PulseExploreJournal ClubDebatesTrendingResearchersJournals
Instagram
HomeExploreJournal ClubTrending
Synapse
⌘+K
Synapse
June 1, 20260 citationsOpen Access

Connected Speech in Japanese EFL Learners: Obstacles and Instructional Pathways in Pronunciation Acquisition

TITakeshi ISHIHARA

Key Points

  • This study investigates the challenges faced by Japanese EFL learners in acquiring connected-speech features in English.
  • Analyzed controlled read-aloud data from learners at A2–B1 levels of CEFR under uninstructed and instructed conditions.
  • Compared various connected-speech phenomena including assimilation, linking, and elision.
  • Nasal assimilation, high-frequency linking, and /h/ and schwa elision occurred naturally in uninstructed conditions.
  • Stop assimilation, yod-coalescence, cluster elision, and polysyllabic reduction required explicit instruction.

Abstract

This study examines how Japanese learners of English as a Foreign Language (EFL) acquire English connected-speech phenomena, with a focus on assimilation, linking, and elision. Due to typological differences between English and Japanese phonology, learners often produce conservative, syllable-by-syllable articulation, resulting in reduced fluency and intelligibility. Using controlled read-aloud data from learners at the A2‒B1 levels of the Common European Framework of Reference for Languages (CEFR), the study compared productions under uninstructed and instructed conditions. Results showed that some processes (nasal assimilation, high-frequency linking, /h/ and schwa elision) surfaced naturally, whereas others (stop assimilation, yod-coalescence, cluster elision, polysyllabic reduction) required explicit instruction. Findings highlight the value of corpus-informed pedagogy that consolidates accessible features while systematically targeting instruction-dependent processes.

Ask AI
Helpful
Bookmark
Share
View Full Paper

Cite This Study

Takeshi ISHIHARA (2026) studied this question.

synapsesocial.com/papers/6a1d22bb02fbce9130638594https://doi.org/10.20669/0002000446
Ask AI
Helpful
Bookmark
Share
View Full Paper