Components of spatial memory : a developmental analysis
How is spatial information represented in the developing human brain, and how are different representations combined to solve spatial problems Neuroscience indicates that mature spatial representations depend on networks of specialised mechanisms that normally interact but are experimentally dissociable. The present studies applied this framework of dissociable components to understanding the development of spatial memory. Two kinds of search task were used. Disoriented adults and 18-24 month olds used conjunctions of colour and room shape to find hidden objects. Three- to six-year-olds searched for toys hidden in an array bordered by landmarks, where the availability of different frames of reference was manipulated by moving subject and / or array before retrieval. Ability to resolve conjunctions of colour and room shape did not depend on language, either in children or in adults, but was mediated by disoriented subjects' erroneous sense of orientation. In small square enclosures, 18-24 month olds were able to use wall colour to reorient. These results argue against previous claims for information encapsulation, but supports the thesis combining different kinds of visual information is difficult in early development. In the array task, core components of adult spatial representation, including parallel egocentric and nonegocentric representations of space, were present as early as three years, and three-year-olds had and greatly favoured representations that were not purely egocentric. Five-year-olds solved viewpoint-independent recall dependent only on local landmarks, showing emergence of an "intrinsic" frame of reference. Distinctive patterns of impairment in use of these frames of reference were found in individuals with Williams Syndrome, and in children born premature or with focal lesions. These patterns, and correlations with other tests, provide evidence that cognitive and neural substrates for spatial memory are dissociable in development, and that children's early spatial errors can be explained by differential rates of maturation for different substrates of spatial representation.