Modes of Linearization in the Description of Spatial Configurations

Abstract

Speakers or writers who have to describe a spatial configuration to other people are faced with the following problem: The object they have to describe is two- or three-dimensional, but their verbal output is highly constrained by the one-dimensional, linear structure of language. Translating a multidimensional entity into a linear output thus requires the construction of a linear structure, which itself requires a series of cognitive decisions. Some important questions are: Which planning procedures do describers use to produce a description ? Can descriptive strategies be identified in speakers’ or writers’ productions ? The study reported in this chapt deals with unconstrained descriptions of a spatial configuration. Subjects were presented with the map of a fictitious island bearing nine landmarks and were asked to produce a written description of the map. The descriptions were analyzed using an ATN-based system designed to construct a representation of descriptions and to classify them according to a typology. The texts contained a variety of descriptive sequences, most of which reflected highly systematic structures. The great majority of subjects produced descriptions from a survey perspective. Introductory statements providing addressees with a spatial framework were produced more frequently by subjects who adopted highly systematic descriptive strategies. These subjects also tended to use absolute modes of landmark location, as well as canonical spatial terms referring to cardinal points. Descriptive sequences were considered to reflect the structural organization of the subjects’ mental representations of the spatial configurations they had to describe.