Introduction

The work of Horace Barlow, his colleagues and his students has not only made us consider the efficiency of the visual system in encoding significant properties of the visual world it has also focused attention on the question: how did the visual system get that way? Initially, the most promising approach appeared to be through experimental manipulations in the visual development of animals, and the continuing fruits of that approach can be sampled elsewhere in this volume. However, in the last decade and a half it has also been realized that human infants can tell us a great deal about the development of their visual systems, if we use the right techniques. It turns out that this development is particularly rapid over the first few months of life. One aspect of this is a quantitative improvement in many aspects of visual performance. However, there are also radical qualitative changes, which we will argue show the progressive emergence of functions of the visual cortex.

The basis of developments in spatial visual performance

Visual scientists' first questions about infant development were concerned with acuity and contrast sensitivity, and used measures of grating detection from behaviour in forced-choice preferential looking (FPL) and from visual evoked potentials (see reviews by Dobson & Teller, 1978; Atkinson & Braddick 1981; Atkinson, 1984; Banks & Salapatek, 1983). There are some differences still to be resolved, particularly between earlier estimates of newborn performance and those recently reported from the ‘sweep VEP’ method (Norcia & Tyler, 1985).