ABSTRACT
A higher level implies a lower level. What is the measure of level, the cognitive altimeter? Those who teach vision to students soon learn that higher level refers to the bits of visual perception that are complicated, in some sense more interesting, and poorly understood. Faces, objects, words, and consciousness are high level; edges, acuity, flow fields, detection thresholds, and visual reflexes are low level. Is it ever possible truly to understand higher levels before understanding the lower levels on which they depend? Those who take the existence of different levels of explanation for granted may find this a curious question but its selfevidence is not Ubiquitous. Thirty years ago, in an invited lecture in Cambridge, England, David Hubel described his experiments with Torsten Wiesel on the receptive field properties of visual cortical neurons. Two Cambridge physiologists were overheard to comment on the futility of studying the visual cortex
before we had understood the retina. We still do not fully understand the retina, nor how simple and complex cortical cells acquire their properties, but the leap from lower to higher was surely worthwhile. As T. H. Huxley said, "It's only by going beyond the truth that one gets anywhere near the truth." Let there be no apology for studying higher level vision before the eye surrenders its secrets.
