ASPECTS OF COMPLEMENTATION IN THREE BANTU LANGUAGES

Carolyn Perez

Ph.D. Dissertation, 1985

Abstract

This dissertation analyzes aspects of verbal complementation in three of the Bantu languages of Africa, Shona (Zimbabwe), Kikuyu (Kenya), and Kiruundi (Burundi), within the framework of the Government-Binding theory of syntax. The starting point of the analysis is the observation that these languages permit NP-movement from the subject position of a tensed complement clause. The data thus constitute apparent counterexamples to Government-Binding theory as formulated in Chomsky (1981), particularly the subtheories of abstract Case and Binding. The intent of this dissertation is to explore the consequences of an analysis of these data which incorporates changes in the theory as applied to Bantu which were inspired by this initial observation. Chapter 1 presents the primary observation about NP-movement and additional data to show that this movement is optional, that no infinitive complements are permitted, and that a dependency relation exists between the matrix and embedded subjects. The theoretical interest of the data for the subtheories of Case and Binding is explained. Also, a brief Appendix explains points of Bantu morphology and syntax in order to make the data in subsequent chapters easier for the non-Bantuist to understand. Chapter 2 argues that the subtheory of abstract Case is not operative in the three languages, basing this conclusion on the data from NP-movement and impersonal passives. Ungrammatical data which could be ruled out by Case theory are given alternative accounts, and it is proposed that the connective particle need not be interpreted as a Case-assigner. The theory of Binding is considered in Chapter 3, where it is proposed that Principles A and B apply in different structural domains, based on data from NP-movement and overt pronominals. This analysis leaves open the possibility of governed PRO, and it is suggested that this possibility is realized in Kiruundi. Chapter 4 examines various uses of infinitives, primarily in Kikuyu. It is proposed that infinitives may be divided into two groups, nominal and verbal, based on properties which are attributed to the infinitival subject.