Majority-biased learning: In humans and chimpanzees knowledge is transmitted within a group by means of a majority principle

Researchers of the Max Planck Institute for Evolutionary Anthropology in Leipzig and the Max Planck Institute for Psycholinguistics in Nijmegen have now discovered that chimpanzees are more likely to copy an action performed by a large number of individuals than an action that was performed more frequently. Two-year old children consider both the number of individuals and the frequency of the action demonstrated. For orangutans, however, none of the factors play a role.

In many animal species, behaviours and strategies are passed on from individuals to their conspecifics and potentially across groups by social learning. In chimpanzees and orangutans, whose behavioural repertoires differ from population to population, knowledge is also "transmitted" amongst individuals. In their current paper, researchers Daniel Haun, Yvonne Rekers and Michael Tomasello of the Max Planck Institutes for Evolutionary Anthropology and Psycholinguistics show how human children and chimpanzees pass on knowledge through social learning.

Initially, the researchers wanted to find out whether children and apes are more likely to copy a behaviour that has been demonstrated more often or one that has been demonstrated by more individuals. In the relevant experimental setting, 2-year-old children, chimpanzees and orangutans could receive a reward from an apparatus consisting of three differently coloured subsections if they dropped a ball into a hole. Four individuals then demonstrated an action: One individual dropped a ball into the same section three times; the three others -- one after the other -- dropped their balls into another section. Finally, the observers were also asked to drop a ball into one of the three sections. The result: Most of the chimpanzees and 16 children chose the section that the majority of individuals had also chosen. Orangutans appeared to select a section quite randomly.

In the second part of the study, the researchers analysed whether the frequency with which a subsection was chosen by the demonstrators had an influence on the result. The set-up was similar to the previous test, with one exception: now it was only two children, chimpanzees or orangutans who demonstrated an action. One individual dropped three balls into one of the coloured subsections and for doing this received one reward per ball. The second demonstrator dropped one ball in a differently coloured section and received one award. The result: Chimpanzees and orangutans seemed to choose randomly whereas most of the children chose the subsection into which more balls had been dropped.

"Taking the results of the two studies together, chimpanzees seemed to consider the number of demonstrators more strongly than the number of demonstrations when deciding which information to extract from their social environment. Children considered both. Orangutans considered neither," says Daniel Haun. Interestingly, children and chimpanzees copied the majority behaviour while orangutans did not. One possible explanation: Contrary to humans and chimpanzees, orangutans live together in lose group structures. Social learning beyond the mother-child-relationship might therefore not play an equally important role.