Re: Underdetermination and Scientific Realism

First it must be granted that we are concentrating on scientific theories and not on mathematical ones. It seems reasonable that a theory of unobservables, to warrant assertability, must offer predictions of observables. If a theory cannot offer such testable predictions it is inefficacious and can be discarded, but if it can reliably do so then it is proven, at least for the moment. The proof of any scientific theory is qualified by Popper?s falsification principle, by which empirical evidence can only ever falsify a theory, never conclusively prove it (and this falsification is what theories should strive for). So if a theory can offer reliable predictions it is not proven but simply adopted through pragmatic considerations. This principle entails a certain kind of underdetermination of theory by evidence. But Quine radicalizes the extent to which a theory is underdetermined, and in the process he must discard one important part of Popper?s principle, namely that part by which it gets its name. Quine holds that evidence cannot conclusively prove a theory, but nor can it falsify a theory.

A theory does not normally itself imply predictions. Rather the theory (T) is given in conjunction with certain hypotheses (H) that together lead to a prediction (P). It may still seem natural to assume that if the prediction is shown to be false by the observation of data then the theory of unobservables must be false. However, in this case the theory itself can be defended in the face of any evidence to the contrary because the auxiliary hypotheses can be adjusted without detriment to the integrity of the theory itself. This can be shown clearly in the simple formal language of Propositional Logic, thus:

Premise 1: (T & H) É P

Conclusion: ¬(T & H)

The negation of the conjunction of T and H is all that follows logically from not-P. This alternative argument is invalid:

Premise 1: (T & H) É P

Conclusion: ¬T & ¬H

To conclude the negation of both T and H from not-P is to ignore the scope and order of the functions If one theory can be retained in the face of contrary evidence, then so can another theory. Given that this second theory may be incompatible with the first, the underdetermination of theories by evidence follows. The required adjustment of the auxiliary hypotheses (web of belief) may be great, but the retention of the theory is always possible if the need for such retention is sufficiently great. The underdetermination of theory by evidence need not necessarily lead to an antirealism about reality (or the objects that constitute it). David Papineau points out that there is nothing inherent in the underdetermination of theory by evidence to say that all competing theories are equally supported or equally plausible, in which case theory-choice may legitimately be facilitated by applying Ockham?s razor. Thus i agree that Nick's statement (b) is true; UT has nothing to do with considerations such as beauty and elegance - typical reasons for choosing between competing theories - because it is only pertinent as regards evidence. Antirealism is not entailed, but a weakened realism and a fallibilism certainly are. The closest Quine might come to antirealism here is if he formulated an ecumenical response to the question of which of the incompatible theories would be true; that is, if he held that both were true. In the end Quine adopted a sectarian response of the kind which implied that only one would be true. (This is an issue further confused by Quine?s deflationary view of truth).

The problem that i think remains, and which this version of UT brings out nicely, is the justification for issues such as beauty, ellegance etc as reason to choose between supposedly valueless theories. If issues such as these are all that can be relied upon to choose a theory then the theories are extremely value laden, and this brings problems for the scientific reaslist stance.