It’s feeding time at the Myakka City Lemur Reserve, a leafy 90 acres about 25 miles east of Sarasota, Florida. On this steamy morning in early March, baskets containing pellets of monkey chow, fruit, and other treats drop from the trees onto a clearing. Three ferretlike brown lemurs, the pint-size alpha dogs of the preserve, make a beeline for the food. They elbow aside their less aggressive—and slightly affronted—brethren: the red-ruffed lemurs, sporting stiff Elizabethan collars of fur, and the hypervigilant ring-tailed lemurs, whose facial markings and luxurious striped tails give the impression of svelte raccoons.
Yale University psychologist Laurie Santos squats on the ground surrounded by half a dozen of these curious primates. They regard her quizzically while she takes pictures of them with a digital camera. The photos aren’t keepsakes of her field trip but will be used to help her comprehend the social structure of lemurs, specifically whether they have an affinity for forming social cliques.
Lemurs, from the island of Madagascar off the coast of Africa, evolved in isolation for some 30 million years. Despite this long separation from the rest of the primates, they now have something in common with a group of rhesus macaques in Puerto Rico and capuchin monkeys living at Yale: They all contribute to Santos’s wide-ranging study of our primate relatives, offering an unparalleled glimpse into our evolutionary past. Monkeys, it turns out, have many of the same survival skills that we do, from a predilection for forming groups to a knack for taking risks and deceiving adversaries.
Santos’s studies show that monkeys also possess many of the quirks and foibles once considered uniquely human. Like us, they intuit what others are thinking by reading social cues, a skill that may be millions of years old, hardwired into the primate brain. They make the same errors in economic reasoning that we do, suggesting our sometimes irrational attitude toward money might once have conferred an evolutionary edge. Through a series of groundbreaking experiments, Santos has seen in her primates a humanlike propensity for hoarding, larceny, and competitiveness. By exploring the inner lives of primates, she has offered persuasive evidence that monkeys are capable of sophisticated insight, complex reasoning, and calculated action.
Although the 33-year-old New Bedford, Massachusetts, native grew up surrounded by cats and dogs—her mother is a pet rescuer —Santos originally aimed for a career in law. The switch to studying animal behavior was, she says, “utterly serendipitous.” Shut out of a prelaw seminar in her freshman year at Harvard, she took a psychology class that ultimately led her to the study of nonhuman primates. She was captivated. Now an associate professor at Yale and head of the university’s Comparative Cognition Laboratory, Santos explains to DISCOVER how she learned to think like a monkey—and, in the process, came to understand more about how humans think too.
You first observed monkeys with your mentor at Harvard, the evolutionary psychologist Marc Hauser. What intrigued you? Initially it was just the opportunity to go to his field site on this island near Puerto Rico. But when I saw the monkeys, I got hooked. Watching them, you can’t help but see that they have the same kinds of issues that we do. They play. You can see their social striving. You can see their battles, their caring. They want to make friends and mate and be on top. They care about their kids, and they want their friends to do well.
One day I was alone on the beach and I picked out a nice spot to eat. A monkey came and sat beside me, and he had his monkey chow and was eating. We sat there together and I wondered what this monkey was thinking. Did he think the spot was beautiful? And how would I know that he did? That got me thinking about how they’re a lot like us, but on the other hand so different. They’re navigating these complicated motor situations, jumping from tree to tree, dealing with the stresses of new environments, and foraging. They do all of this minus language, minus the kind of social and cultural learning we humans have, yet they live, they mate, and they do some extremely complicated things, particularly in the social domain. How they achieve these same ends minus computers, BlackBerrys, or tape recorders is fascinating.
Your research with rhesus macaques in Puerto Rico suggests that they are capable of discerning human intentions. How did you realize that they were studying you even as you were studying them? The monkeys were good at deception. Nowadays when out in the field we eat lunch inside a cage that we bring along, in part because the monkeys have gotten very good at stealing our lunch. And it’s not just the food we eat but the fruit we use as stimuli in our testing sessions. Many days we went home early because the monkeys ate all the fruit. We would be doing number experiments where one plus one is supposed to equal two. But with only one lemon left, they can’t add one plus one and we can’t test their addition skills. In order for the monkeys to have taken the lemons when we didn’t notice, they must have paid attention to what we could see and what we couldn’t see. They’d have to wait for the moment we were no longer looking, and step in and take the lemons away.
So they were watching you intently? Not just watching us but specifically paying attention to the cues that were related to whether or not we could see, whether our eyes were pointed toward the lemons. You’d turn to write in your notebook and you’d look up and there was a monkey running up the hill.
That led us to develop a series of experiments where we directly asked whether or not the monkeys were good at using cues, such as where eyes are pointed. The real question is this: Are the monkeys just good at reading our behavior? “Oh, she’s turned away now.” Or are they thinking about our perceptions, mainly “She can’t see now.” And this is still a question of much controversy in the field.
Do you think that, in a very rudimentary way, these monkeys were able to read your mind? To the extent that they’re reading behavioral cues that are correlated with mental states—things they can’t see—yes, they are mind readers. But do they know that you have a mind in the first place? Do they understand that we have thoughts?
One of the most basic mental states that we think about in others involves perceptions—what you can see or hear—and that builds up to what you know or don’t know. The ability to discern these states in others was probably enormously powerful back in the evolutionary day, when we grew up as social primates who lived and died on the basis of how well we were able to predict others’ social actions: Are you going to be a friend to me? Are you going to back me up later?
How do you go about exploring monkey awareness of human perception?
The simplest study uses two experimenters standing on either side of a platform holding two portions of food. One of the experimenters can see the food and the other can’t because they are facing in opposite directions. They get into position near a monkey who is all by himself, and the monkey gets one trial to steal the food. And the question is, who does he steal the food from? Consistently the monkeys steal the food from the guy who’s facing away. Even without training, they realize that when eyes are pointed at them, it’s probably not a good idea to do it. But as soon as eyes are turned away, all bets are off.
Other studies suggest that monkeys pick up on what we can hear. We have a single experimenter who has two boxes. One is covered in jingle bells, so when it moves it makes a lot of noise. The other one looks the same because it’s covered in jingle bells, but we’ve taken the balls out of all the bells, so the monkey can hear that when the experimenter puts food in the box it’s not making any sound. So when the experimenter freezes and looks away, the question is, which box does the monkey steal from? What we find is that they sneak over and take food out of the quiet box and avoid stealing things from the box they know is noisy.
You’ve also begun researching economic behavior in capuchin monkeys. How did this study come about? Some of the capuchins in our colony at Yale came from a lab where the males had already learned how to trade tokens for food. It was funny when the monkeys first arrived because one of the caretakers said we had “genius” monkeys. “They hand their orange peels back to us so we don’t have to clean the cages,” he said. And then I realized it was because they had been trained in token-trading behavior.
So they were using every available scrap to get more food? Yes, they were pretty crafty—trying to negotiate even before we set up market experiments of our own.
How did you establish a barter system among the capuchins? We set the monkeys up with different human traders who offered things at different rates. When I started at Yale, I was lucky enough to have a colleague, Keith Chen, who is an economist and had already been collaborating with Marc Hauser at Harvard. We got together for drinks one night and started talking about ways to set up these experiments. We started with a series of studies examining whether capuchins pay attention to the “price” of food and whether they switch their purchasing behavior when items go “on sale”—if they buy more of the foods when they’re on sale to maximize the amount of food they’re getting. The easiest way to do these studies is to find two objects the monkeys like about equally, like apples and Jell-O cubes. The monkeys have their little wallet of 12 tokens, and they can spend them any way they like. There are two “salesmen,” one selling apples and the other selling Jell-O. What we found is the monkeys spend half their budget on apples and half on Jell-O. The next day, one experimenter has the regular-size apples but the guy selling Jell-O is offering two cubes for the price of one. And lo and behold, just like a human would, they buy more Jell-O. We showed that the monkeys’ market sense is very similar to our own.
But you also began to focus on the monkeys’ mistakes. I was interested in the errors that people make and looking for ways to study those errors in the monkeys.
Why focus on human error? Humans do a lot of systematically dumb things. The way we categorize things is biased and leads to stereotypes. Our judgments are biased when we deal with finances, but why? We’ve hypothesized that maybe some of these errors were built in from the beginning, which is why they’re so hard to get rid of. And if that’s the case, then these kinds of biases might be shared with other primates, which leads to the question of whether they are really biases in the first place. Maybe they’re smart for something that is just wrong for our modern environment.
What kinds of errors are we talking about? One source of error is an innate aversion to loss. Basically it’s the same experimental setup: There are two salespeople. One of them looks like he’s offering one piece of apple, but when the monkey pays him, that experimenter delivers a bonus—an extra piece of apple. So the monkeys get two, which they think is pretty good. The second salesman looks like he’s going to give the monkey three pieces of apple because he’s holding three pieces. But when the monkeys pay him, he takes one of these pieces away, so the monkeys see a loss of one piece of apple.
The important thing is that, on average, the monkeys get two pieces of apple from both experimenters. So in practice they shouldn’t care and should trade randomly, but that’s not what we see. The monkeys prefer to trade with the experimenter who’s giving them the bonuses and avoid the experimenter who’s giving them the losses.
What does this tell you about irrational economic behavior? Economists assume that our decisions are based on what’s going to give us the most wealth or happiness. But the monkeys violate the assumption that all they care about is the amount of food they get. They also seem to care about how the amount of what they get differs from what they expected.
Have the monkeys surprised you in any other ways? We did a similar study in a risky context, where both experimenters start out offering three pieces of something. The first experimenter appears to offer three items but each time ultimately gives the monkey two, so the monkey gets a loss, but it’s a safe, consistent loss. The second experimenter starts out offering three but introduces more risk: Sometimes the monkey gets all three, but sometimes it gets only one. We find that the monkeys prefer to go with the second experimenter. They prefer to risk losing more because there is also a chance they will have no loss at all. That is just what humans do.
It sounds like the falling housing market. People stay in because it could ultimately go up, but meanwhile, the whole thing might crash and burn. Where is the evolutionary advantage in this? My guess is that the strategy worked in the kinds of unstable, variable environments in which we evolved. Say you’re foraging for nuts and you’re in a really bad way—if you don’t get five you’re going to die. If you haven’t found five nuts yet, you need to take a lot of risk to get them because the alternative is really bad. As seen in the housing market, people take more risk when facing loss than they do when dealing with gain.
There was a famous experiment done by the psychologists Daniel Kahneman and Amos Tversky back in the 1980s involving a hypothetical Asian disease that is expected to kill 600 people. There are options for fighting the outbreak. In one scenario, 400 people will die for certain. In another scenario, there’s a one-third chance that no one will die and a two-thirds chance that everyone, all 600, will die. What do you pick? When framed that way, people will choose the riskier scenario because they want the slim possibility that there’s not going to be any loss. The housing market is another case in point. You’ll risk an even bigger loss just on the chance that you might not have any loss at all.
You’ve also studied another kind of bias: the tendency to devalue what we don’t have, a “sour grapes” attitude. Human experiments suggest that when forced to decide between two things we like equally, we end up devaluing the object we chose against. The Aesop’s fable of the sour grapes is a prime example of wanting something but when you find out you can’t have it, you decide you never liked it anyway.
The way we tested this in monkeys was to offer different-colored M&Ms, which the monkeys liked equally. We gave the monkeys a choice between, say, red and blue, and they picked blue. Then the question was, what happened to the M&Ms that the monkeys didn’t choose? Did they irrationally dislike those more than they had before? To get at that, we then gave them a choice between red—the one they chose against—and a novel color, like green.
The prediction was that the monkeys would pick more of the green. And that’s just what they did, which suggests they’ve irrationally devalued red M&Ms because they’d had to make an arbitrary choice about those a few seconds earlier.
And how might this bias provide an evolutionary advantage? Often we make a decision and then change our beliefs to fit the decision. When things don’t match up, we may try to resolve the dissonance through rationalization. It might be the case that your cognitive system strives for consistency in a way that leads to correct inferences about the world later on.
You’ve said that monkeys make the same reasoning errors humans do. Do you think monkeys can reason? I do. They obviously can’t verbally reason things out, but their choices are extremely similar to the ones humans make. So in that sense, I think they have to be reasoning.
Why do you study lemurs when other primates are far more like us, both genetically and behaviorally? Lemurs are important to study because they’re so distantly related to humans. If we see the same cognition in lemurs that we also see in macaques, capuchins, and chimpanzees, it allows us to learn more about how old that kind of cognition is. The other thing is that the lemurs developed in such a different environment on the island of Madagascar. So if we see the same kind of abilities in them, those abilities are probably widespread across primates.
Sort of like behavioral carbon dating? Yes. That is the logic we use. If we find the same capacity in lemur species and in our capuchin monkeys at Yale and in the macaques we’ve been working with in Puerto Rico—if all these branches at the top of the tree have this capacity—then it probably didn’t evolve separately, and it is probably something that we all shared back in the day.
For instance, we found lemurs are as good as a six-month-old human baby at adding and subtracting objects. We showed the lemurs one lemon entering a box, and then a screen came up so they couldn’t see what was in the box anymore. Then we showed them another lemon being placed in the box. Then we would remove the screen and reveal either the correct outcome of two objects in the box or incorrect outcomes of three objects or one object.
Our measure of whether the lemurs expected this was based on how long they looked at it. The assumption was that if it was what they expected, they’d be bored and turn away. But if it was something that surprised them or that they found unexpected, they’d look longer. And that’s just what they did. All the lemur species looked longer at events where we showed them the incorrect outcome.
You also discovered that lemurs can use tools, even though they don’t use them in the wild.
It turned out the lemurs were very good at picking up on the features that were relevant for the tool’s function. While they lack the capacity to set it up from scratch—to create a tool—they still understand how to use it.What does that finding suggest for human evolution?
One of the working hypotheses is that we have a set of simple physical and social capacities that help us function and make sense of how things work. And that provides the scaffolding for learning that can take place when you’re in an environment that has lots of tools. These are basic building blocks for all the stuff we learn later.
This set of capacities seems to be shared broadly across the primate order and seems to come online in humans early in their development. Whether you grow up as a hunter-gatherer in Africa or a modern-day human in Sarasota or a baby lemur in the forest, you seem to have these simple capacities.
Will we eventually discover that nonhuman primates have a complicated inner life? How much they feel is still a big open question. Do they feel loved? Do they feel emotional pain? We still don’t have good measures. But in terms of how they think, we use the same measures that we use in humans, and our studies suggest primates are reasoning about things they can’t see and interpreting mental states. Just a few years ago, we didn’t think primates could do this. Animals’ mental lives are richer than we used to think.
Fifty years from now, are we going to look back on the way we handled other primates and think of it as barbaric? The hope is that seeing more of ourselves in them will help us treat them more humanely and work to conserve them and their habitats in future generations. Our experiments suggest that these primates can understand our intentions. Can they morally reason about what we do? Can they understand someone’s intent to hurt them? When we get the full package of how they pay attention to the world, it might end up informing how we treat them.
