Observational learning in chimpanzees and young children was investigated using an artificial fruit designed as an analog of natural foraging problems faced by primates. Each of 3 principal components could be removed in 2 alternative ways, demonstration of only one of which was watched by each subject. This permitted subsequent imitation by subjects to be distinguished from stimulus enhancement. Children aged 2-4 years evidenced imitation for 2 components, but also achieved demonstrated outcomes through their own techniques. Chimpanzees relied even more on their own techniques, but they did imitate elements of 1 component of the task. To our knowledge, this is the first experimental evidence of chimpanzee imitation in a functional task designed to simulate foraging behavior hypothesized to be transmitted culturally in the wild.

Contrary to recent scepticism, systematic use of an experimental approach suggested by HAYES & HAYES (1952) shows that chimpanzees can imitate novel actions. Two chimpanzees imitated 13 and 17 novel arbitrary gestures, respectively. They were first taught to reproduce 15 gestures on the command, "Do this!", and then presented with 48 novel items. Using a rigourous coding system, two independent observers correctly identified a significant number of the chimpanzees' imitations (p < 0.0001). We conclude that after a period of tuition chimpanzees can go on to imitate arbitrary actions, providing evidence of a basic cognitive capacity for perspective-taking and cross-modal matching.

No abstract

Social learning in 11 human-raised capuchin monkeys (Cebus apella) was investigated using an artificial fruit that was designed as an analogue of natural foraging problems faced by primates. Each subject observed a human model open each of 3 principal components on the fruit in 1 of 2 alternative ways ("morphs"). The capuchin monkeys reproduced, to differing extents, the alternative techniques used for opening 1 component of the task (poking vs. pulling while twisting out a pair of smooth plastic bolts) but not the other 2. From the subjects' actions on the bolt latch, independent coders could recognize which morph they had witnessed, and they observed a degree of matching to the demonstrator's act consistent with simple imitation or object movement reenactment (A learns from watching B how an object, or parts of an object, move). Thus, these capuchins were capable of more complex social learning than has been recently ascribed to monkeys.