Linguistics Home

Workshop on Scalar Meaning

University of Chicago
May 19-20, 2006

The ability to establish orderings among objects and make comparisons between them according to the amount or degree to which they possess some gradable property (such as height, brightness, beauty, etc.) is a basic component of human cognition. Natural languages reflect this fact: all languages have syntactic categories that express gradable concepts, and all languages have constructions that are used to express comparisons between objects with respect to gradable properties, measurement of such properties, and claims about the degree to which such properties hold of an object. Adjectives like tall, bright, beautiful and the constructions in which they appear (10 centimeters taller, extremely bright, as beautiful as a spring day in Rome, etc.) are canonical examples of such expressions, and have formed the central empirical domain for studies of scalar meaning in natural language.

A central conclusion of the research on these expressions over the past thirty years is that the semantic component of human language must include mechanisms for relating objects to abstract representations of measurement, or `scales'. This conclusion gives rise to a number of important questions, central among which are the following:

  • What are the linguistically significant features of scalar representations, and what factors determine the sort of scale that a given linguistic expression uses?
  • Do the scalar representations that have been claimed to underlie the semantic properties of gradable adjectives play a role in the semantic analysis of other grammatical categories?

The first question concerns the interaction between language and cognitive representations: how can the semantic properties of linguistic expressions that are used to convey scalar meaning be explained in terms of features of scalar representations, and what principles determine the kind of representation a particular expression is associated with? One of the core semantic properties of a gradable adjective is whether it has a context sensitive meaning or not. For example, large is highly context sensitive: it is used to make very different claims about 'degree of size' in large atom, large spider, large automobile, large city, large continent and large planet. Full, however, is not context-sensitive: it is used to make the same claim about 'degree of containment' in full glass, full tank, and full reservoir. Recent work (by e.g., Rotstein and Winter, Kennedy and McNally and others) has shown that this distinction correlates with a structural feature of the measurement scale used by the adjective - whether it is open (lacks endpoints) or closed (includes endpoints) - but the precise explanation for this correlation remains open, as does the question of what other features of scalar representations might be similarly important.

The second question addresses the broader role of scalar representations in natural language semantics. Answering this question is important because gradability is not just a property of adjectives, but of nouns, verbs, and prepositions as well. This point is made quite clearly in several important early studies of grading (in particular by Sapir and Bolinger), but it has only recently begun to be addressed in depth and detail by contemporary work in linguistic semantics.

In an effort to address these questions and to further advance our understanding of the role of scalar representations in natural language semantics, the Department of Linguistics at the University of Chicago will host a Workshop on Scalar Meaning that will bring together a group of researchers in linguistics and related fields (philosophy of language, cognitive science) who are actively engaged in the investigation of these issues, in a format that maximally supports interaction and discussion and the development of new ideas.