185 | Arvid Ågren on the Gene’s-Eye View of Evolution

0:00:00.0 Sean Caroll: Hello everyone and welcome to the mindscape podcast. I'm your host, Sean Carroll. Remember when we talked about the social construction of reality, wasn't that fun? We contrasted this human-scale way of thinking about the world, the social world, where you make up categories, and then they're based on what really happens. But there's a lot of freedom there. We contrasted that with the physics level with the electrons and the photons, where it's pretty clear how you want to think about the pieces out of which reality is made, but there are a lot of layers in between the physics level and the social level, the human level right. There's all of biology, for example.

0:00:37.8 SC: Today we're gonna dig into a puzzle or at least an important research area within biology that involves exactly this, what is the right way to conceptualize the world of species and populations and Evolution, famously Richard Dawkins in the 1970s, popularized a view known as The Selfish Gene, back in the old days and the beginning of the theory of evolution, you might have focused on individual organisms, and imagined these organisms wanted to reproduce their own genomes, they wanted to have kids and have their genetic heritage be passed on to future generations. Then came the math, then came a way of thinking about population genetics that pointed out that you can pass on the genes inside of you, even without you being involved, if all of your relatives are very good at passing on genes.

0:01:31.7 SC: So it became not only Dawkins, but other people pointing out that it's as if it's the genes that wanna pass on their heritage, not the individuals carrying them, in fact, you could say that the individuals are just kind of a bus full of genes, they're a vehicle that carries on the genes because individuals die, organisms are born, they have a life; they die less than a century for most species, whereas the genes can live on a very, very long time being passed from generation to generation. So surprisingly, not everyone agrees, not so surprisingly, of course, this is a complicated thing, what is the best way to conceptualize levels of selection, this is what we call this particular debate in biology, is it at the level of genes of species of populations of individuals, and it's exactly, this kind of...

0:02:14.5 SC: How do you divide up reality kind of question, the answer from our guest today, Arvid Ågren, who's worked as a biologist on this problem is he's actually pretty pro The Selfish Gene view, he calls it the genes-eye view of evolution, but he completely admits that it's not the end-all. Biology is more complicated, as we've often found out here on Mindscape than something easy like physics is, there is leakage between the levels of genes and molecules up to organisms, up to populations, so there's insight to be gained by thinking about things from the Selfish Gene point of view, there's also insight to be gained by thinking about higher levels of abstraction, it's a wonderful real-world example of the philosophical questions that we have about emergence, complexity, how to chunk the universe up into pieces that we can analyze and theorize about.

0:03:18.3 SC: So I think it's a great kind of a discussion, it's very Mindscape, as I make the joke in the podcast, both me and the evil Sean Carol, the biologist, who was a previous Mindscape guest are referenced in Arv id's book, so it's a very natural kind of conversation for us to have here. Let's go.

[music]

0:03:51.5 SC: Arvid Agren welcome to the mindscape podcast.

0:03:53.8 Arvid Agren: Thank you for having me.

0:03:55.3 SC: So I think we can assume that most of the audience is familiar with Darwinian evolution theory, natural selection, and maybe also a little bit of Mendel and genetics, but let's start with the groundwork at the synthesis of these two great ideas, the modern synthesis. It still boggles my mind a little bit that Darwin didn't know about genetics. He went pretty far without knowing about that, but tell us how that came together in the 20th century.

0:04:20.6 AA: Yeah, so this is really is one of the striking moments in the history of biology, what has become known as the modern synthesis of biology in the 1920s, 1930s, and as you say it is came together by bringing two different insights into one theory. On the one hand, it was Darwin's theory of evolution, by natural selection, a theory about individual organisms, organisms vary in how well they survive and reproduce, and if any of those traits are heritable, they will become more common over time. Now, the big problem for Darwin then was as you say that, he had no functioning theory of inheritance, and while he had some high ideas of his own, none of them really came to work and that remained a conundrum for him and is viewed as a weakness for a long time, of his theory. This was in parallel then with the second insight that was bought through, and that was a function theory of inheritance, and that was provided by Gregor Mendel's work where the realisation that inheritance function with this inheritance of discrete particular entities which we now refer to as genes.

0:05:39.4 SC: And so if we didn't know about genes and we didn't know about how discrete they were, clearly there is some degree of inheritance, everyone can see that in hair color and things like that, but you might have thought that things would just blend together every time, and then you would never diverge, everything would just become some kind of homogeneous, unified organism or something like that, in terms of sharing the same equilibrated characteristics, but with discrete units, all sorts of funny things can happen.

0:06:07.2 AA: Exactly, and indeed that Darwin probably believes in is sort of blending inheritance and that the realization that you will quickly run out of variation for natural selection to act upon was a kind of a serious issue for him.

0:06:21.2 AA: The problem then with his new theory of inheritance in the form of discrete entities was that how could that be reconciled with a gradualism of, on the one hand, variance as we see it in the natural world, but also with a gradualism as envisioned by Darwinian evolution, like the gradual change over time. And this is really what the emergence of what we know referred to as population genetics. The idea that you can describe evolution as changes in certain variance of genes or allele frequencies over time, so that is what, kind of what evolution almost has come to be defined as. If you open an evolution textbook today, that the evolution is change in allele frequencies in a population over time. The change is inheritable...

0:07:05.6 SC: I guess I don't quite see how that answers the question of the apparent smoothness of variation. I would have just said, the variation is smooth, 'cause there's a lot of genes and you can just change one at a time, it looks almost like it's continuous.

0:07:18.8 AA: And exactly, that is what the solution that population geneticists frowned strongly upon, and particularly the English statistician and evolution geneticist, Ronald Fisher, showed that you could get the appearance of gradual variation from the... If a trait was caused by many alleles with small effects, you could get... You can reconcile the gradualism with the discrete inheritance. And that really represented his major achievement that was part of... And it is generally a synthesis, or bringing together multiple parts of biology into one, where population genetics...

0:07:52.8 SC: I'll ask you... And I think you already explained it, but I'm gonna ask you again to explain the word allele because it's not exactly the same as the word gene, right?

0:08:00.6 AA: It is not, no. So allele is like a version of a gene. So for example, we may think of a gene for, say, eye color, but then you can have multiple alleles... So one for blue eyes, and one for brown eyes, for example. So while we have... We can say, we have one gene for that. As a human, we have one allele from our father and one allele from our mother that we have inherited.

0:08:22.7 SC: So I think probably the person on the street would talk about a gene for blue eyes, or a gene for brown eyes... But really, there's one gene for eye color, and there's an allele for brown eyes or blue eyes, yeah?

0:08:33.7 AA: Yeah. And then in most of the cases, you have many genes with multiple alleles underlying most traits most of the time.

0:08:40.8 SC: Okay, very good. And what is a gene? [laughter]

0:08:46.0 AA: Ah.

0:08:47.0 SC: I ask everyone, I ask every biologist this question, they always get uncomfortable when I ask it. [laughter]

0:08:53.8 AA: It's a practically little word in biology that has come to mean quite different things depending on what corner of biology you are from. In the theoretical population genetics, it is simply something that can be stably inherited something that you can say effects a trait. To a molecular biologist, it has a much more physical or materialistic definition of it, that is, the sequence of DNA that often has a function or encodes a protein or RNA that does something. And indeed, this multiple definitions of genes within biology has often led to misunderstandings and disagreements.

0:09:34.2 AA: So I guess even when we had Mendelian genetics, Mendel was still in the 19th century, but then it was glued on or incorporated into evolutionary theory in the 20th century... But we still didn't have DNA at that point, right. So happily, we have DNA now, did that change the way we thought about this a lot or did it just get... Serving as an underpinning for the existing discussion about genetics?

0:10:00.5 AA: I think, in the first approximation changed a lot in what we think or what genes are, we can study it in a completely different way, we have a sense of the number of genes that a species has and how they evolve. And at the same time, it is important to remember that population genetics as an approach as a field, as you say, emerged before we knew what the material basis of heredity was, before we knew that it was DNA, before we knew that DNA was a double helix. And those mathematical models that we still use, in evolutionary biology, were developed before that, before we had any idea of those. And there you rely on a definition of a gene that's completely agnostic about the material basis. And I do think it's rather remarkable that they work so well, despite being completely ignorant about what the material is actually is.

0:10:49.2 SC: Except that it let us invent a new meaning for the word gene, right? So you had the word gene, and then we found DNA, and then they invented a new meaning for the word gene, namely, a sequence of DNA that encodes for a protein.

0:11:01.3 AA: Yes, yes... I mean, the population genetics gene, as we use it, still in population is much closer in a way to the original Mendelian gene as it was originally. But it is certainly one of those words, one of many in biology that had similar enough meaning that people can think that they are talking about this same thing, but they are different enough to cause frustration.

0:11:23.6 SC: Well, you've written a book called The Gene's-Eye View and what do you mean by the word gene in the title of your book? Or does it not matter?

0:11:31.0 AA: So The Gene's-Eye View uses the word gene in this much more old fashioned way of simply defining it as something that's stably inherited, a part of a chromosome that's stably inherited across a generation. And this means that it can be almost of arbitrarily long a gene, it can be anything that's inherited together, which means that the parts of the Y chromosome, for example, that never recombines with the X can be thought of as one gene or large swaths of chromosome or indeed, very small part of a chromosome is also thought of as a gene here, but it is this definition, agnostic about any sort of material basis.

0:12:15.7 SC: And how do we know... How much do we know about the genetic composition of let's say, the human genome? Does everyone agree that it's divided into so and so many genes? The human DNA?

0:12:29.8 AA: Just as an approximation, yes. And I think everyone is in the same order of magnitude. You may... I don't know exactly what the latest number is. When I was taught it, it was somewhere between 20... And 25,000 genes. I think someone who...

0:12:44.4 SC: With the latest updates of the human genome, you're gonna find a much more precise number. And I think I'm sure experts disagree but to first approximation, I think everyone agrees that this is the number.

0:12:55.4 SC: This is all off-topic a little bit. But since I have you here, I'm just satisfying my curiosity about a bunch of these questions. Do we have differential data about the persistence of different genes over time? Are some genes very malleable over evolutionary timescales and others stick around?

0:13:13.7 AA: Yeah. Now, certainly, we know that... We have something called pseudo-genes, which are genes that are still around, are still recognized as genes, but they've lost their function in humans. So for example, my favorite example is that we have the gene that allows you to synthesize vitamin C, from scratch, which is present in most mammals.

0:13:37.5 AA: Yeah.

0:13:38.2 AA: So we have that gene, it's clearly recognizable, but doesn't work.

0:13:40.6 SC: We don't do it.

0:13:41.3 AA: Lost the function. So it's in us, some of our primate relatives, and I think in fruit bats that have lost that ability and carry this around. So they're around, as in selection hasn't gotten rid of it, but they're also certainly, example of where, if you compare species that are distantly related, you can see that they've retained some genes in common, but sometimes have lost genes or gained other genes.

0:14:06.0 SC: So in my naive physicist's way of thinking about things, I would imagine that in any stretch of DNA, there's a chance per unit generation that a mutation happens. And in those stretches, that really matter to our functioning. If a bad mutation happens, it gets weeded out it gets selected out of the population. And in places where it doesn't matter that much, those can just continue to mutate away. So is it true that certain stretches of DNA maintain their integrity longer than we would think because they're really, really important to us?

0:14:41.8 AA: Yeah. And indeed, that is often have, the heart of the methods that are used to estimate how much natural selection is acting across the genome. So if you think of all the DNA in an organism, we can ask questions like, how much of this is under constant natural selection? And how much of it is evolving just by chance, or neutrally, as we say, which is gonna be a function of things like mutation rate and population size, essentially. And this is something like how much... Questions, how important is natural selection versus evolution by randomness? So genetic drift used to be a quite contentious debates, in population genetics, but one that has matured a lot in light of the abundance of data that we have now. And it's much more empirically informed, which is much more insightful now, 'cause we should simply go out and ask how much of this genome in this species is on the selection and so on, which we couldn't before?

0:15:35.9 SC: And let me guess, the answer is that some parts selection is really important. And other parts, it's not.

0:15:41.1 SC: Yeah. And in some species, most of it is the selection in some large [0:15:45.2] ____ sorts is not, and...

0:15:47.4 SC: And am I right to think that maybe in the back of our minds, when we think about selection pressures, we're thinking about some species going out there and hunting and becoming faster or stronger, and therefore competing, but really, a lot of the selection pressure is just that if you get the wrong mutation, you die, or you're stillborn, it's not really that detailed capacity to get food or have sex or anything like that.

0:16:11.7 AA: Yeah, I think that represents obviously a crucial part of fitness for any organism, but there's so much more going on in the lifecycle of any organism, a lot of it is that we can't see or it happens at a molecular level. And often those machineries are evolutionarily old and often quite sophisticated in keeping check that mutation rates are not too high or to repair mechanisms are in place.

0:16:37.6 SC: Well, I guess so is it it... Again, my naive physicist's thing is every base pair would have a random chance of mutating, and then the selection acts upon those mutations. But is it... Are there even more sophisticated repair mechanisms than that? Are there ways biologically or molecular biologically, that the really important parts of DNA are protected from mutation?

0:17:03.7 AA: That I don't dare answer that in a confident way how that varies. Certainly, selection is stronger on important parts of the genome. And I think for example, that we carry around this gene for vitamin C where you can kind of accumulate almost any sort of mutation without selection, doing anything about it. How this... The efficacy of the DNA repair mechanisms varies across the genome. I must say, I don't know.

0:17:33.1 SC: Okay. I'm just wondering there's, in quantum computing, we care a lot about what we call error correction, right? Or even Classical Communication Theory. I can imagine that it would be evolutionarily advantageous to develop a cellular biology level mechanism that would just say, "No, no, I'm not gonna let any mutations happen in this part of the genome." But I have no idea whether or not that's actually true.

0:17:58.6 AA: Yeah, and now I'm curious too.

0:18:01.6 SC: Okay, good. We'll bring you back if you invent one. Okay, so I think that I completely interrupted you multiple times there while you were trying to explain the modern synthesis. What is the modern synthesis at the end of the day?

0:18:12.5 AA: At the end of the day, the modern synthesis was, in a way, the founding of the field of evolutionary biology as we know it today. It is an event that brought together rather quite disparate fields of biology, everything from Paleontology, to Ecology to Genetics, into one cohesive science, the science of evolution. On the one hand, it was kind of this crucial event of showing that Mendelian inheritance and Darwinian evolution are compatible and part of the same process. But in many ways, you can also think of it as a big institutional event. That's when the Society for the Study of Evolution was founded in the journals and the starting of the emergence of evolutionary biology as departments in university and so on. So the modern feel of evolutionary biology was really born in that event, as well.

0:19:04.0 SC: It's interesting, I bet you don't always see, I'm just saying, again, I'm completely speculating. I wonder how often I should say you see a parallel evolution of scientific knowledge and institutional support or institutional organisation for thinking about it. So you're saying that in the case of evolutionary biology, they went hand in hand?

0:19:26.7 SC: I would think so. I've recently come around to this view of that part of the history of the field of evolution, and the increased importance of the institutional aspects of building a new field. And especially that this was... So this cropped up in the 1940s so is just before the emergence of molecular biology and in some ways, you can think of this is almost as assault on old fashioned Natural History Organismal based biology, but it's more modern science, which was considered to be the future. And the recently deceased Eduardo Wilson at Harvard discussed this very well. And he's... When he was hired at the department, was then the department of biology at Harvard, around the same time as Jim Watson, the co-discoverer of the double-helix. And their battles of what biology ought to be like. And Wilson described it all rather nicely in his autobiography, Naturalism, and that combined with that where we recently had the 75th anniversary of the Journal of evolution or the evolution journal. So that made me reflect upon that part and the importance of people like Wilson.

0:20:41.1 SC: Yeah, it is really fascinating, because, of course, we hope that science gets at some true facts about nature, but we all know that it's done by human beings. And it's not just done by individual human beings, getting support from an institution, having jobs, getting funding, all this stuff matters in a really interesting way. And, of course, the other thing that matters that happened, I think, around that time is math, right? Once you were beginning to be thinking about evolution, in terms of population genetics, population is a number and you talk about the fraction of different genes and suddenly a bunch of equations burst on the scene.

0:21:19.0 AA: It's very much so, and population genetics remain in a way to be the theoretical backbone of evolutionary biology. Evolutionary biology, I would say is part of biology that is perhaps the most mathematicised, and population genetics often borrowed models from physics, statistical physics, and similar models describing changes in allele frequencies from how physicists had described motion of objects. Of course, there are other traditions, evolutionary biology happily stole game theory from political science and economics and incorporated, perhaps even more successfully, as a way of doing things. And they're also other quantitative methods, for example, in what now is quantitative genetics, where you simply describe changes in phenotypes over time, for example, used, in animal breeding, and so on. But I would say the emergent population genetics and during the modern synthesis was instrumental in some ways in making evolutionary biology be viewed as a more exact science and something to be taken seriously.

0:22:23.5 SC: It's certainly true that once you write down equations, even if your substantive content is exactly the same, you suddenly look more respectable, 'cause now you have equations in your papers, you can go, "Oh, okay, now you're doing serious work."

0:22:35.0 AA: Especially when you convince others, you are not to share the same obsession as you are that you have to be taken seriously. That has proven to be quite important.

0:22:43.7 SC: And when we say that there was math, what are the quantities that are being followed by these equations in physics, we have the position and velocity of particles. So what are we keeping track of and deriving differential equations for in population genetics?

0:23:00.5 AA: Population genetics has typically been considered the frequencies of a specific allele in a population. And you can say if you have an allele with these properties, and has, say this effect on fitness, under what circumstances can it invade a population. So can it become more common as the generations go by? And also things you're interested in keeping track of all things like mutation rates, how often does this mutation happen? A specific mutation. Population size, so in large population, natural selection will be more effective, whereas in small population chance will play a bigger role, if it's strongly advantageous, or is what is known as the selection coefficient. So how, how strongly is the selection acting on this specific allele? Obviously I'll come back to the most basic parameters and then you can introduce all sorts of complexities in terms of population structure and relatedness between individuals interacting assumptions about the life cycle whether they reproduce sexually or asexually and how mating is determined and almost any aspect that you're interested in. But in traditional population genetics, these are the basic parameters: Population size, mutation rates, section coefficients, and so on.

0:24:20.9 SC: Well, that's interesting. I think we're, we run immediately into some philosophy of science questions here. Because things like the frequency of alleles, or the mutation rate, those sound very measurable. If I were a positivist, I would say, "That's good, I can see what's going on there." But then you have some model right, some theoretical framework that it fits into. So you already talked about selection coefficients or fitness, maybe explain a little bit more about what fitness is supposed to be, and how empirical are those ideas versus concepts we need to introduce to make sense of our equations?

0:25:00.3 AA: One of the architects of population genetics, the Brit, J. B. S. Haldane once described fitness as a bugger, it's one of those central entities in evolutionary theory. And in many ways you can say we only can have one trait discipline, at the end of the day, that is what matters. At the same time, it is awfully hard to agree on a proper definition and even harder to measure in complex what really matters in natural population. Usually, we think of it as something along the lines of the genetic contribution that a given individual makes to the next generation. So how can the genes in certain individuals, how much they contribute to the next generation? But then we also talk about fitness or specific mutations. So then we talk about the genetic level. We have a long debate whether it's best to define fitness singly, only at the individual level, or sometimes we have notions we also should account for the interactions we have with individuals around us this notion of inclusive fitness.

0:26:06.1 AA: This is the even before we try to measure anything, so if you does this study evolution in a wild, so for example, you wanna compare... If you have say... You have sampled individual plants from two different population and you put them in a third one and see which one have a higher fitness. Well, then you often have to rely on like, made something like number of seeds that they produce, or maybe even something like size and there's always like indirect measures of fitness, which means that it is the central entity that has been a lot of both conceptual debate how we ought to define it theoretically, but also a bit of how do you measure it and what are good indirect measures, because as usually what we have to rely up on at the end of the day, if you wanna study it in natural population.

0:26:52.4 SC: When I wrote my book "The Big Picture", I talked about fitness landscapes a little bit, I talked a little bit evolution, and when I ran it by some biologists, some thought it was fine. Others are like, why are you talking about fitness? This is an outmoded concept, we can't measure this. They thought that it was one of those things that gives you the illusion of understanding, but when you get right down to it, it's hard to really quantify in a reliable way that everyone would agree on.

0:27:15.7 AA: Yeah, I think it's often like to do it well, you would have to keep track of it. If it just not only of the focal individual that you're interested in, but also others in a population in a way that's really quite hard. In the long run, especially if you study anything other than plants or things in the lab, like you can do it really well with fruit flies or microbes in a controlled lab setting where you can keep track of anything, but if... Like I do believe that at end of the day, what truly matters in by all these try to understand organisms in natural populations. It is much harder.

0:27:49.7 SC: Well, for one thing, there are things like floods and earthquakes and asteroids hitting the planet, and it's hard to predict ahead of time what your fitness is going to be in those circumstances?

0:27:58.9 AA: Yeah.

0:28:00.3 SC: Pandemics. [laughter] Okay. But anyway, to get back to the main line of thought here, we've mathematized a little bit of population genetics on the basis of the Modern Synthesis, and what we realize it... Correct me if I'm wrong. My impression from looking at your book and other places is that it's not just about me and my offspring in some sense, if what if you can speak of these goings on in slightly overly anthropomorphic language, what wants to happen is that your genetic heritage wants to be shared and one way to do that is to have kids, but the other way to do it is for your brothers and sisters to have kids, right. And so we get the idea of kin selection and inclusive fitness.

0:28:49.0 AA: So this is kind of an insight that appears [0:28:53.0] ____ starting at the generation after the architects of the Modern Synthesis, so in the 1960s, and at the heart of this is the English Biologist, William Bill Hamilton, who already as a struggling graduate student, writes these two tremendously important papers. One published in 1963 and the others are two part majestic piece, in the following year in 1964. And there, he introduces both concepts, both kin selection, which is not a term that he coined, but it was coined by someone else, but which is simply the idea that selection involves you and the relatives around you, but at the heart of it was this idea of inclusive fitness, and to kind of get a sense of why he thought we needed an update of the basic notion of fitness, you have to think of situations where the classic definition doesn't really work, and there's often happens in the context of social behaviors that we had known for a long time, that in social insect like in bees and ants, lost... You have individuals that are completely sterile and instead devote their life to helping other members of the same population to reproduce to be with the queen.

0:30:04.2 AA: This was recognized already by Darwin who described as is one special difficulty of his theory, and what Hamilton realized that one way to explain this is to think of that you shouldn't just think about the offspring that a specific individual has, but you should also account for other offspring that this focal individual is causally responsible for, and you should add those to our focal individual fitness scale by the relatedness that you have to those individuals, so in a way that if your brother has it, that counts more than if your cousin does it, and so on, and then also has to do this to do the math properly also then have to subtract from this sum... The part of the your or the focal individuals, personal fitness that someone else is causally responsible for and doing these kind of sums is where it can get messy pretty quickly, but that is basically the heart you end up with this quantity inclusive fitness that you can show that individual can appear to maximize that. Then under certain circumstances, inclusive fitness can be maximized, even though this individual does not personally have any offspring, but they can help enough others to do so and then still, you can have the process work whereby natural selection maximizes that kind of fitness... That is a new definition of it. This is kind of what emerged in the 1960s.

0:31:33.2 SC: And does the witty remark from JBS Haldanes that he would lay down his life for two brothers or eight cousins... Right?

0:31:39.2 AA: Exactly, exactly. That is kind of predates actually this... But it's very much... Sums up the basic insight very well.

0:31:48.4 SC: By the way, just for people who might get confused out there too. I wanna mention two things, one is, you mentioned William Hamilton, who is not the famous mathematician, physicist William Hamilton, who lived 100 years before... Who invented Hamiltonian mechanics and the quaternions and things like that. Two different William Hamiltons. Also, there are two different Sean Carrolls, and there's a biologist, Sean Carroll. And when I got your book, I think he will laugh at this. When I got your book, I realized that this is a rare book, given that you wrote about levels of selection and sort of philosophical things that I care about, but also biology and evolution that the other Sean cares about... This is the rare book that might have both of us in the index, and indeed...

0:32:31.3 SC: There is one entry for Sean Carroll in the index, but it refers to two different appearances, and one of them is about me and the other one is about the other one. [chuckle]

0:32:39.4 AA: Oh, is that true? That's great.

0:32:42.5 SC: So it's a very Sean Carroll-centric kind of a topic that we're talking about here. That's what I like to say.

0:32:47.9 AA: Yeah.

0:32:48.2 SC: Okay. So yeah, we get the idea that the motivation for thinking this way, maybe the motivation is that it's true, but part of the motivation is we see in nature, individual organisms acting in ways other than just trying to have the most offspring as possible, and maybe this kin selection or inclusive fitness can account for that.

0:33:10.0 AA: Exactly. That has been a tremendously influential way to think and to make sense of these observations. It is also an approach that has been criticized from the get-go, often from mathematically minded parts of the field and the kind of a criticism that continues to this day, that the way that you... It's not as mathematically robust of a construct as some people with perhaps a more mathematical strong background would like it to be, given the centrality that it has in evolutionary biology. It has survived because it provides a tremendously powerful rule of thumb and it is something that empirical biologists and a few of us can go out and try to guide their experiments.

0:34:00.4 AA: And to cut this long debate short, my impression of it is it's partly what you want models to do in evolutionary biology, how kind of comfortable you are with these kind of mathematical limitations, because at the end of the day, this kind of summing up what... At the editivity [0:34:16.6] ____, or summing up what you are responsible versus other parts of the population, often it becomes prohibitively difficult. In other cases, the most simple ones. Similarly you have to make assumptions about genetics and how strong selection are and so on to make it work. But it has pointed us to something really valuable and that... Perhaps one way to sum up is perhaps the third leg of the kinda Hamiltonian contribution to evolutionary biology is the kin selection, inclusive fitness, and sometimes what are known as Hamilton's rule that a trait will be favored if the benefit times [0:34:53.1] ____ related is greater than the cost, which is kind of summed up in JBS Haldane's "I would lay down my life for two brothers or eight cousins" quip.

0:35:00.9 SC: The cost being that he would lay down his life.

0:35:04.8 AA: Exactly. And that as a guiding principle has been very helpful.

0:35:08.0 SC: Okay, but are you alluding to the criticisms by mathematicians... In that, are you alluding to Martin Nowak and Tarnita and Wilson? They had this paper about group selection and new sociality a few years ago that really stirred things up.

0:35:23.3 AA: It really quite did. And I thought in some ways quite an exciting way. I had just started graduate school when the paper came out, and so it's kind of been with me...

0:35:30.9 SC: Your whole life, yeah.

0:35:31.9 AA: Ever since. It was one of the first papers we read in journal club when I was a new graduate student, and I was really taken by these kind of conceptual debates and really re-invigorated my interest in that part of the field. So they, in some ways, represent the latest iteration of the debate. They are far from the first to make these kinds of criticisms of inclusive fitness. They were made from the get-go, in a way, given the centrality that it has taken, it's not as mathematically robust as you would have wanted it to be, but it's a debate that perhaps has calmed down a little bit the last few years, and I think the field, once the dust has settled, is better off. I think we're more on the same page of what the limitations are and you can have I think a more interesting discussion about when can you live with those limitations and when can you not, and what the benefits of the approach are. And then I think I'm quite comfortable with biologists at the end of the day using different approaches, depending on their temperaments and what they want their model to achieve in a specific situation.

0:36:43.4 SC: But I guess I'm kind of confused a little bit because the inclusive fitness idea, part of what it wanted to do, at least in principle, was to explain not only why certain ants don't reproduce and nevertheless contribute to the life of the colony, but also maybe even altruism in human beings or other mammals and so forth. It sounds like it is quintessentially mathematical, like Haldane's joke is very mathematical... Two brothers or eight cousins... But you're saying that the criticisms have come from the more mathematically inclined. Why is that? Is that explicable?

0:37:23.6 AA: So I think that there are... Part of the concept is kind of that relatedness is gonna be really important, and like kind of two brothers is in a way mathematical, but it's also quite simple intuition, so it led a lot of... Like you can go out and measure what are the relatedness between interacting individuals in this specific population, and the studying that in a specific population in the wild. You can do it probably genetically by comparing... If you look at this group of species that have developed this quite complex social behavior, do they have a population structure of higher relatedness... So you have interaction generally between relatives... Compared to this part or this group of species on another part of the our genetic tree. Do they have less relatedness among interacting individuals? So it has kind of guided us a lot in those kind of approaches, and that kin selection idea is there, I think has been... They are mathematical, but they're not too complex in a way. They just kind of point us in the right direction.

0:38:25.6 AA: Now to get inclusive fitness... So, if I can take a step back, inclusive fitness has, to its proponents... One uniquely good property of it is that it's a property of the individual organism, and it can be modelled as a property that an the individual organism should appear as if trying to maximise. And this is good because then you can kind of treat it as the design maximin or design principle of evolution or natural selection that... And this kind of goes back to some part of evolutionary biology, the key problem that we should try to explain is the appearance of design, that like the fact that organisms appear so perfectly suited for the environment that they live in. That's kind of adaptation or the appearance of it. That is the problem we should try to explain. And you then take a step back and say in order to do that, we ought to have a principle that shows like what should organisms appear designed for? And to its proponents then, inclusive fitness is that, that organisms should appear trying to maximise that, and you can derive a model by which you can get the expectation that that is what organisms "should try to do".

0:39:39.8 AA: And that is like when you try to get inclusive fitness to have that property, that it should be both, should try and be maximised and be the property of an individual organism, in order to get there, the maths is a little shakier in a way. You have to rely on more restrictions, either to get it to work or the assumption... I consider it to be almost the opposite problem, they are too general to really become useful, 'cause they almost end up in this kind of truism where it's always true. But there is where I think even more mathematically minded people have had its criticisms. And I should say, all of these people are at the mathy end of the spectrum within biology, but if you come there from a biologist by training, interested in theory, versus someone who has their original training in say maths or physics and then come into biology in graduate school or even later on, degrees of mathiness.

0:40:34.7 SC: And so again, just to get the landscape as it were perfectly clear, is the alternative to inclusive fitness, or is the alternative to kin selection, group selection, or is it more complicated than that?

0:40:48.7 AA: I'd say it's an alternative. Historically, that has often been viewed as the two contrasts.

0:40:55.0 SC: And what does it mean, group selection?

0:40:57.0 AA: Group selection, again, it's one of those terms that mean different things to different people and that is part of the...

0:41:01.9 SC: Fun.

0:41:03.1 AA: The problem. And it has historically meant two slightly different things, again coming back to how we measure fitness. So by now... So kind of historically, the first way we thought of group selection was in kind of... Whereas now often even by its proponents referred to as naive group selection, this belief that individuals may sacrifice their life for the good of the group, really for the good of the species, and that's often done in a rather unreflexive way. But then you often define group fitness in terms of groups producing new dotted groups, that like you measure group fitness in terms of new groups produced. Today, most models of group selection, so if you just pick up one of the current journals and someone prints out the group selection model, fitness is usually measured in terms of individuals producing other individuals, but that you account for some sort of group property, some part of the group that's not really reducible in a meaningful way into individual-level properties. That is what is now these days is meant by group selection. I can sometimes think that is a little unfortunate that that has become known as group selection, but that is... Me and others lost that debate a long time ago, that this is what they've settled on as a field.

0:42:22.5 AA: But those small things can often account for the same things, and often rely on similar properties that a kin selection model would do. So one observation or insight that in some ways calmed things down in this kind of ongoing debate between inclusive fitness and group selection is that you can show that often they are formally equivalent, that they give the same prediction. And I think that that is a good thing to realise that often they have the same properties. There's an interesting kind of, I think more philosophical point there that often this equivalence claim relies on generating rather abstract statistical models that you get almost quite far removed from the core thought processes happening in order to achieve that kind of equivalence, and so I'm kind of in two minds on this. Sometimes I think, "Oh, this is great that we've shown that they're equivalent." Sometimes I'm a little concerned that this equivalence relies on such an abstract notion and that surely we are interested in what causally happened in a specific event, whether it is better described in terms of inclusive fitness, which is an individual-level property, or in terms of group-level properties, but...

0:43:36.9 SC: Good. Well, that's a perfect segue, because I do want to finally hit the pay-off here, which is that once you start thinking about things in this way, evolution in this way of population genetics and equating my life to that of an equivalent number of other peoples sharing the same genome, that's when you get to the gene's eye view, that's when it begins to maybe make sense that what we should think of as competing in the world of evolution is different genes rather than different organisms.

0:44:11.1 AA: Yeah, I think the origin story of the gene's eye view, which is a perspective that really comes to its own in the 1960s and 1970s, stand on precisely these three topics that we have covered. It's an approach to evolutionary biology that takes adaptation as the central problem that we're trying to explain. It builds on the insight from population genetics that evolution can be described as changes in the allele frequencies. And then the third debate from which it emerges is the one over group versus individual-level selection or what is known as broader levels of selection debate. And in a way, it kind of combines these three into one perspective, thinking that adaptation is what we are trying to explain and then goes about thinking, "Well, when we say that adaptation is for the good of something, what is that something?"

0:45:02.0 AA: And the answer to that then is the gene 'cause genes then are conceptualised as these entities that are passed on intact from one generation to the next, whereas the main alternative organisms or sometimes groups, don't have that evolutionary longevity required to play that role. An organism by this reasoning is a unique combination of its genotype and environment, and their interaction is here in one generation, gone in the next, whereas a gene then is passed on. So by this reasoning, the gene is the ultimate beneficiary of any adaptation.

0:45:40.9 AA: And kind of lastly, it does this. It kind of takes this insight on population genetics. The evolution is lineages of genes over generation, but because it's very much gonna emerge out of the study of social behaviour, becoming group selection debate it combines that with a form of anthropomorphise or gentle thinking I think. If I was a gene, what would I do in this situation? And this is where it comes that we expect genes to behave selfishly simply meaning here trying to maximise their own representation in the next generation.

0:46:11.7 SC: And just to give credit where it's due. The idea obviously became famous and was developed by Richard Dawkins with The Selfish Gene, but it was essentially the idea itself was already there.

0:46:22.5 AA: Yeah, so I think most of us learn about the genes that is through the writing of Richard Dawkins, especially The Selfish Gene. 10 years earlier, the American George Williams published a book called Adaptation and Natural Selection, which I hold as one of the most important books in evolutionary biology in the second half of the 20th century. But is this a book primarily directed towards other biologists and on whom it did have kind of a profound influence, it really sharpened the discussion about adaptation, the study on adaptation. But it had the fate that successful academic books have is read by the peers and but not so much more, whereas there are a few terms in modern science that had actually the reach that The Selfish Gene has had but just to say many of those arguments are... They're already 10 years earlier, but they are laid out in an even more explicit and a remarkable forceful way in The Selfish Gene.

0:47:23.2 SC: It's a much better title. Come on.

0:47:25.8 AA: Yeah, it's an incredible title and...

0:47:27.5 SC: It matters. Right.

0:47:28.6 AA: Except that we're still arguing about what selfish and what gene means.

0:47:32.4 SC: That still works. That's okay. The purpose of the title is to make people buy the book, so it's an incredibly successful title in that sense.

0:47:40.1 AA: Yeah, that you have to admit.

0:47:41.4 SC: And just to be just to clarify what it means to say the sort of... I like that you call it the genes... Sorry, what did you call it? I've forgotten it.

0:47:51.8 AA: The Gene's-Eye View.

0:47:52.6 SC: The Gene's-Eye View. Right. So it's a little bit, you're lowering the temperature a little bit from The Selfish Gene way of putting it, but maybe one distinction may be worth drawing is that when we say gene here, and correct me if I'm wrong, we mean the instantiation of every single copy of that particular gene in all the individuals, that's the thing that is competing, it's not like the particular stretch of DNA just in one cell or one individual.

0:48:20.7 AA: Exactly. Most of the time, that is indeed, what we mean they're kind of in a way, the type of the gene type that is present in all these copies in a population, and you think on that kind of whatever that is is trying to maximise its own representation. It turns out that this way of thinking... This of kind of Gene's-Eye View is very effective at also thinking of situations where in the material part of it... Specific genes are in contact with other genes within a body, what is known as the sudden genomic complex in biology where the fitness interest of different genes inside an organism diverge, and this is kind of why there was empirical fear that really was stimulated by the shift in perspective from organism down to the gene level.

0:49:09.7 SC: And one of the reasons why people got their hairs up on it a little bit, and also, one of the reasons why other people love the idea is because it does seem to anthropomorphise a little abstract notion. Richard Dawkins and no one else believes that genes had attitudes or had selfishness or anything like that, so the claim is purported to be that given this mathematical description, a population genetics and kin selection inclusive fitness, it is as if the genes are being selfish.

0:49:44.2 AA: Exactly. And the 'as if' here is crucial. Especially because, I mean, selfish is of course a very loaded term. I think Dawkins practises have some criticism that the use of the word selfish even in the book, slides a little bit, what it means, 'cause on the one hand, you have selfish in terms that all genes are expected to be selfish, which any... You can say that if something applies to everyone, how useful is the term to begin with, because you also then use the same word to describe individual organisms behaving selfish and you contrast that with them say behaving altruistically or mutualistically, and then [0:50:23.7] ____ some of it used slightly differently the essence you kind of get it, but they also kind of slides somewhat, and this not even to begin with the psychological motivations of genes or organisms where again, we use [0:50:39.5] ____ selfish just seems daily language in a slightly though related kind of way.

0:50:43.2 SC: And is The Gene's-Eye View or the selfish gene idea supposed to be exactly equivalent to the existing formalism, but just highlighting things in a conceptually more clear way, or does it actually make slightly different predictions?

0:50:58.3 AA: So most of the time, that is it will be the same, that often... So in the book that was followed The Selfish Gene, The Extended Phenotype, which kind of is the only book that Dawkins wrote that was aimed at professional biologists. All other books, as you know, has been for the public, but The Extended Phenotype for professional biologists and it comes with... Its fully referenced as we can expect that a book like that to be. That being said, it is a book that can be understood by anyone willing to make the effort, but you get less for free that you could compare to his other books. But there his thought really kind of this Necker cube illusion that if you draw a cube on a paper with the straight line, so you can often see from two perspective which lines come first, so to speak, and he kind of said that this is in a way, is the gene-centred versus the inclusive fitness view of the world that there are two ways of looking at the same thing. And in many ways, that is the prevailing attitude, I would say. Earlier on in The Selfish Gene it is written a bit more that like this is an empirical rival to individual or group level selection.

0:52:10.8 AA: But I'd say most of the time is considered that it's equivalent to inclusive fitness, and there is an interesting kind of history about the relationship between the genes-eye view and inclusive fitness, as far as I can tell, the very first time that term selfish genes appears is in the notes that Dawkins developed when he was a graduate student. And his PhD advisor was away on sabbatical and asked him to step in a lecture on animal behaviour and is in the mid-1960s, so the paper on inclusive fitness had just been published, and Dawkins wanted to cover it, but these are tricky papers to get right, so he had typed up his notes and then there he kind of explains the insight of inclusive fitness using this kind of gene-centric approach, how we should expect that genes to behave selfishly or trying to maximise their own. That being said. While Dawkins has kind viewed Hamilton as arguably the greatest Darwinian since Darwin, he's also sometimes thought that Hamilton is one of those people who really never finish their own revolution and never committed fully to the insight that you should just get rid of the organism and go down to the genetic level. So the first answer is that yes, they are almost [0:53:23.7] ____ equivalent. I think the nuanced answer is that there's, I think an underappreciated tension between the genes-eye view and inclusive fitness, that sometimes I think ought to be explored more.

0:53:35.3 SC: Well, so I was gonna ask about how you could get new insights even if they were exactly identical, but maybe I'm not clear on how they're... What the tension is... You know, and maybe there's rhetorical tension, but is there empirical tension?

0:53:50.9 AA: So I'd say the tension arises that if you wanna construct an inclusive fitness model that is an individual organism property, and that often you have to kind of ignore things that happen inside of the organism that the whole part of the organism share the same purpose that has a unity of purpose. Whereas, so I've done most of my empirical research has been on these examples of genomic conflicts and situations where genes differ in their fitness interest. Where that assumption breaks down that you don't have all genes working for the same goal, you have certain genes working for their own transmission, even if that comes at the expense of the fitness of the individual organism that carries it, and these are kind of somewhat confusing terms that they're on a selfish genetic elements, which is certainly close to selfish genes, but here selfish genes is used in a much more explicit way that they are selfish with respect to other genes of the same genome or the fitness of the individual organism.

0:55:01.2 SC: I see, I think I see. So the question is, who are the genes competing with... Is it other alleles? The alleles that are competing, so they are competing with other alleles and other organisms, or if really they just wanna pass themselves on, who cares of other ones are also passing themselves on, they might even fight against other genetic components of their own organism?

0:55:22.3 AA: Exactly, and this is truly a weird and wonderful world, and you have examples like... So you have things like the meiotic drivers that... So normally, you expect that meiosis, so the process of production of sperm and eggs, it's this kind of fair lottery, but you have two copies of specific genes like in humans, and that each of them has a 50% chance of being transmitted every time you produce a sex cell, but then you have these kinds of genes can distort that process. So in general, rather than the 50% of the sex cells stand up in 99% of them, and they can then become very common in a population, even if that by doing so, they reduce the number of offspring in other words fitness of the individual in which this distortion happens.

0:56:12.1 AA: And you have almost any part of the mechanism by which we copy genes and put them into sexes has been hijacked by some form of savage element to improve their chances of being transmitted. And I think these kind of things, it may make perfect sense from gene centred or genes-eye view is often being considered the perfect empirical indication of the approach, and I think the traditional evolution biology that emphasises the individual organism too much, which I think I put inclusive fitness in that category often has to downplay the importance of this kind of phenomenon too much for my liking. It's not that you can't use those approaches, but in a way, it's kind of a measure of temperament I think of what observations you think are important in trying to explain.

0:57:14.0 SC: Well, it's interesting because I forget whether we said this explicitly, but one of the motivations or sort of paradigmatic examples in discussions of kin selection and inclusive fitness or ant colonies, where many individuals have no offspring depending on what kind of ant they are. They work for the betterment of the colony, and the idea being that their genetic heritage is passed down by them serving their queen and their, et cetera, even though they themselves don't have offspring. And so this is kind of a twist on that idea where you're saying that a certain gene can benefit itself, maybe it's the evil twin of that idea, a certain gene can benefit itself by shutting down some of the reproductive capacities of an organism, but guaranteeing that its cells or its heritage will be passed on.

0:58:11.2 AA: Yeah, at the heart, in some ways you're faced with the same kind of conceptual question and that you have an allele that is harmful in a way to the individual that carries it, and you're trying to make sense of how can that ever evolve? That kind of conundrum. And that happens when you have the evolution of sterility in working [0:58:29.2] ____ insects and the situation of genetic conflicts or selfish-like elements that you have the spread of alleles that's harmful to individual characters. So how can we explain that evolutionarily? So to me, there's always been that kind of conceptual kinship between the two kinds of questions.

0:58:50.2 SC: To go back, okay, that was a very, very interesting example there. But even if the two views like Dawkins originally said are more or less equivalent, it still can bring conceptual insight if you have a different way of formalizing the same set of ideas. And so by thinking of genes as the agents in some sense, you think of all biology as slightly different. There's this whole vocabulary of genes as replicators and organisms as just the vehicles, all the organisms are buses to carry on genes from generation to generation.

0:59:26.4 AA: Yeah, in many ways, that is the essential claim of the genes eye view, that evolution by natural selection requires two entities, it requires something to play the role of replicators, which is how information is passed on from one generation to the next, and then something that kind of packages that and transports it around. And that is what is known as vehicles, and this a role typically played by organisms. But in principle it can also be played by things like a cell, in a situation like cancer. So there you have kind of to break down all... You used to have kind of thought, think of it as one vehicle, now we have some vehicles that have broken out on their own.

1:00:05.4 SC: Yeah. Competition.

1:00:07.0 AA: And in principle, it can also be something like the whole group could be the vehicle. But you're actually right, in a way, the choice of words here is important. The vehicle is kind of inherently a passive term relative to the replicator. You clear kind of what is valued in this viewpoint is to become the replicator, and the organism in a way is downplayed. Dawkins was part of formalizing this, but it's also the philosopher of Biology, David Hall, who came up with the same kind of way of formalizing the inside of the genes eye view, but he preferred the term interactor over vehicle. 'Cause you'd think that by calling it vehicle, you're downplaying too much of what's interesting about organisms. But Dawkins famously just said that, he coined the term vehicle not to praise it but to bury it, that he really wanted kind of the field to move its focus down to genes.

1:01:05.9 SC: Some of my best friends are organisms. I do wanna give them credit for doing something, playing some role here.

1:01:12.1 AA: That's right.

1:01:14.3 SC: But this is probably out of that field, but that distinction kind of reminds me of Turing machines a little bit. The idea of some physical or even just Turing machines is probably being too high-brow about it. Computers, we have hardware and software. I've always wondered, going back to Turing and his machines where he has an instruction tape and then some little head that goes back and forth and does the job. How natural is this division of labor into sort of physical doer in the world versus instructions or information about how to do it? At the end of the day, it's all physical stuff. So where and why and how universal is this distinction? It certainly seems central to the genes eye view.

1:02:02.2 AA: Yeah, and it's an interesting layer to that, that Dawkins himself seemed to be very taken by computers. He was one of the first to kinda learn how to program a computer when he was still an active researcher. And in several of his book, he does emphasize that genes or DNA is this kind of information, just like a CD or any kind of outer storage. And the organism is this machine which is [1:02:29.9] ____ RAMS. There is a debate in philosophy of biology to understand kind of the machine metaphor is helpful or not. And in the thinking of organisms as machines goes back a long way, and I think historically, has been quite helpful. I'm a little bit of too much 'cause I think there is something interesting about the organisms. 'Cause on the one hand, of course, they are a physical objects, just like everything else. As you say they're subject to the same laws of physics like anything else and there's nothing kind of magical about I. So they're physical objects just like anything else.

1:03:02.6 AA: At the same time, organisms are physical objects unlike anything else. And that they have this kind of... Sometimes a sense of purpose-ness or goal directness in them, and this is kind of what given rise to kind of biology's uneasy relation with teleology. And how much should we allow ourselves to talk in terms of purpose? And that is something that really runs quite deep in biology. You can rile up even the most senior and sober people on questions like this. Whether this is an embarrassment that we talk in terms of purposes in Biology, whereas others think that it is almost impossible not to do it, especially when we study animals and animals' behaviors. That kind of attempts to try to use some sort of language that's meant to be neutral is not just unhelpful, is downright silly.

1:03:58.6 AA: My attitude is somewhat that... There has been the classic debate that, did Darwin banish Teleology from Biology or did he naturalize it, did he make it okay by kind of natural selection is a way that makes it okay. 'Cause then he can talk about goals and purpose in terms of fitness. And in a way, kind of I think of the [1:04:19.3] ____ in a way coming from that naturalising way that he had kind of something that organism ought to kind of appear as striving towards. So there is no magic here. There's no Elan-Vital, there's nothing kind of spooky going on. But you have kind of allow yourself then to use a kind of language and otherwise isn't really okay in Science. You don't use it in Physics or Chemistry. Part of Science that biology often looks towards. But biology straddles, this stays, on the one hand, it's a physical science, but on the other hand it's a behavioral science, and it's awkwardly trying to balance the two. But it's certainly the part of science where purpose enters into our explanation in one way or another. And I am on the view that we should just accept, but I try to do it in a formalised or try to be so kinda less formal about it as possible in the way we approach it.

1:05:14.3 SC: Yeah, good. I'm completely on your side, I have hopes to write. I've already written in various places toward the side that says that purpose and Teleology can be naturalized. Basically, I would say not just because of evolution, but because of the arrow of time, because of entropy increasing over time. But to do it, I say it can be naturalized. I didn't say that I did it or we have a formal theory of it. And I think that that's something that would be very interesting to understand better than we do.

1:05:43.4 AA: Yeah, so a colleague of mine, Manas Patani, just submitted a paper where we had to defend what we call license anthropomorphizing.

1:05:50.2 SC: Oh, good.

1:05:51.5 AA: Which is a... Content that you're allowed to anthropomorphize and say if I was in gene, I would do so and so. But then you need to back it up with a model, a mathematical model, and that is what provides the license that... And it kinda just goes back to this basic point that it's almost impossible to do biology without using this. But we should try to do it then in an organized or licensed kind of way, and that we discuss things like mathematical population genetics or other part of Mathematical biology provides that license to do so.

1:06:27.0 SC: I would say personally, the same exact things about free will. But let's not get into that debate right now, that's for another discussion.

1:06:33.4 AA: Then there's that.

1:06:34.8 SC: Let me give you one more chance just in case we've missed anything to give good consequences of being a genes eye view person or a selfish gene person, are there other predictions that this viewpoint leads us to or other perspectives that are helpful to biology?

1:06:54.3 AA: So I think that the genes eye view, it's a very powerful thinking tool in biology. I think it's one of the best tools we have to make sense of this messy world. But like with all of tools, to get the most out of it, you must understand where it came from and what it's trying to achieve. So it's an approach then to try to understand adaptations and kind of the logic of natural selections that can give rise to that. So it works especially well when those adaptations that we observe are kind of contra intuitive to the individual organism's point of view, which has been the traditional focus of biologists. So we talked a little bit about genomic conflicts, and I think it's a great example of where the benefits of the genes eye view comes through. The others in the study of social behavior to understand what [1:07:44.8] ____. One that was kind of a new thing that people [1:07:48.8] ____ will handle is the notion of extended phenotype. So these are defined as phenotypes that are located outside of the body in which the genes are kind of 'responsible for' that phenotype reside. So kind of typical example has been the things that animals build like the beavers dam or the nests of birds. So the kind of structure that clearly has a genetic component to them. So in a way, we thought of it as a phenotype, but that phenotype is not part of the organism.

1:08:20.6 SC: So, genetically, do the clothes I wear count as my extended phenotype?

1:08:24.9 AA: Some of these are typically considered to be part and have some influence on the fitness of the organism in kind of a consistent way across time and population.

1:08:34.9 SC: I think it's pretty clear that how well people dress has an impact on their reproductive success. [chuckle]

1:08:40.0 AA: It do. Though humans are famously an abhorrent species to study. [chuckle]

[overlapping conversation]

1:08:48.2 SC: I'm just going to believe this on purely theoretical grounds, don't worry.

1:08:50.7 AA: Yeah. But those kind of built structures of animals. They're kinda thinking of those from a genic point of view, helps to stimulate that. Also kind of individuals, where individuals of one species can manipulate the behavior of another. So kind of this gruesome example like zombie ants where ants who are infected by a particular kind of fungi completely changed their behavior from kind of shying away from parts of the plant where they live to go where it's beneficial for the fungi to live and eventually kinda end up building completely bunkers because they have been infected by this fungi. So here. Again, there's a behavior of the ant that has a genetic component, but the genes for that for, if you will, if you allow us to have that short hand, for that behavior is not located in the ant, is located in another organism. And that, again, is something which kind of extended phenotype are easier to think about from a genic point of view, than trying to shoehorn it in to that of the individual organism.

1:10:00.4 SC: So except, let me bring up that not everyone agrees, this is why you wrote a book. Because there are skeptics and critics of the genes eye view point of view, and at least hopefully, you'll fill me in on what all of the critics are, or the most important critics are but one of them is that even the modern synthesis is no longer how we should be thinking about evolution. There's horizontal gene transfer, there's epigenetic factors, the whole idea that there's just a simple tree and things branch off and never cross-pollinate has become a lot messier in recent years. There's mitochondrial DNA as well as nuclear DNA. So is there any legitimacy to this idea that the genes eye view is just a codification of a very old-fashioned way of thinking about evolution?

1:10:51.0 AA: The genes eye view certainly shares a lot with the modern synthesis, especially kind of ideas so that adaptation and natural selection are important. And kind of framing that in terms of genes. But there are a couple of things that are worth mentioning here. So one, that the modern synthesis is more diverse than what critics of it gave it credit for. There's a percent that is a very kind of strict dogmatic viewpoint. Which it wasn't included in many different kinds of people, in many different kinds of views. Second of all, criticism of it has existed pretty much as soon as it was kind of codified or kind of as [1:11:29.0] ____ expressed that often are the major of people not really thinking that what they studied got the respect that it deserved.

1:11:39.7 SC: Sure.

1:11:42.6 AA: And this has come different kind of versions of this over over time. It is interesting, both Richard Dawkins and George William, so they kind did the two [1:11:50.3] ____ of the genes eye view, did emphasize that what they were doing was to kind of articulate the kind of orthodox view in a new way that they kind of very much align themselves with that kind of traditional approach. So the larger criticism of the modern synthesis and the genes eye view often, is taken to be kind of the poster child of that, is that biology has changed so much that this old way of thinking about it is no longer helpful. And yeah, this is a contentious topic.

[chuckle]

1:12:23.9 AA: And I think many sensible people can disagree on it. I don't tend to view this new observation as being lethal to the modern synthesis in any way. I think modern synthesis is a flexible framework. And evolutionary biology is not as dogmatic as even its internal critics would like to present it. It's been quite... To the fact that incorporating it. We are probably putting too much emphasis on certain things and not enough emphasis on other things. Perhaps the greatest challenge to the genes eye view is this notions that we ought to consider a more inclusive notion of inheritance, that we are giving genes too big of a role and that parents pass on more to their offspring than their genes. So this can be things like epidemic signals, methylation patterns, but also cultural from the cultural inheritance, some other kinds of maternal effects. So that, I think there may well be some truth too. One thing that, I would like to see, I don't know if anybody's done is that the reason why the gene eye view proponents like to talk about replicators rather than genes, is that replicators is completely divorced from any sort of material basis, and that what really matters is that it can make copies of itself. So I wonder how some of these kind of other forms of inheritance can that be incorporated into this traditional framework.

1:13:48.9 SC: Certainly Richard Dawkins is in favor of thinking about memes as an important method of cultural transmission.

1:13:56.8 AA: Yeah, memes also is interesting. Of course, it was coined in the last chapter of the first edition of the selfish gene. It was meant to kind of free replicator from the association with genes in a way, and that was the idea that if organic evolution requires something to pay a replicator and a vehicle, so should cultural evolution. Ironically though, I think that the... Despite wanting to free itself from the material basis of genes, it is actually cutting those kinds of properties that makes the genes having worked so well in organic evolution, but not in cultural evolution.

1:14:36.2 SC: Yeah.

1:14:36.3 AA: There are things that... Where does one gene start and the other one ends? How long is a gene? What does mutation actually look like, or competition between alleles, what are really... A lot of things that... In organic evolution, we try to downplay some of those properties, but in a way, that is kinda exactly why it worked at the end of the day in organic evolution, but not in cultural evolution where those bonds become so poor is that to lose its value as a concept.

1:15:03.2 SC: Yeah, a real biological gene is something at the end of the day, you can hope to point to, there, in the DNA.

1:15:08.4 AA: Yeah, it's anchored in some sort of material reality in a way that I think is hard to do in memes but... So, despite being a very successful meme in of itself, my impression is that it's influence in current study of culture is somewhat limited.

1:15:22.0 SC: That makes sense. I guess, and this is a good place to sort of wrap up. The part of the discussion that is most fascinating to me, it's taken us a while to get here, but it's the level of selection, levels of description discussion. One thing that one can say, it's not a very useful thing to say, but one can say it is, if I could be the Laplace's demon, if I knew the exact state of everything in the universe, and I knew all the laws of physics, and I had infinite calculational capacity, I could just tell you what was gonna happen. And I think an under-appreciated aspect of that is that Laplace's demon doesn't need to know about cells and genes and organisms. It just needs to know about the microscopic state of the world, but I can talk about things in terms of cells and genes, or organisms or societies, et cetera. This is a far too big a question to answer, but say what it inspires you to say, what is the right way to think about the different levels of description that we have to capture biological reality, and can they co-exist, or is there a best one that we should focus in on?

1:16:34.7 AA: I'm very much calm on the view these days that there isn't one best way to do it in biology. And I think that biology is a kind of science that is so messy, and especially departure of biology. Like evolution, where we are trying to understand the historical process, that we are better off trying to retain multiple perspectives. And so I'm quite comfortable that some people prefer to model things in a group section way, and some in an inclusive fitness way, and sometimes in a genomic way, and that is not necessarily gonna be a true answer that... While there may be one, but I don't know if it's meaningful necessarily to spend all their time trying to find that.

1:17:16.2 AA: That being said, I think we ought to spend some time worrying about when we have two different kinds of approaches. So one example is perhaps in population genetics and inclusive fitness methodology. Inclusive fitness, you've often relied on these kind of optimization methods, quite popular in the side animal behavior. But optimization is also something that population geneticists rejected in the 1960s, that we should never expect selection to optimize anything. So things are gonna be in this Bayesian equilibrium, but they're not gonna be optimized in any sort of meaningful way. So some part of our field has rejected it and some uses constantly. And in some ways, I'm fine with that, at the same time, I think we ought to spend some time worrying about it. But on a larger philosophical points, I think we have one reality, but we have many ways of talking about it.

1:18:09.1 SC: Yes.

1:18:09.1 AA: And I think that becomes expressed very well on a day-to-day level in the study of biology.

1:18:15.5 SC: Well, I guess the complication or the interestingness of the question comes in because... Well, for me personally, I don't know about for the rest of the world, but as someone who is trained on pretty cut and dried examples of different levels of description such as atoms and fluids, we can literally derive fluid mechanics from large collections of atoms interacting with each other. And we know that you don't need to understand atoms to understand fluids or vice versa. So there's this almost too good to be true simplicity of autonomy of the different levels, and I get the feeling, I'm sort of slowly dawning to the realization that in biology and in the human sciences, those levels are harder to disentangle from each other.

1:19:03.9 AA: I think that is exactly right, that you have this kind of layers of descriptions that both can work really quite well. So kind of the quantitative genetics that animal breeders use, where you're kind of selecting on phenotypes and you kinda know that there's heritability 'cause you can pair parents and offspring, but the kind of quite precise quantitative methods that you used there, don't really care about DNA sequences necessarily.

1:19:30.7 SC: They don't know, yeah.

1:19:30.8 AA: They can be informed by those. And I think really advance of more genome sequencing and all that, that part has also been improved, but it functions rather well without it as well, but it's not that... So instead you can have derived those kinda of phenotypic approaches from studying genetic sequences or vice versa is not necessarily gonna happen. And I don't really feel that pessimistic saying that, I think that's just kind of the nature of the endeavor. Absolutely, I think biology is kind of perhaps where that... That you can derive one from that, is starting to really break down.

1:20:06.9 SC: Well, I certainly agree with that 100%, but there's sort of a deeper, trickier issue. I think that far too much emphasis has been put on drive-ability because outside the very simplest examples, of course, you can't derive. I cannot derive the solidity of the table in front of me from the standard model of particle physics, but I don't think it's incompatible with that. But what is important, not what is important. What I'm interested in at the moment is, I don't need to know about the standard model of particle physics to describe the behavior of the table, there is an autonomous theory of it. And in biology, forgetting about derive-ability, it's not even clear to me that there's that.

1:20:49.5 AA: No, that's an interesting point. I hadn't thought about it that way, but yes, I think there's less clear that you have that kind of a symmetry. Yeah, no, I think that it's absolutely true that you can kinda go about it... Everything has to be compatible, of course, with what's was going on.

1:21:05.2 SC: Exactly. Sure.

1:21:05.7 AA: At a lower level. But how you even begin to kinda worry about that, is less clear.

1:21:11.5 SC: Well, I have an example in mind which is very contemporary, which is the germ theory of disease. [chuckle] So we can have a theory of how diseases are transmitted through breathing and contact and things like that, without actually knowing that they're microbes. We don't need that lower-level theory, but it is helpful. And so I honestly don't know the answer to this question, but I would like to know, is there some sense in which there is a self-contained theory of, let's say, disease transmission that never refers to microbes? Or if there is such a thing, is it so baroque and backwards that you can always improve your description by just helping yourself to the microbial description?

1:21:56.7 AA: Yeah, I guess part of the answer... Epidemiological models are in some ways very generic in the sense that you rely on, it's just transmitted from one individual to the next, but they can certainly be improved by knowing how it's transmitted. Like learning that the COVID-19 virus is primarily airborne rather than sticking around on surfaces improves models in kind of how you want to become more exact, but you can certainly... It's not necessarily... You can do sound models regardless of how they kind of transmit and what is most important for transmission on this particular virus, but it's gonna [1:22:39.6] ____ wear one part of where the biology of it matters, sort of the reality of it.

1:22:45.0 SC: Yeah, I do think these are good questions we don't know the answers to yet. So I'll give you one last chance, one final question. What does all this bode for the future of Evolutionary Biology? It's interesting to me as an outsider to see the flare-ups, not just intellectual, but emotional when things like The Novak et al paper came out about group selection or just a discussion around the idea of The Selfish Gene. It gets pretty heated. You've mentioned a couple of times that data has sort of stepped in and made things a little calmer, is that the future or are we still in for a whole bunch of heated debates in the pages of The New York Review of Books?

1:23:29.3 AA: I think data certainly has helped a lot, but there's also this issue that in evolutionary biology, we tend to have these flare-ups over time. There was a fun paper probably a couple of years ago simply called What is wrong with Evolutionary Biology? And the argument there is that there are kind of two things, that evolution by natural natural selection on the one end is an abstract principle that it can be applied to almost anything, but also that evolutionary biology is such a diverse field in terms of what we study. So if you can't convince the palaeontologists, maybe you can get molecular geneticists to listen to you. If they don't, maybe you can try the plant ecologists, and that those two things in combination mean that you can always get a small enough audience to keep it going, but you also may have a hard time explaining, kinda committing everyone.

1:24:19.3 AA: The key thing that I hope to achieve with this book is to bring some nuance to the debate over the genes eye view. That I had noticed that one, that people kept having these debates, but also when I was a graduate student, I would ask more senior people in my department about it. You walk to one person and they will tell you, "Oh yeah, this is the correct way to think about evolution and everyone knows that." And then you walk a few doors down and it's like, "Oh no, this was debunked in the '70s and everyone knows that." There was something there that I kinda was frustrated both by the proponents and the critics that I was consulting that they painted kind of an unfair picture of what a debate actually is, kind of the strengths of the case both for and against this way of thinking about the evolution. And so that's kind of what I wanted to capture really, really well.

1:25:06.8 SC: Yeah, no, I think it's great, and I think this is extremely helpful conversation for me in clarifying things. So maybe, in whatever debates we have in the future, they'll be at least on a more informed basis going forward. So Arvid Agren. Thanks so much for being on The Mindscape Podcast.

1:25:20.9 AA: Thank you very much for having me.

[music]