Spatial Representations of Taxi Drivers.

摘要

One of the central questions about cognitive skills is how the necessary knowledge is organized. In the case of spatial skills, there are at least three theoretical viewpoints on the nature of internal representation of large-scale environments. The cognitive mapping approach suggests that this representation is very much like a map in the head. the more recent geographical approach is still very map-like, but places more emphasis on an abstracted representation based on psychologically salient features and their relationships. The cognitive science approach, on the other hand, recognizes that the representation need not be map-like at all, and emphasizes that the process which operate on it are an integral part of spatial skill. This study of expert-novice differences among taxi drivers, involving both laboratory tasks and actual driving in the field, has confirmed the validity of the cognitive science approach. Tasks in which a map-like representation would be of value, such as map drawing and placing locations on outline maps, showed no skill differences at all. The representation which did emerge is a hierarchy based upon geographical areas. At the top level are global features (Pittsburgh's three rivers), then general areas (north side, east end), then neighborhoods and, finally, locations within neighborhoods.