If natural selection implies a struggle for existence between competing individuals, how can cooperative behaviour ever become established? This apparent paradox has been tackled, over the years, by a number of increasingly sophisticated models that have emerged from a branch of mathematics called ?game theory?. Most models assume that cooperation is an all-or-nothing affair: either individuals cooperate, or they don?t. This is unrealistic, say Gilbert Roberts of the University of Newcastle upon Tyne, and Thomas N. Sherratt of the University of Durham, UK, who present a refinement to the cooperation models in the 9 July 1998 issue of Nature
The game most frequently used to study this problem is known as the prisoner?s dilemma. The story is that two prisoners are in a cell with a high window, too high for either of the prisoners to reach on their own. If they refuse to cooperate, neither of them has a chance to escape. But if they both cooperate, one can give the other a leg-up, and that prisoner can then pull his helper up to the window, so they both escape.
This is a preview of subscription content, access via your institution