engleski

Scalar models in a cognitive approach to hyperbolic expressions: With a little help from metonymy

The present paper is concerned with the cognitive status of expressions traditionally referred to by the name of hyperbole. Our chief aim will be to demonstrate that while the scalar model, as developed by Israel (1996, 1997, 1998, and 2001), is indeed involved in most cases of hyperboles, metonymic mappings must be nevertheless assumed to operate on this model to produce the right interpretation. The scalar model, developed primarily for polarity sensitive items, is a structured set of propositions ordered along one or more parameters in a way that supports inferencing. One of the problems with the model, as far as its utility in accounting for hyperboles is concerned, is that the region on the scale that is entailed is too large. If it is to be applicable to hyperbolic expressions, it should have some in-built mechanism that would stop inferences well before approaching default values. It will be demonstrated on large number hyperboles, that it is possible to conceive of hyperbolic expressions as being motivated by the part-for-whole metonymy involving a numerical scale and its upper or lower end, i.e. UPPER-END-OF-A-SCALE-FOR-WHOLE-SCALE or LOWER-END-OF-A-SCALE-FOR-WHOLE-SCALE metonymy. Metonymy seems to be the mechanism that singles out these measure-like, and some other, number terms placed somewhere within the scale (note that number names form in principle, if not in practice, an endless series) and promotes them to end points, practically obliterating all the other numbers between them and the default value. At the same time, the ground seems to be destroyed for all kinds of bizarre entailments that might have arisen if all the other numbers had remained in play. In other words, metonymy virtually constrains the inferencing process and thus makes possible the emphatic hyperbolic interpretation.

yperbole; scalar models; metonymy; metaphor

nije evidentirano