The aim of this essay is to understand the relation between the "structuralist activity" as defined by Roland Barthes in 1963, and the way Boulez considers musical composition, concieved as a double operation of disjunction and organization of units with a view to producing a kind of "simulacrum". In the case of Notation I for orchestra (1978), such a process implies the complete re-interpretation of a distant model, a small piece for piano composed in the mid-forties.