Abstract

The Architecture of Reason aims to develop a unified theory of theoretical and practical rationality along foundationalist lines. The structure of practical reason, Audi argues, is closely parallel to that of theoretical reason: just as the rationality of inferential beliefs ultimately rests on the rationality of basic beliefs, so too the rationality of instrumental desires and of the corresponding actions ultimately rests on the rationality of intrinsic desires. Intrinsic desires are rational if they are well-grounded in certain “objectively valuable” features of experience, much as basic beliefs are justified, and thus rational, if they are well-grounded in certain perceptual or intuitive features of experience. Though Audi’s theory of practical rationality does not strictly require “axiological experientialism”, in point of fact it is “the experientially grounded rationality of intrinsic desire that chiefly concerns [him] in accounting for the foundations of practical reason”. This normative sort of foundationalism, Audi holds, is to be distinguished from merely motivational foundationalism; the latter is compatible with certain instrumental accounts of rationality which view intrinsic desires as ultimate motivational states capable, via appropriate beliefs, of conferring rationality on the actions they motivate without themselves being properly regarded as rational. The normative foundationalism that Audi proposes to defend in opposition to instrumentalism about practical reason holds not only that rational action must ultimately be well-grounded in rational intrinsic desires, but that “there are substantive criteria for the rationality of intrinsic desires”. My aim in this discussion note is to raise some concerns about Audi’s claim to have established a normatively and substantively foundational alternative to instrumentalism, and thus to have provided a unified foundational account of rationality. In closing, I shall also attempt to respond to certain objections Audi has raised against instrumentalism. Much in this wide-ranging book, however, remains untouched by my comments; many of its excellent features would doubtless have been highlighted in the context of a general review.