Personal pronouns and person perception: Do paired and nonbinary pronouns evoke a normative gender bias?

Research on gender-fair language aims to identify language inclusive to a multitude of individuals, for example, increasing the visibility of women by using paired pronouns (he/she) instead of generic masculine forms (he). However, binary presentations like he/she might come with unwanted side effects and evoke what we label as normative gender bias. A normative gender bias is defined as when words lead to stronger associations with individuals with normative gender expressions than with individuals with non-normative gender expressions, thus contributing to making non-normative individuals invisible. In three experiments, we compared the extent to which the paired pronoun he/she (Swedish and English), the neo-pronouns hen (Swedish), ze (English), and the generic pronoun singular they (English) evoked a normative gender bias. Swedish- (N = 219 and 268) and English- (N = 837, from the UK) speaking participants read about individuals referred to with the paired pronoun he/she or with hen, ze, or they. In Experiment 1 (Swedish), there was no main effect of condition on a normative bias, but in Experiment 2 (Swedish), the paired pronouns he/she evoked normative gender bias while hen did not. In Experiment 3 (English), both ze and singular they evoked normative gender bias, although normative associations were lower in these conditions compared to he/she. Furthermore, the normative bias was lower among participants who had knowledge about the use of ze as a nonbinary pronoun. Finally, neither ze nor they evoked a normative gender bias when their use was explicitly stated to be nonbinary. A potential explanation for why singular they did not generally result in less normative associations, despite almost all participants knowing about it, may include its more common use as a generic pronoun. Taken together, our results suggest that neo-pronouns, but not paired pronouns, have the potential to evoke less normative associations, but that they must be both (1) actively created new words and (2) well-known to language users as nonbinary pronouns.