Reduplication in Lusaamia

In Bantu reduplication, disyllabic minimality and maximality constraints often determine the shape of the reduplicant (RED), deriving it from the base; cf. Kinande (Mutaka and Hyman 1990), Kikuyu (Peng 1991) and Siswati (Kiyomi and Davis 1992, Downing 1997). In cases of partial reduplication, in which prosodic constraints determine the size of RED, it is usually clear whether reduplication is prefixal or suffixal: when the verb root is longer than two syllables (or two moras), only part of it reduplicates. Identifying the base and the reduplicant is then simply a matter of identifying the larger and smaller pieces. However, in languages where there is total reduplication, and no prosodic constraint governs the maximal size of the reduplicant, the process of identifying the base versus the reduplicant is more difficult. This paper discusses verbal reduplication in Lusaamia, a Bantu language spoken in Western Kenya and Eastern Uganda, in which there is total verbal reduplication, and in which morphological requirements determine the size of the reduplicant. The morphology, phonology and semantics of reduplication in Lusaamia are described herein, with the goal of identifying the base and the reduplicant. Drawing primarily on evidence from asymmetric patterns in the morphology of reduplication, it is argued that reduplication in Lusaamia is prefixal.