Structural correspondence between theories and convergence to truth (Q542129)

In the first section, the author presents the opposition between scientific intrumentalism and scientific realism, the justification of the latter in terms of the no-miracles argument (NMA), and two counterarguments of the NMA, namely, the pessimistic meta-induction and the no-speculation arguments. The NMA, proposed by Putnam, claims that ``the empirical success of contemporary scientific theories would be a sheer miracle if we would not assume that their theoretical superstructure, or ontology, is approximately true.'' In the second section, the author presents a correspondence theorem which, under certain conditions, allows one to infer the partial truth of a theory \(T\) from its empirical success, relative to a theory \(T^\ast\) which preserves \(T\)'s empirical success and is assumed to be at least as close to the truth as \(T\); essentially, this result states that, given a \(T\)-theoretical expression \(\varphi(x)\), there exists a \(T^\ast\)-theoretical expression \(\varphi^\ast(x)\) with the property that, whenever a system \(x\) is exposed to the circumstances in one of the finitely many subdomains assumed to partition the domain of discourse, then \(\varphi(x)\) holds iff \(\varphi^\ast(x)\) holds. In the third section, the author illustrates his result with \(T\) referring to the phlogiston theory, \(T^\ast\) to the generalized oxygen theory of combustion, \(\varphi(x)\) to pholgiston-richness or dephlogistication, and \(\varphi^\ast(x)\) to electropositivity or donation of electrons, respectively. The interpretation of the result is that, though \(T\) is theoretically incompatible with \(T^\ast\), \(T\) has got a right model about \(\varphi\)'s outer structure, though it has got a wrong model about \(\varphi\)'s inner structure. In the fourth section, the author uses the transitivity property of his notion of correspondence and discusses three possibilities: that a sequence of theories make progress and converge to an `ideal theory' \(T^+\), or that two sequences of theories make progress and eventually merge and converge to \(T^+\), or that two sequences of theories make progress and both converge to \(T^+\) although all members of one sequence are incompatible with all members of the other sequence. In the last section, the author explains how this correspondence theorem justifies only a form of minimal realism, according to which an approximately true theory describing the external reality in a humanly accessible way is possible -- independent of whether humans will ever find this theory. He discusses why the form of realism he defends might be too weak to a metaphysical realist, and still too strong to a non-realist.