bridging the divide: scalarity & modality
Klecha, Peter. 2014. Bridging the Divide: Scalarity & Modality. PhD Thesis, University of Chicago.

Keywords. modality, gradability, imprecision, modification, gradable modality

Abstract. This dissertation interrogates the boundary between scalarity and modality, two phenomena which have lengthy but largely independent histories in natural language semantic analysis. Particularly this dissertation examines cases in which the strength of a modal expression may vary across lexical or discourse contexts, where that variation can be analyzed as correlating with a conceptual scale of some kind. The upshot of this is that this kind of scalarity can occur in many different ways; this dissertation examines three. First, gradable modality, in which scalarity is an inherent semantic property of the modal. Second, imprecise modality, in which scalarity is a pragmatic property of use of the modal in communication. Third, modified modality, in which scalarity is induced extrinsically by another expression. None of these kinds of scalarity is special to modality; this is the major point of this thesis. A given modal may, or may not, exhibit these properties. Gradable modality has seen a recent focus in the literature, with many challenges to traditional models of modality. Against this background I argue that the existence of gradable modality does not fundamentally challenge the prevailing view of non-gradable modality. Imprecision, on the other hand, has not previously been argued to have an effect of modality, so this dissertation makes a novel claim: That modals are in fact subject to imprecision, which should be modeled as a pragmatic phenomenon, and that moreover this can be used to explain several long contested puzzles about modality. Finally, modification of modality has received very little treatment; I argue that modification of a modal can induce scalarity despite a lack of inherent scalarity in said modal, both by semantic and pragmatic means – a microcosm of the larger picture of scalarity in modality. The analysis therefore is heterogeneous – and this is just as we should expect. There is nothing about modals that should make us think that they are especially likely to be gradable or ungradable, precise or imprecise, modifiable or unmodifiable. But there is much insight to be gained by examining this variation.