The use of metaphors and analogical representations in social simulation models

Despite the increasing spread of computer simulations in sociology and social psychology, from a methodological point of view some issues remain open. One of these is related to the relationship between the data of a simulation, on the one hand, and the empirical facts, on the other. In this paper it will be shown how the plausibility of computational models is sometimes justified by means of metaphors and analogical representations. In particular, two classes of model are considered: the models based on cellular automata and those based on the “contagion spreading” metaphor, both of which form the basis of many successful simulation models. Metaphors and analogical representations provide a method for determining an ideal-typical situation, since, through an evocative reference to experiences, based on common sense and shared knowledge, they capture the essential events and eliminate those that are not interesting, and, at the same time, they consider the facts according to new points of view that are heuristically fruitful and meaningful.