Abstract

Understanding neural anatomy and physiology depends on first understanding the behaviour being mediated. Glover, in his review of earlier work suggesting various dichotomies in visual processing, shows how there is a tendency to oversimplification if this approach is ignored. His own new proposals demonstrate the advantages of allowing function to drive anatomical analysis. Nevertheless, the new planning–control dichotomy he proposes, though a valuable advance, is itself an oversimplification of what must be a multi-channel system.