The objectivity of mathematics (Q885536)

This article represents an extended and welcome engagement by one of the most prominent philosophers of mathematics working today with one of the most detailed discussions of objectivity. Shapiro takes as his focus the account of objectivity offered by \textit{C. Wright} in his [Truth and objectivity. Cambridge, MA: Harvard University Press (1992)]. Wright presented four different tests for whether or not an area of discourse was objective: (i) epistemic constraint, (ii) cosmological role, (iii) cognitive command and (iv) response dependence. Shapiro argues, with a few important exceptions, that applying these tests strongly supports the contention that mathematics is objective. Epistemic constraint applies to a domain when all truths in that domain are knowable. Shapiro argues that this fails for mathematics unless we make strong idealizing assumptions about what is knowable. The failure of epistemic constraint supports the view that mathematics is objective. The rest of the paper supposes that epistemic constraint is met by mathematics and considers in what sense (ii)--(iv) could also be met, thereby suggesting that mathematics is not objective. The most detailed discussion concerns cognitive command, which obtains just in case any two appropriately situated agents disagree on claim \(p\) only when there is a locatable reason to blame one of the agents (360). For mathematical claims, the applicability of cognitive command turns on delicate issues about the objectivity of logic and the status of axioms like \(V=L\) that extend common axiom systems like ZFC. Based on the view that ``foundational matters are decided on holistic grounds'' (367), Shapiro tentatively concludes that ``cognitive command fails'' (368) in such cases. However, rather than taking this to be evidence that these areas of mathematics are not objective, Shapiro suggests that here ``Wright's criterion is at odds with the underlying intuitive notion of objectivity''.