Using AR filters in L2 pronunciation training: practice, perfection, and willingness to share

Traditional L2 pronunciation teaching puts too much emphasis on explicit phonological knowledge ('knowing that') rather than on procedural knowledge ('knowing how'). The advancement of mobile-assisted language learning (MALL) offers new opportunities for L2 learners to proceduralize their declarative articulatory knowledge into production skills more efficiently through systematic practice. Within the framework of skill acquisition theory, and using a mixed-methods design including production tests, articulatory awareness quizzes, and a semi-structured interview, this study aims to investigate the effectiveness of using an augmented reality (AR) filter app during pronunciation practice to improve Chinese L2 students' segmental production and articulatory awareness of the targeted English consonants (dark/e <</,/o/and/theta/), and their attitudes towards this application. The experimental results indicate that the app with AR technology, facial detection technology, and digital zoom technology played a positive role in promoting students' segmental production and raising their articulatory awareness. The interview results show that the AR filter helped to alleviate the learners' speaking anxiety and embarrassment related to interacting with the teacher for corrective feedback, and encouraged daily practice by providing more fun. Combining all the qualitative and quantitative results, we conclude that an AR filter app can offer academic, psychological, and social support for L2 pronunciation skill learning.