In this paper we present the traditional (which we call canonical) model used to explain self-deception and show some of the problems it has to face. Then we present a different model known as the deflationary model or the non-intentional model and defend it is a better alternative to explain self-deception than the former. Finally, we present a problem the deflationary model has to face and that, when treated, leads to new considerations about the model and to see it under a new light. This problem, however, does not debilitate the model but makes it stronger in the end. We defend that the problem arises from the fact that the model has been thought on the light of perceptual evidence but not including social interactions. The inclusion of such interactions makes it necessary to modify slightly the model in order to accommodate them and the behavior of the self-deceived agents. In each of the sections of this paper we resort to literary examples that depict the essential characteristics of each case.
