INFORMATION TO USERS This manuscript has been reproduced from the microfilm master. UM! films the text directly from the original or copy submitted. Thus, some thesis and dissertation copies are in typewriter face, while others may be from any type of computer printer. The quality of this reproduction is dependent UJWD the quality of the copy submitted. Broken or indistinct print, colored or poor quality illustrations and photographs, print bleedthrough, substandard margins, and improper alignment can adverselyaffect reproduction. In the unlikely event that the author did not send UMI a complete manuscript and there are missing pages, these will be noted. Also, if unauthorized copyright material had to be removed, a note will indicate the deletion. Oversize materials (e.g., maps, drawings, charts) are reproduced by sectioning the original, beginning at the upper left-band comer and continuing from left to right in equal sections with small overlaps. Each original is also photographed in one exposure and is included in reduced form at the back of the book. Photographs included in the original manuscript have been reproduced xerographically in this copy. Higher quality 6" x 9" black and white photographic prints are available for any photographs or illustrations appearing in this copy for an additional charge. Contact UMI directly to order. U·M·I University Microfilms tnternauonat A Bell & Howell tntorrnatron Company 300 North Zeeb Road. Ann Arbor. M148106-1346 USA 313/761-4700 800:521·0600 Order Number 9506222 The linguistic and psycholinguistic nature of kanji: Do kanji represent and trigger only meanings? Matsunaga, Sachiko, Ph.D. University of Hawaii, 1994 Copyright @1994 by Matsunaga, Sachiko. All rights reserved. V·M·I 300 N. ZeebRd. Ann Arbor,MI48106 THE LINGUISTIC AND PSYCHOLINGUISTIC NATURE OF KANn: DO KANn REPRESENT AND TRIGGER ONLY MEANINGS? A DISSERTATION SUBMITIED TO TIlE GRADUATE DMSION OF THE UNIVERSTIY OF HAWAll IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR TIlE DEGREE OF DOCTOR OF PHILOSOPHY IN EAST ASIAN LANGUAGES AND LITERATURES (JAPANESE) AUGUST 1994 By Sachiko Matsunaga Dissertation Committee: David Ashworth, Chairperson John DeFrancis Timothy J. Vance Martha E. Crosby Peter Dunn-Rankin iii © Copyright 1994 by Sachiko Matsunaga All Rights Reserved iv To the memory of my parents v ACKNOWLEDGMENTS I wish to express my sincere appreciation to my committee: Dr. David Ashworth, Dr. John Defrancis, Dr. Timothy Vance, Dr. Martha Crosby, and Dr. Peter Dunn-Rankin for their valuable comments and continued encouragement. Without their guidance and support, this dissertation would neither have begun nor been completed. I would also like to thank Dr. James Unger and Dr. Agnes Niyekawa for their useful advice in the early stage of the dissertation research. Their comments and suggestions helped me view the research topic in a larger scope, and thus understand it better. The challenges presented by their comments are greatly appreciated. My gratitude should be expressed to Dr. John Haig, who willingly agreed to serve as a proxy for one of the committee members on short notice. I would like to thank him for his kindness. My deep appreciation goes to my research assistant, Hisaaki Shimizu, who spent a memorable summer with me conducting experiments, which required a great deal of patience, waiting for each subject's arrival. Without his assistance, data collection would not have been possible. I am grateful for the assistance and cooperation of many people and organizations in making this research project possible. In particular: the Center for Japanese Studies and the National Foreign Language Resource Center for their financial support for this study; the Department of Educational Psychology for the use of the eye-tracker; Dr. Wesley Peterson for the use of his software; the offices of the English Language Institute, the New Intensive Course in English, and the Hawaii English Language Program at the University of Hawaii at Manoa, Rainbow Gakuen, Hawaii Loa College, and Hawaii Pacific University for the vi search for potential subjects. My heartfelt gratitude is of course expressed to all of those who participated in the study. It is a pleasure to express my appreciation to the two librarians at the Library of Congress, Yoko Akiba and Fumie Tateoka, who were especially helpful for searching and finding materials relevant to my research. During my short stay in Washington, D.C., I was able to collect the maximum amount of research materials because of their help. Finally, I must thank my husband, Geoffrey Middlebrook, who checked my English at every stage of preparation of this dissertation. My deepest love and appreciation go to him for his patience, understanding, and unceasing support. vii ABSTRACf There have been persistent beliefs among scholars and nonscholars alike that kanji, or Chinese characters, are picture-like symbols representing not sounds but ideas or meanings of words in Chinese and Japanese, and that readers of these languages obtain the meanings of kanji without relying on sounds. Are these beliefs valid? This is the research question asked in this dissertation, and an answer was pursued by examining the nature of kanji both linguistically and psycholinguistically. Linguistically, the literature review revealed that kanji do not represent only meanings, but they also represent sounds in a significant manner. Psycholinguistically, the literature review showed that there is sufficient evidence to suggest that kanji trigger not only meanings but also sounds, at least at word- and sentence-level processing. Since no study was found to provide a clear picture of the role of sounds in reading kanji at the discourse level, a study was conducted by the author investigating the occurrence of phonological coding, by asking native Japanese readers to read Japanese texts into which homophonic and nonhomophonic kanji errors were inserted. This study used an eye­ tracking methodology, measuring the number of fixations the readers placed on the kanji errors when they read the texts for comprehension. In the present study, following Horodeck (1987), two hypotheses were tested: the first hypothesis, to be rejected, was that fluent readers of Japanese fail to notice homophonic and nonhomophonic kanji errors regularly and equally; the second hypothesis, to be supported, was that when they do notice, they do so more frequently with nonhomophonic errors than with homophonic errors. By rejecting the first hypothesis and supporting the second hypothesis with eye-movement data, this study provided strong evidence for the occurrence of phonological coding in reading kanji at the discourse level. V111 The findings from this dissertation research were discussed in terms of four issues: the linguistic issue of the primacy ofspeech, the psycholinguistic issue of inner speech, the issue of applying the concept of primacy ofspeech in teaching Japanese and Chinese, and the issue of reforming written Japanese and Chinese. ix TABLE OF CONTENTS ACKNOWLEDGMENTS v ABSTRACT vii LIST OF TABLES xi LIST OF FIGURES xiii CHAPTER 1 INTRODUCTION: OVERVIEW AND RELATED ISSUES 1 A. Overview ,... ........ ....... ............... .......... .......... ...... 1 B. The Issues 3 1. Primacy of Speech 3 2. Inner Speech 5 3. Teaching Japanese and Chinese as Foreign Languages 12 4. Indispensability of Kanji in Orthographic Reform 14 C. Summary 18 CHAPTER 2 THE LINGUISTIC NATURE OF KANJI: DO KANJI REPRESENT ONLY MEANINGS? : 20 A. Introduction 20 B. Are Kanji Pictographs? 20 C. Are Kanji Ideographs? 22 D. Are Kanji Logographs? 32 E. Are Kanji Morphographs? 34 F. Summary 39 CHAPTER 3 THE PSYCHOLINGUISTIC NATURE OF KANJI: DO KANJI TRIGGER ONLY MEANINGS? 040 A. Introduction 40 B. Word-level Processing 40 1. Stroop Tests and Tachistoscopic Experiments 41 2. Naming Tasks and Lexical Decision Tasks 46 3. Memory Tasks 51 C. Sentence-level Processing 52 D. Discourse-level Processing 57 E. Summary 61 CHAPTER 4 THE STUDY: METHOD AND RESULTS 63 A. Introduction 63 B. Method 63 x 1. Subjects 64 2. Materials 65 3. Hypotheses 69 4. Equipment 70 5. Questionnaire 71 6. Error-marking Sheet. 71 7. Passage-rating Sheet.. 72 8. Procedure 72 C. Results 74 1. Bio-data, Level ofEducation in Japan, and Length of Residence in the United States of the Subjects 74 2. Percentages of Kanji Errors Noticed and Detected by the Subjects 77 3. Eye-movement Data Analyzed by Error Type 82 a. Design 82 b. Analysis 84 4. Eye-movement Data Analyzed by Passage Type 88 a. Design 89 b. Analysis 90 5. Effects of the Content and Linguistic Aspects of Passages 95 6. Effects of the Frequency of Kanji.. 98 7. Effects of the Degree of the Sound Differences between the Erroneous and Correct Kanji. 103 8. The Relationship between the Percentage of Errors Noticed and the Number of Fixations Placed on the Noticed Errors 105 D. Summary 109 CHAPTERS CONCLUSION: SUMMARY, IMPLICATIONS, AND SUGGESTIONS FOR FUTURE STUDIES 112 A. Summary 112 B. Implications , 114 1. Primacy of Speech 115 2. Inner Speech 116 3. Teaching Japanese and Chinese as Second Languages 119 4. Indispensability of Kanji in Orthographic Reform 122 C. Suggestions for Future Studies 126 APPENDICES 129 APPENDIX A: PASSAGES AND COMPREHENSION QUESTIONS " 129 APPENDIX B: ENGLISH EQUIVALENTS OF PASSAGES AND COMPREHENSION QUESTIONS 136 APPENDIX C: SAMPLES OF EYE-MOVEMENT DATA 143 APPENDIX D: QUESTIONNAIRE 149 APPENDIX E: ERROR-MARKING SHEET 150 APPENDIX F: RATING CATEGORIES FOR EACH PASSAGE 151 APPENDIX G: ENGLISH EQUIVALENTS OF RATING CATEGORIES FOR EACH PASSAGE 152 REFEREI-TCES 153 xi LIST OF TABLES Table Page 1 Summary of the Types of the Passages Read by the Two Groups 66 2 List of Errors 67 3 Bio-data, Level of Education in Japan, and Length of Residence in the United States of the Subjects 75 4 Correlation Matrix for Bio-data, Level of Education in Japan, Length of Residence in the United States, and Eye-movement Data of the Subjects 76 5 Extent to Which Errors Were Noticed by the Subjects 78 6 Percent of Subjects Who Noticed More Nonhomophones (NH) Than Homophones (H) Compared to Percent of Subjects Who Did the Opposite , 79 7 Comparison between Noticing Errors and Detecting Errors 81 8 Average Number of Fixations per Error.