The paper presents the mainstream methodological reflection in the field of art history, shaped by the reception of Karl Popper’s philosophy of critical rationalism from the 1940s to the 1980s. A key role in this process was played by various attempts to respond to the deductive-nomological model of scientific explanation. Referring to Popper’s ideas, Gombrich developed the project of deductive iconology, associated with the conventionalist approach to the principles of image representation and communication. In dialogue with Gombrich’s views, alternative and mutually contradictory versions of the adaptation of the DN model for the methodological explanation of images were put forward by Oskar Bätschmann and Michael Baxandall. Michael Fried and Norman Bryson proposed opposing versions of viewing the image as a form of response to the objective and fundamentally fixed initial conditions of contact with the viewer. The divergence and incommensurability of the methods of art history facing Popper’s methodology revealed the inherent paradox of the notion of fact, on the one hand treated realistically and opposed to theories, and on the other depending on the interpretive perspective and theoretical assumptions.