The Perception of (Trans)masculinity in Speech: Effects of Acoustic Characteristics and Rater Identity.

Purpose: Gender-affirming communication services are based on studies of speech produced and perceived by cisgender men and women. The current study examined the perception of gender and gender orientation (i.e., whether someone is cisgender or transgender) in the Palette of Voices, an openly available corpus of the speech of transgender and cisgender men, by cisgender heterosexual men (CHM) and cisgender heterosexual women (CHF), and a group of gender and sexuality expansive (GSE) listeners. We examined how both the acoustic characteristics of speech and listener identity affect gender and gender orientation categorization.

Method: Participants (n = 199) categorized the gender and gender orientation of 240 sentence productions produced by 20 male talkers in an online experiment, including tokens whose fundamental frequency (F0) and formant frequency scaling had been altered, and unmanipulated tokens.

Results: Consistent with previous research, productions with lower F0 and lower formant frequencies were more likely to be categorized as male than ones with higher F0s and formants. The weighting of these variables differed systematically across listener groups, with the GSE group weighting these variables less than the CHM and CHF groups when categorizing gender, but more when categorizing gender orientation.

Conclusions: The relationship between the acoustic characteristics of a talker's speech and the categorization of their gender and gender orientation is highly variable across and within groups. The perception data and speech samples in this study are openly available. Suggestions are given for how they might be used to supplement existing gender-affirming communication services.