Social learning from video demonstrations in chimpanzees (Pan troglodytes), children (Homo sapiens), and ravens (Corvus corax)
Abstract
The capacity for cumulatively complex and constructive technologies is a
hallmark of human cognitive ability. Many species use tools, yet the breadth,
adaptability, and inventiveness of human tool use distinguishes us. Such complex
technological adaptations require the causal understanding to invent and perfect new
techniques and the ability to copy perceived behaviours. This has led to considerable
research comparing the social learning abilities of chimpanzees and young children on
tool-using tasks, yet experimental studies directly investigating tool modification are rare.
The studies outlined in this thesis sought to assess the social learning abilities of
chimpanzees and children by manipulating both the complexity of a tool modification
method and the amount of information available in a demonstration. Video
demonstrations of conspecifics were used in lieu of live models, in order to manipulate
the quality and quantity of information directly. Both chimpanzees and children presented
with complete information about the modification process learned to combine two tools
together to make a more efficient tool significantly more than those provided with less
information. Unlike chimpanzees, children presented with a more perceptually opaque
method of tool modification (twisting and extending an internal rod) were also able to
socially learn the task, despite the fact that none of the children in the control condition
successfully solved the task. Both children and chimpanzees who solved the task after
seeing a demonstration also persisted in using the socially learned method two-weeks
later, even when it was no longer necessary. These results identify potent social learning
effects in both chimpanzees and children, however, children proved superior to
chimpanzees in observationally learning finer manipulative techniques.
This thesis also provides the first analysis of video stimuli in ravens. To assess the
feasibility of the methodology, ravens were first presented with different types of video
stimuli, varying in terms of the subject identity. Ravens showed a preference for video
footage of other ravens over different species of birds. In a second study, ravens who saw
a conspecific solve a two-step task were more likely to attempt a solution than those who
had not. This represents the first evidence of social learning from a video demonstration
in ravens and further supports the use of this medium to dissect social learning
mechanisms in a range of species.
Type
Thesis, PhD Doctor of Philosophy
Collections
Items in the St Andrews Research Repository are protected by copyright, with all rights reserved, unless otherwise indicated.