Transcript 4 | Tools for Thought, Artifacts from the History of Comparative Psychology with Dylan Kerr
Kate Armstrong
Okay, so I think that we can almost start. We have a lot of people joining us. Good morning, good evening, and good night, depending on where you are in the world.
It's my pleasure to introduce you to this Interspecies Internet lecture which is part of our public and scientific programming here at Interspecies Internet. My name is Kate and I have recently joined the organization to steward the public programming. I'm really excited that this lecture is one of the first that we've done in a while. We are really looking forward to launching more scientific and public programming in the near future. So we have a very interesting lecture today as part of the Interspecies Conversations series. So this series is really about bringing emergent ideas and research from the field to you all as a community and allowing a forum and a space to actually discuss this and get these ideas heard and kind of mixed with yours.
So the aim of today's session will be to hear a 30 to 40 minute talk from our speaker, Dylan Kerr, and then we'll have time for some questions and answers. And you can put the questions that you have either in the chat or you can hold them until after and we can have that discussion live. And I think before we start, and I pass it over to Diana maybe just to say thank you for joining us. I know that it's Memorial Day weekend for you in the US and also thank you to some of our trust for joining us as well. We have Vint here and I can't see anyone else, but I'm sure they're going to join us in due course. So, Diana, I pass it over to you to introduce Dylan. Please keep your microphones closed during the time that we're having the talk. And then of course, when we have the conversation, you're willing to open them. You're able to open them and I will moderate you. Great. So over to you, Diana.
Diana Reiss
Well, thank you so much, Kate, and good morning everybody.
And again, this is a great time if you're in the United States at least to start this holiday weekend. And today's lecture is really special for me. This is again a really special pleasure for me to welcome you all here today for this talk by Dylan Kerr. I'm the director of the Animal Behavior and Conservation Master's program and certificate graduate programs at the Hunter College in New York City. It's part of the City University system and I've really had the good fortune and pleasure to have served as Dylan's MA Thesis advisor. And today Dylan is going to be presenting some of the work he did for his thesis project. Dylan's a writer, an editor and a researcher based in Brooklyn, New York. And he holds a bachelor's degree in media studies from Vassar College. He also did a minor in religion, there, and also he now has a Master's degree in animal behavior and conservation from Hunter College.
Dylan is really one of the students that came into our program with a really unusual background. And again, I was really happy to. We try to get a lot of diversity in terms of the backgrounds of the students that come into our program. Dylan worked as an art reporter and an editor for Art Space magazine and other publications and he currently works as a copy editor for the New Yorker. He's written about animal translation technology, the history of live animals in modern and contemporary art, vegan art supplies artists use of object oriented ontology and the evolutionary basis of art making and appreciation. He's also conducted many interviews with artists, scientists, curators and more. He kind of speaks to our community then. At the same time that he started his work at the New Yorker, Dylan contacted me and he was really interested, more of an educational background in the world of animal studies, animal communication, animal intelligence and communication. And he applied and was accepted into our ABC program. So again, he was sort of doing double duty on this one. And the focus of what you're going to hear today is the focus of his thesis and it includes the history of comparative psychology and its relationship to the history of art, tracing the development of animal behavior and science over the past century. And in line with Dylan's curatorial writing interests and background, he originally intended to do his master's thesis as a walkthrough art exhibit. He had a whole thing planned where he actually found the space he was getting ready to go. And then like for so many of us, the pandemic hit and changed our plans. So I'm not going to tell you too much more. I'll leave that up to Dylan and he'll tell you the rest of the story in his talk about Tools for Thought, Artifacts from the History of Comparative Psychology. So it's all yours, Dylan.
[Historical Significance of Animal Behavior Research]
Dylan Keer
Thanks so much, Diana. I'm going to attempt to share my screen now. We'll see how this works.
Great. So like Diana said, my. Sorry, just a second.
My project is about the Apparatuses used in some of the seminal animal behavior research of the past century. And you can see on this slide here, many of you, I'm sure, are familiar with these experiments. You can see the original rat maze, the Skinner box, Yerkish lexigrams, and we'll get into a bunch of those today.
My argument here is that these artifacts are worth looking at in their own right, not just as a means to a scientific end, not just as part of the experiments that they're involved with, but as objects that have something to tell us about the history and the future of comparative psychology. As you'll see, they also have all kinds of connections to other areas of human culture, whether that's art, architecture, mythology, economics. And my position is that these commonalities are worth exploring, especially insofar as it might inspire future lines of research. But before we talk about those devices, I'd like to talk quickly about animal traps. Now, I know this might seem a little odd in this context. I imagine most of us are animal lovers. We're more interested in learning about animals than in trapping, killing, and eating them. But I think that these traps can help us look at experimental devices in a new way. And I'm inspired by the anthropologist Alfred Gell, who wrote in a 1996 essay, the question that you see here, and what he's getting at here, is our traps just utilitarian objects that show us how much we like to eat meat. Or could looking at these traps closely, the way we might with historical artifacts or works of art, let us in on new complexities about the traps?
So Gell goes on to argue that there's much more to read in a trap, that traps are kind of text. In the case of this rat trap, you can see if you do a sort of close reading on it, but the cultural background that informs its design, the lore and traditions about what kind of trap to make, how to set it, where to set it, why you would want to set it, that sort of thing. But also can see an idea about what the rat might think. There's a conception that the rat likes a certain kind of food that's attracted to dark, enclosed spaces. And so the design of the trap reflects both the human side, but also people's conceptions, what the animals might be thinking. These traps show complex interactions between the creator, that is, the trapper, and what we might call the audience, or the animal itself.
And the animal behavior is vitally important to the trap, because otherwise it won't work. If you don't know how a rat thinks, at least on Some level, a rat trap is going to be hard to design. So in this sense, Gell says these are texts on animal behaviors and that they embody a theory about how animals behave that's informed by both history and practice. Gell goes even further in this essay, though, and he argues that we might even think of these traps as a kind of artwork, a documentation of something like a performance between an animal and a human. Granted, that performance ends very badly for one of the participants. Gell writes that both traps and artworks are capable of synergizing and drawing meaning out of the other. They're not the same, but they're not entirely different or incommensurable either. So I think if we accept this idea that traps are maybe not a bona fide artwork, but are doing some of the same things that artworks might do, I think it's interesting to compare them, these traps, to proper works of art from human history and other forms of human culture to tease out their commonalities and their divergences. So you probably already see where I'm going with this. My main kind of thrust here is that devices from the history of comparative psychology also can be read as a type of text. They show both the history of the science that led to the experiment and the specific results of the experiment themselves. You need the device to do the experiment. You can see here Thorndike's box here, which we'll return to in a second. But also you can look at the traps as illustrating all kinds of connections between human culture, including art, but also religion, economics, architecture, and more.
These traps and devices are objects that are informed by human history as much as by the scientific theories that they illustrate. They also have interesting connections, even if they aren't directly informed by. And we'll explore all that in just a second.
Starting off here, finally, back to our devices. The rat maze, one of the most famous of all scientific apparatuses, has become almost synonymous with, or kind of shorthand for, experimental science itself. You see cartoons and stuff like that that reference rat mazes. But it's easy to forget that someone had to actually come up with the idea of putting a rat in the maze in the first place. That person was Willard Small. Now, he wasn't the first person to come up with the idea of a maze. Mazes and labyrinths have had this long history in human culture. And we see here on the write just one example, the Minotaur and the labyrinth. But almost every culture has some kind of labyrinth or maze iconography that they use. Whether that's a physical maze or not is A different question. But Small wasn't inspired, at least directly, by any of these. He had a kind of more prosaic inspiration, and that was Scribner's Magazine. In his letters, he describes reading a Scribner's Magazine story about kangaroo rats and seeing this image that you see here, which is a diagram of the rat's nest in the wild. So Small, he's working in 1900, 1901, right at the dawn of the experimental psychology. And so he starts to think, how can I replicate this in the lab? Like any good grad student, he talks to his advisor. And his advisor suggested building a physical maze. One, but two, basing it on the Hampton Court maze. The Hampton Court Maze, the oldest extant hedge maze. It's in England, and it was constructed around 1690, and it's part of Hampton Court palace, and it's now a tourist destination. You can go and walk the maze yourself. And just to emphasize this is a real physical hedge maze made out of plants. So Small and his advisor corrected the trapezoidal maze. And you can see the image on the left there is the image they used from the Encyclopedia Britannica. And in his letters, he actually says what edition of the encyclopedia they consulted. And this is an image from that edition. They corrected the trapezoidal maze into a more rectangular format, which is what you see on the right there, and constructed it at a rat scale out of plywood and wire. And thus the rat maze was born. Small found that the rat's performance in the maze improved over time. That is, it could start and finish faster with fewer errors. And for Small and others, this suggested a trial and error approach to learning, which amounts to a kind of proto behaviorism. He was working before behaviorism had been fully codified, but we can see how his experiments in Thorndikes, which we'll talk about in a second, lead into that behavior. Now, of course, this wasn't the last rat maze. As we said, it become a kind of shorthand for this sort of experimental science. And it's also a thriving research tradition. It's expanded in the modern era to water mazes, elevated platform mazes, tea mazes, and these kinds of wild, to me, maze shapes that you see here. These are just a few of the hundred examples. And of course, scientists are still using mazes to learn different things about animals. So from that magazine story, a whole research tradition has been sparked.
One of the most interesting examples of this for me is Tolman's experiments. And this is a way that a Kind of behaviorist or proto behaviorist means, that is, the original rat maze was used to get at what we would now think of as kind of cognitive psychology. He trained the rats in the maze on the left to get to the goal box, and you see that on the upper right hand corner. But then he put them in the right maze, the one with all the radiating paths. He found that the rats didn't just blindly memorize the movements to get to the box left, right, left, right. They actually chose the correct path among the radiating paths, suggesting that they had some mental image of where the goal was in space. Tolman, even back in 1948, called this cognitive map. And that's a result that couldn't easily be fit into the behaviors paradigm of his time in the 40s. And we'll see that again as we move through that. Results using older means sometimes generate new kinds of data that call for new kinds of theories.
[Comparative Psychology: Thorndike and Skinner]
My next example. We're back to Thorndike's famous box. He's credited even before Small as being one of the first properly experimental comparative psychologists. And I'm sure many people are familiar with this experiment. But the basics is put a hungry cat into this box repeatedly and tracked how long it took the animal to get out. Crucial points here. There's no obvious or easy way to escape. The cat had to perform a complicated series of maneuvers, whether that's pulling a string, lifting a latch, doing stuff like that, in order to escape. And it had no prior experience in the box. The first time it went in, it had to figure out how this mechanism worked by trial and error. And Thorndike found that the time in the box was drastically reduced once the cat landed on the solution. That is, it got faster and faster and faster getting out of the box the more times it got put in. For him, this was more evidence of the association between stimulus, that is getting put into the box. In response, figuring out how to escape the box. I include this quote on the right here, partially because I think it's funny, partially because I live in Brooklyn, was born in Yonkers, and partially because it really illustrates the historical period that Thorndike was operating in and what he was kind of coming up against. That last bit there let one dog find his way from Brooklyn to Yonkers, and the fact immediately becomes a circulating anecdote. That word anecdote is crucial. Thorndike was frustrated with the anecdotal approach to animal behavior that had really been at the forefront of the science since Darwin, and also with Romanes and others. Where story, kind of one off, individual stories about animals was how this sort of comparative psychology was done. Thorndike wanted to generate hard data for hard science analogous to chemistry or physics, which were also really coming into their own at this time, not just stories. The way he did this was through replicability. People often forget that that box which is now in the center, the one we saw in the last slide, is one of the 15 different boxes that Thorndike designed for this experiment. And this is all part of his effort to get away from anecdotes. One cat getting out of one box is kind of just a story. 15 cats getting out of 15 different boxes and doing it over and over again. That starts to get at what we now think of as scientific evidence.
I see this replicability, and here's the art connection as analogous in some ways to the modernist series which many people say Claude Monet sort of spearheaded, interestingly, at almost the exact same time that Thorndike was working. They're both in the second half of the 1890s. And you see here on the right Monet's grain stacks where he painted 15 different views of these grain stacks in all different weather conditions, times of the year, lighting, et cetera. And we can see a kind of connection there. One view of a grain stack is just an impression. Monet, of course, is an impressionist. 15 different views in different lighting, different times of the year starts to get at something deeper about the grain stack. The art historian John Klein writes of Monet's series paintings that the artist sought to regulate the ephemeral effects he recorded on the individual canvas with the rationalized structure and discipline of a system. And we can see a similar thing happening in Thorndike trying to get away from the one off and then to replicability and kind of generating data in this way. Thorndike's experiments and Monet's painting can be read as part of the same modernist impulse toward quantification and objectivity over what we might think of as a more romantic subjectivity. And I think this is an interesting example of how the pursuits of scientific and aesthetic truth, as different as those two things are, sometimes converge upon some of the same methods.
Like the mazes, the puzzle box also sparked its own research tradition, can read the puzzle box and make many other comparative psychology experiments as being part of this sort of research tradition. Sciences, of course, take inspiration from their predecessors in order to make new innovations in the field. This is one is chosen more or less at random. It's a puzzle box that requires tools to get into it. So, unlike Thorndikes, you're not trying to get out of the box. If you're the subject, you're trying to get the treat that's in the box. And this particular experiment looked at how keas and New Caledonian crows learned how to get into the box. Like I said, two of the options required using a tool. The object is sort of similar, kind of novel box that you have to figure out, but used for very different purposes. And I see here an analogy to the history of painting. You see the Monet and the Rothko there on the right, both using pigment on canvas. Not too much more complicated than that, but they're being used to very different ends and with very different effects. Nevertheless, I think we can see the lines of influence between both the experiments and these paintings, how they fit into a tradition.
Probably the most impactful descendant of Thorndike's box is, of course, the Skinner box, the infamous operant conditioning chamber. It's designed in simplest form, and there's been many, many, many more complicated forms that have been developed. And in simplest form, to provide a single stimulus and allow for only a single response. And that way it's similar to Thorndike's boxes in the maze. And that there's only one way to correctly solve the puzzle. In this case, the puzzle being if the light turns on, you press the lever. And it's part of this steady push to simplify and purify behavior into its simplest form and also to automate data reporting. We'll talk about that a bit more in a second. Those of you that recall the Behaviors manifesto, the stated goal of behaviorism is to control and predict behavior. And this device allows for exactly, and I would say only that there's no creativity, no choice or control on the animal's part. It's just one input in, one output out. Skinner critically adds something new to this box paradigm, and that's automation. Skinner was raised in the context of automated mechanized production. We can see the assembly line there, and we can see that influence in his work. The Skinner box was designed to automatically record the data, kind of leaving the scientists to go about his business, while the rat kind of produces the data on its own. We can see the similarities between the automated Skinner box, allowing for only one output in response to one input, and the automated assembly line, in which workers respond in one way to a single stimulus. As they work on the assembly line, one thing comes in front of them. They do their one little bit to the wheel or whatever it is, and then it moves on to the next. And they just repeat that over and over again. Ford found this was a much more efficient way of building a car. Than having one person do it from start to finish. And Skinner similarly found this is the most efficient way to predict and control behavior. He turned his rats and his pigeons, his subjects, into a kind of factory worker. The product being this behavioral data. Skinner would go on to claim that this operant process underlies all behavior, including the complexities of human behavior. And people's discomfort of this is part of what leads to the cognitive revolution in the late 50s, early 60s.
Next experiment is the seminal mirror self recognition experiments by Gallup. These are usually thought of as one of the kind of premier ideal examples of a cognitive comparative psychology experiment. That's dealing with ideas of internal information processing in the animals. The actual story is a little more complicated and I think a little more interesting than that. We know from the science that's been done since Gallup. That most animals do not appear to recognize themselves in mirrors. But humans have a long history of fascination with their own reflection. Just a few examples here. The ancient Egyptian mirror. On the top right we see the myth of Narcissus on the left. And many, many paintings and works of art employ mirrors. Even the bean in Chicago can be thought of as something like that. Gallop's mirrors, though, comes out of his engagement with a sort of behaviorist psychology. As a doctoral student, he used mirrors as a cheap and reliable reward. Analogous to the food tree in a Skinner box. Monkeys would open the door shown in the figure on the left. More often if it contained a mirror, Though they would eventually grow habituated to it. And that stimulus didn't work as well for the monkeys. When he tried the same kind of experiment with chimpanzees, Though, he noticed that their behavior toward the mirror was different to that of the monkeys. Though they initially responded with social behaviors like the monkey thinking that the mirror image was a conspecific. Repeated exposure found them turning to the sorts of contingencies, testing and mirror directed behaviors. That we would associate with people in front of the mirrors. Checking their nose, checking their tongue, looking at parts of the parts of their body that they couldn't otherwise access. We see examples of that on the right here. This observation led him to test the now famous mirror self recognition experiment. Which employed the mark test, in which the animals were sedated, Marked with either a clear or visible dye, and then presented with the mirror. And he found that the animals with the visible dye would check out the mark in the mirror. Much in the way we would if you saw a little mark on your face, like, hey, what's that on my face? Like Tolman's mazes. These results provide another example of experimental evidence pointing towards more complex explanations than behaviorist theory could explain. These mirrors were able to reflect the internal processes happening inside the chimpanzee outwards.
[Animal Cognition and Communication studies]
Now this, like the others, sparked all kinds of different further experiments. One of the most interesting, if kind of questionable ones, comes from Skinner himself, who sought to refute Gallup's conclusions by showing that a pigeon could be trained to do the exact same thing through an operant method. His position was this is just operant behavior all over again. That's the only explanation for behavior. But I would argue, and many others have, this actually served to prove the cognitive conclusions that Gallup drew that seem to be more reasonable in his case. With the chimpanzees, there was no training needed. And there therefore seemed to be some form of information processing happening not through training, but in the mind of the animal. This was later repeated in species as disparate as dolphins and elephants. I think some of listeners are very familiar with these experiments, but the same technique was also used in other species that have failed the test. For me, this is almost just as interesting as those that pass because it suggests that there's some neural machinery that's needed for something like this. You need a certain kind of brain in order to be able to process something like a reflection as yourself. These negative results are actually examples of the strength of this research paradigm and a challenge to behaviorism. Not all species have the same mental abilities.
One of the most interesting in my mind, if also somewhat questionable growth of this, is the olfactory mirror. Alexander Horowitz reason that dogs use their sense of smell more like apes use their eyes. So it might be reasonable to test recognition using that sensory modality. Here, the mark is a modified version of the dog's own scent. And Horowitz found that they attended to the smell more than their own unmodified order or that of another dog, suggesting that they were attending to the mark and way similar to that of Gallup's chimpanzees. Here we see the idea of a mirror and of reflection itself abstracted almost beyond recognition. Though some people, including Gallup himself and others, may disagree with Horowitz's conclusion, the experiment itself illustrates how something as simple as a mirror or something as commonplace as a mirror can be transformed to produce very different kinds of scientific results.
The last big example that I'd like to discuss are the symbolic language experiments of the second half of the 20th century. The early 20th century, and really through the end, saw a number of experiments that were centered on teaching animals some form of human language, whether that's spoken language in the very early years or sign language. And they did this with varying degrees of success. Of course, those languages are themselves human cultural creations that have inspired science and surely are worth time of their own. But here I'm focusing on physical apparatuses. So I thought Premack’s symbolic language is a good place to start. He constructed a symbolic language, and you see some examples here for Sarah the chimpanzee, Based around magnetized plastic symbols attached to a board, this physical written language had the benefit of being more objective than a spoken or signed language in the sense that what is put on the board is what the statement is. You can take a photo of it. There's no interpretation involved. And in addition to that was more readily controllable by the researchers because they could limit the number of plastic pieces that Sarah had access to. So you could be very sure if she put the symbol for apple on the board, that was what she meant. If we recall Gell's idea that traps can be regarded as texts on animal behavior, here we can see animals are starting to be the ones that are generating the text. And we'll see that in our other examples, too. You may notice that the symbols used by Sarah don't resemble the objects or actions they represent. You can see the apple is a blue triangle, banana is a red square. These images don't correspond to the objects that they're referring to. This was a conscious decision by the researchers in order to focus the attention on what mattered to them. Not matching pictures to objects, but actually seeing if the chimpanzee could understand and use the idea behind the words in a more or less human mode. When we say apple, that doesn't resemble anything to do with an apple. But we're getting at the idea of an apple, the appleness of an apple. In attempting to get at ideas in their purest form. It's another example of both artists and scientists coming to sort of similar means. The early modernists, looking to purify their art form, get away from representation, saw the advantage of abandoning simple representation in favor of abstraction. And we can see a Kandinsky painting on the right there that I think has some similarities to the symbols that Sarah used on the left. Both groups, both the scientists and the artists, were in search of a means by which to express pure ideas, unhindered by the messy contingencies of representation. And both found abstraction to be an effective solution. This quote from the art critic Alfred Barr could easily apply to Premack or the other scientists that I'm going to discuss now: “Since resemblance in nature is at best superfluous and at worst distracting, might as well be eliminated”. And that's exactly what these experiments did.
Similar, somewhat simplified approach was taken by Diana Reiss, my graduate advisor, and her experiments with vocal mimicry. And we see on the left there her keyboard, and on the right, these works by Malevich. The keyboard work used two modalities, sound and a symbol. In order for dolphins to access a toy, so they'd have to press the keyboard, it would play a sound, and they'd get a toy or a rub corresponding to the symbol. This experiment not only found that they used the keyboards effectively, they figured out how to use them. But they also began to spontaneously mimic the artificial whistles in appropriate context. So they whistle the ball, whistle before and after getting a ball while playing with it. As with the other experiments, I can't help but thinking of the abstraction used in early modernism. In this case, as I said Malevich, in both cases, a resemblance to real objects would have gotten in the way of the results they were looking for. This is not to say that the scientists and artists had exactly the same goals. The artists explicitly thought to excise any reference to real world objects, whereas the scientists use abstractions as symbols for precisely those real world objects. But the graphical and theoretical similarities for me bring up interesting questions about the function of abstraction and pursuit of something like truth, and whether that's scientific or aesthetic.
Last example here is Yerkish, I think one of the most complex of these artificial languages that have been developed. And that complexity allows for even more flexibility. The basic idea is that the chimpanzees and later bonobos would use the lexigrams, which are also non-iconic, to communicate first with a machine that would provide treats. And we see that on the right there. But then they realized that in order to really get the animals to use the language, they had to have a human interlocutor or social interlocutor. And that would suggest that language use is an inherently social activity, which I think makes a lot of sense. To make a long story short, they found that the chimpanzees and bonobos could indeed use this language and can do so in incredibly complex ways. They found they could respond to novel questions, combine words in meaningful ways, describe actions in the future and objects not immediately visible. Some of the things that we thought were solely capacities of human language. This language eventually began to resemble what I would argue is humanity's greatest cultural artifact, and that's language itself. We can see here the many, many, many symbols that are been used in Yerikish. And we're starting here now to get to the some of the complexity of what I'm doing right now. Talking, using words, talking about things that are displaced in time and space in complex sort of combinatory ways.
[Innovative Experiments in Comparative Psychology]
So those are the main experiences I wanted to talk about. But there's several others that are sort of one offs that I think are just interesting to mention while we're talking about connections between art and science. One of my favorites is this, the Aesop's Fable experiments. That's based on Aesop's fable of the crow in the pitcher, where a thirsty crow finds a pitcher of water where the water level is too low for it to reach the water. So the clever crow realizes it can drop rocks in the water, raise the water level and thus get a drink. And so researchers decide to see if they could put this to the test in the real world. And they did so here with New Caledonian crows. Now, the crows had to be taught to drop the rocks into the tubes that you see there. So there's some questions as to how much insight they were really employing. But they found that they did use the rocks to raise the water level and get the treat, whereas they didn't use the rocks in the sand one that you see on the right hand side of that picture, because the displacement with sand obviously wouldn't work the same way with the rocks. To the researchers, this suggested some kind of insight into how the physics of those two substances function.
Another example, from a little bit further back is Harlow's controversial experiments on maternal deprivation. And we can see here the echoes of other maternal figures from across cultures and really from across millennia. That image on the left is from pre dynastic Egypt. The image on the right is a more contemporary rendering of the Virgin Mary. And we see there's some kind of echoes between these two things. The open, welcoming arms, the softness of the body. And indeed the results of this experiment that the babies preferred the soft mother to the wire mother, even if they can only get nourishment from the wire mother, show that the form of this maternal figure is probably more important for the monkey, or at least as important than the material needs. So I think looking at these representations of motherhood up against Harlow's constructions provide for some interesting sort of resonances. Another one is portraiture, rendering a recognizable image of another person, which of course, is a very deep history. The examples here are from Roman Egyptian mummies. These are portraits that are on the outside of the mummies. I include them because I think they're very beautiful. And these are not by any means the earliest example of portraiture. There's much, much deeper. But in recent years, there have also been experiments that use a kind of animal portraiture to see how animals recognize other individuals, both conspecifics and also members of other species. And we see in the center there, the sheep could recognize fellow sheep, but also individual humans and dogs. One of my favorite recent studies in this mode found that chimps can also do this with portraits of other female chimpanzees rear ends. And actually that they were better at figuring out whose rear end is who than they were at recognizing faces. Suggesting to me that what's salient about a portrait may be very species dependent. And if we recall Horowitz's olfactory mirror, we think thinking about how different animals respond to different sorts of stimuli and what they respond to. The last one that I want to talk about is a recent experiment that I just came across this week, and I was too good not to include it. Steve north has designed a pair of robotic horse ears. Ear movements are important means of communication for horses and for many other animals. So north sought to be able to replicate some of those movements artificially using these animatronic horse ears. What's interesting in the context of this discussion is where he got this idea from, which is from anime, specifically the idea of Kemonomi, in which characters possess various animal characteristics like ears and tails. He named his project Umamimi, which means horse ears, after this Japanese term. So I think this is a fun example and also shows that there's really no limit to where the inspiration for scientific experiments can come from.
So just to conclude here, Gell opens his essay with the question, do animal traps and their bare decontextualized presence tell us no more than that human beings like to consume animal flesh? I think here, if we replace traps with experimental apparatuses and ask whether they reveal anything more than the results of the experiment, I think the answer is yes. Looking at these devices does tell us something far more than just what the results of the experiment is. We've seen how these comparative psychology experiments have been inspired by many different areas of human culture and how these ideas live on as research traditions that continue to influence the field. My hope here for this audience is that this project will help encourage scientists to look for inspiration from all areas of culture, our species has developed this almost incomprehensibly rich store of various communication activities, strategies, forms, ideas. I hope that as the brilliant scientists that are listening to this continue to develop their ideas, they'll continue looking beyond the labor and using this vast cultural resource we have to further their research. That's that. Thanks very much.
Vint Cerf
Wow. Fantastic.
Kate Armstrong
Awesome. Thank you. Dylan, do you want to. If you stop sharing your screen, that's perfect. Wow. I'm sure everybody has a lot of hands clapping and I'm sure that that means that there might be a few questions of the fantastic intervention from Dyl. Diana, do you have any feedback or thoughts you wanted to share first?
Diana Reiss
I think I'll let other people go for it.
Kate Armstrong
Okay. Any questions from the house or feedback? Go for it. Vint, please.
Vint Cerf
First of all, I can't believe you managed to cram that much into like 35 minutes. That's astonishing. Just thinking a little bit about the animal mazes and ability to learn. Do we know very much about serial learning and remembering things in serial form? Because it strikes me that a lot of people remember things in sequence better than they do out of order. It's sort of like picking up a song and getting the next words because you got the first few. Is there is any indication that that's an important part of our ability to remember and repeat?
Dylan Kerr
So I'm not at all an expert on this, but I do know there's been many experiments that, that deal with almost exactly that question. And it does seem like sequential learning is an important part of how we move through the world. You know, take a left here and a right there and then you're at work. And Shirley is also with animals. I can't bring a specific experiment to mind right now, but I know this has been looked at. That's kind of part of what Tolman was looking at in those cognitive mazes. Although the results that he found actually showed that there was more than just remembering. Left, right, left, right.
Vint Cerf
So Baruch has his hand up.
Kate Armstrong
We can't hear you unmute. Do you want to unmute?
Baruch Fischhoff
There's probably a lesson in learning there as well. I would say that if you could read my lips, it was really enjoyable and thought provoking it what it reminded me of. So I have a PhD in psychology, but I did Math as an undergrad and I was in a graduate program that had very few courses, so my knowledge is quite spotty. But I did take one course in comparative psychology and your talk reminded me of a paper that we read there 1970 Psychological Bulletin paper by Elliot Valenstein and others, in which, as I remember, I fetched the paper while you were. While you were speaking. So I may have misread this 50 years ago and misremembered. But he concluded that at that time there had. As I encoded it, people believe that you could stimulate different parts of, I think, the hippocampus and evoke different kinds of behavior. His reconstruction was that the animals were put in such a stimulus in such a deprived environment that all they could do was things that could be recorded as that behavior. If you thought you had the thirst center and the animal was grooming, you didn't see it, right?
Dylan Kerr
Interesting.
Baruch Fischhoff
An animal naturally would go and eat rather than drink. You just. You had no opportunity to see it. And eventually the animal would do what you were looking for. And they did a reconstruction that there are areas of the hypothalamus that weren't connected as far as I knew, that produced different behavior and the other way around. And so in my own work, I was very influential for me, even if I misinterpret it. And in my own work, I studied decision making. And one of the things that I find in our field is that we set up experiments in which we determine what's the right way, that what's the right behavior. And if people do something else, we assume that they're irrational rather than rational, rather than like, listening to them and seeing whether they actually have some other goal. So thank you for the talk.
Dylan Kerr
Well, thank you very much for that. And you're getting at a really interesting point that I didn't focus on here. But I do talk about more in the written thesis itself, which is the design of these experiments with animals. Now, not people, they're designed to limit the types of behavior that can be shown. Because you're looking to see how they move through space, how they get out of a box, how they associate a light with pressing a lever. So the whole point of them, especially the earlier behaviorist experiments, was to reduce the amount of behavior that is sort of acceptable from the animal. And what I think we see as we get into the cognitive work later in the century, is that the more choice and control that the animals have, the more interesting behaviors that they can see. So that balance between limiting behavior so you can measure it, important part of the science, but then also allowing for enough sort of creativity or more broadly like unexpected behavior, is in some ways is how this. The science has moved forward. And the mirror experiments are a great example of that. So thank you Very much.
Kate Armstrong
Vince, did you want to ask about something that you put in the chat?
Dylan Kerr
I just read that. I think the answer is no.
Kate Armstrong
Steve, go ahead. If you have a question you wanted to raise.
Steve Crocker
Thank you. So I haven't studied as much psychology as most of you, but I do remember vividly sitting in an introductory psych course. And one of the requirements was that we subject ourselves as subjects to the grad students who were running experiments. And one of them was to predict whether or not they light was going to go on to the left or the right. And I found myself working very hard at trying to make that prediction by remembering what all the past responses were. And I was a little nonplussed when I learned later that it was completely random and that it was nine to one on one side and that I could have optimized my score by simply pushing the lever all the way to the side that was the most dominant. I would have gotten a higher score. But in fact, my statistics matched exactly what was predicted, which is 9 to 1. That's not what was going on in my head. However, what was going on in my head was I was trying to build a pattern, a mathematical model, if you will, of what that sequence was going to be. It didn't work. But I have to believe that at some level, even in other species, there's a degree of modeling going on that is sorting out what the possibilities are and trying to do a bit of matching in addition to just recording what the past behavior was. Any comments?
Dylan Kerr
I think that's absolutely part of what's going on in these cognitive experiments, pattern recognition type of thing going on. Whether or not there is in fact a pattern human psychology, which is not at all my area of expertise, really furnishes some interesting results of this. But the example that comes to mind is Skinner's work on superstition. So he found that pigeons would first they would start, they would, you know, press the lever when the light went on and get their treat. But if there was more latency between pressing the lever and getting the treat, they started to do all these weird things like they preen the feathers on their left side and then their right side. And the longer the latency was, the more of these kind of strange, what Skinner called superstitious behaviors would come and come in into being. And Skinner's idea was that the pigeon doesn't realize what of its actions is resulting in the tree, right? So it's only the lever, and it'll take 10 seconds to get the tree or whatever it is. But the pigeon's not sure if it's. Is it the lever? Is it pruning on the left? Is it pruning on the right? Is it scratching my toe? Is it doing this? And so it starts to perform these superstitious actions, kind of searching for what this pattern for getting rewarded is. So I don't know if that really addresses what you're getting at, but I think that pattern recognition is definitely part of it.
Steve Crocker
It does in part. And also makes me recall the lecture on partial reinforcement where a closely related experiment was putting a bunch of chickens in a pen and just randomly dropping food. And we come back in the morning, each one of them has developed its own model and they're going through various dances and so forth, none of of which are directly related. But they worked part of the time.
Dylan Kerr
Absolutely.
Steve Crocker
Yeah. And as you can see, I have not yet figured out, despite trying how to separate that background picture that I wanted to display when I'm not versus there. So there's a complete failure in learning of a totally constructed system here.
Dylan Kerr
I think it's kind of cool, for what it's worth.
Steve Crocker
Well, thank you, but I'll stop the video now.
Vint Cerf
Sorry, it's Vint again. I don't see any of hands raised, so I'll grab the open microphone again. Two things. First of all, in the chat, I've always been amazed at how frenetic dogs are when they spot another dog. And then I was trying to figure out, well, how do they know it's another dog? Because they all look different. I mean, the breeds are so dramatically different, but dogs seem to be able to figure out it's another dog in spite of the fact they all look very different. Is there any light you can shed on that?
Dylan Kerr
Yeah, it seems like Diana has an answer. I don't know that much about this question.
Diana Reiss
Yeah, there's some very nice work. I'm forgetting the person who did it, unfortunately, but not too long ago, as Dylan had mentioned, there have been many studies done on birds being able to recognize differences in other. Between different bird faces, you know, match they use. Often we use a match to sample paradigm. But that. And it's been done with many animals, including dogs. And the study that. It's funny because I use it in my class. I just can't remember the author. Somebody else pop in if you remember it. But where they trained dogs to recognize dogs out of. From other animals and then recognize dogs when the species were so diverse, from Chihuahuas, you know, from dogs that barely had any fur, to big, you know, big fluffy dogs. And it's remarkable. There seems to be the concept of dogginess or doggish. So it's quite interesting. Yeah, I think what's. What's remarkable is that, you know, we've gone from this time of not thinking that animals are even thinking creatures to finding this richness in the ways that they think, the ways that they perceive. And you know, how often this is a. And this is just echoing Dylan, what Dylan's talking about, that, you know, our science and where we go is often a reflection of how we're thinking about others. You know, we wouldn't even conceive of doing certain kinds of experiments many years ago because we didn't think animals were capable of things. So it's exciting time.
Dylan Kerr
One other quick thing I'll add to the dogs is I did read a study that came out in the past few years about how things like cropping dogs tails or dogs that have very different ear or facial kind of layouts as a result of selective breeding, might be limited in their ability to communicate with other dogs because facial expressions, ear movements, tail movements are so vital to their communication. So that's different from recognizing dogs, which is what you asked about in Diana addressed. But I think that is an interesting and potentially deleterious consequence of kind of the line breeding of dogs. I'm thinking of French bulldogs, things like that.
Diana Reiss
I just was going to make. Oh, sorry. Go ahead, Vin. Did you have. No, I think somebody. I see another hand. Steve's hand is up.
Steve
I think Vince seems to be ahead of me, but I'll just do what I have mine. From a utility point of view, it seems to me there are three things that every species is interested in. One, from a mating point of view, they better know whether somebody is close enough in their species to mate with, either identical or very, very close. Second is a cannibalism issue, I think, in terms of whether or not they want to attack somebody in the same species or not. And then even more closely, whether or not there's a familial relationship, whether or not it belongs to this family or to some other family. I don't know any more about that. But those would be the motivations from a. You know, if you look at it from a utility point of view, usefulness, what needs to be learned, what needs to be acquired, what skills and so forth, either built in or learned in some fashion. The dog question is really, really very interesting. How do two dogs know each other? And they do. I mean, you walk down the street and they know each other. Directly completely different from toddlers of the same size, for example.
Vint Cerf
I find it really astonishing how frantic dogs can be when they discover another dog. They're quietly walking down the street and another dog shows up some yards away. And I don't know whether it's olfactory response or visual or both, but they go nuts. And it's only if it's another dog. I had one other. The chicken thing made me think of ritual behavior in general. And we see this in programming where you are introduced to a new programming lab way back in the early days in the 60s when we use punch cards and if you were going to run a program, they would hand you this deck of cards and they said stick this on the front of your program. You had a bunch of job control language from IBM for example. And then when it didn't work, somebody figured out why and stick another card on the deck and eventually you get this big stack of cards. And nobody knew why it was important to stick them on the front. All they knew is that that was how you got your programs to run. And so it became a kind of ritual programming which I thought was pretty weird. Now this one other thing that I just saw, a video today of TED Talk that was done in 2015 by a guy who was experimenting with alternative ways of delivering signals to the human brain instead of our normal senses. He was using touch, but it was activated by sound. So he had an array of sensors that essentially touched his back and in response to sound. And eventually people who were deaf wearing this vest were able to distinguish words. The brain was essentially eventually able to make sense out of repeated patterns. And of course we take advantage of that in machine learning all over the place because we try to adjust all the various parameters in order to see if we can generate a correct response. So I'm beginning to think that the brain is this unbelievably general purpose pattern matching system. And a great deal of what we're seeing, all the various behaviors that we're trying to discover are a consequence of this very general purpose system discovering.
Dylan Kerr
Totally. Yeah. And that that device that you just described, my like mind is turning thinking about the applications for, for the animal research stuff that I'm interested in. I feel like there's someone please come up with that experiment. It's not going to be me.
Vint Cerf
The guy's name is Eagleman and he's at Stanford University and I just sent him an email because after seeing this seven year old TED talk, I want to know what happened in the last seven years to how well does the brain work? So if you look up Eagleman Stanford.
Kate Armstrong
That's a great example. Bill, I'm going to come to you in one second. I just wanted. Maisie, you had a very interesting question in the chat, and I was wondering if you wanted to bring it to Dylan, because I think this is. I was also very inspired by this example. I don't know if you would like to read it or you would like me to do that for you.
Maisie
No, that's fine. Oh. Oh, sorry. Turn myself on again. Yeah, I just. Thank you very much for a very interesting talk. Dylan, I just wondered if you could speak any more about any artists who were particularly inspired by animal behavior. I was intrigued by. And I was just looking up the Mimi, sorry if I've not pronounced that correctly, robotic horses. And I just wondered if you could say any more from your research between those interconnections about artists who were inspired by animal behavior.
Dylan Kerr
Yeah, so this is a great question. I'll say off the bat that my research here is focused on more on the science side than the art side. So it kind of is unidirectional in that way. I wasn't looking so much at how the science went on to influence the art more the other way around. I know there was a fellow British artist, I believe, who constructed a artificial sheep suit. You guys may have seen headlines about this. If you Google artist sheep suit, it'll. It'll pop up. I forget the guy's name.
Maisie
Not goat man. Not.
Dylan Kerr
Yeah, yeah, yeah, yeah. Sorry. I said sheep. Goats. Very different. So that's one example. There's, like I say, I have this background in contemporary art, and I know there's so many instances of people, in some cases, drawing directly from, like, the Yerkish lexigrams. I know there was a. An artist who had a show that. Where she had painted the Yerkish lexigrams in different ways, and I think it was the same woman who she made. So I'm sure most of us are familiar. Chimpanzees use tools to get at termites and things like that. So she had gotten the actual sticks that chimpanzees had used and then cast them in silver to kind of like silverware. Is sort of the idea that these are like eating utensils. It kills me that I can't remember a name. But if you. If you look that stuff up, you'll see. And I saw Irene put in the chat music inspired by birdsong. I mean, I think like so much of human culture is inspired by, if not scientific, animal behavior than observations of animals. And we see this going back to, you know, cave paintings, where they're very concerned with animal life and including, you know, breeding, hunting, stuff like that. So I'm kind of stumbling my way through this because I haven't done a ton of research on this. But to answer your question, there absolutely are so many examples out there. I unfortunately don't know all of them.
Maisie
Thank you. That's very interesting.
Kate Armstrong
Bill. And Bill, who are you with? You had another.
Bill
Yes, I'd like to introduce everyone to Atka. Well, if I can. Here she is. As you can see, Atka has no eyes. She's a Siberian husky who's been completely blind since the age of 5 months. And Atka is normal in every way. When we walk down the street and meet other dogs, she does all the completely normal behavior that you would expect. She does a playbow. She greets another dog appropriately. They just decide whether or not they like one another. If they like one another, they play, they break off. They do a play bow, they play again. One thing I've realized with Atka is that we humans are so key to visual stimuli. Our language, everything we do is centered around vision that we can't even conceive of the world that Atka lives in. She's completely normal in every way. Not having eyes has not hindered her behavior. It's not hindered her socialization. She's got a very high prey drive. She's caught a bird. Okay. I can't come anywhere near close to catching a bird. I mean, so I think we really need to open up our minds to what it means to perceive the world, because she perceives the world in ways that we can't even conceive of. And if you'd like to know more about Atko or her story, there's a video on YouTube called Boundless Dogs. Atka's story. And we were out camping at one point, and one of the people in the campground was so amazed at ability to navigate through the world that she even made a video about her. And, I mean, just to give you an example, when we go hiking with Atka, she's 15ft out in front of me, goes down the trail, never bumps into anything. When she comes to a log, she jumps over it or jumps on top of it. How does she know how big the log is that she needs to get over it? Yet she does it with 100% accuracy. Thank you.
Diana Reiss
Yeah, thanks, Bill. I Just wanted to interject. It's really remarkable because when just what you said then and you said Etka was born without eyes. Right. This was.
Bill
She was not. She actually was born with both eyes. She had glaucoma. And then at the age of 5 months, she had both eyes removed and we adopted her at the age of seven months. And I thought, I've made a horrible mistake. I'm taking in this dog. She's going to be a burden. I'm so active. This blind dog can't possibly keep up with my active lifestyle of hiking, etc. She's the first one down the trail. Nothing stops her. And it's really remarkable to watch.
Diana Reiss
Yeah. I just wondering when you said that if she had been born blind because you said she was doing play bows and that would suggest that play bows might be more hardwired, but she may have learned that earlier on. So that was the question I had. But I also know a dog who's totally blind and has been so for so many years. And when I first saw the dog, I could not believe the dog was blind. The dog was as active as any of the other dogs, if not more so. So it's fascinating. Thank you so much for sharing that with us.
Kate Armstrong
And Bill, if you are part of the Slack group, and if not, I will be put the slack in the chat. It would be fantastic to see the video that you were, that you were discussing. And also I think, Inga, you had a great conference that you're hosting, which would also be fantastic to have in, in the Slack. I don't know if you wanted to have a special word about that, Inga. It looks very interesting.
Inga Hamilton
Hi.
Yes. And thank you for talking about glued ears and docktails. This is our. One of our rescue pups and she's got both. So it's interesting to know she might have communication problems with other dogs for that. But I'm probably okay. I'm a sculptor and I'm doing my PhD in the things that animals make and their sculptural intelligence. And a group of us doing our PhDs in philosophy, archaeology, psychology are putting on a conference on animal and human tool use and technology, bringing together all the different disciplines. So it's, it's an in person one. It's in Durham on the 6th June in the UK and we have a couple of keynote speakers, one of whom is Lambros Malefusis. If I say that right and I'm not even going to attempt, I think it's Catherine's name. She's a Chimpanzee, behavioralist as well. So all welcome. You just need to register beforehand. It's free. You get a good, good vegan lunch. And I should be talking to people about your work as well. Thank you, Dylan. It's really interesting.
Dylan Kerr
So very cool. Great.
´
Kate Armstrong
Thanks so much, Inga. I think, I think that the community will definitely agree that today's been a real interesting and very complementary discussion to the work that you've been doing throughout the last discussions. And I think that it brings a different perspective to what the interspecies Internet is really about. I personally have been absolutely fascinated. There are so many fantastic links in the chat and if anybody does want to share those, the slack is a really good place to do this because we have different, different channels there and you can share any, any updates or any interesting information you come across. Diana Vint, do you have any closing words? And Dylan, maybe you as well.
Diana Reiss
Well, I was going to, I was going to thank Dylan. I was really thrilled when I saw his master's presentation. I learned a lot and I think that for me, I have an art background as well, before I went into science. And I was always fascinated about how scientific, scientific. I'm losing my language, how scientific thinking can influence art and vice versa. And I remember many, many years ago I was thinking about how Duchamp's work as an artist with Nude Descending a Staircase was happening at the time of relativity. And, you know, it's this mutual influence I find just fascinating. So I thank you again, Dylan, for enlightening us about some of these ideas. I think there's a lot more there. I hope you write a book about this. I keep on saying write something. It'd be a really fascinating book and you're a wonderful writer. So thank you for sharing that with us today.
Dylan Kerr
Thank you so much, Diana. And thanks so much for, to you and to everyone else as part of Interspecies IO for inviting me. This was really, really fun and great questions all around and definitely get in touch if you have more ideas about this kind of stuff. I love, love talking about it.
Kate Armstrong
Great. Thank you, Dylan. Thank you, Diana. I think we will close it there for this session and please keep an eye on the website, the social media, as I said, the Slack. Because we'll be announcing our next lecture programming and also some other programs coming up to bring the scientific and public programs together. So we really hope that we'll see all of you there. And please, if. If you don't also have a subscription to our newsletter, you can also receive information via the newsletter. So it's been an absolute pleasure. I hope those in the US Enjoy the Memorial Day weekend and the nice vacation that you're going to have with everybody else. I'm sure you'll enjoy Saturday and Sunday as well. Thank you very much for joining us, and we'll see you next time.
Diana Reiss
Thanks everyone.
Dylan Kerr
See you soon.
Vint Cerf
Bye for now.
