ABSTRACT
This chapter presents a new computational model which attempts to deal with the spatial prepositions in, on, over, under, above and below and extends processing of the visual scene to include functional factors parasitic upon object knowledge. One possibility is that object knowledge can be used as a means of weighting parts of geometric processing, as is suggested by T. Regier, L. A. Carlson and B. Corrigan. Describing the position of objects in space necessitates a mapping between the spatial representation(s), computed by the visual system, and the language processing system. However, it turns out that spatial description is influenced not only bywhere objects are in space, but also by the functions that objects afford, and the functional relations between objects.
