Prenasalized consonants in Bantu: a cluster or a unit in Optimality Theory

Prenasalized consonants in Bantu: a cluster or a unit in Optimality Theory

The present study attempts to investigate the status of prenasalized consonants in Bantu within the framework of Optimality Theory. Following Herbert (1975, 1977, 1986), Hayes (1991), and Hyman (1992), this paper proposes that the prenasalized consonant, i.e., the sequence of a nasal consonant and the following oral consonant (NC), is in principle a cluster in the underlying representation, while they are unified and then realized as a unit segment in the surface representation. In this view, the prenasalized consonants in their underlying representation are ambisyllabic. In other words, in the underlying representation the nasal component of prenasalized consonants is the coda of one syllable; the oral consonant of prenasalized consonants is the onset of the next syllable. Several phonological phenomena, such as vowel lengthening, the nonexistence of NC-initial and NC-final stems, and the syllabification of NC-initial words, may afford a basis for this analysis, providing that the prenasalized consonants are two distinctive segments. By facilitating the constraint-ranking theory (McCarthy and Prince 1993, Prince and Smolensky 1993), this unification of the underlying nasal and following consonant sequence can be successfully accounted for.