Abstract

This paper revisits the concept of fiction employed in recent debates about the reality of theoretical entities in the philosophy of science. From an anti-realist perspective the dependence of evidence for some scientific entities on mediated forms of observation and modelling strategies reflects a degree of construction that is argued to closely resemble fiction. As a realist’s response to this debate, this paper provides an analysis of fictional entities in comparison to real ones. I argue that the distinction between fictional and real entities is reflected in their different relations toward their representations. This is particularly evident when it comes to the investigation of properties not explicitly given in a representation but that rely on knowledge external to it. A comparison of the resulting difference in the interpretation of fictional and real entities is then shown to provide guidelines for the assessment of when a realist claim can be made for model-based inferences to theoretical entities in science. At the end of this paper I advocate a pluralistic view on scientific realism by showing that representational pluralism, far from posing a problem for a realist interpretation of scientific practice, serves as an indicator for the reality of scientific entities.