Robert Boyd & Peter Richerson — How Ice Age Climate Chaos Made Humans Cultural Animals (#157)

Robert Boyd and Peter Richerson are anthropologists based in the US. Their partnership was central to the development of Dual-Inheritance Theory, a framework that applies Darwinian evolution to culture and explains how genes and culture have intertwined to shape our species.

This is their first ever joint interview.

Video

Transcript

JOSEPH WALKER: Today, it is my distinct honour to be speaking with Peter Richerson and Robert Boyd.

They are two anthropologists widely regarded as the co-founders of the field of cultural evolution. And they’ve made probably the most significant augmentation to Darwinian evolution since the neo-Darwinian synthesis, which merged Darwinian evolution and Mendelian genetics. So it’s really special to speak with them today.

And in fact, it’s actually a triple honour because I think this is also – we were speaking beforehand – but I think this is your first ever podcast interview, maybe even your first ever joint interview.

ROBERT BOYD: I think that’s right, yep.

WALKER: So, Pete and Rob, welcome to the podcast.

PETER RICHERSON: Thank you.

BOYD: Happy to be here.

WALKER: So, first question: The oldest tools we’ve found are about 2.6 million years old, the Oldowan tools.

RICHERSON: Some people think about 3.1, 3.3 at …

BOYD: 3.3, at Lomekwi.

WALKER: For the Oldowan tools? 

BOYD: They’re not Oldowan. They’re even simpler than Oldowan tools.

WALKER: Interesting.

BOYD: They were probably made the way Kanzi made tools: by flinging stones at a hard surface …

RICHERSON: And picking up the sharp flakes.

BOYD: Picking up the sharp flakes and using the flakes.

WALKER: Right. Knapping?

RICHERSON: No, not knapping, in the sense that knapping involves using two hands to knock off … You hold the core in one hand and knock off flakes. That’s an Oldowan.

BOYD: There’s a term for this flinging and I forget what it is.

WALKER: Right, so even simpler than knapping.

BOYD: Even simpler. The cores are huge, they’re like two and a half feet across. And there was some controversy originally about whether they had the dates right because of geological considerations at the site. But now I think pretty much everyone agrees that there are tools there at 3.3 and there are bones with cut marks just a few kilometres away at another site in Ethiopia. And so: 3.3.

WALKER: 3.3. That’s fascinating. I didn’t know we’d found evidence of tools from that long ago, but I did certainly know about those cut marks. They found cut and scrape marks on the bones of a bovid and an ungulate in this site in Ethiopia. 

BOYD: Yeah, Gona.

WALKER: Right. So something was using a tool to cut flesh from bones 3.4 million years ago.

BOYD: Right.

WALKER: We also know from fossils found about 3.2 million years ago that australopiths had evolved precision grip in their hands. So clearly something was applying selection pressure on their hands to become more dexterous. So my first, I guess, substantive question is: how likely is it that cumulative cultural evolution was already underway with australopithecines?

RICHERSON: There is – I just read a paper, I forget the authors – they decided, it seemed kind of arbitrary to me, if a cumulative culture, the signature of it, is more than six steps in the manufacture of a tool (Now, why six rather than two? I don't know. It wasn't convincing to me) But by their account there, by the Oldowan times or later, even, the tools were too simple. There's something that somebody could reinvent for themselves. That's the argument. Now, I'm not sure I quite believe it. But that's a kind of approach that people are taking to trying to find a signature for cumulative culture.

BOYD: The other bit of evidence. So I think it's a mistake to think of cumulative culture as either there or not there. I mean, when the psychology evolved, presumably there's a bunch of steps, and so it's a fact that there are lots of australopithecine sites with no cut marks and no tools and australopithecine fossils. So one model you could have that fits the data is there was some kind of psychology that let little bubbles of technology evolve and then they would persist for some period of time, and then they'd be lost. And lots of models have that property. If the error rate of learning is high enough, you can easily go backwards.

RICHERSON: And maybe the technology was only useful in certain habitats. That's also possible. I mean, we know that chimpanzees and other tool making animals sometimes have two or three or more steps involved in making the tools. By my way of thinking, it's just a continuum, and great apes are pretty smart. Australopithecines must have been pretty smart, and they probably made all sorts of tools that leave no record. I mean, the stone tools are sort of like looking at the cumulative culture record through a keyhole. They're wonderful things because stone is so durable, but they're a tiny window into probably what they were doing just to look at chimpanzees and other tool-making modern creatures. And as you say, the hands of australopithecines became rather modern looking. So to me, it's a big mystery what most australopithecines were doing. As Rob said, most of them weren't making any tools – that we recover anyway.

BOYD: Like stone tools.

RICHERSON: Exactly. So what were they doing with those hands?

BOYD: Well, chimps make all kinds of stuff.

RICHERSON: They don't have hands that are really well adapted to making things like stone tools.

BOYD: And when you get to early Homo – so that's, say, 2 million years ago, 1.8 million years ago or something like that – then you never find fossils without stone tools. So there's a difference between the australopithecine archaeological record and early Homo erectus, depending on how you want to … It's unclear how to categorise all those fossils, but …

WALKER: So what were australopithecines doing with their hands?

RICHERSON: That's an excellent question. I mean, pretty clearly, at least sometimes they were using them to make tools, but they might have been using them to carry things. Rob had this idea a long time ago that australopithecines are small, not very fast, out in fairly open country. So what they might have been doing is going around in big mobs carrying sticks and stones to defend themselves against lions and giant hyenas and the other nasty predators that were out there. Karl Butzer has an old book in which he has a plate of a drawing of a chimpanzee with a great big rock. I mean, a chimpanzee, an australopithecine with a great big rock threatening to chuck it at something.

BOYD: There is this idea about throwing. So – what's that guy's name at Stony Brook?

WALKER: The stone-throwing hypothesis?

BOYD: Yeah, exactly. My proper noun module is failing on me. But people like Leslie Aiello and people who know the anatomy will tell you that the shoulder and back have been reorganised by early Homo and maybe australopithecines for the purposes of throwing: overarm throwing. And you could tell lots of stories. We don't know, but I often thought … When I was a young lad, we had rock-throwing wars. There was this pond near my parents house, and different groups of kids would start throwing rocks at each other, and you can really get hurt with a sizeable cobble. And if I were trying to scavenge from lions or hyenas, throwing rocks at them would be a good strategy because they don't want to come chase you away because they want to stay there eating. There's a big free rider problem, and you could stand off and make trouble for them, but just, who knows?

WALKER: So, Pete, you mentioned somebody has this argument that it's unlikely australopithecines had achieved cumulative cultural evolution, because the tools that we can conclusively connect to them are things that seemingly could be created in the lifetime of one individual. So individual learning is kind of sufficient to explain those innovations

RICHERSON: Or at least it's at the margin.

WALKER: At the margin. So my rebuttal to that, if I wanted to take the other side of that argument, would be: sure, but their brains were much smaller than early human brains. Their brains were only a little bit larger than that of a chimpanzee. And so perhaps those tools weren't cultural creations that could be devised in the space of an individual lifetime for an individual with a brain of that size. What would you make of that rebuttal?

RICHERSON: It's quite possible, it seems to me. And also there's a kind of a quantitative. Maybe 1 in 100 australopithecines would invent a particular tool, but the other 99 per cent of them got it from cultural transmission or from mum. 

BOYD: But no accumulation.

RICHERSON: But not much accumulation, at any rate. So it just seems to me that we'd expect this to evolve gradually and there wouldn't be any cut point. We put it on cumulative culture, be a bit arbitrary. So my picture of it is that as the brains got bigger, more information processing was possible. And of course, that would also mean that Homo erectus was probably more inventive. They could probably master things individually that an australopithecine, the average australopithecine at least, might depend upon acquiring culturally.

BOYD: And their brains were – it depends who's doing the estimating, but about 30 per cent bigger, corrected for body size, than chimps. But they show no trend. Australopithecine encephalisation is completely flat from three and a half million years ago to 2 million years ago. And as soon as you get to early Homo, there's quite a steep slope. So by a million years ago, you've got creatures with brains, 1000 cc, 1100 cc, something like that. So something happened at the transition between australopithecines and early Homo. Early Homo is a real mixed bag. There are little tiny ones that are sort of australopithecine size and there are big ones, and it's not completely clear what was going on in terms of one species, several species: that's all up in the air right now.

WALKER: Yeah.

RICHERSON: One of the things that is a potential causal factor is the onset of Pleistocene climate fluctuations. We still don't have any high-resolution data from between 3 and 1.5 million years ago. There's just a recent paper that reports a core that's 1.5 million years old. And it's pretty clear that the millennial and submillennial scale variation is present 1.5 million years ago, but it's gotten more intense across the whole Pleistocene, or at least the last two-thirds or three-quarters of the Pleistocene. And we don't know anything about the high frequency variation before the onset of the Pleistocene. There's a big gap in our knowledge at that critical period when the transition occurred. But my speculation is that it was the onset of that millennial and submillennial scale variation that started to push the encephalisation.

WALKER: Yeah, it's such a fascinating speculation. I want to come back to that, and I have a few questions on that. But before we do, a couple of other things first. I wanted to test an intuition with you. Sometimes you hear people talk about the prospect of chimp scientists and whether cumulative cultural evolution could allow the possibility of having chimp or australopith scientists. But my understanding of the way the process works is that that is not how it would play out, because as soon as the autocatalytic feedback loop of social learning to cumulative cultural evolution spins into motion, that puts selection pressure on brains to grow larger. And then by the time whatever the creature was has evolved into being a scientist, the species is no longer what it originally was, whether that was a chimp or an australopith or whatever.

BOYD: Is that a correct science? So you mean science? Science, as we know, it is only a few hundred years old.

WALKER: Yeah, I guess so.

BOYD: It's a particular …

WALKER: That's a high standard. That's a high standard.

BOYD: It's a particular social institution that has a bunch of rules about credit and priority and other things that generated the world we live in, basically. But before that there are people doing experiments and cumulative culture all the way back. So that's what you mean by science.

WALKER: So could you have a creature with a small brain of 400-cm3, like a chimp or an australopithecine, that because it evolves cultural learning it can simultaneously be a small-brained hominid and evolve some kind of cultural complex as intricate as the scientific method. And I guess the intuition I had was that those things aren't compatible, because by the time it evolves the intricate cultural complex it's already a different species because of the pressure that feedback loop would place on its brain size.

RICHERSON: So do you know the argument of this South African student of tracking, Louis Liebenberg?

WALKER: I don't think so.

BOYD: He'd be a great guy to interview.

RICHERSON: Yeah, I don't know him.

BOYD: I've had some … Yeah, go ahead.

RICHERSON: Anyhow, he studied tracking by the Kalahari San people, and he argues that the roots of science are in the kinds of practices that they have. So usually it's two, three, four guys tracking an animal. And the signs are very ambiguous and slight: bent grass blades and a strike of a hoof here and a bent twig where something brushed past a bush, or something like that. And so these guys are talking to one another all the time, and their natural history knowledge is quite good. Very good. And so they know the animal that they're tracking. They know what species it is, they know how it moves and what it's likely to do. And so they're using that background knowledge to generate hypotheses about what this particular animal is doing right now.

And they are using this to reach a decision about: do we continue to pursue this animal? If they're tracking it, they don't have it in visual, they don't have it in sight. So they're depending upon these thin clues to try to infer what it's doing. Now, if they think it's wounded or old, then they might continue this pursuit for quite some time.

BOYD: Days.

RICHERSON: Days even. And on the other hand, if they think it's a perfectly healthy animal and can run four times as fast as they can, they say, ‘yeah, give up on this one, we'll never catch it’. So they're prepared to spend a huge amount of effort if they think they've got a reasonable chance of success, but not otherwise. And they're just talking it out as they're running along, looking at one sign after another.

So this, he thinks, is … The big thing is that the background knowledge; they've got a theory about what this animal might be doing, and they're testing that theory with this rather skimpy evidence that they're running across.

WALKER: Right. So they're using a form of deductive inference?

BOYD: Yeah, it's a scientific method, kind of, on the fly out in the bush, chasing wildebeest or something. They hunt with these little low-powered bows: tiny little 40-, 50-pound pole bows and poisoned arrows. So they often have to track animals that are dying from the poison for long periods, for hours and hours. And so it's a big part of their hunting strategy. It's not like other people – there are those that kill the animal on the spot, but not the Hudson, not the Kung. Actually, it's not the Kung, it's a central Kalahari people.

RICHERSON: I can't remember the name of the group.

BOYD: I can't pronounce it. Exclamation point X O.

RICHERSON: !Xó.

WALKER: That's so interesting. So, last point for now on the history of early humans. But what's your lower bound for when cumulative cultural evolution was achieved in our lineage? When can we say with some certainty that it's clear in the record?

BOYD: Three to 500,000 years ago.

RICHERSON: Middle Palaeolithic.

BOYD: Middle Palaeolithic. You get prepared core tools and lots of evidence, I think. People know how to make fire.

WALKER: And this is with Homo erectus?

BOYD: No, a little later than that, probably. Who this is – you can get in fights about what to call them, but these are these robust, sometimes people call them archaics, proto–Neanderthals, proto-Denisovans.

WALKER: Right. Is this sometimes called heidelbergensis?

BOYD: Yeah, sometimes people call it heidelbergensis. In certain circles, you can get in a big fight about what to call these guys.

RICHERSON: But you're not going to find Rob and I in that fight.

BOYD: No. I  mean they were, morphologically heavyset compared to us. They're quite distinctive. But what they were like cognitively: their brains were a little smaller, but not very much. And by the end, they were bigger, actually, than our brains. And so, if I had to bet, if somebody introduces a time machine and makes me bet, I would bet at least a million.

WALKER: At least a million.

BOYD: At least a million. But I think the evidence … There's this one site in Israel. 

WALKER: Gesher Benot Ya’aqov.

BOYD: Yeah, exactly.

RICHERSON: That has all the organic as organic remains.

BOYD: There's a skeleton of an elephant. They're eating acorns.

WALKER: Fish, turtles.

BOYD: Fish, turtles, but eating acorns. Depending on what kind of acorns they were, that can involve quite a bit of detoxification.

WALKER: Yeah. You've got to leach them.

BOYD: Leach the tannins out of the acorns. Now there are acorns without … sweet acorns. So it could be that. We don't know for sure. But anyway, it's a site. It's unique. It's this anoxic site that preserved all this stuff. And I watch guys making hand axes. We have some very skilled knappers in my department. And it looks pretty complicated. I don't think we know, actually. 

RICHERSON: Yeah.

WALKER: Interesting. Okay, so I'm going to break all of my podcast rules by getting you to answer a pretty basic question. And not only that, but a three-part question.

BOYD: Oh my goodness. So the audience has to remember three parts.

WALKER: Yeah, exactly. And you guys too. So we don't have to spend too much time on this, but just to give everyone context, before we move on, could you briefly outline, firstly cultural evolution, secondly cumulative cultural evolution, and thirdly gene-culture coevolution.

RICHERSON: Sure.

WALKER: You can tag team if you want.

BOYD: Like in the wrestling. Tag me and you can jump in the ring. So cultural evolution occurs when an organism gets useful information from other organisms through some kind of social learning process. So I think chimps have cultural evolution; lots of songbirds do. Lots of animals have some kind of cultural evolution. 

Cumulative cultural evolution occurs when the process is sufficiently faithful or accurate, the social learning process, so that some individual invents something and then some other individual learns enough about what that individual made and now changes it, and other individuals can learn the new variant. So lots of social learning processes don't involve the transmission of variation within a population. 

It's a great story. Should I tell you? So Hans Kumar was a very famous primatologist, and he had a macaque colony outside of Zurich, and there were crabapples along the outside of the cage, and the crabapples would drop. And the monkeys were very interested in these crabapples. And one of the monkeys learned to get a stick to drag the crabapples into the cage and eat them. This attracted a lot of attention, and the monkeys got very interested in sticks. So this is what's called stimulus enhancement. The fact that the sticks were associated with something that monkeys really wanted – the crabapples – made them ... but no other monkey could put it together. And the first monkey used the stick and dragged it back.

WALKER: This is a really rudimentary form of social learning.

BOYD: It's called stimulus enhancement. It's really common in non-human animals.

RICHERSON: Probably common enough in humans.

BOYD: It's common in humans too. Right. But the thing is that unless I see … So you've innovated something. Unless I can copy that innovation – not the fact that you use sticks or the fact that there's crabapples to eat, but the fact that you did a particular thing – that's what you need for accumulation. Because now somebody else makes another improvement, somebody else makes another improvement …

RICHERSON: Somebody makes a stick into a rake.

BOYD: Yeah, and you end up with something that's beyond the inventive capacity of individuals. There's a very colourful psychologist, Claudio Tennie, who has studied this a lot, and he calls it the zone of latent solutions. The zone of latent solutions is that space of behaviours that individuals can generate based on the behaviour in their lives with some social input, sticks are important, but humans are able to get the details or the variants that somebody else invents, and that leads to cumulation. And gene-culture coevolution happens anytime. Culture of either kind changes the selective regime so that gene variants that weren't favoured when in the acultural, non-cultural, species are now favoured. And that doesn't have to be due to cumulative cultural evolution.

RICHERSON: So, there are migratory ungulates that have to learn their migratory patterns by some form of social learning. And there are these natural experiments where some migratory route has been stopped for some reason – somebody built a fence, and then the fence gets torn down, and it takes them a few generations to recover their original migratory pattern. So this is something that's very important to their lives. And yet it's still a very simple form of culture. How far do you migrate and when you do it.

BOYD: Domain specific.

RICHERSON: Yeah, rather domain specific. And then it seems to me that gene-culture coevolution comes in two flavours. One, originally pointed out by E.O. Wilson, is that culture puts pressure on genes to control the cultural evolution, to influence cultural evolution so that it increases genetic fitness. And then there is, I think of it as culture-led gene-culture coevolution, where the cultural variation is impressed on the genes because the cultural environment generates selection on the genes, it favours the cultural system. Modern penal institutions have this effect that if you're incarcerated, you're less likely to reproduce. I think of chimpanzees as basically a society of psychopaths. In humans, we've got psychopathy down to a low roar. Something like 2 or 3 per cent of humans are clinically diagnosed as psychopathic. And that's because for hundreds of thousands of years, if not longer, we have punished people who don't cooperate and rewarded people who do cooperate. And this has transformed our social psychology. 

The difference between chimps and humans is that humans are adapted to make massive use of cultural adaptations. And to do that, you need to have large societies. Two things are critical to mobilising resources. One is technology and the other is social organisation. So we make a lot of profit out of being able to cooperate in ways that our competitors can't.

BOYD: Yeah, and on a much more simple scale, a big fraction of the new genes that have been … So you can identify genes that have been underselected by looking for long haplotypes that show evidence of a gene being favoured by selection and carrying a bunch of the lateral genes along with it.

RICHERSON: Linked genes. Yeah.

BOYD: So there's a bunch of linked genes, and if there's some highly favoured gene, it drags a bunch of linked genes with it. And by looking for those long segments of chromosomes, you can tell which genes were under selection.

RICHERSON: And how long ago.

BOYD: And how long ago. And a lot of the ones that have been under recent selection, in the last 10,000 years, are genes that the cultural evolution of farming and urban life created the new selection environment. So genes for dealing with starches, genes for dealing with pathogens that are characteristic of a village rather than hunter-gatherer life and stuff like that. It's like two-thirds of the genes that have been under recent strong selection are ones that are plausibly related to environmental changes that were culturally evolved.

WALKER: That's really interesting, because so often I'll hear cultural evolutionists make a claim like ever since our species crossed the Rubicon into cumulative cultural evolution, maybe that was a million years ago or whenever…

BOYD: Yeah.

WALKER: … ever since that point, the culture we've accumulated has become the main selection pressure on our genes. But my question is: how is that statement actually quantified or quantifiable? But maybe that suggests a way that you could.

BOYD: Yeah. So we can't detect selection. Well, there are ways, but recent selection is a lot easier to detect than selection 50,000 years ago. And that's just because recombination breaks up all these long haplotypes.

RICHERSON: Long haplotypes are reduced to …

BOYD: You can't see them. There are other methods, but they're much less reliable, I think. So it could be that in spreading across the world 60,000 years ago there was lots of selection for all kinds of stuff, but we have a hard time seeing it, detecting it with genomic methods. I mean, I think almost certainly moving from the Horn of Africa to the High Arctic must have selected for some stuff, right?

RICHERSON: I mean there are the morphological differences, cold adaptations. Cold adaptations. The high altitude adaptations is another good story, but who knows how much historical information geneticists will eventually be able to figure it out to extract? The innovations have been spectacular. Ancient DNA, for example. Spectacular innovations in the last few decades. 

BOYD: Yeah, if we had populations of ancient DNA, which is not out of the realm of possibility, then we could do long haplotypes with a DNA.

WALKER: I'm aware of that study Kevin Laland and a few others did a few years ago, where I think they found over 100, maybe it's more now, but they'd found over 100 genes that plausibly had been shaped by or selected by culture. 

Just to quickly go back to the definition of cultural evolution, how much loss would there need to be in transmission before you said, okay, that's no longer an evolutionary process?

BOYD: Yeah, I mean, that depends, right? So in experiments, there's a lovely experiment by a young cognitive scientist at Berkeley, Bill Thompson. And he had people learn to do a sorting algorithm, and then he offered them in one treatment – well, there were three treatments. One, they had to do on their own, only a few people figured it out, a few per cent, 10 per cent or something. And then they had social learning and they could see other people. And then he gave them social learning plus the payoffs that the individuals who solved – everybody's payoff depending on what algorithm they used. And the error rate in that experiment was about 50 per cent. So 50 per cent of the people who copied – cause he knows who copies who from the way the experiment was run – failed to get it right.

BOYD: But if they were given the payoff information, the best algorithm spread throughout the whole population, because the flow of information in was high enough to compensate for this very high error rate

RICHERSON: Mutation rate. 

BOYD: Whereas if you didn't get payoff information, the best algorithm just percolated along at what the invention rate was.

RICHERSON: So people could use the success rate of others as a basis for biasing their …BOYD: And they just completely focused on people who had the best algorithm and that whole process … So I don't think it's just the accuracy. You have to look at the processes that are degrading information as it's transmitted through time and then processes that are increasing the quality of information.

RICHERSON: So it's just a kind of a budget problem there: ingoing, outgoing, and the sort of equilibrium depends upon the relative rates of those two things.

BOYD: I think you've got cumulative cultural evolution when things get better through time. Now obviously once everybody's learned the good algorithm, they're going to stop getting better, but it'll reach some steady state. But if the potential is there for gradual improvement over generations, then I count that as cumulative cultural evolution. 

WALKER: Interesting. So back to gene-culture coevolution. Are there, for you, any particularly surprising or counterintuitive examples of gene-culture coevolution? I think the one that fascinates me the most is that the sclera of our eyes have probably evolved to aid with teaching, because you can track the gaze of someone who's teaching you if you can see the whites of their eyes. And I guess it helps with social referencing and infants as well. But are there any other examples that. I know that canonic …

BOYD: How do you find that counterintuitive?

WALKER: Not counterintuitive, but just surprising or kind of fascinating.

RICHERSON: So one of my favourite examples – it’s a little bit peculiar, but Japanese people eat seaweed as a fairly large part of their diet, and it's not human genes that change, but they harbour a bacterium in their gut that degrades the polysaccharides in seaweed. And this gut flora that we have is a whole ecosystem that is presumably coevolving with our genes and with our diets. And obviously diets are culturally determined in large part. So that's a whole other field of coevolution, if you want. We've got this giant domesticated ecosystem in our guts.

WALKER: But that hasn't genetically evolved, that's downloaded through culture, approximately?

RICHERSON: Well, who knows? In other words, there's people in Japan are eating these seaweeds with polysaccharides that humans can't digest very well. And is it that they just acquire a wild bacterium that happens to have this capacity, or has a capacity evolved in human guts? I don't know. I don't imagine anybody does. But it's another whole dynamic system that involves gene-culture coevolution, not human genes. And I suppose it's a lot like domestic animals. There's domestic plants and animals are a whole other coevolutionary process that involves human artificial selection. Now, it's not deliberate selection in our gut, I don't suppose.

BOYD: But most domestication isn't either.

WALKER: The Siberian fox experiment’s probably the classic example there, right?

RICHERSON: Yes.

WALKER: In maybe 40 generations – no, way fewer – they got …BOYD: There's a bunch of controversy about that now.

WALKER: Oh, really?

BOYD: Yeah.

WALKER: What's the source of the controversy?

BOYD: Well, the story is that Vavilov, is that his name?

RICHERSON: No, Vavilov was a plant guy. What is his name? I can't remember.

BOYD: Anyway, it doesn't matter. They had one selection criteria, which was flight distance. Basically, people would approach and see how close they would let people get. And they started with wild foxes, and there's been a bunch of stuff recently saying neither of those things is really true, that there'd been a lot of domestication of these foxes already. And I haven't kept up with that carefully, but we know that dogs and cats and cows and pigs, there's been lots domestication, mainly by Stone Age farmers who weren't trying to domesticate anything, right? They were just keeping the ones they liked.

RICHERSON: Yeah, a friend of mine, his father was the superintendent of the National Bison Range in Montana or Wyoming. So he grew up on this federal facility that was trying to propagate bison for release in the wild back in the ’40s and ’50s, when they were trying to build up the herds. And they came to find out – they used, basically cowboys in this, and the same kind of techniques that you used to deal with cattle to deal with bison. The only problem is that the bison are just wilder, compared to a cow. And so they discovered that they were inadvertently domesticating the bison because, particularly the bulls, a lot of them were really scary, and they broke out of fences, they squashed cowboys. And so those are the ones that somehow ended up at the butcher more often. They were trying to maintain the wild type because they wanted to release them into the wild. And ability to deal with predators would have been a desirable characteristic, but their ability to deal badly with cowboys was costing them their lives. So that's a domestication story that I find fascinating. It's hard not to – when you're dealing with wild animals – it's hard not to inadvertently select them for traits that make it easy to deal with them, at least, if not other things.

WALKER: Yeah, that's really interesting. Okay, so I promised we'd come back to Pleistocene climate variation. So this is both fascinating and crucial for the mathematical models you guys developed in the ’70s and ’80s. They were most famously published in the ’85 book Culture and the Evolutionary Process. I've learned a lot about Pleistocene climate variation through reading your work. So you have these macro cycles. In the first 1.5 million years, those cycles lasted for about 41,000 years.

RICHERSON: So 23,000 years was a dominant cycle before the Pleistocene, and then the Pleistocene transition brought on the dominance of the 41,000-year cycle. And then in the middle of the Pleistocene, by about 800,000 years ago, the 100,000-year cycle became clearly dominant.

WALKER: Yeah. And then superimposed on those macro cycles, you had these shorter cycles. And this is so interesting, I didn't realise that every, say, thousand or so years, in the shorter cycles, in what we know as the ice ages, like the last ice age, the temperature could actually go from glacial to interglacial …

RICHERSON: Almost. Not quite. 

WALKER: But what we think of as ice age is we're actually interspersed with quite warm periods, so a huge amount of variation. But I have a bunch of questions about this. But before we get to those, could you just paint a picture of what that would have been like on the ground, in terms of the temperature, the rainfall in Africa, and then the variation in those dimensions?

RICHERSON: Yeah. My picture of it is it was just completely chaotic. The original palaeontologists and palaeobotanists had looked at this. They called them anomalous distributions. So they would find things like reindeer and bison and horses in the same fossil beds. Now, whether these were exactly contemporaneous or not, it's hard to say. But with those millennial and submillennial scale, high-amplitude submillennial and millennial scale variation, the vegetation and the rest of the ecology couldn't come to any kind of equilibrium, insofar as there's ever an equilibrium. So when the transition to the Holocene occurred and the climate got very much quieter, it took a couple of 3000 years before we got the modern biomes organised. So the boreal forest and the tundra and the steppe, these stripes. So the term of art was plaid versus striped patterns.

RICHERSON: And so the Pleistocene, the last ice age, it seems at least, was just hugely chaotic, particularly the last half of it.

BOYD: But it's fair to say that the cold periods were drier, right?

RICHERSON: Well, and I mean, a bunch of things.

BOYD: He's the expert on this.

RICHERSON: Well, I'm going to advance to amateur. So the picture of it is, yeah, it was colder on average, drier on average, and of course, less CO2, which impacts plant growth rates. And it was just a wildly varying system on a fairly short timescale, really short timescale. And so that impacted humans. Now, the last ice age, of course, was the first place that this was discovered. Rob and I talked about saying something about this in the ’85 book, and Rob taught me to out of it because the data was lousy. There just wasn't a decent high resolution record that resolved these millennial and submillennial scale variations until, well, the publications are in the early ’90s.

RICHERSON: The cores, ice cores from Greenland were raised in the late ’80s, and then it took a few years to get the data analysed. So in ’92, ’93, this data came out. And I thought we couldn't ask for better data than that. It was just what the model suggested should be happening. 

WALKER: An amazing vindication.

RICHERSON: Yeah, so were. Well, in retrospect, maybe we should have been bolder. We should have predicted the millennial and submillennial scale variations on the grounds that humans wouldn't have evolved without them.

WALKER: What gave you – because that data, as you said, didn't come out till the mid-’90s – what gave you the hunch?

RICHERSON: Well, there were cores drilled in Greenland, mostly along the edge of the ice, because logistically, to get out on the ice divide in the middle of the Greenland ice sheet is a logistical big project. And so the Europeans and the Americans in the late ’80s organised the resources to go out there and drill those two-mile-long cores down through … So they drill on the ice divide. On the ice divide the ice is not moving horizontally, the layers are getting thinner and thinner as you go down. But away from the ice divide, it's basically turbulent. The ice is flowing and it's overturning and mixing up as it goes along, so that by the time you get to the ice margin the record is lousy. But it did seem to show, a few cores did seem to show, these big excursions. But it was, as I say, it was not the kind of data that would convince a sceptic at all.

BOYD: And also you always argued that there's this variance propagation that the fluid guys have from long timescales to shorter timescales. And we knew that there was much more variation on 100,000-year timescales than there ever had been before. And so it seemed plausible. I mean, Pete talked me into it.

RICHERSON: There are these so-called variance cascades. I learned about this because I collaborated with physical environmental sciences to get tenure, basically. And they taught me a bunch of stuff that I would never otherwise have learned. So in the turbulent motions in the ocean, there's a huge amount of energy in the major currents, like the California current off our coast here. And that energy then forms eddies that break down. And ultimately at small scales, at centimetre scales, friction absorbs all of that energy that's being put into the ocean, basically by the wind. So there's a quite orderly cascade of variance from low frequencies to high frequencies. And so you could imagine the same thing occurring on these 100,000-year timescales, that there'd be a variance cascade that would affect the higher frequencies. But it's a metaphorical argument. It's not real. You're a physicist. It wasn't really a mechanistic argument. It was analogy, if you want.

BOYD: But it turned out to be right.

WALKER: In a big way. So explain how culture is an adaptation to this climate variation.

RICHERSON: Well, so the culture evolves faster than genes. That's the basic starting point of the argument. On the other hand, culture – at least the kind of culture we have, cumulative culture – requires this massive brain that's extremely costly in metabolic terms. And your head's fragile. It could easily get hurt. Some people even think that insanity comes from having too big a brain. It gets confused sometimes. So there are these major costs to having this system of cumulative culture. And so the question that Rob and I posed in our ’85 book is: what could this thing be good for? I mean, the way I phrase it in talks or to students is: the dinosaurs had it right. You should have as small a brain as you can get away with. It's too expensive to mess around with brains unless they're doing some real work for you. So what could the real work be?

WALKER: Brain tissue is twenty times more energetically expensive than muscle tissue.

RICHERSON: Yeah, something on that order, yeah. And it contributes a lot to basal metabolism. In other words, there's not muscles if you don't use them. Like we're just sitting here, we're not. Our muscles are just ticking over, but our brain seems to be generating a lot of metabolic activity all the time.

WALKER: I recall in your book Not by Genes Alone, I think you have a statistic: the brain absorbs about 16 per cent of a human's basal metabolic rate but only about 3 per cent in an average mammal.

RICHERSON: Yeah, 5 per cent, I think. I can't remember the exact numbers. Three per cent in efficient mammals like opossums and the other small-brained … the little-brained mammals. Yeah. So that, the question then is: how do we pay that overhead cost? And so one thing that will pay that overhead cost is adaptation to spatial and temporal variation. If we can adapt more finely to high-frequency variation in time or adapt more efficiently to spatial variation on a finer scale, then that will. In the models we made in the ’85 book, that would be something that would tend to pay that overhead cost.

BOYD: And especially in sort of multigeneration timescales. So the thing about … ordinary learning will do stuff for you on individual life timescales or generational timescales, but it gets lost, right? If you don't have cultural transmission. And a lot of the variance seems to be in these kind of 1000-year, 100-year timescales where culture is fast enough to keep up but preserves things long enough to be useful on those timescales.

WALKER: That's so interesting.

RICHERSON: So, in a sense, then, cultural evolution economises on individual learning. We learn things socially, and the cost of innovation is spread over the whole population. Everybody's doing a little bit of innovating and then sharing with the rest of us, and that will pay this overhead cost.

WALKER: That's really interesting. So if the environment is varying slowly enough, be that in space or time …

RICHERSON: Genes keep up fine.

WALKER: Genes keep up fine. If it's varying too quickly, the only way to keep up is individual learning?

RICHERSON: Or individual phenotypic flexibility of one kind or another.

WALKER: Right. But if it's varying in this sweet spot of a millennial or submillennial scale, then culture is the best way to keep up?

RICHERSON: That's our argument.

BOYD: That's our story.

WALKER: That's really interesting.

BOYD: We're sticking to it, actually.

WALKER: There's maybe one other interesting timescale here, which is if it's varying on an extremely quick timescale, like an hour by hour basis, then that would go back to favouring genetic evolution, right? It would take the average?

BOYD: Oh, well.

RICHERSON: Yeah, that might, yes. So if the temperature of your environment goes up permanently, then … But, yes, if it's varying on too rapid a timescale, then if you can adapt to the average. I mean, if the temperature were going from freezing to 1000 degrees on an hourly basis, I don't think anything could manage that. But if it's something that you can buffer. We have rather large body size, so we can buffer high-frequency variations in temperature just because we have a large body size.

BOYD: But you're thinking about things … I mean, it really has to do with the timescale of individual learning. So, I mean, you think about … our irises adapt to a varying environment on really short timescales because it's a really simple mechanism. Genes have provided us with a photometer that does that. But if it was about something else that took several days to learn, then things that vary on an hourly scale, then you're right. It would just be a genetic adaptation to whatever the average was.

RICHERSON: Right. Or as Rob says about irises of eyes, we have a lot of different phenotypic flexibility mechanisms. So, you know, if you see a bus coming down the sidewalk, you panic and get the hell out of the way. So it doesn't have much to do with culture or individual learning. It's just emergency decisions.

WALKER: Got it, got it.

So I find this functionalist analysis of culture so appealing and interesting, but do you remember how you came up with it? Like, what was the genesis of the idea?

RICHERSON: Well, on my part, I was a freshman faculty member, first-year faculty member in this new department, and it was an interdisciplinary department meant to deal with environmental problems. And one of my senior colleagues was a sociologist and he'd been part of the planning for this new department, and he'd put a course on the books called ‘Principles of Human Ecology’. And so then he was designated to teach it, but he decided he needed a natural scientist to help him teach this course. And I'd worked on a postdoctoral project with him, so I'd done a certain amount of reading in the social sciences and I knew that there were these people who call themselves cultural ecologists who talked about cultural adaptations. And so we had decided to make adaptation a central part of this course.

And so I went off to find out what these cultural ecologists had to say about how cultural adaptations worked. My training in evolutionary biology was fairly adequate, so I sort of knew where adaptations came from. Genetic adaptations came from. So where did cultural adaptation come from? And what these guys … They had the concept of cultural adaptations, but they didn't have the concept of cultural evolution, at least not in anything like the way evolutionary biologists did. Rob and I were teaching a different course together, so I was bugging him about this, and one thing led to another.

BOYD: Yeah, we came upon … It was kind of in the air. There were other people who had kind of related ideas. Don Campbell.

RICHERSON: Yeah.

BOYD: Mark Feldman. You introduced us in a way that some people would disagree with, that we aren't the first. They're the most important, you know: Mark. And so they would think of themselves that way. And I'd read their papers, and I actually don't think they had a theory of cultural adaptation at that point.

RICHERSON: They weren't very interested in that.

BOYD: No, they weren’t. Anyway, so then Pete and I just started talking, and we weren't very … So people like Andrew Vida and Skip Rapaport and Marvin Harris, Julian Stewart, they had lots of data on cultural adaptation, lots of interesting stuff, but no mechanism that was plausible, to us anyway. And so we just started talking, and out it came.

RICHERSON: I think of it as just a sort of Darwinian straight and narrow. We just marched down the same path that the geneticists had in the neo-Darwinian synthesis, in the modern synthesis.

WALKER: So I want to play a counterfactual. Imagine in the next few years, we get some new data that stretches all the way back into the Pliocene that shows that climate variation has actually been continually increasing since the Pliocene, right up through the Pleistocene. From your perspective, what's the most important story that would enable us to tell? Would it enable us to tell some kind of grand story about how climate variation has been driving the evolution of brain size and intelligence on Earth?

RICHERSON: Well, I think that what we expect … in other words, doing what Rob and I didn't do in the ’85 book, turning the question around and arguing that brain size is a kind of a palaeoclimate indicator. In other words, brains are for coping with variable environments on different timescales. Culture being, as you say, in the sweet spot; there is a sweet spot for culture. So what? Well, to go even further back, brain size in mammals generally has been increasing for the last 65 million years.

WALKER: Since the dinosaurs.

RICHERSON: Since the dinosaurs, yeah. And birds probably too. But birds don't fossilise very well because they're so delicate. To fly, you have to have light bones, and light bones aren't very rugged. And so the bird record is poor, and the mammal record is quite poor. A guy named Harry Jerison at UCLA in 1973 in his book, I think, he used naturally occurring fossil endocasts to build a record of brain size evolution for the last 65, 70 million years. But it's a really crude record because fossil endocasts aren't very common. And nowadays somebody could go back and take all the skulls in all of these collections and put them in an x-ray machine and do computerised tomography and develop a much finer-resolution set of data. But I keep looking for this, and so far, nobody that I have run across has actually tried to do this.

BOYD: It's complicated because the other response to climate variation is to go small. I mean, you know that. So apes – there were hundreds of species of apes 20 million years ago, and by 8 or 10 million years ago, most of them were gone, replaced by monkeys. So apes came first, and monkeys replaced apes mainly, not completely. So, if you can't predict, then just make a lot offspring; that’s the kind of invertebrate strategy. And that's going to be overlaid on top of …

RICHERSON: Well, yes. And Jerison’s data, how did he summarise it? Many mammalian lineages got bigger brains, but by no means all of them.

WALKER: So what's been driving that? If it's been happening for 65 million years?

RICHERSON: Well, what I think our models suggest, and models like them suggest is that the world has been getting more variable, on short timescales. Now, the trouble is that there are no short timescale records that I know of – certainly not to cover the whole last 65 million years. Now, I don't think you need to need that. You could just have samples. You could look at fossil tree rings or something like that and develop samples of high-frequency climate variation – old lake sediments, old marine sediments – and not try to have a complete record, just spot samples would give you some indication.

BOYD: You'd like to know the slope of the brain size versus time curve, right? Because I think we just know the average.

RICHERSON: Well, Jerison suggested that there was a long, slow … and then the transition to the Pleistocene was our story.

BOYD: Pleistocene, Miocene, Eocene variation being not that big: warm, wet planet. And then all of a sudden in the late Miocene things go to hell, basically, and it gets colder and drier and more variable. And so until 20 million years ago, there were tropical rainforests in Moscow

RICHERSON: Temperate rainforests in Antarctica.

BOYD: Yeah, exactly. So it was a much more … Now this is, of course about primates who are forest specialists. But then all of a sudden in the Miocene, things dry out. You can't just live in the forest anymore. There’s a lot more variation. And if it's true – I didn't know this about Jerison’s data – if there’s an acceleration …

RICHERSON: Well, again, the crudity of that data might cause some scepticism. But that's what he says.

BOYD: That's what I predict, okay, so I’ll stick my neck out here.

WALKER: I like it. That's so interesting.

RICHERSON: There is a great paper that you might look up by a geochemist at UC Santa Barbara and his co-authors. I think his first name is Zachos. A 2001 paper in Science, I think. If you have trouble finding it, just drop me an email and I'll give you the full reference. And I don't know how he derived it, but he and his co-authors derived it. But they have in the graph that they have a pattern of dots around the mean curve that they use to represent their impression of what the variation in climate has been. They don't specify timescale.

BOYD: I know this. Yeah, I know that paper.

RICHERSON: It's a great paper.

BOYD: I think the dots are just data points.

RICHERSON: I don't know what data they are.

BOYD: It's all from deep sea cores, yeah?

RICHERSON: Could be, yeah.

BOYD: Yeah, I think that's right. I think it's all from deep sea cores. And what's piece, the variant, the cloud of points around the mean gets much wider.

RICHERSON: As you get to … particularly during the Pleistocene, it just explodes. But I think this would be the orbital scale variation. If it's real data, it's orbital scale. Milankovitch. 100,000-year, 20,000-year, 41,000-year, 23,000 years the shortest. They aren't real cycles, they're more complicated than that. They're orbital perturbations that come from the gravitational effect of mainly Jupiter on the Earth. So that to a first approximation, the Earth and the other planets follow these orderly orbits, but under the massive influence of the gravitation of the Sun. But Jupiter has enough gravitational force on the other planets to perturb their orbits in a systematic but not exactly cyclical way. That's where you get the 100,000-year ellipticity and the 41,000-year tilt and the 23,000-year procession of the equinoxes by the gravitational effects. I suppose that Saturn and the other big planets also may have a measurable effect.

WALKER: I didn't know that we could connect the variation back to the orbital perturbations.

RICHERSON: But the orbital perturbations then are somehow filtered by the Earth ocean atmosphere system. So the 100,000-year, the 41,000-year and the 23,000-year quasi0cycles, they sometimes say: they haven't changed. As the Earth moved from the Pliocene to the Pleistocene, the early Pleistocene and then the later Pleistocene, those orbital scales didn't change. Something about the way the Earth responded to them is what generated the dominance of one cycle versus the other. Just exactly what that all amounts to is at least still a mystery to me. I don't know if it's a mystery to the palaeoclimatologists or not, but I think it is. And what generates the millennial and submillennial scale variation? And why has it been increasing over at least the last 1.5 million years?

One hypothesis I saw – speculative, I think – it's that the North American glacier is sort of the dominant ice mass during the high glacial episodes. Over the successive glacial cycles, it's gradually been planing the North American continent flatter, so it reduces the friction. So the reason that the amplitude and frequency have gone up is because the sliding of the glacier off the North American continent has gotten faster and faster.

BOYD: That's sort of, you know, for the want to nail, the kingdom is lost kind of causation, but it easily could be.

RICHERSON: Yeah. It's a mechanistic hypothesis. Whether it's true or not, I don't … I'm not an expert enough to have a. What do I want to say? I'm not enough of an expert to have an expert opinion.

WALKER: Yeah, that's fascinating, nonetheless.

Okay, so everything we've spoken about thus far, for me, raises the question of contingency. So it seems like – you can correct me if this is inaccurate — but for cultural brains to evolve took at least three, if you like, exogenous shocks.

One was apes becoming bipedal about 5 million years ago.

The second was the large amount of predation on the savannas in Africa. I read that there was about twice the level of predation there is today, both in terms of the number and the type of predators.

And then the third factor is the Pleistocene climate variation we've spoken about. Do you see all of those ingredients as just independent factors, and we won the lottery in a sense, in that we had these pre-adaptations, our lineage had become bipedal apes, were under the threat of lots of predation and the selection pressures that was creating for grouping socially, and then, just as that was happening, we entered this period of climate variation on a millennial and submillennial scale?

Or are those three factors somehow correlated? Maybe bipedalism was driven by the climate variation in some sense? Maybe apes had to change their foraging strategies because of the forests growing and shrinking or something like that?

So how contingent was the evolution of cultural brains?

BOYD: I think it's really contingent. I'd add a bunch of other things.

WALKER: What would you add on your list?

BOYD: Well, I think this … So humans, australopithecines, were bipedal, and that placed constraints on brain size, because as brains get bigger, there's more obstetric complications. That led to … So in early Homo, there seems to have been a shift to a more predatory lifestyle. So now we've got apes out on the savannah, they're bipedal. Heads are little, so that's not a problem. Then they start hunting. Females can't hunt with highly dependent offspring, and that leads to changes in social organisation and cooperative breeding, so-called. Cooperative breeding, that probably potentiates cultural transmission because there's more individuals together in social groups.

RICHERSON: Also, it brings males more firmly into …

BOYD: The males have to envision females if …

RICHERSON: Fathers have to teach their kids, their boys to hunt.

BOYD: Yeah. So there's a bunch of things, I think, all of which could have gone some other way. When I teach this in class, I say it's like little Eliza jumping from one ice floe to another. It's just a bunch of stuff happened and it ended up where we are. But it didn't have to be that way at all.

WALKER: Wow.

BOYD: That'd be my story.

RICHERSON: Yeah. I would say basically the same thing. But if you look at any evolving lineage looking backwards, it's one crazy thing after another. So, horses started as these little forest browsers, and as the climate got drier, more open, then you had a whole revolution, Miocene revolution, in mammals. You had the origin of the antelope and bovids and all these efficient grass-eating herbivores that exploited the more open vegetation, browsers that could … If the trees have their leaves way the hell above the ground, then there are no mammals that get up there very much. But if you've got low trees and shrubs and things, then you’ve got a bunch of browsing specialists that can get after the vegetation. And then there are top-down effects of the grazers and the browsers on the ecosystems. So if you look back in history, it's just one damn thing after another. So historical contingency plays a big role.

BOYD: And that's not opposed to an adaptationist perspective at all, I don't think. I mean, Gould kind of back in the day, tried to make those … If you're an adaptationist, then you should think … you always end up in the same place. And I think there's no reason to think that. All selection cares about is now. And on the surface … Wright had this idea of adaptive topography. It's got peaks everywhere, and they're moving around like the ocean, and you're climbing. Selection is causing lineages to climb as best they can. Where they climb depends on where they are. So I just … Yeah. I think.This guy at Oxford, what's his name? Simon Conway Morris. Isn't he the guy who thinks that if dinosaurs hadn't got extinct, we'd have bipedal big-brained dinosaurs? Maybe. I doubt it, though, myself.

RICHERSON: Well, it's interesting that … People challenged me about this; I suppose, Rob, too. I mean, if your story's correct, why don't we have lots of animals with big brains and cultural, cumulative cultural adaptations? All of us are living in the same Pleistocene environment. Why is it just humans that have got this massive …

BOYD: Absolutely.

RICHERSON: And that has to do with something like – you mentioned it – ape pre-adaptations, where apes are already fairly big-brained. They already have a certain amount of culture, we assume, in the last common ancestor, the living apes. The apes all live in fairly large, fairly sophisticated societies. All of these things you can imagine are potential pre-adaptations. And then we get the australopithecines that are bipedal and their forelimbs can turn into technology-making and technology-using devices

BOYD: And carrying. Carrying is a big deal, right? Because you invest in a tool and you’re quadruped, you’ve got to throw it away. Unless you carry it in your mouth, you're done. Whereas, humans could carry a spear around all day and …

RICHERSON: Yep.

BOYD: Chimps actually carry stuff like this. Kevin was telling me.

WALKER: Oh, really? Between the cheek and their shoulders.

RICHERSON: Oh, I see.

BOYD: Yeah, carry … They make sticks and stones they use to crack nuts and stuff, but they can't carry them very far. It's too inefficient. So, I would think: life's like that. Right?

RICHERSON: That's my picture of evolution: it's endlessly dynamic. There is this picture of evolution that it's the gradual … this progressive idea of evolution, which is, I think, widespread in the public, is that it just took a long time to get from humans, get humans starting with bacteria, and everybody would be a human if they could. Every lineage would have a big brain if they found a way to get there. This seems to me to be the wrong way to think about it. We have this extremely expensive adaptation that has evolved in response to these climate variations. And it's not the best adaptation in some absolute sense. It's just something that works in the present-day environment.

WALKER: So I've got some questions about intelligence and artificial intelligence and then some questions about the fertility crisis. But to start with intelligence, one of the interesting things that's been happening with human brains is that they've been shrinking over the last 10,000 years?

RICHERSON: Controversial. The latest paper I saw in that attempted to do the statistical analysis right. There are two, if I understand it correctly, there are two major problems. One is that there just aren't that many Pleistocene, last-ice-age skulls. And even worse, there are very few last-ice-age post-cranial skeletal material. So to correct for bodies. So Pleistocene people were probably bigger. Quite a bit bigger, on average. Meaty diets, hunting and gatherings, athletic occupation. And so they were probably … I think they're known to have been quite a bit bigger, but just how much is a problem. So those are the statistical difficulties that make it really hard to be sure. Any difference is pretty slight. So that, I think, is the problem.

BOYD: Oh, interesting. The Pleistocene fossil record is heavily biased to Europe. And Europeans now are the biggest people in the world, and it's probably a response to cold weather. So I think you have to be really careful with what was going on in Southeast Asia or tropical Africa 30,000 or 40,000 years ago. We have only the slimmest record, because those environments don't produce fossils very readily.

RICHERSON: They don't produce archaeologists very well.

BOYD: That's the other thing: there is the croissant effect, which is it's a lot more fun to work, dig up fossils in southern France than it is in the Congo basin for lots of archaeologists. I read once that there were 50 sites in Europe for every site in Sub-Saharan Africa, archaeological sites. So I would be sceptical. Generalisations about world phenotypes … But go ahead. 

RICHERSON: It’s a good story. It should be true according to our …

WALKER: Theory, according to the cultural brain hypothesis, because culture is substituting for encephalisation.

RICHERSON: So on the other hand, what has transpired in the Holocene, I think, compared to the last ice age, is people have adapted ever more finely to local ecological variation. So we have all of these agricultural systems that are finally adapted to local environments. And hunting and gathering is a much more generic kind of subsistence strategy that isn't so highly tuned to particular environments. So I think that we in the, so to speak, in the Pleistocene, at least in the last ice age, our adaptive effort, so to speak, was mainly adapting to temporal variation. And in the Holocene it switched to mainly being focused not on the slight temporal variation but the geographical variation.

WALKER: As we spread across the earth.

RICHERSON: Yeah, well, were already pretty widespread except for the Americas until the very end of the Pleistocene, but yeah. Particularly agricultural crops and animals are pretty sensitive to the local environment, right down to the variety level. So when you adapt to a new environment, you have to adapt your starchy staples and you have to select for new varieties that will be adapted to the local environment. So that may have had a tendency to keep our brains from atrophying too dramatically.

BOYD: Our teeth have also gotten much smaller over the same timescale. 

WALKER: Because of our addiction to cooking. 

So if the birth canal is the major constraint on encephalisation in humans at the moment, what happens when … So, I think, in the last couple of decades caesareans have increased from about one-quarter to one-third of births in America and they've maybe doubled worldwide? Obviously they still carry a lot of risks. Maybe some even better technology is diffused, like artificial wombs or, who knows, something like that? What would happen at that point to human brain sizes? Would selection pressures continue to push for larger and larger brainstor, or would it break the other way, and because of cumulative culture the pressure for larger and larger brains is no longer what it was? So if we remove the constraint of the birth canal, what happens to brain sizes?

RICHERSON: Well, that's only one constraint. There's metabolic constraints in the other ones that we talked briefly about earlier: fragility of the skull and things like that.

WALKER: But presumably culture helps us obviate those other …

BOYD: Yeah, we're pretty well fed these days in some cultures anyway.

RICHERSON: Right, some countries. Well, famously, evolution is not a very successful predictive science. So I think you're asking a very hard question that … contingencies will intervene. And even if we knew exactly how we would respond to a particular selective pressure, we would then have to predict the selective pressures, and that is hard to do. I mean, we can make short-term predictions about the evolution of pathogens, for example. Covid is … a guy named Paul Ewald. Have you heard his story about this?

BOYD: I know Paul Ewald, but I don't know his story about Covid.

RICHERSON: Well, it wasn’t a story about Covid. It was a story about respiratory diseases in general. This was ten or 15 years ago. He and I …  a student, he had a co-author, and they … The critical variable is how long the infectious agent persists in the environment. If it persists a long time in the environment, then it doesn't have to care very much about the health of the individual that gets sick. So anthrax forms a spore and so it lives for a very long period of time in the environment, and it's a devastating disease if you catch it. Things like the common cold don't persist in the environment very long. So they have an interest in not harming their host too much. And Covid, the initial … The time that Covid spends alive in the environment is, I think, just a little longer than the average flu virus. So the prediction from their model would be that Covid would evolve from being a very virulent disease to one that was much less virulent, something in the order of the seasonal flu, which is more or less what's happened. Those kind of short-term predictions can be quite successful. 

And the evolution of certain kinds of adaptations that are tightly constrained by mechanical principles. Flying and swimming. You can predict what sort of pursuit strategy hawks have from the size and the shape of their wings. In other words, if they're manoeuvering tightly to catch things in a forested environment, they'll have short, blunt wings. If they're diving from great heights, like peregrine falcons on an open environment, they're taking a completely different strategy.

So lots of things are predictable. But brain size in humans, I don't know.

BOYD: And what the selective pressures are …

RICHERSON: What the selective pressures will be. Yeah.

BOYD: Nowadays … I mean, you're going to talk about the fertility crisis. It's not clear what it is that's driving differences in fitness.

WALKER: Right.

BOYD: Whether it's how clever you are or not is …

WALKER: Oh, interesting. I hadn't made that connection, but that makes a lot of sense.

BOYD: Yeah. I mean, when it is how clever you are, then you might think the selection gradient would lead to larger brains and there would be some kind … But it depends – on everything, right. On energy availability and everything. But in modern societies, at least urban ones, I don't think it's so clear what selection is doing.

RICHERSON: Well, I mean, we've got all these crazy scientists running around publishing papers and neglecting their duties as fathers and mothers.

WALKER: Exactly.

RICHERSON: So the big brains are leading people to adopt, so to speak, crazy hobbies that detract from their fitness.

WALKER: Yeah.

RICHERSON: Like climb mountains.

WALKER: Yeah. As you'll be doing next week, I guess.

BOYD: Yeah. Yeah.

WALKER: So maybe because we've segued into the fertility crisis, let's talk about that now. And then I'll come back to artificial intelligence after that. Your way of making sense of the demographic transition is one of the most fascinating and compelling that I've encountered in my reading on the topic. Could you outline how you make sense of the demographic transition? And then I'll ask some more specific, sophisticated questions about it.

BOYD: I think this is your idea, Pete. I think you …

RICHERSON: Okay, well, so in our ’85 book we have a little vignette about the demographic transition in which we attribute it to … basically to careers open to talent. If education becomes a major source of economic success and prestige, then people who spend a long time furthering their education and start their families late will have an advantage in getting into prestige positions, like army officers and teachers and government officials and things like that. So you're sacrificing your genetic fitness for your cultural fitness, so to speak. And that was our basic idea in the ’85 book.

BOYD: And I think that the other thing that's in there is that it has to do with modernisation, right? So it's not wealth per se, income, that is the key thing. It's the rearrangement of the economy and the society such that prestige roles are required; delayed marriage and investment in credentials and stuff.

WALKER: The key difference there being you can become prestigious not through birthright but through effort?

BOYD: Yeah, but I think that goes … I wouldn't think about kings and earls and stuff like that. I mean, think about a village in a modernising society. Who are the prestigious people? You know, the guys who have an outboard on their boat. I worked in Fiji a little bit. And  it's the teacher, it's the local policeman, it's people that have some cash, a salary. It's not the best farmer in town. That's where it was before. It used to be the guy who could really grow yams better than anybody else.

RICHERSON: Feed six kids instead of four.

BOYD: Now it's … And to get those roles, you have to go to high school. In Fiji, you have to learn English. All these things require investments that are different from the investments in producing kids. And so the data suggests that modernisation is a better predictor of the demographic transition than GDP per capita or something like that.

RICHERSON: Right. So the former Soviet countries, for example, never got very rich, but they modernised in the sense that education became universal. Prestige roles were achieved through education.

WALKER: So the key change is not so much moving to meritocracy, but it's opening up those channels of non-parental transmission.

RICHERSON: Exactly.

BOYD: Yeah. I mean, in the old days it was just the village, right? And now people are going to town sometimes and they're seeing who's in charge there and how cool it is. And I don't think … I mean, these are still highly kin-based … Relatedness still matters for a lot of things. But less than it used to.

RICHERSON: Yeah. And, my wife, Leslie Newson – you've read part of our book – and her idea is that – sort of a generalisation of the original argument that we had, I think – social networks have changed dramatically. So the proportion of kin and social networks is declining. And the proportion of factory mates, office mates, friends, that don't have much interest in your reproductive success, in fact that might be a handicap on your friendship with the guys you go to the bar with on Saturday night.

BOYD: You can't buy as many beers.

RICHERSON: Yeah. Those dads that stick at home and don't buy their buddies beers anymore. We're not going to encourage that kind of behaviour. So, as Rob says, I think modernisation is the sort of master variable here. The switch from subsistence farming or nearly subsistence farming to factory work and office work and urban life.

BOYD: One of the predictions that I would love to see somebody test would be: I'd predict the same thing for urbanisation expansions in the past. So if we could go back to Tenochtitlan or Rome and get good measures of fertility, I'd predict the urban people would get sucked up into the same thing because they have these wider social networks. Prestige networks are …

WALKER: I think Rome was really good at just continuing to suck the rural population in.

BOYD: All those cities were, because they were death traps.

RICHERSON: So that's partly …

BOYD: That confounds a little bit, the …

RICHERSON: Moving to the city was a little bit like a suicide, but still.

WALKER: Still is, to an extent. 

BOYD: Still, one of my … I'd like to see somebody … There is demographic data for some of those places. Somebody needs to really work on that …

RICHERSON: It seems to me … Leslie would know better than I. Isn't there some decent data for early modern or mediaeval jews in Italy? That they underwent a precocious demographic transition, that …

BOYD: They would be exactly who you predict.

RICHERSON: That would be the kind of people that you would predict. On the other hand, we have these North American Anabaptists that are going after it. I did a back of the envelope calculation once, and I guessed that, projecting present demographic parameters, in about 200 years half of North Americans will be Anabaptists.

BOYD: They'll be Mormons instead.

RICHERSON: No, the Mormons. I looked at the Mormons when I first got interested in this, and the Mormon demographic transition has mostly happened by now. My former student, who's a Mormon, an observant Mormon, has three kids.

BOYD: Two kids.

RICHERSON: What's that?

BOYD: He has two kids.

RICHERSON: Three kids.

BOYD: Three kids. Adrian.

RICHERSON: Yeah. So he's still a little above the average, but not dramatically.

WALKER: So, one argument about the urban phenomenon is that there's some deep genetically evolved switch in our brains that once we're surrounded by density and crowds of people it causes us to downregulate our fertility in response. Does that seem likely to you?

RICHERSON: Well, one of the observations that people make of hunter-gatherer subsistence camps is they're tightly crowded together. They don't live …

WALKER: Interesting.

RICHERSON: They don't live miles away from each other.

BOYD: In a given camp, there are often 50 people.

RICHERSON: Yeah.

WALKER: Living in each other's hair, so to speak.

RICHERSON: Yeah. Cheek by jowl.

BOYD: Denser than the village.

WALKER: Well, I guess that largely dispenses with that hypothesis.

BOYD: Yeah, it depends. They could be averaging. I mean, they're off … They're quite mobile … I don't know. Yeah, I'm sceptical of those kind of arguments.

WALKER: So, in your model of or your understanding of the demographic transition, there are different cultural forces that are contributing to the spread of the maladaptive fertility-reducing cultural variant, one being content biases, of which Gary Becker's rational-choice model would be a special case.

RICHERSON: Yeah.

WALKER: Am I right in thinking that prestige bias is the most important cultural force relative to the others?

BOYD: We don't know.

RICHERSON: It'd be interesting to try to actually put numbers on that, because Leslie's kin influence hypothesis also suggests that when you're having beers with your workmates rather than with your brother, it's not that your workmates have any particular prestige, but they're diluting your social network and people that are interested in your fitness. In other words, your relatives tend to … They ask you, ‘Geez, you guys should have a kid, shouldn't you?’ You get that thing from grandmothers and mothers and fathers and maybe uncles and brothers and things. You don't get that from your workmate so much. So it isn't necessarily only a prestige thing. So it'd be interesting to try to get a quantitative handle. It might be something you could get a quantitative handle on.

Now, the Anabaptists, to take an example where the demographic transition is not happening, they do both things. They seal off the prestige hierarchy. They have their own internal prestige hierarchy. And they seal their prestige system off from the rest of us quite tightly. Things like movies and … They seal themselves culturally off. And they maintain small communities that are rich in kin. So they are working on both of those things simultaneously.

BOYD: Some, like the Hutterites, are quite well off. I mean, economically.

RICHERSON: Well, yeah, they have to be economically well off to raise seven kids.

BOYD: And buy all the property and buy all the.

RICHERSON: Well, the Hutterites actually … I've read. So the demographers don't do retail science, they're not like anthropologists. They don't go out there and count heads in the villages. They depend upon government data. So the actual data on the Anabaptists is quite poor. But I read it said that the Hutterites are actually reducing their fertility some, because their style of farming is very modern and requires a huge capital investment. They don't only have to buy land, they’ve got to buy tractors and all.

BOYD: Of the pickup trucks.

RICHERSON: Pickup trucks and all of the accoutrements of modern farming. And they have just been growing too rapidly to be able to divide the colony on the timescale that they think is appropriate. So they're scaling back their fertility some. Not a lot, but some. So completed family sizes are more like five or six than seven, something on that order. 

The Amish, at least, have gone into wage labour on a large scale. So what they do is they make a deal with some capitalist entrepreneur that they'll furnish the labour force for some kind of factory. And Amish are pretty skilled people. They grow up on a farm. They can fix machinery and they can build stuff. And so they build things like Winnebagoes and other semi-skilled manufacturing jobs that are consistent with their relatively low level of education. And tourism in some parts of the Amish land, like Lancaster County, Pennsylvania, that is a big absorber of labour. So they've to some extent shifted out of the farming business, maybe in part because of land costs, although some of them have moved further west into Ohio and places like that, Illinois, where they can more easily afford the land. 

WALKER: But they still isolate themselves from 

RICHERSON: The cultural isolation. So the whole deal on these factories is that the entrepreneur will hire only Amish labour, so that the Amish don't have to mix with non-Amish on the job. The entrepreneur sometimes, I think, is not Amish, but other than that they don't have to. Their cultural isolation is not threatened on the job.

WALKER: Right. Yeah. Interesting. I read that their population, I think, is doubling every 20 years or something like that?

RICHERSON: Something like that, yeah.

WALKER: It'd be interesting to see whether the world's Amish in a few centuries.

RICHERSON: Yeah. Well, I mean, the other group that's a little like this is the ultra-orthodox Jews, and they're numerous enough now in Israel to be a political problem – or a political force, let's put it that way. So not to prejudge. Yeah, but who knows what … It's hard to predict what will happen to that whole phenomenon.

WALKER: But I just want to focus on the prestige bias story, because I do find this a very powerful explanation. The way it works is: in modern society, to become prestigious it often helps to have fewer kids, because then you can extend your education, focus more of your time and efforts on career advancement. And the people who become prestigious are the people who are most likely to be imitated. And so in modernity, the people being imitated happen to be people who typically have fewer children than in pre-modern societies. Just so I'm clear, is the imitation happening for reduced fertility per se? Or is it more happening for the package of behaviours that produce success, of which reduced fertility is a byproduct?

RICHERSON: I bet, with number two. 

WALKER: Number two.

RICHERSON: Byproduct, yeah. That's certainly the argument we made in the ’85 book.

WALKER: Interesting. Because at least today, anecdotally, if I speak to female friends they'll tell me that there is a sense that having lots of kids early is uncool, or … calling it socially taboo is probably pushing it too far. But it does seem like there has also been a norm that's kind of congealed around having less kids per se, in addition to the package of things that lead to career success.

RICHERSON: Yeah. Especially in … East Asians seem to have carried this to Koreans. I think Korea has maybe the lowest or one of the lowest TFRs in …

WALKER: The world. It's down to one …

RICHERSON: Point something or even below one.

WALKER: I think it might be 0.9 even.

RICHERSON: Yeah, that's what my recollection is. So, again, there's the prestige angle and then there's the social network effect, and those two are partly decoupled. So that both effects are operating in the same direction.

BOYD: And things get turned into norms. I mean, I don't know if this is ... So there's quite a bit of evidence that people just look around and they say, what's everybody doing? That must be the right thing to do. So common behaviours become moralised or prestigised. And there's guys like Granovetter and there's a bunch of sociologists that have unmotivated empirical data that suggests that's … They don't start from where we're starting. They just …

RICHERSON: I notice this in myself. Somebody tells me that they have five kids, and I go – yeah, you think, wow, that's ....

WALKER: Yeah, yeah.

BOYD: So the data on things like if you tell kids, oh, you shouldn't drink, drinking's bad, that has no effect. But if you tell kids, nobody's drinking, that has a big effect. People have done experiments, and there's a bunch of pluralistic ignorance. So everybody … not everybody, but the average kid on college campus thinks that other people are having a lot more fun than they are and are drinking more and having more sex and all this stuff. So if you tell people what's actually going on, it motivates them to drink less. 

RICHERSON: To drink less.

WALKER: Maybe the process starts with model-based transmission, and then conformist bias puts the final nail in the coffin.

RICHERSON: There's conformist bias, too.

BOYD: Yeah. Conformist bias would just make it more common. But I think – we haven't ever written anything about this – but a good general rule would be if you're a little kid and you're growing up, if everybody's doing something that must be what we're supposed to do around here.

WALKER: Yeah. Just in a Bayesian sense.

BOYD: Yeah, exactly.

PETERSON: Yeah, yeah. 

WALKER: So, I've got three more questions on the fertility crisis. I think these are even the most important questions. So, okay, a few disturbing facts. The demographic transition began in the West about 1870, but obviously it happened unevenly across countries. As far as I know, no Western country has reversed its demographic transition?

RICHERSON: I believe that's true, yes.

WALKER: Moreover, fertility declines have also penetrated into the lower socioeconomic levels of society within Western countries? And then in addition, lots of non-Western countries have begun their demographic transitions?

RICHERSON: Almost all.

WALKER: Almost all. China and India are below replacement. You mentioned South Korea, Vietnam, a bunch of countries in South America

BOYD: Just not Africa.

RICHERSON: Sub-Saharan Africa.

BOYD: Yeah.

WALKER: Africa seems largely unaffected at this stage.

RICHERSON: I'm not sure that's true. The data I've seen suggests that it's proceeding much more slowly in Africa than it has … proceeded very rapidly in East Asia, but it's very slow, if not stalled, in parts of sub-Saharan Africa.

BOYD: Urbanisation is much slower in sub-Saharan Africa too.

WALKER: Does the fact that fertility-reducing behaviours don't seem to be being fixed very quickly imply that cultural group selection has weakened and that we have low variation?

BOYD: Why would cultural group selection favour larger families? That's the assumption behind that question. 

WALKER: Maybe there's a different path into the question. I think the assumption is, if this isn't being fixed quickly, does that imply that we have low variation now on this trait …

BOYD: We do have low variation.

WALKER: And we're in some kind of, like, global macro culture?

BOYD: Oh, I see.

WALKER: I guess the archetype is like the Davos-style elite. Because we're all interconnected through telecommunications technologies.

BOYD: We do have a macro culture in the sense that urban life, bureaucratic government …

RICHERSON: Hollywood movies …

BOYD: Educational systems that take a lot of time, that's shared more or less, to varying degrees, across the world. I mean, development agencies are trying as hard as they can to make everybody else do it too.

WALKER: Am I making an analytical error? Could we infer from the fact that the demographic transition isn't being fixed in all these different countries that cultural evolution is happening too slowly and therefore there must be low variation?

RICHERSON: Well, the trouble is … you mentioned non-parental transmission. Cultural culture, in lots of cases, will result in adaptations that would fix problems with genetic fitness, but that's not guaranteed at all. So cultural evolution can favour fitness-diminishing behaviour.

So selection on cultural variation will not necessarily be correlated with the selection on genetic fitness. Globally, overall, in human evolution, it had to have been true that there was some kind of correlation between cultural fitness and genetic fitness, or else culture wouldn't have evolved, right? It would have been extinguished by selection. But …

BOYD: That doesn't have to be true for even the last 10,000 years.

RICHERSON: No.

BOYD: Especially if you think brains are getting smaller.

I don't see any clear reason to predict that cultural evolution should favour higher genetic fitness, except that our brains have evolved to … We like sweet things and we don't want to die and lots of fitness-enhancing things. But if cultural evolution has finessed that by creating prestige systems, I don't see any reason it can't persist.

RICHERSON: It might persist to extinction of humans.

BOYD: Well, I mean, there's always … there’s the so-called Irish elk case, which is probably not true.

RICHERSON: Probably apocryphal.

BOYD: Sexual selection can lead to ridiculous ornaments that reduce average fitness.

WALKER: Like the peacock's tail, which is the classic example.

BOYD: In theory. I don't think actually there are good examples of this. And the classic example is the irish elk, which is this extinct species of elk that had these truly enormous antlers. And some people have … but I don't … They're long gone. And who knows? Lots of animals became extinct at the end of the Pleistocene.

RICHERSON: Mostly at the end of a spear, I imagine.

BOYD: Yeah. So anyway, I just don't …

RICHERSON: On the other hand, if you want to think about the Anabaptists, the correction is going right along, right? It will only take less than a millennium…WALKER: As long as they can remain insular.

BOYD: Yeah. And get much more powerful. Technologically, much more powerful. I mean, this may happen in Israel, right? That secular Israelis will get tired of this and squash them.

RICHERSON: Yeah. That's one scenario.

BOYD: Who knows?

RICHERSON: Yeah, who knows? But, yes, I don't think there's any law-like process that will force a strong correlation between cultural fitness and genetic fitness. They're, I think, important weak forces. I think of the emotional system, for example, as being the main sort of corrector. Appetites and emotions. You mentioned food, kids.

BOYD: People love kids.

RICHERSON: Yeah, people do love kids. We love our kids. So many of them. Most people who have kids will tell you it's the most important thing they ever did, even if they only had a couple.

BOYD: Yeah. So if selection created a mechanism where you can get a lot of satisfaction from two kids, then it won't work very well … It might have worked fine when there was natural fertility, but it doesn't work anymore. 

RICHERSON: And we got a lot of people in the world, so we’ve got a lot of scope for something happening before we go extinct.

WALKER: There is indeed some variation. There are also a lot of projections that show global population peaking around … There's a UN projection showing it peaking around 2080. How much does that worry you from the perspective of we need a large and increasing collective brain in order to sustain a technologically advanced civilisation? I know, Rob, you did some work on fishing technologies in Oceania, and the islands that were more populous and better connected had a greater number of fishing technology types and more-complex fishing technologies.

BOYD: That's all true.

WALKER: There's also obviously the salutary lesson of the Tasmanian Aborigines in Australia. When the sea levels rose and cut them off from the mainland – fully isolated them – about 8000 years ago, they slowly lost their technologies. By the time Europeans encountered them, they had a toolkit of about 24 tools, which is the smallest toolkit in human history.

BOYD: No boats.

WALKER: Yeah. And as an interesting control, as a natural experiment, just across the Bass Strait, when the local Aboriginal clan or tribe there was … when European contact was made with them, they had hundreds of tools in that toolkit. 

So to the extent that we need a really large collective brain to sustain technologically advanced civilisation, are fertility declines a worry? We're recording this today in the Internet Archive headquarters. Now that we've digitised so much of culture, is that a countervailing force? Or how do you think about that problem?

BOYD: We talked about science a bit earlier, but we now have institutions for innovating and cultural evolution that are evolving on their own. The engine of scientific and engineering progress I don't think depends so heavily on the population size as the diffuse, informal processes of innovation that happen in the village.

RICHERSON: So things like formal education have the effect of greatly increasing the exposure of any given individual to cultural ideas, particularly with regard to things like technology. So somebody who's got an EE degree or a computer science degree has been exposed to a huge amount of technical knowledge that they'd never pick up without formal …

BOYD: And much more efficiently than … 

RICHERSON: At least that's what we college professors claim.

BOYD: Pardon?

RICHERSON: It's what we college professors claim.

BOYD: Yeah.

RICHERSON: Well, yeah, some people are sceptical.

BOYD: ASU is trying to measure what kids actually learn. Seems like a bad idea to me.

RICHERSON: Might not like what you find out.

BOYD: I don't know. The other thing is the models. Joe's model and the Shennan …

WALKER: Joe Henrich. 

BOYD: Joe Henrich has what he calls a treadmill model of this. And Steve Shennan and Mark Thomas and those guys had more like a drift-based, a random loss model. Joe's … Well, it doesn't matter. The point is they both show very striking diminishing returns. So you get a big effect for when populations go from 100 to 1000, a smaller effect 1000 to 10,000.

WALKER: Oh, interesting.

BOYD: So those plots you see in our data are on log scales.

WALKER: Oh, okay.

BOYD: And if you actually plot them out in linear scales, they nose over, because the sampling … They're both based on sampling. Joe's model works. You learn something but, as we know, students don't learn everything from their teachers, so things tend to go backwards a little bit through each cultural transmission process. But it's exactly like the experiment I talked about earlier. By copying the most successful people, that brings the mean back up again. And there's some balance and actually …

RICHERSON: There’s square root of n factor in the models, literally?

BOYD: So Joe's model is very stylised and actually it has a threshold. And either technology goes off to infinity or it goes to zero. But if you put in … It takes longer to learn more complicated things. So Alex Massoudi modified the model in that way. And it shows diminishing return, just like you'd think, to population size. I think 10 billion is a lot of people. And if were down at 100,000, I would be more worried. 

And then couple that with the fact that we've made the system much more efficient by routinising it and institutionalising it and made it less sensitive, I think, to these diminishing returns kinds of things.

I think specialised institutions where incentives have been norms and incentives have evolved to create more efficient accumulation of knowledge. That has a big effect obviously.

RICHERSON: So in my own scientific practice, the availability of scientific literature on the internet has been a big revolution to me. I can read far more papers if I can harvest them off the internet than if I have to traipse down to the library, photocopy them …

BOYD: The proof of that is in the citation rate of articles and books.

RICHERSON: Yeah. Because of the...

BOYD: You can't grab them off the internet, whereas journal articles …

WALKER: So a couple of questions about artificial intelligence. Does your understanding of how human intelligence evolved give you any unique insights into how artificial intelligence might be created?

RICHERSON: Well, I attended a conference on AI about six weeks ago or so, and so I boned up just a little bit. And Alison Gopnik, do you know who she is? She's a developmental psychologist here at Berkeley, and she argues in a paper that artificial intelligence has nailed culture. That artificial intelligence, these large language models have made the accumulated wisdom of the world available in a very efficient way. Now, it's also made the bullshit factor in culture equally exaggerated, right? So that's why the large language models have these flaws, because they're tapping into the craziness of culture as well as the sensible fraction of it. And she argues that … We also have learning models, individual-like learning models, models of innovation, but those aren't integrated with the large language models. And the kinds of success-based filters that we think are so important in making culture adaptive, it's not clear that they're built into the current thinking on AI. And other people make the point that AI has this problem that it's computationally extremely expensive. If we say, with Gopnik, that AI has aced the cultural part of it. What about the cultural transmission kind of part of it? What about the innovative part of it? What about the quality filters part of it? Can we integrate those into AI without making it prohibitively expensive?

WALKER: Yeah.

RICHERSON: Very unschooled thoughts.

WALKER: Interesting, nonetheless. Do you have anything to add to that?

BOYD: AI will be cultural. Yeah, it's got to be. But how it'll all work is a mystery to me.

WALKER: One thing I was contemplating was I think the modal scenario for a lot of people is you basically have billions of copies of AGI agents. For example, if we create AGI, and presumably they will all be interacting both with each other and with humans, and presumably they will be participating in our cultural evolution. But what's interesting is that the forces of cultural evolution might change in that scenario. So presumably they will have less transmission bias than humans? Perhaps content …

BOYD: How their learning psychology gets … how teaching psychology is … Are they teaching each other or are people teaching them? I mean, who knows? It could easily be … They look around and see which bot is the most successful and copy that one. It could be.

RICHERSON: You should read Alison's paper. She's kind of a wicked lady.

BOYD: Oh, she's a smart one, Alison.

RICHERSON: Yeah, she is.

BOYD: And the guy I mentioned earlier, Bill Thompson, is part of the same intellectual stream. And also at Berkeley. And there's a whole bunch of those guys that are.

RICHERSON: Alison is a child development person. So she argues you're trying to replicate babies, and babies are already a much more efficient way of doing this than server farms burning fossil fuel as if it were not causing any real problems. That's why I say she's got a little bit of a wicked sense of humour. You might enjoy that.

BOYD: She's got a great book about child development from a cognitive point of view, with Laura Schultz. It's a bunch of papers too, but the great part of the book is this long intro that she and Laura wrote in the form of an email exchange. It's like a dialogue between a mathematical cognitive scientist and an experimental child development person like Alison. And it's great. I mean, I highly recommended it.

RICHERSON: Yeah, I’ll look it up. I don't know that.

WALKER: Thanks for the recommendation.

BOYD: She's a good one.

WALKER: So I have a million more questions for you guys, which is obviously an indication of just how interesting and rich your work is. But we are running out of time. We're about to get kicked out of this beautiful room that we're recording in. So my final question is, you two, for me, are a very successful scientific dyad. And there are many interesting examples of other scientific dyads, Watson and Crick being a famous one. Marion Pierre Curie. There seems to be something special about the number two that in certain contexts is perhaps more productive than either loan researchers or group sizes of three or more. And firstly, I'd like to invite you to speculate on why that might be from an evolutionary perspective. Is it maybe piggybacking on pair bonding, or is it a culturally evolved institution, the scientific dyad?

And then secondly, I'd like to understand how each of you thinks about the other’s comparative advantage in your scientific partnership. So, Pete, how you think about Rob's comparative advantage. And Rob, how you think about Pete's comparative advantage.

RICHERSON: I think we're friends, right? Forever … for 50 years? Yeah, 50 years.

BOYD: Actually, 54 years. So Bill Hamilton's living room.

RICHERSON: Yeah.

WALKER: You met in Bill Hamilton's living room?

BOYD: No, different Bill Hamilton.

RICHERSON: Davis Bill Hamilton. Rob was a first-year graduate student the year, 1970, I was a first-year faculty member. And we met at a seminar at this eccentric professor's house.

BOYD: This is high ’70s, orange shag carpet, conversation pit, the whole nine yards.

RICHERSON: Well, Bill Hamilton's another whole story. We enjoy each other's company, and I think that's the main thing. I mean, we have this great project that we put together, but it'd be easy to feel resentful, and “Rob got more credit than I did,” or vice versa. And both of us have certain eccentricities that would, in a different context, lead to resentments or “I'm tired of this.” But we never went that way. So I think that is really important.

BOYD: Maybe it's evolution in some deep sense, but I think increasing the number of collaborators increases the diversity of skills and knowledge, and that's very beneficial. And we're going to talk about that in a second, I guess. But the bigger you get, the more the free rider problem raises its head, and tpeople get lazy, people want credit, all those kinds of things. Two may be somehow in the sweet spot between …

WALKER: Well, if someone's not pulling their weight in a partnership, you know who it is.

BOYD: Yeah, exactly. Whereas if it's even three or four, then … 

RICHERSON: Both of us have participated in far larger projects, quite productive ones.

BOYD: So there's that. And I guess I'll start on the second part of your question. We bring, I think, somewhat different – complementary but overlapping, but quite different skills. I'm more of a modeller, and Pete knows more than anybody I've ever met about everything. At least everything important. Maybe not who wrote what poem, but science, especially the more environmental, ‘how the world works’ science. I think bringing those two skills together was quite productive.

RICHERSON: Yeah, I agree with that. I had a friend in graduate school who was a quite accomplished population geneticist. He's in the National Academy by now. And so when I got interested in cultural evolution, I knew my math skills were – not without a hell of a lot of work – not up to what was required. So that's one of the reasons I first tried to interest Rob in this project. Well, it wasn't a project in the beginning. I was just looking for material for one lecture, and it turned into a project.

BOYD: Cavalli-Sforza and Feldman.

RICHERSON: Yep. And so then it turned out that Rob's a hell of a good scientist. You know, he thinks like a scientist is supposed to think, sceptical when he needs to be sceptical, and never takes any bullshit off anybody.

BOYD: Yeah, people are complicated and the fit was good. And then, by chance, I think, we had a good idea.

RICHERSON: Yeah, I just stumbled on this.

BOYD: Yeah, exactly.

RICHERSON: And, you could just see, when we started talking about this, there's all this low-hanging fruit. The population geneticists and the evolutionary biologists were 75 years ahead of the social scientists in this business of what a cultural adaptation was. They were in the 19th century, but they weren't.

BOYD: They were …. They’d decided they weren't going to talk about people. And that’s one of the things … 

RICHERSON: The evolutionary biologists. So the evolutionary …

BOYD: and the population geneticists don't talk about adaptations very much.

RICHERSON: And the evolutionary biologists who do talk about adaptations were in the thrall of the modern synthesis, I think. They couldn't imagine that they needed anything fundamental besides genes as an information carrier. They could see that culture could do some things, but the idea that it could be a fundamental force in human evolution was just beyond them.

BOYD: There are still plenty of geneticists and evolutionary biologists who think that way. But it's changing. The cultural evolution is getting … When we started, it was like chariots of the gods, right? It was crazy talk. What are you talking about, cultural evolution? And now you see it all the time. I mean, PNAS is having a big 50-year celebration about 50 years of cultural evolution coming around on the guitar. So I don't know. It's been a big success.

WALKER: That's really cool.

BOYD: Yeah.

WALKER: When you were talking about your complementarities, Rob, you mentioned that Pete has this encyclopaedic knowledge; you have more of the mathematical.

BOYD: I don't think it's so much… I mean, I'm only a mediocre mathematician. I have some skill at figuring out how to make simple models of things that are.

WALKER: The way someone described it to me was that in the partnership, Pete brings the variation and Rob brings the selection.

BOYD: Yes, I think I've actually said that myself, so I agree with that.

RICHERSON: Yeah. I have a speculative tendency. It can sometimes get out of control.

WALKER: It’s served you well. Well, it's been a great pleasure speaking with both of you today. Thank you so much for coming on the podcast.

RICHERSON: I've certainly enjoyed it.

BOYD: Yeah, it was fun.

WALKER: Thank you.