Contrast Illusions in Perspective

Abstract

THERE seems to be some confusion in Humphrey's account of “Contrast Illusions in Perspective”1. This may be due to over-compression, especially of a crucial paragraph (page 92, para 3), which contains: “The nature of the optical projection to the eye is such that more distant objects on the whole give rise to smaller retinal images than near ones, and distance is therefore a guide to retinal size. The expected size of the further of two objects is less than that of the nearer, and hence if we see two lines which appear to be different distances but have the same actual length, we judge the further to be longer than the nearer”. There are some curious features in this which though perhaps obvious should be spelled out, to avoid an already confused situation being worse confounded. (1) It is clearly not retinal image size which Humphrey supposes is “guided” by distance. Image size is given directly, by the geometry of object size and distance and the optical characteristics of the eye. The intended meaning must surely be along the lines of “perceived size”; but this is a very different matter from retinal image size—hence, indeed, the problem. (2) It is not the expected size of objects, lying at various distances, which is the issue but the image size corresponding to various object distances. (This is given directly by object size and distance, as stated.) (3) Perceived size (assuming this is what is meant, in place of “retinal size”) cannot be given by distance per se. It would have to be given (or “guided”) by available information of distance. This could be real-time sensory data; or it could be stored information, as of the shapes of familiar objects, such as tables.