Models and Theories in Science

Abstract

An important aspect of science is the construction of models and theories. Philosophy of science aims to elucidate this practice by asking various questions, such as: “what is a theory?”, “what is a model”, “how do models and theories relate to one another?”, and “how do models and theories relate to the world?” The so-called syntactic view of theories, which originated in the tradition of logical empiricism and logical positivism in the 1920s, construes scientific theories as axiomatized logical calculi whose nonlogical terms are interpreted in terms of observables. This view came under criticism after World War II and was eventually supplanted with the so-called semantic view of theories. According to that view, theories are sets of models, where models are construed as nonlinguistic entities that relate to reality via either a set-theoretical mapping (such as isomorphism) or similarity. A common denominator of both views is that they see models as subordinate to theories. The syntactic view sees models as alternative interpretations of a calculus, which is primarily of pedagogical interest. The semantic view sees them as being the building blocks of theories. In parallel to these schools of thought, there was always a strand of research focusing on the practice of science, on case studies and the methods in specific scientific disciplines, rather than on overarching philosophical concerns. Heterogeneous in character and orientation, what binds projects in this tradition together is the belief that large parts of science are not in the business of devising exact and all-encompassing theories but rather use a variety of different techniques and ingredients to construct models that are locally adequate. Models are now seen as the center of scientific attention, and theories are relegated to the status of a tool (among others) for model construction. The beginnings of this tradition can be traced back to the 1920s; it gained prominence for the first time in the 1960s and blossomed in the last two decades of the twentieth century. In more recent years, new questions have come into focus, in particular the issues of scientific representation, model-word relations (including idealization and analogy), the use of data, and the role of computer simulation in both modeling and theorizing, the use of models in scientific explanations and scientific understanding, and the problem of multi-model situations. This article provides a guide to these intellectual traditions. In doing so, it sets aside a number of related issues, in particular scientific realism, confirmation, and the application of mathematics.