Recipe knowledge: A tool for understanding some apparently irrational behaviour a,b.

What kind of knowledge about illness determines treatment action? The authors explore this question after encountering a paradox in the course of an ethnographic study on Malaria in Tanzania. Why did Tanzanian mothers who knew about the link between degedege --a local term to describe convulsions in children--and malaria still use traditional practices for degedege , even though they would never use these for malaria? Through questionnaires, in-depth interviews, and ethnographic fieldwork, the authors repeatedly elicited and observed seemingly irrational behaviour (for example, urinating over a child, rubbing the child's body with elephant dung, etc.), which their informants could not explain. Informants routinely commented that "this is what we have always done" or "this is what everybody does". In this paper, the authors suggest that Schu¨tz's (1964) idea of "recipe knowledge"--culturally learned formulas that are automatically activated and remain unquestioned as long as nothing unforeseen happens--offers an explanation for such observed behaviour. In the case of degedege , the "recipes" are embedded in a wider schema for action that combines traditional and biomedical practices, and thereby integrate social values of "tradition" and "modernity". These findings reflect on the limitation of classical knowledge transmitted through health messages for behaviour change, and shed light on the role of historical and social context in knowledge construction and therapeutic action.