The purpose of this paper is to demonstrate that specific spatial and possibly temporal rates of change dominate early infants' looking, that these spatial and temporal events have meaningful and specific empirical correlates in neurophysiology as a function of age, and finally that neurophysiologically constrained models provide testable hypotheses for studies involving infant perceptual development. A model is presented for infant looking duration or pattern preferences depending upon transformation of spatial characteristics of visual stimuli by a developing visual system. It is shown how this model predicts the behavioral data obtained by the author and others and how these behavioral data are reflected in measures of infant brain responses. The inadequacies of a strictly spatial model which force one to incorporate temporal characteristics of stimulation are discussed. (Author/KM)