Concluding Remarks
If the above considerations make sense, to be a selectivist scientific realist is to accept as truthful only the theory-parts that Criterion R+ picks out. I conclude with a few remarks about the selectivist stance proposed in this section.
(i) Criterion R+ picks theory-parts prospectively. Each part selected is both backed by strong reasons for accepting it and free of compelling specific doubts against it.
(ii) The theory-parts that R+ picks lodge most commonly at inferential levels below those of the highest theoretical postulates, as in the example provided by Fresnel's Core. The relevant point is that the selected parts are both substantive and clearly theoretical, i.e. placed at levels significantly above the empirical ground level. Commitment applies most easily to intermediate and low-level theory parts—theory-parts at the highest levels of current theorizing qualifying for comparatively weak levels of realist commitment. This restriction seems a sensible outcome, namely a realist stance that progresses “from the bottom up,” also one seemingly borne out by the history of science.
(iii) If, as it is now widely recognized, the contemporary realism/antirealism debate is primarily about the limits of ampliative inference in science, then Criterion R+ helps scientific realists affirm what antirealists deny, namely that scientific inference is robust enough to support claims about realities beyond the reach of the unaided senses.
(iv) Requiring approval by S2 enhances the explanatory import of realist commitment. The elucidation received by theory-parts in L+ flushes down to the ground level—to the phenomena that the theory in question predicts. In this way, the R+ selectivist (unlike the antirealist) makes the intended empirical domain less mysterious, more intelligible.