47

After a presentation of the book and the context in which it appeared in 1956, Marcel Cohen’s approach, and the methodological and theoretical guidelines he followed, are analyzed. Cohen distinguishes between linguistic phenomena (language in use, as well as individual language systems), social phenomena (a social structure, social situations), and historical phenomena. These distinctions raise still-relevant questions about the recurrent difficulties which have led to the recognition of the importance of taking contrasting viewpoints, of never considering phenomena in isolation, of using varied approaches, and of taking history into account. Factors, particularly social factors, which determine linguistic change are a central area of inquiry, since language systems and language use evolve independently of one another. At this point, a reference to Stalin is made, with the notions of “base”€™ and “superstructure.” Though Cohen does not deny their interest, he follows Meillet in preferring the term “€˜institution€™.” However, he above all emphasizes the need to leave such questions open, while further data are collected, and the potential contribution of other human sciences explored.