Abstract
In this article, I show how Ricœur elaborated the theory of causality that his practical philosophy required. The investigation proceeds in two steps. First, I study the alternative definition (“non-humean”) of causality that Ricœur shapes for the historical sciences, which is characterized by conditionality, processuality and singularity. Then I insist on Ricœur’s effort to justify objectively this new causal category in a logico-transcendental way, by linking the conditions of possibility of an objective knowledge of action not to a funding subjectivity, but to the very dimensions of discourse (semantical, pragmatical, narrative). From an epistemological point of view, causality is a judgement of singular imputation, which plays an intermediary role between structural (“nomological”) explanation and narrative explanation in the historical sciences. From a metaphysical point of view, it allows to differentiate levels of abstraction within eventalism, since it can be specifically used to explain events that are not generic or virtual, but singulars and proper, and that are, for this reason, profoundly linked to a rich context; it therefore sketches a world of non-impersonal and non-atomistic events.