A Neo-Kleinian Approach to Comparatives

INTRODUCTION. Adjectives such as tall get a high degree meaning when used in isolation. Thus, John is tall does not simply mean that John has (a certain) height, but rather that he is taller than average. This led Cresswell (1976) to argue that gradable adjectives are relations between individuals and degrees and to postulate the operator pos which binds the degree variable of the adjective when used in the positive form. Given the fact that pos is invisible, its existence is based on circumstantial evidence. Not surprisingly, it has been claimed that pos does not exist: Klein (1980) argues that pos ‘is merely a device for fixing up the semantics’. In his framework, pos is not necessary as adjectives such as tall are interpreted as the property of being tall, where what counts as tall depends on the context. A major advantage of a Klein-style analysis is that gradability depends on orderings, and as such is predicted to be found outside of the adjectival system, a prediction which is born out (see Doetjes 2008). The distribution of degree modifiers such as less and more in English and for instance hodně ‘a lot’ in Czech or trop ‘too (much)’ in French, that combine with adjectives but also with nouns and verbs, is most easily understood if one assumes that gradability follows from the existence of an ordering, rather than from the presence of a degree variable. However, it has been shown that Klein’s theory cannot account for a number of phenomena (see, e.g. Kennedy 1999, 2001, Bale 2006, 2008). In this talk, we will argue in favor of a modified version of Klein’s theory, which allows us to maintain a non degree based interpretation for gradable adjectives. We will show that a number of important insights from both Kennedy and Bale can be accommodated in this modified version.