The idea of determinism, proposed to extend and generalize physical causality by adding to it the consideration of initial conditions, was built from then on as the ideal reference of all scientific knowledge and was considered as insuperable. However, this ideal was to be challenged in various directions: by the modifications consecutive to relativistic causality; "non linear" deterministic dynamical systems, the behavior of which can lead to totally unpredictable situations; problems with quantum physics ("reduction" of the state function and probabilistic predictions). The two last cases make evident the limits of the notions of causality and of determinism, by showing that actual knowledge can far exceed them, and suggest questions concerning the actual physical meaning of the magnitudes used in the theory. When we consider the magnitudes that are the more physically meaningful from the point of view of the characteristics of physical phenomena, causality appears differently and especially determinism is shown anthropologically centered and insufficient. We are led to another more meaningful category: necessity, which is independent of our conceptual and theoretical choices and has at the same time the capacity to regulate them.
Causality; Science; Completeness; Initial conditions; Determinism; Space; Quantum physics; Physical magnitude; Invariance; Necessity; Relativity; Dynamical systems; Time