Memory for Urban Geography

Open Access
Publisher Website
Google Scholar
Abstract
Consistent patterns of errors are found in estimations of distances and angles along urban routes, even among subjects who know the areas well. These patterns can be used to discover the organization of the knowledge we use to find our way around in everyday life. In the first experiment, 80 undergraduates estimated by ratio scaling the walking distances between pairs of locations in St. Andrews. Routes varied independently in their location, number of major bends, and length. Relative overestimation of length was found with routes in the town centre, with routes having several major bends, and (perhaps as an experimental artifact) with short routes. In the second experiment, 30 Cambridge residents estimated the angles between pairs of roads, by drawing the configuration of roads at their junctions. The real angles were either in the range 60–70° or 110–120°. All the estimates differed little from 90°, regardless of the true magnitude of the angle. The implications of these findings for theories of mental representation of largescale space are discussed. A model is supported in which a spatial area is represented as a “network-map”, consisting of strings of locations forming a net of paths known to be traversable, but vector distance is not preserved.

This publication has 7 references indexed in Scilit: