Phonetic cues, indexical fields, and the perception of gender and sexual orientation

Eckert (2008) asserts, based on empirical findings, that cues exist within indexical fields: Each cue is indexed with multiple potential meanings, one of which is activated by the perspective of the listener as well as the context and content of the utterance. What is less established is how multiple cues, indexed with inter-related potential meanings, are effectuated. For example, when three different cues individually indexed with gender and sexual orientation are simultaneously presented and gender percepts are elicited, which cue is more attended to, and will this cue also dominate in sexual orientation perception tasks? One possible way to answer these questions is to use cue weighting theory and methods, which primarily stem from acquisition research. Cue weighting theory holds that cues are weighted relative to each other (Mayo et al., 2003, 2005; Holt & Lotto, 2006). However, the majority of cue weighting studies is limited to only two cues, and compares two types of listeners (e.g., adult and child listeners) to determine which group uses which cue in phonemic or tone categorization. It is therefore unclear whether cue weighting can account for how cues are utilized in social categorization tasks. In the present study, experimental methods incorporating three cues indexed with gender and sexual orientation are used in perception tasks. The empirical outcomes allow for the determination of cue weighting strategies as well as more in-depth investigations into the relationships of the three indexical fields. Two interesting conclusions are generated. First, highly interactive relationships amongst the cues occur in all analyses: for perceived gender (n=14,099 observations from 105 participants), perceived women’s sexual orientation (n=6071) and perceived men’s sexual orientation (n=8028). This suggests that indexical fields interact, and that cue potential is realized in part by other available cues integral to the given social categorization task. Second, rather than have individual meanings indexed to phonetic cue variants, the results suggest that a gender prototype is compared to the stimulus and cues are interpreted relative to it resulting in markedness features for each variant. Furthermore, cues that affect or block the interpretation of other phonetic cues suggest that there are markedness features of varying strengths for each variant. This analysis predicts perceived gender and women’s perceived sexual orientation results well, however the results are less clear for men’s perceived sexual orientation.