Grice's analysis of conversational maxims and implicatures is examined in the light of Malagasy language and ways of speaking. A cultural contrast in primary assumptions is described. Grician analysis retains usefulness but within the perspective of a comparative typology in which locally valid systems may differ strikingly in what is marked and unmarked. An ethnographic base and ethnological comparison are required. The situation somewhat resembles the situation with regard to grammatical categories addressed by Boas (1911) and Sapir (1921). (Conversational postulates, ways of speaking; English (US), Malagasy (Madagascar)). (DH)