The study of cognitive science is just monkey business, at least when the experiment subjects are Rhesus monkeys.

Jonathan Flombaum GRD ’08, a graduate student in psychology, and Laurie Santos, a psychology professor, recently conducted a series of experiments testing whether Rhesus monkeys can understand what another being can and cannot see, and concluded that monkeys do posses this ability.

The researchers ran six experiments with the same basic setup, Flombaum said. In each trial, two identically dressed experimenters stood equidistant from a monkey. Each experimenter had a grape in front of him, but while one experimenter could see the grape, the other could not because his body was turned away from the grape or there was a barrier in front of his eyes.

In every experiment, the monkeys spontaneously, without training, stole the grape from the person who could not see the fruit, Santos said.

“Contrary to what we might have thought based on other experimental evidence, monkeys do seem to reason about what individuals can and can’t see,” she said. “They do share our ability to reason about the minds of others.”

Santos said the capacity to infer what other people can see by where they are looking and whether their sightlines are obstructed is part of what psychologists call “theory of mind.”

“[Theory of mind is] the ability to understand that another individual has his or her own mental states — his or her own beliefs and intentions,” said David Leiberman ’06, who conducted some of the studies.

Humans constantly use theory of mind, Santos said, often without realizing they are using it. For example, when people watch a soap opera, they interpret the events in terms of the mental states of the characters, she said.

Flombaum said the question the experiment explored was whether monkeys had theory of mind abilities.

“For a long time it really seemed like monkeys and apes didn’t know much about not only what others can know but what they can see,” he said.

Previous studies of monkeys’ theory of mind were not conducted in environments simulating natural competitive situations, Flombaum said.

Cayo Santiago, an island off the coast of Puerto Rico, is home to 800 research monkeys, Santos said. The monkeys have experience with humans and the food they eat, as the island is frequently used for research. The island is inadvertently set up so the monkeys are in competition with the researchers for the researchers’ lunches, she said. The Yale group capitalized on the pre-existing competition to test when the monkeys thought they could steal food from the experimenters.

Flombaum said the group is currently determining whether monkeys can infer mental states more abstract than sight, specifically whether monkeys can infer beliefs.

When psychologists test whether toddlers can represent beliefs, Leiberman said they test whether the children understand that another person can have false beliefs by using an experiment called the Sally-Anne test. In this test, the child watches a puppet show with two characters, Sally and Anne. First, Sally watches as Anne places a cookie in one of two containers. Sally then leaves the room, and Anne moves the cookie into another container. Sally reenters the room, and the children are asked where Sally will look for the cookie. Until age four, Leiberman said, children say Sally will look in the second container. He said they are not able to deduce that Sally has the false belief that the cookie resides in the first jar.

The group used a similar experiment on the monkeys to deduce whether monkeys can undertsand false beliefs, Leiberman said. The group built a ramp with two boxes, one higher on the ramp than the other. In the test, the experimenter placed a grape in the top box and, invisible to the monkey, triggered the grape to roll down the ramp.

In one condition, he said, the monkey saw the experimenter watch the grape roll down the ramp. In the other condition, the monkey saw that the experimenter did not watch the grape as it rolled down the ramp. In this experiment, the monkeys only tried to steal the grape from the bottom box when the experimenter had not seen it roll down the ramp, Flombaum said.

“This was our way of exploring whether monkeys understand that another being, even of another species, can have a false belief, and we find that they are able to understand this,” Leiberman said.