Structure not Selection

Abstract This chapter argues that structural realism should not be thought of as a form of selective realism, because the latter involves criteria that ought to be applicable to our best current theories to select in advance what of their ontology will be preserved. Neither epistemic nor ontic structural realism (OSR) offer such criteria. Ontic structural realism is a metaphysics that also responds to ontological problems for scientific realism other than the problem of theory-change. These include the lack of a fundamental level and scale-relativity. The theory of real patterns addresses these problems, emphasizing the centrality of the idea of objective modal structure that makes OSR a form of realism and differentiates it from van Fraassen’s structural empiricism, making an account of causation and other modal aspects of scientific knowledge possible. The chapter concludes with some remarks about the relationship between structural realism and the historiography of science.