Perceiving the intersection of gender identity, sexual orientation, and gender expression in American English

Abstract

Listeners make assumptions about speaker gender identity (GI), sexual orientation (SO), or gender expression (GE) from speech (Tripp and Munson, 2021). Researchers have identified correlations between acoustic cues such as center of gravity (COG) of /s/ and perceived speaker’s identity. In this study, I investigate how a multitude of acoustic features combine to influence nuanced perception of GI, SO, and GE in speech. In an online study, 197 listeners rated the speech of 66 speakers of American English for perceived GI, SO, and GE. Acoustic measures for f0, vowel formants and dispersion, segment durations, fricative spectra, and creaky voice were correlated with these judgments using analyses such as random forests and PCA. The results indicate that ensembles of acoustic features, including well-known features such as f0 and COG of /s/ but also creaky voice and diphthong formants, contribute to the perception of GI, SO, and GE such that common perceived social categories emerge. I discuss the complex interactions of acoustic cues that describe the ways in which listeners construct a speaker’s identity along traditional binaries such masculine or feminine as well as outside those binaries for queer identities.