The perception of acoustically distorted speech produced with face masks in multilingual multi-talker environments

Abstract

This paper examines the perception of speech produced with face masks in multilingual multi-talker environments. Three groups of participants varying in language background listened to and reported English target sentences produced with or without a face mask in the presence of a competing English or Lithuanian talker. Listeners were monolingual native speakers of English, second language (L2) English speakers with Lithuanian as first language (L1), and L2 English speakers with L1 Mandarin Chinese. In addition, Lithuanian speakers also completed the same experiment but with Lithuanian targets. Results indicate that participants were more accurate with perceiving target sentences in their L1. Targets produced with a face mask were less accurately perceived across all groups regardless of listening in L1 or L2. In general, a competing talker in a language which matches the target (English distractor on English target) had a more detrimental effect on perception accuracy than a mismatched one (Lithuanian distractor on English target). Exceptionally, only when Lithuanian participants—with both English and Lithuanian knowledge – listened in their L1 was there no added challenge from matching distractor and target language. We conclude that acoustic distortions from face masks present an across-the-board difficulty while linguistic knowledge can reduce distraction from competing talkers.