No description available.

This thesis explores three fundamental issues in the phonology and morphology of Ethiopian Semitic languages: mobile morphology, reduplication and epenthesis. In each chapter I draw on comparative evidence from different Ethiopian Semitic languages, an approach which provides greater insight into how the languages vary with respect to these three issues, and how the issues themselves are best analyzed. / The first issue is that of 'mobile morphology' a term I coin to describe the ability of a particular morphological category to be realized on various segments within a stem. The two major types in the South Ethio-Semitic languages are palatalization and labialization. I develop an analysis of palatalization in five different languages which relies on a hierarchy of preferred targets, along with a number of constraints regulating the appearance of palatalization within the stem. / Ethio-Semitic languages have several different types of reduplication. I draw a distinction between phonological and morphological reduplication and argue that phonological reduplication should be viewed as copying rather long-distance geminate structures created by spreading. I also examine the interaction of reduplication with mobile morphology and I present an analysis of double reduplication, showing how languages will avoid the creation of double reduplication relationships. / I develop an analysis of epenthesis which contrasts the behaviour of one set of languages which epenthesize following final consonant clusters with other languages which epenthesize between consonant clusters. I show that while all Ethio-Semitic languages follow the same general pattern, this may be overridden by templatic constraints and more importantly, by sonority considerations holding of adjacent syllables in coda-onset sequences. This last observation is important because it shows that while languages may on the whole violate heterosyllabic contact constraints, in particular circumstances, the constraints will be obeyed, giving rise to an emergence of the unmarked scenario.