| In this dissertation, I look at the relationship between causative and intransitive verbs in English from the viewpoint of Cognitive Linguistics. In Cognitive Linguistics, language is claimed to consist solely of symbolic pairings of meaning with phonological content. What are traditionally considered to be lexical items are in this theory symbols that are highly specified in both the semantic and the phonological poles, and grammatical structures differ from words only to the extent that they are more schematic in their representation. Moreover, linguistic knowledge cannot be viewed as separate from other aspects of cognition, but rather is integrated into our general cognitive abilities and processes. In this dissertation, I show how an account based on Cognitive Linguistics principles can account for the relationship between intransitive and causative forms.; In Chapter 1, I introduce the phenomenon and review several previous accounts. In Chapter 2, I discuss the theoretical preliminaries on which my account is based, including case theory, force dynamics, prototype theory, cognitive grammar, and blending theory. I then present my analysis of the causative as a blend between the intransitive form of the verb and a schematic causative verb. In Chapter 3, I examine the polysemy networks formed by five different verbs (open, float, walk, disappear, and marry) that have both intransitive and causative forms, in order to explore in more depth the interaction between the proposed blend and individual verbs. In Chapter 4, I address four hypotheses that arise from my analysis by analyze a wide variety of English verbs. In both Chapter 3 and Chapter 4, I analyze data from a variety of sources, including books, ATMs, and the Internet, as well as constructed examples. Finally, in Chapter 5, I discuss the implications of my research. First I discuss my justification for proposing a symbolic relationship that is phonologically empty. Next, I propose a taxonomy of verbs which have both causative and intransitive forms, and show how such a taxonomy can be expected if we consider the relationship a blend. Finally, I show how the proposal here differs in certain fundamental aspects from previous accounts of the relationship. |
