next up previous
Next: The concept of Up: Shape Schematization in Previous: Shape Schematization in

Introduction

The concept of schematization is all pervasive in many natural languages although it seems to have been discussed at length only in the context of English spatial prepositions. Examples of such studies include those by Talmy [Tal83] and Herskovits [Her86,Her95]. Discussions of schematization, notably the ones by Herskovits have been concerned with the need to adapt the ideal meaning of spatial prepositions to specific contexts. The purpose of the preliminary study reported here is to demonstrate the crucial relevance of shape schematization for the appropriate choice of classifiers in an Indo-European language called Assamese. This study illustrates the necessity of investigating diverse phenomena in diverse languages for understanding the role that schematization plays in language comprehension.

Talmy describes schematization as [Tal83]

a process that involves the systematic selection of certain aspects of a referent scene to present the whole, disregarding the remaining aspects.
Herskovits characterizes schematization as follows [Her95]:
a process by which a real physical scene, with all its richness of detail, is reduced to a very sparse and sketchy semantic content.
In this paper, we will show that schematization is needed not only for semantic purposes, it is needed for syntactic well-formedness as well. We will argue using examples of a language called Assamese.

Herskovits mentions that schematization involves two distinct processes [Her95]: abstraction and idealization. Abstraction distills away non-relevant characteristics of a scene whereas idealization is a process whereby non-ideal geometric features or characteristics of an object are treated as ideal geometric features.