As described in Chapter 12, if an electrode is passed at a shallow angle through the cortex while the responses of individual neurons to stimulation of one or the other eye are being recorded, detailed assessment of ocular dominance can be made at the level of individual cells (see Figure 12.13). In these studies, Hubel and Wiesel assigned neurons to one of seven ocular dominance categories. Group 1 cells were defined as being driven only by stimulation of the contralateral eye; group 7 cells were driven entirely by the ipsilateral eye. Neurons driven equally well by either eye were assigned to group 4. Using this approach, they found that the ocular dominance distribution across the cortical layers in primary visual cortex is roughly Gaussian in a normal adult (cats were used in these experiments). Most cells were activated to some degree by both eyes, and about a quarter were more activated by either the contralateral or ipsilateral eye (Figure 24.4A).

Hubel and Wiesel then asked whether this normal distribution of ocular dominance could be altered by visual experience. When they simply closed one eye of a kitten early in life and let the animal mature to adulthood (which takes about 6 months), a remarkable change was observed. Electrophysiological recordings now showed that very few cells could be driven from the deprived eye; that is, the ocular dominance distribution had shifted such that all cells were driven by the eye that had remained open (Figure 24.4B). Recordings from the retina and lateral geniculate layers related to the deprived eye indicated that these more peripheral stations in the visual pathway worked quite normally. Thus, the absence of cortical cells that responded to stimulation of the closed eye was not a result of retinal degeneration or a loss of retinal connections to the thalamus. Rather, the deprived eye had been functionally disconnected from the visual cortex. Consequently, such animals are behaviorally blind in the deprived eye. This “cortical blindness,” or amblyopia, is permanent (see next section). Even if the formerly deprived eye is subsequently left open indefinitely, little or no recovery occurs.

Remarkably, the same manipulation—closing one eye—had no effect on the responses of cells in the visual cortex of an adult cat. If one eye of a mature cat was closed for a year or more, both the ocular dominance distribution and the animal's visual behavior were indistinguishable from nor-mal when tested through the reopened eye (Figure 24.4C). Thus, sometime between the time a kitten's eyes open (about a week after birth) and a year of age, visual experience determines how the visual cortex is wired with respect to eye dominance. In fact, further experiments showed that eye closure is effective only if the deprivation occurs during the first 3 months of life. In keeping with the ethological observations described earlier in the chapter, Hubel and Wiesel called this period of susceptibility to visual deprivation the critical period for the development of ocular dominance. During the height of the critical period (about 4 weeks of age in the cat), as little as 3 to 4 days of eye closure profoundly alters the ocular dominance profile of the striate cortex (Figure 24.5). Similar experiments in the monkey have shown that the same phenomenon occurs in primates, although the critical period is longer (up to about 6 months of age).

A key advance arising from Hubel and Weisel's work was to show that visual deprivation causes changes in cortical connectivity (Figure 24.6). The implications of altered circuitry was amply confirmed by complementary anatomical studies. In monkeys, a central aspect of circuitry—the alternating stripelike patterns of geniculocortical axons representing the two eyes that form ocular dominance columns—is already present at birth (Figure 24.6A). As in the case of language development, in which infants exhibit early preferences for speech sounds, the visual cortex is not a blank slate on which the effects of experience are later inscribed. Nevertheless, animals deprived of vision in one eye from birth develop abnormal patterns of ocular dominance stripes in the visual cortex (Figure 24.6B). The open-eye stripes are substantially wider than normal, whereas the stripes representing the deprived eye are correspondingly diminished. The absence of cortical neurons that respond to the deprived eye in electrophysiological studies is not simply a result of the relatively inactive inputs withering away. If this were the case, one would expect to see areas of layer IV devoid of any thalamic innervation. Instead, inputs from the active (open) eye take over some of the territory that formerly belonged to the inactive (closed) eye. Hubel and Wiesel interpreted these results as demonstrating a competitive interaction between the two eyes during the critical period (see Chapter 23). At birth, the cortical representation of both eyes starts out equal, and in a normal animal, this balance is retained if both eyes experience roughly comparable levels of visual stimulation. When, however, an imbalance in visual experience is induced by monocular deprivation, the active eye gains a competitive advantage and replaces many of the synaptic inputs from the closed eye, such that few if any neurons can be driven by the deprived eye (see Figure 24.4B).

The idea that a competitive imbalance underlies the altered distribution of inputs after deprivation has been confirmed by closing both eyes shortly after birth, thereby equally depriving all visual cortical neurons of normal experience. The arrangement of ocular dominance recorded some months later is, by either electrophysiological or anatomical criteria, much more normal than if just one eye is closed. Although several peculiarities in the response properties of cortical cells are apparent, roughly normal proportions of neurons representing the two eyes are present. Because there is no imbalance in the visual activity of the two eyes (both sets of related cortical inputs being deprived), both eyes retain their territory in the cortex. If disuse atrophy of the closed-eye inputs were the main effect of deprivation, then binocular deprivation would cause the visual cortex to be largely unresponsive.

Experiments using techniques that label individual axons from the lateral geniculate nucleus terminating in layer IV have shown in greater detail what happens to the arborizations of individual neurons after visual deprivation (Figure 24.7). As noted, monocular deprivation causes a loss of cortical territory related to the deprived eye, with a concomitant expansion of the open eye's territory. At the level of single axons, these changes are reflected in an increased extent and complexity of the arborizations related to the open eye, and a decrease in the size and complexity of the arborizations related to the deprived eye. Individual neuronal arborizations can be substantially altered after as little as one week of deprivation, and perhaps even less. This latter finding highlights the ability of developing thalamic and cortical neurons to rapidly remodel their connections—actually making and breaking synapses—in response to environmental changes.