Tapping into second language learners' musical intelligence to tune up for computer-assisted pronunciation training

In line with rapid advancements in educational technologies and Computer-Assisted Language Learning, it seems that innovative approaches in Computer-Assisted Pronunciation Training (CAPT) must also be updated in order to address the growing needs of communication and intelligibility in modern-world multilingual contexts and AI-mediated machine interactions. Furthermore, it appears that the affordances of incorporating musical intelligence into pronunciation training, despite the probable links between the two, remain understudied. The present mixed-methods study was an effort to address this gap by developing a technology-mediated melodic approach to teach the pronunciation of 93 vocabulary items to 39 EFL learners. The technique used Adobe Audition and Ableton Live 10 to create recurrent and incrementally progressive melodies based on word pronunciations for practice purposes in the experimental group and Automatic Speech Recognition (ASR) to measure the accuracy of pronunciations. Results of Wilcoxon Signed Rank Test showed a significant improvement in the experimental group while a between-groups analysis employing Quade test did not locate any significant differences between the two groups. In the qualitative phase, thematic analysis of interviews with 15 learners revealed mostly positive attitudes in users, who found the approach melodic, stimulating, engaging, lasting, and convenient. It is concluded that while CAPT technologies are capable of offering positive outcomes, analogous to the ones obtained by teacher-fronted training, their appeal to learners and other advantages can make them viable candidates for promoting autonomous and ubiquitous L2 pronunciation learning.