This article invites us to consider the usefulness of iconography as a method to describe the subjects represented in medieval works of art. In this study, a greater emphasis has been put on investigating the epistemological implications resulting from the use of that method, and the biases that may appear in the process of transforming images into words. The aim is to address what has been called the “semantic gap”, i.e., a kind of barrier that prevents the verbal representation of a non-lexical medium, such as the visual one, in a satisfactory and undiminished way.