The Problem of Unconceived Alternatives
Nearly a century ago, French physicist Pierre Duhem offered a characteristically lucid and provocative articulation of the challenge to the power of scientific methods to uncover theoretical truths about the natural world. What seems to have worried Duhem is the eliminative character of many important scientific inferences: often in science, perhaps even typically, we arrive at a decision to accept or believe a given theory because we take ourselves to have convincingly eliminated or discredited any and all of its proposed rivals or competing explanations of the available evidence. But as Duhem saw, such an eliminative inferential procedure will only guide us to the truth about nature if the truth is among these competitors in the first place. It is argued that eliminative inferences are only reliable when we can be reasonably sure that we have considered all of the most likely, plausible, or reasonable alternatives before we proceed to eliminate all but one of them (or, in the limiting case, simply rest content with the lone contender). But the history of science shows that we have repeatedly failed to conceive of (and therefore consider) alternatives to our best theories that were both well confirmed by the evidence available at the time and sufficiently plausible as to be later accepted by actual scientific communities. The historical record suggests that in science we are typically unable to exhaust the space of likely, plausible, or reasonable candidate theoretical explanations for a given set of phenomena before proceeding to eliminate all but a single contender, but this is just what would be required for such eliminative inferences to be reliable.
Oxford Scholarship Online requires a subscription or purchase to access the full text of books within the service. Public users can however freely search the site and view the abstracts and keywords for each book and chapter.
If you think you should have access to this title, please contact your librarian.