Claude Lévi-Strauss has long been described as the most complete representative of modernist anthropology, branded for its scientificism, immobilizing formalism and deployment of hard* science models, as opposed to history and subjectivity. Recent reviews of his works, particularly the Mythologiques, have emphasized quite different values, stressing more sophisticated concepts of history, a notion of structure based in variation and transformation and a poetical resolution of anthropology’s epistemological purposes. This article sustains that, despite some misinterpretations caused by superficial readings, it is possible to recognize both these sets of traits in the author’s works. Besides, the “two” different Lévi-Strauss that come into view do not represent different moments of a reformulated intellectual project, but rather the alternative between scientific knowledge, limited as per definition, and the much wider universe of that which cannot be expressed scientifically. This science/renunciation pair is a considerable alternative to the – at once confusing and totalizing – hegemonic project of human sciences.