In this paper I propose a version of interrogative logic as a form of representation of the underlying logic of scientific practices—a kind of model-based reasoning, since it takes into account an oracle or model that can be consulted—by means of modified tableaux that can represent a game with two types of moves: logical steps and queries, governed by (modified) tales and a query rule, respectively. In this way, the process of acquiring knowledge is analogous to the justification of a (...) thesis through the method of modified tableaux. (shrink)