Are model organisms like theoretical models?

Models in science are often thought of as surrogates for their intended target systems: one learns about the target indirectly by studying the model and then extrapolating the findings to the target. This analysis is often thought to cover both theoretical models and concrete models systems – simulated economies stand in for actual economies, simple laboratory organisms stand in for more complex organisms and so on.

In this talk I argue that model organisms and other concrete models are in an important sense not like theoretical models. The view that I defend is that theoretical models expand one's inferential abilities by encoding assumptions about a target system in an external medium (like a set of equations on a piece of paper, computer code in computer's memory etc.) such that one can draw conclusions unaccessible to unaided cognition by the application of some explicit inference rules to the modeling assumptions. The reliability of the conclusions rests on the adequacy of the modeling assumptions and the validity of the inference rules – i.e. there is no extrapolation from a model system to the target system. By contrast, the use of model organisms and other concrete model systems involves and empirical extrapolation from the model to the target.

Based on this paper.