| Summary: | This dissertation is a typologically informed, comparative study in the lexical semantics of stativity. The starting point is a tension between standard formal approaches and the cross-linguistic landscape. Formal frameworks commonly adopt on ontology with states, which function as the implicit arguments of verbal predicates and (on some theories) adjectives. Yet patterns of stative predication vary dramatically across languages: N, V, A, and pre-categorical roots are all attested as basic types for stative lexemes, and predication can occur directly or require additional compositional structure. Due to the relative underdevelopment of formal approaches to cross-linguistic semantics, the status of lexical statives and the implications of morphosyntactic variation remain unclear.
Data collected in extensive fieldwork on Wolof (Senegambian, Senegal) is the empirical foundation for a new approach to lexical stativity in semantic modeltheory. Wolof has a dual system for stative predicates: some are lexicalized directly as verbs, while others are built compositionally from mass nouns and possessive morphology. The latter strategy is cross-linguistically widespread, but has received little attention in the semantics literature. The facts that emerge from the Wolof case study point to a unified treatment of statives as a natural class of meanings: while denotations vary predictably based on lexical category, this is mitigated by compositional predicate formation.
Finally, a theory of stativity is proposed which accounts for the relevant constructions in Wolof and their English counterparts. On my approach, the role of states in semantic theory is elevated: besides their traditional referential argument role, equivalence classes of states are argued to function as a proxy for degrees in gradable constructions like the positive, comparative, equative, and superlative. This move is shown to be advantageous in that it naturally links the scalar properties prototypically associated with adjectives to their lexical stativity rather than their lexical category---a move further motivated by adjectiveless languages like Wolof. Moreover, cross-linguistic variation is readily explained with the enlargement of the logical space for representing stative predication: whether lexical or compositional, gradable or non-gradable, the rich diversity of statives is captured in a unified formal system.
|
|---|
