Les conditions d’une science de l’intertextualité: réflexion sur les apories du comparatisme saussurien

Abstract

In his manuscripts about legends, Saussure devises a new approach of intertextuality: instead of looking for a series of influences, he argues that legends are built through a process of displacement and distortion of earlier texts. The aim of the present article is to understand the epistemological obstacles which prevented him from carrying his project through. Saussure’s doubts are all the more interesting that when his notebooks were discovered at the end of the 1950’s, they led Levi-Strauss to define control means to make hisanalysis of Amerindian myths more rigorous. In order to avoid the illusions created by the comparative method, Levi-Strauss set up a structural theory of transformations: differences between two texts can be considered as a form of correlation if, and only if, they correspond to systematic oppositions or metaphorical transpositions. With the concept of transformation group, he demonstrated that the paradoxes inherent in semiologic identity can be solved if the corpus of variants itself is treated as a system.