Morgan Thompson (University of Pittsburgh)
I argue for a second type of robustness analysis that has not yet been expounded upon in the philosophical literature. The two types of robustness analysis are most appropriate when researchers are in different epistemic situations regarding the target system, have different success and failure conditions, are prone to misapplication in different ways, and follow different rules for defining the set of models explored. My account of scope robustness analysis better accounts for robustness analyses that explore growth principles in network neuroscience and biology where well accepted models of the target system already exist (e.g., C. elegans wiring diagram).
