Prefixes, Sound Change, and Subgrouping in the Coastal Kenyan Bantu Languages

Abstract

This dissertation is an attempt to determine a basis of relationship for a cluster of Bantu languages including Swahili, several Pokomo dialects, and a number of Mijikenda dialects, which are spoken along or adjacent to the Kenyan Coast For a considerable "period of time it has been generally felt by Bantuists that these languages are closely related. This has been largely determined through inspection. There is no literature which actually discusses the oasis of this relationship. This dissertation is meant to fill that .gap by specifying some aspects of the morphophonological grammars of these languages, namely the sound changes which have modified the historically underlying system of prefixes, part of the nominal system of every Bantu language. The methodology employed is the Comparative Method and Internal Reconstruction as informed by current thinking and discussion on Generative Morphophonology. This simply entails working between the reconstructed units of the noun class system and the contemporary shapes to specify the ordered set of rules which relate the two sets. This differs from the traditional practice of CM and IR which work from the contemporary units to the reconstructed shapes. at this study attempts then is a reconstruction of the sound changes responsible for the contemporary shapes of prefixes in these languages. It follows a top-to-bottom rather than bottom-to-top model and is particularly interested in the middle ground--the set.of ordered changes which mediate the two points. The result then is a set of ordered rules which formalize the set of historical sound changes "Which are posited to explain the shapes of contemporary forms. Chapter I is an introductory chapter while Chapters II and III discuss and analyze the prefixes of the noun class system from each of the languages of the study. Chapter II concentrates on those changes that have largely affected either the consonantal or vocalic elements . of the prefix, whereas Chapter III discusses changes that have resulted from earlier changes affecting the prefixes. The bulk of the discussion in Chapter II is devoted to both the reconstruction of the Class 5 prefix and the changes affecting it. This departs from the usual approach where Meinhof's (1932) reconstruction is accepted as given. Chapter III is taken up largely by an analysis of the Class 9/10 nasal prefix, and part is given over to a discussion of coarticulated labislvelars in the mijikenda languages. Although the interest in these efforts is historical, a section of Chapter III does go into the implications of some of these changes for synchronic rule formulations. Chapter IV shows how the reconstructed sound changes form a basis J for classifying the languages of the study and how they define these languages as a genetically related group of languages. Chapter IV also surleys other East African Bantu languages to determine the place of xii the Kenyan cluster within this context. The result is the postulation of a Northeast Coastal Bantu group, which includes a large number of Tanzanian Coastal languages. The Kenyan Coastal group is shown to have greater affinities with these languages than with those Bantu languages which are spoken in Central Kenya. This calls into question the Shungwaya hypothesis which posits a common origin for Kenyan Coastal and Central Highlands peoples.