John Benjamins Publishing Company
Chapter 3. A corpus-based comparative acoustic analysis of target-like vowel production by L1-Japanese learners and native speakers of English
-
Abstract
This study combines acoustic phonetics, (applied) corpus linguistics, machine learning, and speech recognition to analyse the production of the monophthongal vowels / ɐ ɒ æ e ɛ i: ɪ u; ʊ ʌ / in the speech of L1-Japanese learners and L1-speakers of English based on transcripts and audio data from the Japanese spoken monologue section of the International Corpus Network of Asian Learners of English (ICNALE). The aim of this analysis is to evaluate what vowels L1-Japanese learners struggle with in terms of target-like vowel production and to provide insights into the determining factors causing divergencies from L1-English produced vowels. The results of a Multifactorial Prediction and Deviation Analysis Using Regression/Random Forests (MuPDARF) show that Japanese learners of English do indeed have difficulties in producing English vowels in a target-like manner but that these difficulties are confined to a relatively small set of vowels (/ ɪ u ʊ ɛ /). In addition, the analysis shows that difficulties are predominantly correlated with language-internal factors while language external factors (the age and gender of speakers) as well as their target variety and proficiency do not significantly correlate with non-target-like vowel production. The results suggest that Japanese learners of English can focus on specific vowels to enhance their target-like vowel production and that difficulties are caused by L1-interference due to a lack of phonemic vowel duration in Japanese and the similarity of Japanese and English vowels leading learners to use their L1 vowels rather than the slightly but notably different English vowel variants. The results can be used to raise awareness of L1-specific difficulties among this learner cohort due to their L1-background.
Abstract
This study combines acoustic phonetics, (applied) corpus linguistics, machine learning, and speech recognition to analyse the production of the monophthongal vowels / ɐ ɒ æ e ɛ i: ɪ u; ʊ ʌ / in the speech of L1-Japanese learners and L1-speakers of English based on transcripts and audio data from the Japanese spoken monologue section of the International Corpus Network of Asian Learners of English (ICNALE). The aim of this analysis is to evaluate what vowels L1-Japanese learners struggle with in terms of target-like vowel production and to provide insights into the determining factors causing divergencies from L1-English produced vowels. The results of a Multifactorial Prediction and Deviation Analysis Using Regression/Random Forests (MuPDARF) show that Japanese learners of English do indeed have difficulties in producing English vowels in a target-like manner but that these difficulties are confined to a relatively small set of vowels (/ ɪ u ʊ ɛ /). In addition, the analysis shows that difficulties are predominantly correlated with language-internal factors while language external factors (the age and gender of speakers) as well as their target variety and proficiency do not significantly correlate with non-target-like vowel production. The results suggest that Japanese learners of English can focus on specific vowels to enhance their target-like vowel production and that difficulties are caused by L1-interference due to a lack of phonemic vowel duration in Japanese and the similarity of Japanese and English vowels leading learners to use their L1 vowels rather than the slightly but notably different English vowel variants. The results can be used to raise awareness of L1-specific difficulties among this learner cohort due to their L1-background.
Chapters in this book
- 日本言語政策学会 / Japan Association for Language Policy. 言語政策 / Language Policy 10. 2014 i
- Table of contents v
- Chapter 1. Introduction 1
-
Crossing boundaries
- Chapter 2. New approaches to investigating change in derivational productivity 8
- Chapter 3. A corpus-based comparative acoustic analysis of target-like vowel production by L1-Japanese learners and native speakers of English 41
- Chapter 4. Digital Dickens 62
-
Crossing boundaries
- Chapter 5. 120 years of reporting clauses 100
- Chapter 6. Establishing a ‘new normal’ 125
-
Crossing boundaries
- Chapter 7. Syntactic segmentation of spoken corpus data 154
- Chapter 8. Short-term diachronic and variety-internal approaches to textual functionality in South Asian Englishes 192
- Chapter 9. Do corpus data on World Englishes inspire tolerance of variation in ELT professionals? 217
-
Crossing boundaries
- Chapter 10. Query a corpus in near-natural language 248
- 日本言語政策学会 / Japan Association for Language Policy. 言語政策 / Language Policy 10. 2014 263
Chapters in this book
- 日本言語政策学会 / Japan Association for Language Policy. 言語政策 / Language Policy 10. 2014 i
- Table of contents v
- Chapter 1. Introduction 1
-
Crossing boundaries
- Chapter 2. New approaches to investigating change in derivational productivity 8
- Chapter 3. A corpus-based comparative acoustic analysis of target-like vowel production by L1-Japanese learners and native speakers of English 41
- Chapter 4. Digital Dickens 62
-
Crossing boundaries
- Chapter 5. 120 years of reporting clauses 100
- Chapter 6. Establishing a ‘new normal’ 125
-
Crossing boundaries
- Chapter 7. Syntactic segmentation of spoken corpus data 154
- Chapter 8. Short-term diachronic and variety-internal approaches to textual functionality in South Asian Englishes 192
- Chapter 9. Do corpus data on World Englishes inspire tolerance of variation in ELT professionals? 217
-
Crossing boundaries
- Chapter 10. Query a corpus in near-natural language 248
- 日本言語政策学会 / Japan Association for Language Policy. 言語政策 / Language Policy 10. 2014 263
