2001 LECTURE SERIES

Wild Minds: What Animals Really Think

Dr. Marc D. Hauser
Department of Psychology
Harvard University
March 27, 2001

Well that was a wonderful introduction. I'm really delighted to be here. I think it's a great thing that Jim's put together here because it's not often enough, I think, that scientists are really put in the position to share their ideas with the public, who after all, does a lot to fund our work and we should be responsible for explaining things at a level that everybody can understand.

When I was an undergraduate, I had the opportunity to go and do some research at a tourist spot in Florida called Monkey Jungle. And the slogan for Monkey Jungle is "where the humans are encaged and the monkeys run wild." Because I didn't have much money at the time, I decided to take on a side job which involved feeding the animals in the cages and cleaning up after them. And one day while I was cleaning around the Spider Monkey's cage, which is this female right here, I noticed that she was looking very intensely at me and I didn't think it was my raking style, so I put the rake down and came close to the cage and she approached as well, and as we got very, very close to each other, she looked me in the eyes, reached both arms through, wrapped them around my neck, looked into my eyes and cooed. Now this is unusual for a couple of reasons: one, I didn't think that I was, you know, that deserving. But more importantly, this was not a pet; this was a wild caught Spider Monkey from South America, and yet she had this peculiar interest in me. Now what happened next is even better. Her mate came over and sat next to her. She took her arms away, slapped him and came right back into position. Now since I was only about 19 at the time, that was sort of a pinnacle in my romantic career and I decided I was going to start keeping her very close to myself, wondering, however, what the heck was she thinking, what was she feeling, what was he feeling -- after all, he'd gotten dismissed by his own significant other. I, the guy without any hair, won. But this really was in one sense kind of what got me going. I was an undergraduate at the time studying animal behavior with a wonderful teacher, Doug Handlen, and it was a time when the field of Socio Biology, the evolutionary interpretation of behavior, was beginning to crest and I was really interested in what were the similarities and differences between animals, including human animals. And this really sort of got me going in thinking about how to think about animals.

Now this is a topic which hits the press all the time. We find titles in covers like in Time Magazine, Can Animals Think -- not Yeltsin's War to Survive, but Can Animals Think. Yes, and maybe even lie and play politics. New studies show how clever they can be and shed light on human evolution. So here is a clan that, yes of course they can think and of course we think, as Tom suggested in the introduction, of course we think our pets can think. Well, if we simply grant that then the question is fine, but what are they thinking about? And if you're interested in animals and interested in the lives they have, what it's like to be another creature, then you don't want to just be satisfied with, yes they think, but what are they thinking about.

Now, we have this tendency when we see animals doing really cool and neat things like catching Frisbees, that they must be really, really, smart. But these are tricks and they really don't get us very far in terms of what's going on inside, how are they feeling about this, what are they thinking about when they calculate something like the trajectory of a moving Frisbee and how to catch it. That problem actually at some level halted many years ago when B. F. Skinner, in the tradition of the animal learning theorists, promoted the fact that there's really not much going on upstairs at all, and this is captured by Gary Larson who says, unbenownsed to most students in psychology, the first experiment was to ring a bell and cause his dog to attack Freud's cat. But at the same time, Culler and many other people working on the variety of animals actually began to think about another alternative, possibly, that in fact animals do have much richer sort of thoughts than we ever anticipated.

In this experiment you see here, what the goal was was there was a banana hanging from the ceiling and these chimps were kind of just hanging around and the question is, could any of them figure out how to get access to the banana. And what Culler reported was that the animals, one animal in particular, Sultan, stacked up some boxes and stood on top of the boxes and reached the banana. Now when he described this, he called it a case of insight, as if it there was an epiphany. The animal saw the banana, figured out there must be a series of boxes piled on top of each other to get access to the banana. It turned out, however, that through a series of trials and errors, basically, the animal figured out the solution. That's not insight, but it fully shows some capacity to solve a problem.

What I want to do for you today is basically three different things. The first thing I want to do is tell you about a series of areas I think we're all vulnerable to, and that's sort of largely captured by a phrase from Hamlet, which I'll talk about in a moment, the notion of a skull or direct line of evolution, and then following the view of anthropomorphism of treating animals, other animals, as if they're like us, and imputing them everything that we think we do when we act in that way. Then what I'll do is take you on tour of what I call mental tools, the kind of machinery that animals have to solve the problems in their environment. And the first thing I'll be telling you about is what I call universal knowledge Ð free capacities to recognize objects, count how many there are and navigate through space that I think are shared across all animals. These are the commonalties that we share with all other creatures. Then I'll talk about a set of specialized tools, and specifically, the sense of self, the capacity to deceive, and the capacity to attribute other beliefs and thoughts and emotion to other individuals with minds. And then I'll comment at the end to a problem that really Cofta raised in his metamorphosis about what it's like to have thoughts and to be able to express those thoughts.

Hamlet said, what a piece of work is a man, how noble in reason, how infinite in faculty and form in moving, how express and admiral, with action how like an angel, in apprehension, how like a God, the beauty of the world, the paragon of animals. What Hamlet is doing here is basically committing a tremendous error in thinking that there is a scale in nature where there are all these other creatures, but humans are on top of that heap. That there's some kind of linear IQ scale that you contribute to all creatures, with of course, us being on top.

This is captured here. We have a situation where we've got insects on the bottom, then we get snakes that goes to frogs, it goes to birds, it goes to ungulates then goes to primates. That's how life is created. Basically through evolutionary time, you get an increase in intelligence. The problem with that view is to think of evolution as a single line. As we now know from many studies in evolutionary biology, evolution does not work that way. Evolution is a tree, it's a bushy tree with lots of different branches and lots of different possibilities.

This is captured, I think, here in our sense in which we want to impose on other creatures exactly our kinds of thoughts and emotions, assuming that everybody's kind of got what we have or something completely different. But we take our lenses and look at the animal world with those lenses. So we have a horse, it's name is talking Ed, where in order to get at what animals are really thinking, we give them our voices so they can speak and tell us what they're thinking about. This is captured here beautifully in the movie, Babe. "Dogs are stupid mom. Not as stupid as sheep, mind you, but pigs are definitely stupid. Excuse me, no, we're not." Now there are dogs telling the pigs they're stupid, not as stupid as sheep of course. But we have the beginnings of this kind of this line of evolutionary intelligence and then we turn it around and we get the planet of the apes, where of course the apes now take over and the humans are the ones caged up and these guys are the smart guys. The error here is to treat animals specifically with the human view, not to consider that each organism evolves in a particular kind of social and ecological environment with the kinds of problems that it faces and has to solve. Here is why I think you need to take this view very seriously. If we were to be put into a room with a bat and we had our eyes closed, and we have to recognize all the objects in the room with our ears alone, we would look like morons and the bats would look brilliant. Turn the problem around. Ask the bat now to recognize objects in the environment with its eyes and not its ears, the bat looks moronic and we look like geniuses. The reason why this problem is silly is because the bats have evolved a unique adaptation. They have evolved a system of bio sonar, whereby they can produce a very loud sound, the sound goes out into the environment, it strikes an object, the object reverberates giving echo back to the bat, that gives it information about what kinds of objects are in its environment. In the same way that they use their ears to detect objects in the environment, we use our eyes. And therefore, we think about the brilliance of a species, it has to be sensed as the kind of problems they face in their environment.

So what I want to tell you about today is that we need to take a look at animals with an eye to comparing them against humans, but also comparing them against other species. In some cases, what we're going to be finding is that humans and non-human animals are very similar and in some cases we'll find they're different. And the differences and the similarities are equally interesting if you're interested in how minds evolve, and that's what I'm particularly interested in telling you about tonight.

So let me turn now to the second part of the talk and tell you about what I mean by a mental tool and then I'll turn to these three aspects that I call universal knowledge. The idea of a mental toolkit is simply this. It simply doesn't make sense to think of the brains of any animals as having just simply one function. What we've now learned from many studies in the neuro sciences is that the brains of all animals seem to be at least at some level, divided into a space where different parts of the brain do different things.

If you take that view of the organization of the brain, and therefore the organization of the mind, then it makes sense to ask, what are the kinds of mental tools that animals have that they use in their day-to-day survival. What problems are they confronting that are ecological in nature, what problems are they confronting that are social in nature, and how does their mind solve those particular problems.

So what I want to tell you about now are what I consider to be the universal tool kit Ð things that all animals have. The first has to do with objects, recognizing what they are and possibly if they're shaped like this, what they can be used for. We know that all animals make distinctions between objects in the world. This is a predator, this is not; this is food, this is not. This is a mate, this is not. This is a young, this is an older animal. Those kinds of distinctions between objects based on certain kinds of features are common to all organisms.

The second one has to do withÉ.Gary Larson will once again give us the answer. Look, you had five bones, right. Your friend Zoopie comes over, stays a while then leaves, now you have four bones, right. You don't have to be a Lassie to figure that one out. The other question is do you have to be a Lassie to figure that one out. Do animals have any capacity to track the number of objects in the environment? What I'll be claiming tonight is yes, in fact they do at a certain kind of level. And that level will tell us something about the problems they face.

The third one has to do with space, and here we know that all animals are equipped with at least two very basic mechanisms for moving around in the world. One is called dead reckoning where the animals use the distance and speed of displacement in an environment to track where they've been and therefore use that information to get back to a certain place, and the second is to use landmarks in their environment, literally tags of where things are to find their way to a particular location. In some cases, you get these extraordinary specializations. This is a Clark's nutcracker. This individual will hide seeds in up to 30,000 different locations and then two to three months later, retrieve them. This is a memory specialist. Now that's an extreme, but other animals have this capacity to hide things or find things based on a recall of where things were placed, what they are and where they are and in some cases, even when they were placed there.

What I'll be focusing on tonight, just to give you an example of how you can move in this area, is to focus on number, one of the areas I've worked most intensely on. Now you might think, look, number is something that's only salient to humans. After all, we are the ones who created mathematics. We are the ones who use formal number words and terminology to anchor the world number in our heads. It's a cultural construction. There's absolutely no biological basis at all. More of the point, there's nothing important about number in the animal world. Well here is why you should think that's wrong. In a number of animals, individuals will form coalitions, allies, where two to three individuals, shown here in dolphins, will gang up against two other individuals to beat up those individuals and then gain access to a sexually receptive female. So numbers matter in terms of power. We know from studies of chimpanzees the same kind of thing is going on, where two to three animals, again, will gang up to beat up yet a third. Studies basing all the way back to studies by Jane Goodall in Tanzania, and her wonderful field site and Gombie in Tanzania have shown the following result. Chimpanzees who confront a foreign individual in their territory or in their home range, will gang up and beat up and kill that individual if, and only if, the ratio of attackers to intruders is three to one. This suggests that numbers matter. It's a power game. It's like countries going to war against other countries. Do we have enough to defeat the other side. Here the ratio is looking like three to one.

Well how can you test this. We could simply be satisfied with our results and say, that's fine. But one of the things I think we've learned from the study of animals is that our observations are often misleading. To take things further, we want to do experiments. One of the things I think that will be a message for many of you tonight, is that there's absolutely no reason why when you observe something in your pet you can't take it further and do some very simple experiments. Now what I'll hopefully give you tonight, are some simple experiments you can actually do with your pet. With the chimpanzees, what we show with an experiment was that if you play back vocalizations from an intruder to these individuals, they will attack the intruder again, if and only if they have a ratio of three to one. So the experiment will support the observations from the natural behavior.

But can we take it further? Can we refine what we understand about animals and their number skills. For this, I want to take you from a different place, which is from the East African coast in Uganda to the tropics in Puerto Rico to an island where I've been working for about 13 years now. It's the island called Khaia Santiago. This small island off the coast of Puerto Rico is home to approximately 1,000 Rhesus Monkeys, this animal here. These animals were brought to the island in approximately 1936, about 400 of them. They've obviously reproduced. There are now 1,000 of them. They've been studied for over 60 years and because of these intense observations, we can get very close to the animals, between one to three feet. They treat us kind of like moving trees. They do their thing. We do ours. Everybody's happy.

Now I want to tell you about a very simple experiment, exactly the kind of experiment you can easily do with your dog or your cat if they'll watch you. Here's the idea. You find an animal who's alone, here's a lone monkey, we're here. Here's one of my students, Rex Lakowski. You show the animal two empty boxes and then into one box you place a piece of apple, and into this box you place a rock. Now you're thinking, this guy's an intelligent person, what a dumb experiment. Of course the monkey is going to go to the box with the apple. Well that's exactly what they do. That's not that interesting yet, but that gives us a very simple experiment. If they'll take one piece of apple over a rock, what about two pieces of apple verses one, three verses two and so forth, and that's the experiment. If you present the monkeys with two pieces of apple verses one, they preferentially take two over one. It doesn't matter if you put two into this box first or two into this box second. Regardless of the order in which you place them, they always pick two over one. They also pick three over two. And they pick four over three. But if five verses four, they basically go 50/50. Now this shows a limitation. It says that when these monkeys are seeing the situation, it's one trial per animal so we're not training them, it's one trial, it's when they go and get to choose. They can do four verses three, but when it's five to four, no difference. Now if you've been thinking carefully about this, you would be saying, wait a minute, why do you call that number. After all, it takes longer to put two pieces of apple into a box then it takes to put one piece of apple. We know that all animals are equipped with basic timing mechanism, so if they're timing the events, they would also succeed to pick two over one, three over two, but that's time not number. So to control for that, what we did was place two pieces of apple into one box, one piece of apple and a rock into the other. It takes equally long to do these and therefore, if they pick two over one, it's number and not time. They pick two over one. They pick three over two. They pick four over three. But they are 50/50 at five verses four. So even when time is controlled for, number is what is salient. So without any training at all in a natural situation, animals preferentially pick the larger number over the limit of up to about four.

Well the question is, after all, when we develop our mathematical skills, maybe we come into the world with some basic tools, but after all, what takes us further is our education, we're trained. Can we give animals an education? Can we take them further than what they can do naturally and spontaneously. And the answer is yes. If you train in a classic situation where you ask an animal to press a lever multiple times to get a piece of food, they'll do that. And what you show in these situations is that animals seem to be representing number in an approximate sense. So if a target number of presses is four, they're very accurate. They'll press approximately four times. Sometimes three and sometimes five, but within that range, they're basically exactly correct. If, however, you ask them to press the lever 20 times to get access to food, now what you see is much more variability. Sometimes they'll press 15 times, sometimes they'll press 30 times. In other words, as number increases, so does the variability, which suggests that with training, the animals can represent large numbers, but only approximately so.

Some animals have even been taught either the verbal words for number, as in the case of Irene Pepperberg's parrot, Alex, who actually knows the words one, two, three, four, five and six, although he says them with a kind of Long Island accent. And what I mean by Alex knowing what those words means is the following: If you show Alex a tray with a bunch of blue squares, a bunch of green squares, a bunch of green triangles and a bunch of blue triangles and say, Alex, how many blue squares, Alex will tell you. In other words, he understands that you have to pick out both the color and the shape and really pick out what the number of those items are.

Now maybe a more spectacular demonstration which is a little video I want to show you now, this is a chimpanzee known as Ai, A-I, whose been studied by the last 30 years by Tetsu Matzuzawa, a close friend and colleague. What Matzuzawa's been able to do is train Ai to actually understand the Arabic numerals from zero to nine. In this particular experiment, I want you to watch very closely. Here is the task. A little circle is going to come up here, as soon as Ai presses that circle on a touch-sensitive screen, between three and five different numbers will come up. Her job is to press those numbers in their order of relationship from lowest to highest. But, as soon as she presses the first number in the sequence, white squares will appear hiding the other ones, so she must remember what they are and where they are and press them in sequence. Now watch very carefully because she's going to do it much faster than any of you. Now, I hope you're convinced that Ai knows her numbers. It's hard not to be convinced that she knows her numbers. But here is where I want you to be very careful with how you interpret what you just saw. It's so common that when we see animals doing things that look like what we do, we assume that they're thinking about it the way we do. But here is why this might be a mistake. I'm not saying it's always a mistake. I'm saying it might be a mistake. If you look and see how Ai actually learned this task, it's very different than when young children learned the number or the count system. For Ai, it took her as long to learn what number one means, as what the number two means, as what the number three means, and so forth up to nine. She never got the ah-hah experience that children at the age of about three and three and a half get.

Here's what I mean. Below the age of about three, when you ask a child, for example, here's a plate of cookies, you have seven cookies, can you please tell me how many cookies there are. He'll go, one, two, three, four, five, six, seven. How many? One, two, three, four, five, six, seven. How many? One, two, three four, five Ð he'll go on forever like this because they don't get that the cardinal value, the last number tags the number in the set. And they won't get that until they're about three or three and a half, but at that point, they've got it, and they've got it and they can run away with the system. That's something this animal never got and will never get. If you want to say, can she learn ten, yes she can. But it will take a long time to teach her that. Can she get eleven, yes, and it will take her equally long to get that too. So here's the critical point. What we have learned from the study of non-human animals is that all animals, us included, have a system, which allows us to do small numbers precisely, and large numbers approximately. But only humans with language seem to have a system for doing large numbers precisely, and that seems to be a unique human adaptation.

So what that tells us is that at some level, we kind of share mental hands with all other creatures, at least in the sense of being able to recognize objects, count how many are there in a simple way, and navigate through space. Therefore, in terms of the number sense, yes there are biological building blocks, and there are at least two of them, which I just mentioned. What culture does is elaborate upon the biological foundation. So it's not as if the sum, the idea of number is a completely cultural construct, rather, the culture built on a biological foundation.

What I've just told you, therefore, has to do with the universal sort of tool kit. What I want to turn to now is something quite different. What are the capacities special to certain species and not others that could reflect something important about the ecology or social environment that these animals evolved in.

Let's talk first about the sense of self. What animals have a sense of self and how might we find out. Well, when we look at ourselves in the mirror, shocked or not, we know it's us. We know that that's what we're looking at, our face and not somebody else's face. What about animals? Well, certainly in a natural environment animals have situations where they will see their reflection. They don't obviously have mirrors, but they have mirror like substances or surfaces and that will give them their mirror image. What do they think when they see this. Well one possibility is what happened with Tigger. When Tigger saw himself in the mirror for the first time, he jumped back thinking, hey I didn't know there were other Tiggers in the world. When animals see their reflection, therefore, do they think it's another creature, like maybe their own species, rather than thinking that it's themselves.

As so often happens in Charles Darwin's life, he was actually the first one that I know of at least, to ask this question, to actually go out and do an experiment. He went to the London Zoo where there were two Orangutans and he put a mirror up to their cage. And what he found was the Orangutans approached the mirror, looked in it, often made some facial expressions and sometimes reached behind as if checking whether there was somebody else behind there. The behaviors in and of themselves were ambiguous in terms of what the orang's thought when they looked into the mirror and what it required for a basically more formal testing, was the experiment that was developed by Gordon Gallop in the 1970s to really get a better handle on what's going on with mirror reflections. Gallop already knew, based on a lot of other studies, that animals could use the mirror image to find things in the environment, and this included even things like fish, who when they see a predator in the mirror would evade their response, evade the area. So we knew animals could use the mirror to find things, but could they use the mirror to detect that it's me looking inside. So what Gallop did was he exposed chimpanzees to their mirror reflection for several days, and then once he had done that, he anesthetized chimpanzees, put them to sleep, and placed two red marks over their eyebrow and their ear, marks that could not be felt or smelled. The question is what would the chimpanzees do when they woke up. What Gallop found was that as soon as the chimpanzees woke up and saw themselves in the mirror, they didn't think they had a hang-over, rather, they touched the mark here, they touched the mark here, not the other ones and very quickly after that, they would use the mirror to see parts of their body they had never seen before. What I want to show you now is just a little quick videotape so you can see what they would actually do, and here is the tape.

"Older chimps need as little as half an hour to understand a reflection. They quickly discover the mirror offers intriguing new sites, parts of their bodies they've never seen before."

So what Gallop found was the following important thing. He claimed that when chimpanzees look in the mirror, they're not seeing somebody else, but rather see themselves, and know that it's them. So he claimed not only do chimpanzees have self-recognition, they have self-awareness. That distinction between self-recognition and awareness, I'll come back to in a moment, let me just say that this study was then run on hundreds of other species, and the conclusion to date was that great apes, the chimpanzees, orangutans and gorillas, although not all gorillas, in fact the puzzling thing about gorillas, I'll just mention it here because a couple people had asked about Cocoa, is of 25 gorillas tested, only one passed the so-to-speak mirror test and that was Cocoa. Here's what happened. Penny Paterson gave the traditional task of the marking to Cocoa and nothing happened at all. Cocoa didn't show any interest in the mirror, so what she did was a different test. She would walk by Cocoa everyday and just simply pat Cocoa on the head and say, nice Cocoa, nice Cocoa, whatever. And then one day she walked by Cocoa and patted Cocoa on the head with some white powder turning her black fur to white. And then somebody walked by with a mirror. Cocoa ripped the mirror out of the person's hands, looked at it, wiped the powder off, Penny sighed, who's that, and Cocoa's sighed, me Cocoa. Pretty convincing evidence. So all these other gorillas failed. But we have noticed a pattern that the apes who passed this task, children only at the age of approximately 18 to 24 months would show this kind of response. For those of you who have young children, please do try this out. It's a very nice experiment. I have friends who have a 15 month old. She's still doesn't get it, no. And they said, look, we're convinced our daughter knows that, you know, that that's her in the mirror. And I said, well, let's do an experiment. So the wife put on some lipstick and gave her daughter a kiss on the cheek and we put her daughter in the mirror and that daughter was looking at it and looking at it and no response whatsoever. It's something that really takes time to develop, so it's a very length developing response even in the human species, about 18-24 months. But the other claim was that all these other species completely failed, dolphins, parrots, all these other monkey species, complete failure. They would respond either by ignoring the mirror or acting aggressive towards it as if it was another species or individual.

Well I happened to work on a species, which was deemed selfless. And if you work with a species that's been told to be selfless, it's kind of boring. So here is the species, it's called a Cotton-Topped Hammer Monkey and this species lives in the forest of Columbia. So we were sitting around one day in my lab with a bunch of my students wondering, you know, why might these species actually fail this task, is there something peculiar about their ecology or their behavior that might give us some insight. So here's something interesting. Unlike the apes, including human apes, monkeys find it aversive to look in each other's eyes. You'll never see monkeys sitting there gazing, you know, starry eyed into each other's eyes. If you stare at somebody and you're a monkey, you're threatening them.

So their first idea was, well, if your first idea or insight into what a mirror is is that if somebody else is looking right back at you, it's pretty aversive. So maybe the monkeys wouldn't look in the mirror and therefore never get the information they need. So we thought, well look, here's this lovely white hair, maybe we should do is do something much more radical and change their hair color. And at the time in my lab, I had a student working in my lab who had orange and green hair and she was very keen on the idea and I said, hey, you know, where'd you get the hair color. And she goes, well, it's called Manic Panic and you can only get it in one store in Cambridge and that store is called Hubba Hubba. Now you can probably guess from the title of the store that Hubba Hubba is a store which specializes in sexual paraphernalia and my students decided that I was the only real adult who could go into the store, so they commissioned me to go up to Hubba Hubba, and so I went up to the store and I got two jars of Manic Panic pink and blue, came back to the lab and one of my very bright students who always catches me on the wrong experimental design said, look, we have all these monkeys. We can't just use two colors. We've got to use a whole bunch of different colors to make it a really clean experiment, so you better go back and get some more colors. So I went back to the store and the woman who had helped me the day before was there again and she said, hey, weren't you in here the other day, you hair is still brown. What do you really want? I said, you know, you wouldn't understand. And she said, you know, excuse me, but with the store I have I understand everything. I said, well, I work on animals and, and she goes, oh, animals, hmm. So she gave me a whole load full of Manic Panic. I went back to the lab and we dyed their hair and what we found was quite spectacular. To our surprise, not only did the monkeys sit in front of the mirror and touch their hair while looking, they actually sat and stared at the mirror for a very long period of time. Something they hadn't done before, as if something had really changed. Now this is interesting because of course now what's happened is that we're tapping something which is species typical. This hair is typical to these species, this species, the Tamarin. There are other species of Tamarins, they don't have these big white puff of hair. We have functionally changed one of their species typical marks. This seems to have gotten their attention.

So this opens the question, well, if we're going to do these tasks, with different species we have to be sensitive to what the species is all about. Their sensory modalities, their ecology and so forth. After all, not all species like primates are visual creatures. Would an electric fish that lives in electric modality, have a sense of self with its electric pulses? What about a bat who has auditory modality?

But more importantly, as I mentioned before, and Gary Larson, and one thing it captures in the animal self-help section is there's an important distinction between self-recognition and self-awareness. And here is the best way for me to capture this distinction. There is a very devastating neurological disorder called Prosamagnosia that some humans experience. It occurs due to damage to the back side of your head, to the occipital temple areas of your brain and when this happens, what you loose is recognition of familiar faces, but not a loss of recognition of familiar voices or any other objects. It's specific to the face. This deficit can get so severe that some people loose the ability to recognize their own image in the mirror. They would fail Gallop's task. But what's so devastating about this disorder is that it hits you where the self matters most. It's as if you don't have a sense of self. It's that entire sense of beliefs and desires and thoughts about others that makes the world so frightening to them. They can't recognize their spouses or their children or their friends. That's what makes it so devastating for the sense of self-awareness. The mirror test simply doesn't tell us anything about that. So although the mirror test is an important test in pulling out differences between species and recognizing an image in the mirror, it tells us nothing about the thoughts and beliefs and desires of those creatures.

To do that, we need to turn to something else, and that's deception. And there are two types of deception. One is actively falsifying information. The second, withholding information. These both work beautifully and we do it all the time. The question is, do animals have these tricks. Do they or have they the capacity for lies of commission as well as lies of omission?

Here is an interesting behavior that's seen in certain species of birds known as Plubbers. It's about as close as I can get to describing it is something like Monte Python's Walk of Phillies. When Plubbers see a predator moving towards their nest with eggs or young, they will dive-bomb at the predator and then flutter around on the ground as if they're injured, thereby luring the predator away from the nest and towards them, and at that point, they fly away. They've gotten the predator out of the way and saved the nest. Now this works beautifully and it is an act of falsification because these animals are not injured. It works beautifully. In fact, if the predator doesn't get word on the first try, they try again and are even more dramatic.

Imagine that two Plubbers are looking for food and then all of a sudden one flies around, distracts them and gets the food. They don't do that. It would be a great move, but they can't generalize. So yes, they can deceive, but it's a very specialized ability. Well maybe one of the problems here is that we're looking at birds with a social organization that's less complicated. What about primates where the socialism is much more complicated, where the advantage of deceiving others in a social situation would be that much greater.

This is a species that I studied in Kenya for several years along with my thesis advisors, Robert Syforth and Dorothy Chaney and my post-doctoral advisor, Peter Marlar and it's a wonderful system and the system is the following. This is the animal called the Vervet Monkey. It lives throughout Africa and we studied it in a population in Kenya. In its population, the Vervet Monkeys are hit by a wide variety of predatory species. When the monkeys see different predators, they produce different sounding alarm calls. One call for leopards, one call for eagles, one call for snakes. If you play back those vocalizations through a loud speaker, the animals respond as if the predator were there. If you play back a leopard alarm call, they run way up into the trees. That's the best place to be if a leopard is around because leopards can't climb that high. But if you hear an eagle alarm call, they look up into the sky and they dive under a bush because if you're up in a tree when an eagle comes, eagles will scoop you right out. If you hear a snake alarm call, they stand by and look around the grass below them. Just to give you an example, here is a male giving a leopard alarm call to a leopard the leopard.

Now, here is an interesting observation. When I was a grad student working in Kenya, I made the following interesting observation of this little group. This is Speak Easy, he's the alpha male in our group. This is Leslie and this is her mom, Borsha, who's the alpha female. Speak Easy had been trying for the entire day to mate with Borsha. He would go over there, he'd lift her back up, she'd squat. He'd lift it, she'd squat. He'd lift it, she'd squat. It was like all day long. No way was this going to happen. He gets pissed off, he slaps her in the head. She screams. She and her sisters all come running and are now chasing Speak Easy all over the territory. He sits down, gives a leopard alarm call, they go running up the trees. No leopard. Now that was the best case of deception I've ever seen. I mean, what a move, you know. I mean incredible, you know, what suckers. I mean, you know, why not use this all the time? Well, one of the things we know about deception is of course if you use it all the time and cry wolf it's not going to work. You've got to do it infrequently. But the other thing that's kind of bizarre, if you think about it, is that if he was really going to be a good deceiver, why was he sitting on the ground. I mean it's the worst place to be if a leopards around. Why didn't he go with them? The other problem here, this is why it's simple observations are very misleading, at least difficult to interpret. We could be satisfied in saying, well Speak Easy obviously was deceiving those females to get them off his back. That is one interpretation, but there are others. Maybe he thought he saw a leopard and made a mistake, you wouldn't want to accuse him of that. Maybe he didn't lie at all, right. Maybe actually there was a leopard and I missed the leopard, I couldn't see it. There are lots of possibilities, so how can we go further with this, how can we take these interesting observations that we all have of our pets and other animals and go further.

Here's one example. This is a species, again I told you about before, the Rhesus Monkey that I studied on the Island of Khaia Santiago in Puerto Rico, and when this animal finds food, they give one or more of different kinds of food calls. Some calls for high-quality rare food items, some calls for low-quality common food items. Sometimes, however, when animals find food, they don't call at all. So what we decided to do was do an experiment. Present an animal who is alone with a bounty of food, coconut in particular, the food they like the most, and see what happens. On about half the trials, animals gave these very characteristic food calls and other animals would come running in. But on some trials, about half of them, animals were dead silent. And what we found was that sometimes the discoverer was badly beaten up by those who found him at the food source. The only factor that seemed to account for whether or not you were hit by severe aggression was whether or not you called. Animals who were silent were pummeled, irrespective of their dominance rank. Higher-ranking animals who usually can out-compete lower ranking animals for food were beaten up if they were quiet. In fact, if you're an animal and you find food and you're quiet, and other animals come in and find you and now you give a food call, well, that doesn't count. You get beaten up anyways. You've gotta announce it before somebody finds you.

Okay. What that suggests is the interesting possibility that if you don't call, you are going to be punished because you are a cheater, not by actively falsifying information, but rather by withholding information. Okay. Now, as you all know, if there is punishment of cheaters, this begins to sound something like how they have kind of a convention, there's an expectation that calling is associated with food, something like a rule of convention. Well in our societies, rules of convention apply to individuals within our own cultural group. All the experiments I just told you about were done with animals within a social group. But there are a class of individuals in Rhesus Monkeys society that basically aren't yet a part of a social group. These are males who have left the group in which they were born. They are now looking for another group to join; they're floaters.

So we ran the same experiment with 25 adult males who were not in a social group, same exact experiment, present them the coconut and what we found was these males never called and when they were caught they were never beaten up. It suggests that the previous result is not some kind of specific kind of thing, but it's sensitive to group membership, that if you're a member of a group and you get caught without having called, you get beaten up. But if you're not a part of my group, aggression is costly. Therefore, getting into a fight with somebody who you may never see again is a very costly thing to do. Well, this suggests that there can be something like deception, something like punishment. We haven't learned yet much about the thoughts in the minds of these creatures. The question is, can animals think about other's thoughts. Do they have what psychologists have referred to as having a theory of mind.

Everybody in this audience has this ability in space. We can think a thought. We can think about somebody thinking a thought. We can think about this person thinking a thought, about that person thinking a thought, and on and on it goes. This capacity is natural to us. However, we're not born with this ability, in fact it takes about four or five years to acquire like this. Children by the age of three don't have this ability. Moreover, autistic people do not typically have this ability. They are what someone referred to as mind blind. They don't get that others have thoughts that are different from their own. That's a very devastating thing to be. It's like living in a bubble and that's what's so horrifying about being an autistic person. Your relationship with the world is just very, very different than we experience all the time. Could it be that other creatures are like three year olds? They don't have a theory of mind. They're very good at dealing with behavior, but they can't make any inferences about the minds of others without experiencing behavior from them.

Here's something that everybody here probably didn't think. You probably didn't think that I'm going to get up on stage and take all my clothes off, right. You didn't think that thought. I can assume that because people don't do that. Now it's not because you know anything about me, but because you make those assumptions about my beliefs, your beliefs and so forth. We go on and on like that all day long. Do animals do that?

Well, here's a question that Danny the Anthropologist asked, about chimpanzees with respect to ignorance. Do animals know that some people are ignorant and some are not? Are some people knowledgeable, some are not? Well here's a very simple experiment. Two people are in a room, this man with a white lab coat and this one with a white lab coat. You show the chimpanzee there are these four boxes here and [inaudible] place in front of them, a screen. The person here in the white lab coat leaves the room. While he's out of the room, this woman baits one of these boxes with food. This guy comes back into the room, the screen is removed, and now both people point to a box. The woman always points to the correct box, where the food is. The chimpanzee's job is to reach out and pick the correct box. Over hundreds of trials, the chimpanzees never get this. They're basing random. Sometimes they pick this guy, sometimes they pick this woman, they don't get it. Eventually they figure out, oh, hey, the blond haired one is the one who is always giving me the reward and they pick her, okay. Now, do a simple test. This guy stays in the room, you put a bag over his head so he can't see, same experiment, their back at random, the chimpanzee's seem not to know what others know based on what they can see. And as we all know, seeing is a proxy for knowing. If I look over here and see all of you following my direction of eye gaze, I know that we share attention, and therefore we share some kind of knowledge. Now it's not because chimpanzees can't follow the direction of eye gaze. This is shown beautifully here. The woman comes into the room, here's a chimpanzee, she looks up, as you can see the chimpanzee's beginning to follow, he follows her eye gaze right into the appropriate corner. So we know chimps can follow eye gaze, but they don't seem to get that seeing is knowing. Well there's a problem. This is asking chimpanzee's to know what humans know. What about whether chimpanzees know what chimpanzees know. A much more natural question.

This is an experiment conducted by a grad student of mine, Brian Hair, a beautiful experiment, that really tests the different kind of question but gets to the same kind of problem. And that is, if chimpanzees are competing for food, a natural situation among each other, they better well know what somebody can see and what somebody can't see. So here's the experiment. There's a dominant animal over here, a subordinate animal over here, and two solid opaque screens. In the first experiment, you place bananas here and bananas here, open the cages, and let them go. The dominant comes out, the subordinate doesn't move. Dominants can't compete for subordinates in chimpanzee's order of food. Dominants are going to get the food, subordinate hasn't a prayer.

Now change the situation. Place the bananas, one pair, behind the screen out of view of the dominant but in view of the subordinate. The other bananas go dead center where both of them can see. Open the doors, if the subordinate knows what the dominant can see, and therefore know, the subordinate should now move here, but not here. That's precisely what happens. The subordinate now moves immediately to the secluded piece, because the subordinate knows that the dominant can't see this and therefore can't get it. Turn the thing around, hide one piece in the dominant side, one in the middle, now release them. What does the dominant do, he goes here first and there to the second. Why, because the dominant knows this guy can't get that one, he never even saw it. And he surely can't get that one because I'm the dominant. So dominant is tracking the vision of subordinate as well. What this tells us is that in a more natural and ecological context, chimp on chimp, chimpanzees do understand that seeing is knowing. And therefore, what we have here are the rudiments of the understanding of how something like a theory of mind evolves. A very powerful capacity because here on your own you can model the world. You can generate predictions about others without actually ever encountering them before. It's a very powerful thing that certainly it seems like chimpanzees have the rudiment itself. Whether other species have it we don't know because these tests simply haven't been conducted on them. It also tells us a very important message. If you want to understand what animals know, you better test them animal and animal, not asking animals to infer what humans know.

What I've given you thus far, we're coming to the end now, is basically a tour of some of the errors people have made I think in interpreting animal behavior, in particular, animal minds and emotions. And tell you sort of in a different way, an alternative way of thinking about the problem. It's not to say that animals don't have rich mental lives. In fact, much of what I told you about is about their rich mental lives. It's simply to question are one of observations about what we see to infer that if it looks like what we do, it must be based on how we think about the world. That may be a mistake.

What I'll turn to now is the last part of my talk, which really is about the evolution of our own species, and what I think is really deep and important about being a human being, and also being an animal.

In [inaudible], Samsa awoke one morning from uneasy dreams, he found himself transformed in his bed into a gigantic insect. The critical line is, what has happened to me, he thought. The critical problem here is that Samsa, now a bug, has all the rich thoughts of being human but can't express them. What I told you today, I think is a story that has deep implications for the evolution of the human mind as well as animal minds, and that is animals have deep, rich thoughts and emotions. But what they don't seem to have are the communicative tools to express them the way that we do. And that's important, because what it looks like may have happened during the course of human evolution is that we were humming along as a species, before homosapiens, with lots of rich thoughts and emotions and so forth, but waited till the evolution of language to liberate those thoughts and to exchange them with other organisms.

So we come back to a question, which the philosopher, Thomas Nagel asked many years ago, what is it like to be a bat, we can ask that question of all sorts of organisms. What is it like to be a dog, a cat, a monkey, dah-dah-dah-dah. And I think what we've learned from the study of animal behavior is that using careful observations and careful experiments, we can get some insights into what it is like to be then a creature. In some sense, we share common thoughts and emotions with other animals and in some cases we don't. But that's also true between different species of animals. The tale, therefore, is one about the evolution of minds, similarities as well as differences. Thank you.