Recently, the underdetermination thesis has reappeared on the philosophical agenda. While Pierre Duhem and Willard Van Orman Quine argued that the thesis — according to which for every given theory there is a contradicting alternative theory which equally matches with the same set of empirical data — follows from the structure of mathematized theories and their interdependence with other theories and auxiliary hypotheses, and by this is a sign for a mature science; the recent discussion focuses on underdetermination as part of science in its development. P. Kyle Stanford, for instance, sees underdetermination as the problem of unconceived alternatives and takes it as a serious threat to scientific realism. I will argue, however, that both discussions, the old and the new one, miss an important point about underdetermination: In discussing metaphysical implications of scientific theories and concepts they fall short of explaining the role these very concepts play within the scientific progress.
In my view a thorough discussion of underdetermination is a perfect example for William James’ understanding of verificationism: In order to understand scientific progress and the scope of theories we should not create metaphysical problems for the sciences but we should rather look out for the problems the different sciences are facing in their everyday business. By taking these actual problems philosophically seriously we will be able to fully cherish the role of pragmatic aprioris in science and we can see how these aprioris can actually change — without destroying the overall frame of the science in question. In other words: Following the pragmatic maxim we will make sense of underdetermination, and this sense will have consequences that make a difference.