Tales about the origins of our species always start off like this: A
small band of hunter-gatherers roams the savannah, loving, warring, and
struggling for survival under the African sun. They do not start like this: A
fat guy falls off a New York City subway platform onto the tracks.
But what happens next is a quintessential story of who we are as human
On Feb. 17, 2013, around 2:30 a.m., Garrett O’Hanlon, a U.S. Air Force
Academy cadet third class, was out celebrating his 22nd birthday in New York
City. He and his sister were in the subway waiting for a train when a sudden
silence came over the platform, followed by a shriek. People pointed down to
O’Hanlon turned and saw a man sprawled facedown on the tracks. “The
next thing that happened, I was on the tracks, running toward him,” he says. “I
honestly didn’t have a thought process.”
O’Hanlon grabbed the unconscious man by the shoulders, lifting his
upper body off the tracks, but struggled to move him. He was deadweight.
According to the station clock, the train would arrive in less than two
minutes. From the platform, O’Hanlon’s sister was screaming at him to save
Suddenly other arms were there: Personal trainer Dennis Codrington Jr.
and his friend Matt Foley had also jumped down to help. “We grabbed him, one by
the legs, one by the shoulders, one by the chest,” O’Hanlon says. They got the
man to the edge of the platform, where a dozen or more people muscled him up
and over. More hands seized the rescuers’ arms and shoulders, helping them up
to safety as well.
In the aftermath of the rescue, O’Hanlon says he has been surprised
that so many people have asked him why he did it. “I get stunned by the
question,” he says. In his view, anybody else would’ve done the same thing. “I
feel like it’s a normal reaction,” he says. “To me that’s just what people do.”
More precisely, it is something only people do, according to
developmental psychologist Michael Tomasello, codirector of the Max Planck
Institute for Evolutionary Anthropology.
For decades Tomasello has explored what makes humans distinctive. His
conclusion? We cooperate. Many species, from ants to orcas to our primate
cousins, cooperate in the wild. But Tomasello has identified a special form of
cooperation. In his view, humans alone are capable of shared
intentionality—they intuitively grasp what another person is thinking and act
toward a common goal, as the subway rescuers did. This supremely human
cognitive ability, Tomasello says, launched our species on its extraordinary trajectory.
It forged language, tools, and cultures—stepping-stones to our colonization of
every corner of the planet.
In his most recent research, Tomasello has begun to look at the dark
side of cooperation. “We are primates, and primates compete with one another,”
Tomasello says. He explains cooperation evolved on top of a deep-seated
competitive drive. “In many ways, this is the human dilemma,” he says.
In conversation, Tomasello, 63, is both passionate and circumspect.
Even as he overturns entrenched views in primatology and anthropology he treads
carefully, backing up his theories by citing his experiments in human and
primate behavior. He is aware of criticism from primatologists such as Frans de Waal, director of Living Links, a division of the Yerkes
National Primate Research Center at Emory University in Atlanta, who has said
Tomasello underestimates the minds of chimps and overestimates the uniqueness
of human cooperation.
Nonetheless, Tomasello’s fellow scientists credit him with brave
experiments and ingenious insights. Carol Dweck, a professor of psychology at
Stanford University, who has done seminal research in child psychology and
intelligence, has called Tomasello “a pioneer.” Herbert Gintis, an economist
and behavioral scientist at the Santa Fe Institute, an interdisciplinary
science research institution, agrees. “His work is fabulous,” Gintis says. “It
has made clear certain things about what it means to be human.”
Every Chimp for Himself
Tomasello calls his theory of cooperation the Vygotskian Intelligence
Hypothesis. It is named for Russian psychologist Lev Vygotsky, who argued in
the 1920s that children’s minds do not automatically acquire skills, but
develop full human intelligence only through cooperative teaching and social
interactions. Tomasello applies this idea to the evolution of our species. He
proposes that as many as 2 million years ago, as climate swings altered the
availability and competition for food, our ancestors were forced to put their
heads together to survive.
Tomasello began his research career at Emory University, working with
apes at the Yerkes primate center. He acknowledges that chimpanzees, like
humans, manage complex social lives, solve problems flexibly, and create and
deploy tools. Nonetheless, “I take it as given that something is different,” he
says. “Humans are doing something on a different level.”
As Tomasello began to study the cognition of chimpanzees and other
great apes, he was influenced by pioneering child psychologist Jean Piaget, who
recognized that children see the world differently. “He looked at children as
if they were another species,” Tomasello says. “That’s the guiding image I
At the Yerkes primate center, Tomasello adopted an experimental method
that he would develop throughout his career: systematically comparing the
cognition of great apes and young children in head-to-head tests. Since the use
of language is an obvious difference between humans and chimps, he began by
looking at the precursors of speech. Great apes often communicate with
gestures. Babies point before they talk. Presumably our hominid ancestors also
gesticulated before they developed language. So Tomasello focused on pointing,
devising dozens of studies to explore how and when chimps and children point.
He found a major difference between the two species. By the time a
baby begins to point, at about nine months of age, she has already made several
sophisticated cognitive leaps. When she points at a puppy and looks at you, she
knows that her perspective may be different from yours (you haven’t noticed the
pup), and she wants to share her information—doggie!—with you.
“We naturally inform people of things that are interesting or useful
to them,” Tomasello says. “That’s unusual. Other animals don’t do that.”
Pointing is an attempt to change your mental state. It is also a request for a
joint experience: She wants you to look at the dog with her.
Chimps, by contrast, do not point things out to each other. Captive
chimps will point for humans, but it’s to make a demand rather than to share
information: I want that! Open the door! They do not understand informational
human pointing, because they do not expect anyone to share information with
them. In one of Tomasello’s experiments, food is hidden in one of two buckets.
Even if the experimenter points to where it is, the chimp still chooses
randomly. “It’s absolutely surprising,” Tomasello says. “They just don’t seem
to get it.”
“We naturally inform people of things that are useful to them. Other animals don’t do that,” Tomasello says.
In parallel experiments, children as young as 12 months have no
trouble understanding an adult pointing a finger at a hidden reward. To
understand pointing, Tomasello posits, you must form a “we intention,” a shared
goal that both of you will pay attention to the same thing. Chimps don’t point because they don’t think in terms of “we.” They
think in terms of “me.” “Cooperatively informing them of the location of food
does not compute,” he says. The chimpanzee world is egocentric: Every chimp for
The idea that chimps don’t work together appeared at first to
contradict what some biologists had observed in the wild. Chimps take turns
grooming one another, for example. They also form group hunting parties to
encircle and kill red colobus monkeys, a favorite food. But these behaviors
don’t require the kind of we intention that Tomasello was finding in even the
youngest humans. Grooming is a tit-for-tat activity that merely requires two
animals to alternate: Literally, I scratch your back, you scratch mine. There’s
no need to jointly focus attention.
Chimps can also hunt together without deliberately coordinating,
Tomasello reasons. If, during the chase, each chimp simply maximizes his own
chances of catching the prey, each will position himself where he thinks the
monkey will try to break out of the circle of predators. “This kind of hunting
event is clearly a group activity of some complexity,” Tomasello writes in his
book Why We Cooperate. “But wolves and lions do something very similar,
and most researchers do not attribute to them any kind of joint goals or plans.
The apes are engaged in a group activity in ‘I’ mode, not in ‘we’ mode.”
Still, it was hard to tell exactly what was going on from watching
wild animals. Lab experiments, where multiple animals can be tested in
controlled situations and their responses measured and quantified, could
clarify whether chimps even have a collaborative mode. Since collaboration
requires that you understand what someone else wants and thinks, Tomasello
explored whether chimps have what psychologists call “theory of mind,” or
insight into what another individual might be thinking.
The consensus was that apes did not have this mental ability, but
Tomasello and his team, including then undergraduate student Brian Hare, began
devising ape-centric experiments to test whether that was truly the case.
Rather than using psychology tests developed for human beings, as many ape
researchers did, they invented new tests that were more relevant to the
Chimps are hierarchical with an alpha chimp getting priority in
feeding. One experiment set a high-ranking chimp against a low-ranking one to
compete for food. The researchers hid snacks in such a way that only the
subordinate animal could see all the hiding places. When both animals were
freed to go after the food, the subordinate dove for the snacks that had been
hidden out of the high-ranking chimp’s line of sight. (Control
experiments showed that if both animals saw where the food was hidden, the
subordinate animal didn’t bother to approach the food.) The reasonable
explanation was that the low-ranking chimp modeled his rival’s thought process
in order to exploit his blind spots. He had a concept of what the other chimp
saw and believed—the basic definition of theory of mind.
This study and others like it in 2000 and 2001 led the group to
conclude that chimps do actually have insight into other individuals’ thoughts.
But they don’t use this ability to cooperate, as humans often do. They use it
“If you try to do something cooperative with a chimp—point out
something, show them where some food is—their attention wanders all over the
place,” says Tomasello. “But if you compete with them over food, they are
zeroed in like a laser. All their cognitive skills are on.” (If chimps
had a self-help bestseller, it would be titled, How to Outwit Rivals and Get
It’s not that chimpanzees are incapable of helping. De Waal describes
one instance in which chimps boosted an arthritic elderly troupe mate up to a
joint perch and used their own mouths to carry water to her so that she could
drink. De Waal says this is one of many examples of animal cooperation. He
predicts the claim that humans are unique because they collaborate to solve
problems “will drop by the wayside.”
Tomasello agrees that chimps sometimes assist each other and help each
other get food, under specific conditions that eliminate all possibility of
competition. In one of his experiments, conducted with wild-born chimps in a
Ugandan sanctuary, two chimps entered separate cages, with fruit placed in a
third. The first chimp knew from previous experience that he could not open his
own cage door to get to the food, but he could help the other animal by pulling
a chain that opened the door of the other cage. Here, with nothing to gain and
nothing to lose, eight out of nine animals pulled the chain so that the other
animal could get the fruit.
“It was a huge surprise to me,” Tomasello says. “My initial reaction
was: Damn. That doesn’t fit very well.” But the more he thought about it, the
more he realized that chimps, again, are acting individually, not in
cooperation with each other. “To help, you just need to know what the goal is,
and then if you’re motivated, you can help,” he says. “It’s not cognitively
very complicated.” Human cooperation, meanwhile, requires two or more people to
have insight into each other’s intentions, formulate a joint goal, assume
specific roles, and then coordinate their efforts. It demands cognitive
capacities that even the most helpful chimpanzees don’t possess.
In this case, the term “helpful” may be a little generous. Alicia
Melis, a former postdoc in Tomasello’s lab, explains that chimps work together
only grudgingly when it comes to obtaining food. In one of her experiments, a
board laden with food is placed beyond the reach of two chimps; they can get
the food only if each grabs one end of a rope attached to the board and they
jointly pull it. In her experiment, the animals worked together only if the
food was already evenly divided so that they did not have to compete over the
spoils. It also helped if they already got along.
“The motivation does not seem to be, ‘Let’s do this together,’ ” says
Melis, now an assistant professor of behavioral science at Warwick Business
School in the U.K. “It’s, ‘Let me try to do this alone, and if I can’t, we’ll
do it together.’ ”
Tomasello has discovered that young children, by contrast, find that
working together can be a reward all its own. When adults deliberately drop
objects in his experiments, babies of 14 months will crawl over to pick them up
and hand them back. Toddlers open doors for experimenters whose hands are full.
They do it without being asked and without being rewarded. Once they get the
idea that they are partnering, they commit to joint intentionality. If a
partner is having trouble, they stop and help. They share the spoils equally.
“They really understand that we’re doing this together, and we have to divide
it together,” Tomasello says.
Evolution of Collaboration
There are no fossils of ancient hominid brains or other physical
evidence that might tell us when and how our ancestors first put their minds
together to collaborate. Without such clues, the question of why we alone
became a collaborative species is difficult to answer, says Hare, who is now a
professor at the Center for Cognitive Neuroscience at Duke University.
“Figuring out what makes us unique is hard as hell,” he says. “But it’s much
easier than the next question, which is the real issue, the Higgs boson of
evolutionary anthropology: How did we get that way?”
In the absence of physical evidence, Tomasello proposes one possible
scenario. During the Pleistocene, about 1.5 million years ago, the climate
became very bumpy, with frequent temperature swings that forced our ancestors
to work together to access new sources of food. Perhaps we became scavengers,
joining forces to ward off bigger, tougher meat-eating competitors. Under these
circumstances, any genetic variation that made it easier to collaborate—maybe
by more accurately reading someone else’s intentions, seeing the whites of their
eyes, or simply being more relaxed about
sharing food—presumably would have helped those individuals survive, and would
have spread through the population.
“Figuring out what makes us unique is hard as hell. It’s the Higgs boson of evolutionary anthropology,” says Hare.
Hints as to how this might have happened emerge from a surprising
place: a fox-breeding farm in Siberia. In the 1950s the Russian biologist
Dmitri Belyaev was interested in how dogs might first have been domesticated.
He paired the most docile, friendly foxes he could find, then chose the
gentlest from each litter and bred them. In a mere 10 generations, the young
foxes acted like puppy dogs. The first time they met a human, they wagged their
tails and tried to leap into people’s arms to lick their faces.
The descendants of those foxes still live in a facility in Siberia,
where Hare, Tomasello’s former student, traveled in 2003 to test their
cognition. He found that fox kits that had spent less than 20 minutes total
around a person already understood human gestures, such as following a pointing
finger to find food. They did not have to be taught. “That blew me away,” Hare
says. His experiments suggested that as the foxes lost their fear of humans,
they became able to repurpose their cognitive abilities to a new human-focused
agenda: relating to us.
The foxes’ ability to read human social cues was now under strong
artificial evolutionary pressure, since only the friendliest animals got the
chance to breed. With Belyaev calling the shots, the foxes were competing
against each other to be more socially perceptive. The outcome, a few dozen
generations later, is a fox that understands human pointing. The small change
in temperament permitted a big advance in social intelligence.
Basically, we domesticated ourselves. We became less aggressive and more willing to share.
Hare and Tomasello suspect our ancestors went through a similar
process. Basically, we domesticated ourselves. When collaborating to find food
became essential because of changes in the climate or changes in the
competition, we became less aggressive and more willing to share. Aggressive
individuals, unwilling to cooperate, would starve and die out. Now that our
temperaments allowed us to put our minds together, we were able to develop communal
inventions like language and culture, and sustain these
innovations by teaching and imitating one another. The ability to crystallize
knowledge in inventions and traditions, Tomasello says, is what turned the
ordinary primate mind into an extraordinary human one.
“Through our collaborative efforts, we have built our cultural worlds,
and we are constantly adapting to them,” he writes in Why We Cooperate.
Us and Them
Today Tomasello is also looking through the prism of collaboration and
beginning to explain some of the uniquely dark and nasty things that humans do.
Maintaining a collaborative social structure encourages us to shun outsiders
and discipline nonconformists. It fosters groupthink—the urge to stifle
dissenting opinions in the interest of harmony and loyalty. Here, his research
connects with that of anthropologists and economists who study social norms and
other psychological underpinnings of group behavior.
Game theory models, which forecast how people behave when their
interests are in conflict, suggest that cooperation can only be sustained in
large groups if members punish anybody who freeloads or behaves selfishly. This
prediction was borne out in 2005 by an anthropological study of 15 societies,
mostly traditional small-scale communities, scattered around the globe.
But our tendency to enforce standards goes beyond simply ensuring
justice. The impulse to formulate social rules and punish rule breakers applies
to all kinds of situations, from whom we marry to how we dress. One possible benefit of these social norms is that
they help us quickly identify who is part of our in-group and who is not; they
also make it easier to collaborate more effectively. (If you hunt the same way
I do, it’s going to be easier for us to work together, and it’s also a sign
that you are a member of my group, someone I can presumably rely upon.)
The downside is that we also tend to blindly adopt arbitrary social
conventions. Unlike other great apes, we are fundamentally conformists,
Tomasello says. We form groups in which everybody dresses and talks the same
way, “and anybody who intentionally doesn’t conform, we wonder: What’s wrong
with them—do they not want to be one of us?” From this perspective, laws,
morals, and religious rules are simply larger and more institutional versions
of the impulse to police social norms, he says: “Human societies are just one
layer of cooperation, or incentives for cooperation, on top of another.”
The open question is whether being natural-born collaborators also
condemns us to be small-minded conformists who fear and distrust outsiders.
Tomasello is now exploring how children understand group membership, testing
how they act while wearing uniforms or being asked to work together. Kids
absorb social norms quickly, he has found, and they enforce them
enthusiastically. They can be little martinets. In one recent experiment,
3-year-olds are shown a game of pretend in which a pen is supposed to be used
as an imaginary toothbrush. A puppet held by a researcher then asks to join the
game, but draws with the pen instead. One child is outraged, barking, No! You must
brush the teeth! The child’s reaction demonstrates the basic urge to
impose social rules even when they are meaningless.
Ultimately, Tomasello’s research on human nature arrives at a paradox:
our minds are the product of competitive intelligence and cooperative wisdom,
our behavior a blend of brotherly love and hostility toward out-groups.
Confronted by this paradox, the ugly side—the fact that humans compete, fight,
and kill each other in wars—dismays most people, Tomasello says. And he agrees
that our tendency to distrust outsiders—lending itself to prejudice, violence,
and hate—should not be discounted or underestimated. But he says he is
optimistic. In the end, what stands out more is our exceptional capacity for
generosity and mutual trust, those moments in which we act like no species that
has ever come before us.
Kat McGowan is a contributing editor at Discover magazine and
independent journalist based in Berkeley, Calif., and New York City. She writes
about neuroscience, genetics, and other science that affects how we understand