The production of Arabic vowels by English L2 learners and heritage speakers of Arabic

摘要

It is known that adult language learners often struggle to accurately pronounce unfamiliar sounds in the target language, but the extent and duration of the linguistic experience is found to affect native-like production of target segments. In order to explore the variability in speech production between language learners, I compare vowel production between heritage speakers of Arabic (HSs) and English L2 learners. More importantly, this phonetic investigation asks fundamental research questions such as: Whether one or two phonetic/phonological systems coexist in the mental organization of developing bilinguals? If early childhood exposure to the target language as experienced by HSs affects phonetic learning later in life, e.g., when the HS is an adult learner in a traditional classroom setting? Moreover, do bilinguals of varying proficiency levels process their languages in the same way? Depending on the linguistic experience, prior studies of speech production show that bilingual speakers may possess one or two phonetic/phonological systems for the two languages. Furthermore, exposure to the L2 in early childhood facilitates attainment of native-like L2 phone production. Specifically, Flege (1987), Mack (1989), and Guion (2002) show that early bilinguals are capable of acquiring fine-grained phonetic detail of their L2 more than late learners. In the present study, 12 HSs— 6 experienced (EHSs) and 6 inexperienced (IHSs(—as well as 12 L2 learners— 6 advanced (AL2) and 6 beginner (BL2)—were compared with 6 native speakers of Arabic. Subjects produced 2 repetitions for each of 114 CVC monosyllabic words, embedded medially in a fixed carrier phrase. Formant measures of F1 and F2 were taken (in Bark) at vowel midpoint. Unlike the L2 learners, the results reveal that HSs have acquired two phonetic/phonological systems for Arabic and English, demonstrating the significance of childhood exposure to target sounds in later phonetic attainment. Specifically, more experience in the target language results in more accurate vowel production as shown by EHSs producing values that are closer to target vowels than IHSs and so is the case for AL2 compared to BL2 learners. Presenting a great challenge for language educators and language programs in the United States, implications from comparing these distinct populations (HSs and L2 learners) are discussed in relation to phonological theory, specifically, the intersection between second language research (SLA) and teaching of heritage language instruction.