The influence of model-based biases and observer prior experience on social learning mechanisms and strategies

Abstract

This thesis investigated social learning mechanisms and strategies relating to the characteristics of a model (the individual transmitting the information) and the prior experience of an observer (the individual acquiring the social information) in children and chimpanzees. Experimental designs that mirrored naturalistic settings enabled an investigation of how social learning mechanisms and strategies were affected by: (1) the characteristics of a model, (2) the prior experience of an observer, (3) continued model demonstrations and (4) repeated observer interactions with the task. If models provided viable novel solutions then their characteristics seemed ineffectual upon children’s copying of these solutions. Yet the characteristics of the model did influence children’s copying of irrelevant actions; children who observed an adult reproduced more causally irrelevant actions than those who observed a child. Furthermore, when a known peer with higher, rather than lower, past-proficiency matched a child’s original solution the child was more likely to continue using this solution. Chimpanzees were biased towards touching the tool seeded by a known conspecific with higher, rather than lower, past proficiency but this bias did not affect which tool a chimpanzee successfully used. Both species showed an ability to learn multiple demonstrated methods of success within their corresponding tasks and to explore beyond demonstrated methods. It is argued that both species show more task-behavioural flexibility than previously thought and the implications for this in terms of cultural evolution are discussed.