Abstract

Nelson Goodman argued convincingly that in order to understand the representation relation one should dissociate it from the relation of resemblance because of the logical differences between the two concepts. Resemblance is reflexive and symmetric whereas representation is not. Furthermore, Goodman suggested that what lies at the core of representation is denotation. According to Goodman, if X represents Y then X must denote Y, but he recognized that by opting for an analysis of representation only based on this idea of denotation we run into problems. There are two reasons for this. The first reason is that X could be considered to be a representation of Y by mere stipulation. The second reason is that in many cases X represents Y but Y does not exist and thus there is nothing that X denotes. Both of these problems are important when addressing questions about scientific representation. First, we do not think that we do justice to scientific practices by thinking that scientific models represent by mere stipulation. Second, some of our scientific models represent what we often label ideal systems or ideal states of affairs, and if such systems do not exist in the actual world then it would also not make much sense to claim that our models denote such systems. I argue in this paper that there is a way to overcome these two problems and explicate “representation” by means of “denotation”.