THE DEVELOPMENT OF ENGLISH VOWELS BY NATIVE SPANISH SPEAKERS by Andrew Jeske B.A. University of Wisconsin-Madison, 2007 MA, University of Pittsburgh, 2012 Submitted to the Graduate Faculty of The Dietrich School of Arts and Sciences in partial fulfillment of the requirements for the degree of Doctor of Philosophy University of Pittsburgh 2016 UNIVERSITY OF PITTSBURGH DIETRICH SCHOOL OF ARTS AND SCIENCES This dissertation was presented by Andrew Jeske It was defended on March 31, 2016 and approved by Claude Mauk, PhD, Director of Less Commonly Taught Languages Center Matthew Kanwit, PhD, Assistant Professor Natasha Tokowicz, PhD, Associate Professor Dissertation Advisor: Alan Juffs, PhD, Professor ii Copyright © by Andrew Jeske 2016 iii THE DEVELOPMENT OF ENGLISH VOWELS BY NATIVE SPANISH SPEAKERS Andrew Jeske, PhD University of Pittsburgh, 2016 Previous studies have shown that exposure to a second language (L2) changes one’s perception and production of L2 sounds to become more native-like (e.g., Flege, 1995; Flege & MacKay, 2004). This change has been documented most commonly among immigrants after they move to an environment where the L2 is the main means of communication (e.g., Jia & Aaronson, 2003). However, many people get introduced to a foreign language in school and it has been found that the amount of L2 exposure provided to students in a foreign language classroom is not equivalent to the amount of exposure experienced by immigrants and, therefore, will not produce the same kinds of benefits (e.g., White & Genesee, 1996). This dissertation aimed to examine the effects different amounts of L2 exposure in a classroom environment can have on the perception and production of English front vowels (/i ɪ e ɛ æ/). The participants for this study were a group 2nd, 4th, and 6th graders from two schools (one bilingual, one non-bilingual) who shared the same L1 (Spanish) and age of first exposure to L2 English, but who differed in the amount of L2 exposure they received each week (5 hours vs. 14 hours). The participants’ perception was examined through a categorical discrimination task and their production of English front vowels was elicited via a picture-naming task. Predictions surrounding the relative discrimination difficulty of certain vowel pairs were made through the Perceptual Assimilation Model for Language Learners (PAM-L2) (Best & Tyler, 2007) and their production was evaluated through the Speech Learning Model (SLM) (Flege, 1995). Results from the perception task found a significant effect for school - the additional L2 exposure provided to the bilingual iv school students beneficially impacted their overall performance on the categorical discrimination task. However, this advantage was not clearly exhibited in their production because participants from each school were able to produce each English front vowel in a significantly distinct way. Further research will have to be conducted to see if the differences in production between the two schools affected the intelligibility of the target words. v TABLE OF CONTENTS PREFACE .............................................................................................................................................. XVI 1.0 LITERATURE REVIEW .................................................................................................................... 1 1.1 INTRODUCTION ............................................................................................................................. 1 1.2 AGE EFFECTS ................................................................................................................................ 2 1.3 QUALITY AND QUANTITY OF INPUT ............................................................................................. 4 1.4 L1 TRANSFER/INTERFERENCE ...................................................................................................... 8 1.5 THE RELATIONSHIP BETWEEN PERCEPTION AND PRODUCTION .................................................. 12 1.5.1 Perception before production ......................................................................................... 12 1.5.2 Production before perception ......................................................................................... 15 1.6 MOTIVATION .............................................................................................................................. 17 1.7 AMERICAN ENGLISH AND SPANISH VOWELS .............................................................................. 20 1.7.1 Perceptual and acoustic similarities of L1 Spanish-L2 English vowels ......................... 21 1.8 METHODOLOGICAL APPROACHES ............................................................................................... 23 1.9 PRESENT STUDY .......................................................................................................................... 24 2.0 EXPERIMENT I – PERCEPTION .................................................................................................. 25 2.1 METHODS I (CHILDREN) .............................................................................................................. 25 2.1.1 Participants ..................................................................................................................... 25 2.1.2 Stimuli ............................................................................................................................. 26 2.2 PROCEDURE (CHILDREN) ............................................................................................................ 28 vi 2.3 METHODS II (ADULTS) ............................................................................................................... 29 2.3.1 Introduction ..................................................................................................................... 29 2.3.2 Participants ..................................................................................................................... 30 2.3.3 Stimuli ............................................................................................................................. 31 2.4 PROCEDURE (ADULTS) ............................................................................................................... 32 2.5 RESEARCH QUESTIONS AND HYPOTHESES ................................................................................. 33 2.6 RESULTS ..................................................................................................................................... 37 2.7 DISCUSSION – PERCEPTION ......................................................................................................... 58 2.7.1 Overall perception scores ............................................................................................... 58 2.7.2 Individual contrasts – by school ..................................................................................... 61 2.7.2.1 /i/ and /ɪ/ ............................................................................................................................................ 61 2.7.2.2 /e/-/ɛ/-/æ/ ........................................................................................................................................... 63 2.7.3 Individual contrasts – by grade ....................................................................................... 64 2.7.4 Individual contrasts – school*grade interaction ............................................................ 65 2.7.4.1 /i/ and /ɪ/ ............................................................................................................................................ 65 2.7.4.2 /e/-/ɛ/-/æ/ ........................................................................................................................................... 67 3.0 EXPERIMENT 2 – PRODUCTION ................................................................................................. 69 3.1 METHODS .................................................................................................................................... 69 3.1.1 Participants ..................................................................................................................... 69 3.1.2 Stimuli ............................................................................................................................. 69 3.1.2.1 English .............................................................................................................................................. 69 3.1.2.2 Spanish ............................................................................................................................................. 70 3.2 PICTURE-NAMING TASK (CHILDREN) .......................................................................................... 70 3.3 RESEARCH QUESTIONS AND PREDICTIONS ................................................................................. 72 3.4 RESULTS ..................................................................................................................................... 76 3.5 DISCUSSION – PRODUCTION ........................................................................................................ 99 vii 3.5.1 Overall vowel production ................................................................................................ 99 3.5.2 Vowel production by school .......................................................................................... 101 3.5.3 Vowel production by grade ........................................................................................... 102 3.5.4 Vowel production by school*grade groups ................................................................... 103 3.5.5 L2 English versus L1 Spanish production ..................................................................... 103 4.0 GENERAL DISCUSSION AND CONCLUSION ......................................................................... 105 5.0 SUGGESTIONS FOR FUTURE RESEARCH ............................................................................. 111 APPENDIX A. STIMULI FOR PICTURE-NAMING TASK ............................................................ 112 APPENDIX B. STATISTICS FOR DISTANCE BETWEEN /æ/ AND /ɑ/ ....................................... 114 BIBLIOGRAPHY .................................................................................................................................... 115 viii LIST OF TABLES Table 1. Distribution of Native Spanish-Speaking Participants. The distribution of NS students who participated in the study by school type and grade level. ...................................................... 26   Table 2. English Stimulus List and Acoustic Measurements Across Speakers. Five English vowels were placed into a single CVC context (/b_t/) for the AX Categorical Discrimination task. ................................................................................................................................................ 27   Table 3. Demographics of Adult L1 Spanish Participants. Demographic information about the adult participants used in the perception study (Jeske, 2012). ...................................................... 30   Table 4. AX Categorical Discrimination Task Stimuli. All the stimuli used in the original perception and production tasks with the L1 Spanish speakers. The current study focuses on English /i/ /ɪ/ /e/ /ɛ/ and /æ/ ............................................................................................................ 32   Table 5. Overall Perception Accuracy on the Categorical Discrimination task. Results from a 2x3 ANOVA show a significant effect for school (p=.024) but no significant effect for grade (p=.857) or school*grade (p=.568). ............................................................................................... 38   Table 6. Discrimination Scores for Bilingual and Non-Bilingual Schools. Mean accuracy scores and percentages for participants at all levels from both the bilingual and non-bilingual school. . 38   Table 7. Perception Accuracy of Individual Contrast by School. (Within-Subject Effects) There is a main effect for contrast and school - participants from the schools performed significantly differently on at least one vowel pairing (*p<.05, ***p<.001) ..................................................... 41   Table 8. Accuracy Score (%) per Contrast by School. Comparison of mean scores by school. Significance determined through Duncan's Multiple Range Test (*p<.05, **p<.01). .................. 42   Table 9. Duncan's MRT - Mean Difference per Contrast by School. Results show no significant difference in performance for /i-i/, /ɪ-ɪ/ and /i-ɪ/. The bilingual school outperformed on /e-e/, /ɛ-ɛ/ and /æ-æ/ but the non-bilingual school outperformed on the /ɛ-æ/ vowel pairing (*p=.05, **p=.01) ......................................................................................................................................... 43   Table 10. Perception Accuracy of Individual Contrasts by Grade. (Within-Subject Effects) Upon comparing accuracy scores of each individual contrast, there is no main effect for grade (***p<.001). .................................................................................................................................. 46   ix Table 11. Perception Accuracy of Individual Contrast by School and Grade. (Within-Subject Effects) Upon comparing accuracy scores for each vowel pairing, there is a main effect for school and a significant interaction between school and grade (*p<.05, ***p<.001). .................. 46   Table 12. Duncan's MRT for /i-i/ (Mean difference by group). Display of school-grade mean comparisons for the /i-i/ vowel pairing (*p=.05, **p=.01) ........................................................... 47   Table 13. Duncan's MRT for /ɪ-ɪ/ (Mean difference by group). Display of school-grade mean comparisons for the /ɪ-ɪ/ vowel pairing (*p=.05, **p=.01). .......................................................... 48   Table 14. Duncan's MRT for /e-e/ (Mean difference by group). Display for school-grade mean comparisons for the /e-e/ vowel pairing (*p=.05, **p=.01). ......................................................... 49   Table 15. Duncan's MRT for /ɛ-ɛ/ (Mean difference by group). Display of school-grade mean comparisons for the /ɛ-ɛ/ vowel pairing (*p=.05, **p=.01). ......................................................... 50   Table 16. Duncan's MRT for /æ-æ/ (Mean difference by group). Display of school-grade mean comparisons for the /æ-æ/ vowel pairing (*p=.05, **p=.01). ....................................................... 51   Table 17. Duncan's MRT for /i-ɪ/ (Mean difference by group). Display of school-grade mean comparisons for the /i-ɪ/ vowel pairing (*p=.05, **p=.01). .......................................................... 52   Table 18. Duncan's MRT for /ɛ-æ/ (Mean difference by group). Display of school-grade mean comparisons for the /ɛ-æ/ vowel pairing (*p=.05, **p=.01) ......................................................... 53   Table 19. Nartural Referent Vowel (NRV) All Participants. (Paired samples t-test). Comparing accuracy scores within vowel pairings to test for order effects on perception of phones (**p<.001). .................................................................................................................................... 56   Table 20. Overall Average Scores of Perception Task. Displays the average scores on the AX Categorical Discrimination task by the children and adults. ......................................................... 56   Table 21. Comparison of Overall Perception Accuracy Scores by School and Adults. (One-Way ANOVA). This table displays the mean accuracy scores for the three groups of participants. There is a significant effect for group (***p<.001). ..................................................................... 57   Table 22. Comparison of Overall Perception Accuracy Scores (by School and Adults). (Post- Hoc). Shows a significant main effect for group - the adults performed significantly better than the other two groups (p<.001) and the bilingual school performed significantly better than the non-bilingual school (p=.046) ....................................................................................................... 58   Table 23. Production Task - Elicited Words. The participants were instructed to produce the following words in (1) a picture-naming task [English] and (2) a word-reading task [Spanish]. . 71   Table 24. Mean Barks Measurements of English Vowels by L1 Spanish Speakers. Mean barks measurements for all 124 native Spanish speaking participants. .................................................. 76   x