On 'explaining' vowel duration variation.
| Number | 159 |
|---|---|
| Year | 1974 |
| Drawer | 3 |
| Entry Date | 06/11/1998 |
| Authors | Lisker, L. |
| Contact | |
| Publication | Glossa, 8, 233-246. |
| url | http://www.haskins.yale.edu/Reprints/HL0159.pdf |
| Abstract | Two well-known and generally accepted formulations on vowel duration are investigated: (1) The duration of the acoustic segment associated with a vowel depends, to a significant extent, on the degree of opening of the vowel; and (2) The duration depends also on the nature of a following consonant. Data are given from studies by Peterson and Lehiste (1960) on mean vowel durations, Peterson and Barney (1952) demonstrating mean vowel durations as functions of representative values of first-formant frequencies, Scharf (1962) demonstrating [(script a)] longer than [u], and Perkell's X-ray finding (1969) illustrating that although the tongue is higher for [ae] than for [(script a)], the mandible is lower for the former. The notion that the longer vowels are longer due to the mechanical inertia of jaw or jaw and tongue runs into difficulties when one considers vowel duration variation ascribable to differences of context. Reasons behind the following propositions are discussed: (1) Vowels are shorter before voiceless consonants; and (2) Vowels are lengthened before voiced stops. It is concluded that closure for the voiceless stop does not occur long after the devoicing gesture begins because the phonetic result would otherwise not be a sequence of vowel+voiceless top, but rather, vowel+aspiration+voiceless stop--a phonetic output unacceptable as normal English. |
| Notes |
Search Publications