More Velar than /g/: Consonant Coarticulation as a Cause of Diphthongization
Adam Baker, Jeff Mielke, and Diana Archangeli
60-68 (complete paper
or proceedings contents
In Western American English, there is no contrast before the velar nasal between short-a ([ae]) and long-a ([ei]); in some dialects, there is also a merger of [ae] and [ei] before [g] (e.g. in hag and (The) Hague). The changes in formants from [ae] to [ei] are consistent with a conventionalization of the coarticulation with the following velar (the "velar pinch"). From attested sound patterns, there is the implication statement: If [ae] > [ei] / __ [g], then [ae] > [ei] / __ [engma], since there are no known cases of [ae] becoming [ei] before [g] but not before [engma]. It appears then that [engma] is more velar than [g]: it has a greater velarizing effect. This study tests the hypothesis that [engma] causes greater velar coarticulation because, in production of [engma], both the tongue dorsum is raised and the velum is lowered. The test is to measure velum height indirectly through nasal airflow and to examine the correlation between velum lowering and F1 and F2. A linear regression shows that as nasal airflow increases, F2 raises and F1 drops, suggesting the precise phonetic motivation for the [ae] > [ei] allophony.
Proceedings of the 26th West Coast Conference on Formal Linguistics
edited by Charles B. Chang and Hannah J. Haynie
Table of contents