Abstract
Speech samples obtained from a group of adult males and females while they articulated the tone produced by a single-frequency electrolarynx were played to a panel of listeners who were asked to determine the sex of each speaker. Formant frequencies for the vowels /i/ and /u/ produced in isolation by the same speakers were obtained by spectrographic analysis. The judges were able to correctly identify the sex of the speakers 88% of the time, and vowel formant frequency averages were closely associated with the degree of male or female quality in these voices. These findings indicate that in the absence of between-subject variations in laryngeal fundamental frequency, a distinct cue to speaker sex may be contained in the individual vocal tract resonances.