Imitation or something simpler? Modelling simple mechanisms for social information processing

In the pine forests of Israel, black rats Rattus rattus have hit upon a novel feeding technique. They strip the scales from pine cones to obtain the nutritious seeds inside (Terkel, 1996). The behaviour seems to be socially learned in some way, rather than genetically inherited: Terkel has shown that rat pups will learn to strip cones if they are born to a naive mother but then fostered to an experienced one. What mechanism is implicated in the transmission of this behaviour? It could be that rat pups are genuinely imitating their mothers, but this would mean that they are solving a complex correspondence problem in translating novel visual input—from observations of cone manipulation—into appropriate motor outputs. Might a simpler mechanism be involved? The social transmission of pine cone stripping behaviour is only one example of the general strategy of gaining information from the behaviour of one’s conspecifics. Every animal is constantly facing the problem of what to do next: given the various cues that it can detect in its environment at any one time, what behaviour is most likely to further its ultimate goals of survival and reproduction? In all but the most solitary of species, the cues available from the environment will include information about the behaviour of other animals. Investigation of the general question as to how animals might learn from or be otherwise influenced by the behaviour of their conspecifics has often been overshadowed by a focus on the specific issue of imitation. This emphasis has been around for some time: Galef (1988), in a historical review, points out that the debate over whether nonhuman animals are capable of imitation dates back at least to the dispute between Darwin and Wallace over whether the human mind is of evolved or divine origin. One product of this long controversy has been a plethora of definitions of imitation (see Zentall and Galef, 1988; Heyes and Galef, 1996), which we lack the space to adequately review. For the purposes of our argument, we will simply take true imitation to be the goal-directed copying of another’s behaviour. As Tomasello (1996) notes in his discussion of “imitative learning” (p. 324), successful imitation—in the sense we are interested in—requires not only perceiving and reproducing the bodily movements of another, but understanding the changes in the environment caused by the other’s be-