Spatial Behavior and Its Alterations: Old and New Concepts

Abstract

In a normal perceptual situation, the eye not only sees but also moves. This crude evidence seems to have been largely neglected in the classical models of space perception, in which all the retinal information was thought to be gathered while the retina was stationary. There is no doubt, however, that the retina works as a system of polar coordinates, with their origin in the fovea. Each retinal locus can thus be assigned a sign, defined by its position on the polar diagram: distance from the origin and angular distance from one of the axes of reference (vertical or horizontal). Each local sign is assumed to represent a given portion of the physical space, and consequently the retinal image might constitute a reliable two-dimensional spatial mosaic. The point-to-point correspondence between the retinal image and a central copy, caused by the spatial organization of the upper visual centers, is well known. Such a design certainly holds true in the foveal pure cone system, in which each cone-bipolar-cell-ganglion-cell wiring is individual and represents a single spatial unit. But it becomes largely an ideal in the periphery of the fovea itself and in all of the extrafoveal retina, in which the cone size (and the intercone distance) increases and the cone-optic-tract wiring becomes more and more diffuse. Under these circumstances, the spatial retinal map would be better represented in three dimensions, by plotting the spatial resolving power or the spatial reliability on the Z axis; this would show a peak at the center of the fovea and a rapid exponential decrease toward the periphery (see Jones & Higgins, 1947; Polyak, 1941). As a consequence of the nonhomogeneity of the spatial retinal map and of the poor spatial discrimination outside the fovea, the eyes move so as to bring