Abstract: "Mana", an emblematic magic notion, constitutes a classic case in social anthropology. The different interpretations of mana by Marcel Mauss and Claude Lévi-Strauss represent a shift of theoretical paradigm from the French school of sociology to the structural anthropology. On the surface, the division between Mauss and Lévi-Strauss seems to be the translatability of "Mana". To Mauss, mana is non-translatable and, more to the point, it shouldn't be translated into modern scientific language at all. For Lévi-Strauss, mana is not only translatable but also must be interpreted in modern scientific language. This paper begins its inquiry with Mauss and Hubert's General Theory on Magic by treating mana as a "practical idea". It is argued that beneath the disagreement over the translatability, there are deeper divisions involving social ontology, definition of objects of social sciences, and relationship between theoretical and practical logic. The paper suggests that the conventional view concerning the relationship between Mauss' holism and Lévi-Strauss' structuralism is an oversimplified argument that ignores the fact that Lévi-Strauss not only disagrees with Mauss about the relation between "social reality" and "symbolism", but their quintessence as well.

Key words: Mana, practical idea, social reality, symbolism, unconscious, efficacy, action