The relationships among L2 fluency, intelligibility, comprehensibility, and accentedness

• Published in: Studies in second language acquisition Vol. 47; no. 1; pp. 282 - 307
• Main Authors: Chau, Tuc, Huensch, Amanda
• Format: Journal Article
• Language: English
• Published: New York, USA Cambridge University Press 01.03.2025
• ISSN: 0272-2631; 1470-1545
• DOI: 10.1017/S0272263125000014

Fluency, intelligibility, comprehensibility, and accentedness are important dimensions of second language (L2) pronunciation proficiency representing global, listener-based intuitions. This study meta-analyzed 49 reports from 1995 to 2023, examining 141 effect sizes (Pearson r) to understand their relationships and possible moderators. Three-level meta-analysis models showed weighted mean correlations of .82, .75, .62, .57, and .32 for fluency/comprehensibility, comprehensibility/accentedness, fluency/accentedness, intelligibility/comprehensibility, and intelligibility/accentedness, respectively. Task types moderated correlations for fluency/accentedness, intelligibility/comprehensibility, and intelligibility/accentedness, with controlled tasks leading to higher correlations. Ratings of multiple dimensions by the same listeners tended to result in weaker correlations for fluency/comprehensibility and comprehensibility/accentedness. The findings imply that having an accent does not mean being unintelligible and support prioritizing intelligible and comprehensible speech over accent reduction. The study also highlights an over-reliance on first language speaker norms in L2 pronunciation research and advocates for more transparent reporting.