Exploring open consonantal environments for at-home testing of vowel perception in advanced L2 speakers

Recent work has called for increased investigation into methods used to explore second language (L2) speech perception (Flege, James E. 2021. New methods for second language (L2) speech research. In Second language speech learning: Theoretical and empirical progress, 119-156. Cambridge: Cambridge University Press). The present study attends to this call, examining a common practice for developing listening prompts in the context of at-home administrations. Vowel perception studies have historically used fixed consonantal frames to determine how well participants can discriminate between target L2 vowels, and the present study compares the effects of employing a fixed consonant-vowel-consonant frame (h-vowel-d) with open (phonologically diverse) consonantal environments using real and nonce words. Thirty-eight Mandarin (n = 31) and English (n = 8) first language speakers participated in a listening experiment and a post-experiment question. Data were framed within Best and Tyler's (2007. Nonnative & second-language speech. In O. S. Bohn & M. Munro (eds.), Second-language speech learning: The role of language experience in speech perception and production. A Festschrift in Honour of James E. Flege, 13-34. Amsterdam: John Benjamins) Perceptual Assimilation Model-L2. Internal consistency and proportion correct were calculated and a generalised linear mixed model design was used to investigate how well performance with h-vowel-d prompts predicts performance with the more diverse prompt types. Results suggest an inflation of scores for the fixed frame prompt and support the use of diverse words for listening prompt designs. Findings have implications for vowel perception researchers as well as computer (and mobile) assisted language learning developers wishing to inform their designs with relevant empirical evidence.