Abstract

A speech may be perceived differently depending on what social information we draw from the speaker, be it correct or misguided (Burgers & Beukeboom, 2020; Edwards, 1999; Lambert et al., 1960; Niedzielski, 1999). The phenomenon where non-linguistic social information about the speaker (e.g., race, occupation, etc.) influences our actual experience with speech is called reverse linguistic stereotyping (RLS; Kang & Rubin, 2009). Although RLS has been well documented for its influence on speech perception (e.g., seeing an Asian face can render a speech less comprehensible and more accented), its broader impact on speakers is still less known. This study, therefore, examined how L2 speakers’ assumptions about examiners influence their speaking anxiety in an oral exam.

Participants included 40 Mandarin-speaking international students in Montreal, who completed two English speaking tests delivered through video prompts. Each test featured a different examiner (Caucasian or South Asian), while the audio remained constant (Canadian English). Participants rated their speaking anxiety before and after each test and evaluated each examiner. Retrospective recall interviews were conducted with eight individuals who showed noticeable difference in pretest anxiety ratings across two examiners. Results revealed no significant difference in speaking anxiety in Test 1. However, a higher pretest anxiety was observed in Test 2 when the examiner appeared South Asian than Caucasian, possibly due to a shift in visual stimuli, which activated the stereotypical association between examiner’s race and linguistic ability. The findings highlighted the importance of creating a more inclusive environment in both language learning and assessment.