The Perception and Production of Lexical Stress among Early Spanish-English Bilingual Children

摘要

The present dissertation analyzed speech production and perception in early heritage Spanish-English bilingual children. Specicially, I investigated how perception and production of lexical stress develops in bilingual children. Current models of second language (L2) speech perception do not include heritage bilinguals or suprasegemental aspects of speech, like word stress. There is nothing inherent about these models that limits their predictions to late L2 learners and segemental aspects of speech, they just have not traditionally been extended to them. By analyzing the development of suprasegemental speech perception and production in heritage bilingual children, we can expand the scope of current models of speech perception. To the best of my knowledge, this is the first study to examine how perception and production of phonology develop in this population of speakers. The project attempts to answer the following questions:1. How do child bilinguals produce and perceive stress contrasts in both of their languages?2. How does the production and perception of lexical stress develop as a function of age and proficiency?3. Is there a discrepancy between perception and production abilities?4. How does perception and production of stress during childhood compare to that of adult bilinguals?5. Can lexical stress perception be improved using pedagogical interventions?6. Does improvement depend on stress pattern and language?In order to answer these questions, this dissertation is divided into three experiments. Experiment 1 examined the production of lexical stress. Spanish-English early heritage bilinguals ages 6-11 and a group of adult bilinguals completed two production tasks: a Delayed-Repetition task and an Elicited Production task. Data showed that duration is the primary cue that both children and adults use to denote lexical stress in English and Spanish. As a secondary cue, intensity has language- and stress- specific differences for adults and children. F0 was also revealed to be a secondary cue, but only for the adults. There were no major changes in production strategies based on age or proficiency, indicating that these two factors do not modulate the production of lexical stress in the ages tested.Experiment 2 explored the perception of lexical stress of English- and Spanish- like pseudowords via an AX discrimination task. The participants were identical to study 1. No effect of language was found, which indicates that participants are able to perceive stress contrasts in English as well they do in Spanish. Adults were shown to be more accurate and more sensitive to stress contrasts than children are. Additionally, age and proficiency were shown to be predictors of sensitivity to stress.Finally, Experiment 3 investigated the impact of a pedagogical intervention on the perception of lexical stress and the relationship between perception and production abilities. Participants and tasks were similar to Experiments 1 and 2, except the participants underwent a week-long teaching intervention via instructional games that aimed to improve sensitivity to lexical stress. Data showed that sensitivity to lexical stress can be improved after pedagogical interventions. Additionally, this experiment showed relatively similar findings in terms of production in that duration is the main cue that children use to produce stress contrasts.Overall, the results of the dissertation project inform models of speech learning and pedagogical models to teaching heritage speakers. This dissertation contributes to our understanding of phonological development in bilingual children while also exploring the importance of a teaching intervention to this development. Together, the three experiments in this dissertation contribute to our understanding of speech learning models in that they support the extension of these models to the perception of suprasegmentals. Furthermore, this dissertation aides in shifting the conversation around heritage spea