Michael Friedman maintains that Carnap did not fully appreciate the impact of Gödel's first incompleteness theorem on the prospect for a purely syntactic definition of analyticity that would render arithmetic analytically true. This paper argues against this claim. It also challenges a common presumption on the part of defenders of Carnap, in their diagnosis of the force of Gödel's own critique of Carnap in his Gibbs Lecture. The author is grateful to Michael Friedman for valuable comments. Part of the research (...) towards this paper was carried out while the author was a Visiting Fellow at the Center for Philosophy of Science at the University of Pittsburgh. The paper was presented to the Center's Fellowship Reunion Conference in Athens in 1992. It was committed for publication in the Proceedings of that conference, but those Proceedings never appeared. By the time it became evident that they would never appear, both the hard copy and the source file had been mislaid. The hard copy re-surfaced in 2007. The literature on this topic since 1992 appears to leave some space for the ideas and arguments presented here. Although the paper has been updated in light of the more recent literature, its basic thesis, presented in 1992, remains the same. Only 3 is new, questioning a basic presumption made by more recent commentators in their presentation of Gödel's criticism of Carnap in his Gibbs Lecture. For helpful comments on the current version, the author is indebted to Robert Kraut, Stewart Shapiro, and Adam Podlaskowski. CiteULike Connotea Del.icio.us What's this? (shrink)

This article examines various dilemmas (or, as I suggest, pseudo-dilemmas) that have dogged epistemology and philosophy of language since the 1940s heyday of logical empiricism. These have to do chiefly with the problem those thinkers faced in overcoming the various dichotomies imposed by their Humean insistence on maintaining a sharp distinction between logical 'truths of reason' and empirical 'matters of fact'. I trace this problem back to Kant's failure to offer any plausible, explanatorily adequate account of the process whereby 'sensuous (...) intuitions' were brought under 'concepts of understanding' through the joint (somewhat mysterious) agency of 'judgement' and 'imagination'. The argument then proceeds, via Quine's (on the face of it) radically anti-dualist critique of logical empiricism, to more recent attempts - by Davidson and McDowell - to locate the residual dualism in Quine (that of scheme and content), and thus to bring philosophy out on the far side of all these vexing dilemmas. I maintain that the problem goes much deeper and re-emerges with full force both in Davidson's coupling of a formalized (Tarskian) truth-theoretic approach with a strain of radical empiricism and likewise in McDowell's revisionist, supposedly 'naturalized', but none the less dualist reading of Kant on the twin powers of 'spontaneity' and 'receptivity'. Its ultimate source - I suggest - is the normative deficit, i.e., the lack of rational and justificatory values that has typified empiricist thinking from Hume to Quine, along with its attendant sceptical outlook as regards the existence of causal powers or the status and validity of causal explanations. My paper concludes by indicating briefly some alternative resources that might point a way beyond this longstanding impasse. (shrink)

Mathematical practice prompts theories about aprioricity, necessity, abstracta, and non-causal epistemic connections. But it is not clear what to count as the data: mathematical necessity or the appearance of mathematical necessity, abstractness or apparent abstractness, a prioricity or apparent aprioricity. Nor is it clear whether traditional metaphysical theories provide explanation or idle redescription. This paper suggests that abstract objects, rather than doing explanatory work, provide codifications of the data to be explained. It also suggests that traditional rivals—conceptualism, nominalism, realism—engage different (...) explanatory projects, and thus are not rivals at all. (shrink)