A model-theoretic realist interpretation of science

Abstract

My model-theoretic realist account of science places linguistic systems and the corresponding non-linguistic structures at different stages of the scientific process. It is shown that science and its progress cannot be analysed in terms of only one of these strata. Philosophy of science literature offers mainly two approaches to the structure of scientific knowledge analysed in terms of theories and their models, the "statement" and the "non-statement" approaches. In opposition to the statement approach's belief that scientific knowledge is embodied in theories (formulated in some (first-order) symbolic language) with direct interpretative links - via so-called "bridge principles" - to reality, the defenders of the non-statement approach believe in an analysis where the language in which the theory is formulated plays a much smaller role than the (mathematical) structures which satisfy that theory. The model-theoretic realism expounded here retains the notion of a scientific theory as a (deductively closed) set of sentences, while simultaneously emphasising the interpretative role of the conceptual (i.a. mathematical) models of these theories. My criticism against the non-statement approach is based on the fact that merely "giving" the theory "in terms of' its mathematical structures leaves out any real interpretation of the nature and role of general terms in science. Against the statement approach's "direct" linking of general theoretical terms to reality, my approach interpolates models between theories and (aspects of) reality in the interpretative chain. The links between the general terms of scientific theories and their interpretations in the various models of the theory regulate the whole referential process. The terms of a theory are "general" in the sense that they are the result of certain abstractive conceptualisations of the object of scientific investigation and subsequent linguistic formulations of these conceptualisations. Their (particular) meanings can be "given back" only by interpreting them in the limited context of the various conceptual models of their theory and, finally, by finding an isomorphic relation between some substructure of the conceptual model in question and some empirical conceptualisation (model) of relevant experimental data. In this sense the notion of scientific "truth" becomes inextricably linked with that of articulated reference, as it - given its model-dependent nature - should be.