This article explores an underappreciated pragmatic constraint on the expression of opinions: When expressing an opinion on a topic that has been previously discussed, a speaker should correctly indicate the cultural standing of that view in the relevant opinion community. This Bakhtinian approach to discourse analysis is contrasted with conversation analysis, politeness theory (Brown & Levinson 1987), and analysis of epistemic modality. Finally, indicators of four points on the cultural standing continuum (highly controversial, debatable, common opinion, and taken for granted) are illustrated with examples from American English usage. (Opinion display, discourse analysis, argumentation, hedges, modality, welfare discourse)* ".. . Prose discourse-in any of its forms, quotidian, rhetorical, scholarly-cannot fail to be oriented toward the 'already uttered,' the 'already known,' the 'common opinion' and so forth." Mikhail Bakhtin, "Discourse in the Novel" (1981:279) I N T R O D U C T I O N When speakers voice an opinion on a topic about which there has been prior discourse or tacit agreement in their opinion community, they are expected to mark the cultural standing of that opinion. "Cultural standing" is my label for the location of a view on a continuum that ranges from highly controversial to completely taken for granted in the relevant opinion community. Cultural standing has been insufficiently discussed in pragmatics, sociolinguistics, and discourse analysis as a constraint on opinion display. Expression of opinion has been analyzed primarily from three perspectives: how conversations are structured (conversation analysis; e.g., Pomerantz 1984, Sacks 1987); how social relations of power and solidarity affect the expression of potentially offensive views (politeness theory; e.g., Brown & Levinson 1987); and what evidence speakers have for their views (Chafe & Nichols 1986). In this article I propose a Bakhtinian