| This thesis stands as a challenge to contemporary analytic epistemology. By showing that images in science have epistemic efficacy I suggest that any theory of knowledge that only scopes over propositional knowledge is incomplete. I offer an analysis of images that treats them as bearers of content, but defies reduction to propositions or symbol systems. Central to my argument is Kendall Walton's theory of representation, which treats representational images as props in games of make-believe. Principles of generation, derived from convention, our cognitive and perceptual abilities and background knowledge constrain the types of imaginings that are authorized for any particular image. Although every image has an intersubjective content, there is also room for all viewers to bring their own idiosyncrasies and interests to the imaginings that are prompted. The epistemic role of scientific images rests on this representational role.; The epistemic role of these images is characterized through analogy with scientific models and in particular, Nancy Cartwright's view that scientific models are mediators that are both crucial to scientific justification, but are also irreducible to either theory or world. Production of such models requires creativity and, I argue, understanding of the capacities and features of the objects under investigation. Representational images are peculiarly well equipped to provide understanding in a way that language cannot. They can relate the objects and states of affairs that they depict both spatially and causally in multiple interrelationships.; As a guide through the argument I focus on one example as a hard case---Feynman diagrams---only extending my theory to other examples in the concluding chapter. I argue that even though they have a notational function they nevertheless also have a representational function, which helps to explain their striking success as scientific images. Furthermore, I suggest that when he was devising the theory in which these diagrams were first introduced, Feynman was relying on a type of physical thinking that I characterize as understanding. |
