisely in 1980 and for a long time. So essentially what they are saying there is. There is a fundamental relation between autopoiesis. Autopoiesis is just the old term for closure of constraint. So this idea of self generating itself. Yes, I think the generates itself. And the fact that. So the fact that biological systems are autopoietic is fundamentally connected to the fact that they are cognitive agents. So in that particular framing, cognition seems to come for free from autopoiesis or from closure. But that picture has been complexified quite a bit since the mid 2000s. Again, there is a landmark paper by Ezekiel Di Paolo entitled Autopoiesis Teleology Adaptivity Agency or Autopoiesis Adaptivity Theology Agency. I don't remember the order correctly. There is another paper from 2004 by Bourgin and Stewart which problematizes the equation between autopoiesis and cognition. Of course autopoiesis is the key is the fundamental requirement for cognition, but it doesn't give you cognition for free. The argument Di Paolo has developed in this paper that was cited with some ground a lot of time is that you need to really account for cognition and for agency. Self production is not enough. You need adaptivity because cognition doesn't happen. So for example, a self producing system in a sense is maintaining its autopoiesis. If I throw myself out of the window now, I'm still maintaining my autopoiesis until I hit the ground. But to understand cognition and agency, so it's not a black and white property, you need need shades of gray. You need a relation to the environment in which you have areas of the environment that are more amenable to self maintenance or areas of the environment that are less amenable to self maintenance. The classic example within the tradition is the monocellular organism, the E. Coli in the sucrose gradient. And Kaufman quite provocatively says when the E. Coli is moving around, what is he doing? He's going to get dinner. So it's a bit Anthropomorphizing too much, but is also provocative in the sense that system, that organism can do, it's a very simple system, can do two things. It can tumble and it can swim. And for example, if there's an increasing sugar gradient, it will probably go towards more sugar because sugar is conductive to the maintenance of autopoiesis. If there's acid on the other side, it will probably move away from that area of the sucrose gradient of the solution it's swimming in. So in that sense, and this is a very basic form of agency, but to really have that, you need to have the, this adaptivity component, which is the ability to evaluate, even in very basic regulatory terms, the environment according to the basic norm of self maintenance. And in a sense discriminating what is good and what is bad for self maintain. [01:34:59] Speaker B: Well, this is directly related to sense making or relevance realization. Right. So I wanted to ask you how that came into the picture. But, but so we're using these terms cognition, intelligence, adaptivity. Right. So cognition and intelligence aren't the same thing unless you define them as the same thing. And they're both like such large terms and fuzzy terms, right? That, and we're using these terms and just kind of jumping around them as if they're well defined and they're not. But even adaptivity, as you were saying that, well, that's like the defining property of neural networks that allows them to do their quote unquote cognitive, intelligent things. Right. And so one would say that sure, they don't have an intrinsic purposeful organization, they don't have autopoiesis, but they definitely have adaptivity. So maybe they're halfway there there. Or is that is. Is plasticity and weights a different kind of adaptivity that we're talking about? Because you're talking about adaptivity almost as relevance realization. [01:36:10] Speaker A: So what is the question here exactly? [01:36:15] Speaker B: Well, I wanted to bring it back to artificial intelligence because like the defining feature of deep neural networks is their adaptivity. So maybe they're, they have this necessary component that DePaulo writes about that, okay, sure, autopoiesis is not enough. It needs adaptivity. So would neural networks have the adaptivity that is supposed to be required for cognition? Or are they lacking? Or they might have that adaptivity. But what, what DePaulo actually means is something about being able to change but also give value. In other words, make sense of the world. In other words, have relevance realization to frame what's even could be important to you. And you need that intrinsic purposeiveness to achieve relevance realization, which is a necessary component. Right, Sorry, that was a real mouthful. [01:37:14] Speaker A: Okay. No, I mean, we, you touched on a lot of different interesting themes. So I'm going to try to bring it all together starting from the end. So, deep neural networks. So I mean, disclaimer. I'm not an expert. I know almost zero about how artificial intelligence work. I've been to a couple of workshops, but by any means, I'm not an authority of any kind on this. So my take it as a grain of salt, but my intuitive answer, Deep neural networks, they might have this adaptivity or they might not, but I can assume that they have this adaptivity component, but they don't have the. They're not autopoietic system. They're not trying to materially maintain themselves in relation to a changing environment. So in that sense, sense, what we've been discussing is the relation between autopoiesis and cognition. And we were saying, well, autopoiesis is the fundamental condition to have something as cognition. But it's not enough in that respect. There was also another paper by Alvaro Moreno and Xavier Barandiaran from 2006 with a very interesting title. The title is on what makes some autonomous systems cognitive. And the main idea that Alvaro Moreno has been defending for many, many years is this. I mean, the core idea of that paper is to really be cognitive, you need a form of dynamical decoupling between the autopoiesis, the systems that deal with the metabolic dimension of autopoiesis, and as another system that is partially decoupled from metabolism and is in charge of regulating metabolism in relation to the environment in ways that promote adaptivity. So you need a basal fundamental decoupling. And then of course, in Alvaro Moreno's evolutionary stories, the strongest decoupling you have is the decoupling of the nervous system, which is connected to the vertebrate body plan. And so in that case, this is where then where you get consciousness, because you need a stronger coupling in which the body is fundamentally, let's say, embrained. So coming back another step about what you're saying is that cognition and intelligence, are they the same thing. What comes to mind to me with respect to that question is what two other contributors to the enactive frameworks named Dan Hatto and Eric Mine. Dan Hatto is a professor in, at Wollongong in Australia, and Eric Mine is a professor here in Antwerp and they wrote two books, one of which is called a very provocative title, Radicalizing an Activism. And the second one is called Evolving in Activism. And I think the biggest selling point of those two books is really trying to provide a distinction between basal cognition and let's say intelligence. The way they call it is content involving cognition, intelligence and, and content free cognition, which is basal cognition. Basal cognition is, according to them, is not contentful in the conceptual sense. It's not contentful in the, in the way a form of cognition that has propositional attitudes is contentful in the sense that you are intentional, you're. It's really, you have a form of aboutness that is conceptually robust and for them, basic cognition. So the essentially non linguistic cognition is content less and you don't need content and representational content to explain the way non linguistic cognition works. And then in the second book, so this is the selling point of the first book, radicalizing and activism. We don't need content, contentful cognition to explain the way embodied body relation to their environment works. And then, but then what we might define as intelligence or contentful cognition is something that comes about and again, this might be framed again in dialectical terms once we have forms of what Tomasello has been calling shared agency or joint intentionality, human intelligence is what characterizes intelligence in the human sense is precisely this. That cognition is always social, is something that we do together, that is sedimented historically that we are born into, is a form of cognition that has a lot of scaffolds that are already there and we're born into and grow into. So in that sense, the way I will respond to your distinction between our cognition and intelligence, the same thing, not their, not cognition. Intelligence is a special kind of cognition that has been going through historical and intersubjective processes of sedimentation and sedimented participatory sense making, so to speak. [01:43:07] Speaker B: I like that. Okay, so we just have a couple minutes left here, Andrea, and I think that if I could summarize just the style of our time here, been you trying really hard to keep us on track while I keep making us, while I keep trying to jump us off the tracks. But you bring it back in so beautifully. So I really appreciate it. I really enjoyed our discussion here. And I can tell that you're a really good teacher and I've enjoyed your works and your talk. We haven't talked about something that was on my list to talk to you about, and that was this paper that you wrote with Moreno, I believe, Alvaro Moreno, about you guys are discussing the difference in body plans versus and what might be more or less amenable to consciousness in terms of vertebrates and invertebrates and I don't know if you want to touch on that briefly or if there's something that I haven't asked you that you would like to discuss or something I have asked you, but since I jumped around so much much you felt like maybe was incomplete. Is there anything that you want to discuss in these last few minutes? [01:44:17] Speaker A: No, I think I can jump on that because I was also a little bit touching upon this in the last discussion. So this is a paper I was mostly just helping out with. I don't have any credit on the conceptual side. I was mostly helping out Alvaro, framing his ideas to respond to reviewers. But I learned a lot in that process. And it was a very interesting way to see, reframe themes that for someone who has been reading Alvaro's work for many, many years, was interestingly applied to the problem of consciousness. And so the setting of the paper starts to say, well, consciousness is probably the most problematic issue or concepts or set of problems in natural science, because nobody even know what it is. Nobody even knows how to define it. And the other strategy there was to say, well, we're not gonna get into the high weeds and try to define what consciousness is. We're going to try to provide. Provide an evolutionary story about what are the organizational conditions for consciousness to be there to appear, or for something like consciousness to appear. And the main idea that Alvar has been defending for quite some times in different ways, so back to 2006, this idea of the dynamical decoupling of the nervous system, the dynamical decoupling of the nervous system as the epitome of this decoupling that you need for an autonomous system to be cognitive. And this is applied. This appears in the framework of a book that just came out last year by Alvaro with Yuli Pareto. It's called An Evolutionary Story of Agency. It's very empirically based on. Based, but it essentially goes. It's a short story because it's a short book series that goes from the origins of life to consciousness and how the particular organization of organisms throughout the major evolutionary transitions contributed to the complexification of agency and consciousness is, if you read it in the. This frame is the ultimate or at least the most recent development in this complic. In this complexification of agency. And for Alvaro, the thesis is defending there with, I have to say, quite strong empirical data, because the paper is analyzing a lot of mostly evolutionary literature. And he's saying the key thesis there is, is the vertebrate body plan, especially land vertebrate, is what is fundamentally conductive to something as consciousness to appear and which implies a form of strong centralization of control with respect. For example, to a great example that is discussed in the paper, which became very famous through a book by Peter Godfrey Smith. I don't know if you know it, it's called Other Minds. And another book, Metazoa, he talks about the octopus. [01:47:58] Speaker B: It's all octopus all the time these days. [01:48:01] Speaker A: Yeah, there's also. There was a very famous documentary on Netflix, I think, on the octopus. [01:48:07] Speaker B: I was gifted the documentary. Oh yeah, my octopus teacher. There's also. There was a movie based on a book that I was gifted about how smart octopi are the octopuses. But it was a fictional story. But, but boy, there's been a lot of octopus the past decade. [01:48:24] Speaker A: I mean, there's good reason for it in the sense that they're very interesting case. And in a sense in the paper is used as a potential counterargument that then we try to argue against because octopuses diverge very early on from our body plan and yet they display forms of intelligence that are quite puzzling. Also almost emotional intelligence strategy. They're able to open stuff, they're able to disguise. They have a form of adaptability that shows an intelligence that is really out of norms for the kind of taxa they find themselves in. [01:49:06] Speaker B: But they have that massively distributed quote unquote brain. It's like, don't they have more neurons in their tentacles than they do centrally? [01:49:14] Speaker A: Yeah, yes, exactly. They have a very high number of neurons, but it's very distributed. And this is exactly according to Alvaro, the, the main point that. So essentially you, you can. There is an evolutionary story that you can tell about evolutionary arms races with fishes, but according to Alvaro, precisely because it's very distributed from it, it's you, you have an evolutionary or fitness explanation for how, how that happened to be the case. But since they have this distributed body plan and which is fundamentally different from the vertebrate body plan, they hit a fundamental bottleneck. So they got there, but they cannot go further in a sense. And this is the other thesis, I don't know how much I think is well argued in the paper. I'm not sure if I will bite myself because I don't know the field well enough. But. So the idea is, well, sure, octopuses are smart, octopuses are interesting, but they're ultimately finding themselves in a evolutionary bottleneck when it comes to cognition. And so really what is really from an evolutionary argument point of view, what is really conductive to consciousness is, is the length vertebrate body plan. It's a bit anthropocentric, in a sense, but maybe it's true. But maybe it's true. But I'm not committed to any of this. [01:50:50] Speaker B: Where's the octopus mind, then? I'm sorry, I don't remember this from the paper. So, as a neuroscientist, we're taught the brain generates consciousness. And part of the argument there is that you have to have an encephalization of this sort of central mass of neural material and activity. But in the octopus case, it's so distributed. So are they, like, super smart without a mind? Is that the idea? [01:51:19] Speaker A: I don't know if they're super smart without a mind, but for sure, the encephalization component is what, at least in that particular paper, is what really is what. All the whole entire argument comes down to that. It's about centralizing control in one center, which is the brain. And I was talking to Luis also about this because apparently, at face value, there seems to be a clash between the fact that, yeah, you need centralization of consciousness, but also the brain is heterarchical. [01:51:54] Speaker B: Yeah, that's what I was going to bring it back to. All right, this is a great place to end it. Go ahead. And then I have one more thing. [01:52:00] Speaker A: And. And apparently what? So I was bringing this potential contradiction to the table in my discussion with Whis, and Whis was saying, well, both things can be true in the sense that you need a strong encephalization, you need a stronger decoupling in the sense that you have the brain that controls the other parts of the body in one way or another. But then you could be heterarchical about how the brain. How. How the. The. The. How you. You can believe in encephalization but then still say, well, but the way the brain works is not the prefrontal cortex is governing everything, but it's. You have different parts of the brain that are doing the controlling work at different times. [01:52:41] Speaker B: Yeah, yeah, that's okay. So great. Because I was going to bring us back to the idea of heterarchy, which is what you just did with your work with Luis. [01:52:52] Speaker A: And. [01:52:52] Speaker B: And I think that I relearned this in one of your talks because I think I already knew it, but I find it wonderful and slightly ironic that I think. Was it Warren McCulloch who coined the term heterarchy? Okay, so this is wonderful because it's the McCulloch Pitts neuron, artificial unit, artificial neuron, which was like a logic gate, essentially, that. That has Given rise to what we know of as artificial intelligence, gave rise to neural networks and up to deep learning and what we have now. And I wonder what McCulloch would think, because he knew, speaking of heterarchy, you know, he knew about the huge amount of recurrence in brains and that they would have to account for that essentially one day. But you know, then they were just building it out from these artificial units and they had like, you know, little diagrams with some recurrence or whatever. But part of what recurrence is naturally aligned to do is form these heterarchies as opposed to a hierarchy. So I just kind of wonder if he would find it ironic that today's AI is looks the way it does and in neuroscience because he wanted to understand like cognition, he wanted to understand the brain and stuff. Right. He wasn't concerned with building AI, but the way that neuroscientists conceive of the brain as super hierarchical when in fact he was arguing it's heterarchical from the very beginning. If you. From that beginning, anyway. So I don't know, just, just comment there. I thought that was interesting. I don't know if you have. [01:54:28] Speaker A: I think, yeah, to close here, it's something that I've been thinking quite a bit about the history of, of the cybernetics movement. And if you ask, for example, my friend Matteo Mossio is also not in print, but in his habitation thesis, it's very straightforward in the fact that cybernetics is just a mechanistic theory of cognition and biology and agency and we don't necessarily need it or interested in it. [01:54:59] Speaker B: There's neo cybernetics now, which is not classical cybernetics that I think is getting away from that. I don't know if you're going to [01:55:04] Speaker A: that, but yeah, I mean, what. There is a very important book by a guy called Dupont. The title in French was something about on the origins of cognitive science. And it's mostly talking about the history of cybernetics. And what he's saying is cybernetics was in many ways a history of lost encounters, lost encounters with biology, lost encounters with phenomenology. And this is exactly what the inactive approach is, in a sense is mixing these, this complex system approach, but putting it together with phenomenology, with embodiment, with the biology of self organization and autonomy. So the way I see it is a ground on which that can go in many directions. You can take it, as many have done, into a reductionist computational direction, computer science or. But it's also the same ground that generates a lot of the things that for me are interesting, including biological autonomy that comes from second order cybernetics. Heinz von Forster was one of the mentors on Varela. So it's a very complex ground that can be taken in different directions and so has very different potential legacies. So yeah, you can take McCalloch into logical gates, but you can take it also in brain networking for sure. Yeah. [01:56:41] Speaker B: All right. Andrea, you have a lot of stamina, a lot of patience with me to take us in these. I take these orthogonal turns frequently, but like I said, you always brought us back. I really enjoyed the discussion. Keep up the good work and thanks for talking with me. [01:56:54] Speaker A: Yeah, thanks a lot. Was a lot of fun and. And yeah, thanks a lot. [01:57:05] Speaker B: Brain Inspired is powered by the Transmitter, an online publication that aims to deliver useful information, insights and tools to build bridges across neuroscience and advance research. Visit thetransmitter.org to explore the latest neuroscience news and perspectives written by journalists and scientists. If you value Brain Inspired, support it through Patreon. To access full length episodes, join our Discord community and even influence who I invite to the podcast. Go to BrainInspired Co to learn more. The music you hear is a little slow jazzy blues performed by my friend Kyle Doniphan. Thank you for your support. See you next time. [01:57:48] Speaker A: It.