Gender Neutralizing One’s Voice: Voice Quality and Non-binarity
Résumé
GOALS: to make clear: 1) What NoBiPho Project means by the working definition of “non-binary” (NB): 2) Why we chose to explore spectral slope ; 3) If and how spectral slope is utilized by assigned-female-at-birth non-binary speakers in the Gender Spectrum Speech corpus (GSS, 2019) to perform their gender.
More accurately, do non-binary (NB) speakers produce a different voice quality (as measured by mean fundamental frequency and spectral slope) to index their gender as compared to binary speakers?
METHOD : Five female non-binary participants (21 to 30 years old ; two of whom have benefited from testosterone therapy), four male (21 to 32 years old) and 5 women (19 to 22 years old) read a text and spontaneously spoke on topic of “voice”. All subjects have a variety of experiences with their vocal training.
RESULTS: The non-binary speakers produced a significantly less negative spectral slope in both the read and spontaneous speech contexts than did the cis women; this suggests that the non-binary speakers have a “brighter” voice quality than do the cis women and potentially perceptually than the cis men (but not statistically different). There was no statistically significant effect of voice training or testosterone therapy. No statistical differences between cis women and cis men were found concerning spectral slope. The mean fundamental frequencies of read and spontaneous utterances of non-binary subjects are generally intermediate between those of males and females. More data is needed to confirm these results.