Medical interests have long been fascinated with the eye, if for no other reason than for the diseases and anomalies it exhibits. More recently, much interest has been concentrated on the areas of psychology and neurophysiology, for which the eye, particularly the retina, offers the opportunity for intimate study of the input region for perception and cognitive processes. The physiology of the eye has been thoroughly examined and its optical nature is well understood. Recently, distinguished work in molecular biology has progressed toward a rather complete story of the light response of the retina. Yet neurophysiological data pertaining to the visual perception of man and other species appears to be only in the early stages of investigation. As the depth of neural processing becomes greater, exploring the mystery becomes more complicated, even with today's sophisticated experimental techniques. With such a brief summary of our understanding of seeing, let me deal with the visual aspects of computer work that relates to our perceptual behavior. Then perhaps we can return to consider our extraordinary abilities in a new light. Most of the developments I will describe have been efforts to solve the problem of having the computer communicate with men as efficiently as possible. The eyes have long been recognized as the richest source of man's knowledge of his world; yet we severely limit this faculty when reading. We compromise the vast capabilities of vision when following text, because of the production techniques that have evolved to produce a linear symbol string. Language is the outgrowth of auditory signaling, later recorded in pictures. The sounds we make are resonant modifications of the noise produced by the vocal cords; we generate graphics with arm and hand motions, guided by touch and sight. Thus the rather primitive muscle systems of the throat, mouth, arm, and hand limit language, and the scope of what we can hear and see is compiomised. Here technology may perhaps offer instruments which, by extending our productive capabilities, may improve upon our present languages, if only we were not so blind to our own perceptual potentialities. There appears to be a natural division into two categories of computer functions in the visual arts. We can refer to them variously as production and recognition, or roughly output and input. First let us consider production of visual material, then the cognitive processes which are currently being explored with computers, and finally examples of those cases where the two are closely coupled.