What experience is required for acquiring tool competence? Experiments with two callitrichids

Cottontop tamarins, Saguinus oedipus, do not spontaneously use tools in the wild or in captivity. Yet, adults and naive infants with no experimental experience with tools show impressive sensitivity to their design features in tool choice experiments. Based on these findings, it has been hypothesized that tamarins are equipped with an innate mechanism for recognizing the functional features of tools (Hauser et al. 2002a, b, Animal Behaviour, 64, 299–311). An alternative explanation is that the initial exposure to one tool during training triggered their sensitivity to a suite of functionally relevant features during generalization. Here, we test this hypothesis, bypassing the training phase, and presenting tool-naive tamarins with a variety of tool choices, each pairing involving novel featural changes to the tool's design. We also tested experimentally naive common marmosets, Callithrix jacchus, a closely related species that, like tamarins, do not spontaneously use tools either in the wild or in captivity. Following the first experimental session, there was no evidence that either species selected tools based on their design characteristics. This finding forces a rejection of the hypothesis that tamarins are equipped with an experience-independent, innate representation of tools. Exposure to several more sessions generally failed to alter the tamarins' discrimination, but caused marmosets to preferentially pick the functionally correct tools, attending to relevant featural dimensions (shape, material). Overall, these results suggest that nontool users may have an innate mechanism that operates over object mechanics more generally (i.e. as opposed to tools specifically), and this mechanism is triggered on the basis of a relatively impoverished input.