The augment and {mu-} reduction in Bantu: An Optimality Theory analysis

Abstract

This article seeks to contribute to typology by presenting a formal comparative analysis of the effects of {mu-} reduction in three southern Bantu languages, viz., chiNambya, isiZulu and siSwati, using Optimality Theory (OT). The focus is on the prosodic structure of nasals in nasal-obstruent clusters derived from the deletion of /u/ following [m] in connected speech. In isiZulu, the resultant nasal is syllabic and does not assimilate to the place of articulation of the following consonant. In siSwati, the nasal is non-syllabic, functions as the coda of the immediately preceding vowel (the augment) and does not assimilate to the place of articulation of the following consonant. In both isiZulu and siSwati, {mu-} reduction does not trigger postnasal hardening and is blocked if the stem is monosyllabic. In chiNambya, the nasal that results from {mu-} reduction is neither syllabic nor functions as a coda. The resultant nasal-obstruent sequence is monosegmental, that is, the nasal is realised as part of the onset of the following vowel. In addition, the nasal assimilates the place of articulation of the following consonant and if the stem-initial consonant is a lateral approximant, postnasal hardening (strengthening) is triggered. If the stem is monosyllabic, the augment is retained for morphosyntactic and minimality reasons. These inter-linguistic differences allow for an OT analysis where a factorial typology accounts for the differences.