Abstract

This essay discusses the possibility of methodological pluralism in the social sciences. As the history of philosophical reflection on the social sciences shows, the moral embeddedness of social scientific practices was routinely under estimated because it has been masked by a set of epistemically and morally illegitimate dichotomies. The central dichotomies that will be discussed in the essay are: methodological vs. anti-methodological accounts of social scientific practices; critical vs. conservative social research, integrative vs. fragmenting arguments for methodological pluralism, transcendental vs. empirical arguments of possibility of social science, assuming the perspective of the interpreter vs. the perspective of the interpreted in evaluating the adequacy and fruitfulness of research, and finally procedural vs. substantive ethics. ;The alternative approach that will be developed in the essay draws on Gadamerian hermeneutics especially in Gadamer's dialogues with Habermas. Since Gadamerian hermeneutics acknowledges the intrinsic moral embeddedness of social science, it suggests that it must be pluralistic, reflexive, and dialogical. These notions are developed in relation to the Aristotelian model of phronesis which involves a double moment of knowledge, that of a universal and that of its embeddedness. While the significance of the reflexivity involved in phronesis is that it brings up the limitations of our interpretive horizons, it is only in dialogue that such limitations come to bear. Only in dialogue can the social scientist place his own prejudices on the line and take the subject of research as a partner in inquiry rather than an object of research. But since dialogue can involve violation, a notion of actual and significant consent is suggested as a minimal condition to counter such a possibility. This shift of perspective places the moral responsibility on the social scientist to genuinely enable the subject of research to consent or dissent from the proposed research. However, consent does not guarantee that, in retrospect, dialogue should not be found to have involved violation. Thus, rather than terminating in providing consent, the dialogical relation between researcher and subject extends to the mutual testing of interpretive suggestions