Voicing of Labiovelar Stops in Yoruba

This production study investigates the correlates of voicing in initial /b, k, , b, kp/ stops in Yoruba. We found “voiceless” /kp/ to be “partially voiced”. It has prevoicing, but at the same time it behaves like a typical voiceless stop in raising the f0 of the following vowel, and in having the tendency to decrease surrounding vowel durations. /kp/ and /b/ are furthermore distinguished from each other with respect to the duration of prevoicing. The single stops /k/ and //, on the other hand, differ not only with respect to voice onset time, f0 and vowel duration, but also with respect to the complete absence or presence of prevoicing. We furthermore focus on looking for voicing enhancement mechanisms in the stop voicing system. Preliminary analysis of EGG data and the intensity during the prevoicing seems to suggest that for voiced labiovelar stops, there might be a voicing enhancement mechanism that is stronger than in the case of voiced single stops.