9221855 NEWCOMBE This research concerns children's ability to remember the locations of objects. Remembering where things are in the world has very clear adaptive importance. Yet, somewhat surprisingly, the well- known work of Piaget claimed that children younger than 9 or 10 years remember location only in terms of whether something touches or is close to a landmark, but not in terms of distance. This description of spatial development has guided much subsequent thinking about curricula in areas such as geography, map-reading, and geometry. However, recent work using new techniques suggests that, by as early as 16 months, children can indeed encode distance accurately. In addition, the small errors they do make show a pattern of systematic bias that suggests that, like adults, children of this age encode location in terms of spatial categories, as well as at a more fine-grained level, combining the two levels when they search for objects. This process can be called hierarchical coding. The first group of new experiments will test the hypothesis that children use hierarchical coding from a very early time, by excluding several other accounts of their bias patterns. The second line of work will use a new paradigm to investigate coding of distance as early as 4 months. Infants of this age cannot crawl, and are just beginning to reach, so finding distance coding at this time would suggest that crawling and reaching are not vital to such coding (as Piaget and others have suggested). A third set of experiments will establish the nature of the categories used in hierarchical coding, in particular, whether the probability and extent of subdivision of larger spaces increases with age. The fourth line of work will focus on biases in spatial judgment, aimed at testing the hypothesis that these occur when spaces are subdivided. Taken together, the work should establish a new theoretical account of spatial development. ***
