Lingüística y Ciencia Cognitiva (LyCC)

Esta conferencia se ha financiado a través de un programa Ramón y Cajal (RYC-2010-0607) del MICINN

Current theories of aspect acknowledge the pervasiveness of verbs of variable telicity, and are designed to account both for why these verbs show such variability and for the complex conditions that give rise to telic and atelic interpretations. Previous work has identified several sets of such verbs, including incremental theme verbs, such as 'eat' and 'destroy'; degree achievements, such as 'cool' and 'widen'; and directed motion verbs, such as 'ascend' and 'descend'. As the diversity in descriptive labels suggests, most previous work has taken these classes to embody distinct phenomena and to have distinct lexical semantic analyses.

In Kennedy and Levin 2008, we suggest that it is possible to provide a unified analysis in which the behavior of all of these verbs stems from a single shared element of their meanings: a function that measures the degree to which an object changes relative to some scalar dimension over the course of an event. Focusing on the case of degree achievements, we claim that such "measure of change" functions are derived from two more basic concepts: an underlying measure function, which we take to be the basic denotation of expressions that are lexicalized in many languages as gradable adjectives, and a general operation mapping basic measure functions into functions which measure the difference between two objects on a scale, which underlies the semantics of comparatives.

However, while there are good reasons to believe that this kind of analysis is the right one for degree achievements, it is less clear how to follow through on Kennedy and Levin's suggestion and extend it to incremental theme verbs. In particular, as pointed out by Pinon 2008, such an extension must explain a crucial difference between the two classes of verbs: it is only in the latter that nominal reference makes a difference. It is therefore crucial that a fully general theory of variable telicity that is stated in terms of general features of scalar change not only explain the role of abstract scalar features in deriving (a)telicity, but also the role of nominal reference, in those cases (like incremental theme verbs) where nominal reference plays an important role.

The goal of this talk is to outline an approach that does exactly this. I begin by reviewing the analysis of degree achievements in Kennedy and Levin 2008, showing how (a)telicity is explained in terms of the scalar properties of the verb, and outlining some predictions about composition with degree expressions that this kind of theory makes. I then turn to incremental theme verbs, showing that they do not compose with degree expressions in the same way as degree achievements, concluding with Gawron 2007, Rappaport-Hovav 2008, and Levin and Rappaport-Hovav 2010 that they do not have scalar meanings. I show that Kennedy and Levin's analysis of degree achievements can be extended to an account of variable telicity in incremental theme verbs if we assume that the change scale comes not from the verb, but from a measure function associated with the incremental theme argument itself, something that is independently necessary to account for the distribution of quantity expressions inside DP (Krifka 1989, 1992; Schwarzschild 2006). Given the assumption that measure functions can in general be converted into measures of change, we derive a scalar semantics for "incremental themes" in which the referential properties of the nominal affect telicity of their verbal predicates in exactly the same way as the scalar properties of adjectives in degree achievements.