Information and Morphology in Cognitive Maps

Abstract

The morphology of cognitive maps is a neglected aspect of behavioural geography. This study provides a methodology for identifying the discrepancies in cognitive maps of any geographical area, isolates the main perceptual distortions in a set of mental maps of Great Britain and attempts to explain these errors in the context of the cognitive-behavioural process. The key role of information in image formation is examined in depth and four major factors found to affect the information basis of cognitive maps. INFORMATION AND MORPHOLOGY IN COGNITIVE MAPS THE available literature suggests that cognitive mapping studies can be divided into two types. The first comprises those concerned with schemata (cognitive maps) which are based on designative perceptions of places and which have no evaluative content (Stea & Downs, 1973). The second consists of appraisive studies or those concerned with schemata based on perceptions which involve some use of the individual's value system. Geographers have shown more interest in the latter type of enquiry with the work of Gould (1965, 1967, 1969), Gould and Ola (1970) and Gould and White (1968) proving to be the cynosure for most empirical investigations. To date, the morphologic or designative characteristics of cognitive maps (the elements of size, shape, distance and direction) have attracted much less attention. Lynch (1i960, 1965, 1971, 1972, 1976), Appleyard (1964, 1969), Appleyard, Lynch and Myer (1970) and other planners and psychologists have investigated these elements at the intra-urban scale but almost no effort has been made to examine the morphology of cognitive maps at the regional or national level. Clearly designative perceptions of the environment form a fundamental part of the behavioural approach to the spatial discipline of geography. The present study examines the morphology of a large set of cognitive maps derived from three populations located in spatially disparate parts of the United Kingdom. The two principal objectives of the study were: (i) to examine the shape of cognitive maps of Great Britain in order to identify major errors, with a view to isolating any discrepancies which may exhibit a degree of consistency throughout the population, and (2) to explain the internal structure of the revealed cognitive maps. THE METHOD OF SHAPE ANALYSIS An extensive range of mathematical techniques has been devised for describing the shape of irregular areas (e.g. Blair and Biss, 1967). A common characteristic of most approaches is the reduction of shapes to representative numbers so that similar shapes are described by similar indices. Such procedures, however, are not suitable for the task of comparing a set of cognitive representations of a geographic area with a cartographic standard. In this case it is more meaningful to superimpose the two shapes directly, without resorting to the intervening process 548 This content downloaded from 157.55.39.191 on Tue, 11 Oct 2016 04:14:11 UTC All use subject to http://about.jstor.org/terms Morphology of cognitive maps 549 of numerical abstraction which of necessity discards part of the information conveyed in the cognitive maps. In order to compare the hand-drawn and standard maps in this study it was first necessary to transform each to a common scale. This was accomplished in the following manner: i. The area of each map was measured three times and the average value taken. 2. The square root of this figure was then calculated and a line of length proportional to this value drawn on the map. 3. This line was then enlarged or reduced as necessary to match a similar line drawn for the standard map of Great Britain. In this way the areas of the hand-drawn maps and of the cartographic standard were equalized. 4. Finally, the adjusted outline of each hand-drawn map was traced anew. This procedure was carried out for each of the 77 maps in the sample. Having converted the outline drawn by each individual to one with the same area as the standard map it was then possible to determine any shape errors directly by superimposition. A visual 'best fit' method of comparing the maps was rejected in favour of a more rigorous solution based on geometric principles. In constructing the hand-drawn maps respondents had been instructed to regard the vertical margin of the page as lying in a north-south plane. This directional line was therefore common to all maps. Since a common point and line is sufficient to permit accurate superimposition only one further location point was required. Following Sanders and Porter (1974) the centre of the smallest circumscribable circle about the shape was taken to be this point, since for any shape the location of this circle and its centre is unique. To measure the overlay deviation between the hand-drawn and standard maps a sixteen-point compass rose was set upon each central point and the errors noted where the radials intersected the boundaries of the two shapes. Each hand-drawn map thus produced a set of observations along sixteen directional variables. The complete data set therefore consisted of a matrix of sixteen boundary deviations on 77 maps drawn by standardized groups of British undergraduates at three different locations.I