Do chimps understand what Jon Stewart (or another chimp) believes?

Take a look at this video from last night's episode of Jon Stewart's "The Daily Show."

If you'd like, you can skip past all the political snark to the 4:47 mark to watch Jon bring cognitive psychology into prime time (or at least latenight cable)! That's right; you saw it: Jon Stewart mentioned the psychological concept of "object permanence" on national TV. Object permanence was introduced by Jean Piaget as a way of measuring the growing cognitive ability of children. Three-month-olds don't have it; most 6-month-olds do. More recently, researchers have investigated similar milestones in animals. Parrots, it turns out, have object permanence, as do chimpanzees. Insects don't.

But what about higher-order cognitive functions? Do chimps understand that others have thoughts distinct from their own? Humans understand this around the age of 1, but the evidence is less clear with chimps. Some chimps will use gestures alone to beg for food from a blindfolded human. Does this mean they don't "know" the human can't see them? Perhaps not, but normally a chimp doesn't expect to communicate with a human. When two chimps are in two separate rooms, but can see into a third room where food is being hidden, the subordinate chimp will behave differently if she knows the dominant chimp saw the food being hidden. This suggests chimps do understand that other chimps have different thoughts from their own.

Juliane Kaminski, Josep Call, and Michael Tomasello set up a more complicated competitive situation for both chimps and human children. Two chimps sat in separate rooms with windows so they could see each other and a table between their rooms, like this:

The table had a movable center with three inverted buckets, each capable of hiding a treat. Each chimp's view of the table could be blocked separately. For each task, the experimenter hid a piece of banana under one of the three buckets. In addition, each chimp had her own bucket which she knew contained a less appealing snack: a piece of apple. The children had a similar setup, except they played against an adult, and they weren't confined to chimp-proof rooms. The kids' appealing reward was a toy, and the less-appealing reward was a wooden block.

For the chimps, the game worked like this: While both chimps watched, the experimenter placed the banana under one of the buckets. Then the treat was either moved to a new bucket, or kept in the same place, while both chimps watched or while one had her view of the buckets blocked. Then the chimps got to pick which bucket they thought contained the treat. Only one of the two chimps was actually being tested, and the tested chimp always picked second--and she did not get to see her competitor making the choice. She always had the option of picking the guaranteed treat on the table next to her, or she could take a chance and go for the banana, a much more appealing treat.

There were four possible scenarios in each game: Both chimps saw the banana being moved or kept in the original bucket, or the chimp being studied saw the banana moved or kept in the original bucket while hidden from view of the competing chimp. How often did the chimps (and kids) try for the more appealing prize? Here are the results:

The 6-year-olds came closest to the optimal strategy. They generally didn't choose the better treat when the competitor saw the treat being moved (or not moved) from one bucket to another. Even though they didn't see the competitor choose a bucket, they guessed that the competitor would have already taken the treat, and therefore the best they could do would be to pick the guaranteed, lesser treat.

However, when the object had been moved from one bucket to another and the competitor didn't see the move, they picked the better treat more than 70 percent of the time, figuring that the competitor was unlikely to have guessed the correct location of the treat. When the treat was kept in the same location, even though the competitor didn't see what happened behind the occluder, they chose it less often, presumably because they figured that the competitor was most likely to believe that the treat was in the same spot it had been left in before.

Three-year-olds, by contrast, pretty much always chose the guaranteed lesser treat, presumably because they weren't sure what their competitor had done.

Chimps did somewhat better: they chose to go after the preffered treat significantly more often when they saw that their competitor hadn't seen it moved (or kept in the same place). However, their decision was the same whether or not the treat was actually moved. This suggests their understanding of their competitor's knowledge is not as sophisticated as a six-year-old's: they didn't behave differently when there was reason to believe that their competitor had a mistaken impression of where the treat was located.

Chimps, it seems, do have some idea what other chimps are thinking. They have less of an understanding of their competitors' mistaken beliefs. I wonder what Jon Stewart would say about that!

KAMINSKI, J., CALL, J., & TOMASELLO, M. (2008). Chimpanzees know what others know, but not what they believe Cognition, 109 (2), 224-234 DOI: 10.1016/j.cognition.2008.08.010