Differentiating between gay and heterosexual male speech.

Abstract

How many phones do listeners need to hear in order to discriminate between gay and heterosexual male speakers? Prior research [Munson et al. (2006) has found that listeners only need to hear a single monosyllabic word to make this determination. In a series of experiments, listeners heard monosyllabic words and portions of these words, and were able to discriminate between gay and heterosexual male speakers. Experiment 1 replicated the prior results. In experiment 2, listeners were above chance in differentiating between heterosexual and gay male speakers when presented with specific phonemes, such as /s/ and certain vowels (/eɪ/, /u/, /i/, /ɛ/, //aelig/). These data extend prior work demonstrating that gay and heterosexual men produce certain phones differently [Linville (1998); Pierrehumbert et al. (2004)]. When presented with other phonemes (/n/, /m/, /f/, /v/, /l/, /w/), listeners were not as accurate in differentiating between gay and heterosexual male speakers. These results suggest that listeners were able to accurately discriminate between gay and heterosexual speakers when presented with specific phonemes, although they were relatively more accurate in their determinations when presented with the entire monosyllabic word.