Verbs, constructions, alternations: Usage-based perspectives on argument realization
My thesis examines to what extent the grammar of verbs (also called argument realization) can be based on usage. The usage-based approach is a recent paradigm shift in linguistics which takes the view that grammar is a dynamic inventory of symbolic conventions that emerges through, and is likewise shaped by, actual language use. Many studies have substantiated the claim that the structure of grammar is ultimately tied to usage, and no less may be said about that particular area of grammar concerned with argument realization. Yet, many models still rely on traditional assumptions about the nature of syntax and lexis inherited from research prior to the usage-based turn, such as the valency of verbs and the abstractness of syntactic patterns. Many of these assumptions have not been thoroughly examined from a usage-based perspective. Besides, while many scholars assume some degree of relatedness between verbal usage and argument realization, they often say little about what this relation exactly consists in. Accounts of the usage basis of argument realization are rather piecemeal, and to this day, there is no real complete theory of argument realization claiming to be usage-based from the start.
This thesis seeks to mend at least part of this gap. Adopting a constructional approach to argument structure and on the basis of English data, I address the question of the usage basis of argument realization at three levels of organization:
- At the lexical level (verbs), I investigate the frequency basis of valency, i.e., whether the traditional description of verbs as decontextualized predicates associated with a stable set of arguments proves adequate when confronted to the actual usage of verbs.
- At the constructional level (generalizations of argument structure over sets of verbs), I examine the relation between the meaning of a construction and the distribution of verbs occurring in that construction. It is argued in usage-based construction grammar that the distribution of verbs determines the constructional meaning and hence the productivity of a construction. Since a constructional approach assumes that constructional generalizations can be posited at several levels of abstractness, the question that I am asking is: at which level(s) can this view be maintained?
- At the cross-constructional level (generalizations over several constructions associated with the same event type), I argue that generalizations of a constructional meaning over formally distinct constructions (as captured by argument structure alternations) are cognitively plausible, although they have been largely under-studied in construction grammar, and I investigate whether there is any frequency basis for these higher-level generalizations.