Influence of cross-linguistic exposure on speech perception under informational masking conditions: a cross-sectional study

Abstract Background Speech perception is often disrupted by competing sounds, which act as energetic or informational maskers. Although informational masking has been studied globally, its effects in multilingual settings where migration and exposure to unfamiliar languages are common, remain less understood. As individuals adapt to new linguistic environments, unfamiliar speech, initially perceived as background noise, can become cognitively intrusive as it gains phonological and semantic relevance. Methods This study aimed to investigate how prolonged exposure to Kannada, a non-native language, influences informational masking in native Malayalam speakers. Two groups of normal-hearing, native Malayalam speakers (102 participants) were tested: one with no exposure to Kannada (NKE) and another with four years of exposure (KE). Participants were asked to identify Malayalam digits presented against five masker conditions: Malayalam babble, Kannada babble, Malayalam reverse babble, Kannada reverse babble, and speech-shaped noise (SSN). SNR50 was calculated under each masker condition and was utilized to derive combined informational masking (CIM). Results SNR50 scores varied significantly across masker types ( p < 0.001), with best performance under SSN and poorest under Malayalam babble (MB). A significant main effect of masker type and an interaction between masker type and language exposure group were observed ( p < 0.001). Independent t-tests showed that participants with KE had significantly poorer SNR50 scores under Kannada babble (KB) and Kannada reversed babble (KRB) maskers, compared to the NKE group. No differences were found under Malayalam or SSN conditions. CIM caused by Kannada maskers was significantly greater in the KE group ( p < 0.001), whereas no group difference was observed for Malayalam maskers. Discussion Findings indicate that masker intelligibility and linguistic familiarity jointly influence informational masking. Although all participants showed maximal masking under native (Malayalam) babble, those exposed to Kannada exhibited heightened susceptibility to interference from Kannada speech maskers. This suggests that even limited non-native exposure (four years) can transform speech maskers from unintelligible noise into linguistically meaningful, cognitively demanding maskers, increasing informational masking. Conclusion These findings highlight how linguistic exposure alters brain’s processing of competing speech, emphasizing the need to consider both listener language background and masker characteristics when assessing speech perception in multilingual environments.