AN OVERVIEW OF SPATIAL LANGUAGE

The prior chapter has provided the conceptual underpinnings to appreciate both the striking variety of spatial coordinate systems to be found in language, and the relatively small set of underlying principles from which they are constructed. In this chapter, I sketch how the three basic frames of reference get instantiated in different languages. Here we will be concerned however not with the detailed grammar, morphology and lexical details of different languages – for that the reader is urged to see the companion volume (Levinson and Wilkins in preparation) – but primarily with the relevant semantic parameters, and how various combinations of these get variably encoded. To set the frame-of-reference facts in proper perspective, it will also be useful to mention other (non-frame-of-reference) semantic fields in spatial language, to make clear how they relate and how they are different from frame-of-reference information.

A serious overview of what is known about spatial language would be a book in itself. It is moreover a field of study dominated by preconceptions based on familiar languages – for example the presumption that the most important aspects of spatial language are encoded in adpositions (prepositions or postpositions). This presumption has been elevated to theoretical prediction by, for example, Landau and Jackendoff (1993), to the effect that spatial relations will express only a few aspects of ‘gross geometry’ of the ground or reference object, and will be coded in just a few closed form-classes, principally adpositions.