The Analytic-Synthetic Distinction

If a concept has an analysis, then a statement connecting the two will be analytically true—it will be true just in virtue of the correctness of the analysis. If there are correct conceptual analyses, then there are analytically true statements—and there are correct conceptual analyses. Thus there is an analytic-synthetic distinction. If there is no analytic-synthetic distinction, then there are no analyses: no concept breaks down into constituents, with the accompanying necessary and sufficient conditions. All concepts (“meanings”) turn out to be primitive, or there are no concepts (“meanings”): some adopt the former view; Quine adopted the latter. Both views seem utterly extraordinary to the naïve viewpoint—like being told that all linguistic expressions (including phrases and sentences) are primitive or that there are no linguistic expressions. To hold that there is no analytic-synthetic distinction is apt to strike the traditional philosopher as like holding that there is no subject-predicate distinction or no true-false distinction. He or she will want to hear a pretty convincing argument before going down that path. This chapter considers whether there is such an argument.