Abstract

Summary. This paper analyzes Frederik Stjernfelt’s recently published Diagrammatology in order to clarify the role of diagrammatic reasoning within an epistemology that focuses on the problem of learning and the growth of knowledge. To achieve this goal, I provide more precise definitions of Peirce’s concepts of “diagram” and “diagrammatic reasoning,” emphasizing in particular the necessity of consistent systems of representation as a precondition for both. The paper starts with a critique of two theses for which Stjernfelt argues based on some remarks by Peirce: first, that it is possible to learn by observing icons and, second, that icons can be defined by similarity