Moving from the seminal contribution of Aristotele--according to whom science is not concerned with the what, neither with the how of reality, but with the why--the article emphasizes some characteristic features of a science: it must be rational, methodologically grounded, and it must have a specific object, depending on which many different sciences can be distinguished. An historical sketch of science is offered, as well as some recent contributions in the field of epistemology, in particular those of Carl R. Popper and of Thomas S. Kuhn. The case of Glottodidactics (Language Teaching)--which has been considered as a proper science only since a few years ago, and which has, since then, regular university courses--is very useful to study the difference between "sciences to know" and "sciences to do", and to outline the structure of an interdisciplinary science; this case--this is the author's claim--is quite similar to that of Nursing Science.