Comparing vowel hiatus resolution in ciNsenga and chiShona: An Optimality Theory analysis

Abstract

This article seeks to contribute to typology by presenting a formal comparative analysis of repair strategies used to resolve vowel hiatus in ciNsenga and chiShona. In these two languages, hiatus resolution is sensitive to phonology and morphosyntax such that hiatus resolution strategies apply differently depending on the phonological and morphosyntactic context. Across the prefix + noun stem boundary and within the Inflectional Stem, V1 undergoes “resyllabification” (Myers 1987:222) in the form of glide formation, secondary articulation and elision. An interlinguistic difference occurs when V2 is MacroStem-initial: in ciNsenga, hiatus resolution is blocked but in chiShona spreading is triggered. We follow Mudzingwa (2010) in proposing that resyllabification in chiShona is blocked at the Prosodic Stem edge by an alignment constraint (AlignL-PStem) that requires the left edge of a Prosodic Stem to align with the left edge of an onset-full syllable. We argue that resyllabification and glide epenthesis in ciNsenga are blocked when V2 is MacroStem-initial because Align (rootverb, l,σ,l) outranks Onset and AlignL-PStem. Crucially, this article demonstrates that whilst vowel hiatus resolution is categorical in chiShona, it is domain-specific in ciNsenga.