As humans, we like to think of ourselves as special. Especially when it comes to our intelligence. New research from Harvard University shows that our ability to infer the intentions of others is an ability that arose as far back as 40 million years ago. The research showed that three different types of primates (New World monkeys, Old World monkeys and apes) could infer rational, goal-direction actions just like humans. Not so special now, are you!

More details about the study after the jump.

We live in a chaotic world, constantly piecing various stimuli together to form a complete picture of our environment. An important part of this task is to make inferences about the world around us. For instance, if I see an ice cream cone lying on the ground and see you crying, I may infer that you are crying because you just dropped your ice cream. If, however, I see a large wooden stake protruding from your leg, I would probably infer you are crying from pain and not crying over the fallen ice cream.

Inferences are a result of joining stimuli together to form a model of the world. Until recently, it was thought to be a distinctly human ability. Humans gain this ability around the age of one. When infants watch an experimenter illuminate a box with her head, they will only mimic this behavior if they see the experimenter’s hands are empty. If the experimenter’s hands are full, the infants used their hands instead of their head to illuminate the box. It is probable that the infants inferred the intention was to illuminate the box with the experimenter’s hands, but since they were full and unable to do the job, the experimenter used her head instead.

In past research, monkeys have demonstrated the ability to distinguish accidental from purposeful actions. For instance, a past study showed chimpanzees becoming visibly frustrated when experimenters appear to tease by offering food then withdrawing it. The chimps do not get frustrated, however, when the experimenter offered the food and clumsily dropped it. Both actions give the same result, no food for the chimps, but only the teasing caused frustration. It is likely the chimps assumed the clumsy handling was an accident whereas the teasing was a purposeful event.

The research team wanted to see if primates are capable of differentiating against purposeful and accidental actions. Crucially, they wanted to see if the primates could infer the goal of the human experimenter while performing these actions. All testing used three different types of primates representing the major groups - New World monkeys, Old World monkeys and apes.

The first test presented the primates with two containers, one hiding food. The experimenter performed one of two actions in front of the primate. He either intentionally grasped the container with his hand or performed an “accidental” hand flop, palm up and in a manner that appears non-goal-directed to humans. After some statistical wizardry, it was found that the primates selectively inspected the container that was associated with an intentional hand grasp far more often than containers associated with accidental hand flops. This test showed strongly that primates could differentiate between accidental and purposeful actions.

The second test was similar to the one performed on human infants. In this test, was in either a “hands-occupied” situation or a “hands-empty” situation. In hands-occupied, the experimenter was either holding an object with both hands or holding an object with one hand, the other hand being behind his back. In hands-empty, the experimenter held an object with just one hand, the other hand being free to manipulate the containers.

For both situations, the experimenter touched the containers with his elbow. All the primates inspected the targeted container more often after observing the hands-occupied situation rather than the hands-empty situation. This held true even for the hands-occupied situation where the experimenter was holding a hand behind his back. The primates were rationally inferring that the experimenter was pointing to the container with an elbow only because his hands were unavailable (either full or hidden). It should be noted that none of these primates naturally use their elbows as pointing devices nor were they ever trained to use their elbows. In addition to inferring the intention of the human experimenter, they were doing so with a device alien to themselves (pointing with an elbow).

These two tests show that primates can not only differentiate between purposeful and accidental actions but also infer the goal of an human experimenter. The primates showed there were capable of understanding goal-driven actions (pointing to food) and able to infer the method of pointing to the food (by using an elbow) was the result of hands being full. This is a very human trait and pushes back our self-perceived specialness. It is common to point to things with your elbow if your hands are full. Demonstrating that primates understand this too blurs the line between our species and others.

References

Justin W, David G, Brenda P, Marc H. The Perception of Rational, Goal-Directed Action in Nonhuman Primates. Science (7 September 2007), 317(5843):1402-1405. 10.1126/science.1144663