University of Southampton LogoProfessor Simon Cox
Home Biography Research Publications Contact Me

Other Areas of Research


Research Studentships


  Cooperative Behaviour and Viewshed analysis in Archaeology

There are three central unanswered questions in anthropology, which may be summarised as “Why isn’t Homo sapiens hairy?”, “Why is sex fun?”, and “Why are primate brains so big?” This work addresses the third of these. The brain is an expensive organ to maintain: in humans it accounts for 2% of body volume, but consumes more than 20% of the body’s resources. This is evidence for clear selection pressures favouring its increased size. Many possible factors have been suggested. For example, frugivores (fruit eaters) have significantly larger brains relative to body weight than folivores (leaf eaters) to allow them to monitor ephemeral resources such as fruiting trees. However it has recently been suggested that there is some evidence for a positive correlation between neocortex ratio and group size in living primates. The neocortex is the “thinking part of the brain” and is considered to be “the main anatomical index of cognitive capacity.”

We have constructed a model for cooperation in groups, in which players are able to base their future play on information stored about previous experiences. The results from this model show that players using a larger database of information are favoured in larger groups. This is consistent with the conjecture that living in larger social groups was one of the selection pressures which favoured increased brain size in primates.

Our investigations focus on the conditions required for the stable evolution of cooperation in large social groups based on the repeated Prisoner’s Dilemma Game. Using a parallel computational simulation we show that cooperation can be sustained in large groups when there are sufficient interactions on each round, and players are able to base future play on their observations of other players’ past actions. Of particular interest is a bifurcation in the model when a large number of interactions may be made on each round: sometimes the group sustains global cooperation, other times defection is rife. This work is in collaboration with Mathematics and Archaeology.

Percentage of cooperation in group as a function of group size (N) and number of interactions per individual (v)


Cox, S.J., Sluckin, T.J., and Steele, T.J., 1999. Group Size, Memory, and Interaction Rate in the Evolution of Cooperation. Current Anthropology. 40 (3) 369-376.

Back to Research ListPrevious RecordNext Record

Back to Top