Date of Award
Doctor of Philosophy
The Duhem problem, as interpreted here, is concerned primarily with the ambiguity of scientific testing--falsifying experimental data can never uniquely determine the theoretical target to be falsified. By reformulating the Duhem problem within an epistemic decision framework recently proposed by Isaac Levi, it is hoped that the Duhem problem can be seen as a problem in epistemic decisions and hence can be solved decision-theoretically.;The motivations for adopting the decision-theoretic approach to scientific inquiry are discussed. Views of Popper, Carnap, Lakatos, Hempel and Levi are reviewed. The structures of practical decision and cognitive decision are critically compared. Epistemological concepts relating to epistemic decisions--contextualism, corrigibilism, epistemological infallibilism, local methodology--are examined. An analysis of the Duhem problem with reference to Duhem's and Quine's work, and major ways to combat the Duhem problem are presented. Furthermore, a detailed and critical exposition of Levi's epistemic model is given. Finally, a decision-theoretic treatment of the Duhem problem is proposed. The Duhem problem is interpreted as a problem in epistemic decisions. More specifically, the problem is transformed into a problem of choosing alternative corpora relative to corpus contraction. Some measure of the utility of the corpus is proposed. The utility of a corpus is understood here as the cognitive performance of the corpus which is defined in terms of its problem-solving (question-answering) capability relative to a domain of problems and cognitive objectives. Indeed, by utilizing the measure of corpus performance, it is argued that a ready solution to a version of the Duhem problem can be provided.
Ip, Po Keung, "Epistemic Decision And The Duhem Problem" (1983). Digitized Theses. 1277.