Development of a vocabulary of object shapes in a child with a very-early-acquired visual agnosia: A unique case

Abstract

We report a longitudinal study of an exceptional child (S.R.) whose early-acquired visual agnosia, following encephalitis at 8 weeks of age, did not prevent her from learning to construct an increasing vocabulary of visual object forms (drawn from different categories), albeit slowly. S.R. had problems perceiving subtle differences in shape; she was unable to segment local letters within global displays; and she would bring complex scenes close to her eyes: a symptom suggestive of an attempt to reduce visual crowding. Investigations revealed a robust ability to use the gestalt grouping factors of proximity and collinearity to detect fragmented forms in noisy backgrounds, compared with a very weak ability to segment fragmented forms on the basis of contrasts of shape. When contrasts in spatial grouping and shape were pitted against each other, shape made little contribution, consistent with problems in perceiving complex scenes, but when shape contrast was varied, and spatial grouping was held constant, S.R. showed the same hierarchy of difficulty as the controls, although her responses were slowed. This is the first report of a child's visual–perceptual development following very early neurological impairments to the visual cortex. Her ability to learn to perceive visual shape following damage at a rudimentary stage of perceptual development contrasts starkly with the loss of such ability in childhood cases of acquired visual agnosia that follow damage to the established perceptual system. Clearly, there is a critical period during which neurological damage to the highly active, early developing visual–perceptual system does not prevent but only impairs further learning.

Keywords:

• Early acquired visual agnosia
• Visual crowding
• Critical period

We thank S.R., her family, and her teachers for their support. The research has been supported in part by the Medical Research Council: Grant R00023095.

Notes

¹ Young and Ellis (1989) noted that some items named by A.R. at 11 years 7 months may not have been known to her before her illness. We analysed the naming data, reported in full in their Appendix 1, using 75% age-of-acquisition norms, reported by Morrison, Chappell, and Ellis (1997), and found that, for the 38/40 items for which norms were available, A.R. named 68% of the 22 objects with age-of-acquisition norms at or below 3 years 3 months and only 6% of the 16 objects with norms above this age (Fisher Exact p < .001). Since A.R. acquired visual agnosia following neurological damage at age 3 years 3 months, there is little evidence that she learned to identify visually presented objects after this age.

² Low collinearity: Although the letter elements, composed of horizontal and vertical lines, were orthogonal to the direction of the letter lines, the endings of the letter elements formed collinear edges. This probably explains why S.R.'s performance in this condition reached 75% correct.

³ The proximity manipulation also affected the number of elements per letter. The similarity manipulation necessarily also made letter elements different from background elements. This could only have been overcome by having a variety of background elements, but this further condition was not included.

⁴ The separation of background from figure by at least one unit of space was necessary in order to separate figures formed by plus signs from the background composed of identical symbols.