Novel Predictions: From Empiricism to Unificationism

Abstract

From Fresnel’s wave theory of light to Einstein’s general theory of relativity, the use of novel predictions has a long history in the method of science. Since predictions concern empirical matters, associated models of the method are usually empiricist ones. However, much of the recent philosophy of science shows a lack of emphasis on novel predictions. The central reasons include the general thesis of underdetermination of theory by evidence and the marginalization of novelty to a narrower issue in theory assessment, the prediction-accommodation distinction. In particular, novelty has become nothing more than code for methods classified as unificationist criteria of theory assessment. In this paper, I will extend Harker’s criticisms to a broader history of novel predictions in philosophy of science. I will then suggest a philosophy of science rooted in empiricist ideas, new experimentalism, to recontextualize novel predictions and their epistemological role. I do so in hopes rehabilitating novel predictions as the core of empirical methods. The literature known as new experimentalism offers a particularly promising context for attempting this because it concerns itself with empirical progress as analyzed through the many epistemic values of experiments. The guiding theme of new experimentalism is the theory-independence of experimental phenomena. Thus, it bypasses all unificatory methods and extra-empirical tools of theoretical science. Through this, I aim to provide a basic characterization of novel prediction as experimental interventions, contrary to its current unificationist formulations.