| Native-like pronunciation is necessary for membership into some social groups and to be considered a legitimate speaker of a language. Language immersion education aims to develop bilingual individuals, able to participate in multiple global communities, and while the lexical, syntactic, and sociolinguistic development of immersion learners is well documented, their phonological skills are not. This study set out to address this gap by investigating immersion learners' pronunciation of Spanish vowels, a sound class known to lead to a foreign accent, comparing the vowel productions of native English-speaking learners in one-way (foreign language) immersion and two-way (bilingual) immersion programs to those of their native Spanish-speaking peers and their teachers.;A total of 85 immersion students participated in this study. A cross-sectional sample of students from each of the program/language groups was taken; students from each of four grade levels (first, third, fifth, and seventh) participated. Students completed an animal picture sorting task in pairs during which their speech was audio and video recorded. Up to twenty tokens of each of the five Spanish vowels, for a possible total of 100 tokens per subject, were isolated and examined via spectrographic analysis in order to measure first and second formant values. The tokens examined for each vowel were balanced for their occurrence in stressed and unstressed syllables. Students also completed a written questionnaire in order to gather data about extralinguistic factors (i.e., attitudes and motivation) that have been shown to influence pronunciation.;The findings indicate that the vowel productions of immersion learners differ from those of native Spanish-speaking peers. In general, the vowel space of the learner groups is larger than that of the native speaker peer group. Over time, the number of differences between one-way NES learners and native speakers increase while the number of differences between two-way NES learners and native speakers decrease. This finding suggests that there may be an effect of program model; however, differences in the ethnic background and exposure to Spanish outside of school between the two learner groups may also play a role and thus make it difficult to attribute differences solely to the effect of program model. Differences in attitude between the groups do not reach statistical significance and do not correlate with more native-like vowel pronunciations. |
