The Human Mind and Brain Program
Dept of Neurobiology - Special Lecture
There seem to be numerous defects of the eye that would be expected to interfere with vision. Examples are the upside down retinal image, the blind spot in each eye's visual field, non-uniform spatial and chromatic resolution, and blur and image shifts caused by eye saccades. In order to overcome such defects scientists have proposed a variety of compensation mechanisms. I will argue that such compensation mechanism not only face empirical difficulties, but they also suffer from a philosophical objection. They seem to require the existence of a "homunculus" in the brain that contemplates the picture-like output of the compensation mechanism. A new view of what "seeing" consists in is required.
The new view of seeing considers seeing as a particular way of actively exploring the environment. This "sensorimotor" approach is subtly different from the idea of "active vision" known today in cognitive or computer science. The sensorimotor approach explains how, despite the eye's imperfections and despite interruptions in the flow of sensory input, we can have the impression of seeing everything in the visual field in detail and continuously.
I shall show how the phenomenon of "inattentional blindness" (or "Looked but Failed to See") is expected from the new approach, and I shall examine the phenomenon of "change blindness" which arose as a prediction from the theory. Finally I examine the question of the photographic quality of vision: why we have the impression of seeing things all over the visual field, why everything seems simultaneously and continuously present, and why things seem to visually impose themselves upon us in a way quite different from how memory and imagining do. To explain these facts I shall invoke four objectively measurable aspects of visual interactions: richness, bodiliness, partial insubordinateness and grabbiness.
