Friday, 26 January 2007
Idealization and Scientific Modelling
Demetris Portides, Department of Classics and Philosophy
University of Cyprus
12:05 pm, 817R Cathedral of Learning
Abstract: The problem of idealization could be divided into two components. The first component concerns the clarification of what the relata are in the theory/experiment relation. In other words, where and how do we set the dividing line between a description that belongs to the side of theory and another that belongs to the side of experimental reports (if it can be clearly set)? In the literature on models we can discern two conceptions of the relata: the structuralist and the non-structuralist views. In the structuralist conception the theoretical description is clearly distinct from other conceptual ingredients that are necessary for relating model predictions to experiment, whereas in the non-structuralist conception the model predictions are the result of a complex amalgamation of theory together with conceptual ingredients deriving from auxiliaries, which within the representational device, i.e. the model, cannot be clearly distinguished. The second component of the problem of idealization concerns the way by which a theoretical description is de-idealized as to bring it closer to the features observed to be present in its target physical system, i.e. as to improve its representational capacity. This is obviously related to the first component in the sense that how one addresses the issue of de-idealization depends upon her understanding of the relata in the theory/experiment relation. In the structuralist conception of the relata de-idealization can only be understood as a process that is used in order to reconstruct the data. If one, however, conceives the relata as in the non-structuralist view, then de-idealization must be understood as a process strongly tied to the quest for improving the representational capacity and accuracy of the model. I argue for a non-structuralist conception of the relata, and attempt to formulate a theory of de-idealization that explicates both theory-driven and phenomenological modelling.