There has been debate among linguists with regards to the semantic view and the pragmatic view of presupposition. Some scholars believe that presupposition is purely semantic and others believe that it is purely pragmatic. The present paper contributes to the ongoing debate and exposes the different ways presupposition was approached by linguists. The paper also tries to attend to (i) what semantics is and what pragmatics is in a unified theory of meaning and (ii) the possibility to outline a semantic account of presupposition without having recourse to pragmatics and vice versa. The paper advocates Gazdar’s analysis, a pragmatic analysis, as the safest grounds on which a working grammar of presupposition could be outlined. It shows how semantic accounts are inadequate to deal with the projection problem. Finally, the paper states explicitly that the increasingly puzzling theoretical status of presupposition seems to confirm the philosophical contention that not any fact can be translated into words.