Vocal attractiveness in homosexual and heterosexual listeners

The human voice is a sexually dimorphic trait that signals biologically and socially relevant information about the person who is speaking, including how attractive they sound to listeners. Biological factors that index reproductive fitness lead a voice to be perceived as attractive by heterosexual listeners. However, such a monolithic view may belie the role of social influences in a dynamic signalling system that is under extensive voluntary control. To the extent that homosexual listeners may not share the same vested interest in assessing reproductive fitness, they provide a valuable lens through which to dissociate biological from social influences on vocal attractiveness. In this study, equal numbers of female and male heterosexual and homosexual participants rated the attractiveness of voice recordings that had been masculinised or feminised, either synthetically by digital manipulation of recorded speech tokens (Experiment 1) or naturalistically by volitional modulation during speech (Experiment 2). We replicated previous findings that heterosexual listeners prefer voices that exaggerate rather than attenuate sexually dimorphic vocal cues. However, data from homosexual listeners provides a richer context, indicative of an overall bias for masculinised voices. Despite differences in how listeners responded to voice modulations, all groups displayed a global preference for unmodulated baseline voices that provide a more honest representation of the speaker. Likewise, there was a strong consensus across sexual orientations regarding which speakers had the most attractive unmodulated voices. Listeners maintain separate sets of preferences based on stable features of the voice that are driven by principles of biological signalling versus the dynamic social signals that it also carries. These findings demonstrate the added scientific value of widening the inclusivity of research.