Scientific Cognition: Hot or Cold?

Abstract

Since the emergence of the “strong programme” in the sociology of science more than a decade ago, sociologists and philosophers of science have been at loggerheads.¹ Whereas philosophers have traditionally attempted to explain developments in science using models of rationality, advocates of the strong programme claim that explanations of scientific change can be purely sociological, emphasizing the “interests” of scientists rather than evidential grounds for acquiring beliefs. My view is that the theories of scientific development offered by both sociologists and philosophers have been paltry, painting science with brush strokes far too broad to produce understanding of how science grows. In contrast, cognitive science, the alliance of artificial intelligence, cognitive psychology, and other fields, offers the possibility of developing much more detailed and powerful models of scientific change. I shall describe two computational models of inference, one in the traditional evidence-based philosophical mode, and one in the interest-based sociological mode. When such models are laid out in detail, it becomes an empirical question which model best fits particular cases in the history of science.