Structural Realism and Abductive-Transcendental Arguments

Abstract

491Abstract The paper deals with an attempt to present an “abductive-transcendental”\rargument in favour of a particular version of structural realism (SR), dubbed\rIntermediate SR. In the first part of the paper the general structure of transcendental\rarguments is scrutinized with a close view on Kant’s original version and\rthe prospect of their abductive variation. Then the role of symmetries in modern\rphysics, especially symmetries without real instantiations and in particular gauge\rsymmetries is discussed. This is combined with a presentation of SR as a promising\rcurrent version of scientific realism. The discussion is supported by various arguments\rfrom gauge theories in modern physics. Intermediate SR, a realist position\rabout all and only structurally derivable entities located between the extremes of\rEpistemic and Ontic SR turns out as the best fit to our current fundamental gauge\rphysics and this finally leads to an abductive-transcendental reasoning concerning\rthis position. 501The first premiss is about the notion of experience: Experience means to recognize\rthe change of the variable in relation to the permanent or persistent. Take this\ras an analytic statement about experience and, hence, empirical science. Experience\rwould be impossible in both a world of absolute change or absolute persistence.\rThis relates to the second premiss: Symmetry is the general mathematical tool\rto analyze the interplay between change and persistence. Any empirical science\rmust allow, at its bottom level, a law-like description of the world in terms of\rsymmetry structures.\rThirdly: ISR is the minimal doctrine satisfying a realist commitment to empirical\rscience. This is of course not to say that ISR is the only reasonable realist doctrine\r(which would be an absurd armchair verdict), but that it is minimal as regards a\rscience of structurally defined entities.\rFrom this it follows that ISR is most plausibly the core of any realist position of\rfundamental empirical science. Here again the conclusion only presents ISR as a\rcore position which must perhaps eventually be embedded in a wider position in\rthe sense that more than structurally defined entities constituted by relational properties\rand intrinsic invariance properties are taken into account. The assumption,\rhowever, is that Occam’s razor (a further tacit premiss in the above argument) cuts\rin favour of ISR. Hence, as an inference from the most basic premisses about\rempirical science to the best realist explanation of its success today, ISR turns out\rthe most plausible and natural candidate.