The Chronicle Review

The Intelligence of Beasts

Cognition researchers move past 'chimpocentric' theories, raising new questions about human uniqueness

Chronicle Review photograph by Scott Seymour

Years of study of elephants, like these in Kenya's Amboseli National Park, suggest they have cognitive abilities that sometimes surpass our own.
June 26, 2011

If you've ever doubted that elephants are contemplative, Joshua M. Plotnik has some video you should watch.

Plotnik, a postdoctoral fellow in experimental psychology at the University of Cambridge, wanted to see if Asian elephants could pass a classic cooperation test designed for chimpanzees. The elephants already knew how to use a rope to pull a food-bearing table within reach. But what if the only way to move the table was two elephants pulling on separate ropes simultaneously?

On one video, one elephant ambles up to the rope and waits patiently, trunk and tail gently swinging, for its counterpart to arrive, half a minute later. Then, without hesitation, they grasp their respective ropes in synchronicity, pulling the food to themselves. In another segment, a young female simply stands on her rope, which—the way the mechanism is set up—forces her colleague to do all the work for the both of them. In 60 trials, all six pairs of elephants waited for their partners, with an average success rate of over 93 percent, suggesting that they easily understood cooperation.

"We were very excited by the results," Plotnik says into his cellphone as he walks home, after midnight, through the streets of Chiang Mai, Thailand, not far from his research sites. "Their behavior was comparable to that of the chimpanzees. We're getting further into understanding how intelligent they are."

For much of the last century, research on animal cognition focused almost exclusively on primates, on account of their relatively close evolutionary kinship with humans. But in recent years, many researchers wishing to understand how higher intelligence evolved have taken a different approach, looking to apparently intelligent species that are only distantly related to ourselves, like elephants, dolphins, or ravens. In the process, many cognitive traits once thought to be exclusive to humans—including some that are considered definitive of human uniqueness—have been found in far-flung provinces of the ani­mal kingdom.

"In understanding evolution, human cognition is like the elephant's trunk: It's a very unusual thing, and it really stands out," says Evan L. MacLean, a doctoral student in evolutionary anthropology at Duke University, which is coordinating a collaborative project to compare 30 species, as varied as dogs and octopuses. "The best source of evidence we have are the closest relatives—chimps and bonobos—but we've been missing a ton of very interesting variation by just looking at primates."

Nine-week-old puppies have been shown to be very good at recognizing human gestures and interpreting them for their own benefit, while adult dogs provide and request information, predict social events, and perhaps even speculate on what their masters are thinking—for example, finding hidden food by picking up on subtle human cues. Dolphins recognize themselves in mirrors, while the New Caledonian crow fashions tools with which to capture grubs. The common raven has demonstrated the ability to test actions in its mind, solving complicated puzzles to obtain food on the first try.

"There used to be this chimpocentrism in the field, with research narrowly concentrated on primates," says Ádám Miklósi, a professor of ethology at Budapest's Eötvös Loránd University and an expert on canine cognition. "It's really great to see research in other species expand exponentially, because we really need the breadth and depth of species to be able to say anything meaningful about animal cognition."

While there are many reasons to study the thinking ability of animals—devising better conservation strategies, opening new pathways in artificial intelligence—the great evolutionary question driving many researchers is this: Under what evolutionary pressures do different types of cognitive abilities tend to develop? If several entirely unrelated species turn out to have a given intellectual ability—mirror recognition in humans, dolphins, and elephants, for instance—are there common denominators in the conditions they confront (membership in complex social groups, for example) that might explain the development? Can the study of such examples of convergent evolution help us understand how and why higher intelligence arises in nature? Researchers have taken only the first steps toward finding definitive answers to both of those questions.

Elephants are fascinating subjects in this regard. They have enormous brains, bigger than those of humans. Like many other animals thought to possess unusual intelligence, they live in complex societies in which individuals cooperate and interact to solve problems. They're also unrelated to humans. Whereas we and chimpanzees had a common ancestor about five million years ago, with elephants (and their relations, the aardvark and manatee), it was 100 million years ago, a relationship even more distant than the ones we have with dolphins and whales.

"If you find an elephant and a human sharing an ability that most other primates don't have, you can be pretty sure" it's an example of convergent evolution at work, says Richard W. Byrne, a professor of psychology at the University of St. Andrews, in Scotland, who has studied cognition in both primates and elephants. "If you can see shared circumstances in which these animals use this ability, that can help show what its biological function is."

Researchers like Byrne are finding that elephants have all sorts of intriguing cognitive abilities comparable to—and sometimes exceeding—our own. They recognize their reflections in mirrors. They act empathically toward fellow elephants in distress and offer them assistance. They mourn and even bury their dead. As Plotnik has shown, they're not flummoxed by cooperative problem solving. And both Asian and African elephants make tools for a variety of purposes, from flyswatters to backscratchers.

Byrne's fieldwork has taken place in Kenya's Amboseli National Park, site of a long-term, multidisciplinary study of elephants, in its 39th year. (That research is described in an edited collection, The Amboseli Elephants: A Long-Term Perspective on a Long-Lived Mammal, published this spring by the University of Chicago Press.) In recent years, Byrne and his colleagues have conducted field experiments that show two additional mental abilities not seen in any other animal.

In one experiment, elephants were presented garments worn by two different human ethnic groups: the Masai (who sometimes spear elephants) and the Kamba (an agrarian people who rarely interact with elephants). Whether given access to smells of previously worn clothing or to visual inspection of unworn clothing, the elephants responded with greater fear or aggression toward the Masai clothing. This led Byrne and his colleagues to conclude that elephants are able to categorize another species, humans, into subclasses.

"We don't know any other animals that categorize dangers in this way, but that may be because they haven't been tested," he says.

In the second experiment, the researchers wanted to test if elephants understand "person permanence" and "invisible displacement"—that is, if they are aware that absent individuals exist and if they contemplate the movements and activities of those individuals in absentia, an ability that human beings develop in infancy. Knowing that the members of the extended elephant family they were working with sniffed one another's urine deposits, the researchers began moving the deposits around. "The nice thing was to see their surprised reaction when they encountered a sample from an individual who was actually a half-kilometer or more behind them, a much greater reaction than if the sample was from someone who was actually ahead of them," Byrne explains.

The results suggested that each elephant was keeping track of the location and expected movements of everyone else in its foraging group—17 individuals, in this case—a remarkable display of working-memory capacity.

"They don't use their eyes in the bright Amboseli sun, but they're still able to keep track of 17 individuals in an environment that they cannot scan," Byrne notes. "Think of a parent with 17 kids roaming around a department store; I think we would find it extremely challenging to keep track in our minds of where everyone is. This is an ability that appears to be rather more than humans can do."

Miklósi, who is using cognition work with dogs to program robots that will, doglike, anticipate, serve, and respond to human needs, suspects that scientists will discover many more instances in which animals are found to have mental powers that humans lack. "So far, no matter how much you read in the literature, there is this notion that humans are always the best at everything," he says. "If the tests were not so anthropocentric, we would probably lose a lot more often."

Some other researchers agree that animal-cognition research will inevitably challenge human presumptuousness. "From ancient times, it's always been us versus the rest of the animal world, and it wasn't really considered that animals were thinking creatures," says Diana Reiss, a professor of cognitive psychology at Hunter College of the City University of New York, who studies elephants and dolphins. "The big message is that we're not the pinnacle of the tree, the only creatures who are thinking complexly. We share the upper branches of the tree with many species, each so beautifully evolved for their own environmental niche and social structures."

But other scientists take a more skeptical view. They include David Premack, a professor emeritus of psychology at the University of Pennsylvania and the father of the influential "theory of mind," which refers to the ability to infer the mental states of others. "I have a prejudice that's not common among people who study animals, who tend to do so primarily to show similarities between the animal and the human," he explains. "That's a noble cause, but I feel it's easily abused, as when they leap from similarity to equivalence."

Humans and certain animals might similarly recognize that a large rock is more likely to break a branch than a small one, but Premack says there's no evidence that this understanding of a physical action is equivalent to causal reasoning. That is, if an animal comes across a large rock lying beside a crushed plant, it won't infer, as a human might, that the rock crushed the plant, he argues.

Microscopic study of the human brain, he notes, has revealed neural structures and forms of connectivity not found in any other animal. And cognitive research has not been able to keep pace with those developments. "Things that the naked eye can see—that the human is a very different entity than the nonhuman—is very much corroborated under the microscope," he says. "If you can't see we're special, you have to be permanently drunk."

Byrne agrees that it would be foolish to jump from similarity to equivalence, but he says comparative-cognition researchers are by no means trying to show that animals are "little humans in disguise."

"None of the interpretations we make of our experiments require one to treat superficial resemblance as deep equivalence," he says.

As for his own research interests, Byrne would like to explore what creatures in other, unstudied corners of the animal kingdom might be thinking. Most species thought to be especially intelligent are social creatures, but what about large-brained nonsocial species?

"I don't know anybody who's looking at bears. They have very large brains—I wonder why," he says. "If someone offered me a long-term study of grizzlies, I'd take it."

Colin Woodard, a longtime Chronicle correspondent, lives in Portland, Me. His fourth book, American Nations: A History of the Eleven Rival Regional Cultures of North America, will be released in October by Viking Press.