Prenasalized stops and speech timing

Abstract

Prenasalized stops raise a number of interesting questions concerning the relationship between phonological units and timing in speech. Fijian is a language in which prenasalized stops are incontrovertibly single segments from the phonological point of view. It has two stop series, voiced prenasalized and voiceless oral. All syllables are open and no consonant clusters occur. Hence, this is a particularly appropriate language in which to examine these questions. Utterance‐initial and word‐medial prenasalized stops at bilabial, alveolar, post‐alveolar, and velar places of articulation were examined in a set of words spoken by 11 speakers representing three main varieties of Fijian. Acoustic duration was measured from onset of nasal murmur to stop release (the stop component of a prenasalized stop is usually very brief and was not measured separately). Initial prenasalized stops are significantly shorter than medial ones. Medial prenasalized stops are comparable in duration to intervocalic laterals and voiceless stops, but initial prenasalized stops are significantly longer than the initial nasals in unstressed syllables which occur in the data set. Vowel length before prenasalized stops is comparable to that before other consonants. These results indicate that the role of prenasalized stops in the timing pattern of Fijian is generally similar to that of other single segments. In particular, they do not act like a consonant sequence.