Observability and Mathematics: The case for logical weakness in the name of epistemic security (original) (raw)

This paper challenges the scientific realist claim that science aims to provide a truthful account of the world, contending that this is too strong a logical commitment, which condemns science to unnecessary and persistent failure (Mortensen and Burgess, 1989: 47). Rather, this examination defends van Fraassen’s (1980) weaker aim for empirical adequacy, suggesting he is right to focus on observability when assessing the epistemic security of scientific theories through his constructive empiricist (CE) philosophy of science (1980: 12). However, van Fraassen’s account of observability has been reproached for its seeming incoherency; Gideon Rosen’s (1994) ‘transcendental’ charge against van Fraassen’s apparent commitment to abstract, mathematical objects, asserts that CE undermines the very essence of empirical adequacy. ‘[His] rejection of “metaphysics”’ Rosen states, ‘in fact presupposes a fair dose of the metaphysics it purports to do without’ (1994: 143). How, then, can van Fraassen’s observability distinction survive? This investigation looks to reconcile Rosen’s critique with CE by synthesising elements of Bueno’s (1999) philosophy of mathematics, which is a take on da Costa and French’s (1990) notions of ‘partial relations’, ‘partial structures’ and ‘quasi-truth’, with Hasok Chang’s (2005) ‘humanist empiricism’. In doing so, this essay argues that Chang’s ‘quality-based’ concept of observability can partner with Bueno’s model, thus motivating CE to dispatch its apparent defects. The conclusion will support Chang’s expanded definition, by extending Bueno’s ‘partial structures’ analysis to unobservable phenomena other than mathematics. It argues that his logically weaker, yet epistemically safer account of science does not pose an ontological threat to scientific practice; it preserves the constructive empiricist’s pragmatic commitment to interpretation, without either snubbing realist intuitions of existence, or problematizing the key features of ‘openness’ and ‘incompleteness’ of scientific knowledge (Bueno, 1999: 480).