Examples of Good Evolutionary Psychology
2025-07-01 12:00:00.000 +00:00

Okay. Uh I I am me and you are you. Yes, and that would work, except there's this thing David called uh common knowledge. You might I'll I'll tell you all about it later. Okay, so you might not know who me is. Exactly. There uh you are uh y yeah. Yes, you are not me is not sufficient. I'm the other David P David Pinsoff. I'm with uh my friend and co-host Dave Petrozowski. That is me, which in this case context is sufficient because it gets complicated. It takes a while to wrap your head around it. But anyway, we're here. Boy, this is really uh interesting conversation. People are gonna be turning podcast if they can't explain this, how are they going to explain like how the mind works, right?

Which is actually a question I ask myself every day. So yeah. And we're doing the evolutionary psychology podcast. Even that I screwed up. Evolutionary psychology, the podcast. It was a good throwback to the first one. Yes. Um and uh we're doing a fantastic job introducing this episode. Hey Dave. Hey David. So we've done a few interviews now. Uh and this is the episode where we get back to just you and me, the Dave's putting their heads together. And I think this was a really fun, wide-ranging conversation. Yeah. This is uh instead of sort of being on our looking backwards, this is sort of looking forwards or positively saying, Well, what is evolutionary psychology and what are some examples of really great work that explain what it is and give meat

to the idea that it makes predictions and gives people insights. And so I think if especially for those people listening who want some concrete examples of insight that can be provided from this perspective. And yes, and not just insight, but important, socially relevant insight. So we spend a good chunk of the episode talking about race and uh racial categorization, racial prejudice, and what evolutionary psychology can tell us about this uh important topic. Uh and we really get into the nuts and bolts of it in a in a fascinating and and mind-blowing way. Um from there we go go into anger, we go into happiness. We cover a lot of stuff in this episode. It was it was really fun. Yeah, absolutely.

Yeah. And um and it was sort of fun because the some of the work is some of some of my empirical work that I do sort of in my day job. Um so um yeah so I'm the horse's mouth in that situation. Yeah. So without further ado, enjoy how to be a good evolutionary psychologist. David, good to see you today. Good to see you too. Excited to get into the nitty-gritty of how to actually be a good evolutionary psychologist. Yeah, that's right. We've had a couple guests on, but we haven't really talked about sort of what evps is positively per se. And I think it would be really helpful, especially for people who are listening who aren't part of the regular community of people, to just get a better idea of sort of what it is.

Yes, I agree. There's plenty of discourse surrounding whether evolutionary psychology is good or bad, or whether we should take an evolutionary approach or what kind of evolutionary approach we should take. But there's very little discourse surrounding the very large difference between good evolutionary psychology and bad evolutionary psychology. And I think any person in the field would agree that there is plenty of bad stuff in there and plenty of good stuff in there. And I think the wider public has no way of telling the difference between those two things. Would you agree? I would agree, yeah. And I think also, I mean, this is the problem with the format of scientific dissemination is a lot of stuff is behind paywalls or it's in sort of boring journals.

I think one of the things that is really cool that I hope people will start to get a sense of is what a treat it is to talk to people who are actually doing the science because they're really cool people. They have lots of cool experiences and they're studying really cool stuff. And I think one thing I tell people, like my students in classes, is that textbooks are usually anywhere from 10 to 20 years behind. Journal articles are, you know, three or four years behind. Because if you think about it, you know, you have an idea, you test the idea, you run the study, or you write the paper. It takes a while to then get accepted. Talking to people, you really catch up. And so, you know, if you're reading a popular book or a textbook, you're 20 years or 15

years behind. If you're reading journal articles, you're three or four or five years behind. But if you're talking to people, you really get the update, up-to-date stuff. And that's a real treat to have. And so that's the other thing is so much of the empirical work and theoretical work is so new. Like David, your we'll talk about your humor work in some episode. I just saw that a couple months ago. I was completely blown away. Um, um, John Tuby, my uh recently um he passed away recently, but he was one of the founders of Ev Psyche, and he would famously sort of say you shouldn't study humor as your first project because it was such a trippy, tricky thing to do. Um, and I think uh humor's always been one of these sort of um things that are difficult in

he thought it was it was tricky in the sense of mysterious or or um people won't take you seriously as an academic if you're no no no no no the first one he he said sort of you know you should start a project with clear selection pressures and an obvious design. So sort of sort of something like, you know, uh not starving to death, you know, reproducing clear survival, reproductive uh benefits, clear selection pressures, humor, things like that. He he put humor in this category of it's there's such a number of layers between clear design and the phenomenon that it would be very tricky, especially for an early career student, to do. And I think seeing your talk, you lifting the veil of actually, there's only a

few straightforward designs involved. Uh, it was just to your credit. Uh I'm sorry, I sort of feel like I'm doing a teaser here. Uh so we should uh but anyway, yeah, no, it was uh really good and cool insight. But I think the point is that's an example of cool new work that it's exciting to talk about or share in this venue as well. Yeah, it's not very widely understood just how young the field of evolutionary psychology is. So it started in the 90s. Uh um, I mean, sociobiology was sort of the early beginnings of it. E.O. Wilson was beginning to bring ideas from evolutionary biology to study the social behavior of animals. And in his classic book, Sociobiology, he had a final chapter on humans, where he started

to apply some of these ideas from evolutionary biology and the Neo-Darwinian synthesis to humans. And of course, it was extremely con controversial when he did that. But uh I think it took a couple decades to find the right way to apply evolution to human behavior. I mean, you know, Eo Wilson got some things right, But I think the the framework of sociobiology was not psychological enough. Um, and that is where Tubian Cosmodes came in in the nineties with evolutionary psychology, um, where the adaptation um in question was not behavior, but psychological mechanisms, cognitive systems. There are many ways to describe them, but the stuff in our head that takes in information and guides behavior, that is the appropriate target of scientific inquiry.

That is where we should be looking for adaptations and for um functional organization. We should not be looking at the level of behavior per se. We should be looking at behavior regulation systems inside the nervous system. That's right. Yeah, the the level, the invariance is is the design. And a good example of that would be something like plants growing towards light. If you tried to have a science of describing where all the plants are growing, it would be complicated. Plants are everywhere, There's they're growing in all kinds of directions. The level of invariance is there's something phototropism, the auxins reduce the plant growth. And so that you can economically describe all plant behavior.

I mean, there's other designs too, but in terms of photophotropism, the invariance is the underlying rule of you know, bend towards light and as you're growing. it's a beautiful analogy because if you were to try to study plant growth as the target of scientific inquiry, you'd be very confused. You'd have, you know, plants moving in uh this direction and that direction, uh to, you know, in this diagonal, that diagonal, and you would have no way of understanding why they were growing in which direction. And you could collect all this data about all the different trajectories of different plants, and you'd be incredibly confused as to what was going on. Uh And it is only when you realize the adaptive function of phototropism that you can

organize all that data in an elegant and parsimonious way. And you can look at the at the at the level of the growth regulation system within the plant that was designed by natural selection to find sunlight. And I think a similar paradigm shift is needed for human psychology. We shouldn't be looking at the uh superficial level of behaviors. We should be looking deeper at what are the underlying systems that were selected for to generate the behavior. That's right. And it's also this is a point that Paul Smaldino makes. And before that, Noam Chomsky, which is that in just as a general point in science, and Noam Chomsky has a nice example of this uh, where you could set up a video camera and record a bunch of leaves falling outside of your window.

Um, you could record that uh get a statistical analysis of how do those leaves behave. And one could probably get a better prediction of how those leaves will behave than even a physicist could do, right? Um, but the problem is that that sort of statistical description of those leaves is brittle and doesn't apply to new contexts. And this is Paul Smaldino's point, where if you have a mathematical description of behavior in a certain context, if a new context in a new context the behavior won't be the same, then you actually lose the predictive power of the model. So for example, those particular leaves falling in that particular place, you might have a good model of that, but you're not going to be able to model something else falling, right?

Because it's just different. And there's nothing in the theory about coefficient, drag coefficients, or things like that. And so one of the reasons scientifically to have an underlying model that is causal is that it allows you to make inferences to new situations. And otherwise, you're sort of stuck with descriptions of a particular set of behaviors, but you lose the predictive power of a scientific theory. That's one of the main differences between just a description. And we talked with Dan Conroy Beam in a previous episode about there are a lot of theories in psychology that are descriptions of what happens either in a study or just general behavioral phenomenon. But if you don't unpack the underlying causal invariances like

phototropism, then you're just stuck with statistical descriptions of those plants growing in those places, but you might not capture the right data in terms of like which direction does the light coming from. Same thing, the plants or the leaves falling and the video camera, if you uh extrapolate to a new context, that model might be useless. So it's very brittle. Um and so the point is a theoretical approach where the level of invariance is not behavior, but the underlying psychological design. I think this is sort of where we are as a field in psychology generally. And I think F-Psych was on the forefront in Lita and John's work in particular, really on the forefront of pushing that idea that the level of invariance is in the mind and the

software and the mind, not at the level of behavior. Yes, absolutely. Um, and certainly psychology is stuck in this rut where there is a proliferation of endless biases and effects that pile up. I think Wikipedia has a list of biases that's like hundreds long. This is the sign of a field that is stuck in a cul-de-sac, I think. Um if you're just collecting biases and effects without a deeper theoretical framework to organize them, um, you're just you're butterfly collecting. Um, you're looking at the wrong thing. Um so I think that is one of the great benefits of evolutionary psychology is it causes you to look deeper into the underlying structure of the mind, into what we might call human nature, for lack of a better term, the reliably developing

species typical uh developmental and psychological architecture um that uh has been produced by natural selection. That is the the place to look for adaptation, for function. Um and uh a science that doesn't look at that as the source of invariance, as you put it, is inevitably going to get lost um in descriptive cul de sacs and overfitting to data rather than finding the the underlying patterns. Yeah, I I I think that's right. I think you're what you're saying is that the evolutionary psychology is a kind of correction, and there are other ways to get there. Like I think there are areas of like AI or computer science or anywhere where people have there's sort of tasks that need to be solved that get at uh aside from just the

behavior, what is the underlying design that I think is is sort of a crucial thing. And we're we're still not quite there. One one thing maybe to shift to to make this concrete for listeners, especially. Uh let me pose a question to you and you can you can uh take this question if you want. So one of the common misunderstandings of evolutionary psychology is that it's just so story. And I'm curious uh if in in well, yeah, what do you how do you typically explain why it isn't that? Or do you have an example of a study that you're aware of, or an example research project, either your own or someone else's, that's a clear example of why it isn't a just so story? I mean, uh, I've been I've seen this this idea over and over again.

It's one of those zombie ideas that just won't die, no matter how much you correct it. Um so just so stories. The idea is that you take any human behavior and you spin a uh uh an evolutionary story about us as cavemen or whatever, um, and and you um make up some reason why it might have helped our ancestors survive and reproduce, and then your work is done. Uh and that's that's the criticism: is that evolutionary psychology does this. And I think it can be a fair criticism of some bad evolutionary psychology. So that's that's um a place uh where bad evolutionary psychology can go. If you don't uh have the right way of thinking and and and the right way of doing task analysis, you can sometimes get lost in this uh just so story trap.

So I think it's it's it's a valid concern and something to look out for. Um but just as an empirical matter, um the idea that evolutionary psychology does not make testable predictions um is false. Um what I find ironic about the just so story idea um is that it itself is often unfalsifiable. Like a lot of people will just lazily say that some idea I present is unfalsifiable, and then I'll say, Well, actually, no, here's how you could falsify it. And I'll I list uh a number of empirical tests that you could falsify. And then I I often engage you know in petty debates with people on the internet when I blog, and and you know, when every now and then when someone says this is unfalsifiable, I'll say, No, here's how you can falsify it.

And then what's funny is uh what I sometimes get is well, no, I don't believe you. Which which suggests to me that the idea that the hypothesis is unfalsifiable is itself unfalsifiable. Nothing could actually falsify it. No matter what I say, you know, the person will continue to believe that it's unfalsifiable. Um so uh I mean the best way to just dismiss this idea is just to like point to, say, the Journal of Evolution and Human Behavior, point to the latest uh, you know, two volume handbook of evolutionary psychology, uh, which I believe is uh over fifty chapters uh covering just an an enormous wealth of empirical research um that has been generated by evolutionary psychologists. Um so clearly they're not just sitting around telling stories.

Clearly these people are doing research, they're testing hypotheses, they're doing statistical analyses, they're doing something. Yeah. Right. Um so that's one thing I might do is just point them to that stuff. I don't know. I mean, there are plenty of cases where evolutionary psychology hypotheses have been falsified. Perhaps the easiest one to wrap your head around is this hypothesis about homosexuality. So homosexuality is kind of mysterious from an evolutionary perspective, and that you would expect uh heterosexual uh mating to be selected for and homosexual mating to be heavily uh selected against, given that you're not uh reproducing and spreading your genes. So, one hypothesis for why genes for homosexuality might persist, despite their uh

selective downsides, is that maybe it's a way, it's like a different fitness maximizing strategy. Maybe gay people are investing more in their kin, in their nieces and nephews. And because fitness isn't just about your direct reproduction, but about the reproduction of anyone who shares genes with you, maybe on net homosexuals are still doing okay evolutionarily evolutionarily if they're investing uh a lot in their nieces and nephews. Um the coin it's like helpers in the nest, I think. Helpers in the nest. No, as as a matter of fact, I don't think this is a very plausible hypothesis. Um but at the very least, it is falsifiable. It is very easy to test. It makes a very clear, bold prediction. And that prediction is that gay people should be absolutely effusive toward their

nieces and nephews. They should just uh absolutely adore them and invest tons of time and energy into taking care of their nieces and nephews, enough to offset themselves not having kids of their own. And you can go out and you can run the tests. You can uh ask gay people how much time they spend with their nieces and nephews, you can compare it, you know, it to uh straight people, and uh the data has come in and the answer is no. Gay people are not particularly uh effusive toward their nieces and nephews relative to straight people. So that hypothesis has been uh falsified. Uh, that is uh no longer a viable hypothesis about the evolution of homosexuality. So that's one clear example of a hypothesis that was falsified.

Right. And and I think there's also, I think this comes back to why people misunderstand evolutionary psychology and not only why, but how they misunderstand it, comes back to what we talked about, which is this idea of it's not about behavior but design. And where evolutionary psychology typically makes predictions is at the level of design. So what you just described was if this idea is right about that um if homosexuality, there should be design in this sort of type to be really strong and allo-parenting. And you don't see that. And other examples of design, a good example of this that is falsifiable but but turns out to be true. I, you know, I know this is just my own work, which is that one idea about like where it

is, why do people categorize people by race? People had shown that people always categorize people by age, sex, and race. This was looked at um using a parallel. So let's we should probably unpack what it means to categorize someone by something, because a lot of people might not know what that what that means. Yeah, yeah, yeah. Great. So categorization is uh a process whereby when you're meeting somebody or you're judging somebody or evaluating somebody, you're not just treating their individual characteristics. Oh, this person is nicer or you know uh has this particular um attribute. They're also an instance of a broader category, which is a set of people. So they might be uh all this might be a woman, this might be

somebody who is uh of a particular tribe or a race, like they might be black or white, whatever the local sort of cultural categories are, an old person, a young person. And this has been an important part of studying social cognition since the 20th century, especially in social psychology, because many bases of discrimination and unfair treatment are treating people not because of their individual attributes, but because of the category membership that they embody or people perceive them as embodying. So, for example, if you apply to my to to be a renter in my apartment, if somebody is using a social category as a basis to not allow them in, that's discrimination, right? And it's not that there's any problem with them as an individual, it's that because they're

an instance of this category. So category based discrimination is a really important social phenomenon to understand. And categorizing people, that is, when a new person is encountered, the cognitive process whereby their Their cues that they embody in their face or the how they look or talk or whatever, because they can be verbal, they can like you know how they talk can be a category. Um, these all activate the representation of the category. And understanding that process is really important because it's the categorization process that's the first step prior to any discrimination. If people didn't never categorize people in the first place, there would be no racial discrimination, sexual discrimination, et cetera, because there wouldn't be categorization

by this feature in the first place. So there was a long history of trying to understand not only the discrimination or the attitudes people held towards these categories, but starting around the 1960s and 70s in social areas of social cognition, people wanted to start to understand: well, what are the cognitive processes that cause people to categorize by social categories? In the first place, what are the social categories people use? And there were new methods to sort of study that. And one of the things I do in my day job as an empirical psychologist is study some of those cognitive processes. And one of the tools people use is something called the Who Said What paradigm, where people are shown photos of people along with a sentence.

So it could be something like you're told, oh, you're going to see a bunch of people, they're just all having a conversation, and then you see a picture of like a person and there's text underneath, they say something, and then there's another person, and they say something else, like, hey, nice to meet you. Like, you know, what are you doing today? And then somebody says something else, and they just this sort of a conversation. Um, and usually participants or or people in a study would sort of see this on a computer. It's been done in real life. You get the same kind of result. And then at the end, there's a surprise recall task. So usually there's a long conversation. Usually it's between like eight people.

You see about 24 statements, whatever. But at the end, there's a surprise recall task where all of a sudden you see all the faces or all the people that you had seen earlier, and then you're asked to try to remember who said which sentence. And it's a surprise task. And it feels really, really hard. Like shockingly, memory for what people have said or who have said what is quite bad. I actually make these stimuli, and I still, to this day, even the stimuli that I make, I'm like terrible at it. And I like it feels like random guessing. But the cool thing is that even though it feels like random guessing to the participants, their choices are not random typically. And what happens is that, so for example, if you present people in the conversation that

differ in their sex, so half of them are male, half of them are female. When people are given that sentence and they look at it and they're like, Hi, how are you doing? Like, what are you doing today? And they're like, Man, I don't remember at all who said this. And you just click on one of the faces you saw randomly. Feels like random guessing. You're not conscious of anything. But it turns out that, and this is a very reliable, replicable finding, is that what ends up happening in people's cognitive processing is that they're categorizing the statement not as just coming from an individual, but also a social category. So for example, this is a statement coming from a woman or a man. And what will happen is in the guessing process, the fact that that

statement was said by a man or a woman will cause them to be more likely to choose one of the men or one of the women, even if they don't remember the correct speaker. So, in other words, like if there's four male speakers, even if they get the male speaker wrong, the actual particular individual one, they'll be more likely to attribute it to another one of the men. So if you and I were talking to somebody and they forgot which one of us said it. Which David, yeah. Yeah, exactly. They'd be like, well, one of the Davids said it, right? Um and Or one of one of the men said it. Exactly. And and so, and it turns out then you can quantify for each participant the degree to which their errors cluster non-randomly, and to the degree that a statement that

had come from a man was attributed to other men, a statement that had come from a woman is attributed to other women, that quantifies in a very precise way to the degree to which that particular participant categorized by sex. And it's good because it allows you to measure categorization by, for example, sex in a way that's non-conscious. So there's no social desirability things. Because if you ask somebody like, oh yeah, no, are you categorizing people by like, you know, race, the people are like, no, no, no, no. but like they might be right they just know that you know that you're I'm not supposed to so you just sort of are you are you likely to confuse what one black person said with what another black person said oh no of course not no exactly but and but if you look at the

data they're they're actually doing that yeah exactly and so this is in this this who said what paradigm was used for basically 40 years and what people found was that categorization by many categories like social class, like sort of more arbitrary things, like maybe if they're a university student or a teacher, even visible things like clothing, styles, these were things that would sometimes be categorized as measured by this paradigm, but sometimes not. Yeah. So like it would just sort of depend on is it relevant or not? You know, like in we all have this experience of like, you know, sometimes we categorize people by like whether or not they are a supporter of a particular team or of a political

party or something, and maybe not other times, you know, like it's just not relevant. But um, but there were certain categories that people always use, um, regardless of whether or not it was relevant to the context. And one of those was um, well, one of them was sex, as we talked about, one of them was age, um, you know, usually broad, broad strokes of age, like old, young, um, kind of thing, you know. And then the third one was race. And uh, especially for race, people were interested in seeing if there was a way to reduce categorization by race. Because again, uh, for something that is socially problematic, like racism, uh changing categorization would be a way to change discrimination by race.

And so people immediately, once they discovered people seem to always categorize people by race. And this is usually like black versus white, like that, that broad stroke. There, there have been other um other kinds of categories that you find similar things. So, for example, um uh an analogous category sort of in Germany is um people who are um sort of a Middle Eastern or particular Turkish descent and that has been found. So like people will categorize by like sort of white European German versus Turkish in Germany, for example. So like visible, physically visible sort of like racial categories in a local context. This is sort of a what people have typically meant by race. Okay. I just want to lay the groundwork for the discovery that I think you're about to get to,

which was made uh by an evolutionary psychologist. So th i i i in this context of when when was this research being done? Like 80s, 90s? The first Who Said What paradigm was done in uh 78. And then from then on uh in up and through early 2000s. So this is two or three decades of research on this paradigm. And people are are finding that that the mind is intuitively categorizing people by race. Correct. And the holy grail that everyone in this field is looking for is how to we how do we get people to stop categorizing by race? That's right. The thought is if we can cut this at the root, and if we can stop people from categorizing by race at the beginning, then it will solve this downstream problem of people discriminating or treating

black people unfairly. All of these social problems that we as social scientists are trying to solve, we can cut them at the root by trying to eliminate this bias in the human mind to categorize by race. So everybody is looking for the holy grail of how do we get people to stop categorizing by race. Is that a fair characterization? That's right. And I think I don't, I'm not even sure they had that, uh, all of that in mind. I think it was some of it was just at the level of well, we're finding it is always a basis of categorization, and it would seem like a good idea to get it to not be. So let's figure out a way to do it. And people sort of stood on their head, I think, as a care a fair characterization.

Um, they would tell they would change the relevance of race. Many theories in social categorization were theories that predicted roughly that if a context is relevant to the category, then it should be categorized. Yeah. So relevance, yeah. Um, or stereotypicality. So is the context one where race is relevant for stereotypes, or is it counter stereotypical? All of these things telling people attend to race. So that part of the instruction would be attend to the race. One instrument another a different set of instructions was don't attend to race. Um you know the default is not nothing nothing changed. So so everything that the current theories of categorization uh were predicting uh weren't working.

Yeah. Yeah. And I also wanted to add that at this point, you can correct me if I'm wrong, people were starting to grow rather pessimistic about the possibility of eliminating this tendency to categorize by race. There was even some talk of maybe this is just a part of human nature, maybe this is just in our DNA, we're innately racist because of our evolutionary heritage. There was some bad evolutionary psychology coming out of social psychology at this time that maybe this stuff is just is hopeless and that we're just uh we're innate racists. right? Yeah, I th I think the the I think the that that um innate racist was sort of outside the professional vein of research. The professional vein of research, however, did was as you it's correct what you're saying,

it was not just resigned to it would be inevitable. They were to their credit trying to make claims about the underlying design. And so one of those claims was well maybe it's built into our psychology. I don't think they were claiming we're sort of innate racist, but um, but it it it sort of amounts to the same thing, which is that race, age, and sex are all really important. And they may be primitives to the human mind and sort of built-in categories to the human mind. And that was the conclusion, reluctantly, uh particularly in the case of race, based on these decades of failures to change categorization by race, correct? Okay. Yeah. All right. Now should we should evolutionary psychology enter the stage?

Yeah. So uh and this is where this is a really nice example of two principles. One of evolutionary psychology not just being a just so story. And the second thing is that having an idea of an ancestral environment as different than modern circumstances, both of these principles are really useful for what we're about to say and were really necessary for what we're about to say. So what ended up happening was, um, and so so Lita Cosmeties, John Tuby, um at the time Rob Kurzman had did some work, and then I've done a lot of this work as well. So sort of all of us as a team are are sort of responsible for this work. Um, and we're just talking about it because I know it so well. Um but but basically um what what especially I think Lita Lita

and John um sort of did in Rob were to was to look at this phenomenon of people always categorize people by race in particular. And then this additional claim of there's sort of software in the mind for categorizing people by race. This idea is a functional claim. This is saying there is a system in the mind that is well designed to categorize people by race. A claim about what the function of a particular piece of software is in the mind has to be consistent with the way natural selection works. And uh and this is just simply because this is the only way that you get complexity, as you put very uh eloquently in previous episodes, David. The only way that complex, non-entropic structure can get built into the mind, just like as in the body, is if that structure

is based around features of the environment that would have been present over evolutionary time. And those structures would have responded to those features that would have been present over evolutionary time in a way that would have been fitness benefiting, i.e., would have generated differential reproductive success on average over evolutionary time. So the idea is that if we think of human software as sort of a bunch of complicated input-output processes, the claim that the social psychological literature was making was that there is a system in the mind that is for categorizing people by race. And this is precisely where an evolutionary approach comes in handy. And I want to interject, this is also precisely the point where bad evolutionary

psychology could have been done. So if I see this pattern of people are naturally categorizing people by race, I could very easily tell a just-so story about how uh in the ancestral environment there was lots of intergroup warfare, intergroup competition, and categorizing people by you know of a different race would have been adaptive in the ancestral environment when we were cavemen and we were encountering people from different regions, and we would it would have been it would have benefited us to go to war with those other tribes that we encountered. So and and that I would end the analysis there. No further empirical work. That's my untestable just so story. my work is done. So that's an example of where you could do very bad evolutionary psychology.

That's right. And you could look across cultures and then say, sort of, look, these people, this is the group basis here. And you say, well, yeah, that's because that's, you know, that's activates their tribal psychology, whatever. Yeah, no, exactly. That's right. Yeah. But but you and and Rob and Leda and John did not do bad evolutionary psychology. You did good evolutionary psychology. Better, yes. It's it's still oversimplified. But yeah, yeah, yeah. So I mean, but but it's so the the claim there's sort of um so yeah, let's just so thank you. The claim then is all right, let's look at the claim from uh this functional perspective. So the claim is that evolutionary um over well, the claim is there's a system in the mind for categorizing people by race.

There's two parts to any evolutionary account. One is is the feature of the world present, And then two is would there have been selection for dealing with that feature of the world in a way that would have been fitness relevant? And if you look at race, even on just the first criteria of would this have been a feature of around which our minds could have evolved? Because something has to be a stable feature of the environment for our minds to uh to have some structure to represent it. A good example is like a mother. Mothers are always present. That's a stable, enduring feature of the sexually reproducing organisms. And especially where your mother is investing in you, there's really good ideas that like you always had a mother.

And so that's a stable feature of the world. And that feature of the world, given that you have a mother who has an identity who's individual, that's an enduring feature that can create selection pressures and mental machinery for identifying who your mother is, right? Which is of critical adaptive importance. Exactly. And that's the so it's a feature of the environment. One, two, it would have fitness relevance, right? A robot, a squishy robot that doesn't identify its mother versus one that does. There's clear, straightforward logic to why one design would do well versus another. right? And um so that so that's that. In the case of race, even the first criteria, was it present over evolutionary time?

Already we're in tricky uh tricky ground, uh shaky ground, let's put it that way. Um so race as we talk about it, if you look uh just from an objective biological perspective, um, you know, human beings have uh superficial variation in response to like climactic um exposure. So what we think of as race is sort of very skin-deep superficial differences between populations of people. Um but the degree to which those physical features track ancestry is actually quite weak because um different populations of people who have different recent common ancestry um can look similar because they've been exposed to similar climates or uh it's sort of you know um things like how much melanin people have, like what their um what their eye shapes are, things like that.

A lot of these things are very superficial um adaptations and they're very recent too. Um white skin evolved something like shockingly recently. I think it's even less than 5,000 years ago. There's actually really good. Yeah, it's incredible. That's that's not enough time. No, no, no. Modern humans, you know, we're 60,000 years ago. A 60,000-year-old, well, not year old, but like, you know, if in a time machine, if you if you met a uh a person from 60,000 years ago, they'd be a much better hunter and runner and et cetera, than you. Um, But they would be different in certain ways, but they're still us, right? So, you know, our current brains were around 60,000 years ago. And, you know, human beings don't come into subspecies.

Like even chimps, for example, have five to six times more genetic variation than we do. And even chimps don't have subspecies. And so we're really related, which is consistent with, you know, there's more recent genetic evidence that there were really uh tight bottlenecks in our ancestral populations. So like as a as a from a biological perspective, we are extremely similar to one another, humans as humans are. So um so the idea that sort of race as we talk about it would have been a feature over evolutionary time. Number one, um very racial quote unquote what we count what the physical features that we think of as race are very recent um things. Also, it's only with recent transportation technology that you would people

would have encountered sort of large-scale systematic differences among these things because otherwise everyone's traveling on foot or even by boat. You're meeting people, you're mating with them, you're, you know, like it's just ecologically valid or non-modern technologically mediated social world is one where basically as the scale of social interaction, whatever scale that is, people are sort of blending together in terms of what they look like. Um just by virtue, yeah. Very unlikely for an ancestral hunter-gatherer to be doing intercontinental travel. Exactly, yes. Um and and even if they were, it would be just them, and then they're just part of that community, and their children look like the community, right?

So it's not uh this this large scale, large numbers of groups that are socially isolated remain so distinct for multiple generations, is really a modern phenomenon, right? And and so race as we think about it, you know, it's it's it would have not have been an enduring feature. Um, And even what counts as a race varies historically, right? So, like, you know, in the United States, being Irish was literally a race. It wasn't figuratively a race. You know, you can if you look across places, you know, Rwanda, Hutu, uh, Twitsi, you know, this was a uh we know the history of this was a hallucinated racial distinction that had no objective basis. Right. And I'm sure if you brought people into the lab at that time period and showed them some

Irish people and some they would say, Oh, it was one of the Irish people said that. And they would have you would have seen categorization by Irishness. That's right. The idea that we have inbuilt software for detecting Irishness is preposterous. So something something has gone wrong here. Yeah. If you showed the same thing to ancient Greeks or Romans, the racial categories would have been different. But the same idea, right? Um, and this is sort of a part of what we talked about, instinct blindness, right? This idea that if we experience something vividly, and we do, like there's a pop-out effect, you know. Like one analogy I like to use is something like you can see the clothing color of people or what a car looks like, but there's a pop-out effect if you

see a police officer or uh an EMT or an ambulance or a police car or military uniform. And it's not just a consequence of the color, but it is something about there's this additional inference. Or another example of this would be like sort of gang colors, right? Or team colors, right? There's just like this additional inference. And that I think is the idea. There's sort of pop-out for these for these categories. But anyway, getting back to this idea, race would not have been a feature in which our minds evolved. What counts as a race is really different and volatile across historical time and even within historical time across cultures. And so the idea that the mind has software for this particular

category is sort of seems strange. And then the logic of an evolutionary approach or adaptationism is that you take the behavior, but then you say, look, any outcome is either a product of adaptations, or it's a byproduct of adaptations, or it's noise, right? And it's some combination of those three things underlie any behavior. Now it might be multiple adaptations, it might be multiple combinations of these, but just by virtue of the logic of how natural selection works, it has to be one of these three categories. Yep. And one question then, and this is part of doing a task analysis, which we may or may not get to here, is then you look at a certain behavioral phenomenon or a particular phenomenon

you're interested in. And then you say, well, you know, what kind of an adaptation or sets of adaptations does this look like it might be either a result of or a byproduct of? And so when looking at how people were treating race, what inferences follow, what kind of thing does it look like? This led to an idea, and this is called the alliance hypothesis of race. So the idea is that there would not have been selection pressures on the mind for categorizing people by race. In contrast, there would have been selection pressures on the mind for tracking alliances. So an alliance is uh people joining together to cooperate. That cooperation can also involve antagonism. So then an alliance can be who's on whose side in a conflict.

It can involve coordination, so who's coordinated with who. And there's a lot of evidence, a lot of theory, a lot of modeling, what we know from the hunter-gatherer literature, and just deductive logic is that and even you know comparative evidence of from other primates, chimpanzees also have intergroup differences. Right. Right. So the idea is that cooperation, competition, coordination are really important, deep principles of our evolutionary past. The ability to negotiate these relationships is super important for fitness, meaning the designs that are successful or not. Designs that could negotiate these relationships will do better than those that cannot. A concomitant of doing well or navigating these relationships is

the ability to track who is affiliated with who. This is a complicated problem. I mean, most of my empirical work is on this problem, and we know very little about how the software for this works. But what the idea is that we have this software. This is a clear set of adaptive problems. The ability to identify who is allied with who is important. It's important for predicting when you see two people come together. Can you predict what their relationship will be? If you see another person, can you predict what their relationship will be towards you? And part of the way that the system would work is just like who your mother is, the groups that exist, the alliances that exist are dependent on local circumstances.

So it's a learning system, it's a learning adaptation. There would have been no consistent way in which groups demarcated themselves throughout our evolutionary history. That's right. In the same way it would be a good idea. Yeah. You have to learn exactly how groups are being differentiated, what the what the markers are. Exactly. In the same way that who your mother is always varies over every generation. And therefore, it's a learning system that has an open parameter that says there are a set of cues that are reliably present that tell you who your mother is, but the system doesn't develop to create an identity Ahead of time. It's not innate, yes, quote unquote, in the system. The identity itself is not in the system.

The rules for identifying the identity that exists in the world is in the system. And that's the same principle in the alliance system, which is what the basis of group interaction or treat social interaction will be will vary, but then you can use cues to learn who's affiliated with who. And the idea is that so there's software in the mind for tracking alliances, and part of that design is a learning system that is looking out in the world for reliable patterns of cooperation and competition. You could do a task analysis of what those are. It might be when you see two people physically hurting one another, encode not only their individual identities, but any any category level features that they have, and see if that

category level predicts future conflict. So, for example, like if if the people who live on this side of the river versus who live on this side of the river or who have this tattoo versus have this other tattoo, it's basically this learning thing of there's the software that's looking out in the world for who's cooperating with who, who's fighting with who, and then what are the cues that are correlating in thus predicting those relationships. All of this has very clear selection pressure advantages, to just put it that way. Would have been enduring. We're talking about it at a fairly abstracted level, but it's the it's the level of invariance, right? We're not talking about what the particular cue is.

We're not talking about the particular relationship, but it's the abstract, relevant categories, right? Of that's constant. And the idea of the question is. Right. Yeah. Um, and what you're saying here throws a wrench into that whole dichotomy because what you're talking about is an evolved learning system. That's right. Yes. And there's been yeah, yeah. Yeah, yeah. And Annie Wirtz has another great we will have her on too, uh, where this is exactly her point where there's another context where just like learning who your mother is, when the invariance is variance, right? The invariance is uh you can't build this in. You can't build in what David Pinsoff mother David Pinsoff's mother looks like to every

human being because except for you, David, or your siblings, that will be wrong. Um and so when the invariance is uh is variation, it builds a learning system, right? Yes. This is Annie's point as well. Yeah. So I just wanted to quickly interject that you're talking about an evolved learning system. So continue with with what you're doing. Yeah. Yeah, thank you. It's good to dispel these sort of simple misunderstandings as we go. Yeah. So this is an evolved learning system. And the idea is that this description of the software, even at this basic level, the idea is that that software could be an explanation for why we see categorization by race. And the argument for that was that in a place where certain physical

features become correlated with how people interact and treat one another, and there's a clear history about what that is in our culture and many cultures and in m most of our current milieus. So certain physical features become associated with how people treat one another, interact, and get along. So this software, this alliance tracking software, and that's just the label for it, When it grows up in a society or a social world in which certain physical features are picked up by this alliance tracking system, the idea is that those physical features are predicting these relationships. The evolved learning system is picking up on those physical features. It's not designed to pick up on physical features per se.

It doesn't care if they're physical or not. The idea is that it's learning which physical or which features. I'm sorry, I don't mean physical features, any cues that predict how people interact or get along with one another. It picks up on those and then it stores those as being relevant for predicting other people's behavior, but then also how they're going to interact with me and how I should interact with them. And the idea is that that such that kind of software could explain why people are categorizing people by race, which has nothing to do with there being a dedicated system for categorizing people by race in the in their heads. Yeah. Yes. And I want to add one more thing that people get caught up on.

So a common criticism of evolutionary psychology is that they think everything is an adaptation. They think our belly buttons are adaptations that so that we could, you know, put nuts and small pieces of food there. They think that, you know, our uh the the ridge of our nose is an adaptation to hold up our glasses. They see adaptations everywhere, even where they don't exist. So here we have a very clear um rebuttal to that criticism. We have a an instance where the hypothesis is specifically a byproduct hypothesis. You are not claiming that we have specific adaptations for categorizing by race. By contrast, you are claiming that we have adaptations for detecting alliance membership. And in the modern world, race becomes a cue of alliance membership.

And so we categorize by race as a byproduct of adaptations for categorizing by alliance membership. So this is a byproduct hypothesis, not an adaptationist hypothesis. I just wanted to state that for the record. Yeah, yeah. And I would even say further: the idea is that what race is is a hallucination in the sense of race doesn't exist objectively. It's that it's a social convention. And the physical features that we think of as race get picked up on. And why they become race is because they're those things that have been picked up by this alliance tracking machinery because of the way that these features have been used historically and as a basis of discrimination. Um, and I mean that in the precise sense, I mean, uh, you know, discrimination, it could be

treating other people better, you know, for example, like, you know, um uh white people um discriminating towards on behalf of other white people. Um, but the idea is that like these categories, again, like the Hutu Tutsi thing, can be uh not based on a sort of objective uh features, right? It's just uh coordination uh hallucination. It's could you speak if everyone has shares this uh is coordinated around this, um, that's that's where you get the category race appearing. It's it's after it has become socially predictive and people are acting as if it's socially predictive. That's that's it's it's it's a social construct in the in the precise meaning of that word. Um So, this explanation of being a byproduct, and thank you for emphasizing that.

Um, what's crucial about it is this makes predictions. And it makes predictions because the claim is there's a well-designed system for tracking alliances. And part of that includes then that it should be sensitive to what uh who's cooperating with who, who's competing with who. And the job of this system is to track these relationships, keep track of the cues that predict them, and then use these cues to generate inferences about who will cooperate with who, who will compete with who, et cetera. So the paradigm that had shown that people always categorize by race this who said what paradigm, what Rob and Lita and John and myself have done, and we've really tested the heck out of this, and I'm quite confident in it.

Um, and I was a bit skeptical at first, um, you know, but I convinced myself, and that was a good um, you know, that was a good, you know, 10 or 15 years of running lots of different studies trying to falsify it, um, was that this system, what it should be designed to do is to track the validity of a cue. So if race is being treated as an alliance cue, in the old studies, there was nothing in the stimuli that told people that there was some other alliance cue present. So even though the conversation wasn't about race or relevant to race, the idea is that alliances, especially alliances that have to do with like people treating each other, you know, at the level of like, you know, conflict, strong levels of

discrimination that are valid across situations, people's lives, um, there's no better cue of alliance in the stimuli except for race. And so the reason why people were categorizing people by race is because this software system, this alliance tracking system, was looking at the people, detected this relevant cue and it wasn't detecting any other countervailing information. And the idea is that the alliance tracking system, if it's well designed, should be able to pick up on which cues are relevant for the current situation. And so the idea is that a well-designed alliance tracking system, it has to, as part of its job, juggle multiple alliances and be able to inhibit when a current alliance,

uh a particular alliance is not relevant and activate one when it is and learn about new ones. These are just the con the implications of a well-designed system. Yes. It would be disastrous if you were to use an outmoded alliance category and mispredict who is going to come to your aid or fight with you. This would be dis a disastrous outcome uh from a Darwinian perspective. We would expect uh design for getting the answers to these questions right of what is the most relevant alliance uh in in the current social environment. Aaron Powell That's right. Yes, exactly. It needs to be a flexible, nimble system that's really sensitive to like what is currently organizing the conflict. And if you get it wrong, uh you're in trouble.

Um And this prediction then is that so you so that's the idea of how the software works. So then what you do is then you think, okay, the software is an input-output system. Current humans have this software. How can we test to see if that's the design underlying racial categorization? Yes. And then you just you take, then say, well, what are the class of inputs that this thing should be sensitive to? And so we've conducted studies where we show people in this who said what paradigm just social groups that are a basis of cooperation. So we tell people, like, look, this is a group of people, they're all working together, traveling, they they really trust and count on one another. And you show them having a conversation where one group is talking about how they all

work together, spend a lot of time together, trust one another, have had difficult, dangerous experiences together. And then there's another group who also is similarly like that. And so you're basically showing people the system cues of this set of people are all highly cooperative. This other set of people are all highly cooperative. And it's, you know, with like text, but you know, the presumably there's like any number of cues that you could use. Yeah. Um, but this was a first-pass test of like if there is an alliance tracking system, these cues of a history of deep engaged cooperation for each group with one group, group A and group B, right? this should satisfy the input conditions of the system.

And then the idea is that if this alliance tracking system is good at not only picking up on new group memberships, it should also be able to deactivate or inhibit the use of cues that are no longer relevant, then the simple way to show that is we show people that race is no longer predictive of who's in group A or who's in group B. So concretely, there's an equal number of black and white people in group A and an equal number of black and white people in group B. And so as the people are shown having this conversation, so basically there's somebody in group A and they're like, Hey, nice to meet you. And the groups are having a friendly conversation, but the people are getting a window into, oh, there's one group here.

they're all uh I think it was something like Habitat for Humanity and like partners in health, like they were charity groups. Um each charity group, those groups really work together. And the people who are in that group are all in alliance, basically, is the idea. Yeah. Yeah. And then that idea then is that there's two important cues in that stimuli, in those stimuli. One is that race is no longer predictive of who's on who who's allied with whom. And then there's a new basis of alliance, which is this team membership. And basically, when you do that and you show people the same stimuli that in the past studies had always found categorization by race. In fact, you get categorization by the new team membership, by charity group

membership, and you get people decreasing, strongly decreasing their categorization by race. And we vary how strong the cues are to charity group membership. And in the conditions where it's very strong, people are often not categorizing by race at all. Say that again. Not categorized by race. Not category by race at all. So you have accomplished the holy grail. You have done what social psychologists have been trying to do, have been hoping they could do, which is to eliminate categorization by race. And how you got there was by doing good evolutionary psychology. And it's hard to see how you could have gotten there without doing that kind of evolutionary psychology. Right? Yeah, I mean I think that's fair.

I mean it's it's a uh yeah I'd say that's fair. Aaron Powell So I I I consider that a big win for evolutionary psychology. It is an empirical discovery that we did not know existed. Doing this kind of evolutionary thinking resulted in this new discovery. And it's an extremely important discovery. We basically learned that this thing that we thought was an ineradicable part of the human condition, which is racial categorization, uh, we we thought it was uh uh impossible to get rid of. And then here come the evolutionary psychologists saying, wait a second, the idea that we would have uh evolved software for categorized by race makes no sense. If you know anything about human evolutionary history, what makes way more sense is that we

would have alliance detection systems. If you uh give cues of uh alliances that are more relevant than race, we would expect the mind to categorize by alliance and not by race. And lo and behold, we run the test. Humans categorize by alliance and not by race. We have eliminated the tendency to categorize by race. The first paper that ran this test, I believe, was with sports teams. They found that that racial categorization was eliminated. They called the paper, Can Race Be Erased? And of course, the answer that the paper gave was yes, it can. So I I find this to be uh a really inspiring success story for the field. Uh, do you would you agree with that assessment? Yeah, and I and to be I I sort of because it's my own research too, I sort of feel like uh

a bit, maybe it's false modesty, but I'm kind of like, what you know, there's still a lot we don't know. Yes. Uh what one one thing that's really important to say because if people are thinking about it, they might have this concern. And I had many concerns, is um we know because of the research, we know a lot about how this works. So there's a couple things that are important to say. One is if you do the exact same manipulation, but instead of with race, you do it with sex or with age, which were the big three categories: age, sex, and race. So, in other words, you instead of crossing race with team, you cross sex with team, meaning there's an equal number of men and women on each team, or there's an equal number

of old people and young people on each team. Yeah. Age and sex are not affected by the same manipulation. And that's really important because you might think well, the reason why people are reducing their categorization by race is because the mind is just shifting attention from one category to another. And so people can only keep track of one category. But it turns out people have no trouble keeping track of alliance and sex or alliance and age. So it really seems to be that alliance is rival with race specifically. And the idea for why we believe that is, and this is consistent with this idea, the data are consistent with this hypothesis, is because, in fact, race is an instance of alliance.

And the idea is that these this pattern of data is the prediction and consistent with that idea that really, as far as the mind treats it, race is an instance of an alliance cue. There's another thing that's important too, which is that, and this is where the evolutionary psychology is smarter than I am, like uh, is that I also wanted to know well, how is racial categorization being reduced? In memory literature, and it's a memory task, the who said what task is a memory task. There's two different ways that information can be lost. So if you're like studying for an exam and you forget it, or or if you know you don't know the answer. One one way is that the information was never encoded in the first place.

So, for example, you're just like you couldn't, you weren't paying enough attention. You just stopped paying attention to race, for example, in this case. The other possibility is that the racial information is encoded, but then it's not retrieved during the task. And I actually didn't know which two uh things it would be. And there's a way to experimentally test between encoding versus recall. So the question is: is the reduction happening at encoding or is it happening at a recall? But there's a way to experimentally test between the two. And I ran that study. And what I found was that categorization by race was being encoded, but it was being inhibited in the system. And like there's a paper, you can people can look it up if they want you to know how you

know that. But basically, those details are sort of kind of in the weeds. But the point is that that's actually a really smart way for the system to work. And that was actually something that I hadn't really thought about. But the idea is that if there's an alliance queue that has a high prior probability of being relevant across situations in your life, and then all of a sudden you're faced with a particular context in which it's no longer relevant, the system has two possibilities. One is to encode the category, but not use it for this particular context. And another is to just ignore the cue altogether. And if you think about what a really well-designed system should do that's has the job of

tracking alliances, if there's a queue that's really high prior probability of being relevant, but it's not relevant right now, then a really good system shouldn't code but inhibit it. We talk about it. The expression was promiscuously encode but selectively retrieve these categories. And so that was a nice example of sort of an empirically generated insight into the design of a system that in retrospect sort of uh is the way a well-designed system would have um, you know, uh would appear. But that's one where even I wasn't, I wasn't even being adaptationist enough about it. You know, I was sort of, you know, and so this is a good example of the system is really well designed for this function, to the point of even me, a hyper adaptationist,

wasn't giving it enough credit. And it was like, no, no, no, I'm actually really smart here. I'm gonna show you this additional design feature. That's really cool. I mean, so I mean, it it it makes sense, you know, on reflection. So, like, let's say we have you know the habitat for humanity, and I don't know what the other charity was. Let's say it's the Make a Wish Foundation. And uh so we have this strong cue that these two groups are in uh a cooperative relationship uh with uh other members of of each group. Um and so that's gonna be more relevant for this situation than race. Um it would be very bad design if if if we just totally forgot about race because it's such a relevant and enduring feature of modern social environments, it would be a bad idea to

completely uh dispense with it altogether, because you never know. Maybe there could be a racial conflict between one of these organizations. Maybe one uh you know African-American accuses the organization of uh discriminating against him or her, or or or uh unfairly treating the black uh staff relative to the white staff. These issues could easily crop up as they do in workplaces around the country. It would be very foolish to forget about race altogether. What you want to do is keep the information online, keep it available, but just suppress it because there's a more relevant cue right now in this particular situation. That's right. Yeah. And so there's really two conclusions, and then we can maybe loop back to like our broader

point about why it's not just those stories. The two conclusions are: this is evidence that even modern skulls are showing the signature of a system that's well designed for an alliance tracking system. And the idea that that's a long enduring ancestral adaptive problem was the reason why those particular stimuli were created because the idea was if that system exists, it should be sensitive to these inputs and should generate this output. In this case, we we've so far just talked about categorization. There's other outputs we could talk about, but anyway. Yeah. Yeah. No, that's all fascinating. And I I I will also add that this that this empirical work is really very much the tip of the iceberg.

Yeah. And in in in in terms of the additional empirical work that could be done. So this is just uh scratching the surface. I mean, yeah, it's like proof of concept. So like it's it's incredibly mysterious. You know, there are probably many do dozens of cues that are are used to track alliances. Um, you just you know gave people a text-based description of you know which group was cooperating with whom, but uh you could do further work on this. You know, maybe dialect is important, uh, maybe dress is is important. Um there are there are lots of really interesting predictions you could make about how the alliance detection system is working and reverse engineering it um based on uh details of the local context versus um other situations you might have encountered in the past, which

cues are going to be uh given higher priority when. Um there's a lot of work to be done here of really nailing down the specifics of how this alliance detection system works. Yeah, and and and the inputs and that's the inputs and then there's other work in psychology, some of which sort of inadvertently is getting at it too, like spatial proximity and like who lives with who. And I actually do have some work on accent too. We could talk about that another time or not. But uh but no, that's right. And that's all on the input end, but then there's also the output end, right? Which is that the the categorization is a byproduct. It's not what the system evolved to do. And so there's this whole other thing of um more ecologically

valid outputs, which is inferring who's going to behave, you know, how they're who's going to support who in a conflict. And my students and I are working on that as well. But that's right. The point is that all of a sudden you're moving from stuck with a behavioral phenomenon, categorization by race. We're standing on our heads trying to get it to change. Yep. And then coming at this from a different perspective, which is to not just be locked into what can I just intuitively think of that might be relevant for categorization by race? But to then come from this perspective of weirdly, we're not allowed to think there's a system for categorization by race. It's got to be a byproduct of another system because of an evolutionary history where race

was not present. And because we're only allowed to stipulate that there would be software in the mind around adaptive problems. And then immediately the hypothesis based on an adaptive problem generated a prediction and you get this result. Now it might be oversimplified. There might be a million adaptations in the social world, and this is just the first pass at it, and it is still oversimplified. But it's a really nice example of the just so story of look, um, it's racial categorization is inevitable, right? Because of our intrinsic tribalism. Or even the just so story of it's not inevitable, it's a byproduct of our alliance tracking. The thing is, is that the sense in which these are not just so stories

is that they're making additional design predictions, right? Yes. Um even this, even the eyeglasses, and so there's this fanciful thing, Dr. Pangloss, the Panglossing thing where, like, you know, the reason why we have a nose is to hold up eyeglasses. People one of the ways people care criticize adaptationism is to say, is to use that example, right? And to say, well, you're just obviously that's um stupid. But the thing that I think is funny about that is they're being an adaptationist to make that point, which is that it's very obvious that noses are not well designed to hold up eyeglasses. If I was gonna build something that was well designed to hold up eyeglasses, I wouldn't make it sloped and greasy, um, such that you had to like have things around your ears,

right? Like something well designed to hold up eyeglasses on your nose. Like a nose is not designed to hold up eyeglasses well at all. It's it's really eyeglasses are well designed to stay on noses, which are not designed well to hold anything. They're super slippery slopes, right? Um, and so the point is when you make a claim about something is for something, it makes predictions about how it should be structured. It makes new predictions in the future. And what adaptationism is and why it's not a just so story is that the logic of what follows from the claim about function. For example, the function is not race per se, but it's about alliances that predicts that the system should start to care about alliances that are not racial and

should stop caring about race. So it's an example of claims about function make predictions, and that's why it's not a just so story. Yes. Form follows function, is my mantra. Um, and if you are trying to develop a theory of the form of the human mind, the best way to do that is to come up with a good theory of what the function of the human mind is or what the function of the part of the mind you're looking at is. Once you have a good theory of what its function is, and that function is necessarily going to be an evolutionary theory because evolution is the only source of antientropic functional organization we know of, once you make a claim about function, all of a sudden all the form falls into place.

You have a beautiful framework for predicting what form that system should take. Right. Yeah. And I think in, you know, the racial categorization work, it's a nice example. But then there's all this work on, for example, uh, work on anger, which is a really important thing to understand. And this approach has also an adaptationist approach has revolutionized our understanding of anger, what elicits anger, what reduces anger, which is extremely important because anger is the basis of a lot of domestic violence. It's the basis of a lot of group based conflict. Um things like um shame or pride, same same analysis. Uh, David, your work on humor, another example, right? Where um, you know, these these things that we phenomena we all know.

Um, you know, obviously we experience race, we experience shame or pride or things being funny, but to then unpack the logic of it such that we can predict what is funny or when people will feel shame or pride um or anger. Um that's that I think is a really good example of the power of this approach. And it's we're just it's sort of just getting started. Absolutely. Yeah. Uh it would be very difficult to try to understand how anger works or how pride or shame works or how humor works without having any idea of what these things are for, of what their function is. Um and once you have a good idea of what their function is, all of a sudden you begin you you're drowning in predictions. That's some one of the things that uh really shocks me when people uh level this accusation

against evolutionary psychology that you know it doesn't make predictions or it's unfalsifiable or whatever. And in my experience as an evolutionary psychologist, uh doing work in this field, my problem uh is not having a lack of hypotheses to test. My problem is being overwhelmed with with predictions and hypotheses to test and just being filled with deep sadness that there aren't enough hours in the day to go out and test all these hypotheses. That is really more is closer to my experience. That it's really uh I'm so d I'm drowning in this stuff and I and I I need an army of people to go out and do this research and that army isn't here. Right. Yes. No, I mean, so I started teaching a grad class, even though UCSB is a really important

hub and one of the origin places of evolutionary psychology. There had never been a graduate seminar in it. And one of the things that I had done, I did in the graduate seminar, was have the graduate students do a task analysis. And we did task analyses of jealousy, um, other things. Um and and basically would fill up a whiteboard and maybe we can talk a little bit about what doing a task analysis means um maybe or we you know we can talk about that but basically you know it's taking an adaptive problem like you know not starving to death for example you know um or taking something like jealousy and then sort of saying well what everybody sort of examine your intuitions or what you know about jealousy.

Like what kind of a um, you know, does it quack like a duck, does it sound like a duck, does it swim like a duck? That kind of thing. Like what is it? What uh what is jealousy compared to everything else? And then at a certain point, it's characterizing that phenomenon like jealousy or that adaptive problem like not starving to death. And then just saying, well, what are all the design features that would be necessary, almost sort of like in a robot, right? To actually achieve, to successfully like navigate a world in which this is the desirable outcome to navigate this thing, to not starve. And then immediately, like if you just and we just fill up a whiteboard where you just sort of say, well, like, okay, an organism needs to not starve.

Well, then it has to figure out, well, what's food? Well, that's a problem. Well, how does it know what's food? You know, and maybe like it's koala bear, there's a template, but then maybe not. And this is like part of Annie Wirtz's point, like for for humans who are extractive foragers, who occupy a lot of different food niches. We're omnivores, we don't have like any reliable cue, and so it's a social learning problem. So maybe there's social cues of like the adults around me are so are alive, they haven't been poisoned to death. They're so somehow like they've learned what's food. And so the learning system, like the alliance tracking system, is a learning system for tracking relationships of like, you know, literally what do people put in their mouth?

Okay, what is that? Intent to that, remember that, extrapolate, like how where did that come from? Um and you can just go through the list of like if you were going to build a robot that could grow up and say, what Very hard. Even even just the simple heuristic of eat what other people around you are eating. If I'm a kid and I see my dad drinking coffee, which my daughter often does, she never wants to t have a sip of my coffee. She's not interested in it. That's right. And the extr and the in even the inference rule of like it was it that particular cup? Is it anything that has that color, right? This is the the the the the um Hume's problem of induction, right? Or it's the uh the the Gavagai problem um in uh in philosophy, which

is that the referent or the the or the rule could be almost anything, right? And so so even just not starving to death and then learning how to extract things or forage and like you know, exerting them, you have to exert a certain number of calories to like get food. So there has to be a calculation for that. There has to be a representation of when do I get this thing? What's the processing to get it? So there's this you know enormous problems. problem. How valuable is this food? How much work should I be willing to expend to exactly. And so, you know, modern environments where you have a whole industry of people sort of pulling the levers of this psychology, like it's a very complicated psychology, and

that's why people can pull the levers on like highly processed food. And it just you know turns off uh stop eating signals. But like, but the idea is there's a really complicated machinery for For not starving to death. And if you map out that problem, or if the pro any other problem of like not losing status, right, or competing in a status game, any any number of things, or coordinating about like making a mistake. Um, the second that you you think about what are all the entailments of this, you start to populate the whiteboard figuratively and literally with a bunch of additional functions and subfunctions. And it really is the kind of thing where any one empirical project just gets at one or two

of these elements, but it would organize an entire career or a whole research discipline of empirical work. And I'm always struck with how little we know about any one of these functions and how little empirical work there is on it. So I completely agree with you that the limiting problem is not predictions. It's not knowing what to look for. having a community of people to actually start to test these ideas. And I would say to the extent they've been tested, for example, the shame, the guilt work, the anger work, the racial categorization alliance work, uh, Annie's work on uh learning about food and plants uh and children. Um, I mean, you just uh name name the area. It's just been extremely um fruitful.

Um And I think this is all uh uh examples of why it's not a just so story and what it is positively, uh aside from sort of addressing the negative ideas about what evolutionary psychology is. Yeah, absolutely. Um should we maybe take a look at one of those things you mentioned? Either anger or shame or pride or humor, just to take a take a dive into one of those? Yeah, what do you want to do? All right, why don't we do anger? Yeah, that's great. Uh I I I love this work. This is uh uh Aaron Sell uh and Leda Cosmodies and John Tubi have done work on this. Um Aaron actually gave a talk at UCLA at our lab. And he had a talk on the 12 triggers of anger. Yes, that's 12. So he had a bunch of participants just tell him stories, just submit stories of a time that

they were mad at somebody about something. And he looked through all those stories and he was able to glean 12 distinct triggers that commonly reemerged. And he was able to dissect each anecdote um in in a way that was so brilliant and inspiring. Like I I am not aware of any social science theory that could look at a specific anecdote of of human behavior and be able to explain each precise detail the way that he was able to. Yeah. And to just and to just take a step back, the idea is not that, uh just to clarify for our listeners, because I know this and you I know you know this, but uh it's not that Aaron was sort of saying, well, I don't know how anger works. I'm gonna just get surveys and then do a a um latent thing of like what clusters

in the surveys. It was the complete opposite, where the idea was there are current theories of anger, some of which are like things like he went through the list. I mean, his dissertation is wonderful. There's a lot of empirical work on this. A good example, though, just to to just sort of give one example was frustration aggression. So the idea is that when your goals are blocked, you get angry. And that seems like an intuitive, reasonable thing enough. But Erin, Erin and I overlapped in the CP, and it was talking about this example. And I give this in classes when I talk about this. That predicts that. So, for example, imagine you're driving and um somebody cuts you off in traffic. Um, and uh, you know, imagine how angry you would be.

Or, like, let's say, even worse, a four-way stop. You stop, and then somebody like doesn't stop and goes through the four-way stop sign. Like, clearly their fault, right? Just think about how angry you would be. Now imagine that that car was just like, you know, uh Mercedes-Benz Silver convertible versus was an ambulance. Right. Yeah. There's a real difference in how much anger you feel between those two kinds of cars. Yeah. And they're both they're both thwarting your goals. And that's exactly Aaron's point, is that they're both thwarting your goals. And if the idea, the function of anger is that you get angry when your goals are thwarted, then there's nothing in that theory that explains why the ambulance is different from the

um the Mercedes, right? And Aaron's idea, I mean, do you want to do you want to give Aaron's I mean Aaron's idea? Yeah. Aaron's idea and David you you'll flesh it out even better is just that um what anger is, is a set of adaptations for negotiating treatment and how people treat one another. So how much people trade off their welfare with each other is a really fundamental important thing. Um, meaning uh when we make decisions, we're deciding is the thing I'm doing benefiting me? Is it going to benefit somebody else? And when there's rivalry between these two things, To what degree do I trade off my welfare for their? So, for example, my son, I am very willing to trade off my welfare for his, right?

Um, so that's very different though than like a stranger, right? So I would not be willing to do that. And we have this trade off between your welfare and somebody else's. And Aaron's idea, and Lita and John and Daniel Schnitzer and other people have been doing work on this. The idea is that what anger is, is it's a program, software in the mind for negotiating and bargaining for better treatment. So it detects situations in which you are not being treated with the correct welfare trade-off. And then it deploys a number of different strategies. And those strategies are often relationship-dependent. So, for example, if you are in a close relationship with somebody and they do something that suggests that they're not taking your welfare into account well,

then the first step would be to like be motivated to find out why they did that thing. That seems like they weren't treating you well because there can just be like, you know, not another. So so one example would be something like, I left the door open and your cat got out. Maybe I didn't know you had a cat, right? And so before you decide that I'm I'm a jerk, you should just be like, hey man, did you know I have a cat and my cat got out? Like it's a low benefit to you, a high cost to me, you know? Um and and my reaction would be extremely different depending on what you said. If you said, Oh, I'm so sorry, I didn't know you had a cat, my bad. I'll next time I'll close the door, my anger would evaporate.

Right. But if you said, uh, yeah, I know you had a cat, whatever, I don't care. Yeah, exactly. Yeah, yeah. Forget about your cat. So in in the idea is that this is again gets at this level of invariance thing, right? Where it's there's benefits to me and cost to you, benefit of me of opening the door, or just the cost to you of losing your cat. And you want to see, look, is Dave not taking my interest into account? Do I need to change either how I treat him because he's a jerk, more of a jerk than I thought? Or do I need to let him know that this particular situation was a bigger cost than I'm willing to put up with? So I want to make sure he doesn't do it in the future. Right. And so that's that's sort of the underlying logic of anger.

And if we have a good relationship, it can be checking on that you know. Well, sorry, if we have a good relationship, we're checking first, did you even know that you were imposing a cost on me? And then uh if we if I did, and I'm not sure that I want to change my behavior, then David, your recourse can be withdrawing cooperation. So, so anybody who's in a romantic relationship will probably recognize this, where if you know your partner is mad, what their partner looking like mad is not what a stranger looks like mad. What a partner looking like mad is just somebody with a stone face, right? Yeah. Like, and that's just a cue of withdrawing cooperation. Yes. Now, for somebody that doesn't have a relationship with you, then uh

the idea is that it's a threat to impose costs. So looking angry, you know, and either social costs or physical costs, or at least the threat of that. Um, and the idea is that those costs we either of withdrawing benefits that are currently existing or imposing new costs actively, um, is a way to negotiate for better treatment. And the idea is that this is the logic of that software. So that's all the background to now David talking about the vignettes that Aaron presented. Um how Aaron was able to extract in a very detailed level, um yeah. Yes. Um so first I I want to see if I can maybe do a quick summary of what you just said, just in case it's still unclear for people in their heads. So um when something uh makes me angry, um

there are four important things that are within that angering scenario. The first is the benefit So let's say, let's take the example of you leaving my door open and my cat running out. Um so I'm representing how much that benefited you, Dave, to uh save a bit of time, you know, closing the door. Right. How much does it benefit you to just be able to walk out without closing the door? Okay. That's the first part that the mind is representing. The second part is the cost that was inflicted on me as a result of that action. So how costly was it for me for my cat to run out of the door and for me to uh chase my cat around? Right. The third part is the ratio between those two numbers. Um so when anger happens, um, the benefit to Dave is

very small and the cost to me is very large. I'll say that again. When anger happens, the benefit to the person making you angry is small, and the cost to you is large. So if you inflict a huge cost on me in order to get a very small benefit for yourself, that is going to piss me off. Okay. And once so I'm my mind is basically trying to figure out what that ratio is. What's the benefit to you? What's the cost to me? If it is too low, then anger gets activated. And what anger is, is a way for me to uh without neurosurgery, changing that ratio in your head. Correct. So that um you are no longer willing to impose those costs on me at that exchange rate to yourself, so that you treat me better in the future.

So the idea is that we all have these ratios in our head for every person in our life. We have a level of cost that we're willing to inflict at them for a level of benefit we're willing to receive. Okay. And the idea is we have to have these ratios, otherwise, we wouldn't be able to make decisions. So everyone we have, everyone in our lives, we have a welfare trade-off ratio. Um and what anger is, is a way of uh trying to recalibrate someone else's welfare trade off ratio toward you. So let's say yours toward me is say one one over four, and I think that's too low. What anger is doing is I'm trying to get that up to a one over three or a one over two or or whatever I think is the correct or uh justifiable ratio or the the ratio

that I deserve based on a variety of factors like my social status or or how kind I am to you or or or history as friends or whatever. So that's what anger is. Um it's a way of recalibrating uh other people's uh welfare trade-off ratios toward you um without doing neurosurgery. Um and and the way you do that is by threatening to inflict costs. So if you don't change, if you don't treat me better, if you don't change your welfare trade-off ratio, I'm gonna punch you um, or I'm gonna uh or I'm gonna withhold benefits, I'm gonna stop talking to you. I'm going to stop uh sharing my food with you, whatever. Uh it's these are these are attempts to inflict costs or withhold benefits, uh uh conditional on the person um recalibrating their welfare trade-off ratio and expressing

that they've recalibrated it through an apology or a show of guilt, the function of which is to change your um welfare trade-off ratio towards someone who you've wronged so that you treat them better in the future. Okay. Um there's one beautiful example that uh Aaron gives, which is that imagine uh Dave and I are uh eating uh French fries together. Um and uh Dave has a bit of ketchup on his lip. And we're both we're we're we're wearing beautiful uh nice suits. And uh Dave reaches across the table and takes my beautiful nice tie and wipes the ketchup off his face with my tie. True story. This actually happened? No, I'm kidding. No, no, no, no. I I'm this might have happened to Aaron. I don't know.

What's so beautiful about that is that it's it's it's the perfect angering scenario. The benefit to you of wiping the ketchup off your face is so small. You could have easily just gotten a napkin. Yeah. And and the and the cost to me, if you're ruining my tie, is huge. The ratio is so out of whack there. Yeah. I would be so pissed off if you did that. Yeah. It's it's a great, it's I love about it that examples really just, you know, like saying the abstract logic sounds really stilted and weird and foreign. But then the second you give examples, uh, like like the door thing, right? Like I left the cat, I let your cat go out. Well, what if like, you know, I had like severed my like artery was running out the door.

You wouldn't you wouldn't think twice. You wouldn't be angry at all about me leaving the door open. Because it was it was a super uh I mean I was there was there it was super high benefit to me um to to just go out the door, right? Um and so you wouldn't care at all about the fact that your cat, I mean you might be, but you're gonna be angry at me, you know. But if I was just like no good reason, then it would be very different. Right. So in that scenario, you would have three different strategies for diffusing my anger. The first one would be to highlight how the benefit to you was actually much larger than I thought. So I didn't know that you were you know gushing blood and you were rushing out the door.

Had I known that, then the benefit of you uh of getting to the hospital quicker and not having to close the door, that was actually a pretty big benefit. And it's probably worth it for you to not close the door. So in that case, your welfare trade off ratio toward me was fine. It's fine that you didn't close the door. It was a really big benefit to you. That's right. I'm not the kind of that's right. I'm not the kind of person who wouldn't ordinarily like take your cats into account, but man, like I was dying. Yeah. Exactly. Exactly. Exactly. So that's one w strategy you could use. The other strategy is to say that the cost inflicted on me actually wasn't that great or what wasn't that huge. And you knew that.

So like yeah your cat always runs uh out the door and always comes back. It's never it's never a big deal. Like, who cares? Why are you getting so bent up about this? Right. That's another strategy that it wasn't that big of a cost to me, so I should chill out about it. Uh, and the third strategy you could use is ignorance that you did just didn't know. You didn't know how big of a cost. You didn't know I had a cat, or you didn't know that the cat has you know run away in the past and has caused me a lot of stress and and pain, or or whatever. Um, and in that case, your behavior is not reflecting your true welfare trade-off ratio toward me. Um, had you known, you would have acted differently.

So this is the your apparent behavior is misleading. It was a product of of you not knowing and you being uninformed. So those three strategies, if successful, could diffuse my anger. Um, okay. Um, other examples. So uh Aaron, when he came in, he gave tons of vignettes of real vignettes, real things that happen to people. And he was able to identify the benefit to the person, the cost inflicted on the aggrieved party, whether it was too high or too low, whether it was intentional or not, whether the person knew the ext magnitude of the cost or the or the benefit that they would be receiving. And he was able to just dissect each individual scenario in a way that was so beautiful and inspiring.

And it it was also beautiful in the way that it was so specific that it's not merely behavior that anger is directed at. So anger isn't about your actions, it's about a number in your head. It's about the ratio in your head, and it is exquisitely fine-tuned to that ratio and not to your overt behavior. So a beautiful example that he gave, uh, this is a real example from someone's life. Um, so um uh a woman and her boyfriend, and um they have had previous interactions where the boyfriend has insulted her driving ability. So she the boyfriend is constantly giving her shit for being a shitty driver, right? Or he perceives her as a shitty driver. He perceives her as a shitty driver. Um, and because our welfare trade off ratio depends on how much we value the person, you

know, if the if the boyfriend thinks she's a shitty driver, he's he's gonna have a lower welfare trade-off ratio toward her and not treat her as as as well as he otherwise would. So that that in and of itself could be a could be a problem that would make her angry. Um so in this scenario, um, she uh misses the exit and cuts someone off in in traffic and you know almost gets into an accident. And um the boyfriend doesn't even say anything, doesn't even do anything. Yeah. And she says, don't. And she got mad at the boyfriend at that moment. Sure. Because she because she knew what he was thinking. Yeah. Yeah. Yeah. So he didn't do anything at all. There the anger was not caused by any overt behavior.

It was caused by a belief about what was going on inside his head. That's right. Which is such a beautiful example of the specificity of this emotion. Yeah. That it's designed to pick up on something inside the other person's head about how much they value you. Um and if it's too low and they're not valuing you high enough, you're gonna get angry. Yeah. Um so that was my favorite example. Uh there are other others that I can't remember. Yeah, but but but it sounds like what you're saying is that you know, a way a way to put it is there's a grammar. So there's Erin, this idea of taking the adaptive problem of how people treat one another, understanding that it's a long-standing problem, understanding that there is a level of

specificity where there are commonalities, there are invariances like cost, benefit, the person not holding a high enough regard for your welfare, check to see if they understood it. Are they ignorant or not? Do I negotiate? Do I need to impose a cost, withdraw benefits, impose active costs? These are all the things that are common across individuals, across time, across generations that are built into the software. And by thinking about all of those invariances, Aaron was able to sort of create a grammar, a social grammar, if you will, that then was common in all of these vignettes. And you could look at the vignette and go in and really lift out in each case the reason why somebody is saying this thing is because they're trying to communicate that

the cost wasn't so great, as in like the cat really didn't go very far, or I had a huge benefit because you know I was bleeding and I needed to actually like get out the door, I would die. Um and so again, it comes back to this thing of if you have a theory of function, It reduces down all of the complexities and all of the noise and blooming, buzzing confusion, sort of uh of the data. Yes. And then you can really compactly and economically describe when somebody's angry, what they will do, in the same way you can compactly describe sort of how a plant will grow towards light. Is that sort of the point? Yes. Yes. Well put. Yeah. So I think the so the anger thing is a really great example. So I think both the anger thing and the racial categorization thing are really nice

examples of taking uh a phenomenon we already know about. Mm-hmm. Anger, racial categorization. I mean, people have been studying anger for for decades. There were different theories of anger. People have been studying racial categorization for decades. There are many theories about why that was happening. Um, and you know the you even really smart people, you know, um it's not enough to just be smart. You have to think about what the thing is for. It's to take an artifact analogy, like if you don't know what a car is for, you can be really smart, but you can come at it the wrong way. And I'm not it's not saying that anyone approaches right, and and there are there can be wrong evolutionarily informed hypotheses, but that's fine because you're at least narrowing

down the possibilities. But in these cases, this welfare trade-off and this alliance tracking idea really cut through all the confusion surrounding these phenomena. And these are just two examples sort of off the top of our head. Yes. The m one of the main benefits I see to doing this kind of task analysis, reverse engineering, um adaptationism, whatever you want to call it, is that it really constrains the hypothesis space once you start thinking in this way. So if you're gonna do a good task analysis of an adaptive problem uh that our hunter-gatherer ancestors uh would have reliably faced throughout our evolutionary history, uh, and you're and you're going to posit a cognitive system that evolved in response to

this selection pressure, and you're going to test for the existence of that cognitive system. Already you have wiped away 99.99% of all the possible hypotheses you could have about the thing you were studying. So, like right off the bat, um, I'm I'm positing an adaptive problem our ancestors would have likely faced. That screens out tons of hypotheses because there were lots of things, like uh racial groups, that our ancestors would not have plausibly encountered. And so, right off the bat, that suggests that any adaptation for racial categorization is implausible. We should not explore that hypothesis. Right? Right. So right off the bat you have constraints about what the ancestral environment would have

plausibly been like, what kinds of challenges our ancestors would have likely faced. Um another Yeah, can I just inter interrupt really quick? One thing I think that's really important to say, because I can imagine people listening thinking this, and it's an important thing to say, is people often say, Well, we don't know, we can't know for certain. But that that is exactly like the um uh criticism of uh just so stories because the nose is poorly designed. It uses the principle to reject the principle, which is of course we can't know certain things for certain. So that's why you are always thinking for the thing that you're most certain about. We can be certain that alliances were present over evolutionary time.

The idea that there was a particular kind of alliance is less certain. It might still be more certain than like even something less certain. But the idea is that we're constantly criticizing our thinking and trying to describe things in a way that would be the most certain. Yes. That you can't get at if you just ignore the issue altogether. So when people sort of say, well, you can't know anything about the evolutionary past, often what they were saying then is that we shouldn't even think about how likely it is that there were certain selection pressures as opposed to others. But that's the exact opposite of their own point, which is that we should actually just try to stay in the range of things like where can we pre certain people impose costs on each

other or confer benefits, you know, and needed to negotiate that. There's the anger thing. That's very likely. So I just want to point out that yes, weirdly, people will often say things like, Well, we can't know anything with certainty, and then therefore, any attempt to be more or less certain is is hopeless. That second thing doesn't follow at all. So I just I don't want to interrupt too much, but I want to say that important, yeah. Crucial point, yeah. We want to build our evolutionary theories off of things that are most certain. To the extent that we're building our theories on a on a shaky foundation of uncertain premises about our evolutionary past, well, that should cause us to doubt the hypothesis,

the more uncertain that that premise is. That's right. Insofar as we're building our theories on foundations or on assumptions that we're not sure about, that should weaken our confidence in the theory? So the anger stuff is based on some pretty plausible assumptions that there were people in the ancestral environment and they could have imposed costs on us uh in order to benefit themselves, and they likely did in various ways. That is a pretty that is that's pretty hard to deny that that was a feature of ancestral environments, right? We know that our ancestors were social, we know that other primates are social, we know that humans are intensely social, and simply by existing in the world, humans have

opportunities to inflict costs on one another, and there would have been selection pressure for minimizing those costs. Okay, so that theory is resting on a pretty solid foundation on uh what uh ancestral selection pressures would have been like. Um similarly with the racial categorization work. We know from evolutionary genetics works work that skin color is a recent adaptation and would uh would have uh not been an enduring feature of ancestral environments. And even even if it was, the idea that our ancestors were engaging in intercontinental travel is implausible. So that theory is resting on a really shaky foundation, and therefore we should have lower confidence in that in that theory that we have adaptations for racial categorization.

And that's part of how we can rule out that hypothesis. We can also rule it out with empirical data showing that we don't, in fact, categorize by race when there are other alliance uh cues present. But another way we can rule it out is just by saying that this is evolutionarily implausible. That's right. And if somebody disagrees with that, they can make that claim and say, but that's the point is you say it explicitly. And then that's the thing is by making it explicit, it's not less falsifiable. It's more because you're actually giving the logic of implicitly or explicitly, I think this is an enduring, long-standing problem. And that's why I'm thinking there's a system that has X, Y, and Z, like the anger system.

And by making that explicit, it's not naive, it's or it's not sort of uh incautious about certainty about ancestral environments. It's actually so by making it explicit, if somebody wants to claim Aaron, for example, like I don't think that's a feature of ancestral environments. By making it explicit, you open yourself up correctly, I think, to even a discussion about that. And there are many examples of within the evolutionary social sciences of different evolutionary hyp hypotheses about certain selection pressures, like how how large were social groups over evolutionary time, for example, or other things where where we can have these discussions. But if you don't make explicit the idea of here is the

claim about the enduring feature of the environment, then it's not more careful about certainty, it's less careful. Exactly. It it is not it is the opposite of unfalsifiable because what it does is it opens up new avenues for falsification and for criticism of the theory. If you're explicit about the assumptions you're making about the ancestral environment, that opens you up to new kinds of criticisms and arguments against your theory. You are making your theory more vulnerable to being challenged or even falsified. To take the example of race, if you're arguing that we have adaptations for detecting race, and I come along and show you some evidence from genetics showing that actually race is a really recent trait and it wouldn't have been present.

Well, you've falsified my hypothesis, or you've at least cast significant doubt on it. So what you're doing when you talk about ancestral selection pressures is you are opening up new opportunities for falsification. Right. So um moving right along. So right off the bat you have to come up with an ancestral selection pressure that would have been plausible, a foundation that you can build your theory on. That rules out adaptations for racial categorization. It rules out tons of other hypotheses that you would be inclined to make. Number two, the system you're positing has to uh be a good solution to that adaptive problem you're positing. Um, and what does it mean to be a good solution? It has to be well designed, It has to actually get the job done of, say, finding food or

categorizing alliances or negotiating for better treatment. And to the extent that it it does those things poorly, then that's another avenue by which your hypothesis can be falsified. What you're trying to do empirically is to show that this system does the job really well. Anger does the job of recalibrating welfare trade-off ratios really well. The the degree to which it is non-randomly organized to fulfill this function of recalibrating welfare trade-off ratios, that level of organization can only be explained by natural selection. It cannot be explained by random noise. This is the best fit for the data. This is the best theory. that's what you're doing. So it has to be consistent with what we know about evolutionary history.

The feature uh to which the adaptation um uh is is adapted to must have been present. The the system you're positing uh must actually solve the problem in question, and it must solve it uh efficiently, reliably, precisely, and effectively. Um just incorporating that into your theory building rules out so many hypotheses. Um there's so much psychology, for example, that argues that the function of a cognitive system is to pursue happiness or to generate positive hedonic states, to uh to make us feel good. So social identity theory argues that we we like to identify with groups and um uh defend the reputation of our groups because we feel good when we belong to a group and we feel good when our group looks good.

And and the function of this system is to produce positive emotional states. If you're doing good evolutionary psychology, good adaptationism, the function of pursuing happiness or increasing positive hedonic states is not a viable evolutionary function. Evolution does not design organisms to get stuff inside of their own heads. Evolution designs organisms to get stuff in the world that would have reliably increased their fitness in ancestral environments. The idea that we have adaptations for increasing happiness is incredibly implausible. Even if you were to assume that happiness, for whatever reason, was correlated with fitness in ancestral environments. Let's just make that assumption. Let's make the dubious assumption that in our hunter-gatherer past, people who were happier

uh were more likely to survive and reproduce. I don't think it's plausible, but let's make that assumption. Even if that's true, that would not result in a motivational system designed to seek happiness. The result would be that we would evolve to be happier. Just as if a squirrel is insufficiently fearful of predators, the squirrel would not evolve to pursue fear as a goal and try to like, they wouldn't be jumping out of the bushes of each other and scaring each other to pursue fear. No, they would just become more fearful. Right? If happiness truly increased our fitness, we wouldn't evolve to pursue happiness, we would just become happier. So, right off the bat, this is ruling out, you know, a huge chunk of existing psychology.

Can I can I push back a little bit? And I think we should have a happiness uh description. Yeah, that would be a good good topic. Yeah. I mean, what about the argument that it's a proxy? I mean, uh the way I think about those claims is that those are fine descriptions. And we even talked about this because you know I study groups. People are motivated, they feel good when they join a group. That's a perfectly fine sort of within the experience of a person description of sort of how they work. But it it but then you can translate into a more evolutionary perspective, which is well, what features of a group um would actually uh why would people feel happy about joining a group, or what should if we think about sort of like that's the

robot expressing this is a good situation. If if somebody says, like so, again, like thinking about people from this level of software, like people are robots. If the robot tells you I feel good right now because I'm a member of a group, you could be an engineer about it by thinking, well, okay, the robot software is registering, it's in a good situation from an evolutionary perspective, given what affordances and functions group alignment membership is, what can we predict about the kinds of cues or the kinds of software this robot would have to detect what counts as a good group, what counts as a bad group, and correspondingly when it should feel good or not good. Um and I think that's the way for me, at least I think about it, right?

So it gives teeth to what is otherwise trapped inside the internal arbitrary thing of just like it's essentializes, right? It's saying like a group is an essentialized category, feeling good is an essentialized category, and they're stuck together, like velcro together. And what an evolutionary approach does is it says group, you're not allowed to think a group exists. It has to be composed of a bunch of evolutionarily recurrent cues and happiness, or like the detection of this is a situation that's good, or maybe more sophisticatedly, a detection of a situation that's even better than is expected for now. Um what would those kinds of situations be? Yeah. And from an evolutionary perspective, and that gives that unessentializes and turns

into an engineering problem. Yes. What is otherwise the bland statement of like, I like to be happy because I'm a member of a group. Like, I don't think it's a problem for people to say that. It's just it's a very limited perspective on the broader phenomenon. Got it. I'm I'm going to do something I rarely do on this podcast, which is disagree with you. Excellent. Yes. I think it is a problem when people do that. Good. Um, and here's why. If I'm designing a robot to say, uh join groups and defend group reputations. Uh and I'm drawing up my blueprints of what the robot needs to be sensitive to and what it should be motivated to go out and get. There's gonna be no spot on my diagram for happiness or for good vibes, right?

I don't need that. I can just program the robot to go out and and find groups and detect them and join them. And if it detects cues that its group is under attack and needs to defend the group's reputation, the robot will spit out a compelling monologue about why the group is actually heroic and brave and making the world a better place or whatever. But I don't need happiness to motivate the robot to do those things. And so we need to ask the adaptationist question about happiness itself, about feeling good itself. What is that for? Why does that exist? And once we ask that adaptationist question, uh people are inclined to give the answer, oh, it's it evolved to motivate us to go out and get things.

That is a bad adaptationist answer. Because if the purpose of happiness is to motivate me to go out and get stuff that increases my fitness, well then how does evolution motivate me to want happiness? Right. I remember that. Does it have to give me does it have to give me more happiness? Yeah, yeah. It's a circular problem. So what is what is completely circular. Yeah, so what's your argument for what it is? Yeah. So my argument is that um what happiness is is a way of recalibrating our motivations and our expectations when something turns out to be better than we intuitively expected. So if I'm uh, let's say making a paella and I'm not confident about it, and uh and I think it's gonna taste bad, and it ends up being

amazing, um, well, my mind has a lot of recalibration to do. I had some mistaken beliefs about the quality of paella and my own ability as a cook. And those beliefs need to be recalibrated. And also my motivations need to be recalibrated. I now need to jack up my motivations to make more Spanish cuisine in the future. And so all that recalibrational work has to be done somehow. And I think the job of happiness is to do that recalibrational work. So what it what happiness does is it forces you to slow down, stop, pay attention to the awesome thing that's happening to you. Figure out how this thing could have happened. What how could how could you have been so mistaken? Right? Where did you go wrong in your expectations?

Where did you go wrong in your priorities and your motivations? Pay attention to the thing and figure out what is causing it, what is unique to this outcome. Once you figure that out, change your priorities, change your expectations, seek out more of this thing in the future. The next time you encounter it, encounter it, expect it to be good or expect it to be better. Yeah. Right? Good. Yeah, yeah. And and so that's excellent. And that is you're right. It's much more sophisticated. I wonder if then so like let's imagine, let's go back to social identity theory. Yeah. Um for me, which is like I think it's right, and it's for me it's a source of additional hypotheses about additional design, but I agree it's not fully fleshed out in the way that

they talk about it. Um imagine you're sort of um just going around and then like you just decide somebody is like you know got involved you in something on on the weekend and you go to it and it's like a really great social experience and you and it's a group, it's some group thing. All of a sudden you feel really happy. I think the the argument would be like all those same functions are there, which is that you know, you were mistaken, you weren't expecting this to be an important part of your uh uh the benefits, the sort of grammar of my representations of like, is this a group or set of people that I'm gonna invest in that I get benefits out of that are worth prioritizing compared to all the other things I could do, that feeling of happiness is

happening now for that group. And so I wonder, and maybe this is just saying the same thing in a different way, but for me, like when I hear social identity theory, I'm translating like people are motivated to feel good or to join groups because it feels good, as um that the feeling good is the thing that is leading them towards of all possible groups, the ones that are in a gradient towards better than what they're currently in. And the step-by-step movement towards better than they're currently in is caused exactly by this process you're describing. Yes. So that is a much better way of uh because then you have some sophisticated predictions about when you're gonna feel good joining a group.

Correct. When the group is better and a better investment than you would have expected compared to the existing alternative groups. And so we have another example of a beautiful prediction you can make by taking an evolutionary perspective. So um And when you should be dissatisfied with the group too. Yeah. Yes, exactly. Um but uh we can also say that uh being part of a group shouldn't necessarily always make you feel happy. Right. Um if if you know the group is is treating you exactly as well as you thought and exactly it's exactly as cohesive as you thought, yeah. There's no reason to feel happy. Your motivations are are set correctly, your expectations are set correctly, you can hang out with the group all day and not feel much happiness at all.

And so that's one case where taking this hedonic perspective of everything is about pursuing happiness will lead you astray. And there are even cases where um defending the group will make you feel unhappy. So being vigilant against threats to the group and and thinking in a conspiratorial way about all the people who might be out to get you or all the all the harms that might befall your group, that's going to make you sad a lot of the time. Uh, but we we should still evolve to do it and still adaptive to do it, because what we're pursuing is not happiness, but group integrity and group success in competitions. Yeah, that's right. And it makes predictions about to the extent that the group membership thing, what you're

getting out of it is social benefits and compared to alternatives or your own intrinsic value with respect to the group, it makes predictions of you should become more dissatisfied with a group to the extent that, for example, if all of a sudden your social status or bargaining power increases and nobody else is in the group does, that should make you more dissatisfied with the group. Or if another group all of a sudden becomes available and it has a better, it's better for you. And the argument is none of this is conscious, quote unquote. It's just the systems detecting this. Then all of a sudden the new opportunity should make you more dissatisfied with the current group. And so these are you're you're right, these are all predictions that are

unpacking what is otherwise a description staying at the level of emotions. It doesn't, and it's like kind of what Jamie Krem said in the other uh a previous episode, which is there's an additional why question, which is and how questions. It's well, what is a group? Why does join a group make you feel good? And and why and how, right? And what are the what are the predictions? And so I think so, yeah. I mean, do you think that that's fair, the what we just characterized? I think we just kind of converged on. Yeah, I think I think we did converge. Totally. Yeah, agreed. Yeah. No, I it's okay to say that we do things um to feel good, I suppose, as long as you're unpacking what it means to feel good from an adaptationist perspective.

That's right. Yeah, yeah. Otherwise, you're stuck in sort of this um uh Marvin Minsky has a book that's called The Emotion Machine. And it's interesting because he's coming at it from an uh uh an AI perspective, but it's very much this engineering perspective. And it's the Emotion Machine is it's his follow-up book to the much more popular Society of Mind book, which I recommend people read if they haven't already. The Emotion Machine book, in some sense, it's very sophisticated, in other ways, it's not, because some of the things he appeals to are sort of not in his wheelhouse, but but it's a very good book in in some respects. And it's about sort of high-level uh cognitive architecture things.

And one of the things he talks about in this book is that if you stay at the level of what we had talked about before is the intentional level, he doesn't use this language, but he sort of stumbles on this idea of motivations. So the person is motivated to like pursue groups that make them feel good. Yeah. It's sort of it really it's not an engineering problem, right? Or if you are an engineer, you just you're stuck with these questions, these deeper why questions that like Jamie pointed out. And so I think this is again another reason why it's not a just so story. In some sense, it's actually people are motivated to join groups they like, is a just so story. It's kind of ironic, right? It's that staying at the intentional level and not translating that into the fleshed out

with teeth, with predictions. What counts as a why should we feel good? Translating emotions and concepts like groups and belongingness um into a set of circumstances and outcomes that is the opposite of a just so story. I agree. Um, I I wanna chime in on that point in particular, which is I have this belief that there is no such thing as non-evolutionary psychology. I don't think that exists. I don't think you can do psychology in a non-evolutionary way. What you can do is do psychology with your evolutionary assumptions unstated and implicit. Um, you can do that. And that allows you to come up with theories that are evolutionarily implausible and have uh uh dubious assumptions attached to them because you don't bother to state those

assumptions and make them vulnerable to criticism. So you can do uh psychology, uh you can do bad evolutionary psychology without knowing you're doing bad evolutionary psychology because you haven't made your assumptions explicit. But at the end of the day, the psyche is uh an evolved organ, it is a product of evolution like any other organ, It is divided into subsystems with specific functions, like all of our organs are. The only way to study the design of that organ is by properly understanding the evolutionary history that gave rise to it, or the functions that were selected for. So when you're doing psychology, there's no way to study the psyche as something other than an evolved organ. If you're s making a claim about the mind, you are implicitly making an evolutionary claim

because the mind is a product of evolution. Any claim about it entails certain evolutionary assumptions. And so what evolutionary psychology is, is just making those assumptions explicit, forcing you to make those assumptions explicit so that they are open to criticism and that they allow you to screen out implausible hypotheses and constrain the hypothesis space and open up new avenues for criticism and refutation and falsification. It's just doing psychology, but you're being explicit about all the assumptions you're making. Yeah, that's right. Yeah. And Dan Nettle and Tom Scott Phillips have this great paper. Is a non evolutionary psychology possible? Um and I think that uh you know roughly is their point as well.

I think that's sort of a a way of putting it. Um yeah. Cool. So uh yeah that so yeah so number one that's why uh number one point is just that's that's why evolutionary psychology isn't a just so story or um as Lita once once said it, one of her students used to say that it's just so, meaning like it just spits everything just so. Um oh by the way, Richard Kipling, like uh, you know, of the just so stories. Rudyard. Rudyard, is that yeah, yeah. Um super racist. Oh really? And not in like a like, oh well, we have to like understand like he's a product of his time. Like uh Mark Twain, uh George Orwell were contemporaries were like scathing about him. Uh yeah, no, no, no. So uh so it's kind of funny people kind of throw Rudyard, how did you say it?

I can't even say Rudyard or Rudyard I don't know how to do it. I don't know how to apologies. Well not really but yeah apologies. Like uh yeah no no no so like people kind of throw that those just so stories um out of psych but they're sort of throwing a very racist imperialist uh person um at evpsy which is i think sort of ironically i i wasn't aware of that that little bit of history yeah yeah yeah uh he has this book called um or it's a poem the white man's burden um you know oh okay uh and then mark twain did a uh spoof on it sort of to the person sitting in darkness or something like that and it's sort of uh so Kipling's point was sort of imperialism you know it's sort of like for the good of the half demons uh

I think is a quote Yeah. And Mark Twain was sort of like maybe not. Okay. All right. Well, that gives us yet another reason to be happy that our discipline is not engaged in just so storytelling and is not uh associated with ruddyard or ruddyard Kipling. Yeah, that's right. And and and maybe check your uh if you're gonna use Kipl just so stories as a as a a wedge or as a cudgel, just check who you're who you're citing before you do that. But yeah, yeah. Maybe maybe Panglossianism would be the better site. Although I'm sure Molière was was racist too. Who knows? Yes, exactly. Yes. Uh things get better. Um. So yeah. Okay. Have we have we done it? Have we explained how to be a good evolutionary psychologist?

Or today. Okay. All right. I mean, this is very much a to be continued. To be that's a p that pun pun intended to be continued. Very, very nice. Okay. All right. Uh well, here we are at the end of the episode. Thank you for listening. Uh, if you like what you're hearing, uh please don't be shy. Uh tell people about it on social media. Uh shout it from the mountaintops. We're a new podcast, we're trying to build an audience and gain support for what we're doing. And and where can people uh support us, Dave? Well, they can support us emotionally however they want, but in terms of behaviorally, they can sort of leave a good review. They can support us on Patreon. And yeah, just listening, liking, subscribing, and leaving reviews

that are neutral to positive is certainly helpful. Absolutely. Yeah. If you feel that way, you know, if you feel like it's you feel negatively about it. Um you know, I'll leave that to your discretion. Yeah. Yeah. You if you hate us, you can send us hate mail too. Uh but you know. I would not maybe no, maybe not hate mail. No, if you hate us, just think about why you hate us and maybe think about something more productive you could do with your time. Yeah, if you have nothing nice to say, then don't say it at all. I think that's a good rules to live by. So but you know, you take that wo take that with a grain of salt. So but uh but yeah, thanks for listening and we'll see you next week. See ya