Software Engineering (SE) is a field without too much historic background. The youth of the SE discipline is resulted in an immaturity of this research field and SE research still lacks suitable scientific precision. Moreover, in SE research there are several objects of study with different nature each of them and, for this reason, different research and validation methods are needed. In view of this situation, we discuss the nature of the SE, trying to identify different kinds of research problems this discipline deals with. As hypothesis, and according to the nature of the research problem to be broached, we propose a classification of the research problems of the SE discipline in: those with a Scientific nature and those with an Engineering nature. We try to justify this hypothesis on the basis of the paradigms and the research process which is in general used for the resolution of these kinds of problems.