Do you have free will? - Part 1

Hello everyone. Thank you for joining us for this episode of Simplifying Complexity. The question I have for you is, did you decide to join us out of your own free will, or were you compelled to do so by the mechanics inside your brand? Well, talk about that today. We are joined by Kevin Michell, Associate Professor of Genetics and Euro Science at Trinity College, Dublin. Now Kevin is going to take us on an incredible journey over the next two episodes. We're going to go all the way back and start with single-celled organisms. And then we're going to follow the path of evolution until we get to ourselves, human beings, an organism that has an ability to reason in a way that no others can. And hopefully, by the end of it all, he'll be able to answer the question whether you picked this podcast because you wanted to or because you were compelled to. This is Simplifying Complexity, a podcast where we explore the underlying principles of complex systems, systems that seem to defy our rational view of the world, like economies, ecologies, or even you, or me. I'm forensic engineer Sean Brady and I'll be your host. Kevin, welcome on the show. Sean, thanks very much for happening. So, why do we start with your background? How do you get into what you're into? It's a long road actually, so I did my undergraduate work in genetics, Trinity College in Dublin. We came very interested in development and then went to do my PhD in Berkeley studying development of the nervous system in the fruit fly. So a nice simple organism where we can do really kind of clean genetics where we can look in the embryo and we can see like individual nerve cells that we they have names, you know, where we can see the same one over and over again and we can see it goes this way or that way and we were trying to figure out what are the instructions in a developing embryo that tell the nerve to go this way instead of that way, because some of them go this way. So trying to figure out the molecules that actually wire up the brain and how the instructions for that are somehow encoded in the genome. And then I went to another lab still in California working on the same problem but in mice where the complexity of the structure is obviously much, much higher. It's a little more difficult to do the genetics but we had some nice sort of tools and tricks to do that with the same idea, you know, trying to figure out what are the instructions to let the brain wire up and came back to Trinity in 2002 and set up my own lab working in that, you know, the development of the nervous system but pretty quickly developed and an interest in the genetics of neurodevelopmental disorders in humans. So the genes that we know are responsible for wiring up the brain, if you have mutations in those genes or those that affect those processes, then that can lead to clinical conditions like intellectual disability or autism or schizophrenia or epilepsy and so on. And it turns out the work that we were doing is really basic science, just curiosity driven kind of science, turned out to be really clinically relevant because it's the same processes underlying those things. And so that got me more and more interested in human beings, which I guess many people start there, that's not where I started. And then in thinking about things like say schizophrenia where, you know, you have disorders of perception disorders of cognition disorders of mood and, you know, ability to think in an ordered way, it really made me think, well, I need to better understand these kinds of processes of what's going on. I need to better understand cognition and the mind. And so that opened up the whole sort of area of thinking about what goes on in brains relates to what goes on in minds. And yeah, so a few years ago I wrote a book called innate, which is about how variation in the genes can lead to variation in brain structures and wiring and manifest as variation in things like personality traits, psychological traits that we all have, but also occasionally in, you know, what we call disorders, so mental illness of one kind or another. So the point of that was that we're not blank slate, right? So we are born different from each other. There's some kind of program in the human genome. And I use the word program very loosely that specifies making a human being with a human brain. But we all have different versions of that program, right? There's genetic variation. And so we all have different versions of the brain. So we're not blank slate. We do have psychological predispositions. And we are in a sense just wired a certain way, or least we come pre-wired a certain way. And people used to ask me sometimes, well, what does that mean for free will? Right? If I'm just wired a certain way and that wiring affects my behavior and it manifests in the way that I behave, then do I really have free will because I didn't choose that? And so that's what eventually drew me to writing this more recent book about free will. So why do you want to write the book? Obviously, you've alluded to explain this. Where do you start and what do you end up with this? I mean, there were a few reasons why I wanted to write it. One is what I just said. It was sort of my own thought process. And that question of do I have free will, I think it bothers everybody to some extent, but I think it really bothers neuroscientists because we're there, like it's sort of in this existential risk that we have in doing the jobs that we do because we spend all our time looking at how the machine works. Right? We can get in there. We can put somebody in a scanner and we see when they're feeling afraid or when they're thinking of serving a tennis ball or when they're looking at the picture of their mother's face, we can see, oh, well, this part of your brain lights up or that part or that part. And that's why you're having that experience. Right? It's just this part of your brain is active. And, you know, in animals, we have amazing technologies now where we can go in and we can control, we can activate sets of neurons and cause the animal to do something. Right? So we can cause it to move in a certain way. We can cause it to try to hunt or try to mate or try, you know, or go to sleep or we can, you know, implant false memories in there. We can change the way that it's thinking. We can up or down regulate its confidence in a decision, all sorts of things. So we can really get in there and mess with the machinery of cognition. And the danger of doing that is that it almost makes it seem like it's just machinery. Right? The mental parts, the sort of subjective phenomenal experience that we have of being a self is just an epi phenomenon. It's not really doing any work. All of the real work, all of the real causation is down there at the level of neurons because I just thought that we can just stick in an electrode in there and cause things to happen. So that must be the case. But, you know, I think that's a simplistic line of reasoning, but it's hard to get away from it. So partly, my motivation was for myself to understand how an entity, an organism can act as a self where the whole thing has causal power. And it's not just being pushed around by events that are happening within it. So it's not just some neural firings in my brain right now that are making me say these words. I'm choosing to say the words. It turns out not to be so easy to think about that or to figure out how where that holistic kind of causal power can actually come from. So take us on a journey. Where do you want to start to bring us true, the story of your book? Yeah. Well, you know, I kind of lay out the worries about free will and there's a lot of different concerns that people have had over how could we possibly have free will that date back millennia? You know, often one of the big ones is that free will is set against determinism. The idea that things in the physical world are just happening according to physical laws. And because we're made of physical stuff and you know, most scientists just accept this. We are just made of physical stuff and there's no magic, right? There's no magical spirit kind of thing. Then isn't the physical stuff within us just regulated by those same kind of physical laws? How could me having a thought sort of transcend all of that physical causation? And especially if the laws of physics are in a sense closed where the equations of physics just have one solution all the time. You know, if that's true, then we're in trouble. And so there's lots of questions like that. So very basic questions. And then there's other questions specifically in humans where it turns out a lot of the debate around free will is not about whether we can control our actions or cause our actions. It's more about whether we can be held morally responsible for those actions. And to me, that gets into a whole different area of thought about where morality comes from, what that means, what does responsibility mean, how do we actually exercise that culturally, conventionally and socially? And so I wanted to sort of disentangle that and leave those questions till the end. And just to ask a much more fundamental question, which is in a world where physics dictates what happens, how can an organism be said to do anything? How can there be such a thing as an action that you attribute to an organism, a whole self, not just physical goings on within it? And that's just much, much more fundamental than anything to do with humans that applies to any kind of organism. And in fact, my argument is that if you want to understand the very, very elaborate systems that we have for controlling our behavior, The best way to do that is to start with the simplest things, understand how they control their behavior, and then scaffold our understanding from there, and that scaffolding naturally follows an evolutionary line. We can just follow how evolution elaborated this and made it more sophisticated. That's the scope of the book. And really, it starts with, it's probably surprising, I guess, to some readers, because it starts that journey really all the way back with the question of, what does it mean to be living organism at all? I'm saying we have to figure out how organisms do something. Well, what is an organism? What is life? And so the question of free will and agency and life, to me, are all part of the same sort of mixture of questions. And if you think about the very origins of life, or what we know about it, then really, single-celled organism is a site where there's lots of complex chemical reactions going on, in a way that maintain themselves, that maintain the whole reaction. So they're all interlocked with each other in a kind of a regime or a pattern that just is the thing. So a little bacterium just is that pattern of chemical activity through time. It's not the physical stuff, it's the processes that are happening, it's the dynamics of it that make the thing alive. Once those dynamics stop, it's dead. Same with you and me, right? If our metabolism stops, we're dead. So life is a dynamical process. And the tricky thing about it is that it keeps going, right? I mean, in a really unlikely way. So if you think just from the second law of thermodynamics, which says things should just get disordered, then why do these living things stay ordered? They stay in this pattern through time. And the reason is they take in energy to do that. They do work. They have to work at it in thermodynamic terms. And so ultimately, we want to understand how does an organism make something happen? And actually, it's grounded in the fact that organisms are constantly, continuously making themselves happen. That's just what it means to be alive is to keep making yourself happen. So that's where we start. And I think if we can grasp that, we're already miles away from simple physics. We're already got kinds of causation there that aren't at play in the non-living world. Because in a sense, we've got purpose. Because the things that are persisting better, the kinds of configurations that you can get that are better at persisting, just persist longer. They stay alive longer, then they can evolve. And so what you end up with is systems that are evolved in order to better persist through time. That's fascinating. We had to think about it. Sarah Walker on the show talking about our origin of life before. And I use it in that it's almost like the fact that we have a continuous process of reactions happening. That's a form of order. And does that mean automatically that's a form of life? Or does it have to be self-sustaining and taking energy? Is that enough to separate? Or I think Sarah was coming from the place of going, once the thing uses information to respond in a way, that's the beginning of life. What's your thoughts on that? It's probably hard in the examples we have on Earth to separate those things, right? Because actually, what I've just been talking about is the system that's trying to take in energy and keep its metabolism going, maybe even replicate and reproduce through time in order to do that in an inhospitable world that's constantly changing. Those organisms have to be able to take in information. So just because they're using up food, for example, or they're producing too much waste products or they're producing too many progeny in the place of getting crowded, right? So the environment is changing partly because the organism itself is making the environment change. But it needs to be able to detect things like, where's their food? If you used up all the food here, where's their food in the environment? Or where's their threats? Where's their chemical toxins or temperatures too high or whatever? So it needs to be sensitive to things in the outside world. Let me back up. The first thing it needs is a barrier. It needs to be bounded by some kind of a membrane or a cell wall that distinguishes the thing from the rest of the world. So that's the first bit. But once it has that, it can't just stay cocooned in there and ignore everything. It has to be able to get information about what's out in the world. So most simple cell things will have these antennae, protein receptors sitting in the membrane that will chemically bind something on the outside of the cell and then give a little wiggle inside the cell. There's had a little conformational change inside the cell that just is a signal that something is out in the world. And again, it's a kind of a causation that's not a physical, there's no physical oomph there. There's no force that's happening. It's just an informational thing. It could be in one confirmation or another, depending on what confirmation it is in that information for the cell. And then the question is, okay, well, what does the cell do with that information? So then it needs to have a system which is not just a sensory system, but a control system. It needs to be able to control either, say, regulate its own physiology. So for example, if there's one kind of a sugar out in the world versus a different kind, if there's oxygen versus no oxygen, you know, many species can regulate their metabolism accordingly without moving around, right? That's just an internal reaction to circumstances. But the other option is just to move, right? If there's no sugar here or there's no oxygen, you can just move to a place where there is some. And so evolution can wire in these control policies in engineering terms, where when a bacterium senses a sugar, it will naturally move towards the sugar, right? So that's the first glimmers of what I think we can call action that's a doing of an organism, not just a happening. Yes. I mean, what's interesting there is, what's the difference between, you know, we could build a robot to do that. You know, they take facing inputs, process them in this way. But is there a fundamental difference between that and what we're talking about here? There are a few differences. One is that the living thing itself, as we talk about already, is just trying to stay alive, trying, putting, trying in, in scare quotes, right? But it's doing work in order to stay alive. And most robots are not like that, although presumably you can make one that is like that. And I think it's really interesting to say, you know, if we look at what we see in living things and we implemented them in artificial systems in the right way, would we actually have a living thing? And, you know, I actually don't see any reason why we couldn't make a living thing. I don't think it would be easy, and I don't think we're close to doing it. But in principle, I think that we could. But the other thing that's really important here is you could build a robot that senses this one thing and has a reaction to it. So like a little reflex sort of circuit, right? And the way I just referred to the bacterium, you can think about it that way. And in fact, a lot of the experiments that we do in the lab, so when people are trying to figure out how does this process of what's called chemotaxis in a bacterium work? How does it follow a gradient towards a food source? Well, you can isolate your bacterium, you can put it in an environment where there's nothing else, but that food source. And when you do that, you can see that you can find the receptor that binds to it internally. You can see the proteins that transduce that signal and all the way to the bacteria, have like a little outboard motor called a flagellum. It spins around and lets them move. And so the direction of movement of that controls where they go in the world. So there's a nice linear pathway of components within the bacterium that are activated. And it makes it look like the bacterium is being pushed around by those parts within it. However, that's a really, really unnatural perspective to take on things because nature does not present itself so accommodatingly to a bacterium like one stimulus at a time. There's all kinds of stuff going on. There's all kinds of information in the world that the bacterium has to integrate and it has to integrate it relative to its own internal state and its recent history and how many cells are around in the world and what the temperature is and what the pH is. And so in effect, what you're seeing is actually a whole system response because in the lab, we take away all that context. We ignore all the context. We reduce it to this one thing and we say, "Look, I have found the cause of. " or the mechanism whereby cells do chemo-taxes. But they don't always do it that way in the real world because there's all this other stuff going on. Like that's one of the simplest behaviors we know about in the world. And even in that behavior, it's very context-dependent. It's very holistic. It's very integrative. It's not completely passive. It's not just a reactive system. The system is endogenously active all the time. And it's accommodating to incoming information. Lots of different streams of incoming information in such a way as to mount the most adaptive response to it. But it's never just sitting there, not doing anything. It's always this endogenously active, dynamically active system that is making its way in the world in a holistic sense because it's trying to stay alive in a holistic sense. Its proteins aren't trying to stay alive because they're not alive. It's the organized pattern that's alive and that's what's been selected for. It's the organized pattern. So then would you say that the bacteria has free will? I wouldn't. Now partly just conventionally free will. I just want to reserve that for humans. It's such a loaded term. But I would say I prefer the just the term agency, the ability to act as a holistic being. Yes, I would say a bacteria has that ability. Now, it's a very, very basic kind of level, right? It's not aware of anything. It has some context sensitivity, but it's not very sophisticated. It's like the most basal level of agency that you could think of, but it gets us off the ground. You've got purpose, you've got information, you've got value. That is, some things are good or bad. It's a good thing to move towards food. It's a bad thing to move away from food. And then in a sense, you've got a kind of a grounded meaning. Things are meaningful to the organism because they're relevant to its survival. And they can be acted upon as information in an adaptive fashion. So by grounding those concepts in that simple way, when we ultimately start to think about the much, much more elaborate forms of control that organisms like you and I have, we're going to rely on those ideas. And we have them in a nice, clear, sort of less complicated scenario where we can build from there. Just come back to one of the things you say, as part of that. The cells themselves, are you saying the cells themselves are the pieces of the bacteria, shall we say, the bits don't have agency in and of themselves. They are quite mechanical. But when you put the organism together, we get this emergent property, if we want to put it in complexity terms, that is agency, which is it wants to stay alive. Yeah, that's the way that I would put it. Now, there's some other people who would argue differently. I don't know if you know Michael Leven's work, but Mike would say that maybe even you could think about the little proteins as little agents and they have little jobs to do. I don't find that particularly useful. So I think agency is, as you say, an emergent property of the whole system. And partly because, like I said, it's selected for at that level, the bacterium as a whole is the locus of fitness, not any of its parts. In evolutionary terms, what I mean is that the way that the whole thing is configured is what evolution is having a say on. That's what evolution is judging. It's like how well does the whole thing survive and reproduce? It doesn't care about any of its parts. It doesn't see any of its parts. So it only is at that high level. So it's a locus of fitness. It's a locus of concern in the sense that some things in the world are good or bad for the bacterium as a whole. And it's a locus of control because the bacterium has to be able to make adaptive choices in the sense that there are many things that it could do. And it has to pick one of them. It can't try to do a aerobic and an aerobic respiration at the same time. It has to pick one of them. It can't choose to go left and right at the same time. And it has to go one way as a whole thing. It can't leave some of its parts go one way and some of its parts go the other way. It has to be coordinated. Is now the right time to talk a little bit about evolution? Because that's a fascinating statement that, and I mean, it makes so much sense when you say it, but I've never thought of it like that, that evolution cares about the whole thing, the system. It doesn't care about the bits of the system. So is evolution essentially rewarding the emergent behavior of the organism in a manner that they're making it onto the next generation, and we're evolving to better fitness. Yes, no, I think you said it really well there. Actually, you can kind of think about as a reward function. Of course, it's a very different kind of algorithm from the sort of systems we see in artificial intelligence that use reward functions. But basically the power of evolution comes from the fact that it's massively parallel. What I mean is that it doesn't work on single organisms. It works across populations. We've been talking about our little one little bacterium that's trying to stay alive, but of course, it's in a big population where there's been loads of division. There's all kinds of bacteria there, and some of them will have mutations that affect how well different parts of the system work. Now, most mutations in an organism are bad because evolution has already done a job of optimizing all of its systems for whatever environment it's evolved in. But occasionally, you'll get a mutation that gives a little advantage, or a combination of mutations that gives some advantage, maybe in a new environment that wasn't the one that was selected for it, but now in a new environment, say there's a new food source, a bacterium that couldn't digest it before now, couldn't metabolize, I should say. So then that bacterium is going to grow a little faster. It's going to divide a little faster. It's going to take over a little more of the population with its descendants, and then the mutation is going to spread through the gene pool of that population over time. So that's basically how evolution works. But like you said, the important thing is evolution is not a knowing actor. It doesn't know anything. That's just the name that we give to the consequence of this tendency. It's not a force. That's just what happens. So nobody knows anything about what's going on inside these organisms. They're not saying, "Oh, this one's better because this or that." It's just mathematically who survives longer, who breeds faster, who has more offspring? That's it. To stay with that for a second, that means that when we talk about the organism has agency just stay alive. That's because evolution is rewarding, they're staying alive. Would we even say that the organism is trying to stay alive in any sort of, it's not even in a thoughtful way? Not at all. It's purely like I'll be here when someone else isn't as a bacterium. So that's it. Evolution, it just starts with a totology, which is things that persist better, persist better. That's it. That's it. It sounds stupid. To say it like that, but that's the whole gist of it. That's true even before we got actual reproduction, even before there's genetic material involved. You can just have one sort of configuration of chemical reactions that's quite stable because they're all interlocking and supporting each other. And then another one that is stable for a while, but maybe a bit precarious and falls apart quickly. And if you look through time, you're going to see the first one and you're just not going to see the second one because it's just not going to stay alive for very long. So now we've got a bounded organism for one of our world. We've got an information sensing ability. We've got a decision processor in there and then we can take some action. Where do we, and evolution is rewarding the mutations that make it survive longer or are making more robust? Where do we go from then and the story of agency? Yeah. So what you could say about bacteria is that in a sense they're doing things for reasons, but they're evolution's reasons. Evolution has packed in these control policies, the very direct ones, but also the context-dependent relations such that that configuration responds well in the environments that these bacteria are normally encountering. What happened next, of course, in evolution is, well, after some long time period, is that we got multicellular organisms emerging. So ordinarily, when single cell things divide, the two cells go off and become single cell things, but of course, what happens in multicellular organisms is that they stay together. They stay together and they make a big thing like you or me. They make an embryo and then the embryo develops into a mature organism of all the different types that we see. Now, the problems facing a multicellular organism are the same as those facing a single cell organism, which is if they want to stay alive in a world that's changing, then they need to know what's out in the world and what should I do about it? So they need sensory systems and they need control systems. And in order to know what's a good thing to do about it, they also need knowledge. So what we get is the emergence of systems. It's really the reinvention of systems on a multicellular scale that we're doing the job in a single cell. And that system is the nervous system. So we have sensory organs. They connect with nerves to the brain or to the spinal cord. And then we have that the nervous system just is a control system. Its job is to control our behavior, to regulate our physiology and to do that in an adaptive way, depending on what's out in the world and what it knows about the world. So it's not just a pre-wired control system. It's a learning system. And that's where the real power comes from. That's the real transition from simple basal agency to a much more individual agency where what an organism chooses to do is inform not just by evolution, pre-wiring these control policies, but by its own experience. So it has learned through its own experience that in some scenario, if it does A versus B, that A turns out well and B doesn't. So it should do A again. And that's from its own lifetime. So a couple of things. So it's a bacteria you don't learn. Is that a fair statement? They have some very short term biochemical adaptation to past conditions. So they do carry a little memory of what was just happening like a second before or two seconds before. And in fact, they have to because that's actually how they tell if they're going up a gradient of a food. They're not measuring the difference between the front and the back end. They're measuring the difference through time. So it's like, okay, what's the concentration now? Is that higher or lower than it was a few seconds ago? So they carry a very short memory, but they don't learn in any real sense. And there's a lot of different kinds of learning that we do, for example, you know, we learn skills like riding a bike where that's called procedural learning. So we get better at controlling ourselves, right? And in a sense, that's the nervous system learning how to predict the outcomes of its own activity. Yeah, right. So it tries some activity, you fall over, try some different activity, you fall over again, and it keeps trying until it learns, no, this is the way. This is the activity pattern I need to produce in order to ride the bicycle, right? And then of course, there's, sorry, there's other kinds of learning as well, like learning about events, learning about categories and associations and so on. Okay. So then we go to multi-salutor and is the defining feature really that you do need to connect up those cells or else they're just a bunch of single cells together? And that's where the nervous system comes in. Well, so probably the nervous system evolved in very simple creatures where you just need to coordinate movement. So you have simple creatures like sponges and hydras and things like that, little marine critters that have very simple, what's called a diffuse nerve net, right? They don't have any concentration of cells into a brain. They just have this spreading net of nerves connecting all of the cells. And the cells of those creatures are kind of like muscles and kind of like skin cells. So their cells that can contract, but you need them to contract in a coordinated fashion for the whole thing to do something, right? So you can't have some bits contracting here and some bits contracting here and then it would just be a squishy mess. If you contract them all in a coordinated fashion, you can for example, change the shape of the whole organism. So you need a system to do that once the organism gets big enough, you have to be able to physically coordinate the movement of the various parts of the animal. And that's probably where nervous systems evolved first was from this skin/muscle epithelium. And then you got some electrical cells that were good at telling everybody what to do. And then eventually, you know, they connected up with sensory organs and then, you know, at later times you have this condensation where you get these intervening layers between sensation and action. There's this intervening layers that are doing lots of internal processing. And that's what we call a brain. And that's where you get away from simple reflexes that are isolated. And instead, you're integrating lots of information. And crucially, you're making judgments on it based on what you have learned in the past. So that's the end of part one. We'll see you in part two. Thanks for listening to simplifying complexity. When we look at the key concepts of complexity science with expert minds from across the world, concepts like emergence, self-organization, adaptation, networks, scaling, tipping points, and much more. This podcast was produced by Brady Hewood and Wevelyn Creative. To make sure you don't miss an episode, be sure to subscribe to or follow the show in your podcast app. I'm Sean Brady and I'll see you in our next episode.