Acoustic adjustments of clear speech in Korean heritage vowel production

Aims and objectives: This study aims to explore the segmental and suprasegmental acoustic adjustments in the clear speech of Korean vowels produced by English-dominant Korean-heritage speakers and native Korean speakers. Methodology: Native Korean students at a university in the United States and Korean-American students enrolled in an intermediate Korean language course at the same university read isolated sentences featuring three representative Korean vowels (i.e., /i, a, u/), once in casual speech, and once in clear speech. They repeated this sequence five times. Data and analysis: The vowel space of the first and second formants (F1 and F2) in the clear speech was expanded for both heritage and native speakers, and the heritage vowel space was more fronted than the native vowel space, suggesting the influence of English among heritage speakers. However, the peripheral extent and direction in the clear speech and the dominant language effect in heritage speakers were inconsistent across the vowels. Also, pitch dynamics were highly significant in clear speech, exhibiting the salient fundamental frequency (F0) rising at the syllable offset for both heritage and native speakers. The vowel length enhancement in the clear speech was also confirmed in both native and heritage Korean vowels. Conclusions: The results confirmed the cross-language tendency in clear speech to induce the modification of vowel quality, pitch, and duration, which also applies to Korean heritage vowel production. Furthermore, the asymmetrical acoustic modifications across the vowels in heritage clear speech support the Speech Learning Model (SLM) prediction. Originality and implications: The findings present new data for the clear speech effects of Korean heritage vowel production and a further enhancement strategy of syllable offset pitch raising in clear speech for both native and heritage groups. The results contribute to the theoretical implication of bilingualism by connecting the SLM premise and heritage clear speech.