PhD Project 18

Processing vague expressions: The interplay between semantics, pragmatics and cognition


PhD-candidate: Arnold Kochari
PIs: Robert van Rooij (WP2) and Herbert Schriefers (WP3)
Start date: 01 September 2015

(last update 2019-06-27)

Research Content

Scalar adjectives such as ‘large, ‘long’, and ‘tall’ are highly context sensitive: their meaning largely depends on which exact nouns they combine with (e.g., the size that the adjective “large” refers to would be different if combined with a “mouse” or a “horse”). These adjectives therefore express different meanings in different situations. Nonetheless, they are some of the most frequent words in language, and humans have no trouble understanding these adjectives when they are used. In order to find out how humans understand their meaning, the current project is taking into account the properties of the human cognitive system for perceiving and retrieving magnitudes – the so-called generalized magnitude system. Specifically, experiments are carried out comparing how people process numbers and scalar adjectives since both numbers and scalar adjectives refer to magnitudes with the difference being that numbers refer to precise quantities, whereas adjectives refer to magnitudes without giving any specific value. The second approach that is taken in this project is investigation of adjective-noun meaning combination processes in case of scalar adjectives and in case of adjectives the meaning of which does not depend on the context (e.g. ‘dead’, ‘square’). Here, the idea is to compare the time-course of combination of meaning of different types of adjectives and a noun: the difference in the requirement for contextual information should be observable in terms of how soon the adjective and the noun meaning can be combined.

Progress 2018

A set of experiments comparing processing of Arabic digits (e.g. 3, 5), number words (e.g., ‘three’, ‘five’) and scalar adjectives (e.g. ‘long’, ‘short’) was finalized and results were written up in a manuscript. These experiments found parallel effects with all three types of stimuli. These results support the hypothesis that both adjectives and numbers refer to magnitudes. This manuscript is currently under revision and a follow-up project is in the planning phrase.

Work in the second line of research also started where the aim is to compare the time-course of processing of scalar and other adjectives. An experiment was designed in which participants will read phrases consisting of a scalar adjective and a noun (e.g. ‘large animal’) and phrases consisting of an adjective not dependent on the context and a noun (e.g., ‘dead animal’). Here, the hypothesis is that the meaning combination will be delayed for scalar as compared to other adjectives. To carry out this project, expertise in MEG methods was needed, and therefore invited LiI postdoc Lewis (see also page 26) to collaborate which substantially improved the quality and widened the scope of the project.

Groundbreaking characteristics

The project involves a novel combination of disciplines, namely empirical linguistics and philosophical logic with experimental psychology. The investigation of vague expressions will give new insights on how these are interpreted by humans.