Approaches to Scientific Modeling, and the (Non)Issue of Representation: A Case Study in Multi-model Research on Thigmotaxis and Group Thermoregulation

Abstract

Recent contributions to the philosophical literature on scientific modeling have tended to follow one of two approaches, on the one hand addressing conceptual, metaphysical and epistemological questions about models, or, on the other hand, emphasizing the cognitive aspects of modeling and accordingly focusing on model-based reasoning. In this paper I explore the relationship between these two approaches through a case study of model-based research on the behavior of infant rats, particularly thigmotaxis (movement based on tactile sensation) and temperature regulation in groups. A common assumption in the philosophical literature is that models represent the target phenomena they simulate. In the modeling project under investigation, however, this assumption was not part of the model-based reasoning process, arising only in a theoretical article as, I suggest, a post hoc rhetorical device. I argue that the otherwise nonexistent concern with the model-target relationship as being representational results from a kind of objectification often at play in philosophical analysis, one that can be avoided if an alternative form of objectification is adopted instead.