Production and comprehension have long been viewed as inseparable components of language. The study of vision, by contrast, has centered almost exclusively on comprehension. Here we investigate drawing—the most basic form of visual production. How do we convey concepts in visual form, and how does refining this skill, in turn, affect recognition? We developed an online platform for collecting large amounts of drawing and recognition data, and applied a deep convolutional neural network model of visual cortex trained only on natural (...) images to explore the hypothesis that drawing recruits the same abstract feature representations that support natural visual object recognition. Consistent with this hypothesis, higher layers of this model captured the abstract features of both drawings and natural images most important for recognition, and people learning to produce more recognizable drawings of objects exhibited enhanced recognition of those objects. These findings could explain why drawing is so effective for communicating visual concepts, they suggest novel approaches for evaluating and refining conceptual knowledge, and they highlight the potential of deep networks for understanding human learning. (shrink)

This article focuses on the difficulty for a general theory of depiction of providing a notion of pictorial content that accommodates the full diversity of picture types. The article begins by introducing two basic models of pictorial content using paradigmatic positions that maximize the ability of the respective models to deal with pictorial diversity. Kulvicki's On Images is interpreted as a generalized projection-based model which proposes a scene-centred notion of pictorial content. By contrast, Lopes's aspect-recognition theory, in Understanding Pictures, is (...) taken to exemplify a recognition-based model of pictorial content. It is then considered how applicable these two notions of pictorial content are to a spectrum of different picture types, highlighting differences between mechanically produced and “freehand” pictures. It is shown that while the projection-based and recognition-based models are each required for some picture types, neither model can be applied over the whole spectrum of diverse picture types. This, it is argued, has the far-reaching consequence that neither the projection model nor the recognition model, nor any unitary combination of the two, can provide the basis for a theory of depiction applying to all pictures. (shrink)