Author's personal copy CHAPTER 14 Development of the Visual System S.P. Johnson University of California, Los Angeles, CA, USA OUTLINE 14.1 Classic Theoretical Accounts 250 14.4.5 Development of Visual Memory 256 14.1.1 Piagetian Theory 250 14.4.6 Development of Visual Stability 258 14.1.2 Gestalt Theory 251 14.4.7 Object Perception 259 14.2 Prenatal Development of the Visual System 251 14.4.8 Face Perception 259 14.2.1 Development of Structure in the Visual 14.4.9 Critical Period for Development of Holistic System 252 Perception 261 14.3 Visual Perception in the Newborn 252 14.5 How Infants Learn About Objects 261 14.3.1 Visual Organization at Birth 252 14.5.1 Learning from Targeted Visual 14.3.2 Visual Behaviors at Birth 253 Exploration 261 14.5.2 Learning from Associations Between Visible 14.3.3 Faces and Objects 253 and Occluded Objects 262 14.4 Postnatal Visual Development 254 14.5.3 Learning from Visual–Manual 14.4.1 Visual Physiology 254 Exploration 264 14.4.2 Critical Periods 254 14.4.3 Development of Visual Attention 255 14.6 Summary and Conclusions 265 14.4.4 Cortical Maturation and Oculomotor References 266 Development 256 The purpose of vision is to obtain information about over 50% of the cortex of the macaque monkey (a phylo- the surrounding environment so that we may plan ap- genetically close cousin to Homo sapiens) is involved in propriate actions. Consider, for example, a stroll on the visual perception, and there are perhaps 30 distinct beach vs. a hike in the Grand Canyon (Figure 14.1). Both cortical areas that participate in visual or visuomotor activities involve locomotion, but each places very dif- processing (Felleman and Van Essen, 1991; Van Essen ferent demands on the perceptual and action systems, in- et al., 1992). cluding the visual system. In the case of the stroll, the This chapter reviews theory and data concerning de- beach is wide, there are few obstacles and little risk. In velopment of the human visual system with an emphasis the case of the hike, in contrast, the path is narrow; rocks, on object perception. As will be seen, infants are pre- vegetation, and abrupt precipices must be avoided. To pared to see objects and understand many of their prop- remain safe, the hiker must know what the risks are, erties (e.g., permanence, coherence) well in advance of and this invariably involves knowing what objects there locomotion, so that by the time infants begin to crawl are in the visual scene. The importance of accurate per- and walk, they have a good sense of what and where ob- ception of our surroundings is attested by the allotment stacles might be, even if the hazards these objects pose of cortical tissue devoted to vision: By some estimates, remain unknown. Preparation of this chapter was supported by NIH grants R01-HD40432 and R01-HD48733. Neural Circuit Development and Function in the Brain: Comprehensive Developmental 249 # 2013 Elsevier Inc. All rights reserved. Neuroscience, Volume 3 http://dx.doi.org/10.1016/B978-0-12-397267-5.00033-9 Comprehensive Developmental Neuroscience: Neural Circuit Development and Function in the Brain, (2013), vol. 3, pp. 249-269 Author's personal copy 250 14. DEVELOPMENT OF THE VISUAL SYSTEM system that is prepared to contribute in important ways to learning, but incapable of perceiving objects in an adult-like fashion. Developmental processes that lead to mature perception and interpretation of the visual world as coherent, stable, and predictable are an area of active investigation and are only beginning to be understood. 14.1 CLASSIC THEORETICAL ACCOUNTS Discussions of the nature versus nurture of cognitive development are entrenched and persistent. Such dis- cussions are particularly vigorous when concerning in- fant cognition and have tended to be long on rhetoric but short on evidence, in part because the evidence has been, until recently, relatively sparse. Research on vi- sual development, in contrast, has tended to focus on de- velopmental changes in neural mechanisms, with much of the evidence coming from animal models (Kiorpes and Movshon, 2004; Teller and Movhson, 1986). Re- search on human infants’ visual development has often been motivated by two theoretical accounts, each of which considers seriously both the starting point for postnatal development and the mechanisms of change that yield stable, mature object perception: Piagetian the- ory and Gestalt theory. FIGURE 14.1 Two visual scenes. 14.1.1 Piagetian Theory The first systematic study of infants’ perception and There is much else to learn. Visual scenes, for exam- knowledge of objects was conducted by Jean Piaget in ple, tend to be very complex: a multitude of overlapping the 1920s and 1930s (Piaget, 1952/1936, 1954/1937). and adjacent surfaces with distinct shapes, colors, tex- According to Piaget, knowledge of objects and space de- tures, and depths relative to the observer. Yet our visual veloped in parallel, and were interdependent: One can- experience as adults is not one of incomplete fragments not perceive or act on objects accurately without of surfaces, but instead one of objects, most of which awareness of their position in space relative to other ob- have a shape that can be inferred from partial views jects and to the observer. Knowledge of the self and of and incomplete information. Is the infant’s visual system external objects as distinct, coherent, and permanent en- sufficiently functional and organized to make sense of tities grew from active manual search, initiated by the the world from the onset of visual experience at birth, child. When the child experiences her own movements, able to bind shapes, colors, and textures into coherent she comes to understand them as movements of objects forms, and to perceive objects as regular and predictable through space and applies the same knowledge to move- and complete across space and time? Or does the infant’s ments of other objects. visual system require a period of maturation and expe- Initially, prior to any manual action experience, in- rience within which to observe and learn about the fants understand the world as a ‘sensory tableaux’ in world? which images shift unpredictably and lack permanence These ‘nature versus nurture’ questions begin to lose or substance; in an important sense, the world of objects their steam when the details of visual development are that we take for granted does not yet exist. Active search examined and explained, because visual development behavior emerges only after 4 months and marks the be- stems from growth, maturation, and experience from ginnings of ‘true’ object knowledge. Over the next few learning and from action; all happen simultaneously months, infants reveal this knowledge, for example, by and all influence one another. Infants free of disability following the trajectory of thrown or dropped objects, or developmental delay are born with a functional visual and by retrieval of a desired object from under a cover II. COGNITIVE DEVELOPMENT Comprehensive Developmental Neuroscience: Neural Circuit Development and Function in the Brain, (2013), vol. 3, pp. 249-269 Author's personal copy 14.2 PRENATAL DEVELOPMENT OF THE VISUAL SYSTEM 251 where it had been seen previously. Later in infancy, in- that development of object perception per se involved fants are able to search accurately for objects even when active manual exploration, which imparts additional there are multiple potential hiding places, marking the information about specific object kinds (Koffka, 1959), advent of full ‘object permanence.’ but the starting point for visual experience is necessarily Piaget placed more emphasis on the importance of quite different on the two accounts. On the Gestalt view, manual search for developmental changes in object per- perceptual organization precedes object knowledge; ception than visual skills, yet the lessons from his theory on the Piagetian view, object knowledge and perceptual for questions of development of the visual system could organization develop in tandem. not be more relevant. Upon the infant’s first exposure to Piagetian and Gestalt accounts specify a starting point patterned visual input, he does not inhabit a world of ob- for postnatal development, and each has particular jects, but rather a world of disconnected images devoid views about how development of the infant’s visual of depth, coherence, and permanence. Building coherent world might proceed. Neither account is wholly on things from these disconnected images comes from ac- one side of the nature–nurture issue, and both accounts tion and experience with objects over time. have offered testable predictions that have guided sub- sequent research, and as will be seen later in this chapter, both accounts have influenced important research on ob- ject perception in infants. Yet neither can be taken as 14.1.2 Gestalt Theory complete, in part because neither took a sufficiently com- Piagetian theory can be contrasted with a coeval, com- prehensive approach to vision. A quote from Gibson peting account. The Gestalt psychologists, unlike Piaget, (1979) helps explain why this is so: The visual system were not strictly developmentalists, but they did have comprises “the eyes in the head on a body supported much to say about how visual experience might be struc- by the ground, the brain being only the central organ tured in the immature visual system. They suggested of a complete visual system. When no constraints are that subjective experience corresponds to the simplest put on the visual system, we look around, walk up to and most regular interpretation of a particular visual ar- something interesting and move around it so as to see ray in accord with a general ‘minimum principle,’ or it from all sides, and go from one vista to another” Pra¨gnanz (Koffka, 1935).