This volume argues that social science and expert systems offer much benefit to one another. Well established social science procedures for data collection and validation can improve expert systems development, and expert systems themselves can benefit the social sciences by providing an important new mechanism for improving modes of thinking. The authors introduce the concept of expert systems development as a model for the acquisition, representation and validation of knowledge about relatively limited domains. Case studies derive from the authors' own development experiences in the social sciences, both with normative and descriptive modelling systems.