Abstract:
This chapter provides a formal analysis of Dagbani vowel phonology, arguing that the surface forms of vowels emerge from: (i) faithfulness and markedness constraint hierarchies based on sonority, [ATR] and height features; (ii) prosodic conditioning and (iii) [+ATR] harmony. In non-final positions, mid vowels become [a] because they are marked in height specification. The preference for more sonorous vowels as syllable nuclei produces a hierarchy in which faithfulness to non-high vowels outranks faithfulness to high vowels. Prosodically-sensitive markedness constraints produce [i, ɨ, a, ʊ] in minimally bimoraic words. In sub-minimal words, an [ATR] markedness constraint hierarchy ensures that [i, e, o, u] are the only non-low [+ATR] surface forms. Rules of [+ATR] harmony produce [+ATR] variants of /a, ɛ, ɔ/ in nonfinal positions. The analyses demonstrate that in spite of the inherent differences between markedness and faithfulness-based approaches, analyses of harmonic patterns may require an eclectic approach.
