Acoustic variability in the vowels of female and male speakers.

Abstract

When vowel formant values for female and male speakers are normalized and plotted in an F1–F2 space, it is striking that the quadrilateral spaces for the females are uniformly larger than those for the males. Data will be presented from seven languages and dialects, including three dialects of English, which indicate that, all other things being equal, female speakers appear to produce vowels in a manner that is more phonetically explicit than males do. It is particularly in the F1 dimension that the females’ quadrilaterals extend beyond the males’; it may be inferred that female speakers articulate vowels with a more open mouth than males do. Can one expect that speakers with a higher F0 automatically have a larger vowel space? Are females making more of an effort to keep vowels distinct, and potentially contribute to greater intelligibility? Or do females over-articulate, avoiding reduced or centralized forms, as a result of social expectations to be conservative ‘‘guardians’’ of the language? Acoustic and sociophonetic answers to such questions will be offered. In addition, data will be presented from three dialects of English which show that female speakers are not uniform in their behavior: some females merge vowels more often than males do, while other female populations appear to differentiate the same vowels more systematically. The perceived social prestige of an accent is offered as one explanation for these disparate directions of change.