Durational variability of schwa in early and late Spanish–English bilinguals

Abstract

Aims: The purpose of this study was to evaluate whether bilinguals categorically displayed shorter or longer schwa durations between fixed word pairs where one word contains a deletable syllable and the other does not (as in “rational” [ræʃənəl] → [ræʃnəl] compared with “rationality” [ræʃənæləti]). We hypothesized that monolingual and early bilingual speakers would produce shorter values for “deletable” schwas whereas late bilinguals would not. Methodology: Twenty native English speakers and 40 Miami-based Spanish–English bilinguals (20 early and 20 late bilinguals) were asked to read sentences containing target words from 10 semantically related word pairs. Analysis: A three-way mixed model ANOVA was performed to determine the relationship between schwa duration and the following factors: language group, word root, and deletable/non-deletable category. Pairwise t-tests were also performed on individual categories to determine if semantically related pairs differed in duration. Conclusions: Results indicated that native English speakers produce significantly shorter durations for schwa in deletable versus non-deletable position. Early bilingual productions are very similar to those of English-only speakers, whereas late bilinguals display much longer durations in both categories. Originality: This study is the first to directly compare word-internal vowel duration patterns in Spanish–English bilingual speakers. Implications: These results indicate that age of L2 acquisition is a predictive factor in determining how native-like an L2 speaker’s pronunciation of reduced vowels will be. Limitations: Given the specific nature of the word pairs tested, a reading instrument was the most efficient manner of eliciting target word pairs. Future studies should try to evaluate this issue using other types of stimuli.