Formats of Representation in Scientific Theorizing

Paper tpresented at the international conference Model-Based Reasoning in Science and Technology (MBR09_BRAZIL), held at the University of Campinas (UNICAMP), Brazil, December 2009This paper is intended to sketch out the definition of a methodological tool - the notion of a format of representation - for the study of scientific theorizing. Theorizing - seen as hypotheses forging and test-ing as well as teaching and communicating - always implies reasoning with and sometimes literally manipulating devices, such as diagrams, equations, graphs, schematic drawings, etc.: in order to draw inferences, one has indeed to reason with some concrete device, even if merely mentally represented. These devices are at the same time computational and representational tools. The notion of a format is intended to help us analyze the various forms under which such devices are displayed, and the influ-ence such variety has on reasoning. Acknowledging that representations are "formatted" means that two representations can contain the same information though convey it in different ways, and that such changes matter to the inferential processes one can draw from it. The notion of a format is intended to characterize the differences between various types of representational devices used in scientific theo-rizing, from the point of view of the inferences they enable cognitive agents to do. By briefly presenting two case studies (the various formulations of Classical Mechanics, and the case of Feynman's dia-grams), I will try to show that an analysis of the different formats that are used in scientific practice should give us a way to understand the double aspect of theorizing (computation and representation)