Roth's (Meaning and Method in the Social Sciences, 1987) advocacy of methodological pluralism does not appropriately sustain the project of social scientific methodology in response to holism and indeterminacy. (..) Quine (): thesis of the indeterminacy of translation. Rorty (1979): the indeterminacy thesis is a consequence of Quine's prior commitments to empiricism and naturalized epistemology (and to holism). Indeterminacy is supposed to follow from the super fluousness of semantic notions for natural science. (..) only a special case of the underdetermination of all theories by empirical evidence. (..) This reading of the indeterminacy thesis and its relation to other Quinean doctrines has been subjected to powerful criticism by Paul Roth (1987). Roth argues that the indeterminacy of translation follows directly from Quine's holism and the indispensability of intersubjectively available stimuli for language learning. (..) Against Rorty's thoroughly pragmatist revision of Quine, Roth claims to have vindicated Quine's insistence that an adequate pragmatism is ineliminably empiricist (Quine 1981). (..) Roth advocates an irreducible "methodological pluralism" for the social sciences. (..) The principal target of my criticism will be the claim that indeterminacy has distinctive importance for the social sciences vis-a-vis the natural sciences. (..) The indispensable Quinean premise for this argument is holism, the claim that theoretical sentences acquire meaning and evidence only as part of a theory. (..) Roth regularly con trasts theories in the natural sciences to semantic theories, as theories about two different sorts of domain, one of which contains matters of fact while the other does not. (..) "Physical object" and "semantic object" remain contrast classes, and indetermin acy specifically concerns theories about meanings as opposed to theories about objects. (..) Alternative: Fact and meaning are not separate epistemic domains, but aspects of every significant utterance in any epistemic domain. (..) The indeterminacy thesis can thus be represented as the claim that there is nothing un language-like "behind" language use which can fix its meaning or reference. (..) The salient features of a shared environment might be neither objects nor sensory qualities, but rather purposive complexes of equipment. (..) A dilemma confronts Roth's claim that the shared environment which resolves the paradox of language learning provides the evidence base for natural science. Either the natural sciences include such phenomena as moods within their domain of explanation, or else a theory of linguis tic development which posited moods as the proximate referents of first language would be ruled out in advance, contra Roth's pragmatic pluralism. Neither alternative seems acceptable. (..) Roth's pragmatism and pluralism are contravened by his residual empiricism, and the force of his arguments suggest that it is clearly the empiricism which must be abandoned. (..) What implications do these arguments have for Roth's own proposed methodological pluralism for the social sciences? Pluralism is first and foremost a negative thesis, the denial that we can make sense of there being "one proper set of rules" for studying human behavior, (..) What about the positive program expressed by methodological plural ism? Roth seems to be making two basic claims. He follows Quine and Feyerabend in "remaining convinced that there is something to be said concerning how inquiry may best proceed even after the demise of traditional epistemology". (..) With respect to his first claim, my arguments serve primarily to soften the contrast he wants to maintain between Rorty and Quine or Feyerabend, by loosening the sense of "methodology". (..) If I am right, then Roth's insistence upon methodological plural ism is a last vestige of the "Rationalit?tstreit" which he otherwise so consistently undercuts. Roth has convincingly interpreted Quine as showing that conversation and consensus can only be achieved via persons' mutual interaction with a shared environment: it is unintelli gible to regard inquiry and its products as "merely social", uncon strained by the world. Yet it is also unacceptable to take that shared environment as securing a domain of "objective evidence" in any sense which would escape negotiation: the identification of evidence within that shared environment, along with its employment for the sake of furthering the practices of inquiry, is an ineliminably social activity which cannot be fixed by any determinate fact of the matter, whether physiological or behavioral. |