Diagrams in the generation and dissemination of new science: some examples and applications

The author identifies a number of functions for visual (not necessarily diagrammatic) representations in science. In particular he distinguishes between the function of an argument and its style (showing that arguments having a similar style can have a very different function in the development of concepts and the transformation of conceptual schemes). He illustrates the heuristic and generative roles of sketches and diagrams and suggests an analogy between perceptual inferences and abduction