Abstract

The debates concerning scientific realism are extremely complex. Different philosophers have defined the key terms differently, making it difficult to trace coherent and well-connected lines of argument. In addition, the debate occurs on different levels: a general philosophical level, the level of science studies, and the level of philosophy of science. I examine major issues in these debates at all three levels and apply a method of analysis that reveals that what is really at issue is not so much the general philosophical problem of realism but instead a gradual shift in our most basic presuppositions about what reality is like and how to characterize knowledge. ;This shift can be understood as a move away from an objectivist orientation and toward a more relational emphasis. That is, objectivist realism can be understood as a version of realism that begins with the assumption that reality is composed of discrete entities and kinds that can be individuated in a stable and absolute way. The goal of knowledge then would be to identify these objects and their properties. Relational realism, in contrast, is a version of realism that emphasizes relationality over objectivism. A proponent of relational realism is willing to allow that the dividing of reality into component entities with stable individuation is not always possible---there is not one absolute and correct way to divide the world. Instead, there are different ways to divide the world depending upon the kinds of relations that are being studied. Therefore, under relational realism it is the relations among objects that are generally regarded as more important than merely the classification of objects and their properties. Furthermore, under relational realism, it is possible to take into account the relationships between knowers and the known. ;The change from objectivist realism to relational realism also suggests changes in epistemology and semantics. In particular, it becomes problematic to characterize knowledge solely in terms of language: the discrete structure of language continues to tempt us back into an objectivist framework. The model-based account of scientific theories serves as a better framework for characterizing knowledge under relational realism