Abstract

In contrast with the development of big theories in the context of social sciences, there is nowadays an increasing interest in the construction of simulation models for complex phenomena. Those simulation models suggest a certain image of social sciences as a kind of, let us say, "patchwork". In that image, an increase in understanding about the phenomena modeled is obtained through a certain sort of aggregation. There is not an application of sound, established theories to all the phenomena of a certain kind, but an aggregation of the structures supposed, and of the results obtained, when particular systems are modeled. The recent case of the "El Farol Bar" problem, and the models built in order to face this problem, are a good example of this. We will analyze that case, trying to make clear what would be implied by the image above mentioned. Special attention will be paid to the need to take seriously the notion of a bounded rationality, linked to the special circumstances generating each decision problem, and to the existence of an irreducible pluralism of models.