Abstract

When philosophers of science speak of a "logic of discovery", what they mean is a systematic method for the generation of scientific hypotheses. The task of this dissertation is to examine arguments for and against a logic of discovery and to defend the possibility of a methodological logic of discovery as opposed to a psychological or epistemic logic of discovery. According to a methodological logic of discovery, there are systematic methods for generating hypotheses that promote the long-term goals of science, such as problem-solving success, probability, truth, or consilience. According to a psychological logic of discovery, there are systematic methods that are actually employed by scientists to generate hypotheses. According to an epistemic logic of discovery, there are systematic methods for generating hypotheses that provide essential justification for such hypotheses. ;The first chapter contains an examination of consequentialism, according to which the only means of justifying a hypothesis is by reference to its empirical consequences. The arguments for consequentialism only threaten an epistemic logic of discovery. The second chapter contains an examination of Laudan's claim that the only rationale for a logic of discovery is epistemic and that there is no epistemic rationale for a logic of discovery. Laudan draws the wrong lesson from the historical shift to consequentialism and his context of pursuit depends upon caricatures of both discovery and justification. The third chapter contains an examination of Popper's attempt to deny a logic of discovery by drawing an analogy between evolution by natural selection and scientific progress, but Popper fails to reject either a psychological or methodological logic of discovery based on this analogy. The fourth chapter contains an examination of Reichenbach's arguments against a logic of discovery, but Reichenbach's version of the discovery/justification distinction is orthogonal to the logic of discovery debate and his defense of induction implies an epistemic logic of discovery. The fifth chapter contains an examination of Hanson's attempt to provide a logic of discovery. Hanson's defense of plausibility as analogy and formal symmetry provides the best prospects for a methodological logic of discovery