I am currently working on a proposal involving bonobos in a study on imitation and teaching. The bonobo, Pan paniscus, is the closest extant genetic relative of humans. The study employs a device I call the "Teaching Machine." True imitation is of interest because it requires a cognitive transformation of the actions or behaviors of another individual into a purposeful novel activity by the observer. This is a cognitive ability heretofore held as uniquely human, pending conclu sive evidence to the contrary.
In the literature on imitation, there is an accumulation of semantic and technical problems which has made progress difficult. The goal of my study is to take a fresh approach to the question of imitation by carefully devising an experimental architectur e which can provide apes with the tools and motivation to exhibit intentional teaching, and to provide me with the proper controls for making a reliable assessment of their behavior. If teaching is demonstrated, I can make a strong argument for the neces sity of the ability to imitate based on the social dynamic interconnecting teaching and imitation. This avoids the interpretive problems encountered by the many studies which have claimed to document imitation in great apes.
In this study, I will use the need for cooperation on an operant task as motivation for one animal to teach another to perform the task. To identify teaching, I will be comparing the interactions of the apes in three experimental pairings: naive-naive, experienced-experienced, and experienced-naive. The "teaching machine" apparatus will outwardly consist of two parallel levers which must be worked simultaneously for the delivery of a food reward. The levers are positioned in such a way that one animal cannot manipulate both at the same time. This creates the need for cooperation between two animals to successfully work the apparatus. When only one member of a pair knows how to extract food from the apparatus, this cooperative feature will provide the incentive for teaching. This approach allows the study of imitation to return to a more ecologically valid forum based on social information sharing, rather than focusing on the theoretical distinction s between learning mechanisms. The animals will be allowed to behave naturally without any attempt to analyze the behavior into mimimalistic single-action components.
The issues of teaching and imitation are key features of human social information transfer. The study of imitation and teaching in apes allows us to study this transfer in a simplified non-linguistic form.