The methodological requirement of clarity has become a theoretical-empirical requirement on the subject matter: reality itself must be clear to be analysed. However, that is inadequate proposal to analyse social life, since one of the basic traits of action is that it is not clear, but action operates inside a world that is confusing. The requirement then needs to become one in which the methodological criteria is not translated to a demand on the analysed reality. The paper explores some possibilities about what it should be sought: a possible clear modelling of a confusing reality.