346 | Erica Cartmill on How Human and Animal Minds Think and Play

0:00:01.4 Sean Carroll: Hello, everyone, and welcome to the Mindscape Podcast. I'm your host, Sean Carroll. One of the topics that we've talked about frequently on the podcast is the nature of intelligence. Especially, of course, because we're thinking these days about artificial intelligence. Is that a good label? At what point are AI programs counting as intelligence? But even when you put that aside, just thinking about human intelligence and other intelligences in the natural world, there are certainly senses in which human beings as a species are different than other species. We're the only species that has podcasts and invented calculus and things like that. But is that a single fundamental difference between human beings and other species? Or is it an accumulation of many things? Or is it just something that we got there first in some sense? Different aspects of humanity have been suggested as the origin of our differences. Tool use, language use, maybe even how we speak and the shape of our throats and esophaguses, sheer brain power, number of neurons, social organization. But none of it seems to be quite the thing that tells the difference. So obviously one attack, one strategy to think about this question is to better understand animals and what they can do, how they are similar to human beings, how they are different than human beings, etc.

0:01:31.8 SC: And not just in intelligence, because intelligence is not just one thing. There's different ways to be intelligent, different skills you can have, and also there are other aspects of social life that are related to intelligence, but not exactly the same thing. Today's guest, Erica Cartmill, is an anthropologist, cognitive scientist, and also animal behavior scientist, an interdisciplinary person, and she studies different aspects of animal behavior and intelligence. In particular, the idea of animals playing with each other, animals teasing each other, animals having a sense of humor. This might seem like a simple and trivial thing at first, but when you think about it, the idea of telling a joke, as we do for human beings, is a pretty sophisticated idea. It's a social construct in some ways, because when you tell a joke, you're setting things up, which gives people, your listeners, your audience expectations, and then you subvert the expectations at the end. So that's actually a subtle move for an intelligent creature to make. Certainly dogs and cats and things like that play with us. They play fetch or whatever, but do they invent games? Do they play with each other in ways that have roles like pitcher and catcher that you could then switch sides and things like that?

0:02:56.7 SC: Does it evolve out of aggression? Is it the same as aggression? Is it different? Do they do it just for fun? Or is there some down-to-earth practical reason for fetching balls and things like that? So these are all fascinating questions and, unsurprisingly, the scientists who've studied them have discovered a lot of fascinating things. And as always, from my perspective, when I have these kinds of conversations, I'm both struck by how similar we are to different species in various ways, and how different. Human beings put things together in a certain way while sharing many, many of the same traits with individual other species. But there's a lot we still don't know about human beings and other animals. So, let's go.

[music]

0:03:57.4 SC: Erica Cartmill, welcome to the Mindscape podcast.

0:04:00.0 Erica Cartmill: Thanks so much for having me, Sean. It's great to be here.

0:04:02.7 SC: So one of the questions that comes immediately to mind, and I'm going to apologize from the start because this is one of those, do you think I'm right? Kinds of questions. But in reading your stuff and in talking to other people, comparing the intelligence and the capacities of other species to humans, one thing... It might have been natural to think that there is a linear progression of intelligence and other species just not as intelligent as us. And they're intel you can relate them to, oh, they're at a two-year-old human level of intelligence or something like that. But what I'm actually getting from your stuff and others is there's just some ways in which they're more intelligent than us, some ways in which they're vastly less. It's just a lot of different capacities that don't necessarily march in lockstep with each other.

0:04:48.0 EC: Yes, you're right. Moving on. No, I'm just kidding. No, I think that's exactly right. There used to be this model of intelligence, but also just evolution, the scala naturae. I know you know about that. Where you think where it's sort of the idea was humans were on top and other animals could be arranged working up towards the perfection that was human beings. And I think what science has done over the last, well, really many decades, has tried to really break that apart and say it's not a linear progression towards humans. Humans are equally as evolved as every other species. And so the things that humans have evolved to be good at are helping humans fit into the social and ecological niches that we have. But the same is true of every other species. And so you wouldn't expect humans to be able to fly or echolocate or see in the dark as well as some other species. And so why do we think about cognitive capacities as still being the sort of linear progression? The way I like to think about it or talk about it, because I think that what humans have is a unique constellation of abilities. Some of those abilities are more developed than other species. Some are less developed. But what we really have is this unique configuration, this unique constellation that pulls them together in a way that really helps them build on each other and gives us something that is really powerful and allows us to do all of the things we do as a species.

0:06:18.9 SC: But it just makes it so hard to have a nice, tidy narrative about our superiority this way.

0:06:25.1 EC: It is difficult. That's true.

0:06:28.1 SC: And it's also... It makes it more work for you folks who are studying this thing, because there's not one magic thing that makes humans different. It's we're one among many and we're all different in different ways.

0:06:40.8 EC: Yeah, it's tricky. Because so one of the things I started out studying language, and that was really always the bulwark of humanity. Humans have language, other animals don't have language. But when you really start breaking down, well, what do we mean by language? Is it syntax? Is it reference? Is it compositionality? The ability to put together small units into bigger units that have different meanings? Each of these things, when you look at a very specific feature of language, you can find at least one other species that has that ability. But what we can't find and I don't think we will find, is another species that has all of those abilities. And so I think that's what I really mean when I say humans have a unique constellation is that we put together all of the pieces in a way that give us something that might in fact be qualitatively different. Although it isn't just a sort of single switch that gets flipped on and says, "Ah, we have language, other animals don't." It's that other animals have pieces, they might have one thing better than we do, they might have little pieces, partial pieces of other things. But we have this... All the pieces fit together into this puzzle for us.

0:08:05.4 SC: I would love it if you could explain this example that you actually showed in a talk that I saw you give on YouTube. I didn't see the whole thing, but it was a chimp doing a task where there were numbers flashed on the screen, and then it was supposed to remember the order they were in. And the chimpanzee was better at this than human beings were, and it was embarrassing for the human beings. Could you explain this experiment?

0:08:29.8 EC: Yeah, so this is actually an experiment done in Kyoto at the Primate Institute there, where a chimpanzee, her name is Ai, and she was really the first one who was trained to do this, where she's trained to sequence numbers. And so, the numbers... So this is something where you have to train her slowly over time, but the eventual version of the task is the numbers one through nine flash up on the screen. She sees them, and then as soon as she... Her task is to press the numbers in order. And as soon as she presses the first one, they all disappear. They get masked, and so she can't see where they are. And so you have to look very quickly form a mental image of where they are and the order in which you need to touch them. And then very quickly, you have to keep that in what's called your working memory, sort of your short-term memory that allows you to continue the task that you're in the middle of. Now, Ai proved to be very, very good at this task, much better than untrained adults, human adults.

0:09:35.3 EC: And so we actually run... This task is something that's up on the computer screens in the chimpanzee exhibit at the Indianapolis Zoo. Chris Martin is the research director at the zoo there, who had worked with chimpanzee Ai on some of these studies, installed these for the zoo visitors. And it's very humbling for them to go and and try it themselves, because they see it and they're like, "Oh, cool, I'm going to try this." And then they see a chimp doing it and they're like, "Oh, I can beat the chimp." And a human has never beaten beaten Ai. However, if you take very young children and you give them, it doesn't even have to be very young children, but if you take someone and you slowly train them up, you can get to the point where you become an expert at this. So really what we're doing is kind of an unfair comparison because we're taking Ai, a chimpanzee expert, and having her compete against human novices. But one thing that... I know this is a little bit off topic, but I actually think it's really interesting and important from a cognitive perspective is that numerosity, the ability to learn, understand, and manipulate numbers is something that seems to differ quite a bit, or in particular ways between humans and other great apes.

0:10:54.5 EC: So one of the very interesting things with chimpanzees is that they can learn to recognize and to sequence numbers, one through nine. And I think the video I showed, and I know Ai can now go up into the teens. But teaching her a new number is just as difficult as teaching her the last number. So learning three is just as hard as learning two. Learning five is just as hard as learning four. And that's very different from what happens with humans. So in human children, there's this period of slow laborious development between learning the number one, the number two, and the number three. But then once they learn four, or maybe five, then they suddenly get it, and they realize that each new number is exactly one number higher than the number before it. It's called learning the cardinal sequence, the cardinal number principle. And so that's something where understanding that underlying pattern and being able to apply it to new examples, new numbers, is something that humans at some point, figure out and have this aha moment. And that chimpanzees don't seem to learn that underlying pattern in the same way, even though if they're trained on each of these different numbers, they can become incredibly proficient at recognizing and sequencing them.

0:12:20.0 SC: Can chimpanzees do simple arithmetic? Can they add?

0:12:25.3 EC: Well, it's interesting. This is not my particular area of expertise. I haven't done these experiments myself. My understanding is that a number of different species can... And when I say species, I'm including mostly mammals, but sometimes also birds. And also pre-linguistic human children. I know they're not a different species, but when we're talking about the kinds of tasks you run with animals, we often run very similar tasks with... Because they're non-verbal tasks, and we have to run very similar tasks with young kids. So the tasks that measure, say, addition and subtraction are not the number four plus the number five equals the number nine. Because they're not manipulating symbols in that way, but it would be showing a quantity four going... Say you have four balls, and they go behind a screen, and then you have five balls, and they go behind the same screen, and then the screen is raised up. Now, what you want to show in, say, a baby watching this, or a orangutan watching this, is when the screen goes up, if there are only three balls behind the screen, it should be surprising, because that's a violation of the additive principle.

0:13:44.7 EC: And so that is, in fact, what you often see with the species that have been tested. Now there are some other things. Again, math is not really my area of expertise, but there are some other differences where the kind of how far apart the numbers are before animals will recognize them seems to matter. Like it's difficult without being able to explicitly count, say, recognizing seven versus eight is very difficult because you have what's called a large approximate number system. And that's true both for young human children and for other species.

0:14:29.5 SC: But there's this... If I'm understanding correctly, yeah, so like, other other great apes can understand quantities and maybe even addition of quantities, but there is some level of abstraction that they're not quite reaching with what you and I would think of as addition and the successor operation from one number to another.

0:14:49.6 EC: Yeah, I think that's fair. It might be that some great apes who have been symbol trained can do simple arithmetic with manipulating symbols, but that's not... As far as I know that would be a real exception.

0:15:11.9 SC: Yeah, okay. And how well do we understand the explanations for these differences in terms of what evolution is asking of us? Can we tell stories? I don't know how testable the stories would be about why chimps are good at some things and less good at others.

0:15:31.2 EC: Yeah. So, it's difficult because, of course, in some ways it is just storytelling. We can generate lots of hypotheses about why humans are good at particular things. There are lots of hypotheses about why humans have language, why we have particular forms of language. Ultimately, it's very difficult to test those. We can ask things like, well, is the structure of human language linked to our ability to see hierarchy in the steps of producing a tool? So there are a lot of theories about, say tool production or material culture and the fact that it has nested hierarchies of you have to do this before you do this, but you have to keep the end goal in mind in order to structure the intermediate goals. And that's very similar to structuring a sentence. So there are interesting theories about dating the evolution of fully structured human language. And I say that maybe connect it to the emergence of certain kinds of stone tools. So we know that apes can make tools, but they can't make the very complex.

0:16:57.4 EC: They're called Acheulean stone tools, where it's not just flaking off a single edge, but it's flaking a shape to make a prototype base shape and then making secondary flakes to give it a sharp edge on one side and not on the other or modifications on top of modifications. And it's that future thinking and planning that some people think is a marker that species that could do that also probably could structure complete sentences. But at the end of the day, all we can really do is say, can other species of great apes that are alive today understand and appreciate structural differences in human sentences? So there are studies with trying to teach great apes language and asking do they understand the difference between the dog chased the cat and the cat chased the dog? And you can also ask do they understand how to plan for future events and how to order different sort of intermediate steps and things? And so I think that's really where a lot of the work on comparative cognition, looking across different ape species and saying well, not just what do these species do in the wild today, but what are their underlying cognitive abilities if they're put in the right... And I don't mean right, and that there's one correct situation, but if they're put in a similar situation, given access to the same kinds of substrates and motivations, can they complete a task in the same way? Can they understand particular sequences or orders? Is there structure to the way they think?

0:19:01.0 SC: Yeah, and among... I presume I know the answer to this, but I'm going to ask you because you're the actual expert. Among the different kinds of great apes, the gorillas, the chimpanzees, the orangutans, the bonobos, are there like again, unequal capacities there? Some of them are better than others, but worse at other tasks? How much do we know about that?

0:19:20.4 EC: Yeah, I think just like I was saying with comparing humans to other species, comparing across the great apes I think really yields a picture of different areas of excellence, if you will. So it used to be the case that people would think that chimpanzees and bonobos, who are the most closely related to humans, maybe were the most sophisticated cognitively. What they are is very good at particular kinds of cognitive tasks. They are the most social of the great apes. So they live in large multi-male, multi-female groups, as humans do. Just like us, they're interested in gossip. They're very interested in paying attention closely to who's doing what with whom, when. And so any things that have to do with paying attention to others' motivations, paying attention to what information others may be keeping from you. These are things where chimpanzees and bonobos really excel because those are things that are important in the their natural environments. Now, orangutans, which is the species I started out studying, they spend a lot of time, the vast majority of their time in a semi-solitary state.

0:19:51.8 EC: So males in the wild are solitary. They have large home ranges, they range around, they look for females to mate with, but they don't spend social time with others outside of courting a female. Females will have an infant and then who grows into a juvenile. And they will have one baby at a time who they travel around with and they spend all their time together. And that is true for 10 to 12 years. So it's a mother-infant pair, travels around, and then when the mother has a new infant, the older offspring is displaced and goes off to find their own way in life. They go to college, if you will.

0:21:34.4 SC: Emptiness.

0:21:36.4 EC: And so that kind of what sometimes people call Machiavellian intelligence thinking about who's plotting against me, who do I need to make friends with in order to overthrow the king, things like that.

0:21:53.5 EC: Those types of intelligence are not areas where orangutans really need to use those skills in their natural environment. They are social. They can pay attention to what another's goals are. They're not completely solitary individuals. So I don't mean to give that impression, but they're not as intensely obsessed with others as bonobos and chimpanzees are. Where orangutans really seem to excel is in material culture, material manipulation. So they are very good at figuring out manual tasks. And they also, as I've worked mostly with orangutans and now with bonobos and chimpanzees, and they also have very different sort of levels of energy and very different personality types. So if you give them a problem to solve, an orangutan will sit there and they'll stare at it for a long time and they might chew on something. They're just looking at it. They might move a little bit, look at it from the side, but they're just sitting there. And then eventually they will get up and go over and they might solve it in the first try. They might not. But they're really kind of... They give it some space and then they give it a shot.

0:23:18.0 EC: And chimpanzees and bonobos, you present them with something and they immediately want to get their hands on it. They're like, "What's this? I'm going to go over here, I'm going to poke it, I'm going to shove it, I'm going to shake it, I'm going to bang it up and down. Look at me." And those are really differences in that underlying motivation that can lead to different conclusions about cognition, but really have more to do with an orientation towards the world than they do, and maybe an orientation that's driven by your social world. If you're a chimpanzee or a bonobo and you have an opportunity to try to get a piece of food, if you're not there immediately you might not get it. Someone's going to get to it before you. If you think about, I don't know if you remember this old Campbell's Soup commercial where they pass the bowl down the table and it's like, from the older kid to the middle kid to the younger middle kid, and it's like he gets all the way to the youngest one. He's like, "Wait, there's chicken in here." It's like, "Now with chicken noodle soup, now with more chicken." And it's that kind of like, if you live in a big social world where you're not the one who gets to eat first, you better get in there when the getting's good. And so I think some of these things about the social and physical environments that species have been adapted to live in can really change the way they're inclined to engage with the tasks that we as scientists give them.

0:24:49.7 SC: As a physicist, I got to think that the orangutans are just theorists and the chimps and bonobos are experimenters. They're in there. They want to knock things around and see what the data are going to tell them.

0:25:00.8 EC: Yeah, no, that's very fair.

0:25:02.6 SC: The other thing I saw in the same video that I was talking about before that it really struck me. I didn't see the whole thing, so I'm not quite sure how it developed. But there was a question about whether chimpanzees can recognize what it means when a human being points at a certain box. And someone asks, "Well, do chimpanzees themselves ever point like that to indicate things?" And your answer was something like, "No, because if chimpanzees have a resource like food, they want to keep it for themselves. They don't ever want to tell anyone else where it is."

0:25:34.6 EC: Yeah, and that is more or less what people have found. I think for a long time, scientists argued that humans were the only species that pointed, and particularly the only ape that pointed, because most of the comparisons about language are between humans and other apes, although, of course, there are very interesting other kinds of models, like in birds and cetaceans, you know dolphins and whales. But the point in question is a very interesting one that changes depending on the environment that the ape is in. And so when apes are in human care, in situations where they are prevented from reaching things themselves due to physical barriers that the humans put up like say, in a zoo, they can pretty easily learn to indicate to someone a thing that is out of their reach that they would like, and the human will give it to them. Now, is that pointing? It certainly looks like it. You can extend an arm towards a thing. There was a lot of nitty-gritty discussion about, "Well, are they pointing with an index finger? Or they just... Is it a modified reach? Are they maybe just reaching to get it?"

0:27:04.6 EC: But sometimes the thing is really far away or... So I'm not convinced that they're just reaching. I do, in fact, think they're pointing. And actually, if you look across human societies, humans point in all kinds of different ways. It's not just that we use this extended index finger. There are places where you point with two fingers or a whole hand. Actually, people get trained in Disney parks to not point with a single finger, because it's considered rude in some cultures. There's also some societies where people point with their chin by stretching their chin towards something or inclining their head. And if you think about it, you might do the same thing if, say, you were carrying something and someone asked you for directions, and you might incline your head and say, "Oh, yeah, it's over that way."

0:27:48.1 SC: But is there also a feeling that humans are more sort of cooperative/trusting in their ways of dealing with the world? Like we sort of rely on other people to cooperate with us to find things in ways that, I don't know, either chimpanzees or any other great apes are sort of more self-reliant.

0:28:07.4 EC: Absolutely. I think human society... It's for all of our not great traits as a species. And as much as we have liars and cheaters and things like that that appear in news stories and... Humans are fundamentally a cooperative species. We share information with others that doesn't benefit us, sometimes it disadvantages us. If someone says, "Hey, what's going on?" Or, "What what are you standing in line for?" Like, why would you tell them? This might be something where I just want this myself. Or someone asks you, "Where's the best noodle place in Poughkeepsie?" And why would you tell them? Because you want more noodles for yourself. But these are things where we will stop and help people we're not related to. We will share information freely. And that information is rarely a lie. It is, obviously sometimes people are very capable of lying, but if you look at the average person and the average interaction, people are very trusting. And I think that that's one of the real hallmarks of our species is that we... In fact, that's one of the things that makes language work is the ability of both the person who's speaking and the person who's listening to trust that the other one is communicating in good faith.

0:29:43.1 EC: If you tell me something, I don't immediately... Not every time anyway. I don't immediately say, "Prove it." Because we're... If in order to talk about things that we talk about that aren't in, just in the here and now, you have to believe the other individual. And so this entire enterprise of, I mean podcasting, for example. That is not something that I think any other species could do. That doesn't mean that they can't communicate at a distance. So chimpanzees make long calls. Orangutans make long calls. Lots of species make alarm calls. Birds have these territorial displays. Anything that's broadcast. It's out there for others to listen to. But it oftentimes is linked to the specific information about the time and place and the speaker or the signaler. If a bird is giving a territorial display, and you hear it, you know something about who they are, where they are. These are actually really... And this is a little bit off topic, but it's actually one of my absolute favorite studies that I always teach in animal cognition, where song sparrows were... So song sparrows have very discrete territories.

0:31:15.0 EC: And one of the, I think, neatest experiments that shows that animals encode information not just about the type of call, but who is making the call and where they're making the call from is that if you take a call made by, say, the neighbor on the right-hand side. Like a right-hand territory. So you think about you live in a house and you have a neighbor on the right and a neighbor on the left. So if you take a call from the neighbor on the right and you play it back, but from the left, the birds will respond. The bird in the middle will respond very strongly because it's like, "Wait, Fred's in the wrong place."

0:32:02.8 SC: Okay, good.

0:32:04.0 EC: And they'll respond basically as if Fred's a stranger. But if you play Fred's call back from where it's supposed to be, they don't respond or they don't respond as strongly because that's what they're used to hearing. And so I think it's really interesting that it's not just that animals are recognizing, "Oh, this is an alarm." Or, "This is a food call." But they're processing information and comparing it to their memory of who is providing that information and where is that information coming from. Yeah, so I think that animal communication is both different from human language, but also more often more complex than we think it is.

0:32:48.1 SC: Well, it leads right into where I wanted to go next, which was the idea of roles as an understood concept in animal societies, if you can even call them societies. Like, presumably there's definitely cases where there's two animals and one sort of is dominant over the other. But to what extent do different species have well-defined niches that individuals play? Obviously, insects have biological differences. One is a queen, one is a worker. But just culturally how obvious is that notion in primates, for example?

0:33:26.9 EC: Yeah, so it differs a lot between species. More broadly across the animal kingdom. I think you have some colony animals that live in colonies, like say meerkats or chimney swifts or again animals that live in a group where you have multiple generations in the same group. In some of those species, there will be a single breeding pair. And so in that way... And everyone else is helping. And so in that way, it's similar to a eusocial insect like a honeybee, and that there's one "queen". They're not as physiologically different as the queen bee is, but the dominant female will suppress reproduction in the younger or subordinate females. And so you do have some of these structures when it comes to mating structure. In non-human primates I think... And you certainly have social hierarchies. Again, depending on the species. I don't want to sort of overgeneralize too much. But, some species have very strict dominance hierarchies in either the males or the females or both. So in some species, you can have a hierarchy where it's mixed genders and females can be mixed in there with males.

0:34:54.1 EC: You have some where it's all the males, but there's a strict order of males, and then all the females below all the males. You can have distinct rights and responsibilities, if you will, that are afforded to the dominant individual. And so it's not just, say, mating privileges. So some species you might have the dominant male is allowed to mate with all of the females whereas the subordinate males are only allowed to mate with other females if the dominant male sort of allows them. And I don't mean gives him a written pass, like a hall pass, but sort of observes the male courting another female and doesn't attack him. And so you can have alliances that are formed. And it's like, "Well, if I'm the dominant male, I can mate with all the females, and maybe I let a couple of my good friends mate with a few of the females, but maybe not my favorite female." So you do have these kind of again, very complicated social rules. What you don't have is you don't have a division of labor where some are collecting fruits and some are weaving baskets and some are collecting water.

0:36:17.6 EC: You do have some division in terms of hunting. So in primate species like chimpanzees that hunt, you will have some individuals that, which is usually the males, but a few females as well, who will go on hunts together, and they will hunt usually for monkeys, but other species too. And they will... Tends to be the same individuals that will go on hunting parties. And then they come back and they will share the meat with favored individuals. So they'll share it with their friends, males will share it with females that they're courting. And so meat is a very sort of precious resource, is high calorie content. And so it's something that is used a little bit like a currency in species that hunt and then food share.

0:37:12.6 SC: So I can imagine if I'm a chimp or whatever, knowing my place in the hierarchy. How much do primates know about the relative hierarchy of other members of their group?

0:37:25.4 EC: Yeah, quite a bit. There's some really beautiful experiments done, I think originally by Robert Seyfarth and Dorothy Cheney at the University of Pennsylvania, but some others have since then, where you can simulate hierarchy reversals by playing back simulated calls. So you can play... So let's say you have a fight where one individual is giving aggression calls and one individual is screaming having sort of victim screams. And then maybe afterwards does some submissive vocalizations to sort of show yes, you're in charge, I yield. And right, Uncle, Uncle. So what you can do is if you record those, you can then cut them up. So you have a call from individual A, who's maybe an aggressive call, a call from individual B, a call from individual C. And maybe those are... That's the order A, B and C. So A is the dominant, B is in the middle, and C is lower down. Now, if you take those calls and you splice them together to create a simulated fight where, say, C is attacking and beats A, and then play that recording back to other individuals in the group, they will be very surprised.

0:39:09.4 EC: So one, they will look longer at the speaker, and then two, they will show more aggression to A afterwards. Because this is sort of a moment where they say, "Wait a second, A is vulnerable. And so maybe this is my chance to improve my spot in the hierarchy." And so they're really paying attention to not just to their relationships with others, but to others' relationships to each other. Because those are the kinds of things that you need to know if you want to get ahead. If you want to have mating opportunities, it's really important, and you're male, it's really important that you make friends with other males who are going to be in power if not now, then in the future. If you're a female and you want to make sure that your offspring are protected and they're not subject to, say, infanticide, if a new male comes into power, you need to have strong relationships, not just with other females who could come to your aid and help protect your infant, but with other males who in the future may one day take over. And so it's very strategic. It definitely needs an HBO show.

0:40:27.5 SC: It sounds like it. It makes me wonder, do animals have a sense of humor?

0:40:33.0 EC: Well, it's certainly something that I have been fascinated with for a long time and something that I have been exploring really over the last five years. I've been trying to figure out whether... Well, particularly great apes, have a sense of humor, have a sort of understanding of surprise or sort of these subversive moments. And I think this comes out of the things that they pay attention to in their natural interactions. But yeah, we've actually been working on an entire research line looking at humor, but really through the lens of what I call playful teasing. So teasing is something that I think is a fascinating opportunity to look at the social minds of animals because it's this gray area. It's sort of right in between aggression and play. And so a lot of times people when... I think when people talk about teasing and the way teasing has been studied in the past in animals anyway, it has really been thought about as a form of aggression. I'm pestering someone, I'm bothering them, I'm doing something to try to improve my stance in the group.

0:42:04.7 EC: It's sort of a low-level form of aggression. But if you look at humans, we use teasing all the time in positive ways as well as negative ways. And so I think also when people talk about teasing in humans, it has a pretty bad rap because it gets thought about as a precursor to bullying. And obviously sometimes it is, I don't mean to say all teasing is positive, but some forms of teasing, I think are actually very loving. I think they build relationships. I think they're critical in developing this sort of complex understanding of how far another individual will go for you.

0:42:55.5 SC: So the short answer is yes, you think that animals do have a sense of humor or is that anthropomorphizing too much?

0:43:03.9 EC: Yeah, I certainly think that they... Well, it's also hard when you just say animal, Sean.

0:43:11.1 SC: You're allowed to narrow it down. You're allowed to change any question I ask into the question you want to answer.

0:43:16.8 EC: Because I will tell you, I think some animals have a sense of humor. I think some animals probably don't. So for example, right now so I'm doing some wildlife rehabilitation and I have a possum, who is an unreleasable... He's an educational animal. And he has a disability that means he can't be released. And he is a fascinating animal and I love possums. I could talk about them all day. But he does not... I do not think that he has a sense of humor. He's very good at what he does. He like Pac-Mans around eating things and he'll cuddle. He's very sweet, but I do not think there's a whole lot going on upstairs. Not to put animals on a ladder again, but I think that is not an ability that I think needed to evolve for his species to survive. Whereas we also have a cat and a dog. I know you have two cats and I certainly think that cats and dogs have a sense of humor in a way. I certainly think that they have things that they know they're not supposed to do, that they enjoy doing because it gets a reaction. And I think that that pushing the boundaries and having these moments of something unexpected or a violation of a social norm or an explicit rule, I think those are the things that are really at the heart of humor.

0:44:57.5 SC: Well, I'm glad you put it that way. It's very helpful because my current cats actually don't do this, but I used to own a cat who did the traditional cat behavior of sitting on a table or a cabinet and knocking something off. But she would do it in a very specific way and I swear I'm not hallucinating. She would move the thing toward the edge and then she would stop and look at you like, "Are you going to stop me?" And like, "No." Okay, you're going to move it closer, stop and look at you and then eventually push it off. As a human being, I want to interpret that exactly the way that you just licensed me to do, that she was clearly demanding attention, annoying me purposefully. And that could be interpreted as a sense of humor, even maybe a mean-spirited one.

0:45:43.6 EC: Yeah, I think it's a very valid question of what, is all humor at its heart mean-spirited?

0:45:48.6 SC: Exactly.

0:45:49.6 EC: But certainly when I started this project I thought that my husband and his brother were very mean to each other. And I was just like, I'm an only child. I grew up in a family that did not tease or didn't tease to a great extent. And my husband is one of five boys and there was a lot of teasing in that family.

0:46:23.0 SC: Not going to escape it. Yeah.

0:46:24.4 EC: Yeah, and I think when I saw him and his younger brother get together and it was just, how can I get under your skin until it really irritates you? And I have to say conducting this line of research in great apes has made me develop a much greater appreciation for the love that they have between them. I'm still not sure I'm going to fully participate in that level of, how can I irritate you? But but I do think that it's fundamentally built on respect and trust. Coming back to this example you gave about your cat with the, "I'm going to look at you and just push it a little closer. Oh, maybe I'll wait. Oh, I'm going to push a little more. Oh, are you coming close? Oh, now." One of the things that started this whole project was a... We were looking for three different behaviors that a psychologist at the University of Portsmouth in the UK, Vasu Reddy had identified in human infants. So she called them not humor per se, but she called them clowning behaviors.

0:47:47.7 EC: So they were sort of non-verbal forms of humor. And these behaviors were offer and withdrawal. So, "Here, would you like this thing? Haha, just kidding." Disrupting others' activities. So you're doing something and I come and jump in the middle of it or make it impossible for you to do the thing you're doing. And then the one with my favorite all-time label, which is provocative non-compliance. I think that cats are masters of provocative non-compliance. But you think about... And so does that mean they have a sense of humor? I would like to leave the door open for that, but not have a strong conclusion. I would say it suggests that they have an appreciation for violating others' rules in that way. But I think it needs systematic exploration in a sort of scientifically objective way where we're saying, "Well, will they do it when someone is out of the room?" They do it in the same way. Are they really looking at your response? Are they trying to get a particular response? So if you put yourself in the sort of role of an experimenter for a moment, and you said, "Well, when they do that, I'm going to turn and walk out of the room." Is that something where they're suddenly not going to be interested? Are they doing it in order to get a response from you?

0:49:10.6 SC: But I love the connection with violating expectations. It's almost, maybe this is pushing too far, but I'm a theorist by inclination. So you can imagine that this behavior grows into telling jokes, which I presume that anyway, cats don't do. Maybe chimpanzees do.

0:49:31.3 EC: Yeah. And I do think that this violation of expectation is really at the heart of what a joke is. Of course, we can push jokes. As humans, we push them to the extreme, and we tell these long and complicated jokes. But I think fundamentally a joke is setting up an expectation and then violating it in some way. And that was actually the thing that initiated this entire project for me was when I was... I completed my dissertation studying gestural communication in orangutans. So I went to zoos in the UK and Europe, and I was looking at how orangutans used gesture to communicate with each other. And during that study, I had this... I witnessed this interaction that I had no idea how to study from a gestural perspective, but I just thought it was fascinating. And it stayed with me. And I came back to it more years later than I would care to admit without giving away my age, when I had the opportunity to apply for a grant in this space. And so in this interaction, there was an infant orangutan who was hanging from a rope over her mother.

0:50:42.7 EC: And she had a stick in her hand. I think it was a piece of bamboo. And she reached down and sort of offered it towards her mother. Her mother reached her, her mother was lying on the ground on her back. And mother reached up and started to grab, reach for the stick. And then the infant pulled it back out of her reach and the mother put her hand down.

0:51:01.9 SC: Hilarious.

0:51:02.4 EC: And I was like, "Oh, haha." And then the infant did it again. And the mother reached and the infant pulled it back. And then the infant did it again. And I just thought, "Oh, this is a cute little sort of game and the mother is tolerating the infant." But then what I thought was even more interesting was that at some point, the infant dropped the stick and seemed to get tired of the game. And then the mother picked it up and started doing it back to the infant. And so to me, that moment, that role reversal really demonstrated that it wasn't sort of an idiosyncratic routine that developed between that infant and that mother. It was a game. It had a structure. It had different roles that could be played. It obviously wasn't really violating an expectation, because if you sort of understand what's going to happen, you know that they're going to pull it back. The mother wasn't reaching for it really fast to try to get it. She was playing along. But within each of those offer and withdrawals, it has this structure of a very basic joke.

0:52:10.5 EC: It has this setup, the offer, and then this punchline, if you will, this violation, which is the withdrawal. And once I started to think about it as a very simple joke, it made me want to look to see where else apes might create these moments of expectation violation for each other. It seemed to be something that they both were enjoying doing. They were both doing voluntarily. And it seemed likely that this was a behavior that you might see more broadly if only you knew what to look for.

0:52:53.1 SC: It reminds me of a recent podcast I did with Rebecca Newberger Goldstein. I don't know if you know her, but she's a philosopher and novelist, and she has a book out on mattering, what it means to matter and why we care about mattering so much. And she defines it, what it means to matter is not to just want attention, but to feel that you are deserving of attention. And certainly, it seems to me, again, correct me if I'm wrong, but some of these games, some of these proto jokes are about asking for attention and you being rewarded by like, oh, yes, I am worthy of it. Maybe there's a benefit to being teased as well as to being the teaser.

0:53:34.9 EC: Absolutely. Yeah. I've been thinking a lot about the function of these teasing behaviors, because I think it's interesting just to document them, of course, and say, "Do apes do these? And what do they look like and when do they occur and who's doing them?" But ultimately we want to know why are they doing them? Is this something that has a benefit to developing a place in a social world? Is it something that helps you understand others' minds more? Is it something that helps you develop social relationships? And so one of the things that we've been thinking about is that they might in fact be serving these playful teasing sort of proto jokes, if you will. Might be serving both to build relationships, but also to test those relationships because I think they have this element of, how far can I push you? Or how far am I willing to be pushed? And those are things that are not transferable across individuals. I can call a good friend of mine a bitch, but I can't call my... Can I say that on the podcast hopefully?

0:54:45.8 SC: You can say that. People have said far worse things.

0:54:48.5 EC: Okay. But I'm not gonna say that to my boss. And so I think that those are things that can serve to strengthen a relationship. You think about early dating. A lot of flirtation is teasing. A lot of it is how oh, I have a little sort of nudge, a little poke, a little like, see how you take it. Testing the waters. Do you laugh it off? Is this like, do we have that kind of relationship? And if we do, then you can go a little bit farther. And I think that's true of of friendships, it's true of romantic relationships, it's true of family relationships. And those kinds of... I think the way that teasing can serve as a sort of test of relationship strength is seeing how much the other individual values the relationship over their momentary annoyance. It's like, "Ugh, fine, but I love you, so I'm not gonna be annoyed." Especially when you look at parents and infants or parents and young children, they put up with all kinds of things that they wouldn't from say... You just look at anyone in a grocery line. If a young child is doing something to a parent, the parent is going to put up with a lot more than they would if the unrelated stranger adult behind them did the same thing. Because it's an honest signal of how much they value that relationship.

0:56:28.5 SC: And one of the things that young children do and also animals do is scope out. I talked to Judea Pearl on the podcast once. Make a causal map of the physical world. Like if I poke this, what happens? And what you're suggesting is teasing and play, play an analogous role in the social world. You're sort of seeing where you are, what the relationships are, how you can interact with different members of the social group in different contexts.

0:56:57.6 EC: Absolutely. And I think that one thing that is interesting about this social hypothesis testing is that it isn't static. It's not that once you learn the rules, you now know them. I think there's a certain extent to which, once you learn the physical properties of an object, you know them, you don't have to keep testing them. I don't think the same is true of social relationships because those change in a way that physical properties don't. Your relationship with your friend or your sibling today might be quite different than it was yesterday. It might be different because they're in a different mood. It might be different because you pushed them too far yesterday. It might be different because something changed about their environment. And so learning how to predict their responses to your actions is very critical in accurately anticipating and evaluating the relationship that you have with them. I remember anytime being a graduate student, and there would be days when you would go into your advisor and and they'd be in a great mood and they were like, "Yeah, this is a great idea, let's go with it."

0:58:24.4 EC: And other days where they were like, "This is a horrible idea, back to the drawing board." And you would rely on the graduate students who had meetings before you to tell you whether this was a good day to pitch a new idea or not a good day. So I do think that the social relationships matter, but this is a landscape that continues to change over time. And so you want to not just learn about the baseline state, but you want to improve your ability to predict. And I think that's one of the things that teasing is doing is it's giving a sort of low stake environment for practicing the attribution of others' reactions to you.

0:59:17.2 SC: So there's teasing and joking and so forth. But is it accurate to put game playing at a slightly higher level of abstraction? We human beings have games with very complicated rules and roles and things like that. Anything analogous to that in different parts of the animal kingdom?

0:59:36.7 EC: I don't think there's anything, say, analogous to chess. But I certainly think there are things analogous to tag and to hide and seek and to some children's games. There's an appreciation for turn-taking that can occur with things like chase, dogs all the time. I chase you, now it's your turn to chase me. And it isn't just that I turn around. And it isn't just that like one dog chases... A chases B, and then at some point A gets tired, turns around and runs the other direction. A will turn around, run a little ways, look back at B, and if B doesn't follow them, A will go forward, will play bow, will bark, will nudge them. And it's sort of a reminder, "Hey, you're supposed to be doing something."

1:00:30.3 SC: We're playing a game here.

1:00:31.8 EC: And so I do think that there are these simple role reversals in social games that animals have. I think it's also possible that there can be non-social games. And that's one thing that I'm trying to look at a little bit. We started a new project. I'm launching a canine cognition lab at IU and our first study, we're doing a project looking at dogs' collecting behavior. So we're interested, it's a collaboration with a developmental psychologist, Martin Zetterstein, who has looked at this in children and looking at children's collections of objects and how those differ around the world, change over time and things like that. And he's found that children basically everywhere collect at least one thing. And I'm really interested in whether any dogs exhibit similar behavior. And so one of the things we're asking in that is do dogs who are interested in a particular type of object, do they manipulate them or say order them or store them in any way that has internal structure to it? So do they put them all in one place? Do they rearrange them?

1:01:48.3 EC: Do they put all the purple ones together? And so I do think that it might be possible that there are a lot of things that we could consider games if we looked at them that could consider essentially private games things that are about setting arbitrary goals and trying to achieve them. The one example I have that's a self-driven game from my dog Bonnie. So she has this little stone turtle that she is obsessed with. She actually has 11 stone turtles, but there's one in particular. They came from a garden store. We used to have them in our garden and she just loved to find them and pick them up and bury them. And so there's one. Sorry. There's one whose name is Darius, by the way, because every time we find him, I go, "Darius." So there we go. I know it's really cheesy, but she likes it. So in this game, I will roll the turtle up in towels and she has to dig through and find him. And she loves this. And she will play it for hours and just go completely...

1:02:58.6 EC: It will bring her out of any bad mood or if she's stressed out about having gone to the vet or whatever. But what I think is really, really interesting, and one of the things that leads me to think that animals, at least some animals again, might sort of play games by themselves, is that she will bury it for herself. So if I've gotten tired or I'm on a Zoom call or whatever, she'll take it and she'll put it on a towel, and she'll put another towel on top of it, and she'll smush it all up, and then she'll sit there for a minute, and then she'll try to find it again. And I really think that she is doing something very similar to what young children are doing when they're playing... I think as much as you can call a game something like, I'm going to see how long I can stand on one leg or how many times I can spin around before I fall down. We call these things game like play in children. They don't have structured, documented rules, but it's about setting goals for yourself that don't have any survival... They don't have any immediate benefit other than it feels good to achieve them.

1:04:07.6 SC: It feels good to achieve them. Yeah. I guess that's where I was interested in going next. What is the... I don't want to use the word motivation, because maybe that's already being too anthropocentric. Is it simply that they do it for the joy of it? Is it an instinct that is kicking in? Is there some tangible benefit?

1:04:28.2 EC: To teasing or to any game?

1:04:31.9 SC: To teasing, to playing? Yeah, you give me the answer that is most interesting.

1:04:35.7 EC: So I do think that... I think it feels good. So one of the things that we're trying to do actually in a different project, but ties in together, is I'm working with philosopher Colin Allen and a number of other animal experimentalists and ethologists at different institutions around the world to look at joy in animals. And we're actually trying to figure out... We're looking across three different, very distinct species, distantly related species. We're looking at kea parrots, looking at dolphins, and we're looking at great apes. And in this study, what we're trying to do is to develop better biometric measures of positive emotion to try to look at when animals experience positive emotion, do they communicate it to other individuals? And how does perceiving those communications impact things like memory, attention, and prosociality? And so with apes, we've been looking at laughter. First of all, humans are also great apes, so I've been saying apes just to mean non-human apes.

1:05:54.7 SC: We get it. Yeah.

1:05:55.1 EC: Yeah, just making sure everyone gets it. But so apes laugh just like humans do. Their laughter sounds a little bit different. And at sort of the risk of making a fool out of myself, I will try to demonstrate. So where I think human laughter is quite musical, chimpanzees and bonobo laughter is a little less musical. So it still has this rhythmic panting element, but it has less melodic qualities. So it's a little more like [1:06:19.3] ____.

1:06:20.0 SC: Okay, I recognize that. Yeah.

1:06:20.9 EC: And it happens a lot during tickling. So, it happens a lot with infants. It happens during play. Adults will tickle infants, infants will laugh. I'm sorry, and gorillas and orangutans will also laugh, but their laughter again is even less musical and more breathy. But all of the species laugh. They laugh in very similar environments. And so what we've been doing is trying to use laughter and using it as an auditory stimulus for a whole set of experiments to say, "How does hearing another's laughter impact the way you think about the future, the way you interact with others, the way you remember things, the way you perceive facial expressions." Things like that.

1:07:24.1 EC: And generally what we found in this line of experiments is that it seems to impact them both physiologically and cognitively. We've done some things with thermal imaging where we're looking to see, which basically is an external measure where you can use a thermal camera to look at changes in blood flow by measuring the skin temperature. And so what we're looking for in great apes is that we're looking for a dip in nasal temperature. So we look for the nose to get cold as an indication of emotional arousal. So you're getting more excited, and when you're excited, the blood gets drawn into your core, so your things in your periphery get colder. It's sort of a fight or flight reflex. And that can happen when you're feeling excited and happy. It can also happen when you're feeling excited and angry. So we want to pair that with other behavioral measures to say, "Well, how do we know the animal isn't just scared or isn't pissed off?" They don't hear this laughter and they're like, "Well, that sucks." And I'm like, "I hate that guy." And so we're pairing it with things like, we did what's called a cognitive bias test, which was developed originally for use in animal welfare as a way of trying to say, do animals become more optimistic or more pessimistic following changes to their enclosure?

1:08:58.7 EC: And so we apply this to great apes. And so in this test, what you do is you first train them on two, we call them anchor stimuli. So in this case, it was actually a black box and a white box. And the black box was always rewarded, and the white box was never rewarded. And so what they did in this study was they pressed a button, and then we showed them at the other side of their enclosure a box. So they press a button, and then another person holds up a box, and then the ape has to decide, do I want to get up, expend energy to walk over and see if there's food in the box? And so they learn, it took some time, but they learn that they should always go for the black boxes and never go for the white boxes. And then once they get that down and they're demonstrating with very high precision, go for the black, skip the white, because they can just press the button and get a new box to skip. And so they're like, "Skip, skip, skip. Oh, this one. Skip, skip, skip. Oh, that one."

1:09:56.8 EC: And so in the actual test, what we do is we throw in a few gray boxes. They've never seen them before, and they're somewhere in between the black and the white. And what we're interested in is using that as a measure of optimism. So, do they treat those intermediate boxes more like the positive one or more like the neutral one or the negative one? And what we found is that great apes, well, specifically bonobos, when they've listened to laughter, they're more likely to approach the gray boxes than when they've listened to a control sound. So I think that's one behavioral piece of evidence that I think suggests that listening to laughter leads to them expecting more positive things.

1:10:48.4 SC: Yeah. And there's no actual connection, between the laughter and whether or not there is something good in the gray box or not.

1:10:58.0 EC: Absolutely. So it's simply, it's kind of a laugh track. They listen to it for a while. It's not providing any useful information. We're interested in whether it's essentially inducing them, putting them into a better mood. And if they're in a better mood, we think that they should expect more positive things than if they're in a not... I don't want to say negative mood because we're not trying to put them in a bad mood. We're just not. We're playing a very neutral environmental water sound.

1:11:33.8 SC: In humans playing and humor are very closely associated with imagination and pretending. That sounds like something that would be hard to study in great apes, but is there anything that we can say about how much the quality of imagination is present when they're doing these things?

1:11:54.0 EC: Yeah. So one of the things that I think people have been very interested in for a long time is play signaling. The communicative signals that different species use to say, "Hey, now I'm playing." This is something that's been looked at a lot in dogs and in wolves. The play bow in dogs is the sort of classical play signal. And what's really interesting about that is it seems to be something that's very species typical. It's not learned. It's something that even young puppies will do. But it's not so much the production of it, it's the interpretation. So when a dog sees another dog play bow, they are less likely to respond to that other dog's pouncing and biting things in an aggressive manner. And so I think that this is an example of a kind of behavior that could be creating this frame. Where it's saying, "Hey, the thing that I'm about to do shouldn't be taken seriously." Now, is that pretense? Is that imagination? Not to the extent of a four-year-old saying this spoon is a scepter and I'm a princess. But it is still marking something as different from the the quotidian, right? Different from the the way this behavior would normally be interpreted. And I think that that's also happening in teasing. In that a lot of teasing is I'm doing something, but it shouldn't be sort of blown out of proportion. It shouldn't be... Like if I steal something from you and I'm teasing, it shouldn't be taken as seriously as if I steal something from you outside of a teasing interaction.

1:14:01.2 SC: Yeah, it's useful social knowledge for everyone to agree on that protocol.

1:14:06.5 EC: Exactly. I think it's really difficult. A fascinating area and I love thinking about these things. And from a theoretical standpoint, I think that's really where philosophy can play a big role is in saying, well, what would be the burden of proof? In a non-verbal way, how can we differentiate these things? Because that's the difficulty in looking at non-human animals is that you can't ask them, "Well, what did Fred mean when he did this behavior?" Or, "Why did you respond differently to Fred than you did to Larry when they came over and tried to bite you?" And in humans, even young children, you can say, "Oh, well, I know Fred was just playing." You can't do that in animals. And so we need to come up with creative ways of looking either at their spontaneous behaviors to see, well, how does the social situation, the social dynamics and the actions in the interaction differ? As well as developing biometric markers of emotion, where we're looking at things, ideally things that you can measure externally in free-ranging animals who are naturally interacting.

1:15:30.0 EC: So things like looking at skin temperature or looking at piloerection, the hair standing on end and things like that. There's some things where you can see behaviorally and externally that are direct responses to internal physiological processes. And I don't think there's going to be a single answer. I do think that these are complicated questions that really rely on sort of a multi-method approach, bringing together experiments where we're asking questions about what happens when you're presented with this stimulus together with observations of natural behavior together with studies of social preference. Do you prefer to interact with an individual who you've seen teasing someone else or not? One of the things that I think we haven't looked at yet, but just thinking of commonalities in humor and sort of the origins of human humor, one of the things that should be more funny is bringing down those in power. It's punching up the hierarchy. And I wouldn't be surprised if we saw that same thing in apes. Like if you say, watched a video where a dominant male slipped and fell in the mud versus or fell off a branch or something, as opposed to subordinate or juvenile. Or you watch a juvenile teasing an adult male versus adult male teasing a juvenile. And I think that it should be more interesting and potentially funnier to see the mighty brought low.

1:17:35.2 SC: I like that as an aspiration. I'm not sure empirically whether or not human beings, we even pass that test. We often like to punch down sometimes.

1:17:45.0 EC: Yeah. Well, I think that's one of the things where we would... In thinking about the comparative approach, the intuitions that we have about, "Oh, here's what humor is like in humans." We obviously have to ask those questions in humans as well to make sure that we're not simply building on our assumptions based on our own past experience.

1:18:10.7 SC: So speaking of the comparative approach, this leads us to sort of the last thing I wanted to ask, and it's the thing I often ask last these days in the podcast, which is something to do with AI. You're studying comparatively humans and other kinds of animals. So those are the data points we possibly have about natural intelligence here on Earth. And now we're faced with the prospect of a different intelligence that we make artificially. Maybe it's not there yet in terms of self-awareness or whatever, but is there anything that we're learning or suggesting or big questions we need to start asking that relate artificial intelligence to these natural ones that you're studying?

1:18:54.1 EC: Yeah, I think this is the potentially, to date myself, the $64,000 question. I was like, "Oh, I really feel like there should be... I should raise that now." So, as someone who has built a career out of studying communication and comparative cognition, the AI question is a fascinating but also irritating one. And I say irritating sort of tongue in cheek, but I think it's because especially LLMs. So, AI obviously can mean many, many different things, but for a lot of people, when it comes to asking questions about what does AI understand? What can it do? Can it pass X or Y psychological test? We're really talking about LLMs things like ChatGPT.

1:19:53.0 SC: At the moment. Right.

1:19:54.1 EC: Right, at the moment. And what is very... I think that we are really in a critical moment where it could go one of two ways. For a long time, language was seen as this defining line between humans and other animals. And even today, I think there are some kinds of questions that some scientists are not happy with people asking in a nonverbal way. So there's still debate about whether animals have things like episodic memory. Can they remember themselves in the past sort of from a first-person or first-animal perspective? And the way you ask questions about that in humans is through self-report. And there are behavioral tasks where you can ask whether animals can use their memory of something in the past to plan for the future. Some people argue that demonstrates episodic memory. Other people argue it doesn't. And so these arguments, I think are important ones, but are really difficult to resolve given that you can't ask most animals verbally. You can ask an LLM, and they will give you a response.

1:21:12.8 SC: An answer.

1:21:14.5 EC: Yeah, they will give you an answer, and it will often be the same answer that a human would give you. But they also have access to essentially all of human language, all of the sort of documented data points and responses and next turns, next word predictions of human history, the the current internet. And I think the big question is, does an LLM understand anything? And this is actually something that we've done some work on where we're advocating for applying more sort of guidelines and methods from animal communication or animal cognition to the study of AI and LLM specifically. Because I think a lot of times because of the methods that we use to study cognition and intelligence in humans when we can ask... When we get verbal report from humans, we assume that that tells us something about what's going on inside.

1:22:33.8 EC: I don't think the same should be true for an LLM. So we can't just take them at their word. And what we need to do is to develop much more sort of thoughtful, systematic ways of probing their levels of understanding by manipulating the way you ask questions, by changing the modality or the situation that you're posing to them. And I think that this is an area where comparative cognition researchers people working in animal behavior and animal cognition, and also developmental psychologists have a lot of experience. Because you can't just ask a one-year-old or a two-year-old, "Hey, why did you think this? Or, why did you do this?" You can't ask a magpie or a chimpanzee, "Why did you do this?" You can ask an LLM and it will give you an answer, but the question is, does that answer actually map onto why they did that? And coming back around to trusting language, this is a question even with human adults, the way that if you ask someone, "Hey, why did you give that answer?" The reason we think we did something might not actually be the reason we did it. And so even with human adults, you have to be some, I don't want to say skeptical, but you have to be somewhat careful in just taking verbal self-report as the final word on the matter. I think with many of these things where we're trying to understand what's going on under the hood, so to speak we really need to bring together... We really need to bring a multitude of approaches to bear on these questions.

1:24:27.9 SC: Do you think that they're doing that? Do you think that the LLM people are reaching out to psychologists and anthropologists and philosophers to get more insight on these things as much as they should?

1:24:39.3 EC: I think some of them are. I don't think any of them are as much as they should. I think this is really an area where I think we will see more collaboration and more development in the years to come. I think when people really hit walls and want to understand, "Hey, why are we getting this error? Why are they not doing the thing that we expected they would do because we thought they were doing the first thing the way humans do it?" And so I think it's going to take kind of... I don't think these are going to be problems that you can engineer your way out of just in terms of improving the algorithm. I think there are things that where you really have to try to find ways of assessing what the LLM is telling you about how it's doing the thing.

1:25:42.8 SC: Yeah, questions that we cannot engineer our way out of. Those are some of my favorite things. So I want to thank you for being on the podcast, but also, you're the scientist who has websites and resources and things that listeners can go poke at. So what should our listeners know that will be fun for them?

1:26:01.5 EC: Thanks so much for asking, Sean. So one of the things that I would love to get people to check out is that we're actually asking people to share their own experiences with animals teasing. And so this could be your dog or your cat. It could be, you watch two crows over a field. It could be something you saw at the zoo. It doesn't have to be... We're collecting any animal experiences of teasing that you might have experienced. And so the website there is observinganimals.org. And we actually have several different sort of public science studies up there, but there's one on teasing. And so if anyone listens to this and has reflections about animals that they've seen teasing, we would love to hear about them. And then more broadly, if people are interested in the work that we're doing at IU or that my lab is doing, we're starting a new center, a new research center at IU, which is called the Center for Possible Minds. So that's possibleminds.org. You can just check out the Center for Possible Minds.

1:27:13.4 EC: Yeah, we'd love to have folks reach out if they're interested in getting involved with any of the things that we're doing there. This is a new initiative launched by myself and Jacob Foster. And we also run the Diverse Intelligences Summer Institute, which again, you can find out more information through our center website. Yeah, thanks so much for asking and for having me on. It's been a real delight having a chance to talk to you about some of this work and to think about it together.

1:27:43.6 SC: You're going to be inundated with so many stories of people thinking their pets have done cute things. There's an endless supply of those. So, Erica Cartmill, thanks very much for being on the Mindscape podcast.

1:27:54.3 EC: Thanks again, Sean.