Intelligibility of speaking styles elicited by various instructions

Abstract

The acoustic-phonetic modifications made by talkers are attuned to the specific communicative situations that listeners are experiencing (Lam and Tjaden, 2013; Hazan and Baker, 2011). The extent to which such modifications are under explicit control remains largely unknown. This study examined the extent to which native and non-native talkers can implement acoustic-articulatory enhancements following specific instructions and the extent to which these changes will improve intelligibility. Ten native and 10 Korean-accented talkers read sentences in various styles including in conversational and clear speech, while imitating a native speaker’s conversational and clear speech, with exaggerated vowels, more slowly, and more loudly. Sentences were mixed with noise (−5 dB SNR) and presented to native listeners. Intelligibility results revealed that nonnative talkers were overall less successful in enhancing intelligibility following different instructions compared to native talkers. Instructions to speak clearly and imitating native clear speech sentences provided the largest intelligibility benefit while the instructions to slow down were least successful in improving intelligibility across talkers. Speaking loudly and exaggerating vowels increased intelligibility only for native talkers. Acoustic analyses will examine which acoustic-phonetic changes were implemented following each instruction. The results have important implications for enhancing intelligibility in difficult communicative situations (e.g., classrooms).