Rejected Posits, Realism, and the History of Science

Abstract

Responding to Laudan’s skeptical reading of history, an influential group of realists claim that the seriously wrong claims that past successful theories licensed were not properly implicated in the predictions that once singled them out as successful. For example, in the case of Fresnel’s theory of light, it is said that although he appealed to the ether he didn’t actually need to in order to derive his famous experimental predictions—as far as the latter were concerned the ether concept was “idle,” “dispensable” or worse. This view, developed by Philip Kitcher in the 1980s and subsequently supplemented by Jarrett Leplin and by Stathis Psillos, has received critical attention over the last decade, but more needs to be said on the subject—or so I suggest in this paper. I bring forward four converging argumentative lines to show how and why, from the days of Fresnel to at least two decades after the Michelson-Morley experiments, the ether functioned and was understood as an “indispensable” posit in physics. My first line draws from Fresnel’s actual deployment of the ether concept and the way he and his circle understood his theory’s achievements. The second line draws from the rise of surprising implication as an epistemological value and its impact on leading theorists in the last two-thirds of the century. The third line draws from discussions of the optical ether in end-of-century reports around 1900. The fourth focuses on entrenched metaphysical assumptions that persisted in the practice of physics until the advent of Special Relativity. Pulling these four lines together shows why attempts at synchronic identification of sound theory-parts (as advocated by Kitcher, Leplin and Psillos) are bound to fail, but also how realists might try to meet the challenge this creates.