The Emergence of the Covering Law Model and What We Might Learn from It

Fons Dewulf (University of Ghent)

Abstract:  Until 1948 scientific explanation was not a topic of interest in logical empiricist philosophy. After the publication of Hempel’s and Oppenheim’s “Studies in the Logic of Explanation” in 1948 it gradually became a major research interest in philosophy of science. In this presentation I reconstruct how Hempel came to be interested in explanation through his contact with American philosophers during the Second World War, notably Ernst Nagel, John Hospers and Charles Stevenson. I show how Hempel’s DN model of scientific explanation both originated in but also significantly diverged from the anti-explanatory understanding of scientific inquiry present in the writings of Carnap, Neurath and Frank, who were inspired by the work of Pierre Duhem and Ernst Mach. Hempel - I claim - broke with the positivist understanding of science in a significant way by conceiving explanation as an aim of science that is distinct from description. I focus on two central assumptions which guided Hempel’s shift in 1948 – assumptions which still shape the debate on scientific explanation today. First, Hempel assumes that there are explanations in science with a shared, uniform structure which can be articulated by the philosopher of science (the objectivist assumption). Second, Hempel believes that his philosophical articulation of scientific explanation should be descriptively adequate, i.e. it should account for and be constrained by cases identified as paradigmatic for the explanatory success of science (the descriptivist assumption). In the anti-explanatory tradition of Duhem and Mach both assumptions were put into question. Duhem and Mach argued that explanation was not an aim of scientific knowledge and that, consequently, there was no given uniform explanatory structure within science to be articulated by a philosopher. This entailed a non-explanatory interpretation of paradigmatic explanation-talk in (the history of) science. Instead of choosing which explanation-talk was exemplary and thus constrained a philosophical model of it, Duhem and Mach showed possible strategies to interpret such talk from an anti-explanatory perspective. I argue that Hempel’s work on explanation did not address the concerns from this anti-explanatory tradition properly, but instead glossed over them. Possibly, this also holds for most contemporary theories of scientific explanation which start from the objectivist and descriptivist assumptions.