Korean L2 learners’ English vowel adjustments during interactions with two ASR systems: An AI assistant and a speech-to-text tool

• 발행기관: 경희대학교 언어정보연구소
• 저자명: Liying Bai Jae Yung Song
• 간행물 정보: 『언어연구』제43권 제2호, 563~592쪽, 전체 30쪽
• 주제분류: 인문학 > 언어학
• 파일형태: PDF
• 발행일자: 2026.06.30

국문 초록

There has been growing interest in using Automatic Speech Recognition (ASR) systems to support second-language (L2) pronunciation practice. This study examined how Korean learners of English adjusted their pronunciation in response to two ASR feedback modalities: auditory feedback from an AI assistant (Amazon Alexa) and visual text feedback from a speech-to-text tool (Google Docs). Nineteen Korean L2 learners produced 36 words containing six English vowels (/i, ɪ, ɛ, æ, ɑ, ʌ/). Following a misrecognition, they repeated the same word up to two additional times. Results from mixed-effects models showed that (1) recognition rates generally increased across attempts for most vowels, with the largest gains occurring between the first and second attempts; (2) when modifying their pronunciation, learners primarily relied on vowel lengthening rather than spectral adjustments; and (3) the two ASR systems appeared to elicit somewhat different adjustment patterns. Overall, ASR provided effective real-time feedback that prompted learners to modify their pronunciation. However, in the absence of explicit guidance, these adjustments were largely confined to temporal changes. These findings highlight the need for ASR activities and feedback to be carefully designed to facilitate pronunciation learning.

목차

1. Introduction

2. Methods

3. Results

4. Discussion

References

키워드

ASR feedback vowel pronunciation phonetic adjustment Korean EFL learners L2 English

참고문헌

• ReCALL
• Language and Speech
• International Journal of English Studies
• Journal of Memory and Language
• Innovation in Language Learning and Teaching
• Journal of Statistical Software
• The acquisition of phonology and phonetics
• Frontiers in Communication
• RELC Journal
• The Routledge handbook of contemporary English pronunciation
• Language Learning in Higher Education
• Computer Assisted Language Learning
• TESOL Journal
• Language Learning & Technology
• RELC Journal
• Proceedings of the 26th Annual Boston University Conference on Language Development
• Computer Assisted Language Learning
• Journal of Phonetics
• Computers in Human Behavior
• Journal of Contemporary Philology
• American Journal of Speech-language Pathology
• Optimizing online English language learning and teaching
• Technological resources for second language pronunciation learning and teaching: Research-based approaches
• Journal of Phonetics
• Computer Assisted Language Learning
• R package version
• Exploring language education
• Languages
• Proceedings of the 2024 Conference on Empirical Methods in Natural Language Processing
• Journal of Memory and Language
• System
• Journal of Psycholinguistic Research
• Proceedings of the 28th Annual Meeting of the Berkeley Linguistics Society
• CALL and Professionalisation: Short Papers from EUROCALL 2021
• Language and Speech
• ReCALL
• Journal of the Acoustical Society of America
• Frontiers in Psychology
• Proceedings of the 2017 CHI Conference Extended Abstracts on Human Factors in Computing Systems
• Computer Assisted Language Learning
• Frontiers in Communication
• Language Teaching Research
• Journal of Phonetics
• Interactive Learning Environments