A mathematical formalism in physics is not “just a matter of description”