Feyerabend and incommensurability

I consider only the semantic claims of Paul Feyerabend's incommensurability thesis. These semantic claims are that incommensurable scientific theories, taken paradigmatically as successive theories: (1) are inconsistent; (2) the terms of one theory differ in meaning to those of another incommensurable theory; and (3) the claims of one theory are largely logically independent of the other. Since the inconsistency claim (1) is essential to Feyerabend's argument (against the Received View on theory reduction and explanation), I claim that (2) and (3) must be understood in the light of (1), and that (3) must be revised to avoid contradiction with (1). Feyerabend's semantic theory supporting (3) is presented and found wanting. Two other main arguments against (3) are also considered. The first is the causal theory of reference (of Putnam and Devitt), including causal descriptive theories advocated by Kitcher and Psillos; none of these theories is found to offer compelling reasons to reject (3). The second main argument against (3) is Donald Davidson's essay 'On the Very Idea of a Conceptual Scheme', and a close reading of Davidson's paper is offered. I find that Davidson does offer convincing reasons for rejecting any implication by (3) that the languages of incommensurable theories are not intertranslatable, or that such theories are closed cognitive frameworks. However, I agree with Larry Laudan that Davidson does not deliver a fatal blow to the semantic incommensurabihty thesis because: (a) incommensurability need not entail nontranslatability; and (b) Davidson's semantic arguments do not succeed in demonstrating that the very notion of a conceptual scheme is incoherent. I present briefly two versions of the semantic incommensurability thesis which are consistent in an interesting way with (1), (2) and a revised (3), namely taxonomic incommensurability and a model of misinterpretation in intractable conflicts.