Simulating Social Complexity

Part of the series Understanding Complex Systems pp 455-495


Evolutionary Mechanisms

  • Edmund Chattoe-BrownAffiliated withDepartment of Sociology, University of Leicester Email author 
  • , Bruce EdmondsAffiliated withCentre for Policy Modelling, Manchester Metropolitan University

* Final gross prices may vary according to local VAT.

Get Access

Why Read This Chapter?

To learn about techniques that may be useful in designing simulations of adaptive systems including Genetic Algorithms (GA), Classifier Systems (CS) and Genetic Programming (GP). The chapter will also tell you about simulations that have a fundamentally evolutionary structure – those with variation, selection and replications of entities – showing how this might be made relevant to social science problems.


After an introduction, the abstract idea of evolution is analysed into four processes which are illustrated with respect to a simple evolutionary game. A brief history of evolutionary ideas in the social sciences is given, illustrating the different ways in which the idea of evolution has been used. The technique of GA is then described and discussed including: the representation of the problem and the composition of the initial population, the fitness function, the reproduction process, the genetic operators, issues of convergence, and some generalisations of the approach including endogenising the evolution. GP and CS are also briefly introduced as potential developments of GA. Four detailed examples of social science applications of evolutionary techniques are then presented: the use of GA in the Arifovic “cobweb” model, using CS in a model of price setting developed by Moss, the role of GP in understanding decision making processes in a stock market model and relating evolutionary ideas to social science in a model of survival for “strict” churches. The chapter concludes with a discussion of the prospects and difficulties of using the idea of biological evolution in the social sciences.